Conscientiousness and neuroticism, self-management practices, and perceived situational constraints were integrated into a model that predicts efficacy and performance. The model was tested using structural equation modeling with a sample of 228 undergraduate students. The results indicated that individual differences exist in self-management practices and perceptions of situational constraints, and that both self-management practices and perceived situational constraints have partial mediating effects on the relationship of personality to self-efficacy and performance.
A myriad of past research has explored the relationships between dispositional characteristics, motivation, and performance. Over the past decade, the dispositional approach to understanding motivational and performance outcomes has received a great deal of interest and support (Barrick and Mount, 1991, Chamorro-Premuzic and Furnham, 2003, House et al., 1996 and Judge and Ilies, 2002). Several meta-analyses have provided strong evidence for the relationships between personality and job performance (Barrick and Mount, 1991 and Barrick et al., 2001) and between personality and motivation (Judge & Ilies, 2002).
With support for such relationships now established, little is known regarding the mechanisms that may underlie these established relationships (Barrick et al., 2001). In other words, now that we have evidence for which individual differences may impact performance, more information is needed regarding how such differences impact motivation and performance (Kanfer, 1991 and Weiss and Adler, 1984). This study investigates two possible mediating mechanisms that may be at work in the relationships between personality, motivation, and performance: self-management and perceived situational constraints.
Means, standard deviations and variable intercorrelations are shown in Table 1. Conscientiousness displayed strong correlations with both self-management (r = .58, p < .01) and academic self-efficacy (r = .51, p < .01). Relatively strong correlations were also observed between academic self-efficacy and both self-management (r = .48, p < .01) and perceived situational constraints (r = −.41, p < .01).
Table 1.
Means, standard deviations and intercorrelations of study variables
Mean SD 1 2 3 4 5 6
1. Conscientiousness 3.62 .60 (.85)
2. Neuroticism 2.82 .67 −.11 (.88)
3. Self-management 3.85 .64 .58∗∗ −.23∗∗ (.74)
4. Situational constraints 2.50 .41 −.29∗∗ .21∗∗ −.35∗∗ (.71)
5. Academic self-efficacy 3.41 .58 .51∗∗ −.25∗∗ .48∗∗ −.41∗∗ (.69)
6. Course grade 2.97 .83 .39∗∗ .09 .20∗∗ −.15∗ .18∗∗
7. General Mental Ability 26.22 2.93 −.05 −.12 −.10 −.09 .01 .20∗∗
∗p < .05.
∗∗p < .01.
Note: Numbers on the diagonal reflect the Cronbach’s alpha reliability.
Table options
A measurement model consisting of all the Time 1 measures resulted in a good fit to the data χ2(38, n = 228) = 89.63 (p < .05), CFI = .95, NFI = .93, RMSEA = .07. To test our hypotheses, we developed a structural model that incorporated the measurement model and allowed the paths shown in Fig. 1 to be freely estimated, as well as a path from general mental ability to course grade. We allowed the correlations between the exogenous variables and the disturbance terms (i.e., the variance not accounted for by the exogenous variables) associated with perceived situational constraints and self-management to be freely estimated ( Kline, 1998). We implicitly tested the mediated relationships contained in our model, by constraining all of the remaining paths to zero. Overall, the model did yield a marginal fit to the data: χ2(78, n = 228) = 181.93 (p < .05), CFI = .90, NFI = .89, RMSEA = .08. Following the procedure recommended by Joreskog (1993), we modified the model by adding a path that was both statistically significant and conceptually consistent with our research model. Examination of the modification indices and correlation residuals indicated that model fit could be improved by adding one such path, from conscientiousness to performance.
The addition of the direct path from conscientiousness to performance is consistent with prior research that has examined the relationship of conscientiousness with performance (e.g., Barrick et al., 1993). Given the complexity of the conscientiousness construct, and the investigation of only two potential mediating processes in this study, it was plausible that conscientiousness may be impacting performance through a more direct route. This change resulted in an improved model fit: χ2(77, n = 228) = 170.31 (p < .05), CFI = .92, NFI = .91, RMSEA = .07.
Path analytic results of the SEM analyses of the modified model are shown in Fig. 2. The standardized path coefficients from conscientiousness and neuroticism to self-management were both significant (=.57 and =−.23, respectively, p < .01), providing support for Hypotheses 1 and 2. The standardized path coefficient from self-management to academic self-efficacy was significant (=.50, p < .01), providing support for Hypothesis 3. The standardized path coefficients from both conscientiousness and neuroticism to perceived situational constraints were significant (=−.46 and = .21, respectively, p < .01), Thus, Hypotheses 4 and 5 were supported. Finally, Hypotheses 6 and 7 were supported, as the standardized path coefficient from perceived situational constraints to self-efficacy was significant (=−.31, p < .01), as was the path from academic self-efficacy to course grade (=.26, p < .01).
Full-size image (25 K)
Fig. 2.
Path analytic results. Note: all path coefficients are standardized; numbers in parentheses represent explained variance; all variables were modeled as latent variables except course grade and general mental ability, which were modeled as observed variables; the path indicated with a dashed line was not included in the original model. ∗p < .05 ∗∗p < .01.
