4 ms·

What are the distributions of the salaries among one language? Mean values alone are completely useless, give us some information about how you found the salari

by quchen 13y ago

What are the distributions of the salaries among one language? Mean values alone are completely useless, give us some information about how you found the salaries out, what their standard deviations are etc. I don't want to imply maliciousness here, so I'm going to call this an example of unknowingly bad statistics (which, sadly, seems to be the default in most places). Bonus points for the bar (!) chart that starts at 90,000 and looks like ActionScript is twice as profitable as the rest of the languages.

dlss 13y ago
He found the salaries out using rapleaf... which didn't know the salaries but was making an educated guess. (You can try out their api if you register an account, FWIW it was way off for the people I've tried it on.) With data like that, I'd imagine the only sort of data you could trust would be the mean, since I'm guessing its much harder to be wrong about relative income levels between large groups than it is to be right about actual income numbers.
- quchen 13y ago
  The only sort of data you can trust is the data itself. Providing the mean of data is providing a sum value a person thinks is useful to convey the shape of the data, and the mean in particular is unsuitable at conveying shape on its own. What the author tries to get at is a precise quantity, namely the average salary for each language. However, we cannot measure that quantity directly, so taking samples (from a good source) is as good as it gets. The easiest assumption about that data is that it follows a normal distribution, which does have a mean, but also a standard deviation; both of these are required to talk about any meaning of the result. (If the normal distribution is not a good fit things may get more complicated.) What I'm saying is that providing just means is not a ballpark way of doing statistics, it is wrong. The statistical ballpark is the normal distribution, and maybe a value ensuring that it is at least somewhat appropriate to use that one.
  - lifeisstillgood 13y ago
    So forgive the statistical ignorance, but how would one display mean and std dev in useful graphical form? (Asking seriously as it would be good to know which setting to press on gnu plot)
    - route66 13y ago
      With boxplots? http://benchmarksgame.alioth.debian.org/u64q/which-programs-are-fastest.php http://benchmarksgame.alioth.debian.org/u64q/which-programs-... (they can show a bit more though, min, max, interquartile range or mean+/- 1 stddev and mean itself) see also http://en.wikipedia.org/wiki/Box_plot http://en.wikipedia.org/wiki/Box_plot .
    - merijnv 13y ago
      Not quite the same, but box-and-whisker plots (https://en.wikipedia.org/wiki/Box_plot https://en.wikipedia.org/wiki/Box_plot) are good for visualising distributions. They show the mean, minimum, maximum and the first and second quartiles. Since they show the minimum and maximum we can judge the overall spread, the mean tells us...well...the mean and first and second quartile lets us judge how close data is clustered around the mean. See http://gnuplot.sourceforge.net/demo/candlesticks.html http://gnuplot.sourceforge.net/demo/candlesticks.html for gnuplot examples.
      - lifeisstillgood 13y ago
        But box plots seem to be needing turning on their side to visualise (assuming normal distribution) I suppose I am looking for mathematically faithful whilst still Omnigraffle good looks Cheers
      - omh 13y ago
        A minor point, but I think Box Plots usually show the median rather than the mean at the centre. It would make sense perhaps to use standard deviations and means for some data, but in cases like this I think quartiles and medians make more sense.
    - deleted 13y ago
- contingencies 13y ago
  Without even looking at potential issues with the source data, it is also the case that most programmers of any calibre work across multiple languages. Also, languages which are 'older' and reaching obscurity (ie. fallen out of favour) are going to be known by older programmers who, if they are still in their career, probably command higher salaries only as a factor of age and experience. Finally, salary before tax is not very representative of income or social circumstance, and people in our industry often opt for time flexibility or remote work in preference to higher salaries on purpose (for example, after having children). All in all, not useful.
runeks 13y ago
> Bonus points for the bar (!) chart that starts at 90,000 and looks like ActionScript is twice as profitable as the rest of the languages. When I make charts like this, I always let the X-axis be zero. I find it confusing to look at a chart and not immediately know the percentage difference without using a calculator. With the X-axis at 0 you can immediately tell the proportional differences. It's like looking at the intra-day variations of a stock's price, and thinking "Whoa! What happened here?!". Then you zoom out and it's a completely straight line.
bpodgursky 13y ago
The bar chart is straight from Google docs, which just auto-scaled however it saw fit. I tried to present the numbers first to be clear about the absolute range.

by quchen 13y ago

dlss 13y ago

He found the salaries out using rapleaf... which didn't know the salaries but was making an educated guess. (You can try out their api if you register an account, FWIW it was way off for the people I've tried it on.) With data like that, I'd imagine the only sort of data you could trust would be the mean, since I'm guessing its much harder to be wrong about relative income levels between large groups than it is to be right about actual income numbers.

quchen 13y ago
The only sort of data you can trust is the data itself. Providing the mean of data is providing a sum value a person thinks is useful to convey the shape of the data, and the mean in particular is unsuitable at conveying shape on its own. What the author tries to get at is a precise quantity, namely the average salary for each language. However, we cannot measure that quantity directly, so taking samples (from a good source) is as good as it gets. The easiest assumption about that data is that it follows a normal distribution, which does have a mean, but also a standard deviation; both of these are required to talk about any meaning of the result. (If the normal distribution is not a good fit things may get more complicated.) What I'm saying is that providing just means is not a ballpark way of doing statistics, it is wrong. The statistical ballpark is the normal distribution, and maybe a value ensuring that it is at least somewhat appropriate to use that one.
- lifeisstillgood 13y ago
  So forgive the statistical ignorance, but how would one display mean and std dev in useful graphical form? (Asking seriously as it would be good to know which setting to press on gnu plot)
  - route66 13y ago
    With boxplots? http://benchmarksgame.alioth.debian.org/u64q/which-programs-are-fastest.php http://benchmarksgame.alioth.debian.org/u64q/which-programs-... (they can show a bit more though, min, max, interquartile range or mean+/- 1 stddev and mean itself) see also http://en.wikipedia.org/wiki/Box_plot http://en.wikipedia.org/wiki/Box_plot .
  - merijnv 13y ago
    Not quite the same, but box-and-whisker plots (https://en.wikipedia.org/wiki/Box_plot https://en.wikipedia.org/wiki/Box_plot) are good for visualising distributions. They show the mean, minimum, maximum and the first and second quartiles. Since they show the minimum and maximum we can judge the overall spread, the mean tells us...well...the mean and first and second quartile lets us judge how close data is clustered around the mean. See http://gnuplot.sourceforge.net/demo/candlesticks.html http://gnuplot.sourceforge.net/demo/candlesticks.html for gnuplot examples.
    - lifeisstillgood 13y ago
      But box plots seem to be needing turning on their side to visualise (assuming normal distribution) I suppose I am looking for mathematically faithful whilst still Omnigraffle good looks Cheers
    - omh 13y ago
      A minor point, but I think Box Plots usually show the median rather than the mean at the centre. It would make sense perhaps to use standard deviations and means for some data, but in cases like this I think quartiles and medians make more sense.
  - deleted 13y ago
contingencies 13y ago
Without even looking at potential issues with the source data, it is also the case that most programmers of any calibre work across multiple languages. Also, languages which are 'older' and reaching obscurity (ie. fallen out of favour) are going to be known by older programmers who, if they are still in their career, probably command higher salaries only as a factor of age and experience. Finally, salary before tax is not very representative of income or social circumstance, and people in our industry often opt for time flexibility or remote work in preference to higher salaries on purpose (for example, after having children). All in all, not useful.

runeks 13y ago

> Bonus points for the bar (!) chart that starts at 90,000 and looks like ActionScript is twice as profitable as the rest of the languages. When I make charts like this, I always let the X-axis be zero. I find it confusing to look at a chart and not immediately know the percentage difference without using a calculator. With the X-axis at 0 you can immediately tell the proportional differences. It's like looking at the intra-day variations of a stock's price, and thinking "Whoa! What happened here?!". Then you zoom out and it's a completely straight line.

bpodgursky 13y ago

The bar chart is straight from Google docs, which just auto-scaled however it saw fit. I tried to present the numbers first to be clear about the absolute range.