Results and Discussion ---------------------- A Chi-Square was computed comparing percentage of correct identifications for the experimental condition (39.2%) and control condition (53.9%), X²(1) = 3.281, p=.07. For participants who failed to make a correct identification (n = 80), a Chi-Square was then computed comparing the percentage selecting the wrong face (misidentification) to the percentage indicating “not present." X²(1) = 0.288, p = .59. 57.1% of control participants made a misidentification versus 51.1% in the experimental condition. We computed a 2 (Condition) x 2 (Correct vs. incorrect lineup ID) ANOVA on confidence ratings. The effect of condition was marginally significant, F (1,146) = 2.805, p = .096 (experimental M = 4.634, control M = 4.313). The effect of lineup identification was non-significant, F (1, 146) = 1.775, p = .185 (correct M = 4.601, incorrect M = 4.346). The interaction between condition and lineup identification was also non-significant, F (1, 146) = .559, p = .456. Finally, we computed the confidence-accuracy correlations for the experimental (n =74), r = .05, p = .671 and control conditions (n = 76), r = .164, p = .156. Study 2 ------------------------ A Chi-Square was computed comparing percentage of correct identifications for the experimental condition (36.1%) and control condition (56.7%), X²(1) = 5.931, p=.015. For participants who failed to make a correct identification (n = 75), a Chi-Square was then computed comparing the percentage selecting the wrong face (misidentification) to the percentage indicating “not present." X²(1) = 0.645, p = .42. 44.8% of control participants made a misidentification versus 54.3% in the experimental condition. We computed a 2 (Condition) x 2 (Correct vs. incorrect lineup ID) ANOVA on confidence ratings. The effect of condition was non-significant, F (1,135) = .357, p = .551 (experimental M = 4.355, control M = 4.492). The effect of lineup identification was non-significant, F (1, 135) = .386, p = .536 (correct M = 4.495, incorrect M = 4.353). The interaction between condition and lineup identification was also non-significant, F (1, 135) = .134, p = .715. Finally, we computed the confidence-accuracy correlations for the experimental (n =72), r = .021, p = .864 and control conditions (n = 67), r = .091, p = .462.