Best practices for teaching an introductory data science course
FacDev22 Workshop
Nicholas J. Horton and Benjamin S. Baumer
June 15, 2022
Plan for the morning
- Overview of inclusive practices and the principles
- Exploration of several of the principles
- Share reflections, resources, and experiences
- Interact (virtually) with Josh Recio (who created the Dana Center principles)
Goals for the morning
- Motivate the importance of inclusive pedagogies and student-centered approaches
- Familiarize participants with the Dana Center Data Science Course Design Framework
- Explore several of the principles
- Integrate them into an introductory data science course
- Share resources and best practices
- Explore next steps
Dana Center Design Principles
Inclusive practices
- Simply implementing new courses is not enough
- Need to eliminate institutional and system barriers to student success
- Design effective, inclusive practices from the start
Dana Center Data Science Course Framework
- Copyright 2021
The Charles A. Dana Center at The University of Texas at Austin - Educators granted a nonexclusive license to reproduce and share copies
Dana Center Data Science Design Principles
- Support students’ Social, Emotional, and Academic Development (SEAD)
- Explicitly attend to system inequities in school and society
- Develop students’ ability to engage in data science practice and process
- Remain true to the core of the course
- A range of pedagogical strategies
- Already doing much of this work (but more is needed)
DP1: Active Learning
The course provides regular opportunities for students to actively engage in data explorations using a variety of different instructional strategies (e.g., hands-on and technology-based activities, projects, small group collaborative work, facilitated student discourse, interactive lectures).
DP2: Growth Mindset
The course supports students in developing the tenacity, persistence, and perseverance necessary for learning data science, for using mathematics and statistics to tackle authentic problems, and for being successful in post-high school endeavors.
Example:
DP3: Problem Solving
The course provides opportunities for students to engage in the entire statistical problem-solving process.
DP4: Authenticity
The course presents data explorations that allow students to address relevant questions that arise in their communities.
Example:
DP5: Context and Interdisciplinary Connections
The course presents data science in context and connect data science to various disciplines and everyday experiences.
Example:
DP6: Communication
The course develops students’ ability to communicate insights from their data explorations and findings in varied ways, including with words, data visualizations and numbers.
DP7: Technology
The course introduces students to current technologies appropriate for data exploration and visualization, and prepares them to learn and use new ones.
DP8: Assessment
The course uses project-based assessments both as formative assessments and to evaluate student progress.
Technology and tools
- if we had more time, we could move more slowly and provide more depth
- our goal is to paint a broad landscape and point you towards promising options
- we don’t expect you to follow everything or adopt our full approach
- success would be if you learned about some new things and generated a list of approaches to explore
Slack 101
- https://CAUSEweb.slack.com
- #facdev
- interact during (and after) the workshop
Slack 101
Best practices: organize discussions in threads
Slack 101
Best practices: organize discussions in threads
Slack 101
Best practices: direct messages (DM)
Slack pedagogical notes
- encourage students to set notifications
- use other systems (e.g., Course Management System) for more formal communication
- encourage use of Slack for less formal interactions
- respond to individual DM to the entire class (while anonymizing the student who posed the question)
- support for simple polls
- Slack particularly compelling if your institution uses it for other purposes
- Other technologies for discussion worth considering
- https://slack.com/resources/using-slack/your-guide-to-slack-for-higher-education
- https://towardsdatascience.com/6-data-science-slack-workshops-you-need-to-join-b0c00952105d
Slack pedagogical notes (cont.)
- a contrarian view: https://rtalbert.org/the-rise-and-fall-of-slack-in-my-teaching-a-cautionary-tale/
GitHub 101
GitHub 101
GitHub 101
- GitHub is a powerful and expert friendly version control system
- an important tool for reproducible analysis and responsible workflow
More on using GitHub in the classroom in Beckman et al, JSDSE
RStudio cloud
RStudio cloud
RStudio cloud
RStudio cloud
- We wanted to provide you with the ability to engage with some of our materials in addition to what Ethan has provided via the Hampshire College server
- Big advantage of the cloud: simplifies package installation and setup
- Not my typical approach (we utilize cloud-based servers for our introductory students, then transition them to their own machines after they are up and running)
- We hope that the free account will work for you
- We suggest that use your GitHub account to log in
Recap of goals
- Our goal was to help you to engage with these design principles
- Our plan was to share examples of how we have integrated these principles into our courses and spark discussion
What’s next?
- Slack channel: please share your best idea and questions
- What ideas will you be bringing back to your colleagues?
- What do you wish we’d discussed but didn’t?
- What question do you still have?
- Discussion with Josh Recio (UT Austin Dana Center and lead for the design principles project)