Dashboard calculations

Use this article to better understand how different calculations take place throughout both the Learner and Cohort profile dashboards.

Click the below links to learn more about each of the following calculations:

• Overall attendance percentage
• Achievement semester rollup
• Assessment weightings
• Performance rank
• 'Like student'
• ZPD (Zone of Proximal Development)
• Growth rates

---

➡️ Overall attendance percentage

Schools have a choice of having average semester attendance being calculated based on either 'Minutes absent from school', or 'Minutes absent from class'. You'll find further information about which attendance codes are included in these calculations through this article here.

➡️ Achievement semester rollup

There is a two-step process involved in presenting this overall achievement result:

1. In Admin → Customisation → Achievement, users with administrative privileges can exclude assessment categories pulled from the source data for achievement (or which can be defined in Data Management → Source Code Mapping)
2. Achievement categories not selected for inclusion will not contribute to a student’s achievement % for a class, which in turn, will affect the overall achievement percentage for a semester.

➡️ Assessment weightings

It is fairly common classroom practice to weight assessments differently, i.e. Assessment 1 may be worth 30%, Assessment 2 worth 30%, and Assessment 3 worth 40%. Calculating a semester total where class-based assessment weightings are concerned works as follows:

• If one assessment in a class is weighted, all assessments must be weighted, otherwise, our calculations will ignore the assessments with no weightings. This is because we can't calculate the weighted value of assessments with no allocated weighting where a teacher has indicated that weightings are being used (in that there is one assessment with a weighting).
• If there are no weightings, the results are averaged.
• It does not matter if the weightings do not equal 100 - we can account for this.
• With weighted assessments, we take the performance percentage for the first assessment, treating it as “all” of the “whole”. When the second is submitted, we re-calculate what the “whole” is based on the second assessment’s weighted value and re-calculate the performance percentage for each, then sum them. We continue this process for each subsequent assessment.

➡️ Performance rank

In the Relative Growth tab inside Cohorts, performance rank (PR) is calculated based on an average of all results represented in the scatter plot, and is representative of the student’s rank against other students who have taken the same assessment(s).

For 'Teacher judgements', PR is calculated based on the chosen student's average of all assessments included in their overall semester achievement calculation (this can be checked/modified inside the 'Settings' area), and it represents the student's rank against others in the same (selected) cohort who have recorded results in the same learning domain(s). It's worth noting that if you were to check a student's PR inside their year-level cohort, and then check again inside a class-level or subject-level cohort that this rank will differ.

Regarding performance rank for ACER testing, ACER benchmarks a student's scaled score against all of their historical data for that age group and that test. Students are subsequently assigned to the appropriate stanine or descriptive band for adaptive tests.

Growth score average is calculated based on the average growth percentage across all results in the scatter plot for which a comparative score exists.

➡️ 'Like student'

'Like students' are referred to in the Relative growth tab of the Learner Profile, where comparisons can be made between an individual student's growth and the growth rates of others within their cohort considered to have similar performance, a.k.a. 'like students'.

• By default, 'Like students' are defined as those within a 10% range of the student.
• If there are no students within the 10% range, the range is expanded to include those students who fall within 5% on either side of the student's result. 
• If there are more than 10 students within the 10% range, the range will decrease to 5% to provide a more accurate comparison.

➡️ ZPD (Zone of Proximal Development)

At a high level, we use a sliding window. The window's size (i.e. number of questions in the window) is proportional to the number of questions in the question bank.

Inside this sliding window, we check how many questions the student answered incorrectly or skipped, and if it exceeds the threshold, we mark the start of that window as the start of the ZPD. Currently, the max window size is 5, and the max threshold size is 3. So for most tests, the first instance the student gets 3/5 incorrect is the ZPD.

➡️ Growth rates

We've written an entire article dedicated to growth rates - take a look here.