As most of you know by now, the last half of the spring term sometimes feels like a downhill sprint. Except in this case you’re less concerned about how fast you’re going and more worried about whether you’ll get to the finish line without face-planting on the pavement.
Well, it’s no different in the IR Office. At the moment, we have four large-scale surveys going at once (the recent graduate survey, the senior survey, the freshman survey, and the employee survey), we’ve just finished sending a year’s worth of reports to the Department of Education, and we’re preparing to send all of the necessary data to the arbiter of all things arbitrary, U.S. News College Rankings. That is in addition to all of the individual requests for data gathering and reporting and administrative work that we do every week.
So in the midst of all of this stuff, I wanted to thank everyone who responded to our employee survey as well as everyone who has encouraged others to participate. After last week’s post, a few of you asked how many responses we’ve received so far and how many we need. Those are good questions, but as is my tendency (some might say “my compulsion”) the answer is more complicated than you’d probably prefer.
In essence, we need as many as we can get from as many different types of employees as we can get. But in terms of an actual number, defining “how many responses is enough” can get pretty wonky with formulas and unfamiliar symbols. So I shoot for 60% of an overall population. That means, since Augustana has roughly 500 full-time employees, we would cross that threshold with 300 employee survey responses.
However, that magic 60% applies to any situation where we are looking at the degree to which a set of responses to a particular item can be confidently applied to the overall population. What if we want to look at responses from a certain subgroup of employees (e.g., female faculty)? In that case, we need to have responses from 60% of the female faculty, something that isn’t necessarily a certainty just because we have 300 out of 500 total responses.
This is why I am constantly hounding everyone about our surveys in order to get as many responses as possible. Because we don’t know all of the subgroups that we might want to analyze when we start collecting data; those possibilities arise during the analysis. And once we find out that we don’t have enough responses to dig into something that looks particularly important, we are flat out of luck.
So this week, I’m asking you to do me a favor. Ask one person who you don’t necessarily talk to every day if they’ve taken the survey. If they haven’t, encourage them to do it. It might end up making big difference.
Make it a good day,
Mark