4 ms·

Because thinking back over all the people I've programmed with or inherited a codebase from So basically, when considering a minuscule, poorly defined and poor

Because thinking back over all the people I've programmed with or inherited a codebase from So basically, when considering a minuscule, poorly defined and poorly measured data-set, you've decided to draw broad conclusions about half the human population. I'm not sure how great a programmer you are, but I'll tell you right now that your logic skills need some work.

lmm 12y ago
Broad conclusions? No, more like weak inferences. Not everything can be a formally-verified proof; a vital skill in programming (for debugging if nothing else) is the ability to make a best guess based on limited data (and to know how much weight to give such guesses). (If there were reliable science on the subject I would of course prefer to go by that, but that's practically impossible for a question that's so politicised)
- vectorpush 12y ago
  No, more like weak inferences Indeed, quite weak; weak to the point of being useless. In ten years of programming, I've only worked with two developers who were comfortable with the use of regular expressions, both were women... what does that tell me about women? Nothing. Similarly, whatever experiences you've had with female programmers who are "worse than the average male" programmer is pretty clearly worthless (for the purpose of drawing conclusions about female programmers), especially when one considers that your definition of average is, at best, a nebulous amalgamation of your own subjective impressions.
  - lmm 12y ago
    You're right that two probably isn't enough for a significant result - though it might be, if the effect size is large enough. How large was the rest of your sample, and how big was the difference? I won't stoop to insults the way you did, but are you really not even curious about these kind of differences?
    - vectorpush 12y ago
      I won't stoop to insults the way you did How noble of you. You're right though. I apologize for disparaging your faculties of logic, it was tongue in cheek but wholly unnecessary. though it might be, if the effect size is large enough. How large was the rest of your sample, and how big was the difference? How can effect size be determined when we don't even know what is being measured? What is the criteria for determining the quality of a programmer? How is that criteria applied? Are there controls for education? exposure to programming as a child? years of experience? quality of experience? quality of tools available to the programmer? quality of the work environment in which we seek to evaluate the programmer's quality? It's reasonable for one to leverage professional experience in judging the skill level of colleagues, but it'd be foolish to think that those observations can be generalized across an arbitrary dimension of the programmer population (such as gender, or race, or political leaning). are you really not even curious about these kind of differences? I guess? I'm always eager to learn more, but I don't really find the question of programmer aptitude as a function of gender very interesting since I can't identify anything unique about programming that would suggest either gender would be qualitatively predisposed to the craft.

lmm 12y ago

Broad conclusions? No, more like weak inferences. Not everything can be a formally-verified proof; a vital skill in programming (for debugging if nothing else) is the ability to make a best guess based on limited data (and to know how much weight to give such guesses). (If there were reliable science on the subject I would of course prefer to go by that, but that's practically impossible for a question that's so politicised)

vectorpush 12y ago
No, more like weak inferences Indeed, quite weak; weak to the point of being useless. In ten years of programming, I've only worked with two developers who were comfortable with the use of regular expressions, both were women... what does that tell me about women? Nothing. Similarly, whatever experiences you've had with female programmers who are "worse than the average male" programmer is pretty clearly worthless (for the purpose of drawing conclusions about female programmers), especially when one considers that your definition of average is, at best, a nebulous amalgamation of your own subjective impressions.
- lmm 12y ago
  You're right that two probably isn't enough for a significant result - though it might be, if the effect size is large enough. How large was the rest of your sample, and how big was the difference? I won't stoop to insults the way you did, but are you really not even curious about these kind of differences?
  - vectorpush 12y ago
    I won't stoop to insults the way you did How noble of you. You're right though. I apologize for disparaging your faculties of logic, it was tongue in cheek but wholly unnecessary. though it might be, if the effect size is large enough. How large was the rest of your sample, and how big was the difference? How can effect size be determined when we don't even know what is being measured? What is the criteria for determining the quality of a programmer? How is that criteria applied? Are there controls for education? exposure to programming as a child? years of experience? quality of experience? quality of tools available to the programmer? quality of the work environment in which we seek to evaluate the programmer's quality? It's reasonable for one to leverage professional experience in judging the skill level of colleagues, but it'd be foolish to think that those observations can be generalized across an arbitrary dimension of the programmer population (such as gender, or race, or political leaning). are you really not even curious about these kind of differences? I guess? I'm always eager to learn more, but I don't really find the question of programmer aptitude as a function of gender very interesting since I can't identify anything unique about programming that would suggest either gender would be qualitatively predisposed to the craft.