Create a Predictive Learning Analytics Model

Search our help articles

To create a PLA model:

  1. Decide what you are predicting. (The most common PLA models predict whether learners will be successful in current courses.)

  2. Choose a PLA report from the Library or create your own.

These reports may be used as-is or edited to match the needs of your institution:

Predictive Learning Analytics Input: Passing by Course End Date - may be used with any LMS that provides Course End Dates, such as Moodle.
Predictive Learning Analytics Input: Passing by Term End Date - may be used with any LMS that provides Term End Dates, such as Canvas.
Predictive Learning Analytics Input: Passing 90 Days (Enrollment Start) - may be useful when courses do not have fixed end dates, but learners are expected to complete the course within a defined span of time after Enrollment Start, e.g. 90 days. This length of time can be adjusted.
Predictive Learning Analytics Input: Passing 90 days (Enrollment Creation) may be useful when courses do not have fixed end dates, but learners are expected to complete the course within a defined span of time after Enrollment Creation, e.g. 90 days. This length of time can be adjusted. Use this report if your LMS does not support a defined Enrollment Start Date.

Your report must contain the following data columns. Other data such as “Course Name” or “Student Name” may be included in the report but will be ignored by the PLA system.

  1. Unique Identifier: Each prediction will require a unique identifier. For example, if you are going to make predictions about students in courses, the best unique identifier is the Enrollment ID.

  2. Success Criterion (Target or Label): You must have one column that defines whether the learner was successful in binary terms, where 1 = successful and 0 = unsuccessful. In our sample reports, a passing course grade of 60% is used. This is defined using a formula in the report.
    Other possible criteria include completion of all required activities, achievement of all assigned competencies, continuous participation to the end of the course, etc. 
    You cannot train a model if all your rows were successful or all failed. Our default reports provide an aggregation in a panel at the top of the report. Check to make sure this value is between .10 and .90.

  3. Case Process Criterion

    You must have one column that indicates whether each row of data will be used to train the predictive model or to generate predictions. This column helps to differentiate between the data used to teach the model how to make predictions and the data that the model will then analyze to make those predictions. (The same report will be used to train the model and to generate predictions.) In our sample reports, this column is labeled "Current" and a value of 1 = a current, active enrollment that can be used to generate predictions; 0 = an enrollment that has concluded and can be used to train the model. This value is set using a formula that references the course end date, term end date, enrollment start date, etc., depending on the LMS and the institutional requirements.

    You will need multiple rows with this value at 0 to train a model. We recommend at least 1000 rows of historical data.

    You will need some rows with this value at 1 to predict outcomes using a model.

  4. Multiple Predictor Variables (Features) 

    You may have multiple "predictor" variables that will be analyzed to determine how they affect the outcome. Our default reports include the following:

    • Visits per Activity

    • Participations per Activity

    • Time Spent per Activity

    • Submissions per Assignment

    • Attempts per Quiz

    • Posts per Forum
      Each of these columns must have a range of values to be used in a model. In our default reports we provide a Range aggregation -- make sure each column you want to use has a Range >= 1.

      Usually it is best to "standardize" these values, because course design can differ so much between courses, programs, and instructors. We recommend dividing by the number of activities in the course. 

 



  1. Create a New Model.

  2. Go to “Apps” tab and select “Predictive Models.”

  3. Click “Create Model” Screenshot 2024-03-20 112147.png button in the top right corner.

  4. Fill out the form and click next. This form is step 1 of 6.

  5. Select “Case Identifier” columns you will be making predictions about. This will be used to relate predictions to entities in the database, e.g. user id, course id. They should be fields of type "ID". You may choose multiple columns. This is step 2 of 6.

  6. Select “Outcome” column to be the success criteria. Choose the data column containing the outcome the model will predict. This will be populated in historical data and blank in prediction data. It must contain a value of 0 or 1. This is step 3 of 6.

  7. Select “Feature” columns. Choose columns that will be used to predict the outcome. These must be numerical and contain a range of multiple values. You may choose multiple columns. This is step 4 of 6.

  8. Select “Model Process Criteria” column that indicates whether the row is historical data used to train the model or current data used to generate predictions. Select a column in the data set that defines whether the data in that row is for training, prediction, or ignore: 1 = prediction, 0 = training, -1 = ignore. This is step 5 of 6.

  9. “Check Data and Create Model.” Review your data and click “Create Model” to create.

Direct link to “Predictive Model” Builder Form here.

Predictive Models.mp4

Predictive Model Builder Form

Predictive Model Form.png
  1. Model name: Name your model to differentiate it from other models and make it easy to recognize.

  2. Model Description: Description that summarizes what the model is based on.

  3. Source Dataset: Select the Dataset that will inform the PLA from the drop down menu.

  4. Algorithm: You have a choice of algorithms to build your model. Currently, IntelliBoard Pro supports two algorithms, Logistic Regression and Neural Network. Neural Network tends to produce more accurate predictions but can require more data. Logistic Regression produces more linear predictions.

  5. Next Step: Once the form is complete, click to precede to the next step.

Frequently Asked Questions:

https://intelliboard.atlassian.net/wiki/spaces/KB/pages/2310471706