Kenexa Test Scores Explained: Percentages, Percentiles and Who Sets the Pass Mark

“What score do I need on Kenexa?” has no single answer, partly because Kenexa, now IBM, sells two families of test that are scored in completely different ways, and partly because IBM does not set the bar at all. Here is what each kind of report contains, what IBM says about cut-offs, and where the commonly quoted numbers come from.

First, work out which kind of test you sat

If you are not sure which you were sent, our guide to telling Prove It and IBM Kenexa tests apart will help you identify it from the questions you saw.

How a skills test is scored

IBM documents its skills report field by field, which is rare. The headline Score is, “in most instances”, the number of questions answered correctly divided by the total. Alongside it the employer sees:

Two things follow for you. First, 70% made up of every Basic question plus some Advanced ones reads differently from 70% that misses Basic items, because the employer can see the level of each miss. Second, these tests are typically untimed, and IBM puts the average pace at about 40 seconds a question. There is no reason to rush, but the elapsed time is on the report, so do not drift either.

Some Kenexa tasks are not scored automatically at all. IBM describes its writing, translation and code samples as non-scoring environments: a person at the employer reads what you produced and judges it.

How a reasoning test is scored

IBM's public validation document is about the skills range, and we have not found a first-party IBM description of the reasoning test report. Prep sites agree on the broad model: your number of correct answers is converted into a percentile against a comparison group, such as other graduates or other applicants for similar roles. A 60th percentile means you outscored 60% of that group. No source we found, first party or otherwise, says Kenexa reports reasoning results as stens or stanines, so treat any site that gives you a Kenexa sten table with suspicion.

The adaptive CAT is the exception to counting correct answers. Because the questions get harder as you succeed, prep sites describe its score as reflecting the difficulty you reached, not just how many you got right. Our Kenexa practice explains how the CAT behaves beside the practice version.

What IBM says about pass marks

This is the part most score guides leave out, and IBM is unusually blunt about it. Its validation document says IBM “does not provide cut scores (pass/fail requirements)”, that interpretation “will vary greatly depending on the purpose of testing”, and that companies “should set their own proficiency levels and cut scores”. Where an employer wants a principled cut-off, IBM recommends a two-pass Angoff exercise, in which subject-matter experts judge how a minimally competent person would do on each question.

Then comes the line candidates should hear: “In practice, most companies set their cut scores as high as they can, given their existing applicant flow.” Put plainly, the bar often depends on how many people applied. A score that clears it at one company in a quiet month may fall short at a busier one.

Personality questionnaires have no score to beat

If your Kenexa link included a personality questionnaire such as the Rapid Personality Questionnaire, there is no overall mark. Prep-site descriptions agree that the RPQ reports a separate result for each of its five areas, compared with a general-population group, and that the longer OPI is usually compared with a UK graduate, managerial and professional group. The employer reads the profile against what the role needs. Our guide to personality tests covers how to answer them honestly and consistently.

What to do with this

Nothing is deducted for a wrong answer on the skills tests, because the score is correct over total, and none of the sources describes negative marking on the reasoning tests either. So answer every question. On a skills test, make sure you collect every Basic question, because those misses are visible by level. On a reasoning test, aim to move up the comparison group rather than hit a number, since you will not be told where the employer's bar is. The pre-employment tests guide puts Kenexa alongside the other tests US employers use.