I'm newish to Qualtrics and have been handed what seems like an intricate project that I can't find an answer to. Hope you can help.
At present we have a baseline, review and exit survey that asks the same questions. It is designed to track clients progress and understand service impact. However, we can't guarantee that all clients will complete every survey or every question. But, we only want to include 'matched responses'. That is, for each question, we only use responses from individuals that have done the same question at baseline/review/endline.
To identify individuals, we can use their client code and can enter it in each survey. Workers who complete the survey with the client will access the survey using a generic survey link.
How do I go about matching individuals' results and reporting matched responses for each question?
Further, is it possible to report progress? E.g. if a client scores 1 on the first baseline question, but then at review for the same question they score 3, we can say it was a 2 point progress. We can then find average progress between time points within the group (i.e. progress is based on individual progress rather than the difference between group averages at two time points). What is your suggestion for the best way forward?
Insights much appreciated!
In terms of your last question, I think you could use a custom metric in the dashboard for that?
I'd be interested if someone has a solution to your matched responses question because we have a similar use-case.
I would definitely recommend custom metrics for this. You can create calculations with various filters applied - for example, you can set a filter that questions at all 3 toupoints had answers, and use that as a flag/filter for the dashboard, etc.
Hi InessaG and Evaluation, is this just to report on progress? That's a good way of showing progress.
I'm still stuck on showing in dashboard 'matched data' - that is, showing results from individuals who have responded to a question at two time points (accounting for missed responses). While the above custom metric shows where data is matched, it unfortunately can't be used as a filter (as far as I can see).
Eliza22 The filter in custom metrics is just to check that there is a response for each data point. In the dashboard, you can create pivots for individuals, or filters at page level for each individual. There are many options depending on how you need to visualize it. Since you'll be cutting with the data for each individual based on above, the custom metrics will adjust for that individual either within the table row or on page if using page filters.
Hi InessaG. I've tried using custom metrics to identify matched surveys. The process works, but you can't seem to use a metric as a widget filter.
I've applied a similar process to embedded data in the second survey (carrying across results from survey one), but if you use embedded data to filter, it only filters responses in the second survey and not both surveys (oddly, as responses have been authenticated). It's as though the system recognises the sources (survey 1 & 2) as separate. From what I can see, the only benefit of carrying data over is to customise the second survey to the individual.
Yes, pivot tables are an option if responses for both surveys are stored in the second. It's the closest I get, but it I don't like the presentation. I can't apply a filter to a page, as a need to filter for question level matches (accounting for missing data).
I think it's Power BI for me.
Leave a Reply
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.