BIS Sensitivity, Negative Affect and Performance: Dynamic and Multilevel Relationships

Abstract

This study examined the dynamic and multi-level relationships among the Behavioral Inhibition System (BIS), negative affect and performance. Participants performed multiple trials of a simulated air traffic control task. A single measure of BIS was taken before practice, whereas negative affect and performance were measured at repeated intervals. As expected, there was a negative relationship between negative affect and performance at both an inter- and intraindividual level of analysis. The key findings from this research relate to dynamic effects. The detrimental effect of negative affect on task performance strengthened across practice, at both levels of analysis. The negative relationship between BIS and performance was also found to change over time. These findings are discussed by drawing on resource allocation theory and reinforcement sensitivity theory. This study emphasizes the importance of examining dynamic and multilevel relationships to understand how BIS and negative affect relate to task performance.

Notes

¹Note that some previous research has referred to the ICC2 as the “reliability estimate” (e.g. Smillie et al., 2006; Yeo & Neal, 2004).

*p < .05.

**p < .01.

***p < .001.

*p < .05.

**p < .01.

***p < .001.

²This model (and follow-up analyses related to mediation) controlled for interactions involving practice given that the effects of BIS and average NA were expected to change over time.

³To improve the model fit, the Level 1 interactions between negative affect and practice were specified as fixed effects because of nonsignificant random effects and low ICC2s.

⁴We thank an anonymous reviewer for this suggestion.

⁵Although Zinbarg and Mohlman (1998) took multiple measurements of punishment expectancies, they did not assess the relationship between BIS and punishment expectancies over time. Moreover, given that they did not use multilevel analyses, their results regarding the correlation between BIS and the linear growth parameter cannot be used to make inferences regarding the dynamic effect of BIS.

⁶We thank an anonymous reviewer for this suggestion.