After a frustrating web-search, I have not found any detailed explanations of
possible tests of bimodality. The allstat archive also comes up with only a
couple of quite old queries on this matter...so have there been any advances
since Martin Bland's 2002 query? The data are "continuous" and abundant -
perhaps 100-1000 values per distribution and measured on quite a fine scale.
Many thanks for your help - I will summarise replies for the list.
Regards,
Tony
Tony Dadd
Senior Statistician
Data Sciences
Unilever Colworth
Sharnbrook
Bedford
MK44 1LQ
Tel: +44 (1234) 222420
Fax: +44 (1234) 222761
|