CollegeBoard: Analyses of CS Study Benefits Shouldn't Be Interpreted As Causal

theodp writes: Code.org, backed by some of tech's wealthiest individuals and their companies, is this close to getting computer science declared a 'core subject' in K-12 public schools. So, when the non-profit recently asked CollegeBoard for more evidence that learning computer science is linked to improved learning in other subjects, it must have been disheartened by the study results. "The purpose of this brief note," wrote the CollegeBoard, "is to document some exploratory analyses linking participation in AP Computer Science to subsequent performance in SAT Mathematics and AP Calculus and Statistics. None of these analyses should be interpreted as causal. Although there appears to be a relationship between AP CS participation and subsequent outcomes, it is highly likely that this is the result of one or more omitted and confounding characteristics of students that are not able to be controlled for given this research design."

Comments

Why children should NOT be taught to code

[XxtraLarGe]

I had submitted this post about two weeks back, but it never made it to the front page. FTA:

I've written about this at some length in my book Beyond Technology. The argument depends upon assumptions about learning transfer -- the idea that learning in one context will automatically transfer across to others. This is to conceive of the brain as a kind of muscle: a good workout in the coding gym will have payoffs when we need our logical thinking skills to solve problems elsewhere. Similar claims are often made for learning the game of chess, or Latin. Yet there is no convincing evidence that learning computer programming enables children to develop more general problem-solving skills, let alone that it will 'teach you how to think', as its advocates claim.

While it seems intuitive that programming develops logical thinking, it may be the case that people who program already possessed that skill and programming merely reinforces it.