# Calculation of Levene's Robust Test of Equality of Variances (not in SPSS!)

#### reillyd

##### New Member
Hi. I'm trying to understand the calculation of a Levene's test of
homogeneity of variance
, for a section of my thesis on sex differences
in variability of cognitive performance (greater male variability hypothesis). As I don't
have the original data I'm analysing, I can't use SPSS to calculate Levene's for me, and
need to calculate it manually (ideally in Excel!).

I've come across this resource on the calculation of Levene's (http://www.itl.nist.gov/div898/handbook/eda/section3/eda35a.htm) which gives the original formula as thus:

where Zij can have one of the following three definitions:
1.

where Yi is the mean of the ith subgroup.
and choices 2. 3. are the median of the ith group and 10% trimmed mean.

What I can't understand (and may be on completely the wrong track about) is where the actual standard deviation of the groups comes in ..... because male and female scores often have the same or negligible mean differences, but different standard deviations (and hence variance).

Can anyone point me to a worked example of a Levene's test, or explain it to me
simply? All I want to do is see if the SD1 != SD2 (hypothesis supported) or if SD1 = SD2 (hypothesis not supported).

Many thanks and eternal grattiude for your help!