Domain-general and math-specific self-perceptions of perseverance as predictors of behavioral math persistence

Abstract

Three studies examine a novel pathway by which the perseverance component of the personality trait grit might predict college students’ behavioral persistence when solving challenging math problems. Specifically, we focus on the intervening role of what we refer to as math-specific self-perceptions of perseverance, which captures students’ perceived tendency and ability to persevere on challenging math problems. Across studies, we found that this math-specific construct was correlated with behavioral math persistence, whereas the domain-general perseverance component of grit was not. Despite there being no correlation between one’s general perceptions of perseverance and behavioral persistence on math problems, we consistently found significant indirect effects of general perceptions through math-specific perceptions of perseverance. That is, in all three studies, grittier students viewed themselves as more capable of persevering on challenging math problems, which ultimately predicted their behavioral persistence at a later time point.

Keywords:

• Grit
• math-specific perseverance
• math problem-solving
• math persistence

Notes

1 For a full list of measures and procedures for all three studies, along with some additional analyses, see the supplementary materials. For more information about the exploratory and post hoc nature of our analyses, about the evolution of our conceptual framing, and about exclusions, also see the supplementary materials. A set of analyses for Study 1 pertaining to participants’ growth mindsets were published in Shen, Miele, and Vasileyeva (2016).

2 The participants who completed Session 1, but not Session 2, did not vary significantly from the final sample in terms of domain-general perseverance, math-specific perseverance, or age. There was a marginal difference between these groups in terms of gender composition (see the supplementary materials for details).

3 Math confidence—a self-perception variable (see Marsh et al., 1988; Pajares & Miller, 1994; West et al., 1980) which seems to overlap with math-specific perseverance to some extent—was one of the additional individual difference variables assessed in Studies 1 and 2. Although math confidence significantly predicted behavioral persistence, it did not exhibit a consistent relation with domain-general perseverance. See the supplementary materials for details regarding the measurement and analysis of this construct.

4 All three studies involved a feedback/challenge manipulation that was administered prior to the persistence measure (see Shen et al., 2016). As detailed in the supplementary materials, there were no main effects of the manipulations on the primary outcome measures in any of the studies, nor did the manipulations moderate the relation between any of the primary predictors in the present paper and the outcome measures. This is contrast to findings by Lucas et al. (2015), who found that grit was only associated with persistence in the failure condition.

5 This includes the math confidence variable examined in the supplementary materials.

6 It may be of interest to some readers to note that openness to experience, which we assessed in Session 1 of the present study (but not in the other studies), exhibited significant correlations with domain-general perseverance and math-specific perseverance and a near significant correlation with behavioral persistence (see the supplementary materials).

7 Demographic information on race/ethnicity were not collected for either of these studies. As explained in the supplementary materials, the mean age reported for Study 2b represents an estimate.

8 In Study 2a, one participant’s rating was coded as 2.5 because he/she marked both 2 and 3.

9 The greater percentage of participants who skipped the choice item (compared to Study 2a) can be attributed to a difference in formatting.

10 It may be of interest to some readers that math value, which we assessed in Session 1 of Study 2a (but not in Session 1 of the other studies), was strongly correlated with math-specific perseverance and significantly predicted behavioral persistence (a similar pattern to what we observed with math confidence; see the supplementary materials).

11 For all three studies, we conducted additional regression analyses to examine the potential interaction effects of math-specific perseverance × task/test performance and/or math-specific perseverance × math confidence on persistence. These analyses yielded inconsistent results. For the results of these analyses and for a description of other regression models we examined across the studies, see the supplementary materials.

12 Relevant to this, we found a significant negative indirect effect of math-specific perseverance on behavioral persistence through domain-general perseverance in Study 1, but not Studies 2 and 3 (see the supplementary materials).

Additional information

Funding

Support for this research was provided by the Buehler Sesquicentennial Assistant Professorship and by Collaborative Grant No. 220020483 from the James S. McDonnell Foundation.