Re: How to deal with biologistis?

Jay Weedon wrote:
Hi folks,

I've recently started to work much more than I used to with
experimental biologists - I'm much more used to working with
physicians, epidemiologists and psychologists. One of the things I
notice most about how biologists summarize data is their penchant for
"normalization".

Example client:

"I've got 4 conditions in an experiment including a control. I've
normalized all the data [presumably by multiplication of every score
by the same constant, though it's often not specified] so that the
control group has mean 100. Then I replicated the study on a second
day, and normalized all those data in the same way. I want to pool the
data across days to discover differences among condition means."

Training & experience tell me that all this normalization is probably
unnecessary from the p.o.v. of inferential analysis - we'd do better
with the raw data; if it turns out that means and/or variances vary
substantially across replications, then we might introduce
replications as a random factor, or transform data.

Yeah, at times you need a decent brick wall to bang your head against.

I think you're right: I try and train my biologist colleagues not to pre-treat the data ("oh, we thought you would prefer these counts of surviving and dead individuals as percentages..."): it's my job to make it horrible complex!

I think the core the problem is that biologists approach statistical methods as cookbooks: they don't understand what they're doing, they just turn the handles. You need to be patient, and train them to see the data in the same way you do. Obviously a few blinding successes will help.

Chatfield has a book called "Problem Solving" which is some help, and I think there are other books on statistical consultancy.

It takes time to build up a relationship, but it does bring dividends in the longer run.

Bob

--
Bob O'Hara
Department of Mathematics and Statistics
P.O. Box 68 (Gustaf Hällströmin katu 2b)
FIN-00014 University of Helsinki
Finland

Telephone: +358-9-191 51479
Mobile: +358 50 599 0540
Fax: +358-9-191 51400
WWW: http://www.RNI.Helsinki.FI/~boh/
Journal of Negative Results - EEB: www.jnr-eeb.org
.