How do I hand calculate MANOVA?

Home > Statistics Every Writer Should Know > The Stats Board > Discusssion

How do I hand calculate MANOVA?
Message posted by Val (via 130.194.13.211) on June 11, 2001 at 3:31 AM (ET)

Unfortunately, I need to know how to do it by hand as well as by computer! Thanks. Val

READERS RESPOND:
(In chronological order. Most recent at the bottom.)

Re: How do I hand calculate MANOVA?
Message posted by JG (via 128.8.23.172) on June 11, 2001 at 10:44 AM (ET)

Before you worry about MANOVA, how much do you know about ANOVA and the special problems of trying to do the required computations without a computer ? Before computers, computational shortcuts were a major portion of ANOVA knowledge - especially using orthogonality as much as possible, etc.

Re: How do I hand calculate MANOVA?
Message posted by Val (via 130.194.13.209) on June 15, 2001 at 12:24 AM (ET)

Hi JG,

Thanks for responding. I am reasonably familiar with ANOVA and other univariate stats. Unfortunately, just when I thought I was doing OK with maths, my uni hit me with multivariate stats :( So, I have an exam in a weeks time and MANOVA will form part of that. Orthogonality is something I don't really understand, so I am feeling somewhat desperate! Val

Re: How do I hand calculate MANOVA?
Message posted by JG (via 128.8.22.1) on June 15, 2001 at 2:20 AM (ET)

Re: How do I hand calculate MANOVA?
Message posted by JG (via 128.8.22.1) on June 15, 2001 at 2:21 AM (ET)

Orthogonality is central to ANOVA computations. It allows one to add SS in a manner that is similar to the fact that variances can be added when we have independence. Without orthogonality we have just another large regression problem. That is, ANOVA is regression with special structure. With MANOVA we also have the problem of combining the dependent variables, etc.

Your $5 contribution helps cover part the $500 annual cost of keeping this site online.