Skip to main content

Table 3 Analysis procedure and statistical tests

From: Systematic review and validity assessment of methods used in discrete choice experiments of primary healthcare professionals

Item Category Number of studies %
Level coding discussed Yes 21 61.8%
No 13 38.2%
Goodness of fit considered Yes 20 58.8%
No 8 23.5%
Unclear or not reported 6 17.6%
Internal or external validity investigated Yes 10 29.4%
No 5 14.7%
Unclear or not reported 19 55.9%
Estimation model Conditional or multinomial logit 13 38.2%
Random effects probit 3 8.8%
Latent-class analysis 2 5.9%
Mixed logit 15 44.1%
Random effects conditional logit 2 5.9%
Other 2 5.9%
Unclear or not reported 2 5.9%
Removed responses Reasons for removal reported 6 17.6%
Unclear or not reported 28 82.4%
Model Akaike information criteria 3 8.8%
Bayesian information criteria 4 11.8%
Likelihood ratio test 11 32.4%
Pseudo Rsquared 1 2.9%
Other 2 5.9%
Not reported 16 47.1%