<h2>Results and Discussion</h2>
<p>n = 111 - 7 were excluded because they were too old, 1 was excluded because she was not Caucasian.
Of the remaining sample (n = 103, 60 female/43 male), 52 were in the experimental group and 51 were in the control group.
Participants had a mean age of 19.7 years (s.d. 1.5), ages ranged from 17 to 24 years.</p>
control group 72.5%
experimental group 71.2%</p>
<p>Chi-Square comparing percentage of correct identification
Chi squared <em>df</em> 1 = .03, <em>p</em> = .9, Phi = .016.
Recording a verbal description had no observable effect on identification accuracy.</p>
<p>False positive identification
control group 71.4%
experimental group 46.7%
Chi-Square comparing the ratio of the percentage selecting the wrong face (misidentification) to the percentage indicating “not present” across the experimental and control conditions.
Chi squared <em>df</em> 1 = 1.83, <em>p</em> = .2, Phi = .251.
There was also no signifiacnt effect of providing a verbal description on rates of false positive identification.</p>
Independent sampless 2 (Condition) x 2 (Correct vs. incorrect/miss) analysis of variance</p>
<p>Significant main effect of accuracy, <em>F</em>(1, 99) = 5.04,<em>p</em> = .03, partial eta squared = .05.
No significant main effect of condition (p = .8) and no significant interaction (p = .7).
Participants who chose correctly were more confident than participants who were incorrect. However, this is a small effect, accounting for only 5% of the variance in confidence ratings.</p>
<p>The Stirling University leg of the Schooler & Engstler Schooler (1990) replication did not not find any evidence of verbal overshadowing effect when the verbal description is provided immediately after viewing the robbery video, with a subsequent 20 minute filler task before providing an identification response.</p>