hi
women were asked whether they thought it was best to have one
person responsible for their care (i have approx 1000 replies)
the women came from approx 20 different locations , at each of the
locations the manager was asked the policy if one person was
responsible for overall care
what i would like to do is to compare the womens response againgst
what actally happens
my thought was for each region to create a variable coded either
yes or no, so that for example in region 1 , the manager responded
yes, the new variable would have a yes for all the women in that
region
eg
woman region yes/no
yes 1 yes
no 1 yes
etc
no 2 no
no 2 no
yes 2 no
etc
i then proposed to use kappa but am unsure if this is the way to go
as eacjh of the regions will have different n's and hence a large
region may have a huge influence on the answer
any thoughts on the method of analysis or whether the above is
feasible would be appreciated
there is one added complication which should not affect the above,
there are five groups (a different variable), i would like to look
at the womens response by the managers response within the group
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
|