hi
i have a data set from a study which had two groups (equally split)
i intend to split by group the data set in two , and use both
regression and logistic regression on the first data set- i am
looking at pain which is continuous but can also be dichotomised
i would then like to apply the results from the regression carried
out on the 1st data set to the 2nd data set - to see if selected
variables are useful in predicting the level of pain and also the
proportion of women classified as showing high pain levels
my question relates to the next stage - how can i show that the
results when applied to the 2nd half of the data set are valid, and
that the variables included in the model are useful in predicting
pain. The results would be used to identify patients who in future
are liable to show high levels of pain with respect to a treatment.
any comments greatly appreciated
regards
ann
----------------------
ann e fitzmaurice
medical statistician
dugald baird centre
department of obs and gynae
university of aberdeen
[log in to unmask]
tel 01224 553876 (direct)
fax 01224 404925
|