The combination of the dependent variable with a characteristic

[Berend Nijhuis]

For a project I need to determine what causes price differences across neighborhoods. In the below table there is the dependent variable (the price), the variable that specifies that neighborhood in which the house is located and the other independent variables. These independent variables are all neighborhood specific.

My question is now how I could measure the effect of the inhabitants for example on the house prices for specific neighborhoods in order to determine if the number of inhabitants is a determinant of price differences across neighborhoods.

Thanks in advance.

[Carlo Lazzaro]

Berend:
you may try interacting -inhabitants- with -local authority- (made numeric) or simply consider -inhabitants- without interaction.

[Marcos Almeida]

Just as a side note, after Carlo's recommendations. If I understood right, you may be facing the fact that "local_authority" is a string variable, hence the regression analysis won't be feasible. In short, you'll get "no observations" message.

The solution could be: first, encode the string variable. The - label - option may be used to present the information labelled according to the strings.

Then you may - regress - using the newly-created predictor. Make sure to use factor notation ("i.") for that matter. Also, you may use the if clause to select a specific category of the predictor.

[Marcos Almeida]

The starting message of the thread led me to think the main obstacles were related to tackling data-management issues in Stata. Therefore, my advice was fundamentally aimed to help you to curb them, such as when we need to transform a string variable into a numeric one when we wish to make a regression analysis .

That said, I strongly recommend to take a look at the FAQ, particularly the topic about sharing data/command/output. Please notice the warning on not to use screenshots. There is also an explanation on the reason for avoiding snapshots.

With regards to the "goal of the research", sorry but the field is not the one I feel acquainted with. What is more, I fear this is an overarching question, I mean, that would potentially encompass the "whole concept of the project", and I believe the best approach is first consulting the ones who are in the coordination (professors, PI, etc), since much talk/debate/reflection/teamwork will be needed.

[Carlo Lazzaro]

Berend:
I do share Marcos' reply.
For what I can get from your query, it seems that a multiple OLS will do the trick.
In econometrics textbook (or assignments?) like yours are covered quite frequently.
You can take a look at https://www.amazon.co.uk/Introductio.../dp/0470032707 (pages 50-55).

[Marcos Almeida]

Just to clarify: ‘in the beginning’, well, there was a message after #3, and my reply in #4 targetted the - now missing - message.