“Chess Instruction Improves Cognitive Abilities and Academic Performance: Real Effects or Wishful Thinking?”, 2022-04-02 ():
In accordance with the outcomes from a number of reports, there are cognitive and academic improvements derived from chess learning and chess playing. This evidence, however, endures 3 key limitations: (1) ignoring theoretical premises about the concept of transfer, (2) several shortcomings regarding ideal experiment guidelines, and (3) an uncritical faith in null hypothesis statistical-significance testing (NHST) statistical analyses.
The present review scrutinized the NHST outcomes from 45 studies describing chess instruction interventions (n = 12,705) in 19 countries that targeted cognitive ability (100 tests) and academic performance (108 tests), with a mean Hedge’s effect-size:
g = 0.572 (95% CI = [0.127, 1.062]). There was a lower average statistical power, a higher proportion of false positive outcomes, larger publication biases, and lower replication rates for the studies in the academic performance domain than in the cognitive ability domain.
These findings raised reasonable concerns over the evidence about the benefits of chess instruction, which was particularly problematic regarding academic achievement outcomes.
Chess should perhaps be regularly taught, however, regardless of whether it has a direct impact or not in cognitive abilities and academic performance, because these are far transfer targets. The more likely impact of chess on near transfer outcomes from higher quality studies remains at present unexplored.
[Keywords: chess instruction, cognitive ability, academic performance]