Hi,
Got the relevant link from Stat group.
http://www.gseis.ucla.edu/courses/ed230a2/notes3/portion.htmlThanks a lot,
Hari
India
"Hari Prasadh" <excel_hariNOSPAM...hotSPAREMEmail.com> wrote in message
news:uXIozHIgFHA.1148...TK2MSFTNGP12.phx.gbl...
> Hi,
> Not exactly an excel question but still posting here.
> I want to do z-test using Excel. I do have the analysis tool pak installed
> (along with Analysis tool-pak VBA).
> Being used to Canned software (The concerned tabulation software terms it
> as Z-Test for percentages) which tell me automatically whether there is
> any significant difference between percentages of 2 groups (at 95% CI), am
> at a loss to do the same in Excel (ultimately using VBA).
> I see that excel data analysis option (within Tools - Data Analysis)
> offers "Z test: Two sample for means".
> When I click on it I get a dialog box in which I fill some of them
> a) Variable 1 range: $A$1:$A$7
> b) Variable 2 range: $B$1:$B$7
> c) Hypothesized mean difference:
> d) Variable 1 variance (known):
> e) Variable 2 variance (known):
> f) Labels - this is a check-box which was already checked.
> g) Alpha: 0.05 - This was already filled
> h) Output Range $K$15
> Could anybody, please tell me what to fill in c) THROUGH e).
> (Please note, though i have mentioned equal variable ranges for variable 1
> and 2 above it sometimes happens that one of variable 1 or variable 2 has
> missing values for some data points. In stat we call this as system
> missing values).
> In the tabulation software that I work on, I just have to specify which 2
> groups (columns/variable ranges in Excel) to be compared along with the
> confidence level at which testing is to be done. I have never had to
> mention Hypothesized mean difference, Variable 1 variance etc. Why is
> excel asking for these?
> Am I looking at a wrong test. Is there any other Z-test which will do what
> I want? (Being deficient in Stat, i dunno even know what is it that I
> want!!)
> Though I have separately posted to a Stat group, I would be glad if
> somebody could help me with this.
> Regards,
> Hari
> India
>