Objectives
The “cheating paradigm” is an often-used procedure that randomly assigns participants to cheat (guilty) or not cheat (innocent). However, not all participants conform to their assigned condition. We investigated the potential impact of including non-conformers in analyses under an intent-to-treat model (ITT) on decisions to confess, plea, and waive Miranda rights.
Methods
We conducted a series of meta-analyses with studies that used the cheating paradigm to study the legal decisions of mock suspects and that provided enough statistical information for all participants.
Results
Overall, non-conforming guilty participants had lower odds of confessing, pleading guilty, and waiving Miranda rights than conforming guilty participants, whereas non-conforming innocent participants had higher odds. Importantly, including non-conforming participants under an ITT model attenuates, but does not eliminate, the effect of guilt status on decisions.
Conclusions
These findings highlight how willingness to cheat influences legal outcomes and that researchers need to more carefully consider non-conforming participants.