[R] Box-Cox / data transformation question

Rick Bilonick rab at nauticom.net
Sun Jan 30 23:47:31 CET 2005


Landini Massimiliano wrote:

>On Tue, 25 Jan 2005 15:42:45 +0100, you wrote:
>
>|=[:o)  Dear R users,
>|=[:o)  
>|=[:o)  Is it reasonable to transform data (measurements of plant height) to the 
>|=[:o)  power of 1/4? I´ve used boxcox(response~A*B) and lambda was close to 0.25.
>|=[:o)  
>
>IMHO (I'm far to be a statistician) no. I think that Box Cox procedure must be a
>help to people that had none experience in data transforming. In fact data
>transforming include other methods that Box Cox procedure can't perform as rank
>transformation, arcsine square root percent transformation, hyperbolic inverse
>sine, log-log, probit, normit  and logit.
>Transformation is not simply an application of a formula to massive data. Is
>preferable decide appropriate transformation knowing deepening how and from
>where data were collected.
>
>
>|=[:o)  Regards,
>|=[:o)  Christoph
>|=[:o)  
>|=[:o)  ______________________________________________
>|=[:o)  R-help at stat.math.ethz.ch mailing list
>|=[:o)  https://stat.ethz.ch/mailman/listinfo/r-help
>|=[:o)  PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
>
>
>
>-------------------------------------------------------------------------------------------------------------------------
>Landini dr. Massimiliano
>Tel. mob. (+39) 347 140 11 94
>Tel./Fax. (+39) 051 762 196
>e-mail: numero (dot) primo (at) tele2 (dot) it
>-------------------------------------------------------------------------------------------------------------------------
>Legge di Hanggi: Più stupida è la tua ricerca, più verrà letta e approvata.
>Corollario alla Legge di Hanggi: Più importante è la tua ricerca, meno verrà
>capita.
>
>______________________________________________
>R-help at stat.math.ethz.ch mailing list
>https://stat.ethz.ch/mailman/listinfo/r-help
>PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
>
>
>  
>
Why are you using a double square root transformation? Is the 
transformation for the response variable? Transfromation is one way to 
help insure that the error distribution is at least approximately 
normal. So if this is the reason, it certainly could make sense. There 
is no unique scale for making measurements. We choose a scale that helps 
us analyze the data appropriately.

Rick B.




More information about the R-help mailing list