Using proc ttest with Farrington Manning

[SR79]

Hello

I am comparing two proportions for non-inferiority and I know I can use proc freq to do farrington manning, but I would like to be able to use proc ttest so that I would have: 

proc ttest data=xx h0=noninf_margin alpha=xx;

class group;

var xx;

run; 

Is there an option to do method=farrington manning in this code?

Thanks!

[DWilson]

---

@SR79 wrote:

Hello

I am comparing two proportions for non-inferiority and I know I can use proc freq to do farrington manning, but I would like to be able to use proc ttest so that I would have: 

proc ttest data=xx h0=noninf_margin alpha=xx;

class group;

var xx;

run; 

Is there an option to do method=farrington manning in this code?

Thanks!

---

This seems to have examples of what you want to accomplish:

https://support.sas.com/resources/papers/proceedings15/SAS1911-2015.pdf

see table 4

[SR79]

Thanks. The prpc ttest can give me noninf but it's not giving me Farrington
Manning CIs. Do you know if there's a function call I can do to incorporate
the FM?

[DWilson]

---

@SR79 wrote:
Thanks. The prpc ttest can give me noninf but it's not giving me Farrington
Manning CIs. Do you know if there's a function call I can do to incorporate
the FM?

---

Just use the CIs output by proc ttest.

Rows 1 and 2 in table 4 discuss how to use the standard CIs to determine if the inferiority hypothesis should be rejected.