By Shiya Cao (Smith College)
Information
This work evaluated the impact of the integration of disability inclusion related datasets and readings into an undergraduate-level introductory data science course on students’ data science skills and on their attitudes toward data science. The author constructed disability inclusion datasets from the National Household Travel Survey and the Current Population Survey-Annual Social and Economic Supplement. These datasets and background readings were then integrated into the three mini-projects (data visualization, data wrangling, and mapping). Data were collected on students’ learning gains through a pre- and post-course survey across three semesters from four intervention classrooms (integrating the disability inclusion datasets and readings into the mini-projects) and three control classrooms (using other datasets that have been used in the introductory data science curriculum (e.g., the fec16 data) in the mini-projects) at a liberal arts women’s college. A total of 127 students completed both the pre- and post-course surveys (86 intervention, 41 control). Using linear models on the intervention classrooms, the author found that there was a statistically significant positive relationship between students’ increased understanding of data science within a disability inclusion context (e.g., conducting exploratory data analysis in a disability inclusion context) and their data science skills/their attitudes toward data science after class. Additionally, through a robust analysis for mixed designs, the author found that there were significant differences in the changes of students’ data science skills and their attitudes toward data science between the intervention and control classrooms.