By Otis Hunt (Cal Poly, San Luis Obispo) and Allison Theobold (Cal Poly, San Luis Obispo)
Information
We hypothesize that the Likert Scale format presented to students in classroom surveys does not cognitively engage the respondent and the response options are not representative of the students' actual experiences. We propose a survey scale constructed of descriptions, targeting potential experiences in the classroom, and experimented with different options for shuffling the order in which these descriptions were presented to students. In our presentation, we describe how respondents interacted with these different response scales and orderings. The context of this study are statistics and data science courses at a medium-sized, public institution. The classes in which the study was implemented contained statistics, computer science, mathematics, biology, and public health students. The class sizes are capped at 35 students. The surveys were given following the collaborative activities students participated in with their partner(s).