Which tests? Parametric or Nonparametric

Home > Statistics Every Writer Should Know > The Stats Board > Discusssion

Which tests? Parametric or Nonparametric
Message posted by Rob (via 216.208.96.103) on August 9, 2001 at 11:39 AM (ET)

I'm analyzing a fake employee data set. I am using regression techniques to predict the impact of certain variables like education level and previous experience on salary, and I'm also comparing male employees to female employees. I've used Pearson's R to correlate certain variables, and Eta to test significance. Question: would this sample of employee data merit parametric tests because it IS the population, or nonparametric tests?

READERS RESPOND:
(In chronological order. Most recent at the bottom.)

Re: Which tests? Parametric or Nonparametric
Message posted by JG (via 128.8.22.170) on August 12, 2001 at 7:08 PM (ET)

We generally use parametric mothods when we are willing to assume an error distribution and non-parametric methods otherwise. In regression, we generally assume that the errors are identical, normaly distributed and independent of each other and the rest of the model.

Your $5 contribution helps cover part the $500 annual cost of keeping this site online.