We have spent a lot of time improving the AREALME IQ Test this year. We removed weak questions, refined the scoring model, and opened a public data page. Now we want to give the data behind the score better context.

Starting with Two Optional Questions

The English test is beginning to ask for an age range and gender before the test starts. Both questions are optional. We do not ask for a date of birth, name, email address, or account, and these answers do not change the score someone receives today.

Why ask at all? A score can look different across age groups, and a good model should be able to check whether a question behaves differently for different groups. Until now, our public sample could not answer that question.

Collecting the fields is only the first step. We will not rush to create a new norm from a small early sample. We want enough completed, uninterrupted tests in each group, followed by a second time window that confirms the same pattern.

Checking Devices Without Overreacting

Phones, tablets, and desktop computers can affect online tasks, especially when speed, screen size, or precise movement matters. That does not mean every mobile score needs a correction.

Our recent data does not show one simple device advantage. Desktop results are slightly higher in the Quick test, while mobile results are slightly higher in the Complete test. That mixed result is a reason to study individual questions, not a reason to create separate mobile and desktop scores.

We will treat device type as a quality check. If the same question shows a meaningful difference across repeated samples, we will investigate the design of that question first.

This approach is informed by research on device effects in online cognitive testing and the ITC/ATP guidelines for technology-based assessment.

A Higher Bar for Scoring Changes

Future calibration work will stay tied to the current question bank. Attempts that were paused and resumed will continue to be excluded from the modeling sample, even though people can still finish the test and see their result.

We will also publish the rules before they affect scoring. A change should have a clear reason, enough data behind it, and a result that holds up in a later sample.

We Will Keep Showing the Work

The AREALME IQ Test is an online cognitive assessment for learning and self-exploration. It is not a clinical diagnosis, but that does not mean the work behind it should be casual or hidden.

As the new sample grows, we will update the Data Disclosure Center with what we find, including the limits of the data. If the evidence supports a scoring change, we will explain it. If it does not, we will say that too.