او برای بدست آوردن پول سخت کار می کند: تلاش ارزش گذاری زمینه ساز تفاوت های جنسیتی در خودناتوان سازی رفتاری

Abstract Research in the area of self-handicapping has consistently demonstrated a robust yet puzzling gender difference in the use of and evaluation of behavioral self-handicaps; women (1) are less likely to use these forms of handicaps, particularly those involving the actual or reported reduction of effort, and (2) evaluate the use of these handicaps by others more negatively than do men. The present research examines several possible explanations for these consistent gender differences and finds that the personal value placed on effort is an important mediator of these effects.

Results Time spent practicing (M = 168.56 s, SD = 131.27) and number of practice items completed (M = 7.54, SD = 6.36) were highly correlated (r = .67) and were therefore standardized and summed to create a single index of practice effort (reliability α = .80), as in prior research (e.g., Hirt et al., 2000). The reliability for the candidate mediators ranged from marginally acceptable (Ability vs. Effort Tradeoff α = .66; Prescriptive Effort Norms α = .60) to good (Worker α = .81). The Worker scale (M = 24.98, SD = 5.92) was correlated r = −.25, p < .001, with the Ability vs. Effort Tradeoff scale (M = 13.32, SD = 4.59) and r = .46, p < .001, with the Prescriptive Effort Norms scale (M = 17.10, SD = 2.33). The correlation between the Ability vs. Effort Tradeoff scale and Prescriptive Effort Norms scale was r = −.29, p < .001. Manipulation check An ANOVA on the manipulation check assessing participant’s beliefs about practice revealed a significant effect of the instruction manipulation, F(1, 184) = 9.89, p < .01, such that those in the practice matters condition believed that practicing was more important (M = 3.84) than did those in the practice does not matter condition (M = 2.99). Thus, it appears that participants properly understood and believed the practice instructions. Importantly, there were no effects of gender on this measure (Fs < 1, ns), indicating that men believed that practice was just as important to performing well on the CFIT as did women. Practice effort To provide additional evidence that reduced practice effort in the practice matters condition reflected behavioral self-handicapping, we controlled for reports of the importance of the exam (α = .86), the perceived validity of the exam, and perceived pressure to practice as covariates. 5 Therefore, a 2 (Gender) × 2 (Practice instruction) ANCOVA, including these three variables as covariates, was conducted on the practice effort index. Adjusted means, after controlling for the three covariates, are presented in Fig. 3. There were no effects of the covariates (all Fs < 1.99, ps > .16). The effect of practice instruction approached significance, F(1, 181) = 2.90, p = .09, η2 = .01. There was a significant effect of gender, F(1, 181) = 11.79, p < .01, η2 = .06, such that men practiced less than women. Adjusted practice index scores (Study 3). Fig. 3. Adjusted practice index scores (Study 3). Figure options In addition, a significant interaction of Practice instruction × Gender was observed, F(1, 181) = 4.61, p < .05, η2 = .02. This interaction revealed that men practiced less than women in the practice matters condition, F(1, 95) = 16.36, p < .001, η2 = .13, but not in the practice does not matter condition, F(1, 83) = 1.70, p = .20, η2 = .02. Put differently, whereas women practiced significantly more when told practice mattered, F(1, 91) = 6.75, p < .05, η2 = .07, men practiced nonsignificantly less when told practice mattered, F(1, 87) <1, ns, η2 = .00. This pattern replicates findings obtained in Hirt et al. (2000) and again demonstrates that men self-handicapped in the practice matters condition by not putting forth adequate practice effort. Mediational analyses Next, we sought to test mediation of the significant gender effect on the practice index within the practice matters condition. In this condition, gender differences were observed on both the Worker and Prescriptive Effort Norms scales (see Table 7). Regression analyses were then used to test for mediation within the practice matters condition. Dummy coding was used for the gender variable (Men = 1, Women = 0). To ensure any observed effects were not due to the importance of the exam, perceived validity of the exam, or perceived pressure to practice, we conducted our analyses on residual practice scores after removing the effects of these variables. There were no gender differences on the Ability vs. Effort Tradeoff scale and preliminary analyses revealed this variable did not predict the practice index, and so this variable was not included in the analysis. Therefore, gender was added as a first step in the analysis, and the Worker scale and Prescriptive Effort Norms scale were added as a second step. The results of these mediational analyses are summarized in Table 8. The initial model including only gender revealed a significant gender effect. In the mediation model, the Worker scale predicted practice index residual scores while controlling for gender differences, whereas the Prescriptive Effort Norms scale did not. The addition of these terms added significantly to the model. The inclusion of the Worker scale reduced the gender effect (Sobel test z = 2.44, p < .05), although not completely eliminating it. Thus, the Worker scale partially mediated gender differences in the practice matters condition, whereas the Prescriptive Effort Norms scale did not. Table 7. Gender effects on candidate mediators within practice matters condition (Study 3) Measure Men Women t η2 Mean SD Mean SD Worker scale 23.33 6.40 27.08 5.84 3.06† .09 Prescriptive Effort Norm scale 16.61 2.14 17.88 1.78 3.23† .10 Ability vs. Effort Tradeoff scale 13.37 4.75 12.51 4.46 1.13 .01 † p < .01. Table options Table 8. Mediational analyses for effects on practice index within practice matters condition (Study 3) Initial model Mediation model β t p β t p Gender −.370 3.94 <.001 −.249 2.76 <.01 Worker scale .469 4.26 <.001 Prescriptive Effort Norm scale −.051 <1 ns R2 = .14, FChange(1,98) = 15.55, p < .01 R2 = .31, ΔR2 = .18, FChange(2,96) = 12.43, p < .001 Table options In an additional analysis, we also tested the unique effects of the Worker scale. The Worker scale was again a significant predictor of practice effort (β = .439, t = 4.99, p < .001), and added significantly to the amount of variance explained, R2 = .31, ΔR2 = .18, FChange(1, 97) = 24.86, p < .001. Furthermore, the inclusion of the Worker scale reduced the gender effect (β = −.241, t = 2.73, p < .01), partially mediating this effect (Sobel test z = 2.57, p < .05).