I lead test development and scoring at IQTests. The short version: I am the person who decides which items go into the test, how your answers turn into a score, and how that score is kept honest over time. I am not a clinician, and I will say so on every page, because the difference matters.
Who I am
My name is Marcus Hale, and I have spent more than a decade working on cognitive measurement and psychometrics, the unglamorous science of turning answers into numbers that actually mean something. I came to it through psychology and a long-running irritation with "free IQ tests" that hand out a flattering figure with no explanation attached. At IQTests I own the parts of the work that decide whether a result is trustworthy: the items, the scoring, and the norms. What I do not do is practice clinically, which is exactly why this site sends you to a licensed professional for anything with real stakes.
What I work on
My expertise is narrow and deep rather than broad, and here is the honest shape of it.
- Psychometrics and test theory. I build the test in line with the field's reference standards, working mainly within classical test theory: assessing reliability through coefficients in the Cronbach's alpha family and watching how individual items behave on real responses. A test is only as good as its weakest measurement property, and that is what I spend my time on.
- Cognitive ability and IQ. I work from the structure of intelligence that the research keeps recovering, from Spearman's g to the Cattell–Horn–Carroll framework: fluid and crystallized ability, working memory, processing speed. I also know the limits of an online estimate, which is why I always report the standard error of measurement rather than a single confident digit.
- The ICAR item set. Our items are modeled on the International Cognitive Ability Resource, the public-domain pool published by Condon and Revelle. I selected our items from it, balanced the four domains, and monitor how each item performs so that weak ones get reviewed or retired.
- Scoring and norming. I own the maths that turns a raw count into a percentile and a deviation-IQ score (mean 100, standard deviation 15), and I recompute the norms as data accumulates so your percentile is compared against a current sample. The full method is on the how we score page.
- Validity. I care whether the test measures what it claims, so I lean on validated items and check that the scores behave the way theory predicts before anything is published.
How I approach the work
My rule is to under-promise and document everything. I would rather tell you a score is "roughly 115" with an honest margin than pretend to a precision a ten-minute web test cannot deliver. I retire items that stop pulling their weight, I refresh norms instead of letting them quietly go stale, and I never sign off an instrument I have not assessed for reliability. Above all, I refuse to claim that an online screen is equivalent to a supervised clinical battery, because it is not, and saying so would be the easiest way to lose your trust. If you want to see the reasoning in full, it is laid out on our methodology page, and the standards I hold it to are in our editorial policy.
Get in touch
If you have a question about how your result was produced, a disagreement with one of our items, or a correction to something I have written, I would like to hear it. The fastest route is [email protected], or the contact page, which reaches a real person. For the wider story of the project and the team around it, see our about page. Profile last reviewed 30 June 2026.