Abstract
This study examined how the public perceives interview styles and investigated the effects of a defendant’s intellectual disability on perceptions of a police interview and decisions about guilt. An online survey was administered to 841 members of the general public, who were presented with a fictional arson case and one of two different interviews, one ‘open-ended’ and the other ‘closed-ended’. Half of the participants received a description of a defendant with an intellectual disability. Participants believed that the open-ended interview was fairer, and that the defendant’s confession elicited through this interview was more voluntary, truthful and credible. Questioning style did not influence participants’ decisions about guilt, but the perception of the interview and confessions mediated between interview style and their decisions. Thus, interviewers are recommended to use open-ended questions during suspect interviews, not only to obtain information, but also to ensure that their confessions are admitted in court.
Notes
1 In the academic literature, when referring to disability person-first language (namely, a person with intellectual disability) was recommended. For example, its use was promoted by the sixth addition of the publication manual of the American Psychological Association (APA, Citation2010). However, as Dunn and Andrews (Citation2015) indicate, advocates of disability culture and disability studies scholars have challenged the exclusive use of the person-first language and have instead promoted the use of identity-first language (namely, a disabled person). In line with this movement, the seventh edition of the APA manual (Citation2020) permits the use of both person-first and identity-first languages. Therefore, person-first and identity-first languages were used interchangeably throughout this study.
2 This survey included only those aged between 20 and 69 years because in Japan a person aged 70 years-old or over can decline to be appointed as a saiban-in. In terms of percentage-wise age distribution of Japan’s population aged 20–69 years, 16.5%, 18.1%, 23.3%, 22.2% and 19.9% were in their 20s, 30s, 40s, 50s and 60s, respectively (Statistics Bureau of Japan, Citation2023). The participants in this study were largely reflective of the Japanese population, although participants in the 20s were slightly higher than those in the Japanese population.
3 The internet survey company ensured the quality of the responses. For example, if a participant provided an irrelevant free response, all of the participant’s responses were excluded. Additionally, when the participant’s response was received in a short time span, all the participant’s responses were omitted. The participants were required to answer all the questions (in a case where they did not want to answer a certain question, they had to choose the option of ‘don’t want to answer’).
4 Although the main effect for defendant type was not statistically significant, univariate analyses showed a significant main effect of defendant type on truthfulness at the p < .0125 level (using Bonferroni corrections), F(1, 837) = 7.60, p = .006, ηp2 = .009. Given the non-significant effect found by MANOVA, this result was not discussed in the text.