Table 3.

Results From Post Hoc Mixed-Effects Models Comparing Unexpected Circa Strike (CS) With Other Threat-Imminence Levels

Outcome and model	Estimate	SE	t	p	95% CI	R 2	σ2	τ00
SCR frequency
A. Unexpected CS vs. no CS	0.48	0.16	3.10	.002	[0.18, 0.79]	.61	20.15	28.35
B. Unexpected CS vs. expected CS	0.55	0.29	1.91	.057	[−0.02, 1.12]	.60	21.05	28.13
C. Unexpected CS vs. combined CS	−0.59	0.21	−2.74	.006	[−1.01, −0.17]	.62	19.07	26.60
SCR amplitude
D. Unexpected CS vs. no CS	0.50	0.12	4.36	< .001	[0.28, 0.73]	.66	11.06	20.27
E. Unexpected CS vs. expected CS	0.56	0.02	2.80	.005	[0.17, 0.95]	.67	10.14	18.48
F. Unexpected CS vs. combined CS	−0.49	0.16	−3.08	.002	[−0.80, −0.18]	.66	10.55	17.91

Note: Models A to C are linear mixed-effects models including random intercepts for skin conductance response (SCR) frequency (number of SCRs) and restricting threat-imminence levels to unexpected CS and each of the other imminence levels. Models D to F are linear mixed-effects models including random intercepts for SCR amplitude (log μs) and restricting threat-imminence levels to unexpected CS and each of the other imminence levels. All models control for effects of time. For each model, the estimate is unstandardized, and the t is standardized. Bonferroni correction should be interpreted as p < .008. CI = confidence interval.