Select completed candidates on the Candidates list and choose Compare. You can compare between two and five candidates, and they have to be from the same assessment.
Reading the grid
Candidates are columns, tests are rows. For each test you see every candidate's score, with:
- the strongest score in that test highlighted, and
- a marker showing the test average, so you can see whether the whole group is strong or the winner is simply the least weak.
At the top, the overall score across the scored tests.
Why the same-assessment rule exists
Comparison only means something when everyone took the same tests, at the same difficulty, with the same time limits. Candidates from two assessments did not, so the product does not offer to line them up.
If you genuinely need to compare across roles, the honest answer is to look at individual test scores that appear in both, and to treat the overall scores as not comparable.
The trap in reading a comparison
The overall score is an average across the tests in the assessment. A candidate who is excellent at the thing the job needs and mediocre elsewhere can come out below a candidate who is uniformly average.
Read the rows before the top line. Decide in advance which test matters most for the role and look at that row first.
Personality in a comparison
Personality columns describe differences, not better or worse performance, and the view says so. Do not read a personality row the way you read a score row.
Exporting a comparison
Export comparison gives you the grid as a file, which is the usual way to get a shortlist in front of a hiring manager who does not have an account. See download and export results.
When one candidate has not taken everything
A test someone did not take shows as not taken rather than as zero. That is deliberate: a missing test is missing evidence, not a bad result. Their overall score covers the tests they did take, so compare it carefully.