Moral Development

bundet

becoming upset when the caregiver leaves, and if they are upset the stranger can console the child as often as the caregiver. Insecure anxious/resistant children appear anxious and stay close to the caregiver, are upset when the caregiver leaves, and are not consoled when the caregiver returns. These babies seem to desire the renewed contact while simultaneously push- ing away from it. Disorganized children appear con- fused and fearful and seem disoriented; they react with little emotion. It has been shown that caregivers of insecurely attached children are inconsistent in their response to their babies’ cries and have little physical contact with them. Children with disorganized attachment are often neglected or abused. Studies have shown that attach- ment breeds healthy social competence, whereas unhealthy attachment can lead to aggression; unfulﬁll- ing, insecure adult relationships; and more serious concerns, including psychopathology. Social Referencing Social referencing refers to infants’ use of their pri- mary caregiver to keep them safe while they explore their social world. As infants encounter an unknown object or person, they will look to their reference point (primary caregiver) to notice their reaction to the same stimulus. If the caregiver looks wary, the infants will be hesitant or avoid the stimulus; if the caregiver looks content or neutral, the infants will be less likely to avoid the object. This becomes a primary method of communication between the caregiver and the infant. Self-Concept Infancy marks the beginning of the individual’s understanding and recognition of himself or herself as a unique person. It is during this time that the individ- ual ﬁrst recognizes himself or herself in a mirror, something only humans and great apes accomplish, which occurs gradually over the ﬁrst year of life. End of Infancy At the end of infancy, children begin to have less distress when separated from their caregiver, and their sense of themselves becomes distinctive. They begin to have standards similar to adults and recognize a vio- lation of these standards, such as when they notice the difference between old, broken toys and new toys. It is at this transition that children begin to develop sec- ondary emotions, like embarrassment, guilt, and envy, which begins to show their realization of how they have an effect on others in their social world. Childhood Childhood continues to increase the individual’s level of self-understanding through engagement in friend- ship and peer groups as well as the changing relation- ship with his or her parents. Self-Concept As children enter school age, they begin to hold descriptions of themselves that include gender, friend- ship/group relations, as well as physical and psycho- logical traits. This is the age at which they begin to compare themselves with other children, which shows their understandings of what is desirable in their social environment as they envy the traits of others. Their views of themselves are important because research has shown many beneﬁcial outcomes for children who have positive self-concepts, including more conﬁdence, autonomy, assertion, sociability, and a greater outlook on life. This age group begins to recognize and expect the stability of these traits and behaviors in themselves and others. Self-Esteem Self-esteem is how positively or negatively a per- son feels about himself or herself, and it is an impor- tant attribute of stable mental health. Positive self- esteem is enhanced by parental relationships that are warm and accepting while having clear and under- stood limits. Self-Regulation Along with their self-conceptions and self-esteem, children develop a sense of self-regulation, which means they learn how to act in socially appropriate ways and how to stop themselves before acting in inappropriate ways. This is a difﬁcult task for the child to develop because they must learn how to stop behavior that has already begun. Play with other chil- dren has been shown to have a large effect on devel- opment of self-regulation because children learn how Psychosocial Development 821 to control their behavior and their emotions while watching and interacting with other children who may be more or less advanced than themselves. Friendship and Peer Groups Friendships and peer groups are an important aspect of psychosocial development because they pro- vide the major social context in which youths interact with each other. As youths begin to separate from their familial relationships these social groups begin to establish greater inﬂuence and pressure on a child’s life. These peer groups and friendships increase the development of important social skills. Playing games requires following rules and engaging in prescribed behaviors, both of which help children to understand larger and more complex rulings in their social envi- ronment. Children also begin to understand their place in the social world through their membership in cli- ques and crowds, which are the most common form of child and adolescent peer groups. Peer Pressure Although peer pressure is a common phenomenon within friendship and peer groups, it has been shown in research that youth are much more likely to go along with prosocial pressure than pressure to engage in antisocial behaviors. Peer pressure is believed to peak at the age of 15. Influence of Parents Two major changes occur as a child begins to associate more with peers and less with parents: (1) Parents’ expectations and demands on domestic chores increase along with their demands and expec- tations on academic achievement, and (2) parental control is achieved less through direct methods and more through discussion and reasoning, sometimes with the inducement of guilt. The relationship and monitoring of parents can have a profound effect on peer relationships and later development. Adolescence Research has identiﬁed four major changes in the structure of adolescent social life: 1. A great deal more time is spent in peer relationships. 2. Guidance from adults lessens in quantity and directness. 3. There are large increases in mixed-sex interactions. 4. The participation in large social groups becomes important. Beyond these changes in adolescents’ social life, their search for identity and intimacy also have effects on their psychosocial development. Identity Identity is the individuality of a person and is com- posed of relatively stable personality traits that are especially prevalent in certain situations (role beha- viors such as being a parent) and is thought of as the primary social concern for the adolescent. Although a development of identity occurs throughout the life course, most focus is placed on the adolescent years. According to Erik Erikson, a person must resolve his or her adolescent crisis before the person can confront the next stage crisis; thus, some sort of solution must be made with regard to identity, making it a rather important topic. Identity development is not a simple task; it is a process of understanding and synthesizing childhood experiences and memories, family history, present self-awareness and acceptance, and future goals or ideals. Beyond this internal process the per- son must also combine his or her relationships with persons and things external to the person. Intimacy Adolescents’ search for intimacy includes their creation of new friendships and dating relationships, and these relationships increase adolescents’ under- standing and development of autonomy. Gender Differences in Intimacy As in childhood there are gender differences in rela- tionship formation during adolescence. Females are more caring and emotionally open in relationships, whereas men are seen as being more independent, self- reliant, and closed off. Males are seen to be more asser- tive and domineering, while showing little emotion, and females are seen as more ﬂirtatious and emotional. These gender differences peak in later adolescence, and by the age of 50, they are much more diminished. Separation from Parents During adolescence, while youths are attempting to forge their own identities, they begin to separate 822 Psychosocial Development from their parents. This causes a decrease in the amount of inﬂuence a parent has on the youth; how- ever, research has shown that most adolescents main- tain their parents’ values, even though they may be more likely to verbally question them. Consistently throughout the research there is evidence that the most beneﬁcial developmental outcomes stem from parents who take an authoritative stance toward their adolescent, meaning they exhibit warmth and respect, while holding clear, ﬁrm rules. Adulthood Adulthood is not always seen as increasing the devel- opment of a person; in fact, the early theorists stopped discussing development at the termination of adoles- cence. There are, however, a number of changes occurring throughout adulthood that affect our psy- chosocial development. Early Adulthood Early adulthood is considered a person’s 20s and 30s and typically begins with a continued search for intimacy and progresses through the creation of a fam- ily unit. Love and Loneliness Erikson stated that this is a time in which the psy- chosocial crisis centers on intimacy versus isolation; another way to put this is love versus loneliness. One way in which young adults obtain this sense of love or intimacy is through their continued friendships, which tend to increase in the amount of emotional closeness, even in males but especially in females. Loneliness is said to be a large concern for college freshmen because it is usually associated with transi- tions in a person’s life. Types of Adult Status Adults comprise six status types, which are not linear in dimension and thus do not form a hierarchi- cal structure. Single adults are becoming an increas- ingly viable option for this age group, with many individuals remaining in this status throughout early adulthood. Many negative connotations of ‘‘single- hood’’ have been replaced with positive ones. Coha- biting adults is another status that has been on a recent rise. Research has shown that this type of relationship has either no effect or, if any, a negative effect on marital satisfaction and divorce. Married adults have been on the decline since the 1960s but still represent many individuals in early adulthood. Divorced adults have started to slow their rapid increase in numbers in the recent years. This is a status that can have large emotional effects on future rela- tionships and intimacy, although research suggests that, for most people, these effects are temporary. Remarried adults typically bring with them a combina- tion of families, creating stepfamilies that can create a large adjustment for all members involved; it has also been found that only one third of all remarriages survive. Lastly, gay and lesbian adults have had much change in recent years concerning their right to be married. Research has shown that these adults are very similar to heterosexual adults. There are cur- rently many misconceptions about gay and lesbian adults, including the belief that they are not good par- ents, when in fact they are similar in this regard, and the majority of children raised in these households grow up heterosexual. Stages of Family Life Cycle The family life cycle is seen as progressing through six linear stages: (1) leaving home and becoming a single adult, (2) joining of f amilies through marriage, (3) becoming parents, (4) families with adolescents, (5) the midlife family, (6) the later life family. Middle Adulthood Middle adulthood creates a new set of challenges concerning psychosocial development when adults must deal with the transition of their children leaving home and in many cases their aging parents moving in. Generativity Versus Stagnation Erikson stated that the 40s and 50s were concerned with the crisis of generativity versus stagnation which meant that the person was left with the choice of help- ing the next generation, which is typically associated with leading a useful life, and not assisting the next generation and remaining stagnant. Life Events Approach Life events approach took a separate approach to Erikson’s theory of psychosocial stages, saying that stagemodelsarenotasusefulinlaterlifebecause events sometimes happen out of order, or are skipped Psychosocial Development 823 in entirety, due to life events. This approach states that life events affect the individual not only due to the event itself, but also to how the person adapts to the event and the context of the event. Marriage and Divorce With regard to the life events approach, marriage and divorce can severely affect the development of a person in middle adulthood. It has been found that those who divorce in midlife are less likely than those in young adulthood to have a negative relationship with their divorced partner. It has also been found that divorce can have a positive effect on the female’s emotional well-being; the opposite was found with men, who have more positive emotional states during marriage. Parents and Children Midlife is typically the period when children leave the home, creating the ‘‘empty nest.’’ However, despite popular belief, this event typically increases the satisfaction of the parents, particularly dealing with marital satisfaction. This can also be a time in which the family must deal with bringing in the aging grandparents who may be in poor health, which cre- ates what is called the sandwich generation (having both one’s own children and one’s parents to be responsible for). Late Adulthood Late adulthood is typically a time of life review, in which a person looks back onto his or her life in either satisfaction or disappointment. Integrity Versus Despair Erikson’s ﬁnal stage of psychosocial development concerned the crisis of integrity versus despair, which was reworked by Robert Peck to include three develop- mental tasks faced by older a dults. (1) Differentiation versus role preoccupation concerns the redeﬁnition of the individual’s worth beyond his or her work roles now that the individual has retired. (2) Body tran- scendence versus body preoccupation concerns the fact that older adults must accept their declining physical well-being. (3) Finally, eg o transcendence versus ego preoccupation concerns the need for older adults to accept the fact that they will die and that they have made an impact on the future of others. Level of Activity and Social Support An important component of older adulthood con- cerns remaining active. Studies have shown that older adults who remain active are more likely to look onto their lives with satisfaction. Social support from fami- lies or peer organizations has also been linked with longer lives and increased physical and mental health. Gender Changes in Older Adulthood A ﬁnal change in the psychosocial development of an individual concerns gender. It has been found that gender differences all but disappear in older adult- hood, and in some cases actually reverse. Males become more nurturing and sensitive while females are more assertive. Life Course Psychosocial Developments Some important changes in psychosocial development occur throughout the life span and thus are discussed separately. Language Development One of the most important developments in a psy- chosocial world is the ability to vocalize our thoughts and emotions. Language development begins in infancy as the child begins to hear speech in his or her social world, and continues through childhood, all the way into adulthood as it becomes more complex and varied. At approximately 9 months, children utter their ﬁrst words, but their ability to use words to com- mand attention does not exist until around 12 months. At 18 months they are forming two-word sentences with a large increase in vocabulary, and ﬁnally at 24 to 30 months, they are beginning to take the listener’s perspective into account and responds to indirect requests. This development increases the individual’s place in the social world dramatically because the individual can now be a direct actor in his or her social world and achieve a numerous variety of goals. Moral Development The development of morals is a rather complex topic that begins in childhood and lasts into adulthood with few adults obtaining the highest levels of moral- ity. Lawrence Kohlberg stated that the early stages of morality are considered preconventional reasoning 824 Psychosocial Development that typically ties together morality and punishment, which stems from an external authority. A child in this stage believes in imminent justice in that if a rule is broken, a punishment will occur immediately regard- less of whether there is a witness. The later stages of postconventional reasoning consist of social contracts and individual rights, as well as universal ethical princi- ples of universal human rights. Gender Construction Gender is deﬁned as the social construction of being male or female. An individual typically has an estab- lished gender by the age of 3. Gender roles are socialized norms of how a person should act and feel dependent on his or her gender. Most children develop gender roles initially based on those of their parents; however, these roles develop and are rewarded or punished by peers throughout life. Defining an Identification Deﬁning one’s identiﬁcation may be confused with deﬁning an identity, although identiﬁcation does cre- ate strong roots for a person’s identity later in life. Identiﬁcation is the way a person attempts to look, behave, and feel, as well as who the person identiﬁes with in his or her social world. Individuals also iden- tify through the creation of gender roles which was discussed earlier. Problematic Psychosocial Development Not all individuals develop in the same way con- cerning psychosocial development. Some develop in a problematic way, typically because of some deﬁcit or excess in their life course. Possible Causes of Problematic Development Some links that have been found associated with problematic development include negative or absent attachment, child abuse, and negative or inconsistent parenting styles. Other important causes outside the parent–child relationship are too much television, which can expose the child to increased violence and diminish creativity and verbal skills, and large work schedules of adolescents, although up to 20 hours per week of part-time work during high school has been shown not to have a negative effect on development. Psychological Disorders Associated With Problematic Development In some cases the psychosocial development pro- cess can have large effects on an individual’s mental health and create psychological disorders, including substance use, antisocial behavior, depression, suicide, autism spectrum disorders, and many others. Matthew J. Davis See also Attachment Disorder; Autism Spectrum Disorders; Cliques; Erikson’s Theory of Psychosocial Development Further Readings Cole, M., Cole, S. R., & Lightfoot, C. (2005). The development of children (5th ed.). New York: Worth. Santrock, J. W. (2002). Life-span development. New York: McGraw-Hill. Steinberg, L. (2005). Adolescence (7th ed.). New York: McGraw-Hill. Psychosocial Development 825 Q Torture numbers and they’ll confess to anything. —Gregg Easterbrook Q UALITATIVE R ESEARCH M ETHODS Qualitative methods are used typically in research pro- jects that take a human-focused perspective in the design and implementation of the investigation. Quali- tative research , as a paradigm or worldview, is designed to explore the human elements of a topic under investi- gation; in this context, qualitative methods (such as in- depth interviewing or ethnographic observation) are used to examine how individuals see and experience the world around them. Researchers typically talk to people directly or observe their behaviors in various contexts to understand what those indi viduals view as important about a particular phenomenon. Although qualitative research is often described in opposition to quantitative research, and although the goals and intentions of these paradigms are quite differ ent, many scholars are now using mixed and multimethod approaches and engaging in interdisciplinary resear ch. In these projects, both qualitative and quantitative paradigms are used to more fully investigate a research problem. Indeed, many researchers now use qual itative methods to inform future quantitative research designs (and vice versa) or to add a richness of experience to an investigation by combining qualitative interviews or diaries with ﬁnd- ings from large-scale quest ionnaires or experimental interventions. The Qualitative Research Paradigm The use of qualitative methods has a rich history in the social sciences, health sciences, and humanities. In the ﬁeld of education, many research problems are best addressed by qualitative approaches, so the use of qualitative methods is very much the norm. In edu- cational psychology, case studies, interviews, and other techniques have been used recently to explore a range of research topics. Although Gary Shank noted in 1994 that qualitative research had ‘‘made very little headway’’ (p. 340) in educational psychol- ogy at the time, the rise in the number of qualitative projects over the past decade points to a positive shift toward including more qualitative approaches. The rise of qualitative methods in the ﬁeld of psychol- ogy itself, as noted in Sean Kidd’s review of core journals, also points to the value of using qualitative approaches to investigate problems in disciplines that have been focused traditionally on quantitatively ori- ented research practices. Research of high quality, regardless of paradigm, demands that appropriate methods are used to address the research problem at hand. To select appropriate methods, researchers must ﬁrst understand the types of data they will obtain with each approach. To assess student satisfaction with in-school counseling, for example, a researcher should start by asking, ‘‘What do I want to know?’’ and ‘‘What method will allow me to 827 address this question?’’ Unfortunately, new scholars (or those unfamiliar with the intent of qualitative meth- ods) often make the mistake of starting a project by saying, ‘‘I want to use an online questionnaire’’ or ‘‘I want to design an experiment,’’ even when that method may not be the best one to provide data for a particular research problem. Understanding the intended goals of qualitative research is essential to selecting appropriate methods and to assessing the results. Qualitative methods are best for addressing many of the ‘‘why’’ questions that education researchers have in mind when they develop their projects. Whereas quantitative approaches are appropriate for examining ‘‘who’’ has engaged in a behavior or ‘‘what’’ has happened (e.g., How many students approached school counselors for help last year? For what types of problems did they seek help?), and whereas experiments can test particular interventions, these techniques are not designed to explain why cer- tain behaviors occur. Quantitative approaches are best used to document characteristics of the world around us, that is, what we see and what the implications are when we test hypotheses related to speciﬁc phenom- ena. As Ted Palys and Chris Atchison note in their text, Research Decisions , quantitative researchers pre- fer the deductive method, whereby a researcher deduces a hypothesis from a theory, gathers data to test the hypothesis, and then revises (or discards) the theory or looks for another situation in which to test the theory again; here, the ‘‘true experiment’’ is the method of choice, as this allows the effects of certain variables to be assessed while all other inﬂuences are held constant. Qualitative approaches, on the other hand, are typically used to explore new phenomena and to capture individuals’ thoughts, feelings, or inter- pretations of meaning and process. Many qualitative texts (e.g., David Silverman’s book, Doing Qualita- tive Research: A Practical Handbook ), provide an overview of the preferences of qualitative researchers: to gather qualitative data that are naturally occurring, to explore meanings rather than behaviors, to reject the natural science model, and to craft studies that are inductive and hypothesis generating rather than ones that involve hypothesis testing. The Nature of Qualitative Data As qualitative inquiry allows a researcher to examine a topic in great depth, the data gathered tend to be very rich in scope. A researcher may compile dozens of pages of ﬁeld notes in completing an ethnographic observation, or capture many hours of video during focus group interactions. Data collection and analysis typically occur over many weeks or months, even in one setting or with only one small group of participants as the focus of the investigation. The depth of the investigation makes qualitative research both costly and time consuming. The sample sizes used in qualita- tive research are usually quite small (in some cases, only one participant or setting may be used; in others, 20 to 30 participants may be appropriate) and may involve relatively homogenous groups. However, as generalizability is not the intent of such projects, small samples that are carefully and deliberately chosen (e.g., using maximum variation sampling techniques) provide the best data possible for this type of work. Some researchers will also use multiple qualitative methods to explore one research problem (in a process known as triangulation ), which can extend the time and cost involved in completing a single project. Dis- semination of qualitative results can extend across many years, as the analysis of the rich data resulting from these types of projects can lead to multiple publi- cations covering a broad range of research questions. Also, many researchers engage in projects that build on the results of prior studies, which can extend the length and breadth of a single project over many years and across various contexts or populations. In delineating the nature of qualitative research, it is also important to note the distinction between the gath- ering of qualitative ‘‘data’’ and the use of a qualitative perspective or ‘‘paradigm’’ in research design and implementation. Open-ended questions used as part of a questionnaire, for example, allow respondents to pro- vide their opinions and attitudes about a particular topic or event. Gathering such data is often done along- side quantitative measures (e.g., Likert scales) to elicit information that cannot be captured by closed ques- tions. However, including open-ended questions does not make a research method inherently ‘‘qualitative.’’ Rather, qualitative research reﬂects a particular orien- tation to research that informs all phases of the research design—from the study’s conception, through data gathering, analysis, and writing. Many qualitative research projects are designed by scholars who see knowledge as socially constructed and inextricably linked to individuals’ backgrounds, personal histories, cultural place, and other contextual elements that deﬁne the human experience. Whereas quantitative researchers strive to eliminate these contextual 828 Qualitative Research Methods elements so that they will not ‘‘contaminate’’ the vari- ables under study (to maintain objectivity and reduce bias), qualitative researchers embrace these elements, to examine phenomena in light of these contextual characteristics. Qualitative projects purposefully embrace bias and subjectivity, which, in turn, often become a focus of the investigation. Qualitative Methods at a Glance Qualitative researchers, across disciplines, use a vast array of methods in their work. Indeed, in education alone, researchers use interviews, focus groups, partici- pant observation, diaries, think alouds, and a host of other qualitative techniques to examine students’, tea- chers’, and parents’ experiences of the school environ- ment. New technologies have also recently changed the research landscape; many qualitative researchers now use such tools as videoconferencing or digital video to capture data or use the Internet as a site of qualitative inquiry. Although the following sections provide brief descriptions of a few methods commonly used by education researchers, this is by no means an exhaustive list. Interviews In-depth, qualitative interviewing is one of the most commonly used methods. These interviews allow researchers to examine issues, at length, from the interviewee’s personal perspective. The data gath- ered typically consist of verbatim responses to the interviewer’s questions, which are designed to elicit opinions, feelings, attitudes, descriptions of personal behaviors, and other elements related to the research problem. The method may use a semistructured approach (e.g., open-ended questions are used to cap- ture participants’ thoughts of a variety of topics), or they may be quite unstructured (e.g., participants are encouraged to explore a topic, such as their personal life story, in their own way). Qualitative researchers (such as Irving Seidman) note that the purpose of this type of research is to understand individuals’ experi- ences and the meanings they make of those experi- ences and to put their behaviors in context in order to understand the actions they undertake. Semistructured interviews typically last from 60 to 90 minutes, although the length and the number of interview ses- sions will vary depending on the scope of the project, the availability of participants, other methods used in the project, and so forth. An unstructured interview may last over a number of hours or across many days or weeks, particularly in ethnographic studies. Com- mon themes and patterns are identiﬁed in the data gathered, with saturation of themes (resulting in trans- ferable data) occurring with 15 to 18 participants. Increasing the number of interviewees is one way to enhance rigor in data collection. However, anomalies in the data—where individuals discuss an experience unlike that of other interviewees—are also extremely valuable, as they can highlight very speciﬁc needs and points of conﬂict. Focus Groups Focus groups also fall into the qualitative inter- viewing category, but in this case, the interviews occur with groups of individuals (typically between ﬁve to eight people, with one or more groups in total). The discussions are generally focused on a particular issue, though many groups are designed to engage in open, exploratory dialogue. This method can also be used to assess the value of a product, service, or tool (e.g., in Web site usability studies) or to gauge indivi- duals’ attitudes, opinions, or beliefs about a social topic (e.g., the inﬂuence of the new media on political decisions). Group interviews can be more challenging to conduct than individual interviews because group dynamics need to be managed (e.g., to ensure that all group members have a chance to speak) and because special equipment (e.g., multiple video cameras) is often used. Focus group interviews are best run by a trained facilitator; they often require a more formal setting (such as a boardroom) and may take more time to coordinate and run. Observation Observing human behavior in natural settings (e.g., watching students as they interact in a classroom) allows a researcher to capture valuable data that cannot be gathered in any other way. Observational techniques may be covert or overt; researchers may be either distanced from, or directly engaged in (i.e., par- ticipant observation), the daily activities under investi- gation. Researchers employing observational methods not only document details about the individuals in the setting (e.g., a school), but they also examine the physi- cal (e.g., desk placement, computer lab setup) and organizational (e.g., school administration) structures Qualitative Research Methods 829 within the setting. Data collection may involve one or many sites and typically extends over a long period of time. Personal Journals and Diaries Asking individuals to document their activities (e.g., how, and for what purposes, students use the Internet to seek counseling support) is an effective way to examine behavior without relying on memory (as with interviews, where people are asked to recall past behaviors). Diaries allow individuals to document quantitative elements of their activities (e.g., how often they use the Internet) and qualitative opinions or attitudes about their experiences. Participants typically need instruction about how much to write, how often, and on what topics, but these journals can produce much more detailed accounts than is possible with other methods. Individuals may keep diaries for a week or more and may write on a variety of topics, which can then be further explored through other, complementary methods (e.g., personal interviews). Many researchers also now provide digital audio or video recorders to participants, particularly where writing a diary may prove difﬁcult, inappropriate, or time-consuming. Critiques of Qualitative Methods Although many texts and journal articles on research methods articulate the varied goals and intentions of the various research paradigms, the use of qualitative methods continues to be controversial in some disci- plines. Some scholars (particularly in biomedical sciences and in very quantitatively oriented disci- plines) dismiss qualitative research outright, arguing that these studies lack rigor in design and execution, are biased, or present only anecdotal ﬁndings. Silver- man notes, for example, that questions are often posed about a qualitative study’s reliability or validity when these quantitative measures of rigor are simply inappropriate for judging qualitative research. Other methodologists note that it falls to the researchers themselves to defend their work in light of charges from those who sit outside of the qualitative para- digm. Yvonna Lincoln and Egon Guba, in their clas- sic work Naturalistic Inquiry , noted that qualitative researchers learn quickly that they will face charges of engaging in ‘‘sloppy’’ research or ‘‘merely sub- jective’’ observations (p. 289). Unfortunately, the concerns raised in Lincoln and Guba’s text about unwarranted criticisms being lodged against qualita- tive researchers remain a concern in some disciplines, today, more than 20 years after the publication of their book. However, as qualitative methods have been employed for decades across different disciplines (e.g., sociol- ogy, anthropology, social work, information science, education, nursing), and as researchers have published extensively on qualitative methods development (e.g., in the International Journal of Qualitative Methods), one might reasonably argue that these types of cri- tiques are grounded more in ignorance or misunder- standing than in substance. Rather, the criteria used to assess the rigor of qualitative studies are quite different from those used in quantitative studies. In quantitative research, validity, reliability, generaliz- ability, and objectivity are the markers of quality work. Quantitative results are intended to be free from bias, to be replicable across contexts, and to generalize from the sample under study to the full target population (e.g., to all undergraduate students requiring counseling services). Qualitative research is assessed using the markers of credibility, transfer- ability, dependability, and conﬁrmability, including techniques such as prolonged engagement, persistent observation, triangulation of methods, negative case analysis, peer debrieﬁng, member checks, or a com- bination of these or other techniques to ensure a study is of high quality. Lincoln and Guba provide an excellent overview of the main tenets of these approaches to rigor in qualitative research. Qualitative Methods and Research Ethics It is also important to note that qualitative methods bring with them very speciﬁc issues and challenges that must be addressed in researchers’ applications for ethics approval. Many scholars (e.g., Will van den Hoonaard) have published texts that examine the research ethics associated with a variety of qualitative methods and point to some of the common issues that researchers must address. In ethnographic studies, for example, where a researcher conducts a project involving participant observation, conﬁdentiality and informed consent procedures may need particular attention (especially if the researcher’s role is not dis- closed to members of the community under study). With Internet studies involving private (subscription 830 Qualitative Research Methods only) chat sites, the process for obtaining and main- taining consent may be problematic, due to the ways that participants sign in and out at will. Using video cameras in public spaces may also pose a challenge to qualitative researchers, as individuals’ faces or voices may be identiﬁable; in these types of projects, a researcher may seek approval to disclose indivi- duals’ identities (with their informed consent) or may need to alter the digital record to preserve partici- pants’ anonymity. All of these ethics-related issues and choices are governed by ethics guidelines (which vary by jurisdiction), as well as by disciplinary norms and strategies developed over many decades. It is important that researchers who plan to use particular methods, techniques, and technologies refer to pub- lished texts or seek advice from established qualita- tive scholars in the implementation of appropriate ethics practices to suit those approaches. Recently, ethics boards have been criticized for requiring researchers to impose inappropriate guidelines or ethics-related practices in their qualitative studies (see, e.g., the Giving Voice to the Spectrum report published in 2004 by Canada’s Social Sciences and Humanities Research Ethics Special Working Com- mittee). It is important that researchers engaged in qualitative work educate themselves on the ethics implications of this type of work and demand that ethics review panels implement appropriate proce- dures and offer appropriate advice on the design of these types of projects. Conclusion Qualitative methods have an important, valuable place in research in educational psychology. However, to select appropriate methods , regardless of paradigm, researchers must focus on th e research problem at hand and choose methods that will provide evidence of high quality to address that problem. When projects are designed based on a preference for particular methods (e.g., when a researcher wants to create an online ques- tionnaire because of his or her interest in new technol- ogy) or based on a belief that a particular paradigm is better than another (e.g., that quantitative data are inher- ently more valuable than qua litative data), the overall worth of a study comes into question. As long as partic- ular methods (and paradigms) are privileged over others, with no regard for context or for the desired out- come for the research problem itself, the quantitative versus qualitative divide that continues to drive some educational research approaches will only gain more ground. Qualitative and quantitative approaches are complementary paradigms that examine the world from two very different vantage points, and each has its own important place in advancing knowledge in the ﬁeld of education. Indeed, projects that are designed to suit the intended goals of a paradigm or that embrace both sets of methods (if appropriate) go a long way in illuminat- ing understanding of the diverse range of human behaviors. Lisa M. Given See also Case Studies; Ethics and Research; Ethnography; Longitudinal Research; Naturalistic Observation Further Readings Kidd, S. A. (2002). The role of qualitative research in psychological journals. Psychological Methods , 7 , 126–138. Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry . Beverly Hills, CA: Sage. Mertens, D. M. (2005). Research and evaluation in education and psychology: Integrating diversity with quantitative, qualitative and mixed methods (2nd ed.). Thousand Oaks, CA: Sage. Palys, T., & Atchison, C. (2008). Research decisions: Quantitative and qualitative perspectives (4th ed.). Toronto, ON: Thomson Nelson. Patton, M. Q. (2001). Qualitative evaluation and research methods (3rd ed.). Thousand Oaks, CA: Sage. Seidman, I. (2006). Interviewing as qualitative research: A guide for researchers in education and the social sciences (3rd ed.). New York: Teachers College Press. Shank, G. (1994). Shaping qualitative research in educational psychology. Contemporary Educational Psychology , 19 , 340–359. Shank, G. D. (2006). Qualitative research: A personal skills approach (2nd ed.). Upper Saddle River, NJ: Prentice Hall. Silverman, D. (2000). Doing qualitative research: A practical handbook . London: Sage. Social Sciences and Humanities Research Ethics Special Working Committee. (2004). Giving voice to the spectrum: Report of the Social Sciences and Humanities Research Ethics Special Working Committee. Ottawa, ON: Interagency Advisory Panel and Secretariat on Research Ethics. Retrieved June 8, 2007, from http:// www.pre.ethics.gc.ca/english/workgroups/sshwc/ reporttopre.cfm Social Sciences and Humanities Research Ethics Special Working Committee. (2007). Qualitative research in the context of the Tri-Council policy statement—Ethical conduct for research involving humans: A discussion paper . Ottawa, ON: Interagency Advisory Panel and Qualitative Research Methods 831 Secretariat on Research Ethics. Retrieved June 8, 2007, from http://www.pre.ethics.gc.ca/english/workgroups/ sshwc/consultation07.cfm van den Hoonaard, W. (Ed.). (2002). Walking the tightrope: Ethical issues for qualitative researchers. Toronto, ON: University of Toronto Press. Q UANTITATIVE R ESEARCH M ETHODS Educational professionals are concerned with learning, achievement motivation, human growth and develop- ment, and human behavior. Consequently, measuring change in these areas is of great interest and need to educational researchers. T he purpose of quantitative research is to ﬁnd solutions to problems worthy of investigation. Results of quantitative research provide objectiveevidenceasabas is for decision making. These decisions may affect i ndividuals, organized pro- grams of study, or entire school systems. For example, new teaching methods may be implemented, new pro- grams established, curricula modiﬁed, new programs implemented, or new evaluation procedures used based on the results of quantitative research. Over time, results of quantitative research can inﬂuence an entire ﬁeld of study; such was the case with Piaget’s research in child development. Quantitative research methods allow researchers to measure certain behaviors or phenomena. Data are represented in the form of numbers. Quantitative research methods require the scientiﬁc method of investigation. Measurement is necessary if the scien- tiﬁc method is to be used. The scientiﬁc method involves an empirical or theoretical basis for the investigation of populations and samples. Hypotheses must be formulated, observable and measurable data must be gathered, and appropriate mathematical pro- cedures must be used for the statistical analyses required for hypothesis testing. If researchers are to ascertain relationships between variables and differ- ences among variables, data must be provided to support objective, valid, and reliable conclusions. The conclusions should be based on predetermined hypoth- eses and statistical testin g. Data may be collected in various ways, such as via the Internet, surveys, tests, inventories, or written documents. There are many quantitative methods that may be applied to research in the behavioral and social sciences. Both descriptive and inferential statistics are used in quantitative research. Quantitative methods depend on the design of the study (experimental, quasi- experimental, nonexperime ntal). Study design takes into account all those elements that surround the plan for the investigation, for e xample, research question or problem statement, research objectives, operational deﬁnitions, scope of inferences to be made, assump- tions and limitations of the study, independent and dependent variables, treatment and controls, instrumen- tation, systematic data colle ction actions, statistical analysis, time lines, and reporting procedures. Stages Problem Statement First, an empirical or theoretical basis for the research problem should be established. This basis may emanate from personal experiences or established theory relevant to the study. From this basis, the researcher may formulate a research question or prob- lem statement. Even though a study may have several objectives, a clear research problem statement or research question should be stated. Often these are teacher observations of classroom behaviors. When teachers formulate theories and engage in research, it is called action research. The research problem, whether anchored in theory or an outgrowth of personal obser- vation, provides direction for the methods to be used. Operational Definitions Operational deﬁnitions describe the meaning of speciﬁc terms used in a study. They specify the proce- dures or operations to be followed in producing or measuring complex constructs that hold different meanings for different people (e.g., beauty, intelli- gence, work values). For example, intelligence may be deﬁned for research purposes by scores on the Stanford–Binet Intelligence Scale. Population and Sample Quantitative methods include the target group (population) to which the researcher wishes to gener- alize and the group from which data are collected (sample). Early in the planning phase, the researcher should determine the scope of inference for results of the study. The scope of inference pertains to popula- tions of interest, procedures used to select the sam- ple(s), method for assigning subjects to groups, and 832 Quantitative Research Methods the type of statistical analysis to be employed. All of these elements in a study affect the extent to which conclusions can be inferred beyond the study unit(s). In experimental research, a study unit is that which receives the treatment, such as people, animals, or plants. When the whole population is used in a study, no inferences need to be made. Usually one or more samples from the population are used in a study. Hypothesis Testing Complex questions to compare responses of two or more groups or show relationships between two or more variables are best answered by hypothesis test- ing. A hypothesis is a statement about the researcher’s expectations about a relationship between variables to be studied. These statements may be written in the alternative or null form. A directional alternative hypothesis states the researcher’s predicted direction of change, difference between two or more sample means, or relationship among variables. An example of a directional alternative hypothesis is as follows: Third-grade students who use reading comprehen- sion strategies will score higher on the State Achievement Test than their counterparts who do not use reading comprehension strategies. A nondirectional alternative hypothesis states the researcher’s predictions without giving the direction of the difference. For example: There will be a difference in the scores on the State Achievement Test for third-grade students who use reading comprehension strategies and those who do not. In quantitative research, the null hypothesis is the hypothesis of interest. Stated in the null form, hypoth- eses can be tested for statistically signiﬁcant differ- ences between groups on the dependent variable(s) or statistically signiﬁcant relationships between and among variables. The null hypothesis uses the form of ‘‘no difference’’ or ‘‘no relationship.’’ An example of a null hypothesis is as follows: There will be no difference in the scores on the State Achievement Test for third-grade students who use reading comprehension strategies and those who do not. Hypothesis testing requires inferential statistical procedures. It is important that hypotheses to be tested be stated in the null form, because the interpretation of the results of inferential statistics is based on prob- ability. Testing the null hypothesis allows researchers to test whether differences in observed scores are real or due to chance or error; thus, the null hypothesis can be rejected or retained. Data Organization and Preparation The organization and preparation stage is important and involves several preliminary steps. Survey forms, inventories, tests, and other data collection instruments returned by participants should be checked for com- pleteness to assure that t here are no missing data points. If the data set is small, the researcher can visu- ally check the returned instruments. For larger data sets, a statistical program or spreadsheet is recom- mended. Missing data may be due to participants’ oversight, unwillingness to answer certain questions, or lack of knowledge about the question asked. The researcher should consider carefully the procedures to be used to handle missing d ata, and the procedures used should be stated in the ﬁnal report. John Tukey suggested that e xploratory data analysis be conducted using graphic al techniques such as plots and data summaries in order to take a preliminary look at the data before the actual analysis is conducted with more formal methods. Exploratory analysis provides insight into the data set and the underlying structure of the data. The existence of outliers, data entry errors, unexpected or interesting patterns in the data set, and assumptions of the planned analysis, such as homogene- ity of variance and normality of error for an analysis-of- variance statistical procedure, can be checked with exploratory procedures. Inferential statistics for tests of hypotheses should be conducte d only after all prelimi- nary analyses are completed. Data Analyses It is important that the researcher use the appropri- ate statistical tests for data analyses; otherwise, the results could lead to incorrect conclusions. Sometimes it may not be clear which test is the best for a particu- lar study. Some important considerations are 1. type of research questions to be answered or hypotheses to be tested, Quantitative Research Methods 833 2. number of independent and dependent variables in the study, 3. number of covariates in the study, 4. scale of the measurement instrument(s) used to col- lect the data (nominal, ordinal, interval, ratio), and 5. type of distribution (normal or non-normal). Examples of statistical procedures commonly used in educational research are t-test for independent sam- ples, analysis of variance, analysis of covariance, mul- tivariate analysis of variance, Pearson product moment correlation, Mann–Whitney U test, Kruskal–Wallis test, and Friedman’s chi-square test. Many statistics text- books, reference materials, and Web sites list decision tree diagrams or tables that provide guidance for select- ing the appropriate statistical procedure. Results Interpretation and Conclusions The level of statistical signiﬁcance that the researcher sets for a study is closely re lated to hypothesis testing. This is called the alpha level. It is the level of probabil- ity that indicates the maximum risk a researcher is will- ing to take that observed differences are due to chance. The alpha level may be set at .01, meaning that 1 out of 100 times the results will be due to chance; more commonly, the alpha level is set at .05, meaning that 5 out of 100 times the results will be due to chance. Using the .01 and .05 alpha levels means that if the null hypothesis is true, a low probability value will be observed. Alpha levels are often depicted on the nor- mal curve as the critical region, and the researcher must reject the null hypothesis if the data fall into the predetermined critical region. When this occurs, the researcher must conclude th at the ﬁndings are statisti- cally signiﬁcant. If the researcher rejects a true null hypothesis (there is, in fact, no difference between the means), a Type I error has occ urred. Essentially, the researcher is saying there i s a difference when there is none. Type I errors can have serious consequences in educational settings; for exa mple, if a program or entire school were to revise a curriculum based on the results of such a study, vast resources will have been wasted. On the other hand, if a researcher fails to reject a false null (there is in fact a difference), a Type II error has occurred. In this case, the researcher is saying there is no difference when in reality a difference exists. The power in hypothesis testing is the probability of cor- rectly rejecting a false null hypothesis. The cost of committingaTypeIorTypeIIerrorrestswiththe consequences of the decisions made as a result of the test. Tests of statistical signiﬁcance provide information on whether to reject or fail to reject the null hypothe- sis; however, an effect size (R 2 ,eta 2 , Phi, or Cohen’s D ) should be calculated to identify the strength of the conclusions about differences in means or relation- ships among variables. Experimental Designs True experimental designs are the most rigorous quantitative research methods in that they allow the researcher to have full control over the experiment and to assign subjects randomly to groups. Full con- trol and randomization strengthen internal validity; however, external validity may be compromised. Experimental designs are useful for establishing cause-and-effect relationships among variables. One or more variables are systematically manipulated so that effects of the manipulated variables on two or more groups or individuals can be observed. The vari- able being manipulated is called an independent vari- able (treatment) and the observed variable (measured outcome) is called a dependent variable . True experi- mental designs are difﬁcult to conduct in educational settings, because it is unrealistic to expect a classroom to mimic a laboratory. True experimental designs sat- isfy the following criteria: randomization, experimen- tal control(s), experimental treatment(s), experimental and control group(s), standardized test instruments, and basic patterns of research designs. Randomization Random assignment of subjects to groups helps to control for bias in the selection process and assures that groups are similar on all important variables that may affect the outcome of the study before the treatment is administered. Probability sampling procedures assure that each participant in a study has an independent and equal chance of being selected for any group. Experimental Control The researcher should attempt to hold all variables constant that might inﬂuence the outcome of the study by allowing only the dependent variable to vary based on participants’ responses to the treatment. 834 Quantitative Research Methods Researchers can control for time of day for a particular class, method of instruction, and same teacher in both classes. Achievement motivation, ability level, and satisfaction with school would be difﬁcult to control, yet these variables could inﬂuence the out- comes of a study. It is difﬁcult to impossible to have perfect control in educational studies because there are a host of extraneous variables that could affect a participant’s response on the dependent variable. Pretesting, matching, blocking, and using covariates are common controls. Experimental Treatment One or more interventions or treatments may be manipulated in an experiment. New teaching meth- ods, alternative testing methods, and different curric- ula are examples of treatments used in educational research. Experimental and Control Groups A researcher could compare reading achievement of two groups of students. The manipulated variable would be method of instruction (whole language or traditional), and the dependent variable would be reading achievement scores. The experimental group is taught reading by the whole language approach, and the control group is taught reading by the tradi- tional approach. After the experiment, both groups would be tested on the same standardized test. Any change in reading achievement scores may be tenta- tively attributed to method of instruction. Standardized Test Instruments A test that meets certain standards or criteria for technical adequacy in construction, administration, and use is said to be standardized. A standardized test provides clear directions for administration and scoring, so that repeated administrations and scoring of the test are carried out systematically. The Ameri- can Psychological Association, American Educa- tional Research Association, and National Council on Measurement in Education have established stan- dards that address professional issues and technical characteristics of standardized test development and use in the social sciences. Test validity, reliability, and established norms are major requirements of the standards. Basic Patterns Donald Campbell and Jul ian Stanley suggested three basic patterns of true experimental research designs: pretest-posttest control group design, posttest- only control group design, and Solomon four-group design. Pretest-Posttest Control Group Design A pretest is administered to a control group and an experimental group prior to the administration of the treatment. After the experiment, a posttest is adminis- tered to both groups, and gain scores from the pretest to the posttest may be compared. Statistically signiﬁ- cant differences between gain score means may be computed using a t -test for independent samples, if only two groups are involved. Posttest-Only Control Group Design When pretesting is not possible or desirable, the posttest-only control group design may be used. Ran- dom assignment of subjects to groups serves to assure equality of groups. Gain scores cannot be computed; otherwise, statistical analysis for the posttest-only design is the same as that for the pretest-posttest control group design. Solomon Four-Group Design As the name implies, the design requires four groups. The design is conﬁgured so that one experi- mental group receives the pretest and treatment: One control group receives the pretest but no treatment; one control group receives the treatment but no pre- test; and one control group receives neither pretest nor treatment. All groups receive the posttest. The Solomon four-group design is a combination of the pretest-posttest control group design and the posttest- only control group design. Essentially, the design requires the conduct of two experiments: one with pretests and one without pretests. The design provides more rigorous control over extraneous variables and greater generalizability of results than either of the previous designs. Because the design does not provide four complete measures for each of the groups, statis- tical analysis is performed on the posttest scores using a two-way (2 × 2) analysis of variance procedure. Thus, the researcher will be able to ascertain the main effects of the treatment, main effects of pretesting, and the interaction of pretesting with the treatment. Quantitative Research Methods 835 Quasi-Experimental Designs Random assignment of subjects to groups may not be possible because it may disrupt school schedules or it may limit the number of classes in the discipline to be studied. In such cases, quasi-experimental designs are useful. Much of the research in education and psychol- ogy is conducted in the ﬁeld or in classroom settings using intact groups. In these cases, researchers ran- domly assign treatments to the non-randomly selected subjects. The lack of full control and nonrandom assignment of subjects to groups pose threats to the internal and external validity of quasi-experimental designs. Matching may be used to control for the lack of randomization. In matching, researchers try to select groups that are as similar as possible on all important variables that may affect the outcomes of a study. Pretests are also recommended to control for lack of randomization. Similar scores on a pretest administered to all groups indicate that the groups were matched adequately. There is no doubt that quasi-experiments are weaker than true experiments for making causal inferences; however, information resulting from quasi-experiments is usually better than no information at all. Nonrandomized Control Group, Pretest-Posttest Design In an ideal experiment, subjects are assigned ran- domly to groups; however, random assignment is not always possible in school settings because of class s

bundet

cheduling conﬂicts. Rearranging students’ schedules disrupts their classes, and obtaining permission from administrators to collect data may be difﬁcult. Admin- istrators are more likely to agree to a researcher enter- ing two or more intact classes to collect data than to rearranging the class schedule of several groups of students. In other words, using intact groups for data collection is not only more convenient for the researcher, but it is also more palatable to the admin- istration. For these reasons, the nonrandomized con- trol group, pretest-posttest design is a popular choice for education researchers. A random procedure is used to determine which group or groups will be the con- trol and which will be the experimental. In these situa- tions, the researcher cannot assign subjects randomly to groups, so preexperimental sampling equivalence cannot be ensured. Similar scores on a pretest for the experimental and control groups denote a greater degree of preexperimental equivalency. After the pretest, the treatment is introduced, and both groups are adminis- tered the same posttest. Differences in pre- and posttest mean scores for the control and experimental groups are more credible when pretest means are similar. Statistical tests that compare mean scores are appropriate for the nonrandomized control group , pretest-posttest design when pretest scores indicate group equivalency. When results of the pretest reveal that the scores are not similar, more involved statistical procedures are required to address the measurement error on the pretest. Time Series Periodic measures are taken on one group at different intervals over an extended time period. Essentially, a time series design involves a series of pre- and posttest measures. A treatment is introduced, and its effects are assessed based on the stability of the repeated measures and differences from one mea- sure to another. Single-Subject Designs The researcher investigates a single behavior, or a limited number of behaviors, of one participant by making initial observations, administering a treatment, and observing and measuring behavior after the treat- ment to study the effects. Pretreatment observations are necessary to establish baseline data. A simple sin- gle-subject design is the AB design where pretreat- ment observations and measures (A) are made to establish baseline data, and then treatment (B) is administered, and observations and measures (A) are repeated to ascertain changes in the behavior. Moni- toring of the treatment usually lasts until it is apparent that the treatment has had some effect on behavior. The sequence of observations and measures (A) and treatment (B), followed by prolonged observations and measures (A), may be repeated as in the ABA and ABAB designs. Factorial Designs Many research questions in education require the investigation of more than one independent variable. Factorial designs permit the i nvestigation of the effect of one independent variable on the dependent variable while ignoring other independent variables (main effect) and the combined inﬂuence (interaction) between two 836 Quantitative Research Methods or more independent variables. Full control and ran- domization are difﬁcult at best in complex factorial designs. In fact, some fact orial designs may require nonrandom assignment of subjects to groups, such as whensubjectsareplacedingroupsbasedonsomepre- determined criterion, such as a bility level. Notation for factorial designs is usually written to identify the num- ber of levels involved in each independent variable. Forexample,a‘‘twobythree’’(2 × 3) factorial design would have two independent variables (factors) with two levels on the ﬁrst independent variable and three levels on the second independent variable. All groups receive each level on each of the independent vari- ables. A 2 ×3 design would require six groups. An example of a 2 ×3 factorial design is a study of method of teaching reading (whole language vs. tradi- tional) and motivation (high, medium, low). Three groups would receive whole language reading instruc- tion, and three groups would receive the traditional reading instruction. The de pendent variable would be posttest scores on a standardized reading test. Interpre- tations of the analysis are m ore complex for factorial designs than for designs discus sed previously. Analysis- of-variance statistical pr ocedures are used to analyze data for factorial designs. Nonexperimental Designs It is not always possible, practical, or ethical to manipulate the research variables of interest. A par- ent’s involvement in his or her child’s education, an individual’s level of job satisfaction, or whether or not a student has transferred from one school to another are variables that cannot be manipulated by the researcher. In these cases, the variable has already occurred. For such studies, nonexperimental methods are used. Three common nonexperimental methods of quantitative research are causal comparative (ex post facto), survey, and correlation. Causal Comparative Causal comparative studies do not permit the researcher to control for extraneous variables or for the randomization or manipulation of independent variables as in experimental studies. Changes in the independent variable have already occurred before the study is conducted; thus, these studies are also known as ex post facto , meaning ‘‘from after the fact.’’ Research of interest may be on differences in subjects on dependent variables when they have known differ- ences on independent variables or the extent to which subjects differ on independent variables when they have known differences on the dependent variables. These studies require that the researcher select partici- pants who already have or do not have the attribute variables being measured. Survey Survey research is a kind of descriptive study that is widely used in applied social sciences research. Sample surveys, unlike census surveys that query the entire population (as in the U.S. C ensus), collect data from a sample of individuals thought to be representative of the larger (target) population to which the researcher wants to generalize. Surveys are used to collect data over a period of time (longitudinal) or to collect data at one point in time. Surveys are designed to collect fac- tual data related to sociological and psychological constructs such as opinions, a ttitudes, values, beliefs, and desires. Data are gathered in different ways, for example, with paper-and-pencil questionnaires, one-on- one interviews, focus groups, telephone interviews, elec- tronic online questionnaires, and computerized kiosks. Responses are aggregated and summarized, and partici- pant identity must be conﬁdential or anonymous. Results of survey data are often reported by frequency and percentage for each response item. Other statistical pro- cedures for reporting survey data are cross-tabulations (cross-tabs), chi-square sta tistic, phi coefﬁcient, Kendall coefﬁcient, and th e gamma statistic. Correlation Correlation research is used to explore relationships between or among two or more variables. Correlation studies are useful for establis hing predictive validity, establishing test reliability, and describing relation- ships. Simple correlation p rocedures involve ascertain- ing the relationship between two variables, whereas partial correlation proced ures are used to control for a variable that may inﬂuence the correlation between two other variables. A multi ple correlation coefﬁcient (multiple regression) indicates the relationship between the best combination of independent variables and a single dependent variable. C anonical correlation indi- cates the relationship between a set of independent variables and a set of dependent variables. The kind of correlation coefﬁcient computed depends on the type Quantitative Research Methods 837 of measurement scale used to collect the data and the number of variables. Marie Kraska See also Descriptive Statistics; Inferential Statistics; Standardized Tests; Statistical Signiﬁcance; T Scores Further Readings Campbell, D. T., & Stanley, J. C. (1963). Experimental and quasi-experimental designs for research. Chicago: Rand McNally. McMillan, J., & Schumacher, S. (2006). Research in education: Evidence-based inquiry (6th ed.). Boston: Allyn & Bacon. Myers, R. H. (1990). Classical and modern regression with applications (2nd ed.). Paciﬁc Grove, CA: Duxbury. Ramsey, F. L., & Schafer, D. W. (2002). The statistical sleuth (2nd ed.). Paciﬁc Grove, CA: Duxbury. Salkind, N. J. (2003). Statistics for people who think they hate statistics . Thousand Oaks, CA: Sage. Tukey, J. W. (1960). Conclusions vs. decisions. Technometrics , 2 , 423–433. Tukey, J. W. (1977). Exploratory data analysis. Reading, MA: Addison-Wesley. 838 Quantitative Research Methods R He who lends a book is an idiot. He who returns the book is more of an idiot. —Arabic Proverb R ANDOM S AMPLE It is not always possible or feasible to collect data from every member of a population because of time constraints, lack of resources, or difﬁculty in identify- ing or locating all members of a speciﬁed group. In these cases, the researcher must select a sample from the larger group known as the target population to whom the researcher wants to generalize. Random samples allow the researcher to use inferential statis- tics to estimate the extent to which the sample differs from the population. The target population is the group to which the researcher wishes to generalize, and the sample is a ﬁnite subset of the population. For example, if the researcher wished to generalize about the effect of comprehension strategies on ﬁrst-grade students’ read- ing ability, ﬁrst-grade students would constitute the population. This group of ﬁrst-grade students may be considered a sample of present and future ﬁrst-grade students (sometimes referred to as a time/place sam- ple) in one school, a school district, or state. Usually one or more samples from a target popula- tion are used in a study. In these cases, researchers should follow systematic procedures for identifying the population, selecting the sample (subjects), and assigning subjects to treatment and control groups if ﬁndings are to be generalized beyond the study group back to the target population. A representative sample helps to ensure that the generalizations or inferences made from the sample to the population are free of systematic bias and present a fair view of the target population. The best way to achieve a representa- tive sample is through random sampling techniques. Different procedures may be used to select a random sample. Random sampling is often referred to as sim- ple random sampling to distinguish it from other kinds of sampling, such as stratiﬁed random sampling, cluster sampling, and systematic sampling. A simple random sample is the most basic kind of probability sample. Probability sample selection pro- cedures are those in which each unit or subject in the population has a known, nonzero chance of being selected. For simple random sampling, each case in a ﬁnite population has an equal chance of being selected, and the selection of any one case is indepen- dent of the selection of any other case. For a sample to be truly random, systematic procedures should be employed to select the subjects. The recommended method for selecting a sample using simple random procedures is to use a table of random numbers. Tables of random numbers are usually computer gen- erated, and they are often included in textbooks and on the Internet. To use a table of random numbers, the researcher ﬁrst assigns an identiﬁcation number to each subject in the population. Next, the researcher randomly selects a starting place on the table. This should be done by blindly placing a pointer on the 839 table to identify the row and column from which to start selecting numbers. Only those subjects whose identiﬁ- cation number matches the number in the table are included in the sample. For example, to select a sample of 100 students from a list of 600, the researcher might number the students from 000 to 599, refer to the table, and systematically move through the table selecting only those students whose identiﬁcation numbers fell between 000 and 599 until 100 students were selected. Stratiﬁed random sampling procedures may be used when there are subgroups within a target popula- tion that the researcher wants to be sure to sample. For example, intelligence level, socioeconomic status, political afﬁliation, and number of children per house- hold may form subgroups within a population. The strata or subgroups are ﬁrst identiﬁed and then proce- dures are employed to select subjects randomly from each of the strata. The number of subjects selected for each stratum may be based on their proportional rep- resentation in the population, or equal numbers of subjects may be selected from each stratum. Cluster sampling involves the use of intact or natu- rally occurring groups within a population. Indivi- duals within the clusters should be similar in terms of the variables of interest to the researcher. Clusters should be selected randomly from a population of clusters, and all members of the clusters should be included in the sample. For example, in educational research, all students within certain grade levels, sub- jects, or classrooms are often used in a sample. Large samples such as those used by national education databases often use cluster sampling. However, in these cases, individuals within clusters are selected at random. For example, a representative sample of all students enrolled in the public schools in the United States may be drawn by random selection from each of the following clusters: public school districts in the nation, schools within the districts, classrooms from the selected schools, and students within classrooms. Systematic sampling requires random selection of the ﬁrst subject and then the selection of every nth subject from the target population list. The size of the interval to use to select subjects from the list can be calculated by dividing the population by the desired sample size. For example, the interval fo r a sample size of 500 from a population of 5,000 could be obtained by dividing the population size by the desired sample size (5,000/ 500 =10). Every 10th subject would be selected for the sample. The starting point on the list should be identiﬁed randomly by rolling a pair of dice, using a table of random numbers, or using computer-generated numbers. If your random number were eight, then the eighth, eigh- teenth, twenty-eighth, and so on, subjects would be included in the sample. One should use caution when using systematic sampling to be sure that the population list is current and complete. Also, care should be taken that the population list is not formed by some distin- guishing characteristic such as geographic location of subjects, which co uld introduce bias. Probability samples are important for quantitative research methods because they permit statistical pro- cedures whereby inferences can be made from a sam- ple back to the target population. Regardless of the procedures used to select a sample, the most impor- tant characteristic of a sample is whether or not it is rep- resentative of the population to which the researcher wishes to generalize. Marie Kraska See also Quantitative Research Methods; Statistical Signiﬁcance Further Readings Johnson, B., & Christensen, L. (2004). Educational research: Quantitative, qualitative, and mixed approaches (2nd ed.). Boston: Pearson Education. Thompson, B. (2006). Foundations of behavioral statistics: An insight-based approach. New York: Guilford. Tuckman, B. W. (1999). Conducting educational research (5th ed.). Belmont, CA: Wadsworth/Thomson. R EADING C OMPREHENSION S TRATEGIES A distinction is often made between skills and strate- gies. Skills are often conceptualized as unitary and automatic. Strategies are conceptualized as complex and effortful. Lysynchuk et al. deﬁned comprehension strategies as ‘‘steps or actions that readers can take to enhance comprehension’’ (p. 460). Other available deﬁnitions are similar, but often have subtle distinc- tions. The National Institute for Literacy deﬁnes com- prehension strategies as ‘‘conscious plans or sets of steps that good readers use to make sense of text.’’ The National Reading Panel of the National Institute for Child Health and Human Development deﬁned com- prehension strategies as ‘‘speciﬁc procedures that guide 840 Reading Comprehension Strategies students to become aware of how well they are compre- hending as they attempt to re ad and write’’ (Sec. 4, p. 5). What is important about these deﬁnitions is that they illustrate the diversity in thinking about strate- gies. However, the deﬁnitions all converge on the notion that strategies are cognitive, involving steps or actions, perhaps involving a number of skills. They also suggest that strategies are directed at enhancing, or making possible, the understanding of text. The National Reading Panel reviewed the experi- mental research literature on comprehension strategy instruction and found a total of 203 research studies involving 16 different types of instruction of compre- hension strategies. Of this total, 7 types of instruction of comprehension strategies were found to have sig- niﬁcant scientiﬁc support in the literature reviewed by the panel. The 7 types are as follows: 1. Comprehension monitoring is a strategy in which readers learn how to be aware or conscious of under- standing during reading. Readers also learn proce- dures to correct problems in comprehension as they arise. 2. Cooperative learning is a strategy in which readers work together to learn from reading. Students may work together on an entire problem or individually on components, sharing the results to complete the understanding. 3. Graphic and semantic organizers are strategies that allow readers to represent graphically (write or draw) the meanings and relationships of the ideas that underlie the words in the text. 4. Story structure is a strategy in which the reader learns to ask speciﬁc questions about the elements of stories, including plot, time line, characters, and events in the stories. 5. Question answering is the most conventional strat- egy in which the reader answers questions posed by the teacher (or the textbook) and is given feedback on the correctness of the answer. 6. Question generation is a strategy involving the active production of questions in which the reader asks himself or herself what, when, where, why, what will happen, how, and who questions. 7. Summarization involves the reader identifying and writing the main or most important ideas that inte- grate or unite the other ideas or meanings of the text into a coherent whole. There were six additional categories of strategy instruction that did not have scientiﬁc support. These categories had too few studies or studies in which the range of student abilities or ages was too small to make judgments about scientiﬁc merit. This is an absence of knowledge rather than a judgment that these are ineffective. The categories were Curriculum (integrating strategy instruction into the normal cur- riculum), Listening Actively (oral language practice in comprehension in memory), Mental Imagery (pic- turing concepts mentally), Mnemonic (using memory devices), Psycholinguistic (instruction in language elements), and Vocabulary-Comprehension Relation- ship (emphasizing word knowledge). Instructional Variables Besides the content of instruction, there are several issues in how the instruction is presented or formatted. One important variable in th e instruction of compre- hension strategies is whether strategies are taught sin- gly or in combinations. Reciprocal teaching is one of the more successful comprehen sion strategy instruction models. It incorporates four strategies: predicting, clari- fying, questioning, and summarizing. Research has shown that the questioning and summarizing are domi- nant contributors to the success of the format. Another variable in instruction of strategies is the relative amounts of student and teacher involvement. Two general models have developed. In the Direct Explanation (DE) approach, teachers focus on helping students to conceptualize reading as a problem-solving task that requires strategic thinking. It also emphasizes the need to learn to think strategi- cally about solving reading comprehension problems. A key element of this is the explicit explanation of the mental processes involved. The Transactional Strategy Instruction (TSI) approach includes some of the same elements that DE includes. However, it also focuses on the ability of teachers to facilitate discussions in w hich students collaborate to form interpretations of text and explicitly discuss the mental processes and cognitive strategies that are involved in comprehension. The emphasis is on the inter- active exchange among learners in the classroom. The preparation of teachers to instruct students in comprehension strategies has also been studied extensively. These ﬁndings, also summarized by the National Reading Panel, show that teacher preparation to teach reading comprehension strategies is necessary Reading Comprehension Strategies 841 and can be accomplished, although it requires inten- sive and sustained effort. Limitations of Current Knowledge Instruction of comprehension strategies has been suc- cessful with students in the fourth grade and above. Very little is known about whether certain strategies are more appropriate for certain ages and abilities. Very little is known about the reader characteristics that inﬂuence successful instruction and which strate- gies, or combinations, are best for younger readers, poor or below-average readers, learning-disabled and dyslexic readers, or second-language learners. It will be important to know whether successful instruction generalizes across different text genres (e.g., narrative and expository) and texts from differ- ent subject content areas. The National Reading Panel review of the research indicated that little or no atten- tion was given to the kinds of text that were used and that there was little available information on the difﬁ- culty level of texts. Michael L. Kamil See also Instructional Objectives Further Readings Block, C. C., & Pressley, M. (Eds.). (2002). Comprehension instruction. New York: Guilford. Lysynchuk, L. M., Pressley, M., D’Ailly, H., Smith, M., & Cake, H. (1989). A methodological analysis of experimental evaluations of comprehension strategy instruction. Reading Research Quarterly , 24 , 458–470. National Institute of Child Health and Human Development. (2000). Report of the National Reading Panel: Teaching children to read: An evidence-based assessment of the scientiﬁc research literature on reading and its implications for reading instruction: Reports of the subgroups. Washington, DC: Author. Web Sites National Institute for Literacy: http://www.niﬂ.gov R ECIPROCAL D ETERMINISM Reciprocal determinism is a phrase coined originally by psychologist Albert Bandura in describing the workings of his social learning theory. It describes the reciprocal relationship between person variables (genetics, individual differences); overt behavior; and the social environment. Simply stated, it posits that a person’s behavior is both inﬂuenced by and is inﬂu- encing a person’s personal factors and the environ- ment. According to this view, a person’s behavior can be conditioned by the environment through operant conditioning (e.g., the use of consequences like reward and punishment), but a person’s behavior can also have an impact on the environment. Thus, it is not the case that individuals are merely shaped and inﬂuenced by their environment; individuals also inﬂuence the environment around them—each affects the other. Theoretical Significance Reciprocal determinism represented a signiﬁcant departure from the long-standing debate in psychol- ogy over whether the person or the situation is most responsible for behavior. This new perspective argues that a person’s behavior is based on his or her evaluation of the situation, and it added person variables and cognition to the factors that must be understood in order to have a full grasp of psycho- logical phenomena. As such, this perspective inﬂu- enced the way in which behavior is conceived, particularly within the realm of social psychology and social learning. It emphasizes the importance of person factors, such as cognition, in understanding the impact of the external environment on indivi- duals as well as the ways in which personal factors affect the environment. Speciﬁcally, this view discussed ﬁve different cog- nitive social learning person variables that affect an individual’s evaluating and interpretation processes: 1. Competency and self-efﬁcacy. Competencies include intellectual abilities, social and physical skills, and other abilities. Self-efﬁcacy refers to the conﬁ- dence an individual has in his or her ability to take action and persist in action. These variables address the question, ‘‘What can you do?’’ 2. Encoding strategies and personal constructs. People differ in the way they selectively attend to information, encode, or mentally represent events, and group information into meaningful categories. The same event may be perceived by one person as threatening, but by another person as challenging. 842 Reciprocal Determinism These variables address the question, ‘‘How do you see the situation?’’ 3. Expectancies. Expectations about the conse- quences of different behaviors guide an individual’s choice of behavior. Before people engage in an action, they consider their beliefs about the likely results of that action. For instance, if you cheat on an exam and get caught, what do you expect the conse- quences to be? These variables address the question, ‘‘What will happen?’’ 4. Subjective values. Individuals who have simi- lar expectancies may choose to behave differently because they assign different values to the outcomes. Two people may expect that donating money to char- ity will impress their peers; however, this outcome may be more important to one person than another, thereby affecting their choice of behavior. These vari- ables address the question, ‘‘What is it worth?’’ 5. Self-regulatory systems and plans. People differ in the standards and rules they adopt for regulating their behavior (including self-imposed rewards for success or punishments for failure), as well as their ability to make realistic plans for reaching a goal. These variables address the question, ‘‘How can you achieve it?’’ These person variables mediate people’s interac- tion with the outside world. People’s perceptions of their environment determine their actions, and the ways their social environment responds to them pro- vides feedback that in turn affects these person vari- ables. The reciprocal dynamics of this perspective recognize the degree to which various circumstances affect individuals, the way in which individuals inter- pret these circumstances, and the actions of the indivi- duals themselves and others around them that serve to further change and deﬁne the situation. Predicting and Understanding Behavior In understanding behavior, reciprocal determinism emphasizes that an event and a response rarely have only a single cause. Furthermore, any events and responses that occur are likely to have an impact on future events as well. This p erspective generates a com- plex view of causality, in w hich previous occurrences affect present occurrences, which in turn inﬂuence future events. This complex causality can best be dem- onstrated by an example. Suppose a child named Billy approaches some other children on the playground and asks to join their game. The other children refuse to let Billy join because they are in the middle of a game. Because of Billy’s own person variables, he interpre ts their response to be hos- tile toward him and retaliates by behaving aggressively toward his peers. In response, the other kids dislike Billy and start to avoid him and talk about him behind his back. Their response makes Billy even more para- noid about his peers’ hostility toward him and rein- forces his desire to aggress against them. Previously innocuous situations now become problematic because Billy no longer trusts his peers, and they in turn want to avoid him and exclude him from their games, lead- ing to further acts of aggression. His persistent aggres- sion toward his peers is observed by the teachers, who form a negative view of Billy and further reinforce his interpretation of the social environment. This analysis illustrates the central tenets of recip- rocal determinism—although the environment shapes, maintains, and constrains human behavior, people are not passive recipients in this process, for they actively choose, create, and change their environments. Thus, reciprocal determinism readily acknowledges that genes and other person factors affect behavior, and that environment strongly inﬂuences behavior as well. However, reciprocal determinism advocates the importance that subjective evaluation and interpreta- tion play in determining behavior and in shaping and changing the environments people encounter. In uti- lizing reciprocal determinism as a tool for research and understanding, an emphasis is placed on dynamic changes over time that occur to all three factors (per- son variables, behavior, and environment) because of the complex interplay between these factors. Edward Raymond Hirt and Timothy Scott Reilly See also Applied Behavior Analysis; Behavior Modiﬁcation; Social Learning Theory Further Readings Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Upper Saddle River, NJ: Prentice Hall. Mischel, W. (1973). Toward a cognitive social learning reconceptualization of personality. Psychological Review , 80 , 252–283. Reciprocal Determinism 843 R EGRESSION Research in educational psychology is often interested in examining the relationship between one or more variables and using this relationship to predict future behavior. Regression is the process of modeling the relationship between variables. In statistical form, regression models can be used to predict the outcome of a particular variable given knowledge of the value of one or more other variables. Regression can be used to answer questions related to performance on standardized examinations in subject areas such as mathematics, reading, and sci- ence. For instance, student performance on a mathe- matics assessment is likely related to performance in a mathematics-related class such as algebra or geom- etry. The relationship between these variables can be used to devise a regression equation that allows a pre- diction to be made. That is, given knowledge of per- formance in an algebra class, the regression equation can be used to predict (with error) performance on the mathematics assessment. The regression equation also explains a certain amount of variability in the outcome measure. Regression models take many forms, but linear models are most common. Linear regression models refer to a pattern of change where the outcome measure or dependent variable (DV), Y , increases (decreases) as the independent variable (IV), X , increases (decreases). Theoretically, the linear regres- sion equation can be expressed as Y = a + bX + e , where a and b are estimated values for the intercept and slope, respectively, and e represents measurement error or the amount of unexplained variability in the DV. The intercept is deﬁned as the value of the DV when the IV is equal to zero, and the slope provides information on the direction and magnitude of the relationship between the IV and DV. The goal in formulating regression models is to minimize the amount of meas urement error by explain- ing as much variability in th e DV as possible. This can be accomplished through either a simple regression model or a multiple regression model. The aforemen- tioned example is considered a simple regression model using a single IV, performance in an algebra class, to predict performance on the mathematics assessment. Including an additional variable, such as performance in a geometry class, would likely result in a better explanation of the variability in mathematics assessment scores. Theoretically, the multiple regres- sion equation takes the same form as the simple regres- sion equation but contains an additional slope value: Y = a + b 1 X 1 + b 2 X 2 + e : The slope in a simple linear regression equation is calculated as b = s XY s 2 X , where s XY is the covariance between X and Y and s 2 X is the total amount of variability in X.The y intercept is a = Y − bX , where Y is the mean of Y and X is the mean of X . The slope is reported in either standardized or unstan- dardized form. Unstandardized values of b explain the magnitude in the relationship between the indepen- dent variable and dependent variable in the original units of the variables. However, if the data are nor- malized, standardized values for the slope can be cal- culated. Using standardized values for the slope results in interpreting the relationship between X and Y in standard deviation units. Standardized values are useful for determining which IV is contributing more toward explaining the variance in the DV. The IVs can also be compared in the amount of variability they are explaining via the coefﬁcient of determination, also referred to as the squared corre- lation coefﬁcient ( r 2 ). The r 2 is a numerical value representing the amount of variability in the DV explained by the IV and is a measure of the good- ness of ﬁt of the regression model. For example, if data from two variables are graphed in a scatterplot, the coefﬁcient of determination will be highest for the regression line that minimizes the distance between the observed data points and the regression line. The coefﬁcient of determination is calculated via the following formula: r 2 = SS regression SS total = P N i = 1 ð ^ Y i − Y Þ 2 P N i = 1 ð Y i − Y Þ 2 , 844 Regression where ^ Y is the predicted score of y from the regression equation. The SS regression (or sums of squares) repre- sents the total amount of v ariability accounted for by the regression equation, whereas the SS total represents the total amount of variabi lity to be explained in the DV. Thus, the more variance explained by the regres- sion equation, the larger the value of r 2 .Ifavalueof 1.0 was obtained, the observed data points would fall directly on the regression line, indicating a perfect ﬁt between the model and the data. The coefﬁcient of determination is interpret ed as the proportion of vari- ability in the DV accounted for by the IV. For exam- ple, suppose regressing performance in a mathematics class on performance on a standardized mathematics assessment yielded r 2 = : 25. This indicates that approximately 25 % of the variance in mathematics assessment scores is explained by performance in a mathematics class. To determine the signiﬁcance of the IV as a predic- tor of the DV, a test statistic is calculated. The test statistic uses information on the amount of variability accounted for by the model and the amount of vari- ability left unaccounted for. The test statistic becomes larger as the model better explains the outcome mea- sure. As more independent variables that contribute to explaining the variance in the independent variable are added to the model, the r 2 value will increase and the test statistic will also increase. In summary, regression is tantamount to one of the principle purposes of behavioral research: to analyze and predict behavior. Linear regression can also be used for more complex research methods. For exam- ple, regression can be used for analyzing nonlinear relationships, analyzing the effects of covariates on a relationship, and determining the relationships between categorical independent variables (dummy coding). Regression also plays an important role in more advanced statistical techniques, such as struc- tural equation modeling. Greg W. Welch and Chris S. Meiers See also Descriptive Statistics; Inferential Statistics; Statistical Signiﬁcance Further Readings Pedhazur, E. J. (1997). Multiple regression in behavioral research: Explanation and prediction (3rd ed.). Fort Worth, TX: Harcourt Brace. R EINFORCEMENT Reinforcement is deﬁned as a consequence that increases the likelihood that a behavior will occur in the future. Understanding the use of reinforcement to teach new behaviors and strengthen existing behaviors is critical to the understanding of interactions between individuals. What may be a reinforcer at one time may not be the next time it is presented. A change in the power of a reinforcer is often based on many factors, such as satiation, deprivation, mood, age, or social interaction. Whether or not an event in one’s environment is a reinforcer is determined by the effect it has on the behavior that precedes it, not on the intent of the person delivering it. An item is a rein- forcer only if it increases the likelihood that the behavior will occur under similar circumstances in the future. Understanding that an item is only a rein- forcer by its interaction with behavior is fundamental, and reinforcement, in this sense, can be related back to B. F. Skinner. Positive and Negative Reinforcement There are two types of reinforcement: positive and negative. Positive reinforcement is the addition of a stimulus that increases the likelihood that the behavior that preceded it will occur again. Negative reinforcement is the removal of a stimulus, such as escape from a task, which also increases the like- lihood that a behavior will occur again. Negative reinforcement is often confused with punishment, but negative reinforcement increases the likelihood that the preceding behavior will occur again, whereas pun- ishment decreases the likelihood. Types of Reinforcers There are functionally two types of reinforcers: primary (unconditioned) and secondary (conditioned). Primary reinforcers are events that individuals need and, by virtue of that, can be used as consequences to increase the likelihood that the preceding behavior will occur again. Examples include food, water, and sleep. Deprivation and satiation are especially impor- tant concepts when discussing primary reinforcers. A primary reinforcer that an individual has been deprived of will have extremely strong reinforcing Reinforcement 845 properties. Deprivation is limiting access to the item so that the individual strongly needs the item. An example of this might be if an individual has not drunk in several hours and, upon seeing water, might say the word ‘‘water.’’ If the person is given the water, that water may act as a reinforcer, increasing the likelihood that saying the word ‘‘water’’ will hap- pen again. In addition, if an individual satiates on an item that typically has reinforcing properties, that reinforcer may no longer have any reinforcing power. Secondary reinforcers are events that become rein- forcers after being paired with other reinforcing events. When delivered as a consequence, they also increase the likelihood that the preceding behavior will occur again. Examples might include stickers, social praise, or a toy. Often, a secondary reinforcer is delivered concurrently with a primary reinforcer or with another established secondary reinforcer. So, when ﬁrst delivering a sticker as a reinforcer, a teacher might also deliver a small bite of food. With time and with systematic fading of delivery of the food, the sticker may have similar reinforcing properties as the food. Another way that reinforcers are often distin- guished is by type. Typically, primary reinforcers might fall in a category of edibles. Secondary reinfor- cers might include a variety of types, such as tangi- bles or activities. Some examples of tangibles might include toys or stickers. Examples of an activity might be playing a game, going to lunch, or going for a walk. Social reinforcers might include tickling, high-ﬁves, a smile, or verbal praise. These are things that are of a social nature, and the interaction is typically what develops as a reinforcer. Examples of generalized reinforcers might include tokens or money. Schedules of Reinforcement When looking at reinforcement, it is also important to look at the schedule of when reinforcers are delivered. Reinforcement can be continuous or intermittent. Intermittent reinforcement schedules can be ﬁxed ratio, variable ratio, ﬁxed interval, or variable interval. Ratio refers to a certain number of responses being necessary before reinforcement is delivered. In a ﬁxed ratio schedule of reinforcement, a set number of responses occur before the reinforcer is delivered. So, an FR6 (ﬁxed ratio 6) schedule would mean that rein- forcement would be given after six correct responses. In variable ratio, the number of correct responses before reinforcement varies. A classic example of variable ratio is a slot machine. Interval refers to a certain amount of time that must pass before a correct response will be reinforced. Similar to ratio, interval may be either ﬁxed or vari- able. In ﬁxed interval, reinforcement is delivered on the ﬁrst correct response after a given interval, and in variable interval, the length varies. Nanette L. Perrin See also Applied Behavior Analysis; Behavior Modiﬁcation; Classical Conditioning Further Readings Alberto, P. C., & Troutman, A. C. (1999). Applied behavior analysis for teachers (5th ed.). Columbus, OH: Merrill. Baer, D. M., Wolf, M. M., & Risley, T. (1968). Current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis , 1 , 91–97. Baer, D. M., Wolf, M. M., & Risley, T. (1987). Some still-current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis , 20 , 313–328. Catania, A. C. (1997). Learning (4th ed.). Englewood Cliffs, NJ: Prentice Hall. Cooper, J. O., Heron, T. E., & Heward, W. L. (1987). Applied behavior analysis. Columbus, OH: Merrill. Iwata, B. A. (1987). Negative reinforcement in applied behavior analysis: An emerging technology. Journal of Applied Behavior Analysis , 20 , 361–378. Michael, J. (1975). Positive and negative reinforcement, a distinction that is no longer necessary; or a better way to talk about bad things. Behaviorism , 3 , 33–38. Skinner, B. F. (1953). Science and human behavior. New York: Free Press. R ELIABILITY The two most important properties of an assessment are its validity and reliability. Validity refers to the meaningfulness of the interpretations and uses of a test score and is the most important property of an assess- ment. Reliability refers to the extent to which test scores are free from errors of measurement. Thus, validity examines the interpretations and uses that can reasonably be made from the consistent part of the test scores, whereas reliability is concerned with inconsistent or random errors of measurement. As a result, reliability is a necessary but not sufﬁcient 846 Reliability condition for validity. That is, there needs to be some level of consistency to understand the meaningfulness of particular uses and interpretations of test scores, but measuring consistently does not guarantee the meaningfulness of the interpretations or uses. Reliability and validity are not global properties of an assessment. Instead, they are properties of speciﬁc uses and interpretations that are made from a set of test scores. A test could be valid for a particular use or interpretation and not for another. For example, a test might measure the curriculum covered in a school without providing valid estimates of student performance because of the length of the tests or the nonequivalence of forms. The same is true for reli- ability. For example, a test might provide reliable scoring without being stable over time. In addition, reliability and validity are a matter of degree. Tests are not considered valid or invalid. Instead, they are valid to some degree. Similarly, a test is not consid- ered reliable or unreliable, but is reliable to some degree. Estimates of reliability are indices that quantify the amount of measurement error for a particular test use or interpretation for a speciﬁed population. Although reliability can be deﬁned broadly in terms of consis- tency or generalizability, speciﬁc statistical indices of reliability will vary depending on the statistical model and the sources of error. The statistical model may be based on classical test theory, generalizability theory, or item response theory. Classical test theory and gen- eralizability theory are based on total scores, whereas item response theory is based on an estimate of a latent trait. In this entry, only classical test theory and generalizability theory are considered. Within each theory, there are multiple indices of reliability based on multiple sources of measurement error, including item heterogeneity, equivalence of test forms, stability over time, and consistency of subjec- tive ratings. Different sources of error would be of concern in different contexts. For example, the test score of a student writing an essay is affected by errors in scoring, whereas the test score from a student taking a multiple-choice test is affected by the hetero- geneity of the items selected to measure the construct. In addition, a test score can be affected by multiple sources of error simultaneously. A student taking the GRE might be affected by the heterogeneity of the items, the form of the test, and the subjectivity of the scoring for the written portion of the test. Thus, there are many types of reliability that vary depending on the sources of error being considered as well as the statistical model or test theory being used. These varying deﬁnitions will be selected based on the par- ticular test use or score interpretation being made, and one type of reliability should not be considered inter- changeable with another. Classical Test Theory and Estimates of Reliability Classical test theory assumes that any observed test score, X , is the sum of a true score, T , and a random error, E.Thatis, X = T + E. The issue of deﬁning the true score is a matter of validity, whereas the issue of deﬁning the random error is a matter of reliability. According to classical test theory, the error is the sum of all random components, whereas the true score is the sum of all consistent effects. Thus, the error is undifferentiated with respect to different sources of randomness, unlike in generalizability theory. Broadly, two indices of reliability are commonly reported: the reliability coefﬁcient and the standard error of measurement. The reliability coefﬁcient ( r)isdeﬁnedastheratio of the true score variance to the observed score vari- ance, or the ratio of the true score variance to the sum of the true score variance and the error variance. Hence, the value of the reliability coefﬁcient is the proportion of variation in test scores that can be attrib- uted to consistent measurement (i.e., the true score). The reliability coefﬁcient ranges from 0.0 to 1.0, with higher values being preferred. At r = 0 : 0, there is no consistency in the measurement procedure and the observed score is equal to random error ( X = E ). At r = 1 : 0, the observed score has no error and is equal to the true score ( X = T ). In practice, the reli- ability coefﬁcient will be somewhere between the two extreme values. The standard error of measurement (SEM) is the standard deviation of the errors of measurement. The SEM ranges from 0.0 to the standard deviation of the observed scores, s x . When the SEM = s x , there is no consistency in the measurement procedures, the reliability coefﬁcient is equal to 0.0, and the observed score is equal to the random error. When the SEM = 0.0, there is perfect consistency in the test scores, the observed score is equal to the true score, and the reliability coefﬁcient is equal to 1.0. In prac- tice, the SEM will fall somewhere between the two extreme values. Reliability 847 The reliability coefﬁcient is an easily interpreted index of the consistency of the test scores because it is in a standard range for all tests. Although the SEM is more difﬁcult to interpret initially, it is in the metric of the test scores that allows for the interpretation of the individual test scores via conﬁdence intervals. Another advantage of the SEM is that it is not based on the true scores, and consequently, it is not inﬂu- enced by sampling errors. The reliability coefﬁcient will be underestimated when the sample range of scores is restricted, whereas the SEM will be largely uninﬂuenced by sampling ﬂuctuations. Types of Reliability Within the framework of classical test theory, there are several types of reliability coefﬁcients based on the source of the random errors. The types of reliabil- ity discussed below are test-retest, alternate form, alternate form test-retest, interrater, split half, and internal consistency. Test-retest reliability is used to examine the stabil- ity of the trait being measured over time. The reliabil- ity coefﬁcient is the correlation between test scores for a sample taking the same test on two occasions. Generally, test-retest reliability is higher when the time span between test administrations is shorter. However, the test-retest reliability should be estimated with a time interval that mirrors the actual use of the test rather than trying to maximize the value of the coefﬁcient. Alternate form reliability is used to measure the equivalence of test scores across two (parallel) forms of a test. The reliability coefﬁcient is the correlation between test scores on the two forms of the test taken by the same sample without a substantial time lag. Usually, half of the sample receives one form ﬁrst (e.g., Form A), and the other half of the sample receives the other form ﬁrst (e.g., Form B) so that there is no order effect. Then, examinees take the form they have not taken yet. Alternate form reliabil- ity is higher when care is taken to make sure that the two forms are equivalent in content and statistical properties (i.e., mean, standard deviation, and distri- bution shape). Alternate form test-retest reliability follows the same procedure as with the alternate form reliability except that there is a time lag between test administra- tions. In this case, the errors of measurement include stability over time and equivalence of the forms. In general, this type of reliability will be lower than alternate form or test-retest reliability, which target only one type of random error. Interrater reliability is used to measure the consis- tency of ratings from subjective scoring. The reliabil- ity coefﬁcient is the correlation between the ratings from two raters on the same sample of writings/ essays. Interrater reliability is higher when standard- ized procedures are used by the raters to score the writings. At a minimum, the standardized procedures should include training of the raters and clearly deﬁned rubrics. Large-scale assessments further stan- dardize the procedures to include benchmark writings, monitoring the process, intervening when ratings disagree, and other procedures to check the rating process. Split half reliability is used to measure the consis- tency within a single administration of a test by exam- ining the relationship between two halves of the same test. The procedure for split half reliability is to administer a single form of the test to a sample. The reliability estimate is then based on the correlation between two halves of the test adjusted for test length. That is, the test is divided into two equivalent halves based on test content and item statistics (often, this can be accomplished by using odd- and even-num- bered items to form the halves), and the halves are correlated. However, the reliability will be less for half of a test than it is for the full-length test. Conse- quently, the correlation between the halves is adjusted upward using the Spearman-Brown prophecy formula. Split half reliability will be higher when the equiva- lence of the two forms is higher in terms of content and item statistics. However, the matching of the two halves should not be completed on the basis of the sample statistics because random sampling ﬂuctua- tions could inﬂate the value of the reliability. Instead, careful matching should be completed based on con- tent and item statistics from a prior data collection. Internal consistency is used to measure the consis- tency of items within a single test form. The proce- dure for internal consistency is to administer a single form of the test to a sample and estimate the internal consistency using item and test statistics with an inter- nal consistency formula. The formula for internal con- sistency has many equivalent forms in the literature, including the Kuder-Richardson 20 (KR20) formula for dichotomously scored items and Cronbach’s alpha. Internal consistency is also easy to compute with most standard statistical software (e.g., SPSS or SAS). 848 Reliability Internal consistency is higher when the items are more homogeneous. Below is a summary of the reliability coefﬁcients and their major sources of error that are reported in classical test theory: Each of the reliability coefﬁcients above differs in its data collection procedure, computation, and major source of error. The ‘‘appropriate’’ reliability coefﬁ- cient should match the intended use or interpretation of the test. For example, when subjective measurements are part of the assessment procedure, interrater reliabil- ity is needed. When multiple items are being used (which should be the case), in ternal consistency or split half reliability should be used. In short, the reliability estimate(s) should include all sources of error that will be part of the test use or interpretation. One type of reliability should not substitute for another. Standard Error of Measurement The reliability coefﬁcient is used to quantify the precision of an assessment for a particular use or interpretation. The index is simple to interpret because it is always based on the same scale (0.0–1.0). However, the reliability coefﬁcient fails to show the amount of error that might be expected in an indivi- dual’s test score. The SEM is the standard deviation of the errors of measurement and can be used to cre- ate conﬁdence intervals for examinee scores. Assum- ing a normal distribution, 68 % of the observed scores will be within one SEM of their true score, and 95 % of the observed scores will be within 1.96 SEMs of their true score. For example, if SEM = 2.00 and an examinee’s true score was 25, upon repeated measurements, 68 % of the scores for that examinee would be between 23 and 27. Note that the conﬁdence interval is around the true score and not the observed score, which leads to the interpretation that 68 % of the time that a conﬁdence interval based on one SEM is constructed, it will contain the true score. As with the reliability coefﬁcient, a SEM can be created for different types of measurement error. In fact, the SEM is calculated using the appropriate reli- ability coefﬁcient so that the appropriate source of error is being used. Thus, the table mentioned earlier can be used for the SEM or the reliability coefﬁcient so that the SEM can be created with each of the different sources of error. Magnitude of Reliability The literature does not provide deﬁnitive guidance on acceptable levels of reliability. However, it is clear that what constitutes an acceptable level of reliability is determined by the use of the test. Uses of the test with higher stakes require higher levels of reliability. For example, reliability for a test being used in theo- retical research may not require the same level of con- sistency as would be required for high-stakes uses of tests such as high school graduation, certiﬁcation, or licensure. How to Increase Reliability It is important to be able to increase reliability when developing instruments. In general, there are two ways that should always be considered when increasing reliability: greater standardization and increasing the number of items. Test administration and test development procedures should be standard- ized so that no random errors are introduced. The effect of the standardization will not only globally affect each type of reliability, but it will also have speciﬁc effects on certain types of reliability. Stan- dardization includes methods to create equivalent forms of a test (alternate form), methods to create homogeneous pools of items (internal consistency), or equivalent halves of tests (split half). Standardization also includes methods to create consistency in scoring through the development of rubrics and standardized scoring procedures (interrater). Another key element to increasing reliability is increasing the length of the test. The Spearman-Brown formula is based on the principle that longer tests are Reliability Type Source of Error Test-retest Stability over time Alternate form Equivalence across forms Alternate form test-retest Stability over time and equivalence across forms Interrater Consistency of ratings Split half Equivalence across halves Internal consistency Equivalence and item homogeneity Reliability 849 more reliable. Assuming that the conditions of testing do not change with increased l ength (i.e., fatigue, bore- dom, or item quality), increasing the number of items always leads to more reliable tests. Thus, short forms of a test or subscores are generally less reliable than the full form. As a consequence, subscores and short forms are also more difﬁcult to interpret. Thus, a test that provides a reliable tot al score for accountability may not be useful when examining the subscores that might be needed for diagnostic interpretations. Reliability and Aggregation Increasing the number of items will increase the reliability of test scores because the scores are aver- aged over more data. Similarly, increasing the number of examinees and averaging across their scores will reduce the SEM and increase reliability. That is, the reliability of the mean will be higher than the reliabil- ity of an examinee. This applies to estimates for the full sample as well as aggregates such as classrooms, schools, or states. Whether averaging across items or examinees, the estimate becomes more stable. Indeed, the SEM will almost always be lower for group means than it is for individual means. (Note: under some conditions, the reliability coefﬁcient could be lower for the group means when the true score vari- ance in the groups is restricted in range. However, even under these conditions, the SEM will typically be lower and the group means will be more stable.) Reliability and Growth Scores The reliability of growth or difference scores, deﬁned as posttest minus pretest, has received consid- erable attention in the literature. Some have argued that the growth score is unreliable and that growth is negatively correlated with the pretest. That is, growth will be higher for examinees with low pretests. How- ever, other researchers have pointed out that low reli- ability for the difference scores does not necessarily result in less power for comparisons among groups, and the difference score may be the construct of inter- est. At any rate, caution should be used in interpreting growth scores at the examinee level. Relationship of Reliability and Validity Classical test theory assumes that an examinee’s test score is the composite of a true score and random error. Validity addresses the true score by examining its uses and interpretations. Thus, any systematic error or bias is part of the true score, whereas only random errors are addressed in the reliability analysis. As dis- cussed above, reliability is a necessary but not sufﬁ- cient condition for validity. This means that there needs to be a true score (with some level of reliabil- ity) to examine validity, but that the existence of a true score does not guarantee that it is not a biased esti- mate of the construct of interest as a result of some systematic error. In addition to this relationship of validity and reli- ability, there may be a tension between the two psycho- metric properties of the test. Higher reliability can be attained by standardizing the testing procedure, which has the potential to reduce the breadth of the construct being measured and, thus, to decrease the validity. For example, higher internal consistency is attained by increasing item homogeneity. To the extent that the construct requires heterogeneity of the items, this will create a

bundet

tension between reliability and validity. As a second example, interrater reliability is increased by standardizing the scoring procedure (e.g., clearly deﬁned rubrics and training) . This standardization can limit the deﬁnition of good writing by not including some types of writing in the rubric and thus, as a conse- quence, limit the breadth of the construct. As a result, it is important to consider the impact of any standardiza- tion on the validity of the test as well as the reliability so that the construct is still clearly being measured. Generalizability Theory and Estimates of Reliability Classical test theory examines errors of measurement with a single undifferentiated error that may represent multiple sources of error (e.g., alternate form test- retest reliability). In addition, classical test theory examines reliability only from a norm-referenced per- spective. That is, the methods rely solely on correla- tions that focus on rank ordering of scores. The correlations are sensitive to differences in rank order- ing but not to shifts in scale. Thus, the reliability can be high even when the scales for the two forms, raters, and so on, are substantially different. For example, Rater A could rate 10 points higher than Rater B, and the reliability would be equal to 1.0 as long as every essay was scored exactly 10 points higher by Rater B than Rater A. 850 Reliability Generalizability theory solves each of these issues by (a) modeling multiple sources of error and (b) dif- ferentiating between errors based on rank ordering (i.e., relative error) and errors based on point estima- tion (i.e., absolute error). Generalizability theory assumes that each examinee has a universe score that is his or her average score across all conditions in the universe of admissible observations. That universe is composed of measurement facets (e.g., raters, items) with a particular level of a facet being a condition (e.g., selected raters or items). The potential measure- ment conditions selected from the study are then considered to be the universe of generalization. Gen- eralizability theory is more complex statistically than classical test theory and is done in two stages: the G-study and the D-study. In the G-study (general- izability study), random effects analysis of variance (ANOVA) is used to estimate variance components for each of the effects in the model. The ANOVA, with the associated variance components, can be esti- mated with one or more facets (e.g., persons by items, or persons by items by raters). In the D-study (deci- sion study), alternative measurement models can be examined to optimize the measurement procedures or to examine a reasonable set of measurement condi- tions. The results of the D-study identify the universe of generalization. Similar to classical test theory, there are two types of indices computed in the D-study that show the con- sistency of the measurement procedure. The general- izability coefﬁcient, or dependability index, is the ratio of the universe score variance to the universe score variance plus the error variance, and it is analo- gous to a reliability coefﬁcient, whereas the second index is analogous to the SEM. The generalizability coefﬁcient shows the ratio when using relative error variance and emphasizes the rank ordering of scores. Thus, it would be used for norm-referenced score reporting. The dependability index shows the ratio when using absolute error variance and emphasizes the absolute magnitude of the test scores. Thus, it would be used for criterion-referenced score report- ing. The second index is the standard error. The stan- dard error also can be computed for relative or absolute score reporting. The use and interpretations of the indices in generalizability theory are analogous to the indices in classical test theory. Absolute error is always greater than or equal to relative error. Consequently, the generalizability co- efﬁcient is less than or equal to the dependability index, and the absolute standard error is greater than or equal to the relative standard error. As a result, more conditions (items, raters, etc.) may be needed when estimating examinee scores absolutely rather than relative standing. Consistency of Classification Each of the reliability coefﬁcients, whether from classi- cal test theory or generalizab ility theory, is based on continuous variables. Ofte n, the measurement proce- dure is based on the classiﬁcation of examinees. For example, the National Assessment of Educational Progress (NAEP) classiﬁes students as Advanced, Pro- ﬁcient, Basic, and Below Basic. Clearly, when the data are nominal or categorical, a different statistical proce- dure must be used to examine consistency. Decision consistency is a method of ex amining reliability and the exact agreement across measurement conditions when the data are categorical. Decision consistency can be calculated for each of the sources of error describedaboveinthesectiononclassicaltesttheory. Twoindicescommonlyre ported for decision con- sistency are the proporti on agreement and Cohen’s Kappa. Proportion agreement shows the proportion of the examinees that are classiﬁed the same across forms, time, and so on. For example, if two forms were given to a sample for the NAEP reading assessment in Grade 8, the proportion agreement would be equal to the sum of the proportions that were Advanced on both forms, Proﬁcient on both forms, Basic on both forms, and Below Basic on both forms. The same calculations could be done with other sources of error (e.g., ratings are the same across two raters). Cohen’s Kappa is the proportion agreement after statistically adjusting for the expected agreement. Thus, Cohen’s Kappa shows the agreement above and beyond chance and generally has lower values than the proportion agreement. M. David Miller See also Assessment; Descriptive Statistics; Evaluation; Testing Further Readings American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (1999). Standards for educational and psychological testing. Washington, DC: American Educational Research Association. Reliability 851 Brennan, R. L. (2001). Generalizability theory. New York: Springer-Verlag. Haertel, E. H. (2006). Reliability. In R. L. Brennan (Ed.), Educational measurement (4th ed., pp. 65–110). Westport, CT: Praeger. Lord, F. M., & Novick, M. R. (1968). Statistical theories of mental test scores. Reading, MA: Addison-Wesley. R ISK F ACTORS AND D EVELOPMENT Optimal child development may be compromised by several risk factors, includi ng poverty, poor nutrition, harsh/inconsistent parent ing, familial substance abuse, and low parental education, to name a few. Children developing within these maladaptive contexts often exhibit poor developmental outcomes across cognitive, social, and behavioral domains, such as low academic achievement and antisocial behavior. Some children exposed to great adversity, however, develop compe- tence, suggesting that some factors may protect (or buffer) high-risk children from experiencing negative developmental outcomes. The purpose of this entry is to (a) describe the theoretical framework guiding the research on the development of risk and protective factors; (b) identify how risk factors at the child, parent, family, and community level negatively affect developmental outcomes; and © describe how protec- tive factors may moderate negative outcomes in high- risk children. This entry also discusses the importance, capacity, and effectiveness of early intervention pro- grams that are designed to increase educational out- comes by reducing the nega tive consequences of risk factors. Ecological Systems Theory as a Guiding Theoretical Framework Ecological frameworks recognize that each person functions within a complex network of individual, fam- ily, community, and environmental contexts that inﬂu- ence the availability of risk and opportunities to avoid risk. Urie Bronfenbrenner is a pioneer of the ecological model of human development, in which he describes development as the composite of individual genetic endowment, immediate family inﬂuences, and other components of the environmental context. According to Bronfenbrenner, the indi vidual is embedded in ﬁve interrelated, nested subsystems that simultaneously inﬂuence the process of human development. The innermost circle represents the microsystem of the individual. Within the microsystem are the indivi- dual’s interactions with his or her immediate settings. Bronfenbrenner refers to these interactions as proxi- mal processes. For most children, the family is the ﬁrst and most important microsystem. As develop- ment proceeds through childhood and adolescence, additional microsystems might be sports teams, youth or church organizations, and work. Microsystems may overlap in that the same person may be a member of more than one system in a child’s life. For example, a friend may be a member of the child’s peer group, sports team, and class in school. Just as the individuals interact with others in their microsystem, separate microsystems interact with each other at the level of the mesosystem. The meso- system incorporates linkages between settings such as family, peers, teachers, and other school personnel. For example, an adolescent’s ability to excel in school may depend more on the interconnections between the school and the home rather than solely on ade- quate performance in the classroom. In this case, Bronfenbrenner posits that the breakdown of connec- tions between family, school, peer group, and neigh- borhood underlie the declin e of academic achievement more so than relationships within each of these con- texts alone. Continuing the progression to increasingly distal inﬂuences on the individual, the exosystem represents external environments that the individual may or may not experience; yet events that occur in these environ- ments affect what happens in the microsystem (an individual’s immediate setting). This system includes features of the community such as availability of ser- vices or employment, access to formal and informal support, and socioeconomic climate. Parental and teacher social networks are part of the exosystem, having the potential to inﬂuence interactions with the child even though they may not directly be experi- enced by that child. The overarching macrosystem represents the larger societal contexts, and inﬂuences differ across socio- economic, ethnic, religious, and other subcultural groups. For example, in the United States, macrosys- tem inﬂuences include American culture and the social policies and programs that affect American families, such as racial prejudice and discrimination, the media, and antipoverty social programs. Finally, 852 Risk Factors and Development each subsystem is embedded with time, the chrono- system, which represents that each of these subsys- tems functions within time and changes over time. Applying this model to one of the most researched child outcomes, low academic achievement, indicates that risks begin in the most proximal system. The microsystem includes risk factors such as impulsivity, poor attention, inadequate school readiness, poor physical health, low school attendance, and so on. In the mesosystem, low academic achievement relates to the interactions between the family and the school, and include such factors as parenting styles, parent involvement in academic endeavors, teaching styles, school environment, and so on. The community con- text also affects a child’s ability to perform well aca- demically, given such factors as community violence, neighborhood crime, parent employment, school poli- cies, transportation availability, and so on. Finally, through cultural values, governmental policies, and the political climate, the larger society also begins to inﬂuence academic achievement. Given the inﬂuences from multiple levels of children’s ecology, it becomes evident that the risks for low academic achievement cannot be understood from one context alone. Risk Factors Across the Ecological System The structure of the ecological model allows for understanding the inﬂuences on child development at multiple levels and informs the process by which con- textual factors inﬂuence child development. Although this framework identiﬁes risk factors for child devel- opment, the concept of risk initially evolved from medical research examining the cause of cardiovascu- lar disease. A medical research study that began in 1948, called the Framingham Heart Study, followed multiple cohorts of individuals across several decades in an effort to identify the factors that put individuals at risk for developing cardiovascular disease. The same logic has been applied to human development and enhanced understanding of why some children facing adversities from multiple ecological levels develop into cognitively and socially competent adults, and others do not. Several longitudinal studies have documented risk factors and the decline of optimal child and adolescent development (e.g., the Perry Preschool Project, the Abecedarian Project, the Milwaukee Project, and Project CARE). As research studies began to investigate the cause of poor developmental outcomes and risky behavior (e.g., low academic achievement, teen pregnancy, juvenile delinquency), three categories emerged: child characteristics, parent/family characteristics, and com- munity characteristics. Although risk factors are often studied and discussed within these categories, the eco- logical framework informs the interrelated nature of each factor and ecological system. Child Characteristics The three main risk factors associated with poor developmental outcomes inc lude low birth weight, tem- perament, and child health. First, compared to infants born at normal weight ( > 5 lbs.), children born with low birth weight (1 to 4 lbs.) are more likely to have continual health and developmental problems (e.g., cerebral palsy); lower IQ; and pervasive neuropsycho- logical impairments in attention, memory, problem solving, reasoning, and language. Marked negative behavioral and academic outcomes are also evident, including higher incidence of conduct disorder, hyper- activity, attention problems, and learning disorders. Furthermore, the developmental consequences of low- birth-weight babies are particularly devastating for the extremely low-birth-weight babies (less than 2 lbs.). The second risk factor at the child level is tempera- ment. Children with difﬁcult temperaments are at much higher risk for poor developmental outcomes than easy-temperament babies. Difﬁcult children have a negative effect on their caregivers and thus are less likely to experience supportive relationships with them, which is a necessary component to counteract exposure to adversity. Difﬁcult babies often develop into difﬁcult adolescents, who have been shown to exhibit more emotional and behavior problems than adolescents who were easy babies. The third risk factor associated with poor develop- mental outcomes is poor child health. Frequent hospi- talizations, chronic medical conditions, and repeated illnesses increase the stre ss level within the family, which has been related to later psychosocial and aca- demic difﬁculties in school-age children. Poor health in childhood has a long-term impact on development across the life span and is related to early mortality rates. Parent/Family Characteristics Research has uncovered several parental and family-level risk factors associated with maladaptive Risk Factors and Development 853 development, including family structure and resources, parental background characteristics, and relationships within the family. One of the most pervasive negative effects on developmental outcomes is a family living in poverty. Research suggests that children raised in persistent poverty (as opposed to transitory poverty) perform much worse academically—as measured by achievement test scores, grade promotion, high school graduation, and completed years of schooling—than more afﬂuent children. Furthermore, children in eco- nomically strained environments are almost twice as likely as nonpoor children to experience learning dis- abilities and developmental delays. Parent, teacher, and child self-reports also reveal that low-income children display more externalizing problems, such as ﬁghting, difﬁculty getting along with others, and impulsivity, in comparison to higher-income children. Finally, researchers report that children of poor families have more emotional and behavioral problems than do children who have never been poor. Vonnie McLoyd reviewed the research on children of low-income families. Results indicate that family income produces the strongest relationship with aca- demic achievement of all other variables considered (e.g., parent’s occupation and parent’s education level). Even when the other indicators are summed together, they do not yield as high a correlation with academic achievement as does income alone. This line of research shows that economic hardship inﬂu- ences the psychological well-being of parents and the relationships among family members. Studies reveal that the inﬂuence of economic stress on children and adolescents is mediated to a signiﬁcant extent by dis- ruptions in family relationships and interactions, espe- cially parent–child interactions. In addition, families headed by single parents are also six times more likely to be poor than are families from two-parent households and therefore are more likely to be affected by poverty characteristics. Single parenthood disrupts the relationship between children and the noncustodial parent, usually the father. This may affect the extent to which supportive, caring rela- tionships can be developed and nurtured. Psychologi- cal problems and behavior problems are more likely in children from single-parent or remarried families. Another risk factor can be conceptualized as parental background characteristics, which includes parental substance abuse and low parent education. Each has been linked to increased internalizing and externalizing problems among children. Also, children with parents suffering from alcoholism are approxi- mately four to six times as likely as the general popu- lation to develop alcohol problems. Furthermore, anxiety, depression, and externalizing behavior disor- ders (e.g., conduct disorder) are more common among children with alcoholic parents than among children of nonalcoholics. In families where alcohol or other drugs are being abused, behavior is frequently unpre- dictable, communication is unclear, and family struc- ture is either nonexistent or inconsistent. Because parents who abuse alcohol or other drugs are more likely to be involved with domestic violence, divorce, unemployment, mental illness, and legal problems, their ability to parent effectively is severely compro- mised. There is a higher prevalence of depression, anxiety, eating disorders, and suicide attempts among children of parents who abuse drugs and alcohol than among their peers. Family background factors, such as parental educa- tion, have moderate effects on child development but larger effects on later adult outcomes. That is, parental education level affects later adult outcomes through transmission of educative values, such as not dropping out of school. Parent inco me and education level are associated with school drop- out rates, which drastically reduce the quality of later adu lt life, increasing the like- lihood of unemployment and poverty. The ﬁnal risk factor to be discussed here includes the quality of the relationships within the family (i.e., parent–child, sibling, and marital relationships). In general, research reports that warm and supportive relationships within the family are associated with positive child and adolescent outcomes, whereas coer- cive and conﬂictual relationships are associated with the development of problems. One of the clearest and most replicated associations in this area is between conﬂictual parent–child relationships and child and adolescent maladjustment. For example, harsh, incon- sistent, and ineffective discipline is associated with later antisocial behavior, and insecure parent–child relationships, including parental rejection and force- fulness, are traced to later internalizing problems. Community Characteristics The community context can also inﬂuence a child’s developmental trajectory. Research has shown that poor neighborhoods can have detrimental effects on individ- ual health status through three types of pathways. First, concentration of poverty and re lated characteristics may 854 Risk Factors and Development create more detrimental social environments (e.g., vio- lence, stress and anxiety, exposure to drugs, limited social control). Second, poorer communities are less likely to have access to adequate health care and social services. Third, the physical environment (e.g., air pol- lutants, hazardous conditions leading to accidents, poorer sanitation) in poor communities may be worse than in more afﬂuent communities. Existing research also points to a powerful connec- tion between residing in an adverse environment and participating in criminal acts. Inner-city neighborhoods provide limited economic, institutional, and social resources for families and adolescents. Disorganized neighborhoods have weak social control networks, which result in isolation among residents, high residen- tial turnover, and criminal activity that goes unmoni- tored. Few opportunities for youth to be monitored (such as programs where youth can contribute to the community), as well as the increased access to alcohol, drugs, and ﬁrearms, increas e the likelihood of youth engaging in risky and destructive behaviors. Neighbor- hoods with high crime, poverty, and physical disorder areassociatedwithyouthengagementinviolence, crime, and drugs and alcohol; low community attach- ment; and a feeling of being unsafe. The Impact of Cumulative Risk Although each risk factor across the ecological system has been related to a negative cognitive, social, or behavioral impairment, the impact of one risk factor tends to be relatively small in relation to the cumula- tive effect of multiple risk factors. That is, research suggests that children can handle some level of adver- sity, but when faced with three to four risk factors, developmental outcomes become drastically and neg- atively affected. Furthermore, the more risks to which children are exposed, the worse the developmental outcomes. Therefore, children’s optimal development is ultimately affected by the sheer number of risk fac- tors present, regardless of the nature of a particular risk factor. As the number of risk factors increases, the negative effect enlarges disproportionately. For example, having four or more risk factors, which relate to the child, parent, and sociodemographic situ- ation, can lead to a 10-fold increase in difﬁculties, a result that has been replicated numerous times across the psychological research. It is important to note that although poor develop- mental outcomes tend to relate to the presence of risk factors, the absence of risk factors does not predict healthy development. Appropriate opportunities must be presented for development to ﬂourish and reach optimal levels. For example, opportunities afforded by high parental warmth, rather than just a lack of harsh and inconsistent parenting, are essential to pro- mote healthy development. Research indicates that development is a balance between opportunities for strengths and risks to develop. Within each of the eco- logical levels, there are risk factors that negatively inﬂuence development and there are protective factors that foster positive development. Risks, cumulative risks, and protective factors interacts generating a complex set of relationships that creates develop- ment. Knowledge of risks alone is not sufﬁcient to understanding child development, although these fac- tors have received the most research attention. Modifying the Impact of Risk Factors on Development The negative impact of cumulative risks may be offset by understanding the conditions that promote success- ful development. Resiliency research has demonstrated that there are certain factors that protect against unde- sirable behavior. Resiliency has been deﬁned as the capacity of the person, family, or community to pre- vent, minimize, overcome, or thrive in spite of negative or challenging circumstance s. There are three kinds of resiliency described in the research literature. The ﬁrst involves an individual who overcomes all the odds, suggesting that this individual has a particular quality that allows him or her to withstand adversity. The sec- ond consists of coping in the face of sustained and acute negative circumstances, such as extremely high family conﬂict. Third, resilience can refer to recovery from trauma, such as the death of a sibling. All require risk to be present for resilience to emerge; however, most research tends to fall within the second group: success in the face of adversity. Identifying protective factors is one way that resil- iency can be measured. Protective factors are those that relate to positive developmental outcomes for youth exposed to high levels of risk or adversity only. If positive outcomes are observed across all levels of risk (high and low), the factor is termed an asset, compensatory, or promotive. Thus, protective factors are assets that particularly matter when risk or adver- sity is high. Protective factors moderate the impact of Risk Factors and Development 855 adversity on adaptation, changing the expected nega- tive child outcome. A landmark study on resiliency, conducted by Emmy Werner, has followed the development of chil- dren born on the Hawaiian island of Kauai since 1955. It has provided a wealth of data on protective factors for positive development in children with high cumulative risk. In this study, the risk group (about one third of the children) was deﬁned by having four or more early risk factors that included poverty, peri- natal stress, family conﬂict, and low parental educa- tion. About one third of these high-risk children developed well in terms of getting along with parents and peers, doing ﬁne in school, avoiding serious trou- ble, and having good mental health. The resilient group had more resources and fewer adversities from an early age, including (a) close attachment to at least one good parent, (b) family harmony (i.e., conﬂict- free environment), © lack of parental psychopathol- ogy, (d) more time before the next child in the family came along, (e) easier temperaments as babies, (f) better intellectual skills and competence, (g) more connections with prosocial adults, (h) fewer separa- tions from caregivers, and (i) better overall physical health. Children were also responsible, self-conﬁdent, and motivated to achieve. On the other hand, the nonresilient group exhibited many of the same negative developmental outcomes associated with high-risk youth described earlier. The follow-up study followed high-risk children who were exposed to chronic poverty, birth complications, parental psychopathology, and family discord into adulthood. With the exception of serious central ner- vous system damage, the impact of adversities during childhood diminished adult adaptation depending on the quality of the child-rearing environment and the emotional support provided by family members, friends, teachers, and adult mentors. Poorest outcomes at age 40 were associated with prolonged exposure to parental alcoholism and/or mental illness. Men and women who had encountered more stressful life events in childhood reported more health problems at age 40 than did those who had encountered fewer losses and less disruption in their family during the ﬁrst decade of life. This study demonstrates the need for early attention to the health status of our nation’s children, especially those who are exposed to poverty, serious perinatal complications, and parental psycho- pathology. The policy implications are clear: Early access to good preventive and ameliorative health services and proper attention to the quality of early child care can result in improved quality of life in adulthood. Implications for Educational Psychology As a group, children who live in poverty tend to per- formworseinschoolthandochildrenfrommore privileged backgrounds. For the ﬁrst half of the 20th century, researchers attributed this difference to inher- ent cognitive deﬁcits. At the time, the prevailing belief was that the course of child development was dictated by biology and maturation. These preliminary results caught the attention of Sargent Shriver, Presi- dent Lyndon Johnson’s chief strategist in implement- ing an arsenal of antipoverty programs as part of the War on Poverty. His idea for a school readiness pro- gram for poor children focused on breaking the cycle of poverty. Shriver reasoned that if poor children could begin school on an equal footing with wealthier classmates, they would have a better chance of suc- ceeding in school and avoiding poverty in adulthood. He appointed a planning committee of 13 profes- sionals in physical and mental health, early education, social work, and developmental psychology. Their work helped shape what is now known as the federal Head Start program. The three developmental p sychologists in the group were Bronfenbrenner, Mamie Clark, and Edward Zigler. Bronfenbrenner co nvinced the other members that intervention would be most effective if it involved not just the child but the family and community that comprise the child-rearing environment. Parent involvement in school operations and administration was unheard of at the time, but it became a cornerstone of Head Start and proved to be a major contributor to its success. Zigler had been trained as a scientist and was distressed that the new program was not going to be ﬁeld-tested before its nationwide launch. Arguing that it was not wise to base such a massive, innova- tive program on good ideas and concepts but little empirical evidence, he insisted that research and evaluation be part of Head Start. Although it is difﬁcult to summarize the hundreds of empirical studies of Head Start outcomes, high-quality Head Start programs do produce a variety of beneﬁts for most children. Although some studies have suggested that the intellec- tual advantages gained from participation in Head Start gradually disappear as children progress through 856 Risk Factors and Development elementary school, some of these same studies have shown more lasting beneﬁts in the areas of school achievement and adjustment. Head Start began as a great experiment that has yielded proliﬁc results over the years. Some 20 mil- lion children and families have participated in Head Start since the summer of 1965; current enrollment approaches one million annually, including those in the new Early Head Start, which serves families with children from birth to age 3. Psychological research on early intervention has proliferated, creating an expansive literature and sound knowledge base. Many research ideas designed and tested in the Head Start laboratory have been adapted in a variety of service delivery programs. These include family support ser- vices, home visiting, a credentialing process for early childhood workers, and education for parenthood. Head Start’s efforts in preschool education spotlighted the value of school readiness and helped spur today’s movement toward universal preschool. The existing literature also provides some guide- lines for program design. Arthur Reynolds suggests eight principles for designing effective interventions for children at risk for low academic achievement: 1. Target children and families who are at the highest risk of school difﬁculties. 2. Begin participation early and continue to the second or third grade. 3. Provide comprehensive child-development services. 4. Encourage active and multifaceted parent involvement. 5. Create a child-centered, structured curriculum approach. 6. Have small class sizes and teacher/child ratios. 7. Provide regular staff development and inservice training for certiﬁed teachers. 8. Evaluate and monitor the interventions systematically. These principles suggest that although it may be useful to intervene before 3 years of age, interven- tions for preschool and for school-age children can also be effective. Thus, the ﬁrst 3 years should not be emphasized at the expense of interventions aimed at older children. Second, the effects of early inter- vention have often been found to be larger for more disadvantaged children, which provides a rationale for targeting very high-risk children in particular. In addition to focusing on low-income children, it might be useful to target other aspects of disadvantage, such as lack of maternal education. Third, the most important aspect of quality is likely to be the nature of the interaction between the teacher and the child. Small group sizes, better teacher training, and other regulable aspects of quality all make such interac- tions more likely. Moreover, even rather loose regu- lation of these observable aspects of quality by Head Start has been shown to be effective in eliminating poor-quality programs. Future of Risk Reduction Programs What we know from the ﬁeld of intervention and pre- vention programming is reﬂected in research on effec- tive, risk-reducing, and resilience-building programs. Effective services provide contexts that both reduce the impact of risk factors and foster the development of new or existing protective factors. These programs build on inherent strengths within families, schools, and communities and enable these institutions to help children succeed. And more than just helping chil- dren, the best programs also support those who care for and provide services to these children, thereby enhancing their capacity to care. These programs address child development at a variety of stages, from prenatal care through postsecondary employment— stages that some would even say are too early or too late for appropriate intervention. These programs repeatedly demonstrate that resilience, rather than being solely dependent on individual characteristics, can be socially constructed. Tiffany Berry and Elise Arruda See also Head Start; Home Environment and Academic Intrinsic Motivation; Longitudinal Research; Social Class and Classism Further Readings Child Welfare League of America (The University of Illinois at Chicago), Reynolds, A. J., Wang, M. C., & Walberg, H. J. (Eds.). (2003). Early childhood programs for a new century. Washington, DC: CWLA Press. Goldstein, S., & Brooks, R. B. (Eds.). (2005). Handbook of resilience in children. New York: Kluwer. Groark,C.J.,Mehafﬁe,K.E.,McCall,R.,&Greenberg,M.T. (Eds.). (2006). Evidence-based practices and programs for Risk Factors and Development 857 early childhood care and education. Thousand Oaks, CA: Corwin. Lerner, R. M., Wertlieb, D., & Jacobs, F. (Eds.). (2005). Applied developmental science: An advanced textbook. Thousand Oaks, CA: Sage. R OSENTHAL E FFECT The term Rosenthal effect is deﬁned in its most gen- eral form as the effect of interpersonal expectations (i.e., the ﬁnding that what one person has come to expect from another can come to serve as a self- fulﬁlling prophecy). This concept is relevant to educa- tional psychology in two distinct domains: the domain of research methodology (the experimenter expec- tancy effect) and the domain of learning and behavior (the teacher expectancy effect). The concept of interpersonal expectation effects has been investigated in a wide array of settings, including the relationship between judges’ expecta- tions and their nonverbal behavior as they address the jury, and juries’ subsequent verdicts of guilty or not guilty; the effects of managers’ expectations for the performance of their employees, and employees’ actual subsequent performance; and the effects of expecta- tions of health care providers for their patients’ subse- quent health outcomes and pa tients’ actual subsequent outcomes. Although interpersonal expectations have been studied in many domains, this entry gives primary attention to those domains of greatest relevance to students of educational psyc hology: the experimenter expectancy effect and the teacher expectancy effect. Experimenter Expectancy Effect The experimenter expectancy effect is one of the sources of artifact or error in scientiﬁc inquiry. Spe- ciﬁcally, it refers to the unintended effect of experi- menters’ hypotheses or expectations on the results of their research. Some expectation of how the research will turn out is virtually a constant in science. Social scientists, like other scientists generally, conduct research speciﬁcally to examine hypotheses or expectations about the nature of things. In the social and behavioral sciences, the hypothesis held by the investigators can lead them unintentionally to alter their behavior toward the research participants in such a way as to increase the likelihood that participants will respond so as to con- ﬁrm the investigator’s hypo thesis or expectations. We are speaking, then, of the investigator’s hypothesis as a self-fulﬁlling prophecy. One prophesies an event, and the expectation of the event then changes the behavior of the prophet in such a way as to make the prophesied event more likely. The history of science documents the occurrence of this phenomenon with the case of clever animals that were cued uninten- tionally to give correct answers in foot taps and in barking to questioners who believed the animals could respond with correct answers to, say, arithme- tic problems. The ﬁrst experiments designed speciﬁcally to inves- tigate the effects of experimenters’ expectations on the results of their research em ployed human research par- ticipants. Graduate student s and advanced undergradu- ates in the ﬁeld of psychology were employed to collect data from introductory psychology students. The experimenters showed a series of photographs of faces to research participants and asked participants to ratethedegreeofsuccessorfailurereﬂectedinthe photographs. Half the experimenters, chosen at ran- dom, were led to expect that their research partici- pants would rate the photos as being of more successful people. The remaining half of the experi- menters were given the opposite expectation—that their research participants would rate the photos as being of less successful people. Despite the fact that all experimenters were instructed to conduct a per- fectly standard experiment, reading only the same printed instructions to all their participants, those experimenters who had been led to expect ratings of faces as being of more successful people obtained such ratings from their randomly assigned partici- pants. Those experimenters who had been led to expect results in the opposite direction tended to obtain results in the opposite direction. These results were replicated dozens of times employing other human research participants. They were also replicated employing animal research sub- jects. In the ﬁrst of these experiments, experimenters were employed who were told that their laboratory was collaborating with another laboratory that had been developing genetic strains of maze-bright and maze- dull rats. The task was exp lained as simply observing and recording the maze-lear ning performance of the maze-bright and maze-dull rats. Half the experimen- ters were told that they had been assigned rats that were maze-bright, and the r emaining experimenters 858 Rosenthal Effect were told that they had been assigned rats that were maze-dull. None of the rats had really been bred for maze-brightness or maze-dullness, and experimenters were told purely at random what type of rats they had been assigned. Despite the fac t that the only differences between the allegedly bright and dull rats were in the minds of the experimenters, those who believed their rats were brighter obtained brighter performance from their rats than did the experimenters who believed their rats were duller. Essentially, the same results were obtained in a replication of t his experiment employing Skinner boxes instead of mazes. Teacher Expectancy Effect If rats became brighter when their experimenters expected them to, then perhaps children could become brighter when expected to by their teacher. Accordingly, all of the children in the ﬁrst study of the teacher expectancy effect were administered an obscure test of intelligence that was disguised as a test that would predict intellectual ‘‘blooming.’’ The test was labeled ‘‘The Harvard Test of Inﬂected Acquisition.’’ There were 18 classrooms in the school, three at each of the six grade levels. Within each grade level, the three classrooms comprised children with above-average ability, average ability, and below-average ability, respectively. Within each of the 18 classrooms, approximately 20 % of the chil- dren were chosen at random to form the experimen- tal group. Each teacher was given the names of the children from his or her class who were in the exper- imental condition. The teacher was told that these children had scored on the ‘‘Test of Inﬂected Acqui- sition’’ such that they would show surprising gains in intellectual competence during the next 8 months of school. The only difference between the experimen- tal group and the control group children, then, was in the mind of the teacher. At the end of the school year, 8 months later, all of the children were retested with the same test of intelligence. Considering the school as a whole, the children from whom the teachers has been led to expect greater intellectual gain showed a signiﬁcantly greater gain than did the children of the control group. Robert Rosenthal See also Halo Effect; Teaching Strategies Further Readings Blanck, P. D. (Ed.). (1994). Interpersonal expectations: Theory, research, and applications. New York: Cambridge University Press. Pfungst, O. (1965). Clever Hans (C. L. Rahn, Trans.). New York: Holt, Rinehart & Winston. (Originally published in 1911) Rosenthal, R. (1966). Experimenter effects in behavioral research. New York: Appleton-Century-Crofts. Rosenthal, R., & Jacobson, L. (1968). Pygmalion in the classroom. New York: Holt, Rinehart & Winston. Rosnow, R. L., & Rosenthal, R. (1997). People studying people: Artifacts and ethics in behavioral research. New York: W. H. Freeman. R UBRICS A rubric is a measurement and instructional tool that communicates instructor expectations to learners and uses explicitly stated criteria to categorize levels of performance regarding various skill levels, behaviors, product quality, and/or process quality. Rubrics are useful when instructors assess work in which learners have constructed knowledge and the work may not be easily scored using an answer key. For example, rub- rics can be used to evaluate a performance of a task, a process such as problem solving, a written paper on a topic, a portfolio, and so on. In addition, a rubric enables both the instructor and the learners to identify desirable aspects of the work and areas needing improvement. A rubric often is a one- to two-page document containing a table or grid outlining criteria in categories that helps the learner identify the instructor’s expectations for a task or project and assists the instructor in evaluating the learner’s work. A rubric contains a Likert-type scale that helps to quantify performance decisions. Rubrics may be generic or task-speciﬁc. A generic rubric is used to evaluate a process such as collabora- tion and may be applied across disciplines or content areas. A task-speciﬁc rubric is used for a clearly deﬁned task and therefore has a narrower application. In addition, rubrics may be classiﬁed as holistic or analytic. Holistic rubrics enable the instructor to judge the overall product or process as a whole without focusing on separate components. This provides a quick, evaluative picture but does not give speciﬁc or detailed feedback to the learner. In other words, holistic rubrics require instructors to evaluate the Rubrics 859 quality of learning through one performance level rather than multiple performance levels. In contrast, analytic rubrics have several indicator categories that instructors rate separately in order to differentiate performance levels within and among categories. Therefore, learners receive speciﬁc and detailed feedback that may be useful to them for improving performance. A scoring rubric is used by instructors exclu- sively for the purpose of assigning grades to lear- ners’ work. In contrast, an instructional rubric helps learners to understand the instructor’s expectations for an assignment and aids learners in focusing their efforts on the assignment. The same rubric may be used for both scoring and instructional purposes if desired. Instructors have found that instructional rubrics enable them to provide learners with more informative feedback without spending inordinate amounts of time. Also, instructional rubrics may help instructors to provide fair and unbiased evalua- tion of learners’ work. In addition, multiple instruc- tors can use the same rubric to attempt to have consistency in scoring. These rubrics can be used for learner self-evaluation and peer evaluation. Two key things to note when using instructional rubrics are that they should be provided to learners at the same time as the assignment and that the instructor should spend time explaining the rubric to the lear- ners. Some instructors co-construct rubrics with the learners by ﬁrst helping them to brainstorm criteria for which their work may be evaluated. Rubrics can be obtained from a variety of print and Internet sources. If an appropriate rubric cannot be located, then an instructor may need to adapt an existing rubric or create a new one. The ﬁrst step in creating a rubric is to examine examples of assign- ments to determine exemplary characteristics. These characteristics then can be grouped into categories. Next, levels of performance or gradations of quality are selected. Usually three to four levels are appro- priate, although a rubric may contain more levels. One disadvantage of having three levels is that when learners use the rubric for self-evaluation, they may have a tendency to place their performance in the middle level. Levels may be labeled in a variety of ways, such as below average, average, and above average, or novice, apprentice, proﬁcient, and distin- guished. Point values also may be attached to each level of performance to help with scoring. Finally, indicators are identiﬁed for each category. Indicators are brief statements that describe the particular char- acteristics of products that demonstrate that learning has occurred. It is important to use language in posi- tive terms that learners, their family members, and other professionals will understand. Describing the best and worst levels of performance and then ﬁlling in the middle levels may be the easiest approach. Rubrics should be evaluated and reﬁned continually if they are to be useful for all intended audiences. In order for a rubric to be of high quality, it must have clear criteria as designated by the performance indicators. In other words, what is to be measured shouldbeclearlydeﬁned.Itisimportanttohave enough criteria (e.g., three to seven categories), but not so many that the rubric is difﬁcult to manage. Using rich, descriptive language enables learners to understand the indicators, and the indicators clearly should differentiate among levels of performance in measurable terms. Also, a rubric should emphasize positive attainment of desired performance as opposed to lack of attainment. A high-quality rubric should be valid (i.e., measure key features related to quality of performance) and reliable (i.e., consistent results for different users). Also, it should be sequenced to corre- spond with the steps in the performance. A high- quality rubric should have clear instructions for scoring. For example, it should be noted whether all categories have equal weight and scores from all categories should be added or averaged, or whether one or more cate- gories should receive more weight. There are many reasons why rubrics should be used in instructional settings. First, they help to make instructor expectations for learners’ work clear and concrete. Second, rubrics provide learners with feedback regarding their strengths and weak- nesses. Third, rubrics help learners to evaluate their own work and foster metacognition. Learners can use rubrics to monitor their own progress as they work on an assignment and use the rubric as a ﬁnal checkpoint before turning in the assignment. The beneﬁts of rubrics to instructors include establishing standards, linking assessment and instruction, pro- viding a consistent and unbiased way of scoring work, and explaining those scores to learners and their family members. Martha J. Larkin See also Bloom’s Taxonomy of Educational Objectives; Evaluation; Grading; Testing 860 Rubrics Further Readings Andrade, H. G. (2000). Using rubrics to promote thinking and learning. Educational Leadership , 57 (5), 13–18. Andrade, H. G. (2005). Teaching with rubrics: The good, the bad, and the ugly. College Teaching , 53 (1), 27–30. Bargainnier, S. (2004). Fundamentals of rubrics. In D. Apple (Ed.), Faculty guidebook. Lisle, IL: Paciﬁc Crest. Jackson, C. W., & Larkin, M. J. (2002). RUBRIC: Teaching students to use grading rubrics. Teaching Exceptional Children , 35 (1), 40–45. Whittaker, C. R., Salend, S. J., & Duhaney, D. (2001). Creating instructional rubrics for inclusive classrooms. Teaching Exceptional Children , 34 (2), 8–13. Rubrics 861 S First you take a drink, then the drink takes a drink, then the drink takes you. —Francis Scott Key Fitzgerald S CAFFOLDING Scaffolding is a process in which support is provided to an individual so that he or she can complete a task that could not be completed independently. The sup- port gradually is removed w hen the individual begins to demonstrate understanding of the task. The concept of scaffolding stems from the work of American psy- chologist Jerome Bruner and colleagues based on Lev Vygotsky’s zone of proximal development. Vygotsky described the zone of proximal development as the dis- tance between the actual developmental level where independent problem solving occurs and the potential developmental level where problem solving can occur with the guidance of an adult or more knowledgeable peer. Key to the zone of proximal development is social interaction and collaborative problem solving. Thus, the zone of proximal development bridges the gap between what an indivi dual can learn and do inde- pendently and what he or she can learn and do with support. The scaffolding process in education bears similarities with the traditional deﬁnition of scaffold- ing, which is a temporary framework that supports workers and materials until a building is constructed or repaired to stand on its own. When scaffolding is used in instruction, learners receive support as needed and then the support gradually is removed as they achieve independence in task mastery. Scaffolding instruction includes several essential elements that do not necessarily need to be followed in order. First, the teacher considers curriculum goals and standards along with student needs to select appropriate tasks. Second, the teacher works with students to estab- lish a shared goal. This involvement may result in stu- dents who are motivated and invested in learning. Third, the teacher actively diagnoses student needs and understandings to ensure that students are making prog- ress. Fourth, the teacher provides tailored assistance as needed through prompting, questioning, modeling, tell- ing, or discussing. Fifth, the teacher helps students to remain focused on the intended goal by asking ques- tions and providing clariﬁcation as well as offering praise and encouragement. Sixth, the teacher provides feedback in the form of a current progress summary and mention of speciﬁc behaviors that contributed to student success. Seventh, the teacher controls frustra- tion and risk by creating an environment in which stu- dents feel comfortable taking risks with their learning without fear of penalty. Finally, the teacher gives stu- dents opportunities to practice the task in a variety of contexts and helps them to be less dependent on the teacher in order for them to internalize the task and eventually be able to perform it independently. The following guidelines can help make the scaf- folding process effective. The teacher can plan instruc- tion by having students begin with tasks that they can perform successfully with little or no assistance in order for them to be aware of their strengths and feel 863 good about their abilities. Helping students to achieve success quickly may alleviate frustration. Then, more challenging tasks can be attempted with assistance. Peer acceptance is important to students, so it is important for the teacher to help students to appear like their peers when possible. Although practicing new and previously learned skills is essential, the teacher should recognize when too much practice may be contributing more to student frustration than to learning. The teacher should help the student with his or her current difﬁculties, redirecting the student’s intentions only if he or she is not using an effective strategy for task completion. The teacher should watch for student clues as to when and how much assistance is needed. The assistance should be pro- vided immediately to help the student perform the task, but also should be removed gradually as the stu- dent demonstrates task mastery. The following lesson framework may help teachers provide scaffolding for their students. When present- ing a new or difﬁcult concept to students, it is likely that they will need more assistance at ﬁrst. Therefore, the teacher may model how to perform the task while the students observe. Next, the teacher and students work together on the task. While the teacher demon- strates the task on the board, students may be per- forming the task on a handout while seated at their desks. The teacher may ask the students questions or prompt them to contribute to the class discussion on the task. At this point, the students gradually are tak- ing some responsibility for their learning. When ready, students will work with a partner or a small cooperative group to perform the task. The teacher monitors the group work while gradually turning over more responsibility for the task performance to the students. Finally, the teacher has the students perform the task individually. The four stages of (a) teacher modeling, (b) teacher and students working together, © students working with a partner or small group, and (d) students working independently do not all have to occur in one class period. It could take 1 day to several weeks or more depending on how long a student needs to master the task. Scaffolding is a challenging but beneﬁcial process in instruction. One of the biggest challenges of effec- tive scaffolding is that it is time consuming, particu- larly for one teacher in a classroom. Another is that judging the zone of proximal development for each student may be difﬁcult (i.e., ﬁnding the area where a student needs help, but making sure that area is not beyond the student’s abilities). A third challenge is knowing student needs, interests, and abilities in order for the teacher to provide appropriate modeling. A fourth challenge may be making sure that the teacher gradually begins to fade or withdraw assistance as the student begins to demonstrate task mastery. Despite the challenges, scaffolding can be beneﬁcial to stu- dents. It may provide a better chance of a student mastering the intended task. Second, the structured nature of scaffolding may ensure more time on task and efﬁciency in task performance. Third, scaffolding engages the student, motivates him or her to learn, and reduces student frustration. Fourth, it provides individualized and possibly differentiated instruction. Martha J. Larkin See also Constructivism; Effective Teaching, Characteristics of; Teaching Strategies; Zone of Proximal Development Further Readings Hogan, K., & Pressley, M. (Eds.). (1997). Scaffolding student learning: Instructional approaches and issues. Cambridge, MA: Brookline Books. Larkin, M. J. (2001). Providing support for student independence through scaffolded instruction. Teaching Exceptional Children , 34 (1), 30–34. Larkin, M. J. (2002). Using scaffolded instruction to optimize learning (ERIC Digest No. 639). Arlington, VA: ERIC Clearinghouse on Disabilities and Gifted Education. (ERIC Document Reproduction Service No. EDO EC 02 17) Retrieved from http://eric.ed.gov/ERICDocs/data/ ericdocs2/content_storage_01/0000000b/80/2a/38/e0.pdf Lipscomb, L., Swanson, J., & West, A. (2004). Scaffolding. In M. Orey (Ed.), Emerging perspectives on learning, teaching, and technology. Retrieved from http:// www.coe.uga.edu/epltt/scaffolding.htm Verenikina, I. (1998). Understanding scaffolding and the ZPD in educational research. Retrieved from http:// www.aare.edu.au/03pap/ver03682.pdf S CHEMAS Although no universally agreed-upon deﬁnition of schemas exists, schemas are generally considered to be well-learned cognitive patterns of domain-speciﬁc information that are used as templates by individuals to help them explain, interpret, perceive, encode, and respond to complex tasks and experiences. Schemas also allow for predictions about what to expect in 864 Schemas future situations relevant to the particular schema. They create meaning from situations, data, and events by organizing and determining the patterns in com- plex sets of information. Schemas actually have a reciprocal relationship with data in that schemas may modify the meaning of information, but informa- tion or data may also lead to modiﬁcations in sche- mas. Both educators and counselors have interest in schemas because schemas help them understand how both informational and emotional learning occur. Various types of schemas have been postulated, such as schemas about other people (including role and person schemas), one’s self ( self-schemas), the sequence of various events ( script schemas ), context ( place or location schemas ),andthemeaningofdata ( information schemas). All of the various types of schemas facilitate the efﬁcien t understanding and inter- pretation of information by organizing and assigning meaning to that informati on. As a concrete example, suppose you heard someone talking about seemingly highly disparate pieces of information, such as arran- ging things into groupings; making decisions about color; dealing with tedium; deciding about capacity and facilities to employ; avoiding mistakes; timing of mechanisms; sorting types; determining what could not be dealt with by one’s current equipment, which neces- sitates outsourcing; settin g temperatures in such a way as to avoid catastrophe; dealing with voluminous out- put; enlisting aid from oth ers; and making measure- ments of necessary additives. This list might sound rather convoluted, meaningless, and difﬁcult to remem- ber unless you were ﬁrst told that the pieces of infor- mation all concerned ‘‘doing laundry.’’ During the initial learning process, deliberate con- struction of schemas requires the use of signiﬁcant amounts of working memory (WM) resources. Work- ing memory represents the brain’s capacity to tem- porarily hold limited amounts of information while manipulating that information. However, with practice and repetition, the use of schemas constructed during the learning process becomes virtually automatic. Thus, the development of schemas allows for a sub- stantial reduction in required WM resources as the schemas direct and guide individuals’ attention and focus. The result is often an increase in expertise or skill level within a particular knowledge domain. Schemas also provide an overall executive guidance system during high-level cognitive processing. Without this guidance (or without exter nal instruction), indivi- duals often default to weak pr oblem-solving strategies, such as trial and error and means-ends analysis. Strate- gies such as these can be both time consuming and inefﬁcient, and thus interfere with the construction of new schemas because of the workload imposed on WM resources. Schemas are stored in long-term memory (LTM), which is virtually unlimited in both its capacity and dura- tion and allows individuals to process, organize, and retrieve vast reservoirs of know

bundet

ledge. Once schemas are formed and stored in LTM, working memory is freed up to process, interpret, and ultimately store new schemas into LTM. By and large, when schemas are needed in WM, they are dealt with as a single piece of information, although they contain a rich array of data. Therefore, many educators consider schema formation to be an important focus for instructional design. The concept of schema acquisition has signiﬁcant applicability to instructional design. Often, it is recom- mended that one assess students’ current knowledge and activate relevant schemas prior to teaching new information. Analogies, metaphors, and comparisons can be useful in this regard. Schema acquisition is also facilitated by providing al ready completed, worked-out examplespriortoaskingstudentstoworkoutentirely new problems all on their own. Consideration should also be given to the cultures from which students have come, in that culture provides a backdrop and context for their existing schemas. Thus, asking a student to learn new material in a highl y individualized, competi- tive atmosphere might impede schema development for a student who comes from a culture that empha- sizes cooperation, mutuality, and group membership. Optimally, one wishes to develop schemas that have maximal ﬂexibility and applicability to a wide range of situations and contexts. Therefore, it is often recommended that one attempt to facilitate the devel- opment of schemas that are not overly constricted to speciﬁc situations. Such ﬂexible schemas may be acquired through exposure to similar knowledge that is applied to a range of contexts other than the one in which learning originally occurred. The development of more ﬂexible schemas also has relevance to the emotional r esponses of students. Thus, students’ inﬂexible self-schemas (which have also been referred to as early maladaptive schemas)tendtobe associated with stronger negative affect, which can readily interfere with l earning. By using a schema- based approach to students’ dysfunctional schemas and subsequent behaviors, the school counselor can create an atmosphere based on a nonpathological model and Schemas 865 focus on the ways in which a student has learned inﬂexible, intensely emotional responses to certain types of situations. Inﬂexible schemas can be examined and moderate, ﬂexible schemas developed to take their place. Such moderate, ﬂexible schemas are likely to be associated with less intense emotions and improved learning. Charles H. Elliott and Tam Chandler See also Cognitive Behavior Modiﬁcation; Cognitive View of Learning; Culture; Long-Term Memory; Working Memory Further Readings Chen, Z. (2006). Generalization and transfer of problem solving strategies. In A. V. Mittel (Ed.), Focus on educational psychology (pp. 217–234). New York: Nova Science. Elliott, C. H., & Lassen, M. K. (1997). A schema polarity model for case conceptualization, intervention, and research. Clinical Psychology: Science and Practice , 4 (1), 12–28. Kalyuga, S. (2006). Instructing and testing expertise: A cognitive load perspective. In A. V. Mittel (Ed.), Focus on educational psychology (pp. 53–104). New York: Nova Science. Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect. Educational Psychologist , 38 (1), 23–31. S CHOOL C OUNSELING School counseling, a crucial component to students’ achievement, is a comprehensive program that facili- tates students’ academic, ca reer, and personal/social development within the school setting. Professional school counselors have a minimum of a master’s degree in school counseling. In order to facilitate the development of all students, professional school coun- selors implement a wide range of therapeutic interven- tions, including classroom guidance lessons on topics such as anxiety management and bully prevention, group and individual counseling, career testing and planning, parent and teacher consultation, and advo- cacy for systems change. Research has shown that these school counseling services improve students’ academic success. School counseling is an important topic in educational psychology because it promotes students’ academic, career, and personal/social achieve- ment in the educational settings of elementary, middle, and high schools. This entry provides details of school counseling by (a) reviewing the history of school coun- seling; (b) explaining the American School Counseling Association’s National Model for school counseling; and © describing key components of a comprehensive school counseling program including guidance cur- riculum, counseling, career development, consulting, coordination of resource s, leadership and advocacy, promotion of a safe and respectful climate, accountabil- ity, management of legal and ethical issues, and profes- sional development. History The forerunner of modern sc hool counseling was voca- tional guidance, a preventiv e educational approach that taught students how to deal with life events. In 1908, Frank Parsons, known as the ‘‘Father of Guidance,’’ founded Boston’s Vocational Bureau, where he helped young people with career decisions. Concurrently, in 1907, Jesse B. Davis implemented weekly vocational and moral guidance lessons during English classes in Grand Rapids, Michigan, which led to a systematized guidance program in the public schools. The school guidance movement strengthened as Harvard Univer- sity began education courses for counselors in 1911; the National Vocational Guidance Association was established in 1913; and the Smith-Hughes Act of 1917 funded vocational edu cation in public schools. Over the 100-year history of school counseling, program focus and duties evolved in response to changing trends and needs. In the early 1900s, school counselors focused on scheduling student courses that would lead to careers needed in the Industrial Revolution. In the 1910s, psychometrics became another focus when guidance workers used the mili- tary’s Army Alpha and Army Beta intelligence tests to identify highly capable students. In the 1920s, sec- ondary school guidance personnel were trained simi- lar to college personnel because of limited training programs and thus acquired some administrative and disciplinary duties, similar to college deans of stu- dents. In the 1930s, school guidance personnel fol- lowed E. G. Williamson’s approach of enhancing normal adjustment by helping individuals to set goals and teaching them needed skills. In the 1940s, Carl Rogers’s nondirective emphasis of listening and accepting clients without judgment resulted in school 866 School Counseling counselors providing client-centered counseling to students, rather than just guidance. In the 1950s, after the Soviet Union launched its ﬁrst space satellite, Sputnik I, the United States funded the National Defense Educatio n Act (NDEA). As a result, school counselors focused on student career testing to channel students with high math and science abilities into college. In addition, NDEA funded elementary school counseling so that talented elementary students could be identiﬁed. The 1960s group encounter move- ment inﬂuenced school counselors to offer small group counseling. Concurrently, C . Gilbert Wrenn advocated that school counselors expand their focus to the devel- opmental needs of all students rather than just the top or bottom percentage. Hence, the focus shifted to the developmental guidance approach of promoting positive individual growth and preventing problems. Despite declining school enrollment and economic problems in the 1970s and 1980s, school counselors continued to expand the developmental guidance focus to students’ self-understanding and adjustment as well as career development from kindergarten through 12th grade. In 1998, Norman Gysbers and Patricia Hender- son published Developing and Managing Your School Guidance Program , which provided guidelines for a comprehensive developmen tal guidance program. In the 1990s, multiculturalism became prominent in coun- seling, inspiring school counselors to pay more atten- tion to the varying needs of students from different ethnic and socioeconomic groups. Spanning from the end of the 1990s through the new millennium, concerns of school violence, bully- ing, and crises emerged due to a rash of school shoot- ings and the September 11, 2001, terrorist attacks. School counselors responded by focusing on bully prevention and developing crisis counseling teams. Because of limited resources, the prominence of man- aged care within the health care system, and account- ability requirements of the 2001 No Child Left Behind policy, school counselors began to focus on accountability by providing data to prove that inter- ventions led to student success. American School Counseling Association National Model In order to stabilize the changing focus of school counseling, in 2003, the American School Counseling Association (ASCA) developed a national model to provide consistent, comprehensive guidelines for school counseling programs and professional school counselors’ duties that would promote success for all students throughout the country. A detailed descrip- tion of the ASCA National Model is provided in The ASCA National Model: A Framework for School Counseling Programs, Executive Summary . The ASCA National Model supports the school’s overall mission by promotin g academic achievement, career planning and persona l/social development. It serves as a framework to guide states, districts and individual schools in designing, developing, imple- menting and evaluating a comprehensive, develop- mental and systematic school counseling program. (ASCA, 2003, p. 1) This approach to school counseling programs beneﬁts students, parents, teachers, administrators, and the overall community. It is an integral part of each student’s achievement. It is systematically delivered to every student and is not just for high achievers or at-risk students. Systematic delivery of the ASCA National Model encompasses four interrelated components: foundation, delivery system, management systems, and account- ability. The foundation is composed of (a) beliefs and philosophy on which all personnel agree; and (b) a mis- sion statement that highlights the program’s purpose, which aligns with the school and district’s mission. The delivery system entails four methods needed to systematically deliver the school counseling program to all students. The ﬁrst method of the delivery system is guidance curriculum. The curriculum consists of structured classroom lessons that provide knowledge and skills at the appropriate developmental level for kindergarten through 12th-grade students. The second method is individual student planning in which pro- fessional school counselor s meet with individual stu- dents to help them identify goals and future plans. The third method is responsive services to meet individual students’ immediate needs th rough counseling, consul- tation, referral, peer mediation, or provision of infor- mation. The fourth method of the delivery system is systems support via administration and management of the total counseling program. The management system, the third component of the ASCA National Model, incorporates organiza- tional processes to make sure the counseling program is aligned with the school’s needs. Agreements about the school counseling program’s organization and School Counseling 867 goals are negotiated with school administrators. An advisory council of students, parents, teachers, coun- selors, administrators, and community members is established to review counseling program results and make recommendations. Data are used to decide what activities are needed to promote students’ academic, personal/social, and career achievement. Action plans are developed to achieve every desired competency and result. These action plans describe in detail the competencies addressed, activity components, data indicating the need for the activity, time line, respon- sible party, evaluation methods, and expected out- come. School counselors’ time should be carefully guardedsothat80 % of their time is spent on direct service contact with students. Their duties should be limited to program delivery rather than noncounseling activities. Calendars are developed and published so that students, parents, teachers, and administrators know the school counselor’s schedule of activities. Accountability is the ﬁnal component of the ASCA National Model. To hold professional school counse- lors accountable, data are used to link school counsel- ing activities to student achievement. Results reports ensure that programs were implemented, analyzed for effectiveness, and modiﬁed for activity and program improvement. Immediate, intermediate, and long- range reports are shared with stakeholders. School counselor performance standards are used to evaluate the school counselor and school counseling program. Finally, program audits are conducted to guide future action within the program. ASCA’s national standards for student academic, career, and personal/social development outline com- petencies that students wil l obtain or demonstrate as a result of the school counseling program. Regarding academic development, students will (a) acquire atti- tudes, knowledge, and skills for effective learning; (b) complete school with academic essentials needed for postsecondary options; and © understand the rela- tionship between academics and career as well as life in the community. Regarding career development, stu- dents will (a) acquire skills to investigate work options and self so that they can make informed career deci- sions; (b) utilize strategies to achieve future career goals with success and satisfaction; and © understand the relationship between pers onal characteristics, edu- cation, training, and career options. Regarding personal social development, students will (a) acquire knowl- edge, attitudes, and skills to understand and respect self and others; (b) learn to make decisions and establish and achieve goals; and © develop safety and survival skills. Professional school counselors must ensure that their programs help students accomplish each of these competencies. In order to implement school counseling programs as described above, school counselors’ duties and responsibilities must be delineated clearly. Although professional school counselors are team players, they cannot be fully effective when they perform noncoun- seling administrative or clerical activities such as developing the master school schedule, acting as test- ing coordinators, providing coverage for classrooms and detention rooms, disciplining students, or per- forming clerical duties. Rather, appropriate school counseling responsibilities include, but are not limited to, the following: • Implementing individual student academic programs • Analyzing and conveying cognitive, aptitude, and achievement tests • Counseling students with disruptive problems • Collaborating with teachers on stimulating guidance curriculum lessons • Analyzing grade point averages and achievement tests • Interpreting student data • Consulting with teachers on study hall management • Collaborating with the school principal on resolving student issues A detailed description of school counselors’ essen- tial functions follows. Key Components Developmental Classroom Guidance Curriculum Professional school counselors implement a develop- mental classroom guidance curriculum to all students in an effort to prevent problems in students. The curric- ulum addresses common concerns that are identiﬁed by needs assessments of students, faculty, and parents as well as national standards related to academic, personal/social, and career development. Professional school counselors present structured, planned lessons to a large group of students to meet students’ devel- opmental needs. Topics of typical guidance lessons include the following: • Academic skills support • Organizational, study, and test-taking skills 868 School Counseling • Education in understanding self and others • Coping strategies and crisis management skills • Peer relationships and effective social skills • Communication, problem solving, decision making, conﬂict resolution, and bullying prevention • Career awareness, exploration, and planning • Substance abuse education • Multicultural/diversity awareness Although the topics may remain the same across school levels, developmental approaches vary based on school level as toys and puppets may be used in elementary school whereas games and role-playing may be used in secondary schools. Counseling When students need more intensive help beyond classroom guidance lessons, school counselors pro- vide small group counseling and individual counsel- ing. During small group counseling, professional school counselors meet with two to eight students at a time to provide therapeutic intervention that meets participants’ individual and common goals. In ele- mentary schools, typical goals include improving social skills and general behavior, adjusting to family changes of divorce or death, or resolving underlying personal problems that interfere with academic suc- cess. In secondary schools, typical goals include stu- dent success skills and career decision making. Professional school counselors employ group counsel- ing skills of facilitating open communication, linking group members together, providing feedback on inter- actions, confronting disruptive behavior, and encour- aging positive interaction with other members. When providing individual counseling to a student, professional school counselors may follow a particu- lar counseling theory such as person-centered (Carl Rogers); Adlerian (Alfred Adler); reality therapy (William Glasser); or cognitive behavior therapy (Aaron Beck, Donald Meichenbaum, G. Terence Wilson). However, many professional school counse- lors are eclectic and use skills from various theories to meet an individual student’s needs. In secondary schools, solution-focused counseling (Steve de Shazer, John Murphy) provides a quick and simple approach of helping students focus on small changes and reasonable goals. The steps are as follows: 1. Establish a cooperative, ch ange-focused relationship. 2. Develop small, meaningful goals. 3. Encourage students to do more of what works and use their resources. 4. Change the ‘‘doing’’ by trying a different strategy or changing location, frequency, intensity, or sequence of behavior; and change the ‘‘viewing’’ by seeing it from a different perspective. 5. Evaluate and maintain progress. In elementary schools, play therapy or ‘‘counseling with toys’’ (Garry Landreth) allows students to express their feelings, thoughts, and behaviors and resolve con- ﬂicts through their natural medium of communication, play. Professional school counselors create a safe envi- ronment with specially selected toys and provide thera- peutic responses of tracking play behavior, reﬂecting feelings and content, returni ng responsibility, encourag- ing, building self-esteem, s etting therapeutic limits, facilitating understanding , and expanding the meaning of the child’s play. Numerous research studies have demonstrated that play therapy decreases children’s behavior problems and increases their mental health, which in turn helps them succeed at school. Career Development Career development and guidance in the schools is based on the goal that each student will complete high school prepared for a variety of workplace and post- secondary options, including 2- and 4-year colleges, technical schools, or military service. Career develop- ment begins in preschool and continues beyond high school. According to Edwin Herr and Stanley Cramer, children often make a tentative commitment to a voca- tion in the ﬁrst 6 years of school. Elementary school counselors focus on career awareness that transcends socioeconomic levels and gender roles. Common career awareness activities are career days, in which parents and community members present information on their careers; ﬁeld trips to local industries, banks, hospitals, stores, and so on; and videos of a wide range of careers. Middle school counselors focus on career exploration. They help students explore work options in light of their own strengths, weaknesses, interests, talents, and skills. Middle school students are taught how to use computerized career informa- tion delivery systems to conduct assessments and occupational searches and obtain occupational and educational information. High school counselors focus on career planning by linking students’ interests and skills with occupational and educational options. They School Counseling 869 provide students with information, emotional support, reality testing, and planning strategies. High school students beneﬁt from experiential activities such as attending a career fair, visiting a college campus, com- pleting ﬁnancial aid applicat ions, and participating in a career internship. High school counselors must also give speciﬁc care and information to students who are potential high school dropouts. Consultation Consultation is the process in which professional school counselors (consultants) assist teachers, par- ents, administrators, and community members (con- sultees) with problems related to a student and the system (client system). Typical consultation issues include teachers’ management of students’ classroom behavior and teaching strategies for students with dis- abilities, parents’ discipline of unruly children and motivating their children to do homework, administra- tors’ concerns regarding low-performing teachers or decreasing violence and prejudice at school, and com- munity members’ desire to mentor students or provide resources. Michael Dougherty described a generic model of consultation with four stages. During the entry stage, the school counselor (consultant) enters the relation- ship with the consultee by discussing roles, agreeing for action, and communicating conﬁdentiality. In the diagnosis stage, the consultant assesses the consultee characteristics that may have an impact on the prob- lem, the client characteristics, and the environment. Together, they clearly deﬁne the problem and develop measurable and realistic objectives. In the implemen- tation stage, the school counselor and consultee select acceptable strategies and identify resources within the system. The strategies are implemented, the consul- tant observes and gives feedback, and strategies are adapted and adjusted as needed. In the disengagement stage, the consultant and consultee evaluate the effec- tiveness of the strategies in resolving the problems. Finally, they terminate the consultation by discussing the ending, gradually reducing involvement, and con- ducting a summary conference. Coordination and Resource Management Professional school counselors coordinate and manage resources to meet various needs of students, teachers, and parents, such as the need for a mentor, clothing, classroom supplies, computers, and so on. Clearly, professional school counselors cannot directly meet all students’ needs alone. However, they can coordinate resources provided by parents, teachers, human service agencies, community members, and business partners. For example, the Parent Teacher Association may be able to organize a clothing closet, tutoring, and parent education classes. Business part- ners may be willing to provide career mentors and funding for new computers. Professional school counselors maintain resource lists and contacts with human service agencies such as after-school boys and girls clubs, a domestic violence shelter, public health clinics, and so on. Training peer mediators to help resolve student conﬂicts is also an effective use of resources. In doing so, discipline referrals are decreased and a positive attitude toward school is increased. Leadership and Advocacy Professional school counselors are collaborative leaders and advocates within the school system. As part of the leadership team, professional school coun- selors actively participate on school improvement teams to create a system in which all students can experience academic, career, and personal/social suc- cess. Professional school counselors are also leaders by working in partnership with principals and other key stakeholders, creating a positive school climate, conducting staff development for teachers, developing high aspirations in students, and using technology to track data. Professional school counselors are advocates who question the status quo, challenge rules and regula- tions that deny student access, protest changes that hinder underrepresented groups, empower people who need strength, and promote needed changes in the sys- tem. They work to close the achievement gap for low socioeconomic students and minority students of color who have lacked support in achieving at the same aca- demic level as majority White students. They advocate for these students by removing barriers that prevent their achievement in rigorous academic courses such as calculus, chemistry, or Advanced Placement English. Professional school counselors also advocate for safer school environments for students and teachers perse- cuted and oppressed because of their sexual orientation. For example, they may sponso r a Gay/Straight Alliance student group. 870 School Counseling