add log-binomial edits

remlapmot · Sep 27, 2021 · b0a50f5 · b0a50f5
1 parent 9afe45f
commit b0a50f5
Show file tree

Hide file tree

Showing 3 changed files with 3 additions and 3 deletions.
diff --git a/README.md b/README.md
@@ -13,7 +13,7 @@ Stata package of useful programs for one-sample Mendelian randomization (MR) / i
 The `ivtsps` and `ivtsri` commands implement the following link functions for the second stage model: 
 
 * identity (i.e. linear regression) - for a binary outcome this estimates a causal risk difference 
-* logadd (log additive, i.e. Poisson regression) and logmult (log multiplicative, i.e. gamma regression) - for a binary outcome these estimate a causal risk ratio
+* logadd (log additive, i.e. Poisson/log-binomial regression) and logmult (log multiplicative, i.e. gamma regression) - for a binary outcome these estimate a causal risk ratio
 * logit (i.e. logistic regression) - for a binary outcome this estimates a causal odds ratio
 
 ## Installation

diff --git a/ivtsps.sthlp b/ivtsps.sthlp
@@ -58,7 +58,7 @@ moment condition to the {help gmm} command.
 {opt link(identity | logadd | logmult | logit)} specifies the link function for the second stage model. 
 {cmd:identity} means the second stage model is a linear regression 
 (which for a binary outcome estimates a causal risk difference). 
-{cmd:logadd} means the second stage model is a Poisson regression 
+{cmd:logadd} means the second stage model is a Poisson/log-binomial regression 
 (which for a binary outcome estimates a causal risk ratio).
 {cmd:logmult} means the second stage model is a gamma regression 
 (which for a binary outcome estimates a causal risk ratio, 

diff --git a/ivtsri.sthlp b/ivtsri.sthlp
@@ -58,7 +58,7 @@ moment condition to the {help gmm} command.
 {opt link(identity | logadd | logmult | logit)} specifies the link function for the second stage model. 
 {cmd:identity} means the second stage model is a linear regression 
 (which for a binary outcome estimates a causal risk difference). 
-{cmd:logadd} means the second stage model is a Poisson regression 
+{cmd:logadd} means the second stage model is a Poisson/log-binomial regression 
 (which for a binary outcome estimates a causal risk ratio).
 {cmd:logmult} means the second stage model is a gamma regression 
 (which for a binary outcome estimates a causal risk ratio,