Video Games and IQ: The 2.5-Point Gain Was an Association, Not a Promise

The widely shared 2.5-point finding remains published, but it does not show that gamers generally have higher IQs than non-gamers. A Karolinska Institutet research summary explains that US children who played more than average showed an approximately 2.5-point larger intelligence gain over two years; they were not more intelligent at the study’s starting point.

Evidence added since the original 2025 article has not converted that association into a universal brain-training claim. A 2025 prospective Ontario cohort study instead associated any reported video-game use with lower grade 3 reading achievement, while finding insufficient evidence of an association with its other combined-sample gaming outcomes. The two results concern different children and different measures, so neither cancels the other.

What the 2.5-point result actually measured

The intelligence finding came from a longitudinal analysis of US participants in the Adolescent Brain Cognitive Development project. Children initially aged nine or ten completed a battery of cognitive tasks, while they and their parents supplied information about time spent gaming, watching videos and using social media. Participants with follow-up data repeated the cognitive assessment two years later.

The researchers combined several tasks into a general intelligence measure and adjusted their statistical model for factors including socioeconomic circumstances and genetic differences associated with cognition. Gaming had almost no relationship with intelligence at the initial assessment. The positive result appeared in the change between the first and second assessments among children who reported more gaming.

That distinction matters. The figure was an estimated difference in cognitive development relative to the sample average, not a fixed bonus awarded to every player and not a comparison showing that all gamers had higher IQs than children who never played.

The design also remained observational. Researchers did not randomly assign children to gaming and non-gaming groups, and playing time was self-reported rather than continuously measured. Statistical controls can reduce some alternative explanations, but they cannot capture every relevant difference in family life, motivation, previous ability, game selection or activities displaced by gaming.

A sibling analysis within the same research produced no statistically significant relationship between gaming and intelligence change. That secondary result does not erase the primary model, but it adds uncertainty to a causal interpretation. The defensible conclusion is that greater reported gaming was associated with a modest additional gain in one intelligence composite.

Other cognitive research found differences, but not higher IQ

A separate comparison examined children who reported substantial weekly gaming and peers who reported none. The corrected version of the 2,217-participant cognitive study found statistically significant but very small advantages for gamers on tasks involving response inhibition and working memory, alongside different patterns of brain activation during those tasks.

Those findings are narrower than a claim of superior general intelligence. Faster responses on an inhibition task and better performance on a working-memory exercise do not establish an advantage across language, reasoning, knowledge and other cognitive domains. Brain-imaging differences likewise show that groups processed the tasks differently, not that one group had a better brain.

The comparison was cross-sectional, with gaming habits and cognitive performance assessed at the same stage. It therefore could not establish whether gaming influenced task performance, children with particular cognitive strengths were more attracted to games, or another characteristic affected both.

The article also underwent retraction and replacement to correct analytical errors, followed by further clarification. Its current language is restrained: the reaction-time differences were very small and lacked clinical relevance. Gamers had higher scores on several measures of attention and mental-health symptoms as well, but the same cross-sectional limitation prevents treating gaming as their demonstrated cause.

Why the school-results evidence is different

The newer Canadian analysis followed children from earlier parent reports of screen use to standardized assessments in elementary school. Its negative gaming association was limited to reading in grade 3 in the combined sample. It did not establish a corresponding combined-sample relationship for the other gaming and school-subject combinations examined.

This result complicates claims that a cognitive association automatically produces better classroom performance. An intelligence composite, a response-inhibition task and a curriculum-based reading assessment measure related but non-equivalent abilities. A child can become faster at a particular computerized task without necessarily acquiring more vocabulary, reading more accurately or performing better on school assignments.

The studies also differed in country, age, exposure definition and timing. The US intelligence analysis focused on reported playing time among preteens and change in cognitive scores. The Canadian study examined earlier parent-reported use and later provincial achievement levels. Calling their results contradictory would ignore those methodological differences.

Academic performance is also shaped by instruction, attendance, sleep, reading practice, household resources and numerous other factors. Neither study isolated gaming from every competing influence, and neither identified a mechanism by which a particular commercial game improved or reduced learning.

What the evidence supports

The available research supports a limited proposition: gaming has been associated with modest advantages or gains on selected cognitive measures in some groups of children. It does not support the stronger statement that gamers as a class have higher IQs than non-gamers or that increasing a child’s playing time will reliably raise intelligence.

The evidence also cannot identify an ideal genre, title or schedule. Broad measures of “gaming” often combine action, strategy, creative, cooperative and repetitive experiences that place very different demands on players. Results based mainly on time cannot be confidently transferred to any one game.

Three boundaries keep the interpretation accurate:

• The 2.5 points describe an estimated additional change associated with greater gaming, not a guaranteed IQ increase.
• Small advantages on response-inhibition or working-memory tasks are not evidence of superior cognition in every domain.
• Cognitive scores must be kept separate from academic achievement, sleep, mental health and social development, which require their own evidence.

Video games can contain demanding problems, rapid feedback and opportunities to practise attention or spatial reasoning. That makes cognitive benefits plausible, but the current evidence does not justify prescribing more gaming as a general intelligence intervention. The scientifically supported claim is a modest, conditional association—not proof that playing video games makes every child smarter.

Also read:

• 11 Ways to Gain Customer Trust and Boost Sales
• New Mexico’s AI Teen Decoy Was Evidence Against Snap, Not an Arrest Sting
• 7 Mind-Blowing Statistics that Show Why Video is the Future of Content Marketing