invalid name

[Tuan Nguyen Anh]

Dear all the statalists,
Today, when I have tried to analyze the propensity score following the instruction of Mark Lunt's file, I figured out that pbalchk command is used to check covariate balance achieved by a propensity score. However, when I type this in my do-file, the following error was appeared:
. pbalchk t x1 x2 x3
==1 invalid name
r(198);
Can you help me to figure out what error I met and how to fix it.
Thank you in advance.

[Carlo Lazzaro]

Tuan:
welcome to the list.
This is my first time I hear of -pbalck-; however,

==1

without an -if- before the first

=

sounds as suspiciously illegal.

[Tuan Nguyen Anh]

Thank you Mr Lazzaro,
Let me explain more for you. The pbalchk is a command provided by Mark Lunt in his instruction on propensity score analysis to check covariate balance. In his writting, the suggested command for pbalchk is :
pbalchk treatment covariate1... covariate#, wt(norm_weights)
I have just applied it into my do-file with treatment=t, and covariates are x1, x2, x3. However, the achieved result was:
. pbalchk t x1 x2 x3
==1 invalid name
r(198);
I had no idea what the error is. Can you help me with any further answer for this problem? Thank you in advance.

[Carlo Lazzaro]

Tuan:
despite I cannot be more helpful technically, I would check:
- whether your copy of Stata is fully updated;
- whether your version of -pbalchk- is the last available;
- whether the cause of the problem is running the .do file line by line (instead than letting it run from the start to te end automatically).

[Tuan Nguyen Anh]

Thank you Mr Lazzaro,
I think it might be a problem of my current Stata version. I will take your suggestions to solve my problems.
Thank you again with your help.

[William Lisowski]

Let me point out, in case it was overlooked, that the final line in the output of help pbalchk is

Please email [email protected] if you encounter problems with this program

so perhaps he will be able and willing to provide some insight to you directly.

[Tuan Nguyen Anh]

Thankyou for your help Mr Lisowski,
I have emailed to this address and waited for the answer of author.