Government and charity funding of cancer research: public preferences and choices

Table 4 Random effects model results^a

	Best-fitting model (all respondents)			Best-fitting model (sub-sample randomised to ordering 12345)			Best-fitting model (sub-sample randomised to ordering 12534)			Best-fitting model (all respondents) plus variable for ordering
	Coefficient		Std. Err.	Coefficient		Std. Err.	Coefficient		Std. Err.	Coefficient		Std. Err.
Constant	47.39	***	4.54	50.94	***	6.76	43.87	***	6.21	43.38	***	4.70
Treatment effects (dummy variables)
Scenario 2	4.59	***	0.78	5.05	***	1.09	4.14	***	1.12	4.59	***	0.78
Scenario 3	9.55	***	0.78	10.47	***	1.09	8.64	***	1.12	9.54	***	0.78
Scenario 4	8.75	***	0.78	9.86	***	1.09	7.66	***	1.12	8.75	***	0.78
Scenario 5	2.87	***	0.78	2.94	***	1.09	2.79	**	1.12	2.87	***	0.78
Personal experience of cancer (YES = 1)	5.69	**	2.81	4.71		4.29	6.29	*	3.74	5.53	**	2.82
Personal donation to Cancer Research UK in the last year (YES = 1)	10.38	***	2.57	9.37	**	3.75	11.28	***	3.56	10.34	***	2.57
Personal donation (any type) in the last year (YES = 1)	−7.58	*	4.11	−12.28	**	5.77	−1.81		5.89	−7.42	*	4.11
Guess of percentage of government spending on medical research that is on cancer research	9.95		6.33	17.27	*	9.06	0.64		8.91	9.94		6.32
Order in which scenarios were presented (12345 = 0; 12534 = 1)	Not included			Not included			Not included			2.06		2.46
Number of observations (panel data)	2005			990			1015			2005
Number of respondents	401			198			203			401
Likelihood-ratio test – χ²	219.85	***		136.10	***		92.10	***		220.55	***

Significant to the: ***=1% level; **=5% level; *=10% level.
^aFurther details of the regression and the results of other variables tested are available on request.

ISSN: 1478-4505