Skip to main content
. 2017 Apr 25;2(1):5–17. doi: 10.1007/s41669-017-0025-4

Table 4.

Discrete-choice experiment methodology in tobacco-control research

References Design Design plan Design source Method of creating choice set Number of choice sets Questionnaire Estimation method
Marti [31] Fractional factorial Main effects only and interactions Website Orthogonal array 10 Interviewer administered All models comparison
Goto et al. [8] Factorial Main effects only with interaction N-Logit Version 3 Orthogonal planning method 8 Unclear Mix logit model and simulation
Pesko et al. [27] Balanced Interaction with all variables Unclear D-efficient design
D-efficiency
12 Unclear Linear probability model with sensitivity analysis
Goto et al. [19] Fractional factorial Main effect only with interactions N-Logit version 4.0 and Stata 11 Orthogonal planning method 8 Unclear Random parameter logit model
Paterson et al. [30] Fractional factorial Unclear Expert panel Orthogonal design 4 Internet Random parameter logit
Czoli et al. [26] Balance incomplete block Main effect only with interactions SAS v. 9.4 Orthogonal design 20 Online/internet Multinomial logit regression
Salloum et al. [28] Fractional factorial Main effects with interaction SAS v. 9.4 Unclear 9 Internet-based (tablet) Multinomial logit regression, nested logit model
Salloum et al. [25] Fractional factorial Main effect and alternative SAS v. 9.3 Unlear 8 Interviewer Conditional logit models
Goto et al. [34] Fractional factorial Time risk preference: survival analysis NLOGIT 3.0 Orthogonal planning method 8 Interviewer administered Mixed logit model with simulation
Morgan et al. [32] Fractional factorial Main effect and subgroup analysis SAS v. 9.1.2 D efficient design (co-variance matrix) 24 Web-based online Conditional logit regression model
Kotnowski et al. [29] Fractional factorial Main analysis and interaction SAS v. 9.3 (D-efficiency 98%) Orthogonal and balanced choice set 10 Web-based online Multinomial logic model
Hammar and Carlsson [33] Fractional factorial Main analysis SAS D-optimal design 4 Unclear Standard random effects binary Probit model