In addition, to ensure that the order of these three blocks (implemented punishment, non-implemented punishment and a standard game) did not have any effect on the contribution levels, the order of three blocks was randomized (see details below).
Therefore, the order of implemented punishment versus non-implemented punishment blocks had no effect on choices in the PGGs.
However, in one condition participants were actually punished if they transgressed the rule (implemented punishment), and in the other condition they were not punished even if they transgressed the rule (non-implemented punishment).
It supports the prediction that those high in BIS-anxiety made higher contributions in the non-implemented as opposed to the implemented punishment conditions.
Finally, lower BIS-anxiety should lead to high levels of free riding (lower cooperation) under conditions of non-implemented punishment in comparison to implemented punishment, whilst higher FFFS should lead to increased contributions under implemented conditions.
A difference score between levels of contributions in the implemented versus non-implemented punishment blocks was calculated (contributions in the implemented punishment condition minus contributions in the non-implemented punishment condition: a positive score indicates that more was given in the implemented condition than the non-implemented condition).
There were no significant differences in the contribution levels between those who experienced the implemented punishment block before the non-implemented punishment block as compared to those who had experienced these blocks in the reversed order: the implemented punishment block first (t (77) = −1.41, ns), the non-implemented punishment block first (t (77) = −1.28, ns).
a Not all participants received a punishment in the implemented punishment block, as the rule for being punished was contributing less than group partners, which was 16 or 17 MUs.
It could be possible that participants who experienced an implemented punishment block prior to the non-implemented punishment block (N = 40) were more likely to expect punishment in the non-implemented punishment block as compared to those who had experienced these blocks in reverse order (N = 39).
Conversely, compared to the implemented punishment condition, those low in BIS-anxiety gave less in the non-implemented punishment condition.
