Table 5.

Inferential Statistics for 2 × 2 χ² tests comparing endorsements (agree versus disagree) in experts and laypeople (df = 1)

Statement	n	χ2	p	Adjusted p	φ
1. High stress impairs*	177		>.999	.078	.124
2. Police interview	168	0.058	.810	.078	.019
3. Stress impairs retrieval	168	2.959	.085	.078	.133
4. Stress enhances encoding	165	5.361	.021	.078	.180
5. Children less affected	142	12.857	<.001	<.001	.301
6. Detail type	156	15.613	<.001	<.001	.316
7. Test type	150	0.959	.327	.078	.080
8. Faces affected differently	130	22.472	<.001	<.001	.416
9. Professionals less affected	172	44.109	<.001	<.001	.506
10. Victims more affected	163	1.837	.175	.078	.106
11. Repression	165	94.295	<.001	<.001	.756
12. Violent events	148	6.463	.011	.078	.209
13. Emotional better remembered	149	1.992	.158	.078	.116
14. Moderate stress	155	20.789	<.001	<.001	.366
15. Severe stress	163	0.984	.321	.078	.078
16. Short crime	156	24.480	<.001	<.001	.396
17. Abstractness	151	7.704	.006	.078	.226
18. Immediate retrieval enhances	134	5.612	.018	.078	.205
19. Retrieval timing	145	1.575	.210	.078	.104
20. Misinformation protection	135	19.721	<.001	<.001	.382
21. Older adults less affected	126	2.984	.084	.078	.154

Note. Adjusted p = Holm–Bonferroni adjustment for multiple comparisons. Bold = adjusted p significant at the .05 level. * = Fisher’s exact test instead of chi-square test (when expected cell sizes <5)