Scientific American on image-based tests: what does a click reveal about personality?

You see a photograph of a lively team meeting. Does this describe you? You click without much hesitation. Your colleague chooses exactly the same answer. Yet you mean that you enjoy developing ideas aloud, while your colleague sees a familiar group in which everyone is heard. The same click can have different reasons.

In Scientific American on 24 January 2018, Heather Myers advocates image-based personality tests. Then chief psychology officer at assessment provider Traitify, she writes as a contributing author.

This is an interesting starting point for separating three questions: how pleasant is a test to complete, how well does it measure personality, and what can you do with the result? A convincing answer to the first question does not yet answer the other two.

A recognisable image has more than one meaning

Consider an imaginary test with a photograph of a tidy desk. It might be intended to represent planning and structure. But someone might choose it for the daylight, the absence of colleagues or the feeling that the working day is over. A caption can clarify the intention, but you still need to investigate how the image and text are understood together.

That does not make images unsuitable. It makes the connection between a response and a trait something to establish. A written statement can also be read differently. The relevant question is which interpretations a particular test invites and whether the resulting score adequately reflects the intended characteristic.

During a demonstration, therefore, ask for more than the chance to click through the test. Ask which response pattern leads to which conclusion, how that conclusion was investigated and what uncertainty accompanies it. A smooth experience does not reveal those things by itself.

Recognising quickly and assessing yourself are different tasks

Potter and colleagues found above-chance detection of named scenes presented for just 13 milliseconds. Their study concerned visual recognition, not personality measurement.

Recognising a photograph and deciding whether it describes your usual behaviour are different tasks. For the second, you may need to consider whether you are answering about your current job, your free time or the person you would like to be. That step does not disappear when an answer comes quickly.

The practical inference: treat completion speed as a feature of the user experience. Claims about measurement quality require research into the scores themselves. A quick click can be considered; a slow one can signal uncertainty. A stopwatch cannot tell you which is happening.

Assess the instrument as well as its underlying model

The Big Five describes five continuous personality dimensions. A test using that model still needs to show how its own questions, images and calculations produce useful scores.

A 2022 validation study by Hilliard and colleagues found associations between image choices and Big Five questionnaire scores. The authors called for further validation. This supports investigating image-based tests, not blanket approval.

Keep three kinds of evidence separate when evaluating an instrument:

A pleasant tool for a coaching conversation does not thereby have proven value for employee selection. Agreement with another personality questionnaire does not, on its own, establish who will perform well in a particular role.

Let the participant add meaning

Return to the photograph of the team meeting. Suppose the result prompts a conversation about contact and meetings. Start with a concrete question: “When does talking help you develop a good idea?”

The first colleague answers: “I often only know what I think when someone responds to my proposal.” The second says: “I enjoy contributing when I have been able to read about the subject beforehand.” Their shared image choice need not contain those practical differences.

Then ask for a recent example and an exception. In which meeting did this preference help? When did a different approach work better? This gives the result context and lets someone correct an interpretation. The profile supplies language; the participant supplies information about their own work.

That next step matters in Personal User Guide too. A description that feels familiar does not yet tell a colleague when to call, send something ahead or make space.

Make the next step more concrete than the test

Turn the conversation into an agreement both colleagues can follow. For example:

We share the question to be decided on the afternoon before the meeting. Everyone first writes down a proposal of their own. We then discuss the ideas aloud and record the decision.

This agreement makes room for preparation and thinking together. You do not have to prove that one working style is better. After two meetings, however, you can check whether more ideas were heard and whether the decision became clearer.

When you create a personal user manual, use the same approach: describe a situation, add an example, agree on mutual behaviour and set a time to see whether it helps.

For me, the interesting evolution of personality testing ultimately lies in that whole sequence. An accessible test opens the door. Good research determines which claims are justified. A useful manual makes clear what you can agree to do together tomorrow.