Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

>These charts show how statistics can be used to mislead.

Someone says this every time a chart with a scale other than zero shows up somewhere. It's untrue, generally. The context of the information is what is important.

In this case, what is being emphasized is the difference between salaries of positions. What would the point be of having a mass of whitespace at the bottom of the chart? What relevance is the salary level in relation to zero? Scaling to zero would diminish the information being conveyed.



No, you're saying "emphasized".

It's more exaggerated.

When our lizard brains look at charts, we're conditioned to check how relatively apart the lines are. By removing the zero scale, we're diminishing the lack of the absolute proportions of the salary scales. 8% doesn't seem like much, but 10k seems like a whole lot.

Also, this is a sensationalist article. They did not get into their methodologies, or whether their data points were actually representative of the industry. Somebody getting paid by the column inch found an easy article to write.


Edward Tufte himself says to use a non-zero axis for time series.

If the extent of your analysis is "10k is a lot, 8% isn't" then I'm not really concerned how you interpret my chart.

I have no comment on the quality of the article.




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: