indicators of statistical properties

[John P. Morgan]

On Mon, 4 Aug 2003 10:58:21 +0100
Peter Cameron <p.j.cameron at qmul.ac.uk> wrote


>
>On the other hand, I am not entirely convinced that mathematicians are
>not interested in partially balanced designs. There is an extensive literature
>on partial geometries and their generalisations; these are very specific
>partially balanced designs, but would not be turned up by a search for
>"partial balance" since usually this is not mentioned.

I am not trying to argue that "mathematicians are not interested in partially 
balanced designs." Rather, the topic has content of primary interest in both 
disciplines (in so far as they are separate) and reasonable activity in both, 
thus qualifying partial_balance_properties for the status suggested. This is 
different than for research on association schemes, which really does seem to 
have become the domain of combinatorial mathematicians. 

Underlying the proposal is a conceptualization of combinatorial_properties
as containing counts for incidences and groupings, and of statistical_properties 
as containing information on variances. Because the concept of partial_balance 
speaks to both, the proposal is to not subsume it in either.

I am also concerned that most statisticians will find it awkward/silly to have 
partial_balance_properties under combinatorial_properties.

Leonard and Peter D, would you like to weigh in on this? 

JP