topic Re: Is there a SAS equivalent of qnorm() for non-standard normal distributions? in SAS Programming

Is there a SAS equivalent of qnorm() for non-standard normal distributions?

svh — Mon, 22 Apr 2024 20:19:28 GMT

How does one in SAS 9.4 generate the 20th percentile of a normal distribution with, say, a mean of 500 and standard deviation of 50? In R I call: qnorm(.20, 500, 50)

I have a bunch of percentiles of reading achievement scores, and I need to convert them to to scale scores.

I'm looking throughout the documentation and can't seem to find a way...unless I need to use PROC IML first, but I'm not very good with that.

Re: Is there a SAS equivalent of qnorm() for non-standard normal distributions?

ballardw — Mon, 22 Apr 2024 20:34:54 GMT

Does the value of X look like what you want:

data junk;
   x = quantile('normal',.2,500,50);
run;

The SAS Quantile function takes as the first parameter the name of the distribution, there are about 25. Then the quantile of interest and then distribution parameters. If you leave off the mean and standard deviation the distribution for Normal will assume 0 and 1.

Re: Is there a SAS equivalent of qnorm() for non-standard normal distributions?

svh — Mon, 22 Apr 2024 20:44:24 GMT

Nice! I read about that function, but did not get that I could implement it that way.

Re: Is there a SAS equivalent of qnorm() for non-standard normal distributions?

ballardw — Mon, 22 Apr 2024 20:47:57 GMT

@svh wrote:
Nice! I read about that function, but did not get that I could implement it that way.

Understand. Most of those other distributions really want one or more parameters and the examples using Normal in that function don't show the optional mean and std dev.