Game-based assessment vs traditional psychometric tests
A traditional psychometric test asks candidates to rate themselves or answer items with a right answer. A game-based assessment records what they do in short games: their choices, their speed and how they react under time pressure. Games are usually faster to score and better liked by candidates, while traditional tests have a much longer research record, so a sensible split is to use games to shortlist and keep a structured interview for the final decision.
Key points
- Self-report questionnaires can be faked: job applicants score themselves higher than non-applicants, most of all on conscientiousness and emotional stability.
- Games measure behavior, so there's little to rehearse. Independent research on faking in games is still thin, though.
- Candidates tend to react a little better to game-based and gamified assessments, and see the employer as more attractive.
- Traditional tests have decades of validity research behind them. Game-based assessment is younger, so ask for each vendor's own validation.
- The two work well together. You don't have to switch everything at once.
Choosing a test for your next MT intake?See how DEUS assesses Management Trainees
By DEUS Human Capital Services ·
What counts as a traditional psychometric test?
By a traditional psychometric test (psikotes), we mean one of two kinds of test, or both:
- Self-report personality questionnaires. Candidates read statements such as "I enjoy meeting new people" and rate how well each one describes them.
- Ability tests with right answers. For example, number series, verbal analogies, abstract reasoning. Candidates work through items on paper or on a screen, often against a time limit.
Both are well understood, and the best of them have strong research behind them. The trouble starts when the stakes are high and the volume is large, which describes most Management Trainee and campus intakes.
How do they compare?
| Factor | Traditional psychometric test | Game-based assessment |
|---|---|---|
| Faking | Self-ratings can be shaded toward what the employer wants, and answers to common tests are a quick search away. Ability items can be coached or memorized. | No statements to agree with and no answers to memorize. The data is how the candidate plays. |
| Drop-out | Long, repetitive tests lose candidates before they finish. | Short games with a clear goal. |
| Speed | Scoring, compiling and formatting results can take an HR team days. | Scores and reports are generated as each candidate finishes. |
| What gets scored | Right or wrong answers, and how candidates describe themselves. | Choices and accuracy, plus speed, consistency and hesitation. |
| Candidate experience | Familiar. Long versions can feel repetitive. | Better liked in research, and seen as fairer. |
| Research record | Decades of validity studies and published norms. | Promising, but the research is young. |
Faking
- Self-ratings can be shaded toward what the employer wants, and answers to common tests are a quick search away. Ability items can be coached or memorized.
- No statements to agree with and no answers to memorize. The data is how the candidate plays.
Drop-out
- Long, repetitive tests lose candidates before they finish.
- Short games with a clear goal.
Speed
- Scoring, compiling and formatting results can take an HR team days.
- Scores and reports are generated as each candidate finishes.
What gets scored
- Right or wrong answers, and how candidates describe themselves.
- Choices and accuracy, plus speed, consistency and hesitation.
Candidate experience
- Familiar. Long versions can feel repetitive.
- Better liked in research, and seen as fairer.
Research record
- Decades of validity studies and published norms.
- Promising, but the research is young.
The rows need some explaining.
Faking. A meta-analysis of 33 studies compared the personality scores of job applicants with those of people who weren't applying for anything. Applicants scored higher on four of the Big Five traits. The gaps were largest for conscientiousness and emotional stability. In some jobs, such as sales, applicants inflated the traits that looked most relevant to that job. Games remove the questions that make this possible. To be fair to the evidence, though, a 2022 systematic review of game-related assessments found only one study that had tested faking directly. Its result was positive, but the review says the claim still lacks scientific support.
Drop-out. There's little independent research comparing completion rates for games and questionnaires of the same length. One test publisher's data from hundreds of thousands of assessments suggests about 80% of candidates finish assessments up to 40 minutes long, with more drop-off beyond that. Whatever format you choose, keep it short and run it in a browser.
Speed. This is where the difference is largest in practice. When a national telecom group ran its Management Trainee intake on DEUS, more than 1,000 candidates in 30+ provinces were assessed at once, with results within a day.
Candidate experience. In one experiment, applicants who took a gamified situational judgment test reported more satisfaction with the process than those who took the text version of the same test. They also rated it fairer and found the company more attractive. The 2022 review found the same pattern across studies, though the gain was usually modest, and noted that negative reactions usually come from technical problems, not the assessment. Our own participants have told us it didn't feel like a psychotest at all.
Where does the traditional test still win?
Honesty helps here, because you'll be defending the choice to your own management.
- Track record. A 2022 re-analysis of the selection research still places cognitive ability tests among the top predictors of job performance, at a validity of .31. Game-based assessment doesn't have a body of evidence that size yet. The 2022 review found no evidence that game-related assessments predict performance better than other methods.
- Familiarity. Your in-house psychologists know how to read the results, and so do your hiring managers.
- Small numbers. For a handful of senior hires, the scoring time saved is small, and a well-run structured interview matters more. The same re-analysis ranked structured interviews first, at .42.
Where does game-based assessment win?
- Volume with little time. Invitations, reminders, progress tracking, scoring and reports all run on the platform, so an intake of a thousand doesn't need a bigger HR team.
- Less to rehearse. The games measure behavior, so there's little to look up.
- How people behave. The DEUS Neuroscience games show motivation, appetite for risk, effort allocation and emotion recognition through behavior people don't consciously control. Motivation in particular is easy to misjudge in a one-hour interview.
- Local norms. Every DEUS assessment is normed on the Indonesian population, so a percentile compares your candidate with Indonesians instead of a Western reference group.
- Your employer brand. Candidates come away seeing the company as modern and serious about talent.
Can you use both?
Yes. The 2022 review found that game-related assessments seem to benefit from being used alongside conventional tests.
A practical split for a Management Trainee intake:
- First pass: the DEUS Cognitive and Neuroscience games for the whole intake, with a cut-off score that flags candidates below it automatically.
- Final round: the Simulation module, where finalists analyze a business case and allocate a budget against a live timer, in a scenario set in your industry.
- Decision: a structured interview with the people still in the running.
If you want personality in the process without a self-rating questionnaire, DEUS Discover is one route. It's a situational judgment test delivered as a story, built on the Big Five, and its results can be framed around your own competency framework. Candidates choose what to do as the story unfolds, rather than rating themselves.
How do you switch without the risk?
Don't replace your whole process in one go. Start with a free trial, or run a pilot with one division, one role or one batch. Running the games and your current test on the same pilot batch gives you a direct comparison, in your own company, before you commit. DEUS has no long-term contract: pay month to month or per batch, from IDR 69,000 per candidate.
Questions about game-based assessment and psikotes.
Is a game-based assessment a kind of psikotes?
Yes, in the sense that it's a psychological assessment for hiring. It's built like one too: the DEUS games go through theory, item development, validation and norming, guided by psychometricians. The difference is that candidates play games instead of filling in a questionnaire.
Which is more accurate?
Neither format is accurate in itself. Accuracy depends on how a test was built and validated for your use. Traditional tests have more published research. Our cognitive games have been validated against traditional cognitive tests, showing convergent validity.
Can we keep our current test and add games?
Yes. You can start with one module, such as Cognitive, in the first stage of an intake, and add others later.
Do candidates need to install anything?
No. DEUS runs in any phone or laptop browser, and every session starts with device checks and anti-cheat logic.
See the four DEUS modules, and what each one measures.
See how the games workSources
- Birkeland, Manson, Kisamore, Brannick and Smith (2006), "A meta-analytic investigation of applicant faking on personality measures," International Journal of Selection and Assessment (opens in a new tab)
- Ramos-Villagrasa, Fernández-del-Río and Castro (2022), "Game-related assessments for personnel selection: A systematic review," Frontiers in Psychology (opens in a new tab)
- Millet (2022), "How much testing is OK?", Criteria Corp (opens in a new tab)
- Georgiou and Nikolaou (2020), "Are applicants in favor of traditional or gamified assessment methods?", Computers in Human Behavior (opens in a new tab)
- O'Shea and Fox Luscombe, "Is cognitive ability the best predictor of job performance?", on Sackett, Zhang, Berry and Lievens (2022), TIP, SIOP (opens in a new tab)