**Results and Discussion** Figures and Tables can be seen and uploaded in files. **Results: Verbalization versus Control** A chi-square test, comparing the proportion of correct identifications from the lineup between the experimental and control conditions was not significant, X2 (1, *N* = 113) = 2.08, p = 0.15. Participants in the control group correctly identified the robber from the lineup more frequently than participants in the verbalization group, consistent in direction with Schooler & Engstler-Schooler’s (1990) findings (see Figure 1). A chi-square test comparing the proportion of misidentifications (i.e., selection of an innocent person from the lineup) relative to misses (deciding that the perpetrator was not in the lineup) was significant, X2 (1, *N* = 41) = 4.78, p = 0.03. Verbalizers’ errors were primarily misses (75%), whereas controls’ errors were primarily misidentifications (59%), in contrast to Schooler and Engstler-Schooler (1990), who found no significant difference in error type. The comparison of errors between conditions is presented in Figure 2. A 2 (condition) X 2 (answer type: correct vs. incorrect) ANOVA on confidence ratings revealed a significant main effect of answer type, F (1, 113) = 6.99, p = .009. Participants who correctly identified the robber from the lineup reported being more confident in their selection than those who erred (Correct M = 5.13, Incorrect M = 4.52). There was no main effect of condition, nor an interaction (both Fs<1), consistent with Schooler and Engstler-Schooler’s (1990) findings (see Figure 3). **Results: Suppression versus Control** The three primary analyses were repeated when comparing the suppression condition to the control condition. A chi-square test comparing the proportion of correct identifications between the suppression and control conditions was not significant, Χ2 (1, *N* = 115) = 1.67 p = 0.20. Participants in the control condition correctly selected the robber from the lineup nonsignifcantly more frequently than participants in the suppression condition. The comparison of correct identifications between conditions is presented in Figure 1. A chi-square test comparing the proportion of misidentifications relative to misses was not significant, Χ2 (1, *N* = 41) = 0.001, p = 0.98. Participants in both the suppression condition and in the control condition misidentified faces more frequently than selecting “not present" (58% and 59% respectively) (see Figure 2). A 2 (condition) X 2 (answer type: correct vs. incorrect) ANOVA on confidence ratings revealed a significant main effect of answer type, F (1, 114) = 9.11, p = 0.003. Participants who correctly identified the robber from the lineup reported being more confident in their selection than those who erred (Correct M = 5.15, Incorrect M = 4.41). There was no main effect of condition, nor an interaction (both Fs < 1). The results are displayed in Figure 3. **Results: Suppression versus Verbalization** A chi-square test comparing the proportion of correct identifications between the suppression and verbalization conditions was not significant, Χ2 (1, *N* = 114) = 0.03, p = 0.87. Participants in the suppression condition correctly identified the robber at approximately the same rate (59%) participants in the verbalization condition (57%). The results are shown in Figure 1. A chi-square test comparing the proportion of misidentifications relative to misses was significant, Χ 2 (1, *N* = 48) = 5.49, p = 0.02. Participants’ errors in the suppression condition, like those in the control condition, were most often misidentifications (58%), whereas participants’ errors in the verabalization condition were most often misses (75%). The results are presented in Figure 2. A 2 (condition) X 2 (answer type: correct vs. incorrect) ANOVA on confidence ratings revealed a significant main effect of answer type, F (1, 113) = 7.33, p = 0.01. Participants who correctly identified the robber from the lineup reported being more confident in their selection than those who erred (Correct M = 5.11, Incorrect M = 4.52). There was no main effect of condition, nor an interaction (both Fs<1). Refer to Figure 3 for these results. **Other Analyses** Participants in the suppression condition were instructed to ring a bell whenever a thought about the bank robber intruded. The number of bell rings ranged from 0 to 13 (M = 3.54, SD = 3.19). The median was 3 and the mode was 0. There was a non-significant correlation between bell rings and identification accuracy, r(58)= -0.228, p = 0.085. A X2 analysis comparing the proportion of correct identifications between “successful” (no bell rings) and “unsuccessful” (at least one bell ring) suppressors was significant. Successful suppressors were more likely to correctly identify the robber in the lineup (82%) than unsuccessful suppressors (53%). See Figure 4 for results. The other two analyses were not repeated due to small sample size.