CENSUS, NOT A SAMPLE

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

CENSUS, NOT A SAMPLE
Message posted by Jim on January 27, 2000 at 12:00 AM (ET)

I have recently conducted an attitudinal survey of employees. Of the 350 employees, 241 took part - a response rate of 69%. Can I assume that any differences I detect in the results, say between males and females, are real - irrespective of magnitude? Or should I apply the usual significance tests even though the sample represents such a large proportion of the population?

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

Re: CENSUS, NOT A SAMPLE
Message posted by bon bon on January 30, 2000 at 12:00 AM (ET)

It appears that you should apply the usual procedure to test for significant difference (t-test, ANOVA, etc) as the percentage of the population covered by your sample and tests of significance are two different things.

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