A great example of the utility of R

I am in the process of moving over to R as my baseline statistics applications.  Recently I have been working largely in STATA and then moving some results over to R for its fantastic graphic options.  While this has worked, it has significant limitations.  One of the strengths of R is its creation of objects that can be fed automatically into graphics commands (which STATA can also do -- but in a less comprehensive way).

The following blog post is a great example of how R works and the advantages it provides.  After the break, I will have a short commentary on why I think this example illustrates the utility of R so well.

http://rforpublichealth.blogspot.com/2013/08/exporting-results-of-linear-regression_24.html

Re-launch for Research in Public Management

It has been quite awhile since I posted -- but it is just because I have been reorganizing and resting up.  I plan to re-launch the site this week with a much broader web presence.  This blog will house my blog posts related to the research side of public management -- working papers, comments on papers/book I read, methods tutorials, etc.

I will launch a second blog for my teaching related activities at www.EBPubMan.blogspot.com.  There you will see links to news articles and other blog posts relevant to emergency management, evidence-based public management, and public management issues generally.

I have also launched a twitter feed.  Follow me @Robinson_OU for my professional twitter posts along the range of research and teaching activities reviewed above

Some Scary Simulations on the Robustness of Logit

Based on my training, I was always believed that logit models were relatively robust.  Specifically, the rule-of-thumb had been that having more than 50 observations was probably ok for simple MLE models like logit.  Some sticklers prefer a floor of 100 observations.  However, some recent research suggests that the floor may be need to be higher -- much higher.

Big Changes Coming

There are big changes coming to the blog.  I will try to ramp up posting in anticipation of the changes -- but I am not sure how well that will work.  I am moving to the University of Oklahoma's Department of Political Science to become the Bellmon Chair of Public Service and becoming the editor of Risk, Hazards, Crisis, and Public Policy starting in the fall of 2013.  I am considering a blog for each of these new responsibilities.

Randomized Response Surveys -- Curious but Unconvinced

I have run across an interesting strategy for survey design -- that may interest others as well.  A common criticism of surveys (especially those on potentially sensitive subject -- or anything with a social affirmation bias) is that people may not answer honestly and there is no way to ensure that they do.  One strategy to overcome this is randomized response.

Video resources for STATA

As I have posted here, there are a variety of resources to help with LaTeX and R available on youtube.  STATA has now created its own channel of short tutorial videos.  I am hoping this initiative takes off and inspires more official videos as well as some unofficial videos on more niche subject.  The channel is available at:

http://blog.stata.com/2012/09/26/stata-youtube-channel-announced/

Comparative Histograms in ggplot2

I have been doing a lot of work with limited dependent variables lately.  With my push for better visualization as a means to communicate findings (as opposed to relying on tables of coefficients), I have had to develop some strategies for building effective comparative histograms.  I did some of this in my recent JPART and continue to use the same code for other projects (mostly, my risk perception and trust work).  I thought some might like to see how simple it is to create nice comparative histograms for categorical dependent variables: