How to get replies on my post?

[Lucy Ty]

Hi, I am relatively new to posting on this forum but have used it extensively to add my econometric knowledge.

I am currently not getting any replies on a post from earlier today. As time goes on and more recent posts are displayed, it is less and less likely I will receive a response. Do you have any advice how I can increase the number of responses (from zero!) - is it the subject heading, a lack of clarity in the question or the subject content?

I have copied my previous question below. The original question can be found here : https://www.statalist.org/forums/for...-iv-regression

Thank you so much in advance for any advice you can give !!


Previous Question: Testing if coefficients differ across subsamples in IV regression

Good Morning,

I am trying to determine whether my endogenous variable’s coefficient differs across two subsamples of my population. I have used previous forum answers to try and solve the problem myself.

1^st attempt to solve I received an error after the first line of code:

ivreg2 Y a c d (e = z1 z2 z3) if t==0, pscore(est1)
ERROR “option pscore() not allowed”
est store est1
est change est1, scorevars(est1)
ivreg2 Y a c d (e = z1 z2 z3) if t==1, pscore(est2)
est store est2
est change est2, scorevars(est2)
suest est1 est2

2^nd attempt to solve (change ivreg2 to ivreg) I received an error after the final line of code”

ivreg Y a c d (e = z1 z2 z3) if t==0, pscore(est1)
est store est1
est change est1, scorevars(est1)
ivreg Y a c d (e = z1 z2 z3) if t==1, pscore(est2)
est store est2
est change est2, scorevars(est2)
suest est1 est2
ERROR “ivreg is not supported by suest”

3^rd attempt to solve:

ivreg2 Y a c d (e = z1 z2 z3) if t==0
est store est1
ivreg2 Y a c d (e = z1 z2 z3) if t==1
est store est2
gmm est1: Y-{b1}a -{b2}c -{b3}d -{b4}e-{b0} ///
ERROR “must specify parameters() or nparameters()”
gmm est2: Y -{c1}a -{c2}c -{c3}d -{c4}e-{c0} ///
instruments(est1: a c d z1 z2 z3) ///
instruments(est2: a c d z1 z2 z3) ///
onestep winitial(unadjusted, indep)
test ([b4]_cons = [b4]_cons)

Please may you help me to understand why my previous codes have not worked and, if possible, suggest a way forward. Thank you so much.

[Richard Williams]

Hi Lucy. To phrase Qs more effectively, see p, 12 of the FAQ. See especially the part about code tags.

Unfortunately I don’t know the answer to your Qs. It appears you have syntax errors, so read the help files.

[William Lisowski]

I agree with Richard's advice to review the Statalist FAQ linked to from the top of the page, as well as from the Advice on Posting link on the page you used to create your post. Note especially sections 9-12 on how to best pose your question.

For your first problem, Stata tells you what your problem is. Reviewing the output of help ivreg2 shows that pscore() is not among the options included in the syntax.

For your second problem, Stata tells you that suest following ivreg estimation is not supported, and help ivreg tells us that command has been obsolete since Stata 10. Although ivreg supports the pscore() option, I have no way of knowing what that option does in order to see if ivreg2 can be coerced to do the same thing to solve your first problem.

For your third problem, help gmm shows that your syntax is clearly incorrect. You need something more like the following, although as a non-user of gmm, I cannot warrant that it will produce the output you need, or indeed any meaningful output at all.

gmm est1: Y -{b1}a -{b2}c -{b3}d -{b4}e -{b0}  ///
    est2: Y -{c1}a -{c2}c -{c3}d -{c4}e -{c0}, ///
instruments(est1: a c d z1 z2 z3) ///
instruments(est2: a c d z1 z2 z3) ///
onestep winitial(unadjusted, indep)

The reluctance to address your question likely stemmed from the fact that it was lengthy, seemed to require particular knowledge that not all users (certainly not myself) have, and it was clear you had not reiviewed the Statalist FAQ as you were asked to do on the page from which you created your post.

The more you help others understand your problem, the more likely others are to be able to help you solve your problem.