That's not to say I design perfect surveys, or am the ultimate authority when it comes to methodology - but there are a few quick tips to constructing a better survey. One of the easiest pitfalls to find yourself in is a double-barreled survey question. This tends to happen when the survey designer lumps concepts together in their head (for example - library hours and locations), or when a survey designer is trying to limit the number of questions they ask in the survey. What should be two separate questions can easily become one question that can be complicated for respondents to answer. Using my previous example, it's easy to see how a survey designer could come up with the following question:
Do you find the library hours and locations satisfactory?
-Strongly Disagree -Disagree -Neither Disagree/Agree -Agree -Strongly Agree
A respondent might be completely satisfied with the location of the library, and be completely dissatisfied with the hours of the library - but has no way to convey that to the surveyors. The default answer might be the "neither" option - but then their responses will be congruent with those that have no opinion on the subject or are not knowledgeable about those details. This question should be split into two different questions, so the survey designers can analyze the results and determine if the hours or the location (or neither or both!) are causing dissatisfaction among respondents.
Before releasing your survey - do a close read of each question. Ensure each inquiry is asking a single question, and that there is only one way for the respondent interpret the question .