# The relative importance of affect variability and mean levels of affect in predicting complex task performance

## TL;DR

This study explores how emotional fluctuations and average emotions affect performance on complex tasks.

## Contribution

The study introduces a novel comparison of affect variability and mean emotion levels in predicting task performance.

## Key findings

- Mean levels of emotion were the strongest predictors of off-task attention and performance.
- Flux in negative emotions also significantly contributed to predicting outcomes.
- Researchers should consider mean levels when studying affect variability's impact.

## Abstract

Although research indicates affect variability—the extent to which an individual’s emotions fluctuate—is associated with behavioral outcomes related to adjustment and adaptability, it is unclear to what extent findings make important contributions to the literature when past research has failed to account for the role of mean levels of emotion. Accordingly, we conducted a repeated-measures laboratory study of college students (N = 253) learning to perform a complex computer task to examine the relative importance of affect variability indices (i.e., spin, pulse, and flux) compared to mean levels in explaining variance in off-task attention and task performance before and after changes in task demands (i.e., skill acquisition and adaptation). In doing so, we also disentangled valence and arousal (i.e., activating versus deactivating) aspects of emotion. Relative importance analyses showed mean levels of emotion were the most dominant predictors (i.e., explained the most variance)—negative deactivating emotions for off-task attention and positive activating emotions for performance. However, flux in negative activating and negative deactivating emotions also explained enough variance to be considered important, suggesting that flux has been overlooked in empirical research. Our findings also highlight that future research must account for mean levels when examining relationships between affect variability and outcomes of interest.