The cohort
Counts everyone who has recorded any activity in this namespace. Students working without signing in are recorded locally and appear here only once they identify themselves.
Do these three things before the next session
Generated from the data below, ranked by how many students each affects. If nothing appears, the cohort is not stuck anywhere the data can see.
Module drop-off
How many started, how many finished, and how long it took the median student. A module with high time and low completion is not hard — it is unclear.
Stall points — the section where people stop
Within each module, the share of students who started it and completed each section. The first sharp drop is where the page loses them. Section ids are as recorded by the page itself.
Exercises that mislead people
From the in-session “try it” classifiers. A wrong pick here is a named misconception, not a score — these are the items where students reached for the wrong answer first, ranked by how many of them did.
Workbook fields nobody answers
Fill rate across the cohort, worst first. A field left empty by most students is usually a badly-worded question or one that needs something the session never gave them — not laziness.