interpreting coefficient of Poisson regression, when dependent variable is a count variable

Asked Jun 23 '23 at 14:37

Active Jun 24 '23 at 13:49

Viewed 47 times

I have a question about the economic interpretation of Poisson regression.

What is the economic interpretation of the coefficient when the dependent variable is a count variable?

For example, for "Patent Count = a + b × R&D investment + c", what does 1-unit increase in R&D investment mean for Patnet Count?

I know that when the dependent variable is log(Patent Count), a 1-unit increase in R&D investment, then the predicted Patent Count is multiplied by e^b, holding all other variables constant.

However, after reading all similar questions, I am still confused about what is an economic interpretation of the coefficient if the dependent variable is a count variable.

for Stata code poisson Patent_count R&D_investment , what is the economic interpretation of b?

Many thanks in advance.

edited Jun 24 '23 at 13:49

asked Jun 23 '23 at 14:37

Alex

In Poisson regression the left hand side is log(Patnet) not Patnet. Does that help? – mdewey Jun 23 '23 at 14:52
https://stats.stackexchange.com/search?q=interpret+poisson+regression – whuber Jun 23 '23 at 16:40

interpreting coefficient of Poisson regression, when dependent variable is a count variable

0 Answers0