Table 2.

Analysis of the brief-access lick data in Figure 1

		Latency to initiate licking			Lick rate
Test solutions	Source of variation	F-ratio	df	P-value	F-ratio	df	P-value
E and water	Test solution	22.8	1, 15	<0.0003	78.9	1, 15	<0.0001
	Consecutive trials	2.4	4.5, 67.2	0.051	7.6	4.4, 66.3	<0.0001
	Interaction	0.9	4.8, 72.6	0.515	12.9	4.7, 70.2	<0.0001
E and E + SS	Test solution	11.9	1, 15	<0.005	20.9	1, 15	<0.001
	Consecutive trials	1.3	5.3, 80.6	0.279	3.0	5.0, 75.3	0.016
	Interaction	1.2	4.7, 69.9	0.287	1.2	3.9, 59	0.300
E and E + SSM	Test solution	2.1	1, 15	0.173	10.4	1, 15	0.006
	Consecutive trials	2.8	2.2, 33.5	0.067	6.4	4.7, 71.1	<0.0001
	Interaction	1.1	3.8, 57.1	0.350	4.0	5.5, 81.9	0.002
E + SS and E + SSM	Test solution	3.2	1, 11	0.102	27.0	1, 11	0.0003
	Consecutive trials	1.9	5.9, 54.9	0.105	7.0	3.3, 36.1	0.0006
	Interaction	1.6	5.4, 59.6	0.158	0.7	2.4, 26.9	0.519
E and E + SSM2	Test solution	13.5	1, 15	0.002	123.1	1, 15	<0.0001
	Consecutive trials	1.8	5.4, 81.3	0.114	5.8	4.9, 73.5	0.0002
	Interaction	1.4	5.8, 88.1	0.223	8.4	4.7, 70.3	<0.0001
E and E + SSM4	Test solution	27.2	1, 15	0.0001	42.4	1, 15	<0.0001
	Consecutive trials	0.9	4.4, 66.5	0.467	0.9	3.9, 58.7	0.449
	Interaction	0.3	4.6, 68.7	0.877	1.4	2.9, 44.1	0.265

In each lick test, the rat was offered two test solutions. We examined how (a) latency to initiate licking and (b) lick rate for each test solution changed across the initial 20 trials of the lick test (mean ± standard error). We alternated the presentation of each of the test solution, resulting in a total of 10 trials per test solution. For each dependent measure and pair of test solutions, we ran a 2-way repeated-measure ANOVA. We controlled for sphericity by adjusting dfs with the Greenhouse–Geisser correction.