In a word, no
The sample size is reportedly 100 kids, all from Berekey. Hardly a representative sample of the nation at large. The correlation was 0.27, a very poor correlation as shown here. In short, it's yet another example of a poorly executed study that only serves to reflect the author's bias.
