STUDY 1 ---------------------- Results ---------------------- **Recognition accuracy** Recognition accuracy was analyzed in the same way as in the original (Schooler & Engstler-Schooler, 1990) paper. Describing the previously seen face verbally did not affect recognition performance. The target face was correctly picked by 72.7% of participants in the Face Verbalization condition and 69.1% of participants in the control condition, *χ²* (1) = .18, *p* > .05. The proportion of misidentification and misses did not differ among both conditions. The relative ratio of misidentifications to misses was the same for the two conditions. Approximately 26.7% of the errors in the Face Verbalization condition were misidentifications compared to 41.2% of errors in the Control condition, *χ²* (l) = .74, *p* > .05. **Confidence** A 2 (Condition: experimental vs. control) x 2 (Correctness: correct vs. incorrect/miss) ANOVA on confidence ratings revealed a significant main effect of correctness, *F*(1, 106) = 8.76, *p* < .01, *η²* = .08. Participants were significantly more confident in correct responses (*M* = 5.49; *SD* = 1.07) compared to incorrect responses (*M* = 4.84; *SD* = .99). Neither main effect of condition, *F*(1, 106) = 0.12, p > .05, nor interaction effect of condition and correctness were significant, *F*(1, 106) = 1.36, *p* > .05. Mean confidence scores in responses depending on experimental condition and response correctness are presented in Table 1. Table 1. Mean confidence scores in responses depending on experimental condition and response correctness (SD in parentheses) ![Table 1][1] **Additional Analyses** We have calculated the number of words in each of the descriptions provided by participants in the experimental group. These numbers were used as a dependent variable to probe the relation between accuracy, type of error and elaboration of the verbal description. Because of unequal number of participants in the groups, we used non-parametric tests. The results showed that elaboration measured by the number of words was not related to the correctness (*U* = 296; *p* > .05). Also the type of error (misidentification or miss) was not related to elaboration measured by the number of words (*U* = 15; *p* > .05). STUDY 2 ---------------------- Results ---------------------- **Recognition accuracy** Recognition accuracy was analyzed in the same way as in the original (Schooler & Engstler-Schooler, 1990) paper. The target face was correctly picked by 57.4% of participants in the Face Verbalization condition and 74.5% of participants in the control condition. This difference approached significance, *χ²* (1) = 3.03, *p* = .08. The proportion of misidentification and misses did not differ among both conditions. The relative ratio of misidentifications to misses was the same for the two conditions. Approximately 40% of the errors in the Face Verbalization condition were misidentifications compared to 58.3% of errors in the Control condition, *χ²* (l) = 1.01, *n.s.* **Confidence** A 2 (Condition: experimental vs. control) x 2 (Correctness: correct vs. incorrect/miss) ANOVA on confidence ratings revealed a significant main effect of correctness, *F*(1, 90) = 12.53, *p* = .001, *η²* = .12. Participants were significantly more confident in correct responses (*M* = 5.24; *SD* = 1.11) compared to incorrect responses (*M* = 4.50; *SD* = 1.14). The main effect of condition was not significant, *F*(1, 90) = 2.14, *n.s.* However, interaction effect of condition and correctness was also significant, *F*(1, 90) = 8.82, *p* < .01, *η²* = .09. In the Control condition participants were more confident in the correct responses (*M* = 5.40; *SD* = 1.06) than in the incorrect ones (*M* = 3.83; *SD* = 1.27; *t*(45) = -4.20; *p *< .001). In the Face Verbalization condition confidence did not differ between the participants responding correctly and incorrectly, *t*(45) = -.45; *n.s.* Moreover, participants who did not correctly recognize the perpetrator were more confident in their responses in the Face Verbalization condition (*M* = 4.90; *SD* = .85) than in the Control condition (*M* = 3.83; *SD* = 1.27; *t*(32) = 2.85; *p* < .01). Confidence in the correct responses did not differ between the conditions, *t*(62) = -1.28; *n.s*. Mean confidence scores in responses depending on experimental condition and response correctness are presented in Table 2. Table 2. Mean confidence scores in responses depending on experimental condition and response correctness (*SD* in parentheses) ![Table2][2] **Additional Analyses** We have calculated the number of words in each of the descriptions provided by participants in the experimental group. These numbers were used as a dependent variable to probe the relation between accuracy, type of error and elaboration of the verbal description. The mean number of words used in the descriptions of those participants who were correct consisted of the similar number of words to those who were incorrect. However, the type of error (misidentification or miss) was related to elaboration measured by the number of words on the trend level (*U* = 26; *p* = .09). Descriptions of the participants who misidentified the perpetrator consisted of higher number of words (*M* = 56.63; *SD* = 19.11), than descriptions of those participants who did not indicate any of the pictures (*M* = 42.58; *SD* = 20.55). DISCUSSION ---------------------- The results of Study 1 show that verbalizing did not affect recognition accuracy. Verbalizing participants’ memories of the face did not impair their ability to pick the correct one, and the percentage of correct responses was as high as in the control condition. These results were similar to those obtained by Schooler and Engstler-Schooler (1990) in control conditions (experiments 1, 2 and 4). On the other hand, the results of Study 2 reveal weak effect (tendency) of verbalizing on recognition accuracy. These results are consistent with previous analyses reporting null or weak verbal overshadowing effect (Meissner & Brigham, 2001; Memon & Bartlett, 2002). As the meta-analysis of Meissner and Brigham, (2001) also revealed, verbal overshadowing may be moderated by a type of instruction: the effect is stronger when elaborative instructions are given and when identification task immediately follows description task (Meissner & Brigham, 2001). In our replication the two studies differed in terms of procedure: in Study 1 description and identification task were separated by crossword puzzle but verbalizing immediately followed encoding and in Study 2 recognition immediately followed verbalizing task, however encoding and verbalizing were separated by an additional task (crossword). These results suggest that verbal overshadowing effect may be more likely to occur when verbalizing is delayed and immediately followed by recognition. Moreover, in both studies elaboration in terms of number of words used in the descriptions did not differ between these participants who correctly recognized the face of the perpetrator and those who did not. However, the results of Study 2 show that descriptions of those who misidentified the perpetrator were more elaborated than descriptions of those who did not choose any of the individuals presented in the line-up. In line with original results, we also found in both studies that correctness was related to confidence judgments. In Study 1 these participants who identified the face correctly were also more confident of their recognition. In Study 2 this effect was significant only among participants in the control condition, while these participants who were asked to provide verbal descriptions of the perpetrator’s face were equally confident in both correct and incorrect answers. Such results are in contrast with the results reported in eyewitness literature (e.g., Kassin, 1985; Kassin, Rigby, & Castillo, 1991) showing negligible relation between accuracy and confidence (which however may be strengthened by increasing time spent on studying perpetrator’s face as suggested by Bothwell, Deffenbacher, & Brigham, 1987) and other forensic psychology domains (e.g., deception detection; DePaulo, Charlton, Cooper, Lindsay, & Muhlenbruck, 1997 for a review). To sum up, this line of studies shows no or weak verbal overshadowing effect, what contradicts the original results of Schooler and Engstler-Schooler (1990). Importantly however, our results also suggest that this effect depends on the exact procedure and task order (Meissner & Brigham, 2001). The discrepancy between Study 1 and Study 2 emphasizes the importance of replication studies in order to estimate the actual size of verbal overshadowing effect under various types of instructions and order. From the point of view of law-enforcement practitioners, it is crucial to know whether the commonly used procedure of obtaining a description of a suspect before a line-up presentation (Memon & Bartlett, 2002) produces systematic identification errors and if it does, under what circumstances these errors are most likely to occur. **References** Bothwell, R. K., Deffenbacher, K. A., & Brigham, J. C. (1987). Correlation of eyewitness accuracy and confidence: Optimality hypothesis revisited. *Journal of Applied Psychology, 72*, 691-695. DePaulo, B. M., Charlton, K., Cooper, H., Lindsay, J. L., & Muhlenbruck, L. (1997). The accuracy-confidence correlation in the detection of deception. *Personality and Social Psychology Review, 4,* 346-357. Meissner, C. A., & Brigham, J. C. (2001). A meta-analysis of the verbal overshadowing effect in face identification. *Applied Cognitive Psychology 15*, 603–616. Memon, A., & Bartlett, J. (2002). The effects of verbalization on face recognition in young and older adults. *Applied Cognitive Psychology, 16,* 635–650. Kassin, S. M. (1985). Eyewitness identification: Retrospective self-awareness and the accuracy-confidence relationship. *Journal of Personality and Social Psychology, 49*, 878-893. Kassin, S. M., Rigby, S., & Castillo, S. R. (1991). The accuracy-confidence correlation in eyewitness testimony: Limits and extensions of the retrospective self-awareness effect. *Journal of Personality and Social Psychology, 61*, 698-707. Schooler, J. W., & Engstler-Schooler. T. Y. (1990). Verbal overshadowing of visual memories: some things are better left unsaid. *Cognitive Psychology, 22*, 36–71. [1]: http://imageshack.com/a/img585/346/oa2t.jpg [2]: http://i61.tinypic.com/vg51rb.jpg