Implanting rich autobiographical false memories: Meta–analysis for forensic practice and judicial judgment making

• Published in: International journal of clinical and health psychology 2023-10, Vol.23 (4), p.100386, Article 100386
• Main Authors: Arce, Ramón, Selaya, Adriana, Sanmarco, Jéssica, Fariña, Francisca
• Format: Article
• Language: English
• Subjects: Expert testimony; Fabricated memories; Memory wars; Original; Recovered memory; Suggested memories
• ISSN: 1697-2600; 2174-0852
• DOI: 10.1016/j.ijchp.2023.100386

The implanting of rich autobiographical false memories is crucial for judicial decision–making, and the forensic evaluation of a testimony. In order to assess this issue, a meta–analytical review of the probability of implanting rich autobiographical false memories was performed. A total of 30 primary studies analysing the probability of implanting rich autobiographical false memories were retrieved. Random–effects meta–analyses correcting the effect size for sampling error were performed. The results revealed a significant, positive, generalizable (the lower limit for the 80% credibility value was d = 1.13), and more than large mean effect size (d = 1.43[1.33, 1.53]) for the implanting of false memory. The moderating effects of stimulus type showed that the effect of the probability of implanted false memory was significantly higher in experienced events (d = 2.03[1.63, 2.43]) than in false narratives (d = 1.35[1.23, 1.47]), and in doctored photographs (d = 1.29[1.06, 1.52]). A similar effect for memory implantation was observed in both the underage (d = 1.44[1.29, 1.59]), and in adults (d = 1.36[1.22, 1.50]). The moderator techniques for implanting false memories revealed a significantly lower probability of implanting false rich memory with non–directive instructions (d = 0.90[0.53, 1.27]) than with guided imagery (d = 1.45[1.32, 1.58]), or with pressure to answer (d = 1.56[1.17, 1.95]) instructions. The event emotional valence moderator exhibited the same effect for positive (d = 1.27[1.09, 1.45]) and negative valence events (d = 1.30[1.17, 1.43]). The implications of the results for forensic testimony evaluation, police interrogations, and judicial cross–examination are discussed.