Hello,
I would like to link several study parts using a unique identifier and was wondering which option would be best depending on the study design. It will be a longitudinal study with two treatment conditions and one control group. In total having 4 waves.
in t0: all participants go through the same survey and receive an official invitation to participate in the study.
in t1: participants will be randomly assigned to either the control group, treatment 1 or treatment 2 group
in t2: participants need to be directed to one specific version of a follow up questionnaire,depending on what they have seen in t1.
in t3: participants again need to be directed to one specific version of a follow up questionnaire.
a) Each participant should have a unique ID which allows me to link the different data sets corresponding to each participant.
b) The questionnaire should be anonymous.
c) How can I handle to direct participants to a specific questionnaire depending on the experimental condition in which they participated earlier (e.g. in t1 or t2)?
Thanks!
Item (b) isn't clear. Should the data for all the surveys (t0-t3) be anonymized (so there is no PII in the response data) or will an anonymous link be used for t0?
Thanks for the quick reply, Tom.
Data for all the surveys should be anonymized and we are planning to use an anonymous link for t0.
In all your surveys (t0-t3) click Anonymize in survey options.
In t0, create a unique id using a webservice or JavaScript and save it as an embedded data field. For the people you will invite, redirect to another survey and pass the unique id as a parameter. In that survey ask for their email (and any other PII you need to save such as name). Use a contact trigger to save their info to a contact list, saving the unique id as the ExternalDataReference.
In t1, use a survey flow randomizer to set an embedded data variable as control, treatment 1, or treatment 2. Then branch using the 'group' variable as needed. Use a contact list trigger to update the contact list with the 'group' variable.
In t2/t3, use the 'group' variable to branch as needed.
Leave a Reply
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.