GOOD RESEARCH PRACTICES FOR THE APPLICATION OF CONJOINT ANALYSIS IN HEALTH – A CHECKLIST FOR PUBLISHING IN OUTCOMES RESEARCH

How was the conjoint analysis designed?

• Do you have a clearly stated research question/s and/or hypothesis/es?
• Have you articulated the study perspective?
• Have you explained why you chose your particular attributes for the study?
• Have you explained why you chose your particular attribute levels for the study?

How were the conjoint analysis tasks developed?

• Why was the experimental design chosen and were its properties tested?
• Was an opt out/Status quo option available? Why/why not?
• Were any limits made on the experimental design to ensure that all scenarios were affordable/plausible?
• Was the instrument piloted?  If so, what were the results and how was the instrument changed?

How was the conjoint analysis implemented and data collected?

• How were the task administered and how was the chosen method justified?
• How were the study subjects chosen (inclusion/exclusion, recruitment etc)?
• Can you justify the sample size?
• Did all respondents get the same questions and in the same order?

How was the data analyzed?

• How did you choose/justify the statistical modeling/techniques used?
• How to test for validity/reliability?
• How to test for rationality and consistency?
• Was sub-group analysis/clustering needed/performed?

How was the data interpreted?

• How is the importance of the factors assessed?
• How are WTP/welfare estimates calculated?
• What types of policy simulation are needed?
• How are utility estimates/MRS calculated?

Publishing the results

• What is presented in the abstract?
• How are the results presented?
• How much discussion of methods/theory?
• What is the appropriate use of abbreviations and jargon?

Generally Recognized Steps in Conjoint Analysis:

1. Experimental design
2. Analysis
3. Willingness to Pay – Welfare Estimations
4. Policy Simulation
5. Attributes
6. Attribute Levels
7. Administration
8. Perspective
9. Subjects
10. Sample Size
11. Research Question – Focus
12. Rationality, Consistency, Validity, Reliability
13. Presentation of scenarios: # of questions, tasks split-up across respondents
14. What wording is needed
15. Pilot testing – set up most meaningful
16. Presentation of results
17. Theory models, etc