Please enable JavaScript.

Coggle requires JavaScript to display documents.

Studying the effect of AI Code Generators on Supporting Novice Learners…

- - - - 3 weeks
      - 10 90-minute sessions
      - 1 session/day
    - - remote
      - google meet
    - - 25 multiple-choice questions about Scratch
        programming
    - - 45 programming tasks and 40 multiple-choice questions
      - topics
        
        basics (including variables, input/output, operators,
        and random numbers)
        
        data-types
        
        conditionals
        
        loops
        
        lists
      - instructors were trained to o guide the learners to the specific parts of the documentation to resolve their issues
      - learners received personalized feedback for
        each of their incorrect submissions that were rejected
      - Each programming task in the learning phase had two parts: authoring code (Figure 6b) and modifying code (Figure 6c).
      - Learners were required to continue working on the task before advancing to the next task
        
        learners were also able to skip tasks after exceeding the minimum amount of time (which varied between 3 to 12 minutes based on task difficulty)
      - For each submitted task that was incorrect, graders were instructed to initially just explain what was wrong with the submission. However, as the number of re-submissions or the amount of time that a learner had spent working on a task increased, graders were instructed to provide more direct feedback that included some hints
    - - The immediate post-test was conducted one day after the training phase.
      - learners did not have access
        to either the code generator or the Python documentation, or any
        feedback after submitting the tasks
      - The tasks in this phase included five code-authoring tasks, five code modification tasks, and then 40 multiple-choice questions
      - retention test, was conducted one week
        later and included a new set of tasks and questions with a similar number and order of tasks to the immediate test
    - - 69 learners
      - Participants received a $50 gift card as compensation and the study protocol was approved by our institution’s Research Ethics Board
      - To accommodate all learners with different time constraints, three
        sections were offered each day of the study and learners could
        choose which one to join each day
      - participants were divided into two groups
        using a matched-groups design
- - - - The reason why the experimental
        research was continued for five weeks is that the students in the
        experimental group absorbed the intervention well.