Analytics

The Analytics tab aggregates every session in the task into cohort-level insight. Where Test Sessions answers questions about individuals, Analytics answers questions about the whole exam: how did the population perform, and how does it compare to everyone else who has taken the same test.

Participation Totals

The header shows the task's raw volume at a glance — total participants, total test takers, and total submissions — with a note of when the analytics data was last updated.

Per-Step Analytics

Each step gets its own analytics card with a score summary and two tabs:

• Test analytics — the cohort's candidate count and score breakdowns by section (count, average, min, max) and by dimension (plus standard deviation), each compared against the norm group average — the average of everyone who has taken this test across all tasks. That comparison is what tells you whether your cohort is strong or the test is easy.
• Question analytics — the per-question view for this step's test, linking into the full question analytics report.

Steps running a marketplace test are marked as external — their analytics reflect your cohort, while question-level detail belongs to the test's owner.

Task Analytics vs Test Analytics

This tab is about this task's population. The quality of the test itself as a measuring instrument — item difficulty, discrimination, reliability — lives in the test editor's Test Analytics, aggregated across every task the test runs in. Use task analytics to judge the cohort; use test analytics to improve the questions. See the Analytics chapter for how the underlying reports work.