Using binary variable as moderator

[Daphne Sural]

Hello,

I am a student and I need to complete my master thesis. I dont know which command to use in order to have a moderated relationship with using binary variable as moderator.

My dependent is value, my independent is r&d and I have 2 controls which are also dummies: payment and type of acquisition.

Thanks

[Carlo Lazzaro]

Daphe:
welcome to this forum.
With such scant details (eg, what does techically mean that your dependent variable is value?? Is it a continuous variable? Else?), I can only recommend you to take a look at -fvvarlist- and, last but not least, to read the FAQ about how to post effectively. Thanks.

[Daphne Sural]

Hello STATA experts,

I am writing to you with regards to my analysis.

I'm running a simple regression model. I have 5 variables and I am using a moderator variable as a dummy.

My dependent variable is value which is a continuous variable. Also, my independent variable(rd) is continuous. I have 2 controls which are dummies(stock, crossborder) and my moderator(related) is also dummy.

I ran the following:
regress value rd i.crossborder i.stock i.related c.value#i.related

My coefficient is 0 and under std error is written omitted.

I am using stata 14.0

I am using the right regression formula? what should I use?

Thank you very much for all help and comments in advance,
Daphne

[Carlo Lazzaro]

Daphne:
thanks for providing further details.
However, your code cannot work, as you actually interacted (perhasps inadvertenty) your dependent variable (value) with a predictor (-related-).
What have you planned to do in order to analyze your data?

[Daphne Sural]

Dear sir,

Thanks for your answer. I found out about a SPSS function called Process and I was googling around to find a Stata function suitable for my analysis.

Do you have any suggestions to continue my analysis?

Thank you in advance.

Kind regards,
Daphne

[Carlo Lazzaro]

Daphne:
the command you used in Stata (-regress-) is suitable for what you're after.
The problem is that your code does not make sense for the reason explained in my previous reply.
Hence, the advice is to scrutinize your regression code and fix it according with the OLS logic.
As an aside, please call me Carlo, as all on (and many more off) the list do. Thanks.