Analysis of data from a question and answers format questionnaire in terms of degree of oligopolistic evaluation: Analysis using a zipf distribution

Abstract

The present study focused on collecting students' opinions using a class-evaluation questionnaire with a questionand-answers format as a method for clarifying what the students considered to be the most important knowledge that they had learned in the class. The question-and-answers format questionnaire was used in order to allow the students to focus on their first evaluation and impression. The oligopolistic (dominant) degree of knowledge gained at this time could approximately be described by a Zipf distribution. In addition, the cumulative probability calculated from the Zipf distribution was used to evaluate the degree of saturation of the observed findings, together with the different elements (evaluations, impressions) necessary to achieve a specific saturation. From the Zipf distribution data, 2 kinds of methods were devised: (a) where the number of findings obtained from the Zipf distribution of the questionnaire responses was the number of unknown types, and (b) where the number of findings obtained was limited. An index was proposed that was used to examine whether less specific knowledge was the dominant finding from the free descriptions obtained with the class-evaluation questionnaire.