PRISM Brownbags Archive

Spring 2006

Introduction to HLM6
Spring 2006 Quarter
Derby Hall 0125 (Basement Computer Lab)
Presenters: Roman Ivanchenko and Lyndsey Young

Post-Estimation Techniques in Statistical Analysis: Introduction to Clarify, S-Post, and Graphing in Stata
May 19, 2006
2:00-3:30pm
Derby Hall 0125 (Basement Computer Lab)
Presenters: Roman Ivanchenko and Lyndsey Young

Winter 2006

Introduction to R
February 17, 2006
9:30-11:00am
Derby Hall 0125 (Basement Computer Lab)
Presenters: Roman Ivanchenko and Lyndsey Young
Powerpoint presentation
Data files

STATA Data Management and Programming
January 20, 2006
9:30-11:00am
Derby Hall 0125 (Basement Computer Lab)
Presenters: Roman Ivanchenko and Lyndsey Young
Powerpoint presentation
Data files

Autumn 2005

Introduction to Stata
Friday, November 18, 2005
3:30 – 5:00 p.m.
Derby Hall 0125 (Basement Computer Lab)
Presenters: Sean Williams and Lyndsey Young
Powerpoint presentation

Spring 2005

Introduction to R
Tuesday, May 31, 2005
3:30 - 5:00 p.m.
Derby Hall 0125 (Basement Computer Lab)
Presenters: David Darmofal and Corwin Smidt
Presentation (in PDF)
Supporting Documents
Commands
Data

Winter 2005

Introduction to LaTeX
Thursday, February 17, 2005
3:30 - 5:00 p.m.
Derby Hall 0125 (Basement Computer Lab)
Presenters: Corwin Smidt and Zach Mears
Presentation (in PDF)
Supporting Documents
sections.tex    sections.pdf
example.tex   example.pdf
table.tex         table.pdf
hello.tex         hello.pdf

Autumn 2004

Post-Estimation Techniques in Statistical Analysis: Introduction to Clarify, S-Post, and Graphing in Stata
Tuesday, November 16, 2004
3:30 - 5:00 p.m.
Derby Hall 0150 (Basement Computer Lab)
Presenters: Dave Darmofal and Corwin Smidt
Clarify and S-Post Presentations (in PDF)

Introduction to Stata
Tuesday, October 5, 2004
2:00 - 3:30 p.m.
Derby Hall 0150 (Basement Computer Lab)
Presenter: Corwin Smidt
Presentation (in PDF)

Spring 2004

Saving Time With Prudent Data Management: Working with Data and Programming in Stata
The presentation is an overview of the resources and methods available to efficiently manage one's data in Stata. It covers general methods and tools such as generating variables, sorting variables, expanding and reshaping datasets, and dealing with date and time functions. An introduction to programming within Stata follows with applications demonstrating its benefits as a more efficient means of managing data.
Presenters: Brandon Bartels and Kevin Sweeney
Presentation (in PDF)

Advanced Programming in Stata
The presentation follows up on the introductory programming practices covered in the Saving Time With Prudent Data Management brownbag. The first section instructs one on how one can program one's own likelihood function, followed by the demonstration of specific applications in Logit and Probit, Heteroskedastic regression, and split-population duration models. The second section introduces the simulation programming capabilities of Stata with a focus on bootstrapping, Monte Carlo simulation, and programming post-estimation simulations not covered by canned programs like Clarify.
Presenters: Kevin Sweeney and Brandon Bartels
Presentation 1 and 2 (in PDF)

Teaching PS 585
This session was a PRISM hosted roundtable that encouraged faculty and graduate students to share their philosophies and strategies in teaching the undergraduate-level research and statistical methods course. Syllabi were exchanged and those with experience in teaching 585 expressed what difficulties they have encountered, what strategies they have used, and what practices they felt worked.

Winter 2004

Introduction to Stata
Broad overviews of Stata, including information on getting started, as well as details on how to do basic data management, generate descriptive statistics commands. The session is targeted to those who have never used Stata, but those who have a vague applied understanding of Stata may also find the session useful.
Presenter: Kevin Sweeney
Presentation (in PDF)

Post-Estimation Techniques in Statistical Analysis: Introduction to Clarify & S-Post
Clarify (by Michael Tomz, Jason Wittenberg, and Gary King) and S-Post (by J. Scott Long), both free and easy to download software packages, offer researchers the means of easily interpreting and presenting the effects of independent variables from maximum likelihood models. Common post-estimation techniques include: 1) calculating the predicted probability of an event occurring given a covariate profile; 2) calculating the change in the probability of an event occurring given a particular change in an independent variable of interest (while holding other variables constant); and 3) producing various types of graphs, including the option of graphing the probability of an event occurring as a function of a particular independent variable of interest (while holding other variables constant). These techniques, among others, greatly aid an analyst in the most efficient and powerful presentation of results from a statistical analysis.
Presenter: Brandon Bartels and Kevin Sweeney

Endogeneity in Political Science Research: a panel on the issue of endogeneity in political science research.
Endogeneity exists in a model when an explanatory variable is not truly exogenous, but instead, is caused (or affected) by other variables within the system. In general, unaccounted for reciprocal causation leads to biased results and potentially misleading inferences. The panel examines endogeneity from three angles: as a theoretical issue, an issue of research design, and the issue of dealing with endogeneity statistically and its possible problems.
Panelists: Professor Janet Box-Steffensmeier, Professor Dean Lacy, Professor Irfan Nooruddin, Professor Brian Pollins, Brandon Bartels