I have a relatively long questionnaire currently evaluating health states. I Have finished 2 parts of the questionnaire (CBC & MaxDiff) so far and now I have one part left. Up to this point, I have 66 questions in the questionnaire and the third part is done by the free formatting and composed of more than 250 questions (in total but only a small portion will be seen by each respondent) this part is designed to test certain health states compared to full health using a method called "Time Trade-Off" where the number of years spent in full health is changed until it matches the utility of living in the tested health state for a certain amount of years (10 for example).
Regarding this part I have the following questions:
1- Is it better to add the whole questionnaire to my original questionnaire (add the 250+ to the already 66) or refer the respondent to this questionnaire and upload it separately? Note a) that the respondent has to come back to the original questionnaire to complete other parts. B) I want later on to randomize the parts that the respondents could answer first (i.e. some respondent should answer CBC>MaxDiff>TTO, others CBC>TTO>MaxDiff and so on up to 6 different combinations.
2- If the referring option is better, how should this be managed?
3- How could I randomize the respondents for the order of answering the tasks such as that some of them answer the CBC first, others the MaxDiff first & a third part the TTO first and then also the second and third task should be randomized, in total having 6 different combinations?
I have a question in the third part that I will ask in another post.
Thank you very much