Hi everyone,

I am working at my final paper and need some guidance.

I want to look at the party identification with a specific party_x in the time period 2014 to 2017. My dependent variable is party identification with party_x (binary coding: 0= if identification with another party, 1= if identification with party_x). My independent variables are theoretically based and about right wing extremism.

I would like to answer two questions:

First I want to look at the general trend regarding my indicators and the party_x identification. I am planning to do so with:

– logit party_x independent _variable_1 independent _variable_2 independent _variable_n ,or –

My idea is to look at the Odds Ratio for getting information about the general trend over that time period in regard to the independent variables.


Second I want to look if there is an interaction term by looking at just one independent variable and the interaction with a binary coded time-dummy variable (1= 2016/2017 and 0=2014/2015):

– logit party_x independent_variable_1##time-dummy independent_variable_2 independent_variable_n –

Than I would look at the – margins independent_variable_1, at (time-dummy=(0(1)1)) – for the logit coefficients and plot this with – marginsplot – .

The idea behind that is to visualize the different potential interactions with the gradients of the logit coefficients.

Is that a suitable way to achieve my goals?

I am really thankful for any help or advise

Best