Yes, I read the report after that. RPM is cool and everything, but runs straight into the issue that it's easy to train RPM improvements of just 0.13 stddev - just do visual exercises and manipulations. (This is a big issue with interpreting dual n-back results.) And what is a laptop...?
Fundamentally, IQ tests are validated by their correlations with real world results (like the strong correlation with academic performance), and so any gains on IQ tests also require real world results if there's even a small chance of them being "hollow". (My standard example: you can improve your score on subtests by memorizing vocabulary, and this would improve your overall score. But has your underlying fluid intelligence increased? Probably not...)
And that's exactly what the small/null result for the academic scores shows.
And further, what we see especially (I'd say infamously) often with kid interventions is that the gains fade out within months or years. When tests are run in a few years, will we see even a 6-month gain on the RPM?
Fundamentally, IQ tests are validated by their correlations with real world results (like the strong correlation with academic performance), and so any gains on IQ tests also require real world results if there's even a small chance of them being "hollow". (My standard example: you can improve your score on subtests by memorizing vocabulary, and this would improve your overall score. But has your underlying fluid intelligence increased? Probably not...)
And that's exactly what the small/null result for the academic scores shows.
And further, what we see especially (I'd say infamously) often with kid interventions is that the gains fade out within months or years. When tests are run in a few years, will we see even a 6-month gain on the RPM?