ChatGPT and IQ: Test Scores by How Often People Use AI

Everyone took the same free online reasoning test on MyIQtested. After the last puzzle, and before the score appeared, the site showed one survey card. The order was: the 33 puzzles, the score calculated in the browser, an age card, one survey card, then the results page, where the score is first shown.

The question is chosen at random for each taker from the 4 questions live at the same time, among those their age group qualifies for. Each question stays live for 7 days, then the next question in the queue takes its place. The card can be skipped, and a row is only written when someone taps an answer or Skip: a card closed without a tap writes nothing.

This question was live from 22 Sep to 29 Sep 2026. It was shown 33,677 times and answered 30,491 times (90.5%).

The test has 33 multiple-choice puzzles, each with four options and one correct answer: 20 matrix reasoning puzzles (ICAR-based, from the International Cognitive Ability Resource framework), 6 series and logic, 4 verbal and 3 numerical. Every puzzle must be answered and there is no time limit; the median first attempt takes about 9 minutes.

The score is a fixed conversion of the number of correct answers: 51 + 3 × correct, with a floor of 55 and a ceiling of 150, so scores move in steps of 3. It is calculated in the taker’s browser and is not an empirical norm. That is why this page reports differences and ranks, never a level.

The verbal puzzles are re-authored for each language; the other 29 keep the English structure and answer key. The language versions are therefore not exactly the same instrument.

Shown the question

33,677

Answered

30,491 (90.5%)

Skipped

3,186 (9.5%)

Window

22 Sep to 29 Sep 2026

Age

Of the 30,491 answers, 57% came from people under 18 (including 21% under 13), 40% from adults and 3% from people who did not give an age. This question had no age limit, so under-13s were shown it.

Country

The largest groups were Indonesia (60%), United States (9%), Japan (8%), India (6%). 13 countries have their own column; the rest, 8% of answers, are pooled as Other countries.

Language

The page the question appeared on was in 12 languages; the largest were Indonesian 58%, English 30%, Japanese 8%, Russian 2%.

• No row is filtered out when the figures are built: no bot, duplicate-visitor or score-range filter is applied. The same rows feed our internal report.
• Automated browsers and local development are kept out at the source: the survey is never sent from localhost, or from a browser reporting itself as automated. The answer service rejects unknown questions, answers that are not one of the options, and age-gated rows.
• Repeat takers are kept. 2,547 of 33,677 rows (7.6%) come from a visitor with an earlier row for this question, each a separate attempt. Removing them moves the group averages by between 0.3 and 0.8 points.
• People who skipped the question scored 3.8 points lower on average than people who answered it (not adjusted). Skipped cards are not in any figure except the answer-rate bars.
• Countries with fewer than 150 answers are pooled into Other countries, as are rows with no country. Age not given pools “prefer not to say”, a skipped age card and missing values.
• Cells with fewer than 5 people are pooled into Other countries before publication (138 cells, 339 people), so no cell can identify one person’s score. The aggregate CSV suppresses those cells instead.
• Any group or subgroup with fewer than 30 people is shown greyed as “n too small” and no interval is drawn.

Every figure is relative. Nothing on this page states how high anyone or any group scored on the scale, only how groups differ from each other and where people rank among people of their own age and country.

Every range is a 95% confidence interval, and every difference we point out holds at 95% confidence after correcting for the number of comparisons (Holm method).

Strata and the standard population

A stratum is one age band within one country. Ten age bands and 14 country columns make the strata. With Adjust for age and country on, the standard population is the current filter’s own stratum mix: w_s is the share of everyone in the filter who is in stratum s. With it off, everyone is pooled.

Gap versus all respondents

Coverage is the share of the standard population that the answer has people in; strata where nobody gave the answer are dropped and the rest re-weighted. Coverage is shown in the table when Adjust is on.

Intervals

The factor (1 − p) is there because the reference group, everyone, contains the answer. A variance estimated from fewer than 5 people is replaced by the variance of everyone in the filter. With Adjust off there is one stratum and this is the ordinary two-sample formula.

Effect size

Cohen’s d compares an answer with everyone else: d = (mean of the answer − mean of the rest) / pooled SD, using the same strata weights when Adjust is on. It is in units of the spread of scores, not in points.

Percentile within own age and country

Scores are grouped in 3-point bins and each bin is treated as a single score value. Within a stratum, everyone who shares a bin is tied and shares the range of ranks that bin covers, so the average of their percentiles is the mid-rank percentile. A person’s percentile is their rank among everyone who answered in the same age band and country, whatever they answered. Strata with fewer than 30 people are left out of the percentile figures.

The line in each hero column is the group’s average percentile, and the median in the table comes from the same distribution. With Adjust on, each answer’s distribution is the stratum-weighted average of its per-stratum distributions.

Personality figure

This question also ran on SeeMyPersonality, after its Big Five test: it was shown 1,847 times and answered 1,685 times. Each trait is the percentile (0 to 100) the person saw on their results, against SeeMyPersonality’s own empirical norms. The gap is computed as for Figure 1, on those percentiles, for all ages and countries. With Adjust on, the standard population is the mix of five age groups (under 18, 18 to 24, 25 to 34, 35 and over, not given) and 3 countries plus Other among everyone who answered. Only the gaps and their intervals are published, not the cells they come from, and a group of fewer than 30 people shows no numbers.

Answer rate

Answered divided by shown, by age band, with a Wilson 95% interval.

• The people are volunteers who chose to take a free online test. They are not a sample of any population, and country and age mixes are those of the site’s visitors.
• Each answer is a single self-report item with four options. People may read the options differently.
• The score comes from one unsupervised sitting of a 33-puzzle test. It is not a clinical measure, and its scale is a fixed conversion, not a norm.
• 57% of answers are from people under 18, including 21% under 13, who may read the question and the test differently from adults.
• Country is recorded by the site from the visit; it may not be the person’s nationality or where they grew up.
• People who skipped the question differ from those who answered, and the answer rate for age not given (48%) shows that the group who answered is not everyone shown.
• Differences are shown with intervals, but an interval covers sampling noise only, not the ways the sample was selected.
• The personality figure comes from a different group of volunteers, people who took the SeeMyPersonality Big Five test, and from 1,685 answers, so its intervals are wider.

Dataset version

2026-09-29.1

Generated

2026-09-30

Responses

22 Sep to 29 Sep 2026

Licence

CC BY 4.0

Change log: 2026-09-29.1, first release. If the data or the method changes, the version changes and this line says what moved. Cite the version you used: myiqtested.com/research/chatgpt-use-and-iq.