Abstract:
|
Data science courses are in high demand, and statisticians are often tapped to teach them. Efforts are lacking in data science education research to rigorously measure the effectiveness of teaching methods on student learning and attitudes in data science. Do student attitudes matter as much in data science as they have been shown to matter in statistics? (Ramirez et al., 2012; Kerby & Wroughton, 2017; Pearl et al., 2012) We leverage our work on motivational attitudes toward statistics to develop a family of instruments to measure attitudes toward data science. We will discuss our pilot Student Survey of Motivational Attitudes toward Data Science (S-SOMADS). This includes the theoretical framework, based on Expectancy-Value Theory (Eccles et al., 1983; Eccles & Wigfield, 2002), as well as the item development, subject-matter-expert review, pilot data collection, and exploratory factor analysis results. We will conclude with future directions of the project, including development of instructor and environment instruments, as well as how current or future data science instructors and educational researchers can get involved. This project is supported with NSF funding (DUE-2013392).
|