Editor’s note: Brad Patton is chief software architect at Researchscape, and Jeffrey Henning is chief research officer at Researchscape International.
If you do a lot of survey research, you should investigate automating repetitive tasks. Before automating, you need to implement time-usage tracking to understand what needs to be automated.
The first step is to implement time sheets for your staff, tracking those items you may wish to automate: in our case, we have tracking codes for questionnaire design, survey programming, fielding, data cleansing and analysis, among other common tasks.
The second step is to create a feature database, which can be as simple as a spreadsheet or as complex as a case management system. As subtasks come up, create a case for a new issue or upvote a case that isn’t. For instance, in survey programming, we had a list of manual steps the programmer would take for “Choose all that apply” questions, such as making sure there was an exclusive choice like “None of the above” and setting its properties appropriately.
By tracking staff’s time and tasks, you can prioritize what you want the IT department to automate. For instance, early on in our survey programming automation, we automated many details related to “all that apply” questions and “choose one” nominal questions.
You’ll also need to review if your survey platform supports the automation that you will need. Based on our own experiences automating all aspects of survey research, across five major platforms, our team pulled together a list of features and capabilities a research firm should look for in choosing a survey solution.
Because our clients use a range of survey platforms, we’ve learned the ins and outs of these different APIs. This led us to be deliberate about the platform we choose for our own needs.
Thanks to our careful study of our team’s time usage, and our subsequent investmen...