sampling the sample

[Javier]

Following the recent comments on limiting a search and counting
words, I would appreciate if someone knows a simple way to extract a
random set of utterances (not just words) from a given CHAT file.

For instance: (1) we have an original sample of 1234 child utterances
(2) we get rid of 234 utterances -randomly!-, and (3) we end with a
sample of exactly 1000 utterances (ideally preserving CHAT format for
ulterior analyses).

This would be helpful for some reasons (i.e., equalizing different
samples in size).

It seems an easy process, but I don't know about an easy tool to
solve the problem (included in CLAN or not).

Thanks,
Javier


Javier Aguado Orea         = . . =
School of Psychology           ,
University of Nottingham
UK
--
__