In an experiment, a group of people who have never programmed before have been showed how to code. In the end, their skills were evaluated in a coding test. The expectation was that they would be roughly normally distributed. However, the outcome was that the students were clustered in two groups (within each of which you see the expected normal distribution). The students belonging to the first struggled while the second group fared relatively well. The researchers figured out the cause: the students that did better managed to create a mental model of what a variable is (a container or a box that can hold values) while those that struggled didn’t manage to do this. Are the students from the second group inherently less intelligent?
Yes, because they did not manage to find the right model on their own. No, because once they were provided a straightforward explanation, they were able to code just as well as the students in the better group.
I’ve heard a similar story, but different in some details, and in that story the second group did not end up competent programmers. And I think I’ve also heard that the story I first heard didn’t replicate or something like that. So I don’t know what to think.
Probably this post; this claim has been highly controversial, with the original blog post citing a 2006 paper that was retracted in 2014, and whose original author wrote a meta-analysis that supported their conclusions in 2009. Here’s some previous discussion (in 2012) on LW. Many people have comments to the effect of “bimodal scores are common in education” with relatively few people having citations to back that up, in a way that makes me suspect they’re drawing from the original retracted paper.
I really appreciate you looking this up. Now it is clear that these claims are controversial.
Nevertheless, the question still is what learning ability refers to: Is it the ability to comprehend learning material that explains the topic well, or is it the ability to come up with the simple explanations yourself? It seems that the OP refers to the latter. The first kind probably has lower variation. Also, this means that when measuring learning ability, you should ensure that all involved people have access to the same “source”. I would be interested in hearing the OP’s thoughts on that.
In an experiment, a group of people who have never programmed before have been showed how to code. In the end, their skills were evaluated in a coding test. The expectation was that they would be roughly normally distributed. However, the outcome was that the students were clustered in two groups (within each of which you see the expected normal distribution). The students belonging to the first struggled while the second group fared relatively well. The researchers figured out the cause: the students that did better managed to create a mental model of what a variable is (a container or a box that can hold values) while those that struggled didn’t manage to do this. Are the students from the second group inherently less intelligent?
Yes, because they did not manage to find the right model on their own. No, because once they were provided a straightforward explanation, they were able to code just as well as the students in the better group.
Do you have a citation for this?
I’ve heard a similar story, but different in some details, and in that story the second group did not end up competent programmers. And I think I’ve also heard that the story I first heard didn’t replicate or something like that. So I don’t know what to think.
I couldn’t find it quickly, but I think that I read this on codehorror.
Probably this post; this claim has been highly controversial, with the original blog post citing a 2006 paper that was retracted in 2014, and whose original author wrote a meta-analysis that supported their conclusions in 2009. Here’s some previous discussion (in 2012) on LW. Many people have comments to the effect of “bimodal scores are common in education” with relatively few people having citations to back that up, in a way that makes me suspect they’re drawing from the original retracted paper.
I really appreciate you looking this up. Now it is clear that these claims are controversial.
Nevertheless, the question still is what learning ability refers to: Is it the ability to comprehend learning material that explains the topic well, or is it the ability to come up with the simple explanations yourself? It seems that the OP refers to the latter. The first kind probably has lower variation. Also, this means that when measuring learning ability, you should ensure that all involved people have access to the same “source”. I would be interested in hearing the OP’s thoughts on that.