The OnlineJobs.ph IQ Test: What Hiring Tests Measure and What They Don’t

If you searched for the OnlineJobs.ph IQ test, you probably want to know what it asks and what a good score is. OnlineJobs.ph, the Philippines-based job board where employers hire remote workers, appears in search listings with pages titled “IQ Test” and “Do I Need To Take Tests To Get Hired?”. Its pages returned a bot-check to our tools, so we could not read them, and this article does not state the site’s test provider, number of questions, time limit or scoring. What we can do is explain what the cognitive tests employers use in hiring measure, how their scores are reported, and what research says a score predicts. Before relying on any figure quoted elsewhere for the OnlineJobs.ph test, look for it on the site’s own page.

What do the cognitive tests employers use actually measure?

Publishers document their formats, so two well-known examples can be checked. The Criteria Cognitive Aptitude Test (CCAT) gives candidates up to 50 questions in a 15-minute limit, and most do not finish all 50. It measures three abilities: spatial reasoning, verbal ability, and maths and logic. The report shows a raw score out of 50 plus a percentile, with a percentile for each of the three areas; in Criteria’s sample report a raw score of 24 sits at the 50th percentile. Criteria describes the test as cognitive aptitude, often referred to as general intelligence or g. It draws on a large item bank, so candidates get different question sets, and it can be taken unsupervised on any device.

Wonderlic’s cognitive test is described on the publisher’s page as 50 questions with 12 minutes to answer, covering simple maths, basic logic, language comprehension, spatial reasoning and pattern identification. The page we read does not state a score scale or call the result an IQ; figures such as an average score of about 20 circulate on third-party sites, which we could not verify. Both tests reward speed as well as accuracy, which is why most people cannot finish. Our piece on Wonderlic scores and IQ shows how far a raw score can be stretched into an IQ.

Is a hiring-test score an IQ?

Not by itself. A raw score out of 50 is a count of correct answers within a time limit. A percentile ranks you against a comparison group chosen by the publisher or the employer. An IQ is a standard score with a stated mean and standard deviation, usually 100 and 15, taken from a norm sample. A percentile can be turned into an IQ-scale equivalent through its z-score (see our guide to IQ z-scores), but only if the comparison group is the general population:

• 50th percentile: z = 0.00, IQ-scale equivalent 100
• 75th percentile: z = +0.67, about 110
• 90th percentile: z = +1.28, about 119
• 95th percentile: z = +1.64, about 125
• 98th percentile: z = +2.05, about 131

A percentile against a pool of job applicants, or against people in one occupation, sits at a different place on the IQ scale, so the same percentile can mean a different IQ. The percentile calculator shows the conversion for any score on the standard scale.

We also found no published norm data for Filipino test takers on either the CCAT or Wonderlic. A percentile from one of them therefore says nothing certain about how you compare with other applicants in the Philippines, so ask the employer which group the percentile refers to.

What does research say a test score predicts?

The classic figure comes from Schmidt and Hunter (Psychological Bulletin, 1998), who reported an operational validity of .51 for general mental ability, corrected for measurement error in performance ratings and for range restriction, and .28 before the corrections. Range restriction is the statistical problem that people who were hired scored higher than typical applicants, so a study sees only part of the range. Sackett, Zhang, Berry and Lievens (Journal of Applied Psychology, 2022) argued that the validity of many selection procedures had been substantially overestimated, with mean validity estimates reduced by .10 to .20 points, and that structured interviews, not cognitive tests, ranked first. They traced much of the gap to over-correction for range restriction.

A 2024 paper by Sackett, Demeke, Bazian, Griebie, Priest and Kuncel (Journal of Applied Psychology 109(5), 687-713) looked at 21st-century studies only: 153 samples and 40,740 people, a mean observed validity of .16 and a corrected validity of .22. The authors put the 2022 estimate for general cognitive ability at .31 and concluded that cognitive ability relates to job performance but by less than earlier estimates suggested. Our own summary of the 2022 debate is in what a cognitive ability test at work predicts.

How should you use a test score?

• Applicants: find out which test it is, who publishes it, whether the result is a raw score, a percentile or an IQ, and which group it is compared against.
• Applicants: a large item bank means different candidates can get different questions, as Criteria describes for the CCAT, so a memorised list of answers is of little use; our piece on the practice effect covers what repeat attempts do.
• Employers: treat the score as one input and pair it with a structured interview, which ranked first in the 2022 reanalysis.
• Employers: check that a test does not screen out one group at a substantially different rate, as the four-fifths rule explains.
• Both: taking a test in a second language can under-measure ability, as our piece on second-language test scores shows.

The safest reading of any hiring test is narrow. It samples a few reasoning skills under time pressure and ranks you against a group. To see how a percentile lands on the IQ scale for yourself, take the IQ test, then use the score converter to see it on other scales.

Common questions