WHAT'S IN A SURVEY QUESTION?

Last month, several Boise School District Trustees attended the Boise Chamber of Commerce 2018 Leadership Conference in Sun Valley. A couple of Board members went to a presentation by Boise State University personnel in which they showcased their latest survey (December 2017, 1000 respondents, +-3.1% error margin) results. Here's one of the slides the BSU folks used with respect to preparation:

Wow, not good at all. 29.3% of statewide respondents categorized college preparation efforts as "excellent" or "good". The percentage of "excellent/good" responses for the Boise Metro area was even lower, at 29.0%.

Interestingly, BSU's survey used "fair" as a category, and fully 39% of respondents chose that category for the college prep question. What does "fair" mean to you? Decent? Okay? We aren't sure it's an appropriate category, or that it should be grouped with "poor". But maybe that's just splitting hairs.

However, in the run-up to the 2016 bond measure, we did some polling (300 parents + 300 voters, +- 5.6% margin of error)) in the community with the help of a professional polling group. When they asked a similar question, the results looked very different:

So in this poll the choices are better differentiated. "Very well" was the top choice, "Pretty well" was next. And the two choices associated with "dissatisfaction" were "not too well" and "not at all well".

When the choice of "don't know" is included in the data, 68% of those surveyed chose one of the two "positive" choice. When we factored out the "don't know" responses, the percentage choosing the two "positive" choices rose to 85%.

Admittedly, this poll question asks about "preparing students for a career", so it's a bit less focused than the BSU question on furthering their education". However the difference in response patterns is stark.

However, it's tough to know which of the polls was a more accurate reflection of patron/citizen feeling, since the results differed so dramatically. Perhaps the wisest course, especially considering the accuracy (or lack thereof) of recent state and national polls is to regard polling data with healthy skepticism!