A Summary of Debriefing Feedback on Our Python Lesson

This post originally appeared on the Software Carpentry website.

Last month, we discussed results from a survey of how our instructors are teaching Python. We now have a summary of the feedback we've received in our bi-weekly debriefing meetings. The recurring themes are:

  1. A greater choice of exercises and multiple choice questions would allow instructors to select domain specific examples and cater to varying levels of learner.

  2. Some instructors added an explanation of the Jupyter notebook or Spyder IDE environments.

  3. Some instructors added an explanation of basic Python datatypes before presenting the lessons.

  4. Comments that there is too much material to fit into a workshop and that some sections seem rushed.

  5. Request for a better explanation of the advantages of the Anaconda Distribution at the start of workshops and resources for post workshop learning.

  6. Comments on presenters style were positive for funny and entertaining examples and negative for highly mathematical examples.

The details are listed below, with '+1' showing points that came up several times. We plan to use all of this when we redesign our basic Python lesson.

  • Departures
    • Changed lesson to be more relevant to audience1

    • Instructor used own examples to change emphasis of lesson to loops

    • Re-organised the flow of lesson

  • Additions
    • Explained differences between Bash, Python in the shell, Jupyter notebook

    • Short Jupyter Notebook introduction

    • Demonstrated Spyder IDE as well as Jupyter notebook

    • Basic data types explained before starting +1

  • Requests
    • More exercise options in repo +1

    • More multiple choice questions

    • Directions to post workshop resources needed

    • Explain that Anaconda will install in a separate folder and not affect existing Python

  • Comments
    • Testing section seemed rushed

    • Lesson material too long to fit workshop length +1

Jupyter Notebook

  • Using Jupyter Notebook with Git is challenging

  • Lack of extensive history in notebook (compared to shell) hinders learners +1

  • Starting with an empty notebook better than using template code

  • Notebooks should be saved to repo in a cleared state, split up monolithic Jupyter Notebook (i.e. clear output) +1

    • implies not live coding.

Instructor Style

  • Complaint: prime numbers example too mathematical

  • Complaint: not enough discussion prior to challenges. (Learners: "I don't know where to start!")

  • Compliment: funny anecdotes received well

Dialogue & Discussion

Comments must follow our Code of Conduct.

Edit this page on Github