Work Samples
Asking somebody to do a piece of the job. The best-supported method available and the one most often designed badly.
The ordering between methods is more reliable than the numbers attached to them, and it is stable across analyses.
8 notes in this section
Asking somebody to do a piece of the job. The best-supported method available and the one most often designed badly.
The same questions, asked the same way, scored against criteria. A large improvement over conversation, for very little extra work.
The most-used method and one of the least predictive. Why the feeling of insight is so strong, and what to do if you cannot remove them.
Tests of reasoning and problem-solving. Reasonably predictive, legally fraught in some jurisdictions, and easy to use badly.
Popular, readable, and weakly predictive when used alone. What they can honestly contribute and what they cannot.
Scenarios with response options, scored against expert consensus. Middling prediction, good candidate reception, and easy to build badly.
Evidence the candidate brings rather than evidence you generate. Useful, cheap, and carrying a specific verification problem.
The methods against the things you might want to know, so the choice follows the requirement rather than the catalogue.
Start here
Independent guidance on skills assessment, selection design and fair hiring practice. External tools are included for practical comparison; evidence from the job remains the basis for decisions.