I have been running a draft of my SA’s past a junior collaborator. She is so damn smart. She reminded me of something that I know in the teaching of statistics context but had forgotten or overlooked in the grant-writing context.
Many common words used in statistics have a different meaning in clinical science, in ecology, etc.
One that has multiple definitions is: population. In statistics it is the hypothetical, unknowable, entity about which you would like to know. The sample are the things you measure to make inferences about the population. On the other hand what you measure in epidemiology or ecology is a population.
The one I had overlooked is “treatment”. In statistics a treatment is something you do to a group in your sample. Factors are the “independent” or “carrier” variables. Factors are divided into levels. Thus, if dealing with males and females, sex is the factor and males is one level, and females another. If you have two factors, sex and species (with three levels, being Triadobatrachus massinoti, Casuarius casuarius, Lemur catta), then the interactions i.e. (female-L.catta, male T. massinoti) are treatments.
I am writing a quasi-clinical grant (using animal models). I have been referring to treatments as the different things I will do to animals within the factor of “insult” crossed with age. But what my jr. col. made me realize is that treatments in a clinical world are usually what you do to make something better. Interventions, insults, and other more specific words make for a better proposal.
It is little things like this that make a difference in readability of a grant. If you get it right no one notices and its cool. Get it wrong and the reviewers get irritated by either not understanding or figuring it out late in the game.