Table 6.

Zero inflated negative binomial model of the validation behaviour

	Count model (1) Corrections/day		Zeros model (2) Correction/day > 0
Constant	0.744	(0.032)${}^{\ast \ast \ast }$	1.504	(0.046)${}^{\ast \ast \ast }$
Phase 2	0.047	(0.014)${}^{\ast \ast }$	0.050	(0.020)${}^{\ast }$
Age (decades)	− 0.024	(0.003)${}^{\ast \ast \ast }$	− 0.014	(0.005)${}^{\ast \ast }$
Male	0.074	(0.012)${}^{\ast \ast \ast }$	0.047	(0.017)${}^{\ast \ast }$
Treatment
Control	–		–
Information	− 0.029	(0.022)	− 0.053	(0.032)
Pricing	− 0.083	(0.069)	− 0.335	(0.103)${}^{\ast \ast }$
Education
Mandatory	–		–
Trade/traineeship (baseline)	− 0.098	(0.023)${}^{\ast \ast \ast }$	− 0.220	(0.033)${}^{\ast \ast \ast }$
Higher education	− 0.014	(0.023)	− 0.321	(0.033)${}^{\ast \ast \ast }$
Income (CHF per month)
Less than 4000	–		–
4000 $\le$ 8000	− 0.134	(0.022)${}^{\ast \ast \ast }$	− 0.208	(0.032)${}^{\ast \ast \ast }$
8000 $\le$ 12,000	− 0.203	(0.022)${}^{\ast \ast \ast }$	− 0.324	(0.032)${}^{\ast \ast \ast }$
12,000 $\le$ 16,000	− 0.230	(0.024)${}^{\ast \ast \ast }$	− 0.429	(0.035)${}^{\ast \ast \ast }$
More than 16,000	− 0.124	(0.025)${}^{\ast \ast \ast }$	− 0.360	(0.038)${}^{\ast \ast \ast }$
Interactions
Control * male	–		–
Information * male	− 0.027	(0.028)	0.139	(0.040)${}^{\ast \ast \ast }$
Pricing * male	− 0.004	(0.027)	− 0.001	(0.040)
Pricing * mandatory	–		–
Pricing * trade/traineeship	− 0.113	(0.057)	0.099	(0.081)
Pricing * higher education	− 0.166	(0.057)${}^{\ast \ast }$	− 0.023	(0.082)
pricing * less than 4000	–		–
Pricing * 4000 $\le$ 8000	0.174	(0.059)${}^{\ast \ast }$	0.278	(0.084)${}^{\ast \ast \ast }$
Pricing * 8000 $\le$ 12,000	0.285	(0.058)${}^{\ast \ast \ast }$	0.354	(0.083)${}^{\ast \ast \ast }$
Pricing * 12,000 $\le$ 16,000	0.187	(0.065)${}^{\ast \ast }$	0.456	(0.092)${}^{\ast \ast \ast }$
Pricing * more than 16,000	0.128	(0.068)	0.368	(0.099)${}^{\ast \ast \ast }$
Observations	147,450
Log Likelihood	− 127,206.400

${}^{\ast \ast \ast }p<0.001$; ${}^{\ast \ast }p<0.01$; ${}^{\ast }p<0.05$