Baker, J. A., Bridger, R., Evans, K.
Gifted Child Quarterly
National Association for Gifted Children (NAGC)
Vol. 42, No. 1, pp. 5-15
This article explores the problem of underachievement among gifted students. It discusses the factors involved and explains that most often these issues surface in late elementary school and middle school. The article discusses a detailed study of three models: individual, family and school. It explains how this population can be helped through intervention that combines these three models. Authored by Jean Baker, Robert Bridger, and Karen Evans.
This study explored individual, family, and school-related factors contributing to under-achievement among gifted preadolescents. We explored three simple models of factors contributing to underachievement as well as a model incorporating all three factors. The complex model provided the best fit for the data, yet each of the three simple models was significant, suggesting the importance of an ecological approach to problems of underachievement among gifted students. Intervention approaches deriving from each model were discussed.
Putting the Research to Use
Underachievement presents a complex problem for schools and families. This study suggests that families and schools should think broadly and comprehensively about underachievement among gifted preadolescents. School, family, and individual child characteristics influence the development of under-achievement in preadolescents and could each be the focus of intervention for this problem. Children may benefit from interventions to enhance their organizational skills, self-management, and self-control skills. Families may benefit from consultation regarding issues relating to underachieving students, family organization, and general parenting issues. The data also suggested that school issues, specifically teacher-student relationships and academic quality, should also be the focus of attention when students are underachieving. In general, encouraging child, family, and school systems to work together in a coordinated fashion may provide the most effective means of intervention for gifted underachieving preadolescents.
Underachievement among gifted students presents a perplexing challenge for educators and parents. Underachievement is defined by the gifted child's failure to perform academically at a level commensurate with his or her measured potential. Nationwide, it is estimated that 15%-40% of identified gifted students are "at-risk" for school failure or significant under-achievement (Seeley, 1993).
Currently, we have several models in the theoretical literature attempting to describe the origins of under-achievement among gifted students. Typically, individual, family, and school-related factors are discussed as contributing to problems of underachievement (e.g., Rimm, 1986). These popular conceptualizations often drive educational and psychological interventions. However, empirical testing which could inform decision-making is lacking. This study evaluated the goodness-of-fit of three "simple" models and the added benefit of considering a "complex" model to predict underachievement status among gifted preadolescents.
Individual Etiology Model
Based in the "medical model" tradition within psychology and education, an individual difference approach holds that the roots of underachievement lie within the child. From this perspective, underachievement is usually attributed to motivational or behavioral/emotional problems on the part of the student (e.g., Bruns, 1992).
A number of child-specific variables have been implicated in underachievement among gifted students. Among the most widely studied of these is self-concept. Gifted underachievers generally evidence poorer self-concepts than achieving gifted and regular education students, especially in the areas of academic and intellectual self-concept (e.g., Kanoy, Johnson, & Kanoy, 1980; Whitmore, 1980; Van Boxtel & Monks, 1992). However, the correlational nature of many of these studies makes it difficult to determine if poor self-concept is a precursor to or a result of underachievement and failure experiences (Whitmore, 1980).
A number of other personality and temperament differences between gifted achieving and gifted underachieving students have been noted (Gallagher, 1991). Gonzalez and Hayes (1988) cite studies describing underachieving students as more aggressive, judgmental, hostile, rationalizing, and less persistent than their achieving gifted peers. Underachievement is associated with a fear of failure and perfectionism (Rimm, 1988; Adderholt-Elliott, 1991), emotional sensitivity and stress (Freeman, 1994), and a diminished locus of control, that is, the belief in the ability to control the outcomes of personal efforts or to effect change in the environment (Laffoon, Jenkins-Friedman, & Tollefson, 1989). Underachievement in children, including the gifted, usually accompanies specific mental health problems, such as depression (Whitmore, 1980; Webb, Meckstroth, & Tolan, 1982).
Weaknesses in academic and "school survival" skills also contribute to problems of Underachievement (e.g., Krouse & Krouse, 1981). Gifted underachievers often have deficits in study and organizational skills (Rimm, 1988). Motivational problems are frequently cited in this regard. Students often report poor attitudes toward schooling and low motivation to cooperate with conventional schooling activities (Seeley, 1993). Thus, school readiness may be implicated in underachievement among gifted students.
This is a vast body of literature, and it is difficult to generalize across studies with differing samples and methodologies. Also, there is not complete agreement in the literature regarding the nature or severity of these problems among gifted underachievers. However, this body of work is consensual in considering intrapsychic or individual difference factors as contributing to underachievement problems among gifted students.
Family Etiology Model
In contrast to traditional intrapsychic theories, a family systems perspective understands behavior as influenced by relationships and interactions with others within the family context (Fine, 1992). From this perspective, underachievement derives from the family's inability to adequately support the child or is considered one indicator of family distress (e.g., Rimm, 1986).
Through the last decade there has been an increased interest in using family systems concepts to study underachievement among gifted students (Green, Fine, & Tollefson, 1988). One of these is family structure, or the covert and overt rules that govern family interactions. The family environments of gifted underachievers appear to have a structure characterized by disorganization and unclear guidelines about behavior, including academic performance (Rimm & Lowe, 1988). Although families of underachievers are as active and busy as the families of achievers, they seem to have more trouble managing their schedules (Rimm & Lowe). Thus, the lack of a supportively structured environment in which achievement can flourish, as well as parental modeling of poor organization, may contribute to underachievement from this perspective.
The atmosphere or emotional climate within the home of gifted underachievers has also been the focus of study. Families of gifted underachievers demonstrate a lack of cohesion or more emotional distance between members (Albert, 1978) and a lack of parental agreement relative to parenting (Rimm, 1986). The parents' expectations for conduct are unpredictable, and differences between the parents in expectations and behavior exist (Rimm & Lowe, 1988). Independence is another aspect of the family climate that does not appear to be facilitated within families of gifted underachievers. Self-confidence is not fostered, and a supportive environment that encourages risk-taking is not apparent (Gurman, 1970). Parent-child relations are characterized by a greater degree of opposition, and family relations are more conflictual (Colangelo & Dettmann, 1983; Olszewski, Kulieke, & Buescher, 1987). The families do not facilitate self-expressiveness, and the child finds it difficult to communicate with his or her parents. Raph, Goldberg, and Passow (1966) stated that underachievement is the child's unconscious and indirect means of satisfying hostility that he or she is unable to directly express within the family. There has been inconsistency in the research as Green and colleagues (1988) did not find support for the theory that underachievement is a symptom of family conflict and disturbance. Although these data are equivocal, differences in the family climates of gifted underachievers and gifted achievers seem suggested by this literature.
Families of gifted underachievers are also characterized by mixed messages regarding the value of achievement and the modeling of achievement behavior (Rimm & Lowe, 1988). Although families of gifted achievers and gifted underachievers both stress achievement and its value, families of gifted underachievers seem deficient in "practicing what they preach."
The general consensus within this literature is that families of achievers and underachievers exhibit differences on variables important to supporting children's behavior. The emphasis in this area is on family influences on children's achievement, with underachievement resulting from disturbances within the child's family system.
School Etiology Model
An ecosystemic perspective expands family systems thinking to include interactions within other contexts as potential determinants of behavior (Bronfenbrenner, 1979). After the family, the other major context for children's development is the school. In a large scale study examining school records, Peterson and Colangelo (1996) concluded that the school experiences of gifted underachievers appear impoverished compared to those of gifted achievers on both academic and non-academic variables such as absences and extra-curricular activities. This suggests that schooling may be a contextually different experience for achieving and non-achieving gifted students. From a school-based perspective, a mismatch between the needs of the child and the environment of the school are implicated in problems of underachievement among gifted students (e.g., Whitmore, 1989).
Within the literature some attention has been paid to boredom and inadequate academic programming (Supplee, 1989; Whitmore, 1980, 1989). Boredom is felt to be a cardinal feature of many gifted students' school day (Robinson & Robinson, 1982) and may lead to a host of negative sequela for children, including underachievement (Rimm, 1988). A mismatch between pedagogical approaches and children's learning styles may also promote underachievement if students are not provided with encouragement or viable ways of expressing their talents (Redding, 1990; Whitmore, 1989). Inflexible curricular requirements, age-grouping, and lack of acceleration opportunities are also cited in reference to underachievement (Fehrenbach, 1993). These schooling variables impact student motivation. From this perspective gifted students become disaffected from school and withdraw their participation, resulting in underachievement.
There is an extensive body of literature examining teacher expectations and the effects of the gifted label on students. Kolb and Jussim (1994) describe several ways in which teachers unintentionally influence the achievement patterns of their students. Teachers may fail to recognize talent, especially among minority students or those with behavior problems, and hold low performance expectations for children. Alternately, teachers may hold stereotypes about gifted students and have unreasonably high standards of performance, thereby eliciting withdrawal and failure to perform by gifted students who perceive undue pressure (Pendarvis, 1990). The practice of labeling students as "gifted" may reinforce these stereotypical perceptions by teachers (Robinson, 1986, 1989). Because of these expectations, teachers do not always acknowledge good work from gifted students. They generally grade them harder than they grade non-gifted children and expect gifted children to help tutor those in the class who need help (Clinkenbeard, 1991). Underachievement may accompany a student's recognition that "giftedness" alters the social environment at school in unfavorable ways.
A gifted student's social environment includes peers who may also influence achievement status. Although "being smart" is generally an asset in elementary school, the developmental thrust toward peer conformity in middle and high school may cause gifted students to underachieve in order to deflect attention away from their uniqueness (Brown & Steinberg, 1990; Clinkenbeard, 1991). When gifted children underachieve, peer relationships take on a higher priority than learning, and these students may be positively reinforced for underachievement by the peer group (Compton, 1982; Rimm, 1989). Identification with underachieving peers may be especially common in girls, who are expected to be more social and have different achievement pressure than boys (Dowdall & Colangelo, 1982; Kerr, 1985).
As with the other models, many factors within the school environment have been the focus of study. The unifying theme in this diverse literature is that schools may fail to adequately support gifted students' potential, thus contributing to underachievement within this group.
Emergence of Underachievement
Problems related to gifted underachievement often emerge in late elementary and middle school. From a social developmental perspective, children at this age are able to make comparisons with others and to evaluate themselves through a social comparison process (Selman, 1980). A heightened awareness of the prevailing norms within peer groups and a desire to conform to group standards may make some gifted students deflect attention away from their gifted status through underachievement (Brown & Steinberg, 1990). From an academic skills perspective, later elementary and middle school may present specialized demands (such as time management, study skills, systematic problem solving rather than rote memorization, etc.) that are underdeveloped among students who have been unchallenged and have experienced seemingly effortless academic success in the early elementary grades (Baker, 1996). For these reasons, gifted preadolescents and early adolescents are frequently referred to clinics for problems related to underachievement. Understanding more about underachievement in this age group is warranted because of the frequency with which this group seeks intervention.
Purpose of the Study
Each of these models, individual, family, and school, suggest ways of understanding and intervening in problems of underachievement. However, intervention efforts will be strengthened by an understanding of the relative merit of each of these models of under-achievement. We conducted an analysis of three simple models of factors contributing to underachievement as well as a complex model incorporating all three factors using data from a study of gifted preadolescents and their families. Because of the attention directed toward each of the popular models in the literature, it was expected that each of them would attain statistical significance. Furthermore, it was predicted that a combined model, reflecting a more ecological, systemic approach to underachievement, would also predict achievement status and might afford a more comprehensive approach to understanding this problem.
Participants in this study were 26 gifted underachievers and 30 gifted students achieving well in grades 4-8. All students were identified as gifted in their local school districts using achievement and ability test data (scores above the 90th percentile on standardized tests). The underachieving children were at-risk of academic failure or removal from gifted programs because of poor academic performance. Students were referred by their parents to a research/clinical program pertaining to gifted preadolescents at a university-based psychoeducational clinic in a southeastern state. Parents were made aware of this research opportunity through fliers distributed by the schools and through local parent organizations for the gifted.
The children ranged in age from 9 to 14, with an average age of 12. There were 34 boys and 20 girls. Students were primarily first-borns (65%) and second-borns (25%). Only two children (both in the underachieving group) had existing psychiatric diagnoses (both Attention-Deficit/Hyperactivity Disorder), although parents of three students in the underachieving group and four in the achieving group reported that their children had some "emotional" problems (primarily related to post-divorce adjustment). Only one was seeing a professional counselor. The sample was representative of children likely to be seen in a university-based clinic setting: all students were Caucasian, 82% of families were married, 82% of mothers and 80% of fathers reported at least some college education (26% of mothers and 43% of fathers held a graduate degree). Nine children (five in the underachieving group and four in the achieving group) had experienced a parental divorce.
The gifted achieving group and the gifted underachieving group were comparable on most demographic and socioeconomic status variables. No differences were detected between groups on parents' educational level, marital status, children's birth order, gender, grade, or concurrent mental health or medical problems. There was an age difference between the groups with the underachieving group being slightly older (underachieving group: M = 12.20, SD = 1.40; achieving group: M = 11.13, SD = -1.58; t (54) = 2.66, p < .01).
Students and one of their parents (most typically mothers) completed a battery of psychological and educational tests. The battery was constructed to include the variables discussed above as contributing to underachievement. Reported here are only the measures included in these data analyses. Because of our sample size, only three variables were included in each of the models tested. Again, these variables were included based on the extant literature which associated them with underachievement in this age group.
Individual factors. The measures of individual factors included: parent report of emotional and behavioral problems from the Parent Rating Scales of the Behavioral Assessment System for Children (BASC-PRS; Reynolds & Kamphaus, 1992). The BASC-PRS is a standardized behavior rating scale which assesses problem and adaptive behavior in children. We used the Behavior Symptoms Index (BSI), a summary indicator of problem behavior, which reflects the overall level of externalizing (acting-out) and internalizing (worry/depression) problem behavior seen by parents. The second measure of individual behavior was drawn from the BASC Self-Report of Personality (BASC-SRP) completed by children. We summed scores from the Self-Esteem and Self-Reliance subscales as a self-report of personal competencies. All of the BASC scales evidence good psychometric characteristics (with internal consistency reliabilities above .92 for the BSI and ranging from .61-.84 for the personal adjustment subscales), good test-retest reliabilities, and adequate concurrent validity data reported in the manual. The third measure of individual behavior was a parent report of children's study skills and organization from an informal questionnaire. Parents were asked to rate items such as "Does your child have adequate time management skills when studying?" and "Does your child have adequate skills when completing written assignments?" on a True/False scale. These five items were summed to obtain the study/organizational skills measure. Thus, behavior/emotional problems, personal competencies, and study/organizational skills were the three variables considered for the individual model.
Family factors. Measures of family factors include an assessment of the family's emotional climate, using the combined Cohesion, Expressiveness, and Conflict subscales from the Family Environment Scale (FES; Moos, 1979) completed by the parent, and an assessment of family organization using the combined Organization and Control subscales of the FES. The FES is a widely used measure of the social-environmental characteristics of families. The subscales used in this study evidence adequate internal consistency and test-retest reliability (reliability coefficients ranging from .67 to .86) and validity data reported in the manual. Parents' self-assessment of the adequacy of their parenting skills was the third family variable considered. Parents rated their skills using a four-point Likert-type scale on an informal questionnaire consisting of items such as "Do you feel that you have adequate knowledge about parenting in general?" and "Do you feel that you have the skills you need to parent this child?" The score on these eight questions were summed to form the parenting skill assessment. Thus, the three family variables used were family emotional climate, family organization, and parenting skills.
School factors. Measures of school factors included parents' ratings of the adequacy of several facets of schooling for each year of the child's school career. Unfortunately, independent data from schools were not available for this study. Parents rated their perception of the adequacy of the curricular/academic programming, the teacher-student relationship, and peer relationships on a four-point Likert-type scale for each year of the child's school career. Scores within each domain were summed across each year of schooling that the child had been identified as gifted. Figure 1 summarizes the models tested in this study.
Logistic regression, using the SAS data analysis package and a traditional alpha level of .05, was used to evaluate the adequacy of each of the proposed models to predict group membership (underachieving or achieving status). Logistic regression is a multivariate technique that analyzes nominal-level variables in a regression format and has less restrictive conditions than does predictive discriminant analysis (Morgan & Teachman, 1988). Logistic regression incorporates a log-linear model which is used when the dependent variable in a regression equation is categorical or nominal in nature (see Hosmer & Lemeshow, 1989, for an in-depth review of the procedure). Because of the significant difference in age between the groups, age served as a covariate and was entered first in each of the regression models. Our sample size ensures adequate power to test a medium effect size in the data analyses (Green, 1991).
Descriptive statistics for each of the study variables are presented in Table 1. Pearson product moment correlations were calculated for each of the variables and are presented in Table 2. Although some of the correlations between measures were significant, collinearity is not generally felt to be a problem in regression analyses unless the coefficients are above .70 (Pedhazur, 1982). (Please refer to original article for tables.)
Individual Etiology Model
For this model, achievement status was regressed on age, emotional/behavior problems, personal competencies, and study/organizational skills. The model was significant (Beta2 = 15.541, df = 4, p < .004) and correctly classified 78% of the cases. After controlling for age, which was a non-significant predictor, only the study/organizational skills variable contributed significantly to the model (Chi = .43, p = .05). Neither behavior problems (Chi = .26, p = .20) nor personal competencies (Chi = -.02, p = .89) made significant contributions.
Family Etiology Model
For this model, age was used as the covariate and the three predictors were family emotional climate, family organization, and parenting skills. The model obtained significance (Beta2 = 10.330, df = 4, p < .01) and correctly classified 78% of the cases. None of the independent variables obtained significance in this analysis; however, age approached significance (Chi = .37, p = .09). Of the predictors, only parenting skills approached significance (Chi = .37, p = .09); neither family organization (Chi = .10, p = .57) nor family emotional climate (Chi = .28, p = .14) made statistically significant contributions.
School Etiology Model
For this model, achievement status was regressed on age, academic quality, teacher-student relationships, and peer relationships as reported by the parent. The model was significant (Beta2 = 10.527, df = 4, p < .03) and classified 73% of the cases correctly. After controlling for age, two of the predictors approached, but did not obtain, statistical significance. The quality of teacher-student relationships approached significance (Chi = -46, p = .08), as did academic quality (Chi = .45, p = .06). Peer relationships at school did not contribute significantly to the model (Chi = -.14, p = .46).
We constructed a model using the best predictors from each of the single models to evaluate the relative merit of a multi-faceted conceptualization of under-achievement in this age group. For this model we regressed achievement status on age, child organizational/study skills (from the individual model), parenting skills (from the family model), and academic quality (from the school model). This model was significant (Beta2 = 23.176, df = 4, p < .0001) and correctly classified 86% of the cases. After controlling for age, all of the predictors contributed significantly to this model. Organizational/study skills (Chi = .88, p = .003), parenting skills (Chi = -.50, p = .04), and academic quality (Chi = .66, p = .02) were found to be related significantly to underachievement status.
Understanding the factors contributing to underachievement can help educators and psychologists plan more effective interventions for gifted students. In this study, models derived from the theoretical and empirical literature describing individual, family, and school approaches to underachievement were tested. Each of the models was significant, suggesting that individual, family, and school factors contribute to underachievement. These findings confirm the attention directed toward each of these variables in the theoretical and applied literature on underachievement. However, the findings from the combined model, which incorporated variables from each of these systems (child, family, and school), were the most robust of the models tested. Each of these components seems important to consider when conceptualizing and remediating achievement problems in this age group. These results support the movement in both counseling (e.g., Rimm, 1986) and education (e.g., Baum, Renzulli, & Hebert, 1995) toward using multifaceted intervention approaches when addressing this problem.
The results from each of the individual models considered in this study expand our understanding of problems of underachievement. Of the single models, the individual etiology model was the most robust. The results suggest that underachieving children have deficits both in behavioral control and in organizational skills that could be the focus of direct intervention. Deficits in both strategic problem-solving (Muir-Broaddus, 1995) and in coping skills (Gallagher, 1991) have been noted among gifted underachievers. These students may benefit from instruction in study skills (e.g., Crittenden, Kaplan, & Heim, 1984), from behavioral and cognitive-behavioral strategies to improve self-management and self-control (e.g., Heacox, 1991; Hughes & Hall, 1989), and from interventions to improve coping skills (Gallagher). It is important to note, however, that we did not address students' motivation to use these strategies or to invest in self-improvement within our analyses. The reversal of achievement difficulties is not accomplished by simple skills training (Rimm, 1986). Recent work suggests the benefit of targeting these skills within the context of high interest, meaningful, self-directed activities (Baum et al., 1995).
It was somewhat surprising that internalizing or "emotional" problems, such as anxiety or depression, did not play a more active role within the individual model. The clinical literature suggests that gifted underachievers present frequently with these internalizing problems (e.g., Webb et al., 1982). This may be an artifact of the sample. Externalizing problems are much more likely to result in children being referred for mental health services (Achenbach, 1982). Thus, there may have been an underrepresentation of "emotional" problems within this sample of parent-referred children. Random selection or educating teachers and parents to notice and refer children who seem unusually quiet or withdrawn may correct this potential selection bias in future studies.
Family variables also predicted underachievement status in this study, suggesting the importance of involving families in reversing underachievement. Parent consultation regarding behavior management and general parenting issues is a commonly accepted remediation strategy for underachievement (e.g., Rimm, 1986). Adequate support from parents is a critical variable in the school success of all children (Christenson, Rounds, & Gorney, 1992). Parents of underachieving students may not have the skills to support children adequately given their unusual academic and developmental trajectories. The data were not sensitive enough to enable us to discuss particular skills that differentiated parents of achievers from parents of underachievers. Further work is needed in order to more fully understand parenting issues. However, parent education and consultation to facilitate the academic success of gifted underachievers seem warranted by these findings.
The finding that the family systems variables of structure and emotional climate were not predictive of achievement status was unexpected given the previous research in this area. This finding is probably best attributed to the sample characteristics. These families were relatively affluent, educated, intact families who may not have exhibited the range of family characteristics seen in more heterogeneous samples. Also, the self-selection procedure may have skewed the sample toward parents who were highly motivated and had the resources to seek help and, thus, were potentially better adjusted than non-responders. Alternately, the measures we used may not have captured the range of behaviors and diversity within our families. Further work is needed in this area to clarify the role of family systems constructs in problems of underachievement.
Underachievement is also associated with variables controlled by the school. Interventions at school may afford the most practical and efficient possibility to effect change for underachieving students. The two avenues suggested in this study were teacher-student relationships and academic quality. Poor teacher-student relationships, perhaps as a result of the children's challenging behavior, approached significance in the school model. This suggests the benefit of classroom consultation and addressing issues of teachers' tolerance of behavior. Gifted students with behavioral problems may be especially vulnerable to these relationship problems. Teachers' stereotypes of giftedness often incorporate ideal, or at least mature, behavioral adjustment (Kolb & Jussim, 1994; Whitmore, 1980). Teachers are typically more punishing of students whom they perceive as failing due to lack of effort rather than to lack of ability (Weiner, 1994). Thus, teachers may not tolerate challenging behavior or failure from gifted students and may distance themselves from these children. The importance of relationships with teachers in mediating children's satisfaction with school and achievement is gaining recognition (e.g., Noddings, 1992). High achieving students, and those "recovered" from underachievement (Emerick, 1992), cite the importance of teachers' support in their school success. Interventions that foster social affiliation with adults at school (such as mentorships, e.g., Baum et al., 1995) and that promote encouragement and positive social interactions with teachers may be helpful to address problems of underachievement for preadolescents.
Similarly, poor academic quality approached, but did not quite achieve, statistical significance in the school model, suggesting the need for further research on this variable. This finding suggests that the goodness-of-fit between children's abilities, achievement, and learning style and the academic demands of the classroom is important to explore with underachieving students. It also suggests that classroom teachers need additional support in adapting curricula to meet the needs of gifted students in their classrooms. Actively adjusting the academic environment to promote success and achievement among gifted students is not a new idea (e.g., Fehrenbach, 1993; Whitmore, 1989), but it may be worth emphasizing again in reference to underachieving students.
In our combined model, underachievement was predicted by all individual, family, and school variables, reflecting the complexity of this problem for children, families, and schools. This was the most robust model of those tested in this study. These data support an ecological perspective on this problem and the need for schools and families to think broadly and systemically in order to create solutions for children.
Before generalizing from these data or using them to develop theory in this area, we suggest a replication of these findings. Our relatively small sample size and the university-based data collection method may have introduced a bias into our sample. Because of our sample size we did not have adequate power to explore all of the variables that seemed important from the individual models in our combined model. Additionally, the moderate correlations between the variables within the individual models may have masked the complexity of the problem in those analyses. Larger samples and more independence between the measures may address this problem in future studies. Also, we did not find that family and school variables accounted for much predictive ability in our models, suggesting the need for more sensitive measures in these areas. Direct assessment of school information and other family assessment measures may strengthen our findings in these areas. Our sample did not reflect underachievement which results from internalizing problems, such as anxiety or perfectionism, also suggesting that our sample may not have been reflective of the range of problems associated with underachievement among gifted students. Finally, longitudinal research is needed to accurately address the origins of underachievement. Therefore, it seems important that these findings be replicated prior to generalizing from them.
However, the findings from this study suggest the benefit of ecological and systemic approaches to conceptualizing and intervening with gifted underachievers. Family, school, and individual factors each contribute to problems of underachievement. Interventions targeted at all three of these systems may best support success and adjustment for gifted underachieving preadolescents.
Achenbach, T. M. (1982). Developmental psychopathology (2nd ed.). New York: Wiley & Sons.
Adderholt-Elliott, M. (1991). Perfectionism and the gifted child. In M. Bireley & J. Genshaft (Eds.), Understanding the gifted adolescent. New York: Teachers College Press.
Albert, R. (1978). Observations and suggestions regarding giftedness, familial influence and the achievement of eminence. Gifted Child Quarterly, 28, 201-211.
Baker, J. A. (1996). Everyday stressors of academically gifted adolescents. Journal of Secondary Gifted Education, 7, 356-368.
Baum, S. M., Renzulli, J. S., & Hebert, T. P. (1995). Reversing under-achievement: Creative productivity as a systematic intervention. Gifted Child Quarterly, 39, 224-235.
Bronfenbrenner, U. (1979). The ecology of human development. Boston: Harvard University Press.
Brown, B. B., & Steinberg, L. (1990). Skirting the "brain-nerd" connection: Academic achievement and social acceptance. Education Digest, 55, 57-60.
Bruns, J. (1992). They can but they don't. New York: Viking.
Christenson, S. L., Rounds, T., & Gorney, D. (1992). Family factors and student achievement: An avenue to increase students' success. School Psychology Quarterly, 7, 178-206.
Clinkenbeard, P. R. (1991). Unfair expectations: A pilot study of middle school students' comparisons of gifted and regular classes. Journal for the Education of the Gifted, 15, 56-63.
Colangelo, N., & Dettmann, D. (1983). A review of research on parents and families of gifted children. Exceptional Children, 50, 20-27.
Compton, M. F. (1982). The gifted underachiever in middle school. Roeper Review, 4, 23-25.
Crittenden, M. R., Kaplan, M. H., & Heim, J. K. (1984). Developing effective study skills and self-confidence in academically able young adolescents. Gifted Child Quarterly, 28, 25-30.
Dowdall, C. B., & Colangelo, N. (1982). Understanding gifted students: Review and implications. Gifted Child Quarterly, 26, 179-183.
Emerick, L. J. (1992). Academic underachievement among the gifted: Students' perceptions of factors that reverse the pattern. Gifted Child Quarterly, 36, 140-146.
Fehrenbach, C. R. (1993). Underachieving gifted students: Intervention programs that work. Roeper Review, 16, 88-90.
Fine, M. (1992). A systems-ecological perspective on home-school intervention. In M.J. Fine & C. Carlson (Eds.), Handbook of family-school interventions: A systems perspective (pp. 1-17). Boston: Allyn & Bacon.
Freeman, J. (1994). Some emotional aspects of being gifted. Journal for the Education of the Gifted, 17, 180-197.
Gallagher, J. J. (1991). Personal patterns of underachievement. Special Issue; Update: Underserved gifted. Journal for the Education of the Gifted, 14, 221-233.
Gonzalez, J., & Haves, A. (1988). Psychosocial aspects of the development of gifted underachievers: review and implications. The Exceptional Child, 35(1), 39-51.
Green, S. B. (1991). How many subjects does it take to do a regression analysis? Multivariate Behavioral Research, 26, 499-510.
Green, K., Fine, M., & Tollefson, N. (1988). Family systems characteristics and underachieving gifted adolescent males. Gifted Child Quarterly, 32, 267-272.
Gurman, A. (1970). The role of the family in underachievement. Journal of School Psychology, 8, 48-53.
Heacox, D. (1991). Up from underachievement. Minneapolis, MN: Free Spirit Publishing.
Hosmer, D. W., & Lemeshow, S. (1989). Applied logistic regression. New York: Wiley & Sons.
Hughes, J., & Hall, R. (1989). Cognitive-behavioral psychology in the schools: A comprehensive handbook. New York: Guilford Press.
Kanoy, R. C., Ill, Johnson, B. W., & Kanoy, K. W. (1980). Locus of control and self-concept in achieving and underachieving bright elementary students. Psychology in the Schools, 17, 395-399.
Kerr, B. A. (1985). Smart girls, gifted women. Columbus, OH: Ohio Psychology Press.
Kolb, K. J., & Jussim, L. (1994). Teacher expectations and underachieving gifted children. Roeper Review, 17, 26-30.
Krouse, J., & Krouse, H. (1981). Toward a multimodal theory of academic underachievement. Educational Psychologist, 16, 151-164.
Laffoon, K. S., Jenkins-Friedman, R., & Tollefson, N. (1989). Causal attributions of underachieving gifted, achieving gifted, and nongifted students. Journal for the Education of the Gifted,13(1), 4-21.
Moos, R. H. (1979). The Family Environment Scale. Palo Alto, CA: Consulting Psychologist Press.
Morgan, S. P., & Teachman, J. D. (1988). Logistic regression: Description, examples, and comparisons. Journal of Marriage and the Family, 50, 929-936.
Muir-Broaddus, J. E. (1995). Gifted underachievers: Insights from the characteristics of strategic functioning associated with giftedness and achievement. Learning & Individual Differences, 7, 189-206.
Noddings, N. (1992). The challenge to care in schools: An alternative approach to education. New York: Teachers College Press.
Olszewski, P., Kulieke, M., & Buescher, T. (1987). The influence of the family environment on the development of talent: A literature review. Journal for the Education of the Gifted, 11, 6-28.
Pedhazur, E. J. (1982). Multiple regression in behavioral research: Prediction and explanation (2nd ed.). New York: Holt, Rinehart & Winston.
Pendarvis, E. D. (1990). The abilities of gifted children. Englewood Cliffs, NJ: Prentice Hall.
Peterson, J. S., & Colangelo, N. (1996). Gifted achievers and underachievers: A comparison of patterns found in school files. Journal of Counseling & Development, 74, 339-407.
Raph, J., Goldberg, M., & Passow, A. (1966). Bright underachievers. New York: Teachers College Press.
Redding, R.E. (1990). Learning preferences and skill patterns among underachieving gifted adolescents. Gifted Child Quarterly, 34, 72-75.
Reynolds, C., & Kamphaus, R. (1992). The Behavior Assessment System for Children. Minneapolis, MN: American Guidance Service.
Rimm, S.B. (1986). Underachievement syndrome: causes and cures. Watertown, WI: Apple Publishing.
Rimm, S. (1988). Identifying underachievement: The characteristics approach. Gifted Child Today, 11(1), 50-54.
Rimm, S. (1989). Disappearance of underachievement. Gifted Child Today, 12, 36-39.
Rimm, S., & Lowe, B. (1988). Family environments of underachieving gifted students. Gifted Child Quarterly, 32(4), 353-359.
Robinson, A. (1986). Brave new directions: Needed research on the labeling of gifted children. Gifted Child Quarterly, 30, 11-14.
Robinson, A. (1989). Gifted: The two-faced label. Gifted Child Today, 12(1), 34-36.
Robinson, N. M., & Robinson, H. B. (1982). The optimal match: Devising the best compromise for the highly gifted student. New Directions for Child Development, 17, 79-94.
Seeley, K. R. (1993). Gifted students at risk. In L.K. Silverman (Ed.), Counseling the gifted and talented (pp. 263-276). Denver: Love Publishing.
Selman, R. L. (1980). The growth of interpersonal understanding. New York: Academic Press.
Supplee, P. L. (1989). Students at risk: The gifted underachiever. Roeper Review, 11, 163-166.
Van Boxtel, B. W., & Monks, F. J. (1992). General, social and academic self-concepts of gifted adolescents. Journal of Youth and Adolescence, 21(2), 169-187.
Webb, J.T., Meckstroth, E.A., & Tolan, M. (1982). Guiding the gifted child: A practical source for parents and teachers. Columbus, OH: Ohio Psychology Press.
Whitmore, J.R. (1980). Giftedness, conflict, and underachievement. Melrose, MA: Allyn & Bacon.
Whitmore, J. R. (1989). Understanding a lack of motivation to excel. Gifted Child Quarterly, 30(2), 66-69.
Weiner, B. (1994). Ability versus effort revisited: The moral determinants of achievement evaluation and achievement as a moral system. Educational Psychologist, 29, 163-172.