3x3_ANOVA_MainEffectChance

Contributed by stefaniemeliss on June 13, 2022

Collection: Motivation for near-impossibility

Description: For each participant, the blood oxygen level-dependent (BOLD) response was modeled with the general linear model (GLM) for the following regressors of interest: the high-chance stopwatch cue, the moderate-chance stopwatch cue, the extremely-low-chance stopwatch cue, and the three watch-stop cues (i.e. watch-stop cues with three different speeds; they were separately modeled and then averaged). In addition, feedback for stopwatch trials, error trials (i.e. participants did not press a button for a certain duration; see experimental procedures; for these trials, the entire duration of the trial was modeled), session effects, and motion parameters were included as regressors of no interest. The regressors (except motion parameters and session effects) were calculated using a boxcar function for each stimulus convolved with a canonical hemodynamic response function (HRF) without derivatives. Temporal autocorrelation was accounted for by using a first-order autoregressive model during Classical (ReML) parameter estimation. Results for the three watch-stop cues were averaged and used in the following three contrasts to examine the effects of different cues: (i) a contrast between the low-chance stopwatch cue and the averaged watch-stop cues; (ii) another contrast between the moderate-chance stopwatch cue and the averaged watch-stop cues; and (iii) a contrast between the extremely-low chance stopwatch cue and the averaged watch-stop cues. The resulting contrast images were submitted to a 3 (chance of success: high chance, moderate chance, or extremely-low chance) x 3 (group: no-reward, reward, or gambling) mixed ANOVA with chance of success as within-subject factor and group as between-subject factor. The uploaded image represents the unthresholded F map for the main effect of Chance of Success.

Task View 3D View