I've been wrestling with LimeSurvey for a couple of weeks now, where I've had the time, in order to see if I can use it to replace a system I wrote for conducting telephone surveys.
The surveys we're doing at the moment are sampled by a certain field in the data. We get responses from 10% of the contact set and this is stratified by this field. That is to say that if we get a set of 2,000 contacts, in which this field has possible values "A", "B", and "C", we contact 10% of the contacts for whom this field is "A", 10% of the contacts for whom this field is "B", etc, thus contacting ~200 people over all three strata.
I can see that I can set hard numberical non-relative limits on answers to certain questions, and that I can (if I don't mind having to upload a CSV file where the headings are 'attribute_1', 'attribute_2', etc.) use attributes to pre-answer some questions, but this doesn't allow me to employ stratified sampling. I would have to calculate these figures myself and then if we get a call from our client to say that a group of contacts is not to be contacted, I can remove them from the participant group for that survey but I would then have to manually recalculate the quota amounts for these answers.
Can LimeSurvey allow me to employ stratified sampling based on the contents of an attribute?