Uploaded image for project: 'Moodle QA'
  1. Moodle QA
  2. MDLQA-10999 Moodle 3.4 QA testing
  3. MDLQA-11580

CLONE - Managers can manage analytics prediction models

    XMLWordPrintable

    Details

      Description

      We are interested in hearing your feedback and suggestions for improvement. Feel free to post your thoughts in https://moodle.org/course/view.php?id=17233

      Testing environment:
      It is recommended to the site University of Analytics to test this MDLQA because the setup required to make 'Students at risk of dropping out' generate insights involves a significant effort. However, if you are interested on using your own data set, please note:

      You need past courses and ongoing courses. Past courses will be used as training data and you will get predictions for ongoing courses. The 'Students at risk of dropping out' model has multiple checks and has the following requirements:

      • All courses must have students and significant activity between the course start and end dates
      • Courses used for training must be finished
      • Courses used to get predictions should have a past start date and a future end date as well as a teacher to receive the course insights

      The second part of the test, 'No teaching model'. requires you to have at least 2 courses with teachers and students and 2 courses with students but without teachers. These 4 courses should start tomorrow or later in the week.

      Test:

      1. Log in as a manager and go to 'Analytics models' in the Site administration.
      2. For 'Students at risk of dropping out' select Evaluate in the Actions column.
      3. Verify that a score and information about the model performance using each of the time-splitting methods enabled on your site is displayed (by default Quarters, Quarters accumulative and Single range).
      4. Click continue.
      5. For 'Students at risk of dropping out' select Log in the Actions column.
      6. Verify that a list of previous evaluations is displayed, with version (last time the model indicators were modified), indicators included in the evaluation, time splitting method and reported accuracy.
      7. Click View in the Info column.
      8. Verify that extra information reported by the machine learning algorithm is displayed in a modal window.
      9. Return to 'Analytics models' in the Site administration.
      10. For 'No teaching' select 'Get predictions' in the Actions column.
      11. Scroll down the page listing info and click the continue button.
      12. Verify that for 'No teaching' the Insights column contains a dropdown menu with a single item 'All predictions'.
      13. Select 'All predictions' and verify that the model insights are displayed.
      14. Return to 'Analytics models' in the Site administration.
      15. For 'No teaching' select 'Clear predictions' in the Actions column.
      16. Verify that the Insights column displays the message 'No predictions available yet'.
      17. For 'No teaching' select Disable in the Actions column.
      18. Verify that the Insights column displays the message 'Disabled model' and that the option 'Get predictions' is not available in the Actions menu.
      19. For 'No teaching' select 'Enable' in the Actions column, then select 'Get predictions'.
      20. Scroll down the page listing info and click the continue button.
      21. Verify that for 'No teaching' the Insights column again contains a dropdown menu with a single item 'All predictions'.

        Attachments

          Issue Links

            Activity

              People

              • Votes:
                0 Vote for this issue
                Watchers:
                1 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: