Table 2.

Unstandardized coefficients from multilevel week×condition interaction on daily steps (N=80)^a.

Outcome	Model 1b: steps with baseline				Model 2c: steps without baseline
	β	SE	P value	β		SE	P value
Intercept	5.13	2.61	.05	5.47		2.78	.05
Week	.24	0.05	<.001	–.12		0.07	.1
Age (years)	–.02	0.03	.64	–.01		0.04	.77
Sex	–1.13	0.52	.03	–1.16		0.55	.04
Condition	–.51	0.55	.35	–.57		0.63	.37
Education	–.03	0.10	.72	–.03		0.10	.75
Health	.03	0.01	.07	.03		0.01	.04
Days of app use	.03	0.05	.52	.04		0.05	.4
Week×condition interaction	.02	0.07	.76	.02		0.10	.82

^aDaily steps were rescaled by dividing the number of steps by 1000; therefore, the β value reflects change per 1000-step increase.

^bModel 1: level 1 variance=4.44 (SD 2.11); level 2 variance=6.05 (SD 2.46). Akaike information criterion=11319.5; Bayesian information criterion=11383.1; log likelihood=–5648.8.

^cModel 2: level 1 variance=4.98 (SD 2.23); level 2 variance=5.67 (SD 2.38). Akaike information criterion=9038.6; Bayesian information criterion=9099.8; log likelihood=–4508.3.