This article examines the credibility ratings of various items printed in a questionnaire. In questionnaires, being served as
significant tools for data collection, questions ordinarily have equal coefficients. But the fact is that some questions far outweigh others. Various techniques are employed to achieve the weight coefficient of questions in the questionnaires inter alia entropy and specific vector method. Attempt has been made to primarily
examine the present condition as regards the credibility rating of questionnaires (a case study at Shahrood University for evaluation of the teaching staff performance) and then acquire coefficients for the questions by using AHP Techniques and an OR model.
The results demonstrate that there is a wide gap between the existing credibility ratings (weight coefficient) and the ratings achieved. Lingo and Excell soft wares have been applied in this endeavor to come up with a solution.