I like CIs, but if you report the SE (a) it takes less space, and (b)
people can work them out. But CIs are more informative (who cares
what the standard deviation of the sampling distribution of the
statistic is?)
J
2010/1/8 Pat Burton <[log in to unmask]>:
> I would like to ask people’s opinion on reporting confidence intervals in
> multiple regression. This is not normally something I would do in regression
> reporting but it has recently been suggested to me as important to do. I
> would normally report the SE.
>
>
>
> Pat Burton PhD
> Study Co-ordinator / Clinical Research Associate
> Childhood Nutrition Research Centre
> Institute of Child Health
> 30 Guilford Street
> London WC1N 1EH
> Tel: 0207 905 2397 / 07758864912
> Fax: 0207 831 9903
>
>
--
Jeremy Miles
Psychology Research Methods Wiki: www.researchmethodsinpsychology.com
|