Despite the prevalence of guilty pleas, we know relatively little about factors that influence the decision to plead. Replicating and extending Dervan and Edkins’ Journal of Criminal Law and Criminology, 103, 1-48. https://doi.org/10.2139/ssrn.2071397, (2013), we conducted two experiments to examine the effects of guilt status, trial penalty, and conviction likelihood on plea outcomes using an adaptation of a high-stakes cheating paradigm. Students were led to believe that they were participating in a study examining team versus individual problem solving. Those randomly assigned to a guilty condition were induced to cheat on an individual problem by a study confederate (in clear violation of the study instructions). All participants were later accused of cheating in the research study, and were offered the analogue of a plea deal in an academic context. Across both experiments, guilty participants were significantly more likely to plead guilty than innocent participants. In Experiment 2, conviction probability affected plea rates only among the innocent. The trial penalty manipulation had no significant effect on plea rates. Reasons for pleading guilty differed between the innocent and the guilty, whereas the plea rejection rationales were similar across the two groups. Overall, this research highlights several avenues for further research aimed at improving the current system of pleas to reduce false guilty pleas.

尽管普遍存在认罪认罪，但我们对影响辩护决定的因素知之甚少。复制和扩展Dervan和Edkins的《刑法和犯罪学杂志》，第103卷，第1-48页。https://doi.org/10.2139/ssrn.2071397，（2013年），我们进行了两项实验，通过改编高风险作弊范式来检验罪恶状况，审判处罚和定罪可能性对认罪结果的影响。学生被认为相信他们正在参加研究考试小组而不是个人问题解决。由研究同盟（明显违反研究指示）诱使那些被随机判有罪的人在个别问题上作弊。后来，所有参与者都被指控在研究中作弊，并在学术背景下向其提供了辩诉交易的类似物。在这两个实验中，有罪的被告比无罪的被告认罪的可能性要高得多。在实验2中，定罪概率仅影响无辜者的认罪率。审判刑罚操纵对认罪率没有显着影响。无罪和有罪的认罪理由不同，而两组的认罪拒绝理由相似。总体而言，这项研究突出了旨在改善当前现状的进一步研究的几种途径。认罪的系统，以减少虚假认罪。