Embedded Data and Screening Logic

The change we make that takes up the most space in your survey flow and might look like the biggest change is the embedded data and screening logic. As mentioned in the “Skip Logic” section, we skip respondents to the end of the block when they don’t qualify and we then tag them (embedded data) according to their status and redirect them out (see “Redirects” section after this).

If you look at the survey flow after we make our changes, you’ll notice that things look a bit different. First, we will pull in the id variable(s) that help us identify the respondents on our end for compensation. We pull the IDs separately from the data and do not identify who put which answers. This portion can be as few as one item or as many as 10+ depending on which and how many of our sample sources we are using for the study. These embedded data variables do not affect what the respondent sees or your analysis – they are for our purpose only.

You’ll then see various branching and terminate logic for over quotas (overall quota) and terminates. This ties back to what is described in the “Skip Logic” portion and is where we actually terminate respondents, so it’s critical to our setup. The logic here will generally be simple, such as “If Q2 age is less than 20 then screen out”, with some embedded data tags in there to capture information and then customized end-of-survey elements, which are the elements that actually end the survey at that exact point. These customized elements will redirect people properly (see “Redirects” section) and will also make sure terminates don’t count toward your quotas. If you’re not recording terminate/partial responses in your account, you’ll also see that we have checked a box in the end-of-survey elements that tells the system to not record the responses.

When you look at reports or the downloaded data, you’ll see columns for each of the embedded data fields (e.g., gc, term, rid, psid, PID, rnid, etc.). You can ignore those, as they’re for our purposes only.

However, if you record terminates, this information might be useful to you. Your data – both in your account and downloaded – will have a “gc” value for each participant and a “term” value for terminates. To view only the valid data, filter down to “gc” equals “1” (1=valid complete, 2=terminate, 3=over quota, 4=quality fail). To view the number of terminates per criteria, view the “term” values and their counts.

Embedded Data