What exactly do you want the program to calculate? Just standard deviation of
the mean or errors in linear fit parameters?
If you don't mind it bundled with a graphing program and you are using PC
compatibles, you could try our freeware package called physfit. It is available
by ftp from the directory ftp.kenyon.edu/pub/physics/SOFTWARE/WINDOWS/PHYSFIT.
Read the readme file. There is a 16 bit version and a 32 bit version. Let us
know if you have problems with it.