I have some insight to share with you. The StackOverflow survey is far from reliable. Since it is largely a self-selecting survey, it is not statistically valid. Computerworld says that there are 18.2 million software developers around the world. The 64,000 survey participants represent about 0.35 per cent of the total. This sampling is completely insignificant and negligible.
Thus, the survey results are highly skewed and tell us nothing meaningful. How can we believe that 72.6 per cent of all the developers in the world program for the web? That’s more than 13 million web developers!
The survey can be, and very likely is, quite wrong. It wouldn’t be surprising. After all, last year ALL the political polls predicted that Donald Trump would lose the election. All of them.
So how much credence can we give to a sampling of 0.35 per cent of the world’s developers?