Teaching Students to Use Summary Statistics and Graphics to Clean and Analyze Data

Abstract

Textbooks and websites today abound with real data. One neglected issue is that statistical investigations often require a good deal of "cleaning" to ready data for analysis. The purpose of this dataset and exercise is to teach students to use exploratory tools to identify erroneous observations. This article discusses the merits of such an exercise and provides a team project, problem data, cleaned data for instructors, and reflections on past experiences. The main goal is to give instructors a prepared project for their students to perform realistic data preparation and subsequent analysis. The data for this project involve categorical and continuous variables for subjects age 65 and over testing calcium, inorganic phosphorus, and alkaline phosphatase levels in the blood. The project described in this article involves summary analysis, but the cleaned data could also be used for projects on independent samples t-tests, analysis of variance, or regression.

Computes: Development of an Instrument to Measure Introductory Statistics Instructors’ Emphasis on Computational Practices

“…The influx of data and advances in computing have led to calls by several statisticians to update the introductory statistics curriculum to provide students with the computational tools and data-related capacity imperative for modern practice (e.g., Horton et al, 2014;Nolan & Temple Lang, 2010). Some instructors have already begun integrating more data and computational practices into the introductory course, including relational databases (e.g., Broatch et al, 2019), web scraping (e.g., Dogucu & Cetinkaya-Rundel, 2021), data wrangling (e.g., Hardin, 2018;McNamara & Horton, 2018), data cleaning (Holcomb & Spalsbury;2005), multivariate visualization (e.g., Çetinkaya-Rundel & Tackett, 2020;Kaplan, 2018), reproducibility (e.g., Baumer et al, 2014), and version control tools (Beckman et al, 2021;Fiksel et al, 2019).…”

On the improvement of students' conceptual understanding in statistics education

“…Other research focussed on cooperative learning of statistics, which has been shown to have beneficial effects on student learning (Garfield, 1993;Giraud, 1997;Magel, 1998). Recommendations are primarily given for lectures (e.g., Larsen, 2006), methods for teaching specific subjects (e.g., Walters, Morrell & Auer, 2006), the use of data sets (e.g., Holcomb & Spalsbury, 2005), textbooks (e.g., von Hippel, 2005), and specific exercises (e.g., Kahn, 2005).…”