The Shortcut That Passed the Test

The fastest applicant never opened the catalogue.

A public archive is hiring a researcher. The final exercise gives candidates thirty minutes to find a flood map from 1962 and identify the notes filed with it. Most begin at a computer. June reads the prompt, goes to the service desk, and asks the archivist on duty which collection might contain the map. Eight minutes later she has the right folder and is checking its notes.

Two people on the hiring panel are impressed. The third is irritated. The point, he says, was to see whether applicants could search independently. June has taken a shortcut.

A result is not a method

If the job requires someone who can work alone with a difficult catalogue, he has a point. June has demonstrated that she can ask a useful question and verify an answer. She has not demonstrated how she would proceed without another person available.

But the instructions asked candidates to find the material, not to find it without speaking. Penalizing June now would mean judging her against a rule that became visible only after she succeeded. That is a poor way to test whether someone follows instructions.

The advantage the test forgot

There is a complication in June's favor, too. Researchers work with people. Knowing when to ask a specialist can save hours and prevent an avoidable mistake. A test that rewards solitary persistence above all else might miss a skill the archive actually needs.

Still, not everyone walks into a strange institution equally comfortable asking staff for help. Someone who already knows the building, or knows whom to approach, has an advantage. Calling every shortcut ingenuity could turn unequal access into apparent talent.

A hiring panel should be able to explain which abilities it is measuring, rather than choosing whichever explanation flatters its preferred result.

What to do with June

For this round, the panel can acknowledge what she achieved and what the exercise cannot establish. They can ask every finalist a separate, clearly defined question about independent searching. Next time, they can design two tasks: one with specified resources and no assistance, another that invites candidates to use the archive as they would on an ordinary working day.

That will not produce a perfect ranking. It will at least separate the candidate's ingenuity from the panel's surprise.

The awkward possibility is that June did not break the test at all. She revealed that its designers had never agreed on what passing meant.

If you were on the panel, would you reward the result—or change the test before deciding who did best?