Reliability of the International IQ Test

In this study, three groups of 66,032 different results were randomly selected over three different years (2020, 2021, and 2022) from the International IQ Test database, in such a way as to represent the world population each time (in a world where the population would be 80,000) to evaluate the test’s effectiveness and its IQ score calculation algorithm. Each group includes a proportion of results from every country in the world in order to represent the global population, based on the percentage of the world population that each country’s population represents in 2023.

For example, in 2023, China represents about 18.89% of the population. Therefore, 15,112 results from Chinese users were included in each group (18.89% of 80,000) for the year in question.

The results were filtered before selection, so as to only include authentic results (with no duplicates or bots). The same filter was applied to every country without exception. This same filter is used for the establishment of the IQ ranking by country each year.

The available data were sufficient to represent 82.54% of the world population over these three years within each selection group (66,032 / 80,000).

For the countries corresponding to the remaining 17.46%, there were not enough data to properly include them in the study without significantly reducing the total representation (80,000) in order to lower the minimum required candidates per country, and so they were ignored. However, they should not significantly alter the overall results.

The overall results for the three years, rounded to the nearest whole number, correspond to a standard deviation of about 15 and an average IQ of about 100.

This statistically suggests that the International IQ Test can provide a relevant indication of an individual's IQ score (within a few points) using the Raven's matrices method. However, the results of this test should be interpreted as indicative, and therefore do not replace a psychological consultation.