What is it called when variable response values are unequally distributed?

[fieldsa83]

I have a variable for hourly earnings which respondents can answer with accuracy to the penny. So the mean is like $27, but toward the bottom there is so much small variation in reporting, so it has one person at $11.61 one person at $11.63 two people at $11.67, etc.. but when you get higher there's just clumps of people at $15.00 and $25.00.

So if I graph it, the chart has many more values at the low end and it skews the chart. Is there a way to automatically recode them to all be in clumps increasing by $0.50 or something (without manually recoding hundreds of ranges)?

[Reeza]

ROUND(variable, 0.50)

I thought that's what a histogram did? It bins it automatically? 

[art297 · Accepted solution]

You would still have a skewed distribution, but you could use the round function to adjust the salaries. e.g.:

data have;
  input salary;
  datalines;
11.53
11.48
12.10
12.49
12.74
12.75
;

data want;
  set have;
  round_salary=round(salary,0.5);
run;

Art, CEO, AnalystFinder.com