Date:         Fri, 10 Dec 2004 09:51:24 -0500
Reply-To:     Ian Martin <ianmartin@MAC.COM>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Ian Martin <ianmartin@MAC.COM>
Subject:      Re: Mental block
Comments: To: stephenhampshire@leadershipfactor.com
In-Reply-To:  <PCEMJFNDGBAAKFBKKHBIGEHOENAA.stephenhampshire@leadershipfactor.com>
Content-Type: text/plain; charset=US-ASCII; format=flowed

Hi Stephen,

This sounds like classic analysis of covariance (ANCOVA): testing
differences in salary between sexes after adjusting for the influence
of the common covariate of age. The sexes can have significantly
different ages, as long as the slope of salary with age is NSD between
sexes.

I think you'll find ANCOVA well documented in SPSS help or manual, but
feel free to ask if you have questions.

regards,
cpal

On 10 Dec, 2004, at 8:10 AM, Stephen Hampshire wrote:

> But if there is also a correlation between another continuous variable
> (let's say age) and salary, and there is a difference between the
> sexes on
> this variable as well, how can I take that into account?
>
> I suppose what I'm asking for is a test to determine whether there is a
> gender difference independent of the relationship between age and
> salary.