Hi

I am currently looking into how the election of Trump (main independent variable) affected satisfaction with democracy among Europeans (independent variable). Using Eurobarometer data, there is grounds for an RD-design (i.e., it is a natural/quasi experiment, because respondents were randomly divided between the days just before and just after the shock Trump victory).

I examine whether the Trump-effect was different for different subgroups. One such division is whether the respondent is skeptical or positive about globalisation ('loser vs. winner' of globalisation). I furthermore expect this difference to be most pronounced in countries, where the welfare state is less supportive of the weak (i.e. losers of globalisation). So my design is a 3-way interaction.

Y = Europeans' satisfaction with democracy
X = Trump election
Z1 = Respondent attitude towards globalisation
Z2 = Respondent's welfare state

I want to graphically illustrate this interaction. My variables are dichotome (Trump effect: Yes/No), metric (Z1: An index of globalisation attitude), and ordinal (Z2: Nordic, Central European, South European, Anglo-Saxon, i.e. from most to least supportive).

I suspect marginsplot is a (the?) way to go, but I can't figure it out. Preferably, the x-axis should feature my Z1, X-axis should show (marginal) effect on Y by X (i.e. the "Trump effect"), and four scatter/lines should each represent one type of welfare regime.

P.S. I am new to the forum so would also appreciate general advice, e.g. about conduct here or level/topic of questions.