Printer Friendly

Explanatory style as a predictor of academic and athletic achievement in college athletes.

A longitudinal archival data collection paradigm was used to study the effects of explanatory style on academic and athletic achievement as predicted by the Reformulated Attributional Model of Learned Helplessness and Depression (RAM) and there is a reported relationship between attained freshman grade point average (GPA) and explanatory style (measured by the Attributional Style Questionnaire (ASQ) in Ivy League students when predicted QPA was controlled. Subsequently, partial support for the validity of RAM with athletes on depressiveness, global self-worth, swimming achievement motivation, and trait sports anxiety was shown but its effectiveness as a trait-oriented measure in a sports setting was questioned. This study examined this question using 92 college athletes from a Division I school who completed the ASQ in their first semester; their GPAs during their freshman and junior years served as dependent variables. Six multiple regression analyses including predicted nonscience GPAs, composite positive, negative, and overall ASQ scale scores, and their interactions as independent variables produced overall significant regression models. Only predicted nonscience GPAs produced significant main effects in overall composite ASQ score equations. Three planned t-tests with composite ASQ scales on groups with higher/lower -than predicted freshman GPAs showed nonsignificant results. During their third year of matriculation, continuing varsity athletes (N = 57) were compared to athlete "drop-outs" (n = 27) to examine possible academic and athletic achievement differences. Although initially similar, athletes had significantly higher GPAs than dropouts. Conversely, dropouts had significantly higher composite positive and overall ASQ scores. The results of this study fail to support the RAM predictions on athletic and academic achievement. An analysis of the validity and reliability of the ASQ instrument indicates that sport-specific instruments have greater predictive power in sport settings.

Un paradigme archivistique longitudinal de rassemblement de donnees a ete utilise pour etudier les effets du style explicatirf sur la realisation academique et athletique predite par le "Reformulated Attributional Model of Learned Helplessness and Depression" (RAM) et il y a une relation signale entre la moyenne generale (Grade Point Average: GPA) des notes obtenues en premiere annee d'etude universitaire et le style explicatif (mesure par "l'Attributional Style Questionnaire": ASQ) avec des etudiants de l'ensenble des grandes universites du nord-est des Etats-Unis lorsque le GPA predit etait controlle. Ulterieurement, on ont montre un soutien partiel pour la validite de RAM avec des athletes en depression, ceux ayant une valeur personnelle globale, une motivation de realisation en natation, et un trait d'anxiete aux sports. Cependant, ils ont mis en doute l'efficacite de RAM en tant qu'instrument de mesure oriente vers le trait dans un contexte d'athletisme. Cette etude a examine cette question en se servant de 92 athletes universitaires provenant d'un etablissement de Premiere Division et qui ont complete le ASQ lors de leur premier semestre; leur GPA a servi de variable dependante au cours de leur premiere et deuxieme annees universitaires. Six analyses de regression multiples y compris des GPA non-scientifiques predits, des composes positifs, negatifs, et les scores globales a l'echelle ASQ, et leurs interactions en tant que variables independantes ont en general produit des modeles de regression significatifs. Seulement les GPA nonscientifiques predits onty produit des effets significatifs assentials dans l'ensembles des equations des scores ASQ composes. Trois test "t" prevus avec des echelles ASQ composees sur des groupes avec des GPA de premiere annee plus eleves/plus bas que predits ont montres des resultats non-significatifs. Pendant leur troisieme annee d'immatriculation, des athletes inscrits en fac (N = 57) ont ete compare avec des athletes qui ont abandonnee leurs etudes (N = 27) pour etudier les differences possibles de realisations academiques et athletiques. Quoique initialement similaires, les athletes ont obtenu des GPA predits plus eleves d'une facon significative que ceux qui ont abandonne leurs etudes. Reciproquement, ceux qui ont abandonnes leur etudes ont obtenues des composes positifs et des scores ASQ globales plus eleves. Les resultats de cette etude ne reussissent pas a soutenir les predictions de RAM sur les relisations athletiques et academiques. Une analyse de la validite et de la fiabilite de l'instrument de mesure ASQ dans des contextes d'athletisme est discutee.

The reformulated attributional model (RAM) of learned helplessness and depression (Abramson, Seligman, & Teasdale, 1978) has been used to predict behavioral outcomes in achievement, health, and pyschopathological paradigms. Basically, this model posits that different reactions to past history of successes and failures play a critical role in predicting future positive or negative motivational and performance responses in achievement- and health-related situations. In the achievement realm, individuals who constantly fail begin to believe that they cannot improve--an unchangeable outcome--thereby leading to reduced effort.

According to RAM, people continually try to explain their past behaviors by using three continuous dimensions (internal-external, stable-unstable, and global-specific) as part of their attributional style. In the internal-external dimension, the event is either caused by the person or the situation; in the stable-unstable dimension, the cause is long-term versus short-term; and in the global-specific dimension, the cause influences many versus few situations.

The Attributional Style Questionnaire (ASQ) was formulated (Seligman, Abramson, Semmel, & von Baeyer, 1979) by using six positive and six negative hypothetical events to elicit the nature and degree of subjects' typical causal explanations in these three dimensions. The ASQ has been criticized for low reliability, using composite scores of the three dimensions to boost reliability, and weak evidence for claiming trait-like attributional style (Carver, 1989; Cutrona, Russell, & Jones, 1985).

Nolen-Hoeksema, Girgus, and Seligman (1986) reported that explanatory style (as measured by the Children's ASQ (1984) significantly correlated with California Achievement Test battery scores (r = .26, p |is less than~ .05), teachers' ratings of helpless behaviors (r = -.51, p |is less than~ .002), and mastery behaviors (r = .56, p |is less than~ .002) in elementary school children. Subsequently, Petersen and Barrett (1987) also indicated that college freshman grade point averages partially correlated (r = -.28, p |is less than~ .01) with an Academic ASQ questionnaire score. These results converge with previous research by Weiner (1985a, 1985b) and others (Dweck & Licht, 1980) linking causal explanations to achievement.

Kamen and Seligman (1986) used the ASQ to assess how explanations for success and failure may influence achievement by examining college grade point average (GPA) to see if explanatory style could predict academic achievement more accurately than traditional admissions formulas. The RAM model would predict that individuals who constantly give internal, stable, and global explanations for bad occurrences (e.g., "I failed because I'm stupid") are less likely to excel in college than those who explain these events in terms of external, unstable, and specific causes (e.g., "I failed because I was sick that day"). In addition, it was also predicted that underachievement would occur in individuals using external, stable, and specific explanations for positive (e.g., "I received an 'A' in math because the teacher is easy") versus individuals adopting an internal, stable, and global explanation (e.g., "I'm a genius"). When SAT scores, high school rank, and achievement test scores were controlled, freshman with nondepressive explanatory style for positive events had higher GPAs than predicted by a predictive index of three independent variables, whereas freshman with a more depressive style did worse or the same as predicted. In a second study with upperclassmen, explanatory style for both good and bad events predicted later GPA with the three traditional measures of success controlled. Upperclassmen with high GPAs had a less depressive style than those with low GPAs.

The RAM model and the use of the ASQ instrument are a relatively new focus of investigation in sport psychology and athletic counseling research. Two studies have been published which investigated sport performance with a subjective or indirect measure in an athletic population (Prapavessis & Carron, 1988; Seligman, Nolen-Hoekesma, Thornton, & Thornton, 1990) and generally supported the RAM model. A recent study by Yin, Callaghan, and Simons (1989) using a battery of correlate performance measures only partially supported the RAM predictions for athletic achievement and raised serious questions about the predictive validity of the ASQ in athletic populations. Contrary to RAM theory, none of the behavior variables (particularly athletic performance) was correlated with swimmers' attributional style.

In an age where college athletes' academic performance is coming under increasing scrutiny from within academia and from society in general (Knight Foundation. 1991), it was decided that Kamen and Seligman's (1986) exploratory study with the original ASQ measure should be replicated and extended within an athletic population. Before sport psychologists and athletic counselors adopt the suggested attributional retraining intervention (Petersen & Barrett, 1987) to help college athletes meet their academic and athletic potential, it is necessary to support the RAM predictive validity for academic and athletic achievement within an athletic population.

Specifically, it was hypothesized that athletes who habitually give internal, stable, and global explanations for bad events in many life situations should be less likely to excel in academic pursuits than student-athletes who explain bad events in terms of external, unstable, and specific causes. Similarly underachievement should occur in habitual external, unstable, and specific explanatory styles for good events versus internal, stable, and global styles. To further extend Kamen and Seligman's (1986) findings, subjects' were studied over a year and a half period via GPAs in the freshman and junior years to test the predictability of explanatory style over time since admissions formulas tend to lose predictive validity with upperclassmen. If upperclassmen have accumulated more positive or negative events in conjunction with academic achievement, then explanatory style for negative events could be a powerful predictor of GPA.

A control group that evolved from the cohort group consisting of "dropouts" from varsity teams also enabled a further replication (Seligman et al., 1990) with athletes and non-elite athletes of an indirect measure of athletic achievement. As a second hypothesis, the RAM model predicts that high achievers (varsity athletes) should score higher in external, unstable, and specific composite scales for bad events and higher in internal, stable, and global composite scales for good events than non-elite athletes.



The participants (age range 17-19 years) were male (n = 47) and female (n = 45) student-athletes (N = 92) at the main campus of an NCAA Division I state university (undergraduate enrollment of 26,000) who were enrolled in a required freshman "college skills" course in the fall semester (1987). Approximately 200 ASQ inventories were distributed to student-athletes in the course and 92 participants completed and returned the form (initial admissions data were unavailable for 13 students and 15 other student-athletes did not complete the questionnaire correctly) for a return rate of 60%.

In order to measure indirectly the effects of athletic achievement, a control group was created by grouping freshman athletes who subsequently "dropped out" by their junior year into a "non-elite athlete" sample (n = 27) and the remaining varsity athletes into an "elite athlete" sample (n = 57). Only degree status full-time students were included in the nonathlete sample, and any athletes who had "flunked out" of school or transferred to another school (n = 8) were excluded. Since it was assumed that athletic talent was the primary determinant in whether a student remained on the athletic team in his/her junior year, continued membership on the team was considered an indirect measure of superior athletic achievement relative to the former athletes.

Dependent Variables

The ASQ (Seligman et al., 1979) measures attributional style in six hypothetical positive and six hypothetical negative affiliation- and achievement-oriented events in three separate dimensions--internality, stability, and globality. High ratings on the 7-point scale represent maximum scores in these dimensions. The three subscales for both positive and negative outcomes can be combined to yield a composite score for positive events (CP) and a composite score for negative events (CN). Finally, an overall composite score that accounts for both events can be formed by subtracting the composite negative score from the composite positive score (CP-CN). The three composite scores were used in this study primarily because they have been shown to be more reliable than individual subscales (internal consistency of .75 for good events and .72 for bad events; Peterson & Seligman, 1984). Several reviews (Petersen & Seligman, 1984; Sweeney, Anderson, & Bailey, 1986) have supported the validity of the ASQ in a variety of situations.

Predicted nonscience grade point average (PNGPA) is a measure of expected GPA for a nonscience curriculum in the university formulated by a regression formula calculated by the admissions office. This yearly, confidential formula produces predictions based on a four-point scale and is weighted heavily toward high school GPA (academic subjects in grade 9-11), less heavily toward Scholastic Aptitude Test scores (math and verbal), and minutely toward honors course participation. For the purpose of this study, the PNGPA serves as a measure of academic ability. Change scores for academic achievement for individual student-athletes were produced by subtracting the PNGPA from their cumulative GPA at the end of their freshman year (FGPA-PNGPA) or in the middle of their junior year (JGPA-PNGPA).


Student-athletes enrolled in a required first-semester freshman transition course entitled "Success in College" were asked to complete the ASQ and allow comparison with their future GPAs as part of the research investigation. Ninety-two freshman from the eight sections of the course satisfactorily completed all parts of the ASQ to quality for the participant pool.

Each participant's PNGPA was obtained from his/her admission records; subsequent cumulative GPAs at the end of the freshman year (FGPA) and alter the fifth semester of his/her junior year (JGPA) were gathered from the university records system to serve as dependent variables.



When Cronbach's coefficient alpha was computed, the reliability of the ASQ scales appeared to be acceptable. The six-item subscales obtained a mean reliability of .63 for positive events and .45 for negative events. The two 18-item composite scales, CP and CN, achieved reliabilities of .80 and .64, respectively. The overall composite score, CPCN, consisting of 36 items produced a reliability of .74.

Means and Standard Deviations

Table 1 presents the means and standard deviations for the three measures of academic achievement and the three composite subscales of the ASQ. At-test of the composite positive (CP) and negative (CP) subscales showed greater internality, stability, and globality for good events than for bad events (t(179) = 14.38, p |is less than~ .0001).

Multiple Regression Analyses

Table 2 presents the intercorrelations among all independent variables. The evidence for multicollinearity among the independent variables appears weak.
Table 1
Means and Standard Deviation for GPAs and ASQ Composite Scales
for the Student-Athlete Sample (N = 92)
Dimension Mean Standard Deviation
PNGPA 2.69 0.40
FGPA 2.75 0.54
JGPA (N = 84) 2.79 0.48
Positive (CP) 16.32 1.80
Negative (CN) 12.70 1.60
Positive-Negative (CP-CN) 3.61 2.28
Table 2
Intercorrelation Matrix for Measures of Achievement and
Explanatory Style for the Student-Athlete Sample (N = 92)
PNGPA 1.00
FGPA 0.64(*) 1.00
JGPA 0.67(*) 0.88(*) 1.00
CP -0.01 -0.08 -0.01 1.00
CN 0.14 0.01 0.10 0.11 1.00
CP-CN -0.10 -0.07 -0.08 0.72(*) -0.62(*) 1.00
* p |is less than~ 0.01

The first regression analyses were performed with FGPA serving as the dependent variable. In the first equation, the independent variables were PNGPA, CP, and the PNGPA X CP interaction, entered in that order. The overall regression model was significant, F(3,88) = 20.32, p |is less than~ .0001, and accounted for 40.9% of the variance in FGPA. No significant main effects or interactions occurred.

In the second analysis, the independent variables, PNGPA, CN, and the PNGPA X CN interaction were entered in that order. The overall regression model was again significant, F(3,88) = 20.41, p |is less than~ .0001, accounting for 41.0% of the variance in FGPA. Once again no significant main effects or interactions emerged.

The third regression equation contained the independent variables PNGPA, CPCN, and the PNGPA X CP-CN interaction, entered in that order. The overall regression model was also significant, F(3,88) = 19.82, p |is less than~ .0001, accounting for 40.3% of the variance of FGPA. A significant main effect for PNGPA, F(1,88) = 4.28, p |is less than~ .0001, occurred, indicating that the freshman nonscience grade point prediction was the primary effective predictor of cumulative freshman GPA.

Three further regression analyses were undertaken with junior year cumulative grade point average (JGPA) as the dependent variable (N = 84). For reasons similar to Kamen and Seligman's (1986) analysis, it was believed that since PNGPA predicts FGPA well for freshman but not as accurately for upperclassmen (JGPA), motivational variables such as explanatory style might be more potent influences on academic achievement.

The first analysis used PNGPA, CP, and the PNGPA X CP interaction as independent variables, entered in that order. The overall model was significant, F(3,80) = 21.71, p |is less than~ .0001, accounting for 44.9% of the variance. No other main effects or interactions were significant.

The second equation used PNGPA, CN, and the PNGPA X CN interaction as independent variables, entered in that order. The overall model was again significant, F(3,80) = 22.09, p |is less than~ .0001, accounting for 45.3% of the variance. A main effect for PNGPA, F(1,80) = 1.80, p = .076, was marginal.

The last analysis entered PNGPA, CP-CN, and the PNGPA X CP-CN interaction as independent variables, in that order. The overall model was also significant, F(3,80) = 21.84, p |is less than~ .0001, accounting for 45.0% of the variance.

A significant main effect for PNGPA, F(1,80) = 4.08, p |is less than~ .0001 also appeared, again suggesting that the university's predictive index was still the best predictor of cumulative GPA even in the junior year.

Analysis of Means

Academic Achievement. Kamen and Seligman's (1986) finding that those who did better than predicted by PNGPA had a significantly better explanatory style than those who did worse was then examined. Subjects were selected who did better/worse than their PNGPA by at least one standard deviation of the PNGPA distribution (one SD = .40). A preliminary comparison of both groups' PNGPA (t(24) = -.61, p |is greater than~ .05) indicated that both groups did not differ initially on predicted academic achievement.
Table 3
Correlations for Achievement Change Scores and Explanatory
Style for Athletes (N = 57) and Non-Athletes (N = 27)
CP -0.10 -0.03
CN -0.13 -0.06
CP-CN 0.01 0.02

Three planned comparison t-tests were computed. Student-athletes who achieved above their PNGPA (n = 13) had nonsignificant explanatory style differences for positive events (t(24) = -.90, p |is greater than~ .05), for negative events (t(24) = -.53, p |is greater than~ .05), and for overall composite score (t(22) = -.24, p |is greater than~ .05) than those (n = 14) who did worse than predicted. Identical nonsignificant results occurred when the three ASQ scale scores of all student-athletes who did better than predicted (n = 55) were compared with those athletes who did worse than predicted (n = 37).

Unlike the original study, the six correlations computed between academic achievement change scores (FGPA- PNGPA or JGPA - PNGPA) and the three ASQ composite scores were all nonsignificant ranging from -.13 to .02 (see Table 3), which further substantiates the evidence (Petersen & Barrett, 1987) against generalized ASQ scores predicting academic achievement.

The occurrence of a control group evolving from the original cohort allowed for further direct comparison of elite athletes (n = 57) and non-elite athletes' (n = 27) academic achievement. A t-test of PNGPA for both groups proved to be nonsignificant (t(38) = -.89, p |is greater than~ .05), which indicated that both groups were initially well matched for grade point potential. Two-sample t-tests of FGPA (t(73) = 2.42, p |is less than~ .05) and JGPA (t(55) = 1.84, p = .07) change scores revealed significant differences in academic achievement with elite athletes (M = 2.85) showing an overachievement trend and non-elite athletes (M = 2.73) underachieving during the three year period which seems counter to RAM predictions and societal trends.

Athletic achievement. Since the RAM model purports to predict a variety of achievement endeavors (Seligman et al., 1990), an analysis of athletic achievement was also undertaken to attempt replication of earlier supportive findings. Several t-tests were calculated to assess the RAM prediction that elite athletes should have higher ASQ scale scores for good events (CP) and lower ASQ scale scores for bad events (CN) than non-elite athletes. A significant t-test for CP (t(48) = -2.36, p |is less than~ .05) indicated that non-elite-athletes (M = 16.98) had a higher ASQ score for good events than athletes (M = 15.96). For CN scores, the t-test (t(43) = .77, p |is greater than~ .05) was nonsignificant, which also failed to-support the RAM prediction for athletic achievement.


This student-athlete based study failed to replicate Kamen and Seligman's (1986) findings for academic achievement and Seligman et at. (1990) results for athletic achievement predicted by the RAM model. Only the university-generated PNGPA proved to be a viable predictor of either subsequent FGPA or JGPA; neither attributional style for good nor bad events offered any predictive value for early or later achievement in college;

Furthermore, in contrast to Kamen and Seligman's findings, student-athletes who achieved higher grades than their PNGPA did not have significantly better explanatory style for positive events, negative events, or the total composite score. Finally, a set of correlations between the three ASQ scales and the two freshman and junior GPAs across all student-athletes were nonsignificant (unlike Kamen Seligman's positive results for upperclassmen).

The original supportive findings with non-athlete Ivy League students were not duplicated by this sample of athletes at a large state university in the freshman or junior year. In this case, student-athletes who believe their academic successes are replicable and occur in many situations because of their efforts or abilities are not any more likely to excel than an athlete who has experienced success but takes no credit for it. Future sport research must investigate the viability of explanatory style as a "trait-like" measure powerful enough to predict accurately achievement across a variety of sport and life situations.

The year and a half archival design revealed more surprising findings between the remaining student-athletes and the recently evolved subsample of athletic dropouts which also ran counter to RAM predictions. Current stereotypes about student-athletes and recent exposes would suggest that non-athletes should achieve greater academic success than athletes. On the other hand, it might be predicted that successful Division I athletes should show higher ASQ scores for positive events since their continued athletic successes have allowed them to maintain varsity status over the dropouts. Neither RAM predictions were supported for academic or athletic achievement. In fact, findings opposite to the prediction occurred in the elite athletes' academic achievement and the non-elite athletes' higher CP scores.

Several explanations are viable to explain why this replication and extension of Kamen and Seligman's (1986) and Seligman et al.'s (1990) results failed to support the RAM model. First, it is feasible that the original Ivy League findings were specific to the achievement-oriented nature of an Ivy League institution and these results cannot be readily generalized to large state universities. In fact, inspection of the two university admissions' regression formulas indicates that the Ivy League formula is not as accurate a predictor of academic performance as the state university formula, which would allow other independent variables (including explanatory style) more potential predictive power in unexplained variance.

Another viable explanation may lie in the design weaknesses of the present study. The small participant sample was a nonrandom, highly select group of skilled athletes whose explanatory style and academic achievement may be particular to Division I athletics or to a specific university situation (also a potential weakness of Kamen & Seligman, 1986 and Seligman et al., 1990). Specialized advising and counseling support, higher motivation to retain eligibility, and inflated grades in easier coursework also may have contributed to the overachievement trend found in this study. In addition, the "manufactured" control group of nonathletes may have differed inconsequentially in explanatory style from the original athlete group, and "dropping out" of varsity participation may be an arbitrary indicator of athletic achievement since the number of positive outcomes in sport over the years would be very similar for both groups. Future research must seek random selection of athlete samples at a variety of educational institutions in order to investigate the generalizability of the RAM model.

So far, past examinations of explanatory style in athletic populations have not borne much fruit. While Seligman et al. (1990) concluded that pessimistic explanatory style seemed to predict poor swimming performance, the performance measures lacked an objective criterion, and the manipulation of false failure feedback was unchecked by researchers (Yin et al., 1989). Furthermore, Yin et al. also reported that pessimistic attributional style was associated with low achievement motivation, but none of the behavioral variables (i.e., training progress), measures of athletic competence, perceived sport success, or expectancy for future success showed any strong associations. Similarly. the present study used an indirect measure of athletic achievement; future studies need to investigate the effect of attributional style on performance with more objective and direct measures.

Although the internal consistency of the ASQ seemed acceptable in this study, prior findings (Cutrona el al., 1985; Yin et al., 1989) have questioned its reliability and validity, especially when using composite scores. In the academic achievement realm, Peterson and Barrett (1987) have developed a variation of the ASQ, the Academic Attributional Style Questionnaire (AASQ), which showed potential in accurately predicting GPAs of freshman students at an eastern university.

Recently Hanrahan, Grove, and Hattie (1989) have developed a questionnaire measure of sport-related attributional style. To date (Hanrahan & Grove, 1990), the Sport Attributional Style Scale (SASS) has demonstrated acceptable psychometric properties to effectively predict athletic performance. Future investigations must carefully test the predictive validity of both specific instruments in a variety of settings.

At present, academic counselors and sport psychologists are no closer to better diagnosis of potential student-athlete motivational deficits which could be systematically corrected through educational and psychological retraining. Based on the nonsupportive results of several studies, coaches are also no closer to safely recruiting "overachieving" student-athletes. Until the time when more conclusive research on explanatory style has been completed, selection must continue to be based on available valid predictors and experiential intuition.


Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. D. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, 87, 49-74.

Carver, C. S. (1989). How should multifaceted personality constructs be tested? Issues illustrated by self-monitoring, attributional style, and hardiness. Journal of Personality and Social Psychology, 56, 577-585.

Cutrona, C. E., Russell, D., & Jones, R. D. (1985). Cross-situational consistency in casual attributions: Does attributional style exist? Journal of Personality and Social Psychology, 47, 1043-1058.

Dweck, C. S., & Licht, B. G. (1980). Learned helplessness and intellectual achievement. In J. Garber & M. E. P. Seligman (Eds.), Human helplessness: Theory and applications (pp. 197-221). New York: Academic Press.

Hanrahan, S. J., Grove, J. R., & Hattie, S. A. (1989). Development of a questionnaire measure of sport-related attributional style. International Journal of Sport Psychology, 20, 114-134.

Hanrahan, S. J., & Grove, J. R. (1990). Further examination of the psychometric properties of the Sport Attributional Style Scale. Journal of Sport Behavior, 13, 183-193.

Kamen, L. P., & Seligman, M. E. P. (1986). Explanatory style predicts college grade point average. Unpublished manuscript, University of Pennsylvania.

Knight Foundation on Intercollegiate Athletics. (1991). Keeping faith with the student-athlete: A new model for intercollegiate athletics. Charlotte, NC: Author.

Nolen-Hoeksema, S., Girgus, J. S., & Seligman, M. E. P. (1986). Learned helplessness in children: A longitudinal study of depression, achievement, and explanatory style. Journal of Personality and Social Psychology, 51, 435-442.

Peterson, C., & Barrett, L. C. (1987). Explanatory style and academic performance among university freshmen. Journal of Personality and Social Psychology, 53, 603-607.

Peterson, C., & Seligman, M. E. P. (1984). Causal explanations as a risk factor for depression: Theory and evidence. Psychological Review, 91, 347-374.

Prapavessis, H., & Carron, A. V. (1988). Learned helplessness in sport. The Sport Psychologist, 2, 189-201.

Seligman, M. E. P., Abramson, L. Y., Semmel, A., & von Baeyer, C. (1979). Depressive attributional style. Journal of Abnormal Psychology, 88, 242-247.

Seligman, M. E. P., Nolen-Hoeksema, S., Thornton, N., & Thornton, K.M. (1990). Explanatory style as a mechanism for disappointing athletic performance. Psychological Science, 1, 143-146.

Seligman, M. E. P., Petersen, C., Kaslow, N. J., Tannenbaum, R. L., Alloy, L. B., & Abramson, L. Y. (1984). Attributional style and depressive symptoms among children. Journal of Abnormal Psychology. 83, 235-238.

Sweeney, P. D., Anderson, K., & Bailey, S. (1986). Attributional style in depression: A meta-analytic review. Journal of Personality and Social Psychology, 50, 974-998.

Weiner, B. (1985a). An attributional theory of achievement motivation and emotion. Psychological Review, 92, 548-573.

Weiner, B. (1985b). Human motivation. New York: Springer Verlag.

Yin, Z., Callaghan, J., & Simons, J. (1989). An examination of the reformulated attributional model of learned helplessness and depression in an athletic population. Unpublished manuscript, University of California at Los Angeles.
COPYRIGHT 1993 University of South Alabama
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Hale, Bruce D.
Publication:Journal of Sport Behavior
Date:Jun 1, 1993
Previous Article:Conflict with baseball umpires: an observational study.
Next Article:Spectator effect on team performance in college basketball.

Related Articles
What the basketball prospect wants to know about you.
The Influences of Peer and University Culture on Female Student Athletes' Perceptions of Career Termination, Professionalization, and Social...
Faculty Attitudes Toward Male Division II Student-Athletes.
Advising indicators: operating from an empirical perspective. (On-going Topics).
The athletic, academic and social experiences of intercollegiate student-athletes.
Characteristics associated with superstitious behavior in track and field athletes: are there NCAA divisional level differences?
Identity structure, role discrepancy and psychological adjustment in male college student-athletes.
Establishing an academic game plan for classroom success.
Study strategy predictors of performance in introductory psychology.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters