I have been searching the interwebz about this but have found no straightforward answer -- I realise it is a basic question to nag you with on here =)
So the question: what exactly does the F-value in SPSS's Levene's test for equality of variances represent?
As far as I understand it, it is the result of Levene's test.
Eg. if I am comparing the means of two independent samples using a t-test I have to possibilities: To establish whether the variances are equal at a 95% confidence, I can either use Levene's test or the simple F-test (F=(s1)^2/(s2)^2) and compare both of them with a critical value of F(alpha,m1,m2). Please correct me if I am wrong or the F-value in SPSS's report is something different altogether.
And one more thing: does SPSS ever use Bartlett's test? I understand this is only used for testing the equality of multiple variances?
In SPSS's ANOVA report for multiple groups, the F-value displayed is calculated as (MS between groups/MS within groups). Is this Bartlett? If not, what is it called?
Thanks sincerely in advance!
Last edited by mrhajdi; 04-13-2009 at 10:04 AM.
Advertise on Talk Stats