Data Analyses
For the analyses of behavioral decision-making data, we used a general
linear model. Risk (three levels: low, high, mixed risk) served as
within-subject factor and gender as between-subjects factor for
pure-risk trials. For ambiguous trials, the three levels of risk were
fully crossed with either low ambiguity or high ambiguity. Accordingly,
ambiguity served as a second within-subject factor with these two levels
of ambiguity. In addition, an ANOVA comprising both trial types was used
to examine differences between no ambiguity (pure-risk trials) and low
and high ambiguity (ambiguous trials).
We used Huynh-Feldt corrections, when the assumption of sphericity was
violated. Follow-up Helmert-contrasts and t-tests were conducted to
identify the direction of the effects and tested one-tailed for
evaluating directed hypotheses. Significance level was p=.05, and only
significant p-values are reported.