A comment on "The use-the-best heuristic facilitates deception detection"

Verschuere et al. (2023) examine the effect of using the best available cue (use-the-best heuristic) on deception detection accuracy across 9 studies. In Study 8 they find that focusing on judging one single cue (detailedness) compared to multiple cues (four cues including detailedness, affect, unexpected complications, and admitting lack of memory) increases the percentage of correctly identifying (i.e. accuracy) transcribed statements as true or false (single cue: 58.93% vs. multiple cues: 54.26%, d = 0.41, BF10 = 2.45). Similarly, in Study 9, they find that a single cue treatment including an explicit decisions rule increases accuracy (66.41%) compared to a multiple cue treatment (59.14%; d = 0.48, BF10 = 7.95). We performed a direct replication (N = 549) including both the implicit rule single cue treatment from Study 8, the explicit rule single cue treatment from Study 9, and the multiple cue treatment from both studies by using the same procedures (i.e., methods and analysis) as Study 9 with new data and a broader sample. First, we successfully replicate Study 9 by finding that the explicit rule single cue treatment increases accuracy (67%) compared to the multiple cue treatment (64.2%, d = 0.20, BF10 = 1.23). However, we do not replicate Study 8, by finding that the implicit rule single cue treatment results in descriptively less accuracy (62.4%) compared to the multiple cue treatment (64.2%, d = -0.13, BF10 = 0.06). Thus, we confirm the sign, magnitude and statistical significance of the point estimates for detection accuracy for Study 9 but not Study 8. Second, we test the sensitivity of the results by performing a multilevel model accounting for within-variation in participants and statements and observe similar results in that we replicate effects for Study 9 but not Study 8.