A randomised trial and economic evaluation of the effect of response mode on response rate, response bias, and item non-response in a survey of doctors

Table 4 The effect of mode on response rates (probit regression model)

	All doctors	GPs	Specialists	Hospital non-specialists	Specialists in training
	Marginal effects (standard error)
Simultaneous mixed	0.070***	0.028	0.160***	0.034	-0.030
	(0.020)	(0.030)	(0.036)	(0.046)	(0.060)
Sequential mixed	0.077***	0.053	0.133***	0.072	0.013
	(0.020)	(0.028)	(0.037)	(0.049)	(0.065)
< age30	0.074	0.086	-	0.276	0.108
	(0.046)	(0.156)	-	(0.160)	(0.166)
age30-39	0.037	-0.008	0.029	0.229	0.198*
	(0.027)	(0.040)	(0.048)	(0.163)	(0.085)
age50-59	0.054*	0.039	0.058	0.418	0.009
	(0.023)	(0.031)	(0.036)	(0.245)	(0.144)
age60-69	0.002	0.020	-0.018	-	-
	(0.026)	(0.038)	(0.038)
> = age70	-0.028	-0.063	0.005	-	-
	(0.035)	(0.042)	(0.058)
Age missing	-0.037	-0.034	-0.080*	0.208	0.127
	(0.025)	(0.036)	(0.039)	(0.190)	(0.125)
Female	-0.043**	-0.029	-0.066	-0.048	-0.080
	(0.017)	(0.024)	(0.036)	(0.037)	(0.056)
Specialist	0.053**	-	-	-	-
	(0.018)
Hospital non-specialist	-0.013	-	-	-	-
	(0.027)
Specialist in training	0.009	-	-	-	-
	(0.032)
Inner Regional	0.025	0.024	0.053	-0.050	0.140
	(0.023)	(0.030)	(0.045)	(0.054)	(0.105)
Outer Regional	0.051	0.119*	-0.056	-	-
	(0.040)	(0.052)	(0.064)
Remote	0.231	0.354	-	-	-
	(0.182)	(0.249)
Very Remote	0.517**	0.542**	-	0.459	-
	(0.163)	(0.185)		(0.315)
N	2702	1091	966	430	196
Log likelihood	-1255	-469	-461	-188	-84
χ² (df)	59.7*** (16)	28.3** (13)	37.4*** (10)	12.6 (9)	11.3 (8)
Pseudo R²	0.03	0.03	0.04	0.04	0.05

Marginal effects for discrete change of dummy variable from 0 to 1; * P < 0.05, ** P < 0.01, *** P < 0.001; standard errors are in parentheses.

ISSN: 1471-2288