Diana,
This is not a straightforward question! On the whole (but not
entirely), OddsRatio may be a better quantity to parametrise
the 2x2 table rather than the RR (relative risk), and indeed
for a "Fisher Exact Test" it is the exact parametrisation of
the test. But one must also bear in mind what is the feature
of the two riaks which is of most practical interest. In many
cases this may be the RR though in some cases the difference
of risks (p1 - p2) may be of most interest. However, the 2x2
table is in fact parametrised by two paramters (p1 and p2)
and not (in general) one parameter.
Just one remark you make which needs comment: Given the OR,
the relative risk (RR) is not uniquely determined; nor vice
versa; nor does either OR or RR uniquely determine difference.
(More detail on this point if you're interested).
Best wishes,
Ted.
On 11-Aug-2016 18:39:50 Kornbrot, Diana wrote:
> Dear Colleagues
> What is preferred effect size for 2X2 contingency Table?
>
> Seems to me that ln(OddsRatio) approaches normal asymptote faster than
> ln(HazardRatio) and hence is preferable.
> This would apply whatever the final metric is desired as most interpretable
> by investigators, as ln(OR) may easily be converted to OR, relative risk,
> number needed to treat etc.
>
> What is best practice in view of allstat?
>
> many thanks for help
>
> diana
> ________________________________________
> Professor Diana Kornbrot
> Work
> University of Hertfordshire
> College Lane, Hatfield, Hertfordshire AL10 9AB, UK
> +44 (0) 170 728 4626
> [log in to unmask]<mailto:[log in to unmask]>
> http://dianakornbrot.wordpress.com/
> http://go.herts.ac.uk/Diana_Kornbrot
> skype: kornbrotme
> Home
> 19 Elmhurst Avenue
> London N2 0LT, UK
> +44 (0) 208 444 2081
>
>
>
>
> You may leave the list at any time by sending the command
>
> SIGNOFF allstat
>
> to [log in to unmask], leaving the subject line blank.
-------------------------------------------------
E-Mail: (Ted Harding) <[log in to unmask]>
Date: 11-Aug-2016 Time: 20:10:36
This message was sent by XFMail
-------------------------------------------------
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|