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:

New Paper on Punctuated Equilibrium Theory and Organizational History

I am almost done with the textbook so I am getting back to (or on to) a variety of writing projects.  This is my recent paper getting back in the PET (punctuation equilibrium theory) game.

Paper after the break...

A Resource for TikZ

I have been leaning on TikZ while preparing my lecture notes for our introductory statistics class.  I just ran across this great blog with tons of examples and advice related to creating the sort of figures we often need in social science statistics.  Enjoy -- I am sure I will.

TikZ for Economists

Some Techniques I Would Like to Work On

This is about as self-referential as it gets... but here it goes any way.  There are a few techniques that I would like to learn over the next semester.  I figured I would mention them here in case others are interested in checking whether they might benefit by the same techniques.  I may write more about them in the future as I explore their limits:

• Dot plots (as a replacement for tables of descriptive statistics)
• Mosaic plots (as a replacement for tables for categorical variables -- maybe mostly for presentations)
• Matching -- possibly the integration of matching, estimation, and simulation in Zelig
• CART (classification and regression trees) for something radically different than traditional regression approaches

New Mission for the Blog?

I have been traveling quite a bit while also trying to catch up on writing projects.  As a result, I have been away from my blog quite a bit.  This has given me some time to consider the function the blog can serve.  I have largely used it as a way to post working papers -- with occasional posts on other subjects like job announcement or journal rankings.  There are a few directions I could go with the blog:

1. I stay with this hodgepodge approach
2. I could focus on substantive public management issues by reacting to recent papers, articles, presentations, and even news items
3. I could use the blog to focus on methods to sync with my upcoming PA methods textbook 

I will have to think this over; but, if you have any recommendations please do not hesitate to e-mail me at srobinson @T bushschool.tamu.edu.  

In the mean time, I will work on come content including an update for the journal rankings of PA policy journals and some more LaTeX templates for material like tables.

Norms of Attribution for Data

I am steadily accumulating windmills against which I tilt.  The one I want to bring to public management scholars attention today is attribution for data.

We have strong norms for ensuring attribution of theoretical advances.  Our citation lists overflow with recognition of previous academic publications.  Professional advancement is driven, in part, by the process of getting recognized for these sorts of publications.  However, we could not write many of the articles we do without data collected by others.  It is time data sets get the recognition they deserves.

Theory in Public Management

Following an interesting presentation by Don Moynihan earlier this week, I have been thinking about the state of theory in public management research.  My conclusion is not new -- there does not seem to be much theory to speak of in the field.  Part of the problem lies in the various ways that theory is defined in the field.  Let me debunk a few of these.

[1] Theory is not the same as normative argumentation.  When one self-identifies as a "public management theorist", one often means that you focus your attention on issues of administrative ethics.  This is not wrong but is a limited use of the term "theory" and not what most people mean when they complain about the lack of theory in the field.  I will set this style of theorizing aside for my purposes.

[2] Theory is not having a reference to illustrate that your hypothesis has some support in the literature.  This is the key problem with the public management literature.  Having a citation is not the same as being theory-based.  It is often the case the articles clearly started with a significant puzzle or problem, developed a prediction, and then sought theory to support the prediction.  This is not necessarily wrong but it is a recipe for non-accumulative literature.   In one article, one might draw on a theory from sociology and in another the same author may draw from a theory in economics.  The theoretical connections are an afterthought and represent a cultural genuflection in the direction of theory -- but this is not theory-driven research.

So, most of what we see is not theory-driven.  What would theory-driven research look like?  At the risk of trivializing the subject, a theory is simply a machine that generates hypotheses.  If you start with hypotheses and change machines with frequency, your research may be problem-centered but it is not theory-centered.  If you focus on the machine, see what hypotheses come from it, and then select a domain for research -- you may be conducting theory-driven research.  Starting with the theory and following it where it generates hypotheses allows for a transparent and direct connection between theory and the results of empirical research.

This is why I don't think that PSM (public service motivation) is a theory.  It is a measured construct that is applied in a variety of contexts.  There is nothing internal to PSM to direct where the next study will go.  Instead, it is a variable people use when grafting a variety of theories onto research questions (PSM and red tape, PSM and sector choice, PSM and learning, PSM and performance management).  The statement "PSM matters" is not a theory -- though it may be true.  It may develop into a theory of administrative decision-making but it is not there yet.

This being said, I am not always a theory-driven researcher myself.  I sometimes start with a domain of interest (say, emergency management) and then conduct research to address key practical puzzles.  In other research, I am more traditionally theory-driven (say, in my punctuated equilibrium research).  One of the reasons that I do not conduct theory-driven work more frequently is the difficulty of defining a theory of public management.  I am slowly building this up through a combination on problem-centered and theory-centered research but none stands out right now.  I guess I better get to work on that.  Until then, I will complain about the lack of theoretical grounding in public management research and then proceed along similar lines myself.

Job Opening at NVOAD

NVOAD has an opening in a director of membership services.  The official job ad is after the break.

Revised Version of My Recent Paper on Emergency Management Networks over Time

I previously posted a paper on the evolution of emergency management networks over time.  I have a new version with a shiny new title -- after the break.

New Paper on the Survey Measurement of Collaborative Public Management

I (along with my co-author, Ling Zhu) will present our recent research into measuring collaborative management a the Midwest meeting this weekend. In related news, my previous paper on measuring collaboration is just now out pre-publication from Policy Studies Journal.  Link to our new paper after the break.

A New Blog Series on Visualization

There is a new blog series on visualization techniques -- mostly using R -- starting up.  I recommend checking this out.  When appropriate I may comment or supplement with applications to public management research.

I was hooked with the list of simple advice.  But, there is a lot more here to read and consider.

The Cost of Software

Students often have an inaccurate view of actual software prices because of generous "student pricing" options for statistics software.  I think it is irresponsible to train a generation of students on software that will cost offices thousands per license when there are less expensive options available.  After the break, there is a link to another blog comparing prices of some popular software packages.

Another listing for multiple faculty hires in homeland security (Saint. Louis University)

Saint Louis University (SLU), a Catholic, Jesuit institution dedicated to education, research, service, and healthcare, invites applications for 2 tenure-track faculty positions for Assistant or Associate Professor, in the Department of Health Management and Policy, for its expanding programs in Emergency Management and Crisis Leadership at the School of Public Health. The School of Public Health is the only accredited school of public health in a Jesuit, Catholic university in the U.S., and is committed to social justice, excellence in education, community engagement, and research that leads to improved health. Applicants must have a doctoral degree with advanced education in emergency management and leadership development, as well as, a background in health management, public health, health/public policy, homeland security, emergency preparedness, or related fields. Applicants should have a productive research record and demonstrated ability to secure external research funding. Primary responsibilities will include conducting independent research related to areas such as emergency management and policy development, crisis leadership, systems thinking, complex and adaptive systems, decision science, social network analysis, resilience and continuity modeling and community capacity development. Other responsibilities include teaching graduate and/or undergraduate courses in emergency management and crisis leadership and other areas of teaching such as health management, policy development and public health sciences, and providing professional and community service.

The SLU School of Public Health is nationally recognized for its community based research. Successful candidates will find opportunities for collaboration within the School and with university and community partners. The Department of Health Management offers a Master of Public Health in Emergency Management and Crisis Leadership and a Bachelor Degree in Emergency Management. It also offers one of the top-ranked Master of Health Administration Programs in the nation, a Master of Public Health in Health Policy, and an undergraduate degree in health management. In addition, the School offers dual degrees with several units in the University, including the School of Law, known for its top-ranked health law program.

Inquires may contact the Search Chair by phone (314-977-8162) or send an email including a copy of their curriculum vitae to:

Kate Wright, Associate Professor and
Chair, HMP-EMCL Search Committee
Saint Louis University School of Public Health
3545 Lafayette Avenue, Room 377
Saint Louis, MO 63104
wrightks@SLU.EDU

Saint Louis University is an Affirmative Action, Equal Opportunity Employer,
and encourages nominations and applications of women and minorities.

Process Tracing vs. "process tracing"

I would like to take a moment to clear up what seems to be a common confusion in qualitative research.  The past decade or so has seen an explosion in qualitative research -- and particularly research associated with the Process Tracing techniques championed by George, Bennett, and others.  I would like to take a moment to explain the (considerable) merits of Process Tracing and to distinguish it from "process tracing."

Process Tracing involves the elaboration of a casual mechanism from a theory to identify various points where a theory can be potentially falsified.  Each step in the causal process provides an opportunity for falsification.  Typically, Process Tracing then involves the historical investigation of very specific points in time that can serve to test the original theory.

Notice that Process Tracing requires a strong, detailed theory.  Most of all, you must have a theory that makes detailed predictions about casual processes. You have to set out the specific potentially falsifying moments in the sequence of causal mechanisms as well as the potential evidence for falsification.  Such evidence may be textual accounts of behavior -- the actual type of data is less central to the Process Tracing technique than the elaboration of a causal mechanism.

In the end, your Process Tracing strategy must say "the theory would predict X to happen at historical moment T -- evidence that Y happened instead would constitute contrary evidence."  Of course, not every falsification would necessitate throwing out the entire theory.  However, the theory and the historical test need to be clear enough to draw the theory into question if Y happened instead of (theory predicted) X.

Unfortunately, this powerful tool of investigation has been competing for attention with a seemingly recent innovation -- "process tracing."  As you can tell, one inconvenience is that this new competitor uses the name of Process Tracing and seeks to borrow its legitimacy as a research strategy.  However "process tracing"  seems to be a different beast entirely.  "process tracing" involves maybe (maybe not) presenting the name of a theory and then providing a chronological account of events that occasionally mention terms similar to the name of the theory.  There are no specific predictions to potentially falsify the theory.  Often specific events fade into the background as general themes dominate attention.  "process tracing" is not a theory testing exercise but (at best) a theory motivated narrative of events.

Please do not confuse Process Tracing with "process tracing."

Next up -- replication and reproducibility of public management research...

An Updated Set of Beamer Slides - Now with Tikz and Matrix Examples

I have updated my Beamer slide gallery with a couple of new specialized slides.  You can see the new slides -- one including a node-based graph and one including a matrix.  I will again paste the actual LaTeX code below so you can see how to make or adapt similar slides.