実験 Concealed Information Test（CIT）の外的妥当性について<br/>―メタ分析による実務と実験の比較，皮膚電気活動を焦点に―
-
- 財津 亘
- 富山県警察本部刑事部科学捜査研究所
書誌事項
- タイトル別名
-
- Meta-analysis of mock-crime experimental studies and field studies using the Concealed Information Test
- 実験Concealed Information Test (CIT)の外的妥当性について : メタ分析による実務と実験の比較,皮膚電気活動を焦点に
- ジッケン Concealed Information Test (CIT)ノ ガイテキ ダトウセイ ニ ツイテ : メタ ブンセキ ニ ヨル ジツム ト ジッケン ノ ヒカク,ヒフ デンキ カツドウ オ ショウテン ニ
この論文をさがす
抄録
The Concealed Information Test (CIT), a polygraph test procedure that is adopted in Japan, is a memory detection method based on psychophysiological responses to critical questions. In this study, external validity of CIT experiments was examined by meta-analysis comparing electrodermal activity between mock-crime studies and field studies conducted in Japan. Literatures were searched using the following methods. (1) By referring to lists of journals that are reviewed in the CIT. (2) Computer searches of “Science Direct (http://www.sciencedirect.com/)” and “CiNii (http://ci.nii.ac.jp/)” databases. (3) Manually searching academic journals related to experimental, or criminal psychology. Meta-analysis by mixed effects model showed an overall average effect size of d 2.59 and 95% Confidence Interval (CI) [2.07, 3.11]. Nine mock-crime studies (n=187) showed an average effect size of d 2.35, 95%CI [1.77, 2.93], and 4 field studies (n=116) showed an average effect size of d 3.16, 95%CI [1.99, 4.33]. Furthermore, 3 field studies, without 1 field study as an outliner, showed an average effect size of d 2.47, 95%CI [1.67, 3.26]. Moreover, statistical hypothesis testing of the funnel plot indicated that there was no publication bias in these CIT studies. Furthermore, statistical sensitivity analysis indicated the reliability of the results of the meta-analysis. These results of the meta-analysis indicated that each 95%CI of mock-crime experimental studies and field studies had a considerable overlap, and there were not significant differences in effect sizes between both kinds of studies. It is concluded that electrodermal activity in mock-crime studies have adequate external validity.<br>
収録刊行物
-
- 日本法科学技術学会誌
-
日本法科学技術学会誌 19 (1), 9-18, 2014
日本法科学技術学会