Self-monitoring						Reported calorie intake
	Estimate	SE	df	t	p	Estimate	SE	df	t	p
Intercept	0.66	0.16	315.5	4.09	<0.001***	0.60	0.16	307.8	3.69	<0.001***
Session number	–0.53	0.005	6685	–104.78	<0.001***	–0.53	0.005	6658	–95.04	<0.001***
Discrepancy change	–0.88	0.46	93.65	–1.91	0.06	–0.67	0.35	175.9	–1.92	0.06
Responder status	–3.25	0.29	429.7	–14.55	<0.001***	–3.24	0.22	431.9	–14.47	<0.001***
Clinician-rated adherence	–0.48	0.04	6861	–11.91	<0.001***	–0.45	0.04	6857	–11.15	<0.001***
Discrepancy change × session	0.16	0.02	6651	8.94	<0.001***	0.14	0.01	6608	9.78	<0.001***

* p < 0.05. Discrepancy change scores were computed as the mean difference in discrepancy score from Sessions 1–10 to Sessions 11–20 for each subject and for each adherence measure. Mean change to self-monitoring adherence discrepancy from first to second half of treatment was 0.23 (SD = 0.82); mean change to reported calorie intake discrepancy from first to second half of treatment was 0.54 (SD = 1.18). All models included subject and clinician as a random effect, included a random slope and intercept for Discrepancy Score, used an unstructured covariance matrix and restricted maximum likelihood (REML) estimation. Satterwhaite approximations were used to estimate degrees of freedom and p-values. Session Number ranged from 1 to 20 and responder status was effect-coded (0 = non-responder at either Week 3 or Week 7, 1 = responder). For self-monitoring, AIC = 28543.6, BIC = 28632.73. For reported calorie intake, AIC = 28527.3, BIC = 28616.42.