Table 1 Parameters of test statistics.

From: Repeated behavioral testing and the use of summary measures reveal trait anxiety in preclinical rodent models

Figure / result	Hypotheses testing / model	comp.	var.	Test statistics with df	n	effect size	p	Further information
Figure 1B/C/D	repeated measures ANOVA / rat male main effect	EPM	time(s)	F2,107 = 34.206	54	0.800	0.000
		EPM	freq	F2,107 = 26.517		0.703	0.000	Supp. Fig. 1A
		OF	time(s)	F2,108 = 3.643		0.260	0.029
		OF	freq	F2,108 = 2.235		0.204	0.112	Supp. Fig. 1C
		LD	time(s)	F2,106 = 0.801		0.123	0.452
		LD	freq	F2,106 = 4.119		0.279	0.019	Supp. Fig. 1B
	aversive contrast paired t-test (male rats)	EPM 1 vs 4 (aversive)	time(s)	t76 = 8.007		0.920	0.000
		EPM 1 vs 4 (aversive)	freq	t82 = 6.738		0.742	0.000	Supp. Fig. 1A
		EPM 3 vs 4 (aversive)	time(s)	t94 = 2.898		0.299	0.005
		EPM 3 vs 4 (aversive)	freq	t94 = 2.748		0.283	0.007	Supp. Fig. 1A
		OF 1 vs 4 (aversive)	time(s)	t88 = 2.894		0.309	0.005
		OF 1 vs 4 (aversive)	freq	t106 = 4.643		0.451	0.000	Supp. Fig. 1C
		OF 3 vs 4 (aversive)	time(s)	t102 = 4.043		0.400	0.000
		OF 3 vs 4 (aversive)	freq	t104 = 4.9		0.479	0.000	Supp. Fig. 1C
		LD 1 vs 4 (aversive)	time(s)	t106 = −0.315		0.031	0.753
		LD 1 vs 4 (aversive)	freq	t107 = −0.460		0.045	0.646	Supp. Fig. 1B
		LD 3 vs 4 (aversive)	time(s)	t106 = 0.356		0.035	0.722
		LD 3 vs 4 (aversive)	freq	t105 = 1.078		0.105	0.284	Supp. Fig. 1B
	repeated measures ANOVA / rat female main effect	EPM	time(s)	F2,52 = 3.419	27	0.363	0.040
		EPM	freq	F2,52 = 10.343		0.631	0.000	Supp. Fig. 1A
		LD	time(s)	F2,52 = 0.038		0.038	0.963
		LD	freq	F2,52 = 16.647		0.800	0.000	Supp. Fig. 1B
	aversive contrast paired t-test (female rats)	EPM 1 vs 4 (aversive)	time(s)	t51 = 7.989		0.853	0.000
		EPM 1 vs 4 (aversive)	freq	t49 = 0.595		1.121	0.000	Supp. Fig. 1A
		EPM 3 vs 4 (aversive)	time(s)	t52 = 4.431		0.616	0.000
		EPM 3 vs 4 (aversive)	freq	t51 = 3.609		0.506	0.001	Supp. Fig. 1A
		LD 1 vs 4 (aversive)	time(s)	t50 = 0.594		0.085	0.555
		LD 1 vs 4 (aversive)	freq	t52 = 1.712		0.243	0.093	Supp. Fig. 1B
		LD 3 vs 4 (aversive)	time(s)	t50 = 0.675		0.096	0.503
		LD 3 vs 4 (aversive)	freq	t52 = −2.073		0.289	0.043	Supp. Fig. 1B
	repeated measures ANOVA / mouse main effect	EPM	time(s)	F2,78 = 64.944	40	1.290	0.000
		EPM	freq	F2,78 = 74.446		1.382	0.000	Supp. Fig. 1A
		OF	time(s)	F2,77 = 8.386		0.466	0.001
		OF	freq	F2,77 = 23.279		0.779	0.000	Supp. Fig. 1C
		LD	time(s)	F2,78 = 1.411		0.190	0.250
		LD	freq	F2,77 = 0.971		0.159	0.383	Supp. Fig. 1B
	aversive contrast paired t-test (mouse)	EPM 1 vs 4 (aversive)	time(s)	t59 = 9.806		1.272	0.000
		EPM 1 vs 4 (aversive)	freq	t77 = 12.936		1.478	0.000	Supp. Fig. 1A
		EPM 3 vs 4 (aversive)	time(s)	t62 = 1.192		0.152	0.238
		EPM 3 vs 4 (aversive)	freq	t78 = 1.139		0.129	0.258	Supp. Fig. 1A
		OF 1 vs 4 (aversive)	time(s)	t59 = 1.774		0.230	0.081
		OF 1 vs 4 (aversive)	freq	t54 = 9.281		1.265	0.000	Supp. Fig. 1C
		OF 3 vs 4 (aversive)	time(s)	t74 = −0.289		0.034	0.773
		OF 3 vs 4 (aversive)	freq	t71 = 5.077		0.604	0.000	Supp. Fig. 1C
		LD 1 vs 4 (aversive)	time(s)	t78 = 2.468		0.280	0.016
		LD 1 vs 4 (aversive)	freq	t76 = 3.0342		0.349	0.003	Supp. Fig. 1B
		LD 3 vs 4 (aversive)	time(s)	t76 = 1.054		0.121	0.295
		LD 3 vs 4 (aversive)	freq	t78 = 1.6831		0.191	0.096	Supp. Fig. 1B
Figure 2A	Linear-mixed model’s	complexity	corr coeff	t = 3.597; Est = 0.019; SE = 0.005	54	-	0.001	variable composition: p = 0.0065
								test composition: p = 0.0088
								Residual: 0.0042
Figure 2G	ANOVA	freezing	time(s)	F2,25 = 7.000	28	0.748	0.001
	Tukey post-hoc	conditioned context (ctx) vs safe ctx1		t135 = 2.247; Est = 7.69; SE = 3.42		0.390	0.067
		cond ctx vs safe ctx2		t135 = 3.715; Est = 12.71; SE = 3.42		0.640	0.001
		safe ctx1 vs safe ctx2		t135 = 1.468; Est = 5.02; SE = 3.42		0.250	0.309
Figure 3B	Wilcoxon rank sum test	EPM 1	anxiety score	w = 86	30	0.490	0.143	FDR: non-significant
		EPM 2		w = 74		0.440	0.056
		EPM 3		w = 97		0.050	0.263
		EPM 4(aversive)		w = 34		0.480	0.023
Figure 3C	Wilcoxon rank sum test	EPM 1-2		w = 37		0.640	0.023	FDR: all significant
		EPM 1-2-3		w = 36		0.500	0.034
		EPM 1-4(aversive)		w = 34		0.620	0.020
Figure 3E	Wilcoxon rank sum test	LD 1		w = 41		0.990	0.001	FDR: all significant
		LD 2		w = 50		0.810	0.004
		LD 3		w = 42		0.890	0.001
		LD 4 (aversive)		w = 43		1.030	0.002
Figure 3F	Wilcoxon rank sum test	LD 1-2		w = 38		0.960	0.001	FDR: all significant
		LD 1-2-3		w = 43		0.970	0.002
		LD 1-4(aversive)		w = 33		1.100	0.000
Figure 3H	Wilcoxon rank sum test	COMP 1		w = 52		0.900	0.006	FDR: all significant
		COMP 2		w = 52		0.756	0.006
		COMP 3		w = 64		0.653	0.023
		COMP 4(aversive)		w = 39		0.969	0.001
Figure 3I	Wilcoxon rank sum test	COMP 1-2		w = 37		0.992	0.001	FDR: all significant
		COMP 1-2-3		w = 36		0.944	0.000
		COMP 1-4(aversive)		w = 34		1.089	0.000
Supp. Fig. 3A	Wilcoxon rank sum test	EPM baseline vs aversive	anxiety score	w = 1512.5	54	0.374	0.006
		OF baseline vs aversive		w = 1813		0.659	0.000
		LD baseline vs aversive		w = 1229		0.076	0.575
Supp. Fig. 3B	ANOVA	corticosterone	concentration	F = 44.225	28	1.844	0.000
	Tukey post-hoc	aversive vs resting 1		t81 = 8.407; Est = 426; SE = 50.6		1.870	0.000
		aversive vs resting 2		t81 = 7.854; Est = 398; SE = 50.6		1.750	0.000
		Resting 1 vs resting 2		t81 = −0.554; Est = −28; SE = 50.6		0.120	0.845

Table 1 shows (left-to-right) the result section/figure number corresponding to the statistical parameters, the hypotheses testing method/model that were used, the group comparisons that were made, the variables that were assessed, and the resulting test statistics-values with degrees of freedom, sample sizes, effect sizes and significance levels. We presented the conventional parameters of the particular statistical approaches such as F values in the case of ANOVAs, t or w values in the case of t-tests or Wilcoxon tests, as well as estimates (est) and standard errors (SE) in the case of linear mixed models. Numbers following the test statistics identifier (e.g. F or t) are the within and between group degrees of freedom values, separated by a comma. Effect sizes are Cohen’s f for ANOVA and Cohens’s d for t-tests. In the case of Wilcoxon test, we calculated effect sizes as the difference between the means of compared groups, divided by the standard deviation of those two groups. P values of 0.000 represent values below 0.0005. The column entitled Further information shows additional relevant information that is specific for a particular method, like random variances in mixed models or the results of FDR correction following multiple comparisons.

Back to article page

Table 1 Parameters of test statistics.

Search

Quick links