Adjustment problems, self-efficacy, and psychological distress among Chinese gifted students in Hong Kong.

It is generally acknowledged that gifted and talented students, despite their generally high cognitive abilities, are not immune to social and emotional problems similar to those experienced by their nongifted age peers (see Blackburn & Erickson, 1986; Pfeiffer & Stocking, 2000; Piirto, 1992, Robinson & Noble, 1991; Webb, 1993). Broadly conceptualized, these problems could arise from gifted students' adjustment to being gifted, the quality of educational match, their family environments, and their personal characteristics (Neihart, 1999). Indeed, being gifted might render these students more vulnerable or more at-risk for particular social and emotional problems. In particular, gifted students' uneven or asynchronous development might contribute to their feeling of being different and a misfit among peers (Neihart; Tannenbaum, 1997). For the very highly gifted students, they might even feel more different from their nongifted-age peers, and find it difficult to belong to an appropriate peer group and feel accepted by peers within the desired group (Gross, 1989; Pfeiffer & Stocking). Apart from such interpersonal problems, gifted students might also experience great stress in meeting the high and sometimes unrealistic expectations of high achievement and performance from parents, teachers, significant others, and the students themselves (Freeman, 1998; Webb). Other equally salient adjustment problems might also arise from students' needs for learning and thinking at a pace and level not provided by their educational environments, and from their creativity, energy, intensity, aspirations, and multiple interests and talents (e.g., Genshaft, Greenbaum, & Borovsky, 1995; Gross; Karnes & Oehler-Sinnett, 1986; Swiatek, 1995; Winner, 1996).

Despite the diversity of adjustment problems, it is recognized that different gifted students, given their different personal characteristics, talents, and ways of coping with their giftedness, might experience these problems differently and to different degrees, rendering them more at risk for psychological distress than their nongifted peers. While the majority of academically or intellectually gifted students have been attested to be at least as well adjusted and perhaps better adjusted than their nongifted peers, and they generally do not experience depression, suicidal ideation, or anxiety more often than do nongifted peers (e.g., Baker, 1995; Derevensky & Coleman, 1989, Parker, 1996; Reynolds & Bradley, 1983; Scholwinski & Reynolds, 1985), there is also some evidence suggesting that gifted people might be more at risk for problems with emotional or social adjustment, including specific psychiatric disorders. The evidence is more compelling for the higher rates of mood disorders and suicide among creatively gifted writers and visual artists, highlighting the association of psychological risks with creative giftedness and the pursuit of exceptional artistic achievement (e.g., Andreason, 1987, 1988; Claridge, 1998; Doyle, 1998; Jamison, 1989, 1990; Ludwig, 1994; Neihart, 1998; Post, 1994, 1996; Richards, 1989, 1994). However, whether such vulnerability applies not only to creative adults but could also be associated with creative talent among the younger population needs to be carefully investigated.

In Hong Kong, with the development of a growing number of gifted programs in schools, educators are increasingly concerned with the social and emotional needs of gifted students, their adjustment to being labeled gifted, and the design and provision of preventive interventions to help gifted students with adjustment problems (Chan, 1998, 1999). In addressing these concerns, Chan (2003a, 2003b) developed the Student Adjustment Problems Inventory (SAPI) to assess gifted students' specific adjustment problems. In subsequent studies, SAPI was revised, as a 24-item Chinese SAPI (SAPI-24) to cover six domains of adjustment difficulties: unchallenging schoolwork, poor interpersonal relations, parental expectations, perfectionism, multipotentiality, and intense involvement. While these problems have been found to be common among gifted students in Hong Kong, it is of great interest to explore why some students are more likely to experience psychological distress than others in the face of similar adjustment problems. In this connection, studies that examined personality or personal variables as coping resources in the context of stress-distress relationship are relevant. And among the many personal resources, self-efficacy has been regarded as one of the most influential internal coping resources (see Schwarzer & Greenglass, 1999).

In general, research studies on self-efficacy are grounded in one of the two theoretical orientations. One orientation is Rotter's (1966) social learning theory of internal versus external control. According to this conceptualization, internally controlled students believe that they are competent to manage and have control over their adjustment difficulties, whereas externally controlled students believe that the environment has more influence on their adjustment. The second orientation is Bandura's (1997) social-cognitive theory and his construct of self-efficacy. According to this conceptualization, students with perceived self-efficacy are students who hold optimistic self-beliefs that they can change behaviors by personal action. These efficacy beliefs also affect the amount of effort these students expend to meet external demands, and the persistence to continue striving in spite of barriers and setbacks that may undermine motivation. Thus, self-efficacious students might develop success scenarios in which they see themselves as students capable of coping effectively with their adjustment problems. Nonetheless, the two orientations are quite similar, with the internal control and external control corresponding to perceived self-efficacy and a lack of self-efficacy, respectively. While internal-external control orientation is commonly referred to as a general personality dimension, perceived self-efficacy is usually regarded as more specific (e.g., Schwarzer, 1993).

Drawing on self-efficacy theory, it can be assumed that self-efficacious students also have adjustment difficulties, but they perceive more challenge than threat or loss of control. Moreover, while they do have ups and downs in mood and performance, their perceived self-efficacy or optimistic efficacy beliefs help them manage the negative affective experiences before these experiences become stabilized as symptoms of psychological distress (see Chwalisz, Altmaier, & Russell, 1992; Parkes, 1991; Schwarzer & Greenglass, 1999). Alternatively, a lack of self-efficacy or low self-efficacy might be associated with anxiety, worry, and self-doubts (Schwarzer & Greenglass). Thus, it can be conjectured that control beliefs may mediate the impact of adjustment difficulties on the psychological well-being of students.

With this view, the present study aimed at examining the relationships among adjustment problems, self-efficacy, and psychological distress in a sample of Hong Kong Chinese gifted students. Specifically, this study extended past findings on adjustment problems (see Chan, 2003a, 2003b), relating adjustment problems to self-efficacy and psychological distress. Based on past findings, structural equation modeling procedures were employed to test the hypothesized mediation-effect model that self-efficacy mediated the effect of adjustment problems on psychological distress. For comparison, other competing models, including models that specified direct effect as well as direct and mediation effects, were also tested.

Method

Participants

A total of 207 students (122 boys and 85 girls) nominated by their schools to join the gifted programs at the Chinese University of Hong Kong participated voluntarily in this study. These students were in grades 4 to 13, and were between the ages of 9 and 19 (M = 13.06, SD = 2.91). Specifically, letters were sent to principals of schools all over Hong Kong to enlist their help in recruiting gifted students. In nominating students, principals and teachers were requested to recommend students on the basis that they were judged to be either gifted intellectually (with a high IQ score), academically (with outstanding performance in school subjects), or that they have demonstrated specific talents in other nonacademic areas. Since there are no generally accepted standard measures in Hong Kong schools, and schools generally do not have access to information on specific IQ scores of students, no threshold IQ score was specified, and individual schools would make their own judgments on different aspects of giftedness based on their knowledge of their students. Consequently, gifted students in this study were sampled from a broad age range, and could be regarded as relatively heterogeneous with respect to their gifts and talents.

Procedures

All 207 students who accepted school nomination participated voluntarily in this study with written consent obtained from their parents. Students were then invited to attend group assessment sessions to complete three instruments assembled to assess their adjustment problems, perceived general self-efficacy, and psychological distress. These instruments were the SAPI-24 (Chan, 2003a, 2003b), the WegnerSchwarzer-Jerusalem 10-item Chinese General Self-Efficacy Scale (GSE) as reported in Schwarzer (1993), and the 20-item Chinese General Health Questionnaire (GHQ-20) described in Chan (1993, 1995). All 207 students completed these instruments anonymously, and were assured that the data they provided were confidential and would be used for research purposes only.

Measures

Student Adjustment Problems Inventory. The SAPI-24 assesses gifted students' perceived problems and stressors in their adjustment to being gifted. Six problem areas are assessed through responding to items scored on six subscales: Unchallenging Schoolwork, Poor Interpersonal Relations, Parental Expectations, Perfectionism, Multipotentiality, and Intense Involvement. Each problem area is represented by four items, and respondents respond to each item by rating the extent to which these adjustment problems are descriptive of them or apply to them using a 5-point scale ranging from 1 "not at all descriptive" to 5 "very descriptive." The SAPI-24 has demonstrated sound psychometric properties in past studies (e.g., Chan, 2003a, 2003b), with construct validation and correlation with multiple intelligences, and adequate to moderately high internal consistency for the six subscales (ranging from .74 to .83). (English paraphrasing of the SAPI-24 Chinese items can be found in Table 2.)

General Self-Efficacy Scale. The GSE is a 10-item self-report scale designed to assess respondents' perceived self-efficacy in general. In completing the GSE, respondents are requested to judge how true the items could describe them in a 4-point scale ranging from 1 "not at all true" to 4 "exactly true." A total score can be obtained by summing the item responses, and a higher score reflects greater self-efficacy. The psychometric properties of the GSE has been reported in Schwarzer (1993), demonstrating its typically high internal consistency (ranging from .75 to .90), and good convergent and discriminant validity, correlating positively with self-esteem as well as optimism, and negatively with anxiety, depression, and physical symptoms. GSE has been used in studies with the Chinese population, demonstrating its high internal consistency (ranging from .80 to .88), and significant correlations were found between the GSE and social support as well as psychological distress (Chan, 2002). In this study, the Chinese version was employed, and two subscales of self-efficacy were distinguished based on the substantive consideration that two highly related constructs of perseverance and resourcefulness were assessed by the items. The distinction was made to provide two indicators based on the notion of using two item bundles for the construct of self-efficacy in SEM analysis. (English paraphrasing of the Chinese GSE items can be found in Table 3.)

General Health Questionnaire. The Chinese GHQ-20 assesses students' psychological distress in terms of current nonpsychotic symptoms in the five symptom areas represented by subscales of Health Concerns, Sleep Problems, Anxiety, Dysphoria, and Suicidal Ideas. Respondents are requested to rate each symptom statement on a 4-point scale by comparing themselves during the past 2 weeks with their "usual selves." The 4-point scale is scored 1-2-3-4 to reflect the severity of the relevant symptoms that the respondent experienced (see Goldberg, 1978; Goldberg & Williams, 1988). Five subscale scores can be obtained by summing the ratings on the relevant items, with higher scores indicating greater severity on the relevant symptom areas. The Chinese GHQ-20 was developed based on selecting items that differentiated psychiatric patients from nonpatients (Chan, 1993, 1995). The scale has demonstrated sound psychometric properties, including construct validation and correlation with other psychiatric symptom measures, with adequate to moderately high internal consistency (ranging from .79 to .88) for psychiatric patients and somewhat lower values (ranging from .54 to .88) for nonpatients. (English paraphrasing of the Chinese GHQ-20 items can be found in Table 4.)

Overview of Analyses

The present study employed structural equation modeling (SEM) procedures in testing the relationships among adjustment problems, self-efficacy, and psychological distress of Chinese gifted students using LISREL 8 (Joreskog & Sorbom, 1993). When SEM procedures are used, hypothesized models are tested at the construct level rather than at the level of a single measured variable, thus reducing the effect of measurement error associated with specific instruments. Apart from the confirmatory analyses, SEM procedures also allow relationships among sets of constructs to be examined.

In general, SEM models have two basic elements. The measurement model delineates the associations between measured and latent variables, and the correlations among latent variables. The structural model estimates the direct and indirect effects among latent variables. The present analysis followed a stepwise procedure, one similar to that recommended by Anderson and Gerbing (1988), in which the acceptability of the measurement of constructs of adjustment problems, self-efficacy, and psychological distress were evaluated first before proceeding to an evaluation of relations among constructs. In evaluating the relations among the latent constructs, the mediation-effect model specifying that self-efficacy mediated the effect of adjustment problems on psychological distress was first tested, followed by the testing of competing models that included the direct-effect model and the direct-and-mediation-effect model. Since these models would be rejected by the chi-square test statistic at a conventional alpha level if a large enough sample was used, and accepted if a small enough sample was used, a number of residual-based fit indices and comparison-based fit indices were employed to help determine whether the hypothesized models were well-fitting for these data (e.g., Bentler, 1989; Browne & Cudeck, 1993; Steiger, 1990). Thus, apart from the chi-square statistic, the fit indices used included the Root Mean Squared Error of Approximation (RMSEA), the Standardized Root Mean Square Residual (SRMR), the Expected Cross-Validation Index (ECVI), the Goodness of Fit Index (GFI), the Non-Normed Fit Index (NNFI), and the Comparative Fit Index (CFI). In general, a reasonably adequate and good fit is suggested by RMSEA and S-MSR values below .08 and .05, respectively, and by fit index values approaching .90 and above .90, respectively. The best fitting model among competing models will be the one with the smallest ECVI value or the greatest likelihood that the model will cross-validate across a similar-sized sample from the same population (see Byrne, 1998; Diamantopoulos & Siguaw, 2000).

Results

The relevant item responses of the students to SAPI-24, GSE, and GHQ-20 were first tabulated and scored on six subscales of adjustment problems, two subscales of general self-efficacy, and five subscales of psychological distress. Table 1 shows the means, standard deviations, and internal consistency measures of these subscales. The subscales of SAPI-24 and GSE achieved adequate to moderately high internal consistency values (.73 to .85) whereas those of GHQ-20 yielded somewhat lower and modest values (.59 to .69), which were comparable to those obtained with nonpatients (see Chan, 1995). In general, the data seemed to provide some initial support that these subscales of SAPI-24, GSE, and GHQ-20 could serve as indicators of the constructs of adjustment problems, self-efficacy, and psychological distress for further evaluation using SEM procedures.

Measurement Models of Adjustment Problems, Self-Efficacy, and Psychological Distress

To evaluate the acceptability of these subscale scores as indicators of adjustment problems, self-efficacy, and psychological distress, matrices of item covariances were separately computed from items from subscales of adjustment problems, self-efficacy, and psychological distress for confirmatory factor analyses. Specifically, the six-factor measurement model of adjustment problems, the two-factor measurement model of self-efficacy, and the five-factor measurement model of psychological distress were tested using separate maximum likelihood confirmatory factor analyses. Tables 2, 3, and 4 present the summary of results of confirmatory factor analyses for adjustment problems, self-efficacy, and psychological distress, respectively. The corresponding chi-square values and fit indices for these three measurement models are summarized in Table 5.

It can be seen from the results of the confirmatory analyses as summarized in these Tables that when the relevant items were specified to load on their appropriate factors, the loadings were all statistically significant (p < .01). For all measurement models under consideration, the residual-based fit indices were less than or approached .08, and the other comparison-based fit indices were about .80 to above .90. Comparing the measurement models with the independence or null models where there were no correlations among variables, the chi-square values and ECVI indices were all sizably reduced. Taken together, these results suggested that the relevant subscales could be regarded as reasonably acceptable indicators to represent the constructs of adjustment problems, self-efficacy, and psychological distress. Specifically, the matrices of factor correlations indicated that the six subscales of adjustment problems correlated substantially with one another, as did the five subscales of psychological distress, reflecting the latent variables of adjustment problems and psychological distress could be represented by the corresponding subscales as indicators. As expected, the correlation between the two subscales of self-efficacy was very high, suggesting the two indicators could also be merged into one.

Structural Models of Adjustment Problems, Self-Efficacy, and Psychological Distress

The structural models specifying the patterns of relationships among latent variables of adjustment problems, self-efficacy, and psychological distress were successively evaluated using SEM procedures. Thus, the structural models to be tested specified direct and mediating paths from the latent variables of adjustment problems and self-efficacy to the latent variable of psychological distress.

Specifically, one direct-effect model, one mediation-effect model, and one direct-and-mediation-effect model were considered. In the direct-effect model, the two latent variables of adjustment problems and self-efficacy, assumed to be uncorrelated, were specified to be causally prior to, and have direct effects on, the latent variable of psychological distress. In the mediation-effect model, the latent variable of self-efficacy was hypothesized to mediate completely the effect of the latent variable of adjustment problems on the latent variable of psychological distress. In the direct-and-mediation-effect model, a direct path from the latent variable of adjustment problems was added to the mediation-effect model. All parameters estimated for the paths in the three models were statistically significant (p < .05), suggesting that each of the three models could be tenable. Since the three models under consideration involved the same latent constructs and indicators, they could be compared using the fit indices.

Table 6 summarizes the findings and the various fit indices obtained in fitting the three competing models to the data. It can be seen from Table 6 that the direct-effect model only provided mediocre fit and did not fit the data well. In contrast, the mediation-effect model provided more adequate fit, as evidenced by the sizable decrease in chi-square value when compared with the direct-effect model, the smaller residual-based indices with values approaching .05, and the other comparison-based fit indices with values above .90. Thus, the mediation-effect model could be regarded as reasonably acceptable. Similarly, the direct-and mediation-effect model also had better and improved fit index values comparable to those obtained in the direct-effect and mediation-effect models, and there was a further sizable decrease in chi-square value from that of the mediation-effect model. While this model, which was of greater complexity in terms of an additional direct path from adjustment problems to psychological distress, represented a slight improvement over the mediation-effect model, the direct-and-mediation-effect model emphasized the small, direct, but significant impact of adjustment problems on psychological distress, in addition to the significant role of self-efficacy in mediating the impact of adjustment problems on psychological distress. Figure 1 presents this direct-and-mediation-effect model, which could be regarded as the model of choice among the three competing models. As conventionally presented, the variables in boxes were indicators defining the latent constructs in ovals. All parameters shown are standardized, and are statistically significant (p < .05). The significant paths in this model suggested that students who experienced more adjustment problems tended to have enhanced self-efficacy, which in turn led to reduced psychological distress. The enhanced self-efficacy presumably could arise from students' perception of challenge rather than threat or loss of control. On the other hand, the small but significant path from adjustment problems to psychological distress also suggested that students who experienced more adjustment problems might suffer from greater psychological distress.

[FIGURE 1 OMITTED]

Discussion

This study extended past findings on the adjustment problems of gifted students in response to their being gifted to the relationships among students' adjustment problems, their perceived self-efficacy, and their psychological distress. Based on past findings, a model has been postulated specifying that the effects of adjustment problems on psychological distress were mediated by self-efficacy. However, rather than testing the hypothesized model at the level of observed variables, SEM procedures were used to test this hypothesized mediation-effect model at the construct or latent variable level, thus reducing the effect of measurement error associated with specific instruments. At the same time, this model could be compared and evaluated with other competing models using the SEM procedures.

The SEM procedures first started with establishing measurement models of the constructs to be tested. The construct of adjustment problems defined by indicators of unchallenging schoolwork, poor interpersonal relations, parental expectations, perfectionism, multipotentiality, and intense involvement, was generally supported. Also supported, was the construct of psychological distress defined by indicators reflecting health concerns, sleep problems, anxiety, dysphoria, and suicidal ideas, and the construct of self-efficacy defined by the two indicators of perseverance and resourcefulness. Nonetheless, the measurement models provided reasonably adequate fit to the data, lending support to the validity of these constructs.

In examining the structural models specifying relationships among these constructs, three models were successively tested. The direct-effect model was the simplest, postulating that adjustment problems and self-efficacy had independent but opposing effects on psychological distress. While both of the two estimated paths were found to be statistically significant, the fit indices for the model were less well fitting when compared with the other two competing models. Both the mediation-effect model and the direct-and-mediation-effect model performed reasonably well in terms of fit indices and all postulated paths were found to be significant. However, it was believed that the small, yet significant additional direct path in the direct-and-mediation-effect model should not be overlooked because it showed that the experience of adjustment problems could lead directly to psychological distress. Taken together, the direct-and-mediation-effect model could be regarded as the model of choice.

The direct-and-mediation-effect model has important implications for the promotion of psychological well-being among Chinese gifted students. For one thing, as expected, the experience of more adjustment problems could lead to greater psychological distress in students. However, there is another indirect but important path for adjustment problems to exert their influence. Presumably, when a student looks upon the experience of adjustment problems as a challenge, the experience could lead to enhanced self-efficacy, which in turn might render the student less vulnerable to psychological distress. Whether such enhanced self-efficacy could lead to reduced psychological distress (e.g., through positive and adaptive coping), and whether these findings could be applied to gifted students in non-Chinese cultural settings certainly warrant further investigations in future studies. Thus, to promote psychological well-being or to reduce psychological distress effectively among students, efforts need to be made to help students view adjustment problems positively and to enhance their self-efficacy through promoting their perseverance and resourcefulness. Nonetheless, the important mediating role of self-efficacy in tempering the adverse effects of adjustment problems has to be recognized as it has great implications for the provision of preventive interventions to alleviate psychological distress through its promotion.

This study has tested three models, and the direct-and-mediation-effect model was found to be most tenable. However, one has to bear in mind that this study only provided an opportunity for invalidating the models, and the direct-and-mediation-effect model survived the test. Undoubtedly, there are many competing models that might be proven to be equally viable under similar evaluation procedures. Thus, the meaningfulness of the model must be grounded in theory, and the present model appears to be viable under substantive considerations.

This study has many limitations, including the sampling of gifted students through school or teacher nomination, and the reliance on self-report measures to assess adjustment problems, self-efficacy, and psychological distress. Indeed, it has been said nominations of the gifted by schools are likely to highlight high-achieving academics, who are generally labeled gifted by fellow students and teachers. One might even raise the question whether these students are merely more likable, more polite, more compliant, and more motivated. However, past studies of teacher-nominated students have also suggested that these students generally achieved high scores on tests assessing intellectual abilities such as Raven's Standard Progressive Matrices (Raven, Raven, & Court, 1998), and would be included as gifted based on conventional IQ measures. On the other hand, these students might also be relatively free from adjustment problems and psychological symptoms. In this regard, cross-replication with more heterogeneous samples of students should be helpful in establishing the generalizability of the present findings. Thus, future studies assessing a possibly wider spectrum of adjustment problems with more representative samples using measures not confined to self-reports but also using data from interviews, anecdotal materials and judgments from teachers, parents, and peers might help provide further insight into the impact of adjustment problems and self-efficacy on the psychological well-being of gifted students.

Manuscript submitted July 13, 2004.

Revision accepted February 15, 2005.

REFERENCES

Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103, 411-423.

Andreasen, N. C. (1987). Creativity and mental illness: Prevalence rates in writers and their first-degree relatives. American Journal of Psychiatry, 144, 1288-1292.

Andreasen, N. C. (1988). Bipolar affective disorder and creativity: Implications and clinical management. Comprehensive Psychiatry, 29, 207-217.

Baker, J. A. (1995). Depression and suicidal ideation among academically talented adolescents. Gifted Child Quarterly, 39, 218-223.

Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W. H. Freeman.

Bentler, P. (1989). Comparative fit indices. Psychological Bulletin, 107, 238-246.

Blackburn, A. C., & Erickson, D. B. (1986). Predictable crises of the gifted students. Journal of Counseling and Development, 64, 552-555.

Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen & J. S. Long (Eds.), Testing structural equation models (pp. 136162). Newbury Park, CA: Sage.

Byrne, B. M. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS: Basic concepts, applications, and programming. Mahwah, NJ: Lawrence Erbaum Associates.

Chan, D. W. (1993). The Chinese General Health Questionnaire in a psychiatric setting: The development of the Chinese scaled version. Social Psychiatry and Psychiatric Epidemiology, 28, 124-129.

Chan, D. W. (1995). The two scaled versions of the Chinese General Health Questionnaire: A comparative analysis. Social Psychiatry and Psychiatric Epidemiology, 30, 85-91.

Chan, D. W. (1998). Development of gifted education in Hong Kong. Gifted Education International 13, 150-158.

Chan, D. W. (1999). Counseling gifted students in Hong Kong: A critical need. Education Journal 27(2), 145154.

Chan, D. W. (2002). Stress, self-efficacy, social support, and psychological distress among prospective Chinese teachers in Hong Kong. Educational Psychology, 22, 557-569.

Chan, D. W. (2003a). Adjustment problems and multiple intelligences among gifted students in Hong Kong: The development of the revised Student Adjustment Problems Inventory. High Ability Studies, 14, 41-54.

Chan, D. W. (2003b). Assessing adjustment problems of gifted students in Hong Kong: The development of the Student Adjustment Problems Inventory. Gifted Child Quarterly, 47, 107-117.

Chwalisz, K., Altmaier, E. M., & Russell, D. W. (1992). Causal attributions, self-attributions, self-efficacy cognitions, and coping with stress. Journal of Social and Clinical Psychology, 11, 377-400.

Claridge, G. (1998). Creativity and madness: Clues from modern psychiatric diagnosis. In A. Steptoe (Ed.), Genius and the mind: Studies of creativity and temperament (pp. 227-250). Oxford, UK: Oxford University Press.

Derevensky, J., & Coleman, E. B. (1989). Gifted children's fears. Gifted Child Quarterly, 33, 65-68.

Diamantopoulos, A., & Siguaw, J. A. (2000). Introducing LISREL. Thousand Oaks, CA: Sage.

Doyle, C. L. (1998). The writer tells: The creative process in the writing of literary fiction. Creativity Research Journal 11, 29-37.

Freeman, J. (1998). Educating the very able: Current international research. London: The Stationary Office.

Genshaft, J. L., Greenbaum, S., & Borovsky, S. (1995). Stress and the gifted. In J. L. Genshaft, M. Bireley, & C. L. Hollinger (Eds.), Serving gifted and talented students: A resource for school personnel (pp. 257-268). Austin, TX: Pro-Ed.

Goldberg, D. P. (1978). Manual of the General Health Questionnaire. Windsor, UK: NFER-Nelson.

Goldberg, D. P., & Williams, P. (1988). A user's guide to the General Health Questionnaire. Windsor, UK: NFER-Nelson.

Gross, M. U. M. (1989). The pursuit of excellence or the search for intimacy? The forced-choice dilemma of gifted youth. Roeper Review, 11, 189-194.

Jamison, K. R. (1989). Mood disorders and patterns of creativity in British writers and artists. Psychiatry, 52, 125-134.

Jamison, K. R. (1990). Manic-depressive illness, creativity and leadership. In F. K. Goodwin & K. R. Jamison (Eds.), Manic-depressive illness (pp. 332-367). New York: Oxford University Press.

Joreskog, K. G., & Sorbom, D. (1993). LISREL 8: User's reference guide. Chicago: Scientific Software.

Karnes, F. A., & Oehler-Sinnett, J. J. (1986). Life events as stressors with gifted adolescents. Psychology in the Schools, 23, 406-414.

Ludwig, A. M. (1994). Mental illness and creative activity in female writers. American Journal of Psychiatry, 151, 1650-1656.

Neihart, M. (1998). Creativity, the arts, and madness. Roeper Review, 21, 47-50.

Neihart, M. (1999). The impact of giftedness on psychological well-being: What does the empirical literature say? Roeper Review, 22, 10-17.

Parker, W. (1996). Psychological adjustment in mathematically gifted students. Gifted Child Quarterly, 40, 154-157.

Parkes, K. R. (1991). Locus of control and the demand-discretion model of stress. British Journal of Psychology, 82, 291-312.

Pfeiffer, S. I., & Stocking, V. B. (2000). Vulnerabilities of academically gifted students. Special Services in the Schools, 16, 83-93.

Piirto, J. (1992). Understanding those who create. Dayton, OH: Psychology Press.

Post, F. (1994). Creativity and psychopathology: A study of 291 world-famous men. British Journal of Psychiatry, 165, 22-34.

Post, F. (1996). Verbal creativity, depression and alcoholism: An investigation of one hundred American and British writers. British Journal of Psychiatry, 168, 545-555.

Raven, J., Raven, J. C., & Court, J. H. (1998). Manual for Raven 's Progressive Matrices and vocabulary scale. Oxford, UK: Oxford Psychologists Press.

Reynolds, C. R., & Bradley, M. (1983). Emotional stability of intellectually superior children versus non-gifted peers as estimated by chronic anxiety levels. School Psychology Review, 12, 190-194.

Richards, R. (1989). Compelling evidence for increased rates of affective disorder among eminent creative persons. Comprehensive Psychiatry, 30, 272-273.

Richards, R. (1994). Creativity and bipolar mood swings: Why the association? In M. P. Shaw & M. A. Runco (Eds.), Creativity and affect (pp. 44-72). Norwood, NJ: Ablex.

Robinson, N. M., & Noble, K. D. (1991). Social-emotional development and adjustment of gifted children. In M. C. Wang, M. C. Reynolds, & H. J. Walberg (Eds.), Handbook of special education: Research and practice (Vol. 4, pp. 57-76). New York: Pergamon Press.

Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs, 80, 1-28.

Scholwinski, E., & Reynolds, C. R. (1985). Dimensions of anxiety among high IQ children. Gifted Child Quarterly, 29, 125-130.

Schwarzer, R. (1993). Measurement of perceived self-efficacy. Germany: Zentrale Universitats Druckerei der FU Berlin.

Schwarzer, R., & Greenglass, E. (1999). Teacher burnout from a social-cognitive perspective: A theoretical position paper. In R. Vandenberghe & A. M. Huberman (Eds.), Understanding and preventing teacher burnout: A sourcebook of international research and practice (pp. 238-246). UK: Cambridge University Press.

Steiger, J. H. (1990). Structural model evaluation and modification: An interval estimation approach. Multivariate Behavioral Research, 25, 173-180.

Swiatek, M. A. (1995). An empirical investigation of the social coping strategies used by gifted adolescents. Gifted Child Quarterly, 39, 154-161.

Tannenbaum, A. J. (1997). The meaning and making of giftedness. In N. Colangelo & G. A. Davis (Eds.), Handbook of gifted education (2nd ed., pp. 2742). Boston: Allyn and Bacon.

Webb, J. T. (1993). Nurturing social-emotional development of gifted children. In K. A. Heller, F. J. Monks, & A. H. Passow (Eds.), International handbook of research and development of giftedness and talent (pp. 525-538). Oxford, UK: Pergamon Press.

Winner, E. (1996). Gifted children: Myths and realities. New York: Basic Books.

David W. Chan is Professor in the Department of Educational Psychology and Program Supervisor of the Programs for the Gifted and Talented at the Chinese University of Hong Kong. His research interests include stress and coping, psychopathology and health, creativity and intelligence, learning and reading difficulties, and gifted education and talent development. E-mail: [e-mail]

Table 1
Measures of Adjustment Problems,
Self-Efficacy, and Psychological Distress
(N = 207)

                                    Number          Mean
Measure                            of items        rating

Adjustment problems
  Unchallenging schoolwork             4            11.71
  Poor interpersonal relations         4             6.04
  Parental expectations                4            10.78
  Perfectionism                        4            12.98
  Multipotentiality                    4            11.85
  Intense involvement                  4            14.13
Self-efficacy
  Perseverance                         5            15.69
  Resourcefulness                      5            15.31
Psychological distress
  Health concern                       4             7.14
  Sleep problems                       4             5.21
  Anxiety                              4             5.74
  Dysphoria                            4             5.99
  Suicidal ideas                       4             4.62

                                   Standard      Coefficient
Measure                            deviation        alpha

Adjustment problems
  Unchallenging schoolwork           3.98            .81
  Poor interpersonal relations       2.54            .73
  Parental expectations              4.12            .78
  Perfectionism                      3.62            .73
  Multipotentiality                  3.98            .76
  Intense involvement                3.19            .78
Self-efficacy
  Perseverance                       2.62            .73
  Resourcefulness                    3.20            .85
Psychological distress
  Health concern                     1.92            .68
  Sleep problems                     1.64            .69
  Anxiety                            1.59            .67
  Dysphoria                          1.57            .59
  Suicidal ideas                     1.33            .68

Table 2
Summary of Completely Standardized Six-Factor
Solution of Adjustment Problems (N = 207)

                                                    Factor

Adjustment problems                       1    2    3    4    5    6

Unchallenging schoolwork
  Curriculum not challenging (1)          73    0    0    0    0    0
  Schoolwork repetitive and not           72    0    0    0    0    0
    stimulating (7)
  Curriculum not promoting talent or      67    0    0    0    0    0
    interests (13)
  Curriculum easy and non-innovative      78    0    0    0    0    0
    (19)

Poor interpersonal relations
  Feels being isolated by schoolmates      0   72    0    0    0    0
    (2)
  Difficult to make friends (8)            0   65    0    0    0    0
  Feels people treat one like a misfit     0   69    0    0    0    0
    (14)
  Feels being disliked by classmates       0   51    0    0    0    0
    (20)

Parental expectations
  Higher parental expectations than for    0    0   80    0    0    0
    siblings (3)
  Parents arranging activities for         0    0   42    0    0    0
    achievement (9)
  Parents making special demands (15)      0    0   58    0    0    0
  Parents expecting greater achievement    0    0   87    0    0    0
    (21)

Perfectionism
  Generally feels dissatisfied with        0    0    0   54    0    0
    performance (4)
  Demands perfection in all work (10)      0    0    0   63    0    0
  Disappointed for not getting full        0    0    0   65    0    0
    score in exam (16)
  Concerned about imperfect work (22)      0    0    0   76    0    0

Multipotentiality
  Multiple interests in different          0    0    0    0   54    0
    disciplines (5)
  Uncertain about specific talent          0    0    0    0   49    0
    development (11)
  Multiple talents and uncertainty in      0    0    0    0   76    0
    career choice (17)
  Not knowing directions for developing    0    0    0    0   87    0
    talents (23)

Intense involvement

  Stronger feelings for what is            0    0    0    0    0   67
    happening (6)
  More easily excited than other           0    0    0    0    0   60
    students (12)
  More easily involved in what is          0    0    0    0    0   74
    happening (18)
  More concerned or involved than other    0    0    0    0    0   73
    students (24)

Factor 1
Factor 2                                  28
Factor 3                                  36   31
Factor 4                                  44   24   52
Factor 5                                  28   12   36   35
Factor 6                                  46   26   37   51   45

Note. Decimals are omitted. Item numbers are in parentheses.

Table 3
Summary of Completely Standardized
Two-Factor Solution of Self-Efficacy (N = 207)

                                                          Factor

Self-efficacy                                            1     2

Perseverance
  Always manage to solve difficult problems (1)          44     0
  Find ways to get what one wants (2)                    50     0
  Easy to stick to one's aims to accomplish goals (3)    65     0
  Can solve most problems with necessary effort (6)      68     0
  Usually able to handle what comes one's way (10)       69     0

Resourcefulness
  Could deal efficiently with the unexpected (4)          0    80
  Know how to handle unforeseen situations (5)            0    77
  Calm when facing difficulties (7)                       0    74
  Usually find several solutions (8)                      0    65
  Usually think of something to do (9)                    0    70

Factor correlation matrix
  Factor 1
  Factor 2                                               97

Note. Decimals are omitted. Item numbers are in parentheses.

Table 4
Summary of Completely
Standardized Five-Factor Solution
of Psychological Distress (N = 207)

                                           Factor

Psychological distress           1     2     3     4     5

Health concern
  Did not feel in good health    41     0     0     0     0
  Felt run down                  73     0     0     0     0
  Felt ill                       58     0     0     0     0
  Did not feel energetic         63     0     0     0     0

Sleep problems
  Early awakening                 0    48     0     0     0
  Lost sleep                      0    64     0     0     0
  Difficult to sleep              0    68     0     0     0
  Difficult to stay asleep        0    60     0     0     0

Anxiety
  Feared to say anything          0     0    44     0     0
  Could not make a start          0     0    61     0     0
  Dreaded everything              0     0    60     0     0
  Felt everything on top          0     0    64     0     0

Dysphoria
  Not playing a useful part       0     0     0    22     0
  Not enjoying activities         0     0     0    40     0
  Felt worthless                  0     0     0    68     0
  Not feeling happy               0     0     0    51     0

Suicidal ideas
  Felt life not worth living      0     0     0     0    48
  Thoughts of ending life         0     0     0     0    64
  Wishing to be dead              0     0     0     0    85
  Idea of killing oneself         0     0     0     0    44

Factor correlation matrix
  Factor 1
  Factor 2                       29
  Factor 3                       36    45
  Factor 4                       28    23    65
  Factor 5                       21    40    61    76

Note. Decimals are omitted.

Table 5
Fit Indices of the Measurement Models of Adjustment
Problems, Self-Efficacy, and Psychological Distress
(N = 207)

                                                 Fit Index

Model                       [chi square]    df     RMSEA   S-RMR

Adjustment problems
  Independence model         2003.00 *      276     --      --
  Measurement model           563.57        237    .082    .077

Self-efficacy
  Independence model          852.56 *       45     --      --
  Measurement model            60.15         34    .061    .043

Psychological distress
  Independence model         1033.82 *      190     --      --
  Measurement model           306.33        160    .067    .067

                                           Fit Index

Model                           GFI        NNFI     CFI    ECVI

Adjustment problems
  Independence model             --         --      --     9.956
  Measurement model            .0814       .0795   .0824   0.335

Self-efficacy
  Independence model             --         --      --     4.236
  Measurement model            .0945       .0956   .0966   0.496

Psychological distress
  Independence model             --         --      --     5.213
  Measurement model            .0871       .0799   .0831   1.972

Note. Fit indices are from LISREL analyses (Joreskog & Sorbom, 1993).
[chi square] = Normal Theory Weighted Least Squares [chi square];
RMSEA = Root Mean Square Error of Approximation; S-RMR = Standardized
Root Mean Square Residual; GFI = Goodness of Fit Index; NNFI =
Non-Normed Fit Index; CFI = Comparative Fit Index; ECVI = Expected
Cross-Validation Index.

* p < .001.

Table 6
Fit Indices of the Structural Models of Adjustment
Problems, Self-Efficacy, and Psychological Distress
(N = 207)

                                                 Fit Index

Model                       [chi square]    df     RMSEA   S-RMR

Independence model            627.74 *      78      --      --

Direct-effect model
  AP [right arrow] PD         107.79        63     .059    .102
    & SE [right arrow] PD

Mediation-effect model
  AP [right arrow] SE          95.29        63     .050    .078
    [right arrow] PD

Direct-and-mediation-
effect
  Model
  AP [right arrow] SE          83.80        62     .041    .064
    [right arrow] PD &
    AP [right arrow] PD

                                           Fit Index

Model                           GFI        NNFI     CFI    ECVI

Independence model               --         --      --     3.174

Direct-effect model
  AP [right arrow] PD           .925       .887    .909    0.795
    & SE [right arrow] PD

Mediation-effect model
  AP [right arrow] SE           .934       .921    .936    0.734
    [right arrow] PD

Direct-and-mediation-
effect
  Model
  AP [right arrow] SE           .941       .940    .952    0.688
    [right arrow] PD &
    AP [right arrow] PD

Note. AP = Adjustment Problems; SE = Self-Efficacy; PD = Psychological
Distress. Fit indices are from LISREL analyses (Joreskog & Sorbom,
1993). [chi square] = Normal Theory Weighted Least Squares
[chi square] RMSEA = Root Mean Square Error of Approximation;
S-RMR = Standardized Root Mean Square Residual; GFI = Goodness of
Fit Index; NNFI = Non-Normed Fit Index; CFI = Comparative Fit Index;
ECVI = Expected Cross-Validation Index.

* p < .001