Which Reports to Review for Data Collected Using the Matrix Item

Typically, data collected using the Case Evaluation (such as a post-simulation survey or SET)  is reviewed as group data. That is, we are not concerned with how an individual learner responded, but rather how the group responded.

For this reason, Case Evaluation data is best reviewed in two reports:

• The Responses Report displays:

• • Percentage of learners who selected a response 
  • Point value for each response

• The Case Item Analysis
  • Percentage of learners who selected a response
  • Number of learners who selected a response

Both reports display the  item average and the item standard deviation.

To understand the data in these reports, it can be useful to review the Case Evaluation in the Cases module.

In this example, both reports are for the same Activity/Case.

The Case evaluation has:

• One matrix item with five rows (criteria/questions)
• Radio button responses (learner can choose only 1 response per row)
• Four response anchors (columns),

Each response has the following point values:

• Strongly agree = 2 pts
• Somewhat agree = 1 pt
• Do not agree = 0 pt
• NA = 0 pt

Data Review: Per Response Score

Data Review: Correct Response Explanation

The Responses Report indicates how many learners selected the correct response. But what does “correct response” mean in the context of the matrix item?
LearningSpace treats an evaluation created with the matrix item as a single “question” for scoring purposes.

Therefore, “correct” indicates how many learners selected ALL the responses with the highest point value.

As you can see in the Responses Report below, the Strongly agree response has the highest point value (2 points). Therefore, LS treats Strongly agree as the correct response.

The report shows that only 2 out of 12 learners (17%) responded correctly. That means that two learners responded Strongly agree to each of the criteria.

Data Review: Average

Both reports display the average response score. 

The average is the total number of points divided by the number of students who responded. The average is calculated based on all the responses with a point value (not just the highest point value).

In the Case Item Analysis below, there are 10 responses with a point value greater than 0.

• In this example:
  • Strongly agree = 2 points
  • Somewhat agree = 1 point
  • Do not agree = 0
  • NA = 0
• Therefore, both the Strongly agree and Somewhat agree scores are included when calculating the average, but the Do not agree and NA scores are not.

• Strongly agree = 2 points; the total points for this response is
  
  (2 pts x 3 students) + (2 pts x 3 students) + (2 pts x 5 students) + (2 ts x 5 students) + (2 pts x 3 students) = 38 points

6 + 6 + 10 + 10 + 6 = 38 points

• Somewhat agree = 1 point; the total points for this response is (add the column)
  
  (1 pt x 4 students) + (1 pt x 4 students) + (1 pt x 3 students) = (1 pt x 4 students) + (1 pt x 5 students) = 20 points

4 + 4 + 3 + 4 + 5 = 20 points

• A total of 12 students have completed the SET evaluation (the number of students who submitted responses in each row = 12).

So the average (m or μ) is: 

Data Review: Standard Deviation

Both reports display the standard deviation. The standard deviation is calculated 

• Based on all the responses, whether they have a point value or not
• Based on the number of learners who responded (submitted data)

The formula for calculating the standard deviation is

 

Where

• x = learner score for entire item
• m = average (mean) for the entire item 
• n = number of learners who responded (submitted data)

See below a worked example:

The table contains the data we need to calculate the standard deviation for this example.

• m = average (mean) = 83
• n = number of learners who submitted data = 12

μ = m (mean, average )

Standard Deviation in the Case item analysis Report

The standard deviation is based on the number of students who submitted data (completed the SET evaluation).
(See above for calculations.)

Data review: Histograms

Data can be represented as histograms in the Responses Report by clicking the Show Diagrams toggle switch to the ON position.

Responses Report:

Case item analysis:

Data Review: Filter by Group

Data in both reports can be filtered by:

• User subgroups (created in the User module)
• Evaluator-evaluee subgroups (created on the Assignments tab of the Activity – see below)

Group Filters for Responses Report

Group Filters for Case Item Analysis

Dynamic Subgroup: Case Item Analysis

Select Question/Criterion:

Let’s look at the same data set in the Case Item Analysis: that is, we’d like to see the data only for those learners who selected Strongly agree for “Debriefing contributed to my learning.”

To select a dynamic subgroup in the Case Item Analysis, first select the question/criterion by clicking the radio button on the left of the item.

Select Response:

When you select the item/criterion, a dropdown menu is added to the top of the report which allows you to select the response for the learners whose data set you want to review.

Review Data for Dynamic Subgroup:

The report updates to display the data only for those learners who selected the Strongly agree response for “Debriefing contributed to my learning".

To return to the default data set, select No dynamic subgroups at the top of the report.

For further information visit our Reports Module and Reports - How to pages, and view the Reports Module Training Videos.

Select Question/Criterion:

Let’s look at the same data set in the Case Item Analysis: that is, we’d like to see the data only for those learners who selected Strongly agree for “Debriefing contributed to my learning.”

To select a dynamic subgroup in the Case Item Analysis, first select the question/criterion by clicking the radio button on the left.

Select Response:

Review Data for Dynamic Subgroup:

The report updates to display the data only for those learners who selected the Strongly agree response for “Debriefing contributed to my learning".

To return to the default data set, select No dynamic subgroups at the top of the report.