11
$\begingroup$

Does anyone know of a function or R package that can help me turn z scores into percentile scores? The final goal is to classify or rank a group of respondents into four categories based on the heights of their z scores (20% lowest scores, 30%, 30%, 20% highest scores).

Any ideas? Many thanks in advance!

Improve this question
$\endgroup$
4
  • $\begingroup$ With what assumptions on the original scores? How are the values scaled - are the z-scores computed from sample mean and sample std deviation or from some known/assumed values? $\endgroup$
    – Glen_b
    Commented Apr 14, 2013 at 10:13
  • $\begingroup$ The original scores are raw test scores before i normalized them by using scale(x) $\endgroup$
    – rdatasculptor
    Commented Apr 14, 2013 at 10:17
  • $\begingroup$ Related: Computing percentile rank in R. $\endgroup$
    – chl
    Commented Apr 14, 2013 at 10:37
  • $\begingroup$ chl, thanks! I initially missed that ont, but it turned out to be very helpful. $\endgroup$
    – rdatasculptor
    Commented Apr 22, 2013 at 18:21

1 Answer 1

Reset to default
16
$\begingroup$

pnorm(z) will do it.

> pnorm(1.96)
[1] 0.9750021
> pnorm(0)
[1] 0.5
> pnorm(-1)
[1] 0.1586553

Or if you insist on a percentile, boom. Then try

round(pnorm(1.96)*100,0)
Improve this answer
$\endgroup$
1
  • 1
    $\begingroup$ The usual 95% confidence interval is given by pnorm(1.96) - pnorm(-1.96) with result 0.9500042 $\endgroup$
    – PatrickT
    Commented Nov 14, 2017 at 14:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.