Re: Sample vs Population Question


Tim McCr8 wrote:
> > Is there any strong reason to think that a mean age difference of 7.5
> > years is likely to be of substantive significance given the age
> > distribution of the entire sample?
>
>
> I hope I understand what you're saying. If I look at all respondents my
> mean age is 43.1 and the SD is 13.7 years. That means all my sub-group
> means (salary providers and non-providers) are well within 1 SD of the
> overall sample mean. So, if that's what you're questions asks, I
> suppose I don't have any reason to believe the difference between
> thesub-groups makes a substantive difference vis a vis the entire
> sample.
>
> Part of me says the law of large numbers is in my favor. And the data
> do not suggest some abnormal population here. (My population is a
> profession that notably start at moderately high salaries that plateau
> after 12 years or so. What my data show is a sample that plateaus in
> the 10 to 15 years working range with work experience and degree held
> being the biggest factors in salary.) But I'm finishing grad school and
> the advisor on my thesis has taken the position that this is a fluke
> and that the methodology is questionable.
>
> I like to understand what I'm doing, even if I ultimately wind up
> hiring experts to do the very hard stuff.

Well I'm not a statistician so don't take me too seriously but I was
wondering something the same as your advisor, I think.

However the real point I was trying to make ( I think anyway) was that
if you don't have any theoretical reason to assume that age is going to
affect reporting of salary (and it is not glaringly obvious ( i.e. the
interocular shock test) I'd be wary of interpreting a significant
statistical difference as a substanitive one. Even if you have a a
real difference in the two groups does an unwillingness to reveal
salary level mean as age increased mean anything? Or could it be that
older people for some cultural reason may be more reluctant to report
salary but it is tangential to your enquiry?

I am sure that there are more qualified people here than me to discuss
it though.

Good luck

John Kane, Kingston ON Canada

.

Relevant Pages

  • Re: Bus driver refuses to drive "no god" bus
    ... years of age. ... because until the 'dawn of educational reason' ... invisible sky wizards weren't discounted in the same breath as santa ...
    (uk.philosophy.atheism)
  • Re: The "Id rather have her get cancer than have sex" dept.
    ... Almost all children are sexually active by age ten years. ... "Journal of Sex Research," August, 2005 issue. ... of high school students are "sexually active" in the sense of having ... usually cut off their age cohorts at 13 for this reason. ...
    (talk.origins)
  • Re: C++ vs. C#
    ... MFC is still a power framework that has sort of stood the test of time, but it's starting to show its age more and more (even though there have been some nice additions lately. ... The best reason to use C++... ... these newsgroups. ... "Ajay" wrote in message ...
    (microsoft.public.vc.mfc)
  • Re: Sheff Live 93.2
    ... Since we were discussing real world experience, age is relevant. ... Jamie, stop and think for a minute. ... for the same reason. ... life progresses a bit further and the bad things recede into to the past you ...
    (uk.tech.broadcast)