Article
Separating thinking from reaction speed: how IDoCT scores results
Most cognitive tests reduce performance to a simple total. Cognitron's IDoCT scoring separates a person's cognitive ability from how quickly they physically see and respond.
16 October 2026
Simple totals, such as correct answers multiplied by reaction time, mix several things together: how quickly a person sees and responds to a prompt, how fast their device is, quirks of test design, and their actual cognitive function.
Cognitron uses a scoring method called IDoCT (Iterative Decomposition of Cognitive Tasks). It splits each person's performance into two measures: cognitive ability, which is how well they handle harder problems, and visuo-motor delay, the time taken to see, move, and respond, including any lag from the device.
Speed and accuracy are weighed together, so people who work quickly but make more errors are scored fairly against those who work slowly and carefully. Rather than assuming how hard each item is, IDoCT recalculates difficulty from how hundreds of thousands of people actually performed. Scores are then compared with a large normative dataset, adjusted for age, sex, education, first language, and device.
Validated on data from 388,757 adults, and on brain imaging from 34,927 UK Biobank participants, the method improved test-retest reliability (r = 0.77 vs 0.66) and produced measures linked to distinct brain regions.