Staff Ed: Substandard Surveys

William Dalsen

Lawrence students are bombarded with surveys. LUCC depends heavily upon them to inform and reform its policies, and administrators rely upon them to gauge student opinion on a variety of matters.
Aside from the fact that many students are annoyed with surveys, much of the time LUCC in particular uses less than scientific methods to arrive at its conclusions. Usually LUCC employs one of two methods: members will set up a table at Downer, grab whichever students fall in their general direction, present them with questions that are in at least some cases not unbiased, and use the responses to inform a policy change – in particular, a few Residence Life Committee surveys come to mind. Alternately, mass mailing through Voyager is used, and whichever students happen to respond are fortunate enough to have their responses count.
Even a rudimentary knowledge of sound statistical methods should give us pause before accepting either method. In both cases, those asked to respond to surveys are not randomly chosen from a list of students but rather are a mishmash of students who happen to respond or have a particular interest in responding – for example, they may be mostly super-seniors, mostly not conservatory students, or mostly men. Also, the number asked may be too large or too small to be statistically sound, and the questions may be skewed toward favoring a particular policy outcome.
The recent letters to the editor regarding SOUP provide another possible example of flawed surveys in action. It may be the case that a majority of those who responded to SOUP’s survey regarding the Big Event favored Guster over other bands, but that does not mean that they can soundly infer anything about the desires of the student body from those responses. This is not to side either with Justin Eckl or SOUP on this matter, but rather to point to a situation where sound statistics could settle disputes easily.
Another example may be the LUCC policy change that eliminated super-senior standing from housing selection. If a sound survey was taken that had an equal chance of including super-seniors and underclassmen, LUCC could better guarantee that its policy was truly “what the students wanted.”
In short, there is no excuse for LUCC and the administration not to observe the most basic of sound statistical methods, and there is no reason for students to accept policy changes inspired by surveys that are less than scientific. We hope that the new Snyder administration in particular will account for this during its tenure.