In the words of the authors, clarify does the following:
„First, and most importantly, it can extract new quantities of interest from standard statistical models, thereby enriching the substance of social science research. Second, our approach allows scholars to assess the uncertainty surrounding any quantity of interest, so it should improve the candor and realism of statistical discourse about politics. Finally, our method can convert raw statistical results into results that everyone, regardless of statistical training, can comprehend.“ KING, G., TOMZ, M. & WITTENBERG, J. 2000. Making the Most of Statistical Analyses: Improving Interpretation and Presentation. American Journal of Political Science, 44, 347-361, p.348
Clarify is an awesome tool, and the authors offer an excellent manual. But still, I had to learn the hard way how to create nice graphs like the Figure below. Try, fail, try, fail, try, succeed – but in a nonsatisfactory way, try again, fail abysmally…
Thus, I will try to remedy the situation that confronts clarify beginners and post some lines of code and example graphs here, so that you can get an idea how you can create similar graphs.
Disclaimer No.1: I´m not a good stata coder. I would maintain that I´m a better-than-the-average-user, but still, much of my code will look clumsy to stata sages. That´s why this little tutorial is called „…for dummies like me“. Nevertheless, the code works, and I hope is transparent and accessible in its clumsiness. No shortcuts, no magic, just simple steps. If you have any comments or more elegant solutions, don´t hesitate to share them with me. I´m glad to share some of my hard-learned clarify lessons, but I´m also eager to learn some new tricks. For those of you familiar with clarify, the first examples will be trivial. However, I hope that I can proceed to the advanced tricks later on.
Disclaimer No.2: In the following blogposts, I will often use the most horrendous econometric models, failing to control for temporal dependence in event history models, severely mishandling time-series cross-sectional data and committing further unspeakable sins. However, the purpose of the posts is not to demonstrate good econometric modeling. It´s about producing meaningful graphs with clarify.
Disclaimer No.3: In the following, I presume that you are already familiar with the basics of clarify syntax, and that your only problem is: How do I get these graphs? I will try to put extensive comments in the code, but still…reading the manual is highly encouraged.
And of course, I presume that you already have installed clarify. If not:
net install clarify
And the last disclaimer: I currently use stata 11. So all the code works fine with stata 11, I haven´t tested it with other versions (especially some of the graph commands are sometimes subject to change between stata versions).