4 ms·
The last sentence of this paragraph makes me distrust the whole analysis. Mixing percentages with population-dependent numbers makes no sense. And even if tha
by in_cahoots 6y ago
The last sentence of this paragraph makes me distrust the whole analysis. Mixing percentages with population-dependent numbers makes no sense. And even if that did make sense, the conclusion is exactly backwards. If fewer people go into engineering you would expect the few people who do make it to be of higher caliber. Obviously as you widen your pool you will regress to something closer to the mean programming ability.
- barbegal 6y agoAnd I don't consider being able to "write compilable code" as a good metric for programming ability. You can be a great programmer in many different ways: in optimising for performance, in optimising for testability, in increasing readability or in reducing the time taken to produce working code. Most people I know really heavily on an IDE to help them write compilable code because they want to use their brain power on solving issues the IDE can't help you with.