. 2019 Jul 3;10:1413. doi: 10.3389/fpsyg.2019.01413

Table 3.

Effects on BDS performance from pre- to post environment exposures by study sample, and aggregated time (pre- vs. post environment) by environment conditions (nature vs. urban) by order effects on BDS across study samples.

Study sample			Nature 1st/Urban 2nd		Urban 1st/Nature 2nd		Other		*Time^ Env**		*Time ^ Env ^* Order**
	N	*t-test*	Pre-post Nature (S1)	Pre-post Urban (S2)	Pre-post Nature (S2)	Pre-post Urban (S1)	Pre-post (S1)	F-test	F	*$E t a_{p}^{2}$*	F	*$E t a_{p}^{2}$*
1. Walk, UM, Berman et al. (2008)	37	t	3.343	0.000	3.422	2.452		F	5.670	0.139	3.193	0.084
		df	19	19	16	16		df	1, 35		1, 35
		p	0.003	1.000	0.003	0.026		p	0.023		0.083
2. Picture study, UM, Berman et al. (2008)	12	t	1.369	1.557	2.803	0.667		F	0.486	0.046	0.953	0.087
		df	5	5	5	5		df	1, 10		1, 10
		p	0.229	0.180	0.038	0.534		p	0.501		0.352
3. Walk, healthy sample, UM, 2011.	21	t	3.684	1.701	1.090	1.861		F	0.033	0.002	1.313	0.065
		df	10	10	9	9		df	1, 19		1, 19
		p	0.004	0.120	0.304	0.096		p	0.858		0.266
4. Walk, MDD sample, UM, Berman et al. (2012)	19	t	1.835	−3.000	3.308	1.060		F	18.661	0.523	0.703	0.040
		df	8	8	9	9		df	1, 17		1, 17
		p	0.104	0.017	0.009	0.317		p	0.000		0.414
5. Picture study, UC, 2015.	45	t	1.576	−1.577	−0.436	1.141		F	0.423	0.010	5.049	0.105
		df	20	20	23	23		df	1, 43		1, 43
		p	0.131	0.130	0.667	0.266		p	0.519		0.030
6. Picture study, UM, 2015.	37	t	2.377	−1.513	0.907	2.158		F	1.844	0.050	6.495	0.157
		df	15	15	20	20		df	1, 35		1, 35
		p	0.031	0.151	0.375	0.043		p	0.183		0.015
7. Walk, UC, 2016.	49	t	1.283	−1.489	0.871	1.528		F	0.610	0.013	1.476	0.030
		df	25	25	22	22		df	1, 47		1, 47
		p	0.211	0.149	0.393	0.141		p	0.439		0.230
8. Virtual Reality 1, UC, 2016.	82	t	2.314	0.521	1.513	2.855		F	0.015	0.000	3.621	0.043
		df	39	39	41	41		df	1, 80		1, 80
		p	0.026	0.605	0.138	0.007		p	0.903		0.061
9. Virtual Reality 2- with habituation, UC, 2016.	82	t		−1.657	0.516		3.169	F	2.393	0.029
		df		39	41		81	df	1, 80
		p		0.106	0.609		0.002	p	0.126
10. Composite study- Sounds, UC, Van Hedger et al. (2018)	44	t	1.889			0.502		F	0.692	0.016
		df	21			21		df	1, 42
		p	0.073			0.621		p	0.410
11. Composite study- Pictures, UC, 2016.	40	t	1.022			1.788		F	0.221	0.006
		df	18			20		df	1, 38
		p	0.320			0.089		p	0.641
12. Video study, UBC, Bourrier et al. (2018)	60 (+30)	t	3.467			0.128	1.223	F	3.695	0.060
		df	29			29	29	df	1, 58
		p	0.002			0.899	0.231	p	0.059
13. Picture dose study- Session 1 and 2, UM, 2009.	39	t	3.620		1.651			F
		df	38		36			df
		p	0.001		0.107			p
Total, studies with env as within-subj factor (study 1–8)	302	t	5.833	−0.955	3.467	5.017		F	5.989	0.020	17.78	0.056
		df	148	148	152	152		df	1, 300		1, 300
		p	0.000	0.341	0.001	0.000		p	0.015		0.000
Total, studies with env as betw-subj factor (study 9–12)	226	t	3.793	−1.657	0.516	1.183	3.341	F	4.223	0.019
		df	70	39	41	72	111	df	1, 224
		p	0.000	0.106	0.609	0.241	0.001	p	0.041
Grand Total	528^‡	t	6.966	−1.673	3.320	4.64
		df	219	188	194	225
		p	0.000	0.096	0.001	0.000
		M diff	0.968	−0.228	0.510	0.725
		SE	0.139	0.136	0.154	0.156
	95% CI	Lower	0.694	−0.496	0.207	0.417
		Upper	1.242	0.041	0.813	1.032

F-statistics are from factorial general linear models with BDS as the dependent variable and the factors time

environment, and additionally time^*environment^*order for within-subjects study designs. Time is always a within-subjects factor. Environment is a within-subjects factor in the randomized crossover studies, and a between-subjects factor in the randomized controlled studies. Order of environment conditions is a between-subjects factor in the randomized crossover studies. M diff = mean difference between BDS performance scores pre- vs. post the environment interaction. Statistically significant (p < 0.05) t-values and F-values are marked in bold. UM, Univ. of Michigan; UC, Univ. of Chicago; UBC, Univ. of British Columbia; S1, test session 1; S2, test session 2.

^‡

Excludes the Picture dose study sample.