Federica Zoe Ricci (University of California Irvine), Catalina Medina (University of California Irvine)
Abstract
As statistics and data science classes become larger, automated grading tools are increasingly utilized. However, some type of assignments - like those involving communicating with/about data - go beyond questions with pre-specifiable correct and incorrect solutions and require human judgement for grading and feedback. In this workshop we will teach how to use gradetools: an R package designed to help educators grade open-ended assignments efficiently and consistently while providing submission-specific feedback. Integrating with RStudio, gradetools automates administrative grading tasks (e.g., opening assignments, noting grades on the gradesheet) and saves times with repetitive feedback.
The workshop will feature four hands-on modules:
1. Grading preparation (submissions, rubric, roster)
2. Core grading functionalities
3. Extended grading functionalities
4. Interactions with GitHub (available for those teaching with GitHub)
Participants will be encouraged to practice the package on their own submissions, so they should come with R and RStudio installed on their laptops and with four submissions that can be opened in RStudio (e.g., .Rmd, .R, .py files) and that include open-ended questions (e.g., model interpretation, data visualizations). If interested in the last module of the workshop, they should have a GitHub account connected with RStudio. Software required for installation and an example submission file will be shared one week prior to the workshop at https://github.com/federicazoe/gradetools-uscots2023.