Data Transformation

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

Data Transformation
Message posted by Dave (via 131.247.18.227) on February 1, 2002 at 11:55 AM (ET)

I am reviewing a paper in which the author utilizes power transformations often to transform the data so it appears normal. The problem stems from the fact the selection of the power appears arbitrary and fluctuates throughout the paper. It feels to me the author is picking a transformation given the situation to fit the story he is trying to promote. This feels wrong. Could someone point me to rules that should be considered be for applying transformations. And how one should justify the transformation he has employed.

Thanks

Dave

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

Re: Data Transformation
Message posted by Jack Tomsky (via 12.144.103.66) on February 1, 2002 at 5:15 PM (ET)

If the various sets of data are in the same physical units, then I agree with you that he should use a common power transformation.

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