Table 2.

Contrasts comparing low expectancy (L), partial reinforcement (P), and high expectancy (H).

		Expectancy	Uncertainty	Combined
Stage 1: Fear conditioning
Pupil dilation	t(62)	2.832	2.715	3.918
	p	0.003**	0.004∗∗	0.001***
	β	0.398	0.382	0.551
Dwell time	t(62)	4.000	2.205	4.468
	p	0.001***	0.016*	0.001***
	β	0.232	0.128	0.260
Target fixation latency	t(108)	-2.657	-0.639	-2.427
	p	0.011*	0.262	0.011*
	β	-0.313	-0.075	-0.286
Distracter fixation latency	t(108)	2.596	0.130	2.048
	p	0.013*	0.448	0.021*
	β	0.306	0.015	0.241
Capture frequency (by shape)	t(62)	-3.230	-0.203	-2.575
	p	0.003**	0.420	0.006**
	β	-0.687	-0.043	-0.548
Stage 2: Visual search task
Capture frequency (by color)	t(62)	3.172	0.195	2.525
	p	0.003**	0.423	0.007**
	β	0.249	0.015	0.200
Capture duration	t(21)	1.573	2.210	2.665
	p	0.065	0.019*	0.017*
	β	0.142	0.203	0.241

The expectancy contrast codes for a linear increase L < P < H with coefficients -1, 0, 1. The uncertainty contrast codes for the pattern L < P > H with coefficients -0.5, 1, -0.5. The third contrast explores the possibility that attention is influenced by a combination of both values and has coefficients -0.75, 0.50, 0.25. Tests are one-sided, to test for the specified patterns (and not for the inverse patterns); with ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001. Effect sizes are computed for standardized variables (strongest effect highlighted in bold).