This survey is filled with bias and the results will not be trustworthy

Sign promoting survey at Douglas College, New Westminster

Benjamin Disraeli is supposed to have said, "There are three kinds of lies; lies, damned lies, and statistics". There are a few ways to ensure that a survey gets the results you want. This survey seems to have focused on building bias into the potential responses.

Survey for Chartwell's services at Douglas College

The middle value on a survey that uses a scale with an odd number of potential responses should be neutral. (So, for a five point scale there should be two positive, two negative and one neutral) There is a well documented preference for survey respondents to opt for the middle option. In this case that middle option is "Good",  a positive response.

Survey for Chartwell's services at Douglas College

These questions have a Likert scale. This means that a statement is presented and there are five potential responses for the person completing the survey. If this survey was designed to be bias-free the responses would be; Strongly Agree, Agree, Neither Agree not Disagree, Disagree, Strongly Disagree.  Again the middle option in this survey has a positive bias, "Somewhat Agree" when it should be "Neither Agree nor Disagree".

So, when the results come out no one should be surprised that the dining services are rated higher than the users expect. If you cook the survey you cook the results.