Randomized control trial of Tools of the Mind: Marked benefits to kindergarten children and their teachers

351 kindergarteners (172 Tools, 180 control), mean age 5.2, in 18 public schools = 9 clusters per arm, Vancouver and Surrey, British Columbia

Canada; kindergarten children.

Stratified school-level randomisation with both arms receiving equal training hours and materials funds, which is a genuine design strength. Two weaknesses dominate. There are NO laboratory executive-function measures at all - the outcomes are provincial reading and writing levels plus teacher online-survey responses, mostly analysed as class-level dichotomies with 18 clusters. And the author is the co-author of the 2007 Science paper on Tools and a long-standing advocate, with the paper framing prior positive findings as established.

By May, 33% versus 10% of children were reading at grade-1 level or higher (chi-square = 4.64, p = 0.02, odds ratio 3.25; F = 6.67, p < 0.02, partial eta-squared 0.33), and non-readers were 10% versus 28%. Writing odds ratio 26.18 (p < 0.001). Teacher-reported cliques odds ratio 15.77 (p < 0.001) and teacher excitement about teaching 3.58 (p = 0.02). The lower-income subgroup reading advantage was not significant (p = 0.12).

Low for the randomised contrast; the analyses rest on 18 clusters, which is the binding limitation.

Mixed and hard to reconcile. It measures no executive-function outcome while claiming to vindicate an executive-function curriculum, and it is contradicted on EF outcomes by three independent trials. Its reading result is the one thing the independent trials did not find.