Graphing twoway scatter plots

[Lisa Space]

Dear statalist community,

I currently have a small problem with performing graphs.

In the beginning of my analysis I just want to have some basic graphs before I start with the regressions.

I tried to do twoway scatter plots but they just don't show anything.

My dependent variable in and my variable that should be shown on the x-axis is risk preferences on a scale for 0-10 and my explanatory variables that should be shown on the y-axis are married /non-married, divorced/not-divorced, widow/not

For example, married is a dummy variable and risk preferences is categorical.

twoway (scatter married risk_preferences, sort)

This is what I performed for ever variable and the scatter then always looks the following:

any ideas what I am doing wrong?

If you need more information let me know

ah and btw I haven't told stata that I'm using panel data, so I did not run the xtset command.

Best regards

Lisa

[Dave Airey]

Have you tried changing the opacity of the plotting symbols and adding jitter to the data? You might just be suffering from overplotting.

[Lisa Space]

Dave Airey thank for the quick reply. I haven't tried that but to be honest I also don't know how to do that. I would be grateful for an advice on how to do this in stata.

[Lisa Space]

I tried omething out now but don't know if that's correct:

twoway (scatter married risk_preferences, sort mcolor(%30) jitter(1))

so if that's correct then I have still got the same problem :/

[Danial Hoepfner]

There isn't a problem with the code or data, the problem is that this isn't a good way to graph these variables. All of your married are either 0 or 1 and risk is 1,2,3 etc. Stata faithfully plots all of these points where they are, which is right on top of each other. A larger jitter(random movement of the points) will get more separation, but doesn't really help you understand the data that much better. Try a histogram split by marriage instead.

***What you have
twoway (scatter married risk_preference)
****Only a little better
twoway (scatter married risk_preference, jitter(20))
****Better
hist risk_preference, by(married) discrete

[Nick Cox]

See also e.g. https://www.statalist.org/forums/for...updated-on-ssc

[Dave Airey]

It is true that the given plot is not the best choice for the problem.

clear
input y x
1 1
1 2
1 2
1 3
2 1
2 2
2 3
2 3
end
expand 100
table y x
scatter y x, msymbol(oh) mcolor(%30) jitter(7) ///
 xlabel(0(1)4) ylabel(0(1)3)
hist x, by(y) discrete
// ssc install tabplot
tabplot y x

[Lisa Space]

Thank you so much for all that help !!!!