Although potential concerns exist regarding the effects of fatigue over time and the emotional validity of the IAPS images used as a trial-by-trial emotion manipulation, self-report evidence indicates that the block-based positive emotion induction did elicit the desired effect.
Furthermore, transient pupil activity did not significantly change with trial-by-trial emotion manipulation.
Finally, a test of micro-expression recognition was performed consisting of 14 trials (two trials per emotion), similar to the practice session but without feedback.
Analyses of block- and trial-based emotion effects in pupil data followed the structure used for the reward session data.
Thus, the null effects of trial-level emotion on AX-CPT performance and pupillometric indices should be interpreted with caution.
This contrast indicated that incongruent trials (authentic emotion following the announcement of play-acted emotion and vice versa) induced increased activation in the pSTS and TPJ (Fig. 5a).
In the trial-based emotion contrast (Pos Neut vs. Pos Pos), RTs showed a pattern consistent with increased proactive control (slowing in AY but speeding in all other trial types, although the overall effect was very small, <10 ms).
To examine trial-based emotion effects on performance (in the emotion session), we conducted a 2 × 2 × 2 ANOVA on Pos Neut and Pos Pos trials, with Valence (neutral, positive), Contextual Cue (A, B), and Probe (X, Y) as within-subjects factors.
The analysis of pupil dilation during the cue maintenance period (Figs. 7b and 8b), to examine trial-based emotion effects, again yielded a B > A cue effect [ F 1, 91) = 7.526, p =.007] but no significant effects of emotion condition on transient pupil activity.
In general, across emotions, scores were highly correlated within a scoring method and between the first and second halves of the production trials (see Correlation with Same Emotion Trial rows in Table 1).
This precue manipulation followed the trial-by-trial emotional stimuli used by Dreisbach (2006), in which positive and neutral emotion inductions led to differing patterns of cognitive control.
