Table 1.

Descriptive statistics of stress, negative emotion, rumination and interference by intervention group and burst

		Descriptive statistics										Intervention and burst effects
Variable	Burst	Total					Control		Intervention
		N	M	SD	Range	N	M	SD	N	M	SD		F statistic	p
Stress
Family conflict
	1	1373	6.41	14.60	0–100	777	6.38	14.51	596	6.46	14.74	Intervention	F(1, 48)=2.23	0.142
	2	1115	8.06	16.96	0–100	610	6.12	15.55	505	10.41	18.26	Burst	F(2, 91)=0.56	0.575
	3	953	8.61	17.03	0–100	471	7.07	14.14	482	10.11	19.33	Intv × Burst	F(2, 3383)= 1.05	0.349
Occupational problems
	1	1371	27.93	31.04	0–100	776	24.48	29.22	595	32.44	32.74	Intervention	F(1, 48)=1.90	0.175
	2	1114	29.02	32.91	0–100	610	23.46	32.06	504	35.74	32.71	Burst	F(2, 91)= 1.05	0.353
	3	952	24.18	30.29	0–100	470	19.33	27.18	482	28.92	32.38	Intv × Burst	F(2, 3379)=1.02	0.359
Emotion regulation
Negative Emotion
	1	1381	20.54	21.01	0–100	782	18.00	19.47	599	23.85	22.45	Intervention	F(1, 48)=1.79	0.187
	2	1122	21.01	21.93	0–100	613	17.78	21.80	509	24.91	21.47	Burst	F(2, 91)=0.65	0.526
	3	956	20.59	21.12	0–100	473	17.09	19.00	483	24.02	22.49	Intv × Burst	F(2, 3401)=0.83	0.434
Rumination
	1	1374	27.12	27.25	0–100	778	22.70	25.21	596	32.89	28.72	Intervention	F(1, 48)=3.28	0.076
	2	1116	25.37	25.52	0–100	611	20.95	24.23	505	30.71	26.03	Burst	F(2, 91)=0.76	0.469
	3	953	25.27	25.51	0–100	471	21.51	23.88	482	28.94	26.53	Intv × Burst	F(2, 3385)=0.75	0.474
Interference
	1	1376	24.26	27.36	0–100	779	21.54	25.52	597	27.80	29.24	Intervention	F(1, 48)=1.84	0.181
	2	1119	23.75	26.62	0–100	612	20.19	25.36	507	28.04	27.49	Burst	F(2, 91)=0.85	0.431
	3	954	22.91	26.37	0–100	473	21.55	25.34	481	24.24	27.30	Intv × Burst	F(2, 3391)= 1.76	0.172

Note: Intervention and burst effects were tested in four-level regression models that included random intercepts at the level of the days, bursts and persons and first-order autoregressive covariance structure for the residuals. Models controlled for weekend status, day in burst, history of therapy and gender. Outcomes were square-root transformed in the regression models to reduce positive skew.