Jillian Downey (Gustavus Adolphus College), Brenna Curley (Moravian University)
Abstract
Background. Recent work in statistics education around nontraditional grading includes an overview of the general benefits and challenges of nontraditional grading schemes (Curley et al. 2023) or specific examples of implementation, such as using standards-based grading in a mathematical statistics course (Curley & Downey 2023). In this work we turn our attention to the large cohort of students who take an introductory statistics course. Our goal is to investigate these particular students’ attitudes towards statistics when compared across courses that use traditional versus standards-based grading. Further, given that student performance, and hence their viewpoints towards statistics, may depend on their mathematical, personal, and/or educational backgrounds, we explore whether results vary depending on different demographic subpopulations.
Methods. Data were collected over the course of four semesters (2022-2024) from introductory statistics courses taught at two different liberal arts institutions. For comparison, data were collected for two semesters when the courses used traditional grading and two semesters of data for when the courses had transitioned to nontraditional grading. Collected survey data include demographic information and student attitudes using the Attitudes for Statistics (ATS) survey instrument (Wise, 1985). In addition to descriptive analyses comparing student attitudes across grading schemes and different demographic subgroups, an intended method of analysis will be hierarchical modeling.
Findings. Preliminary results show students were generally positive about the nontraditional graded course compared to a traditional graded course. For example, when asked to compare how confident they were in their learning/abilities in the nontraditional course as compared to a traditionally graded math or statistics course, 61.3% of the students reported “a lot or somewhat more.” When comparing student attitudes towards statistics, there were similar values for central tendency and spread in attitude changes for the two types of grading. However, alternative grading had more extreme outliers in attitude change scores (both in the positive and negative direction).
Implications For Teaching and For Research. Our prior work on nontraditional grading focused on introducing these practices to the statistics education community and providing example implementations that instructors could adapt in their own courses. In this work, we focus on a more in-depth analysis of student survey data. An overarching goal of all these works is to encourage instructors to implement nontraditional grading practices in their statistics courses. However, we recognize that pedagogical shifts take time and effort. So which, if any, of these types of work motivate us to make those changes; and, how can it inform us as statistics education researchers moving forward?