Dear all I was looking at the aov documentation page and came across the following which seems like a contradiction to me: " This provides a wrapper to |lm| for fitting linear models to balanced or unbalanced experimental designs." (I presume 'This' refers to aov) and "|aov| is designed for balanced designs, and the results can be hard to interpret without balance" So, do I use it for unbalanced designs? Thanks, AJ -- Dr Albertus J. Smit (Marine Ecologist/Algal Biologist) School of Biology and Conservation Science Faculty of Science University of KwaZulu-Natal Westville Campus P/B X54001 Durban 4000 South Africa Tel. (031) 260 7472 Fax (031) 260 7364 -------------------------------------------------------------------- Please find our disclaimer at http://www.ukzn.ac.za/disclaimer -------------------------------------------------------------------- <<<<gwavasig>>>>
On Mon, 25 Oct 2004, AJ Smit wrote:> I was looking at the aov documentation page and came across the > following which seems like a contradiction to me: > > " This provides a wrapper to |lm| for fitting linear models to balanced > or unbalanced experimental designs." (I presume 'This' refers to aov) > > and > > "|aov| is designed for balanced designs, and the results can be hard to > interpret without balance"and that is part of a Note, not even a whole sentence of it. In particular, the note has extra force for the multistratum designs mentioned there.> So, do I use it for unbalanced designs?You _can_ use it, but the results _can_ be hard to interpret. That is not a `contradiction' -- perhaps you need to consult your dictionary? You may do better to use something else, like lme(), hence the `Note'. -- Brian D. Ripley, ripley at stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UK Fax: +44 1865 272595
AJ Smit <smitaj at ukzn.ac.za> writes:> Dear all > > I was looking at the aov documentation page and came across the > following which seems like a contradiction to me: > > " This provides a wrapper to |lm| for fitting linear models to > balanced or unbalanced experimental designs." (I presume 'This' refers > to aov) > > and > > "|aov| is designed for balanced designs, and the results can be hard > to interpret without balance" > > So, do I use it for unbalanced designs?If you know how to interpret the results... In general, I would avoid it if there is unbalance in the random effects, but some designs with unbalance in the fixed effects are manageable. -- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk) FAX: (+45) 35327907