Posts | Kenneth Benoit

Quantitative Text Analysis (TCD 2018)

Thu, 15 Feb 2018 15:30:30 +0000

When and where: 16 February, 23 February, 9 March, 16 March, 23 March 2018; Arts Building 3020.

Instructor: Professor Kenneth Benoit, kbenoit@tcd.ie. Assignment marking by Stefan Muüller.

Overview

The course surveys methods for systematically extracting quantitative information from political text for social scientific purposes, starting with classical content analysis and dictionary-based methods, to classification methods, and state-of-the-art scaling methods and topic models for estimating quantities from text using statistical techniques. The course lays a theoretical foundation for text analysis but mainly takes a very practical and applied approach, so that students learn how to apply these methods in actual research. The common focus across all methods is that they can be reduced to a three-step process: first, identifying texts and units of texts for analysis; second, extracting from the texts quantitatively measured features - such as coded content categories, word counts, word types, dictionary counts, or parts of speech - and converting these into a quantitative matrix; and third, using quantitative or statistical methods to analyse this matrix in order to generate inferences about the texts or their authors. The course systematically covers these methods in a logical progression, with a practical, hands-on approach where each technique will be applied using appropriate software to real texts.

Objectives

The course is also designed to cover many fundamental issues in quantitative text analysis such as inter-coder agreement, reliability, validation, accuracy, and precision. It focuses on methods of converting texts into quantitative matrixes of features, and then analysing those features using statistical methods. The course briefly covers the qualitative technique of human coding and an- notation but only for the purposes of creating a validation set for automated approaches. These automated approaches include dictionary construction and application, classification and machine learning, scaling models, and topic models. For each topic, we will systematically cover published applications and examples of these methods, from a variety of disciplinary and applied fields but focusing on political science. Lessons will consist of a mixture of theoretical grounding in content analysis approaches and techniques, with hands on analysis of real texts using content analytic and statistical software.

Detailed Course Schedule

Week 1: Introduction and issues in quantitative text analysis

This session will cover fundamentals, including the continuum from traditional (non-computer assisted) content analysis to fully automated quantitative text analysis. We will cover the conceptual foundations of content analysis and quantitative content analysis, discuss the objectives, the approach to knowledge, and the particular view of texts when performing quantitative analysis. We will also discuss issues including where to obtain textual data; formatting and working with text files; indexing and meta-data; units of analysis; and definitions of features and measures commonly extracted from texts, including stemming, and stop-words.

Required Reading

Vignette and instructions at https://docs.quanteda.io
Tutorials at https://tutorials.quanteda.io/
Grimmer and Stewart (2013)
Manning, Raghavan and Schütze (2008, 117–120)
Krippendorff (2013, Chs. 9–10)
Dunning (1993)

Resources

Week 1 slides
Getting Started with quanteda and R and the .Rmd source, which you should save and try to knit
Assignment 1: Working with Texts in quanteda and .Rmd source

Week 2: Quantitative methods for comparing texts

Here we focus on quantitative methods for describing texts, focusing on summary measures that highlight particular characteristics of documents and allowing these to be compared. These meth- ods include characterizing texts through concordances, co-occurrences, and keywords in context; complexity and readability measures; and an in-depth discussion of text types, tokens, and equiv- alencies. We will also discuss weighting strategies for features, such as tf-idf. The emphasis will be on comparing texts, through concordances and keyword identification, dissimilarity measures, association models, and vector-space models.

Required Reading

Krippendorff (2013, Ch. 10)
Lowe et al. (2011)
Manning, Raghavan and Schütze (2008, Section 6.3)

Resources

Week 3: Automated dictionary-based approaches

Automatic dictionary-based methods involve association of pre-defined word lists with particular quantitative values assigned by the researcher for some characteristic of interest. This topic covers the design model behind dictionary construction, including guidelines for testing and refining dic- tionaries. Hand-on work will cover commonly used dictionaries such as LIWC, RID, and the Harvard IV-4, with applications. We will also review a variety of text pre-processing issues and textual data concepts such as word types, tokens, and equivalencies, including word stemming and trimming of words based on term and/or document frequency.

Required Reading

Neuendorf (2002, Ch. 6)
Laver and Garry (2000)
Rooduijn and Pauwels (2011)

Resources

Week 4: Machine learning and scaling for texts

Classification methods permit the automatic classification of texts in a test set following machine learning from a training set. We will introduce machine learning methods for classifying documents, including one of the most popular classifiers, the Naive Bayes model. The topic also introduces validation and reporting methods for classifiers and discusses where these methods are applicable. Building on the Naive Bayes classifier, we introduce the “Wordscores” method of Laver, Benoit and Garry (2003) for scaling latent traits, and show the link between classification and scaling. We also cover applications of penalized regression to score and scale texts.

Resources

Required Reading

Manning, Raghavan and Schütze (2008, Ch. 13)
Lantz (2013, Ch. 3–4)

Week 5: Scaling models

Building on the Naive Bayes classifier, we introduce the “Wordscores” method of Laver, Benoit and Garry (2003) for scaling latent traits, and show the link between classification and scaling. We also cover the unsupervised “Wordfish” scaling method of Slapin and Proksch (2008).

Required Reading

Evans et al. (2007)
Laver, Benoit and Garry (2003)
Slapin and Proksch (2008)

Resources

Exercise 5: Classifying legal documents and legislative speeches.

Final Assignment

Project Guidelines

The final project is a written analysis of approximately 4,000-5,000 words. The project replaces a final exam, and is designed to give you a chance to analyze a set of texts that you have chosen, reflecting (hopefully) your own research interests. This can be, and probably makes the most sense to be, textual data from something you are already studying. Which texts you choose, what question you investigate, and how you analyze the texts is your choice, but you must justify the choice.

Content

Your content should include the following:

A cover sheet including the title and your name.
An abstract page, with an abstract of no more than 200 words.
Introduction. An expanded version of your abstract, which introduces the question, states the rationale for trying to answer it, briefly describes your corpus, identifies the methods you apply to the texts, and summarizes the findings.
Motivation. Why have you chosen to analyze this topic? Is there a compelling social reason? Does it contribute to scholarship? This can include a “literature review” (but don’t overdo it).
Description of your corpus. You are free to choose any corpus you wish, of any size, although you must include a justification for the choice of texts, and acknowledge the source. You will need to document any format conversions or pre-processing steps you have applied to the texts prior to analysis. You should also present some basic summary statistics about the texts, prior to analysis.
Description of your methods. What techniques will you apply to analyze the texts? Is there a precedent (in previous scholarly literature) for applying such methods to texts similar to yours? Defend why the application of the methods is appropriate.
Results. Apply the methods, present the findings. Be sure to be explicit about any steps taken.
Conclusions. What conclusions on the question can we draw from the results?

Formatting

There is no rigid set of guidelines, but you should use a Chicago manual of style compatible referencing system (parenthetical references rather than footnotes). Tables should be formatted rather than consisting of pasted output from a statistical package. Think of trying to look (roughly) as if you were formatting a journal article.

Deadline

30 April 2018, 5pm, by email to kbenoit@tcd.ie.

References

Barberá, Pablo. 2015. “Birds of the Same Feather Tweet Together: Bayesian Ideal Point Estimation Using Twitter Data.” Political Analysis 23(1):76–91. doi: 10.1093/pan/mpu011.

Beauchamp, N. 2017. “Predicting and Interpolating State‐Level Polls Using Twitter Textual Data.” American Journal of Political Science 61(2), 490-503.

Beil, F, M Ester and X Xu. 2002. Frequent term-based text clustering. In Eighth ACM SIGKDD international conference on Knowledge discovery and data mining. pp. 436–442.

Benoit, K. and M. Laver. 2008. “Compared to What? A Comment on ‘A Robust Transformation Procedure for Interpreting Political Text’ by Martin and Vanberg.” Political Analysis 16(1):101–111. doi: 10.1093/pan/mpm020.

Benoit, Kenneth and Paul Nulty. 2013. “Classification Methods for Scaling Latent Political Traits.” Presented at the Annual Meeting of the Midwest Political Science Association, April 11–14, Chicago.

Blei, David M. 2012. “Probabilistic topic models.” Communications of the ACM 55(4):77. doi: 10.1145/2133806.2133826.

Blei, D.M., A.Y. Ng and M.I. Jordan. 2003. “Latent dirichlet allocation.” The Journal of Machine Learning Research 3:993–1022.

Chang, J., J. Boyd-Graber, S. Gerrish, C. Wang and D. Blei. 2009. Reading tea leaves: How humans interpret topic models. In Neural Information Processing Systems.

Choi, Seung-Seok, Sung-Hyuk Cha and Charles C. Tappert. 2010. “A Survey of Binary Similarity and Distance Measures.” Journal of Systemics, Cybernetics and Informatics 8(1):43–48.

Clinton, J., S. Jackman and D. Rivers. 2004. “The statistical analysis of roll call voting: A unified approach.” American Journal of Political Science 98(2):355–370. doi: 10.1017/s0003055404001194.

Corley, Courtney and Rada Mihalcea. 2005. Measuring the semantic similarity of texts. In Proceedings of the ACL Workshop on Empirical Modeling of Semantic Equivalence and Entailment - EMSEE ’05.

Däubler, Thomas, Kenneth Benoit, Slava Mikhaylov and Michael Laver. 2012. “Natural Sentences as Valid Units for Coded Political Texts.” British Journal of Political Science 42(4):937–951. doi: 10.1017/S0007123412000105.

DuBay, William. 2004. The Principles of Readability. Costa Mesa, California. http://www.impact-information.com/impactinfo/readability02.pdf.

Dunning, Ted. 1993. “Accurate methods for the statistics of surprise and coincidence.” Computational Linguistics 19:61–74.

Evans, Michael, Wayne McIntosh, Jimmy Lin and Cynthia Cates. 2007. “Recounting the Courts? Applying Automated Content Analysis to Enhance Empirical Legal Research.” Journal of Empirical Legal Studies 4(4):1007–1039.

Gilardi, F., Shipan, C. R., & Wueest, B. 2017. “Policy Diffusion: The Issue-Definition Stage.” Working paper, University of Zurich.

Ginsberg, Jeremy, Matthew H Mohebbi, Rajan S Patel, Lynnette Brammer, Mark S Smolinski and Larry Brilliant. 2008. “Detecting influenza epidemics using search engine query data.” Nature 457(7232):1012–1014.

Grimmer, Justin and Brandon M. Stewart. 2013. “Text as Data: The Promise and Pitfalls of Automatic Content Analysis Methods for Political Texts.” Political Analysis 21(3):267–297. doi: 10.1093/pan/mps028.

Gurciullo, S., & Mikhaylov, S. J. 2016. “Detecting Policy Preferences and Dynamics in the UN General Debate with Neural Word Embeddings”. Working paper, University College London.

James, Gareth, Daniela Witten, Trevor Hastie and Robert Tibshirani. 2013. An Introduction to Statistical Learning with Applications in R. Springer Science & Business Media.

Jürgens, Pascal and Andreas Jungherr. 2016. “A Tutorial for Using Twitter Data in the Social Sciences: Data Collection, Preparation, and Analysis.”

Klašnja, M., Barberá, P., Beauchamp, N., Nagler, J., & Tucker, J. 2016. “Measuring public opinion with social media data.” In The Oxford Handbook of Polling and Survey Methods.

Krippendorff, Klaus. 2013. Content Analysis: An Introduction to Its Methodology. 3rd ed. Thousand Oaks, CA: Sage.

Lampos, Vasileios, Daniel Preotiuc-Pietro and Trevor Cohn. 2013. A user-centric model of voting intention from Social Media. In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (ACL).

Lantz, Brett. 2013. Machine Learning with R. Packt Publishing Ltd.

Laver, M. and J. Garry. 2000. “Estimating policy positions from political texts.” American Journal of Political Science 44(3):619–634. doi: 10.2307/2669268.

Laver, Michael, Kenneth Benoit and John Garry. 2003. “Estimating the policy positions of political actors using words as data.” American Political Science Review 97(2):311–331. doi: 10.1017/S0003055403000698.

Loughran, Tim and Bill McDonald. 2011. “When Is a Liability Not a Liability? Textual Analysis, Dictionaries, and 10-Ks.” The Journal of Finance 66(1):35–65.

Lowe, W. 2008. “Understanding Wordscores.” Political Analysis 16(4):356–371. doi: 10.1093/pan/mpn004.

Lowe, William and Kenneth Benoit. 2013. “Validating Estimates of Latent Traits From Textual Data Using Human Judgment as a Benchmark.” Political Analysis 21(3):298–313. doi: 10.1093/pan/mpt002.

Lowe, William, Kenneth Benoit, Slava Mikhaylov and Michael Laver. 2011. “Scaling Policy Preferences From Coded Political Texts.” Legislative Studies Quarterly 26(1):123–155. doi: 10.1111/j.1939-9162.2010.00006.x.

Lucas, C., Nielsen, R. A., Roberts, M. E., Stewart, B. M., Storer, A., & Tingley, D. 2015. “Computer-assisted text analysis for comparative politics.” Political Analysis, 23(2), 254-277.

Manning, C. D., P. Raghavan and H. Schütze. 2008. Introduction to Information Retrieval. Cambridge University Press.

Martin, L. W. and G. Vanberg. 2007. “A robust transformation procedure for interpreting political text.” Political Analysis 16(1):93–100. doi: 10.1093/pan/mpm010.

Metaxas, Panagiotis T., Eni Mustafaraj and Daniel Gayo-Avello. 2011. How (not) to predict elections. In Privacy, security, risk and trust (PASSAT), 2011 IEEE third international conference on social computing (SocialCom).

Mikolov, T., Chen, K., Corrado, G., & Dean, J. 2013. “Efficient estimation of word representations in vector space.” arXiv preprint arXiv:1301.3781.

Neuendorf, K. A. 2002. The Content Analysis Guidebook. Thousand Oaks CA: Sage.

Pennebaker, J. W. and C. K. Chung. 2008. Computerized text analysis of al-Qaeda transcripts. In The Content Analysis Reader, ed. K. Krippendorf and M. A. Bock. Thousand Oaks, CA: Sage.

Roberts, Margaret E, Brandon M Stewart, Dustin Tingley, Christopher Lucas, Jetson Leder-Luis, Shana Kushner Gadarian, Bethany Albertson and David G Rand. 2014. “Structural Topic Models for Open-Ended Survey Responses.” American Journal of Political Science 58(4):1064–1082. doi: 10.1080/01621459.2016.1141684.

Rooduijn, Matthijs and Teun Pauwels. 2011. “Measuring Populism: Comparing Two Methods of Content Analysis.” West European Politics 34(6):1272–1283.

Ruths, D., & Pfeffer, J. 2014. “Social media for large studies of behavior.” Science, 346(6213), 1063-1064.

Seale, Clive, Sue Ziebland and Jonathan Charteris-Black. 2006. “Gender, cancer experience and internet use: A comparative keyword analysis of interviews and online cancer support groups.” Social Science & Medicine 62(10):2577–2590.

Slapin, Jonathan B. and Sven-Oliver Proksch. 2008. “A Scaling Model for Estimating Time-Series Party Positions from Texts.” American Journal of Political Science 52(3):705–722. doi: 10.1111/j.1540-5907.2008.00338.x.

Steinert-Threlkeld, Z. 2018. “Twitter as Data.” Cambridge University Press.

Tausczik, Y R and James W Pennebaker. 2010. “The Psychological Meaning of Words: LIWC and Computerized Text Analysis Methods.” Journal of Language and Social Psychology 29(1):24–54.

Young, L., and Soroka, S. 2012. “Affective news: The automated coding of sentiment in political texts.” Political Communication 29(2), 205-231.

Yu, B., S. Kaufmann and D. Diermeier. 2008. “Classifying Party Affiliation from Political Speech.” Journal of Information Technology and Politics 5(1):33–48.

Zumel, Nina and John Mount. 2014. Practical Data Science with R. Manning Publications.

Quantitative Text Analysis (TCD 2016)

Thu, 04 Feb 2016 21:48:30 +0000

Meeting times

5 February, 12 February, 11 March, 18 March.

Course Handout

Short Outline

The course surveys methods for systematically extracting quantitative information from political text for social scientific purposes, starting with classical content analysis and dictionary-based methods, to classification methods, and state-of-the-art scaling methods and topic models for estimating quantities from text using statistical techniques. The course lays a theoretical foundation for text analysis but mainly takes a very practical and applied approach, so that students learn how to apply these methods in actual research. The common focus across all methods is that they can be reduced to a three-step process: first, identifying texts and units of texts for analysis; second, extracting from the texts quantitatively measured features—such as coded content categories, word counts, word types, dictionary counts, or parts of speech—and converting these into a quantitative matrix; and third, using quantitative or statistical methods to analyse this matrix in order to generate inferences about the texts or their authors. The course systematically covers these methods in a logical progression, with a practical, hands-on approach where each technique will be applied using appropriate software to real texts.

Objectives

Detailed Course Schedule

Session 1: Introduction and Issues in quantitative text analysis

Required Reading:

Vignette and instructions at http://github.com/kbenoit/quanteda
Grimmer and Stewart (2013)
Manning, Raghavan and Schütze (2008, 117–120)
Krippendorff (2013, Chs. 9–10)
Dunning (1993)

Required Reading:

Krippendorff (2013, Ch. 10)
Lowe et al. (2011)
Manning, Raghavan and Schütze (2008, Section 6.3)

Session 3: Automated dictionary-based approaches

Required Reading:

Neuendorf (2002, Ch. 6)
Laver and Garry (2000)
Rooduijn and Pauwels (2011)

Session 4: Machine Learning for Texts

Required Reading:

Manning, Raghavan and Schütze (2008, Ch. 13)
Lantz (2013, Ch. 3–4)
Evans et al. (2007)
Laver, Benoit and Garry (2003)

Final Assignment

Project Guidelines

Content

Your content should include the following:

A cover sheet including the title and your name.
An abstract page, with an abstract of no more than 200 words.
Introduction. An expanded version of your abstract, which introduces the question, states the rationale for trying to answer it, briefly describes your corpus, identifies the methods you apply to the texts, and summarizes the findings.
Motivation. Why have you chosen to analyze this topic? Is there a compelling social reason? Does it contribute to scholarship? This can include a “literature review” (but don’t overdo it).
Description of your corpus. You are free to choose any corpus you wish, of any size, although you must include a justification for the choice of texts, and acknowledge the source. You will need to document any format conversions or pre-processing steps you have applied to the texts prior to analysis. You should also present some basic summary statistics about the texts, prior to analysis.
Description of your methods. What techniques will you apply to analyze the texts? Is there a precedent (in previous scholarly literature) for applying such methods to texts similar to yours? Defend why the application of the methods is appropriate.
Results. Apply the methods, present the findings. Be sure to be explicit about any steps taken.
Conclusions. What conclusions on the question can we draw from the results?

Formatting

Deadline

30 April 2016, 5pm, by email to kbenoit@tcd.ie.

Data Mining and Statistical Learning

Fri, 23 Jan 2015 10:49:27 +0000

2015, Trinity College, Dublin, Department of Political Science

Instructor: Prof Kenneth Benoit, LSE

Details: Class meets MONDAYS in Feb-March from 14:00 – 16:30, with one exception on Day 2 (see below)

Rooms: See specific dates below.

Note: As the class proceeds, I will add resources (slides, R code, text datasets, problem sets) to each session below.

Main Texts:

James, Gareth, Daniela Witten, Trevor Hastie, and Robert Tibshirani. 2013. An Introduction to Statistical Learning. Springer Science & Business Media.
Lantz, Brett. 2013. Machine Learning with R. Packt Publishing Ltd.
Zumel, N., & Mount, J. 2014. Practical data science with R. Shelter Island, NY: Manning.

Detailed Schedule

Day 1 Working with data and data structures

(Mon 9 Feb, 14:00-16:30, Room 201 Pheonix House)

Datasets, databases, data formats, transforming and organizing data. Review of R data structures, SQL and alternatives.
Required readings: Lantz Ch. 2; Zumel and Mount Ch. 2;
Recommended readings:
- Introduction to SQL
- Introduction to reshape2
- [A non-programmer’s introduction to JSON](A%20Non-Programmer’s Introduction to JSON)
Exercise 1, due Wed Mar 25. Answer key here.

Day 2 Rethinking regression as a predictive tool (Wed 25 Mar, 10:00-12:30, Arts Block 3025)

Revisiting prediction for the classical regression model, including logistic regression. Prediction v. association and causation.
Required Readings
- James et al, Chs 3-4
- Lantz, Ch. 6
- Recommended readings:
- Conway, Drew, and John White. 2012. Machine Learning for Hackers. O’Reilly. Chapter 5, “Regression: Predicting Page Views”.
- Zumel and Mount, Ch. 7
Exercises: None, due to the short week, prediction methods will be rolled into the exercise for week 3.

Day 3 Introduction to machine learning (Mon 2 Mar, 14:00-16:30, 206 Pheonix House)

Naive Bayes classifier, k-Nearest Neighbour, Support Vector Machines.
Readings:
- James et al, Ch 4
- Lantz, Chs 3-4
- Recommended Readings:
- Manning, Raghavan and Schütze (2008, Ch. 13)
- Evans et al. (2007)
- Statsoft, “Naive Bayes Classifier Introductory Overview,” http://www.statsoft.com/textbook/ naive-bayes-classifier/.
Exercise 2: Machine Learning and Prediction. Answer set here.

Day 4 Shrinkage methods (Mon 9 Mar, 14:00-16:30, 206 Pheonix House)

Ridge regression, the Lasso
Readings:
- James et al, Ch 6
- Recommended readings:
- Conway, Drew, and John White. 2012. Machine Learning for Hackers. O’Reilly. Chapter 6, “Regularization: Text Regression”.
Exercise 3:
1. Using the dail2002.dta dataset, select a random subset of 80% of the candidates, and then stepwise methods to discover the or a model that maximizes the variation explained in this training dataset. Then predict the fit to the 20% that you left out, and report the RMSE.
2. Following the worked examples from James et al Ch. 6, do Problem 9 from p263 using the College dataset. You can get this from the “ISLR” package.

Day 5 Unsupervised learning (Mon 16 Mar, 14:00-16:30,Aras an Phiarsigh Room 2.04)

Principal components, clustering methods.

 review the last part of James et al, Ch 6 on principal components regression

```
 James et al Ch 10
```

 Bond, Robert, and Solomon Messing. 2015. &#8220;Quantifying Social Media’s Political Space: Estimating Ideology From Publicly Revealed Preferences on Facebook.&#8221; _American Political Science Review_ 109(01): 62–78.

 Weller, Susan C, and A Kimball Romney. 1990. _Metric Scaling: Correspondence Analysis_. Sage.

Day 6 Working with text (Mon 23 Mar, 14:00-16:30, Room 201 Pheonix House)

Applications to textual data, with a focus on social media text mining from Twitter.
Readings:
- Grimmer, J, and B M Stewart. 2013. “Text as Data: the Promise and Pitfalls of Automatic Content Analysis Methods for Political Texts.” 21(3): 267–97.
- Benoit, Kenneth and Alexander Herzog. In press. “Text Analysis: Estimating Policy Preferences From Written and Spoken Words.” In Analytics, Policy and Governance, eds. Jennifer Bachner, Kathyrn Wagner Hill, and Benjamin Ginsberg.
Exercise 4: Working with Textual Data.

How to install the R package topicmodels on OS X

Tue, 02 Dec 2014 20:10:21 +0000

Many people have reported problems when attempting to install the R package `topicmodels` on R when using OS X Mavericks or Yosemite. The problem is that the binaries are not yet built for these versions of OS X, and you need additional software installed in order to build the source. Once you have built the package from source, however, it seems to work fine.

Here is the solution that worked for me:

Install the free XCode from the App Store. After installation, you need to open the application XCode from the Applications folder, so that it can install the needed components.
You need the gsl (Gnu Scientific Library) installed. One way is to get the latest version from [ftp://ftp.gnu.org/gnu/gsl/](ftp://ftp.gnu.org/gnu/gsl/" rel=“nofollow). A better solution (in my opinion) is to install using homebrew. From Terminal, type:
```
 ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
```

Then install gsl using:

 brew install gsl

Now, from R, try

 install.packages("modeltools") # a dependency of topicmodels
install.packages("topicmodels", type="source") # should compile from source

If for some reason the second command fails, then try

 install.packages("http://cran.r-project.org/src/contrib/topicmodels_0.2-1.tar.gz", repos=NULL, type="source")

The Quantitative Analysis of Textual Data (NYU Fall 2014)

Sat, 06 Sep 2014 16:09:27 +0000

Sponsored by: NYU Department of Politics 2014

pdf version of Course handout

Instructor: Prof Kenneth Benoit, LSE

Details: Class meets TUESDAYS 10:00 – 11:50 in Room 217

Note: As the class proceeds, I will add resources (slides, R code, text datasets, problem sets) to each session below.

Detailed Schedule

Day 0 (16 Sept): Course overview and introduction to the quanteda R package

slides

Day 1 (23 Sept) : Quantitative text analysis fundamentals

slides
examples code (revised 26 September)
Exercise 1

Day 2 (30 Sept): Descriptive Statistical Methods for Texts

slides
examples code, plus text data for Russian texts, Spanish tweets, and the Excel example.
Exercise 2 (on the way)
Additional recommended readings about lexical diversity: (see Dropbox link for the file, or get it from the link here if you have access):
- Labbé, Cyril, Dominique Labbé, and Pierre Hubert. 2004. “Automatic Segmentation of Texts and Corpora.” Journal of Quantitative Linguistics 11(3): 193–213.I wanted to recreate their Figure 8 and perform some tests, so I located the corpus files from the Oxford Text Archive at http://www.ota.ox.ac.uk/desc/2466 (and kudos to Labbé et al for making these available, and with such good documentation). I wrote make_deGaulle.R, which uses files from the corpus as noted, plus this dataset I created from their notes: deGaulleData.csv. If you are looking for a tutorial on how to construct a corpus, this is a good example, and I’ve extensively commented the code.
Additional recommended readings about collocations:
- Manning, Christopher D, and Hinrich Schütze. 2000. Foundations of Statistical Natural Language Processing. Cambridge, Mass: MIT Press. Ch. 5, “Collocations”.
- Bautin, Mikhail, and Michael Hart. “Significant Phrases Detection.”
- Pecina, Pavel. 2005. “An Extensive Empirical Study of Collocation Extraction Methods.” In Association for Computational Linguistics.

Day 3 (7 Oct): Quantitative methods for comparing texts

Day 4 (21 Oct): Dictionary Methods

slides
examples code
a useful regular expressions “cheat sheet”
Exercise 3
Additional recommended readings about dictionaries:
- Graham, Jesse, Jonathan Haidt, and Brian A Nosek. 2009. “Liberals and Conservatives Rely on Different Sets of Moral Foundations.” Journal of Personality and Social Psychology 96(5): 1029–46.
- The LIWC-formatted dictionary for this article is available here.

**Day 5 (28 Oct): Classifiers and supervised scaling **

slides
examples code
Exercise pending
Additional recommended readings about supervised scaling methods:
- Lowe, Will, There’s (Basically) Only One Way to Do it (August 30, 2013). Available at SSRN: http://ssrn.com/abstract=2318543 or http://dx.doi.org/10.2139/ssrn.2318543

Day 6 (4 Nov): Unsupervised scaling models for text

Day 7 (18 Nov): Clustering and topic models

Day 8 (2 Dec): Mining social media

demonstration of text cleaning and Crowd-sourcing (Ken Benoit)
Database structures (Jonathan Ronen)
SQL notes (Jonathan Ronen)
SQL dump of some tweets (Jonathan Ronen)
Social media (Pablo Barberá)
code: Social media (Pablo Barberá)

How to batch convert pdf files to text

Thu, 31 Jul 2014 10:39:36 +0000

Frequently I am asked: I have a bunch of pdf files, how can I convert them to plain text so that analyze them using quantitative techniques? Here is my recommendation.

Download the xpdf suite of tools for your platform. This includes the part we will use, pdftotext.
Alternatives are the Apache PDFBox Java pdf library, and the Python-based PDFminer.
[Windows only – Mac and Linux/Unix have this built in to the Terminal or shell already]: You will need a bash shell for your platform. (It is possible to do what I suggest below using the Windows shell, but it’s been so long since I programmed in the Windows DOS/command line script language that I won’t even attempt it now.) The main options seem to be win-bash and Cygwin.
Create a folder called pdfs in your home folder (for this example – of course it can be elsewhere). Copy your pdf files to this folder.

In a text edtor, create a text file called convertmyfiles.sh with the following contents:

 #!/bin/bash
FILES=~//pdfs/*.pdf
for f in $FILES
do
echo "Processing $f file..."
pdftotext -enc UTF-8 $f
done

(I am not providing a link because if you cannot create a text file and copy this text to it — and crucially edit it slightly for your own needs — then you probably won’t have much luck with these steps anyway.)

 * Open the bash shell (Terminal.app or win-bash or equivalent) and execute the following: cd pdfs

./convertmyfiles.sh

 Now you will have a set of text files (ending with .txt) converted as a set. These will probably need tidying up, as the conversion tends to include cruft like headers, page numbers, etc. Sometimes the layout: single
pdftotext -h
Note that in the file provided, the extracted text is given a UTF-8 (Unicode) character encoding, which is what you should be using whenever possible.
Example: (from Terminal.app on my Mac)
Last login: Thu Jul 31 11:29:44 on ttys001
KBs-MBP13:~ kbenoit$ cd pdfs
KBs-MBP13:pdfs kbenoit$ pwd
/Users/kbenoit/pdfs
KBs-MBP13:pdfs kbenoit$ rm *txt
KBs-MBP13:pdfs kbenoit$ ls
11centerpartiet2004.pdf
11folkpartiet2004.pdf
11kristdemokraterna2004.pdf
11kristdemokraterna2004_300k.pdf
11miljopartiet_de_grone2004.pdf
13radikale_venste2004_ENGL.pdf
13socialdemokraterne2004.pdf
21Ecolo_programme_2004.pdf
21Mouvement_Reformateur_100_propositions_pour_2_Θlect_Vlaams_en_europe.PDF
21SPA_europeesprogramma2004.pdf
convertmyfiles.sh
KBs-MBP13:pdfs kbenoit$ ./convertmyfiles.sh
Processing /Users/kbenoit//pdfs/11centerpartiet2004.pdf file...
Processing /Users/kbenoit//pdfs/11folkpartiet2004.pdf file...
Processing /Users/kbenoit//pdfs/11kristdemokraterna2004.pdf file...
Processing /Users/kbenoit//pdfs/11kristdemokraterna2004_300k.pdf file...
Processing /Users/kbenoit//pdfs/11miljopartiet_de_grone2004.pdf file...
Processing /Users/kbenoit//pdfs/13radikale_venste2004_ENGL.pdf file...
Processing /Users/kbenoit//pdfs/13socialdemokraterne2004.pdf file...
Processing /Users/kbenoit//pdfs/21Ecolo_programme_2004.pdf file...
Processing /Users/kbenoit//pdfs/21SPA_europeesprogramma2004.pdf file...
KBs-MBP13:pdfs kbenoit$ ls
11centerpartiet2004.pdf
11centerpartiet2004.txt
11folkpartiet2004.pdf
11folkpartiet2004.txt
11kristdemokraterna2004.pdf
11kristdemokraterna2004.txt
11kristdemokraterna2004_300k.pdf
11kristdemokraterna2004_300k.txt
11miljopartiet_de_grone2004.pdf
11miljopartiet_de_grone2004.txt
13radikale_venste2004_ENGL.pdf
13radikale_venste2004_ENGL.txt
13socialdemokraterne2004.pdf
13socialdemokraterne2004.txt
21Ecolo_programme_2004.pdf
21Ecolo_programme_2004.txt
21Mouvement_Reformateur_100_propositions_pour_2_Θlect_Vlaams_en_europe.PDF
21SPA_europeesprogramma2004.pdf
21SPA_europeesprogramma2004.txt
convertmyfiles.sh
KBs-MBP13:pdfs kbenoit$

Update 12 November 2015 for Windows (thanks Thomas)

For Windows, one way to do the is to use Windows PowerShell ISE (Integrated scripting environment) in Programs/Accessories as follows:

 cd mypdffolder
$FILES= ls *.pdf
foreach ($f in $FILES) {
C:\Program` Files\xpdf\bin32\pdftotext -enc UTF-8 $f
}

Quantitative Text Analysis 2E, Essex 2014

Tue, 01 Jul 2014 15:02:55 +0000

Quantitative Text Analysis 2E

Essex Summer School 2014

Course handout

Instructor: Prof Kenneth Benoit, LSE

TA: Dr. Paul Nulty, LSE

Day 1: Quantitative text analysis overview and fundamentals

slides

demonstration .R code using US inaugural address speeches (downloaded from https://archive.org/details/Inaugural-Address-Corpus-1789-2009)

exercise 1 instructions and starter code

Texts you will need for **exercise 1 – **on the S: drive

Day 2: The Elements of Textual Data

slides

Zipf’s law demo code in Stata; in R (you need your own output .csv files)

exercise 2 instructions and starter code

exercise 2 solution

Day 3: Descriptive Statistical Methods for Texts

slides

exercise 3 instructions and starter code

exercise 3 solution

Day 4: Quantitative methods for comparing texts

slides

exercise 4 instructions

exercise 4 solution

Day 5: Automated Dictionary-Based Approaches

slides

exercise 5 instructions

exercise 5 solution

Day 6: Document classifiers

slides

exercise 6 instructions

exercise 6 solution

Day 7: Supervised Models for Scaling Texts

slides

additional slides from Tom Mitchell’s Machine Learning textbook

exercise 7 instructions

exercise 7 solution

Day 8: Unsupervised scaling models for text

slides

exercise 8 instructions

exercise 8 solution

Day 9: Topic Models

slides

exercise 9 instructions

exercise 9 solution

Day 10: Working with Big Text Data: Twitter

slides part 1

slides part 2

exercise 10 instructions

exercise 10 solution

You might also be interested in a large corpus of 1.6 million english tweets available from http://cs.stanford.edu/people/alecmgo/trainingandtestdata.zip (see the general web page at http://help.sentiment140.com/for-students), if you want to play with a large dataset.

Quantitative Text Analysis (TCD)

Wed, 08 Jan 2014 13:34:44 +0000

Short Outline

The course surveys methods for systematically extracting quantitative information from text for social scientific purposes, starting with classical content analysis and dictionary-based methods, to classification methods, and state-of-the-art scaling methods and topic models for estimating quantities from text using statistical techniques. The course lays a theoretical foundation for text analysis but mainly takes a very practical and applied approach, so that students learn how to apply these methods in actual research. The common focus across all methods is that they can be reduced to a three-step process: first, identifying texts and units of texts for analysis; second, extracting from the texts quantitatively measured features—such as coded content categories, word counts, word types, dictionary counts, or parts of speech—and converting these into a quantitative matrix; and third, using quantitative or statistical methods to analyse this matrix in order to generate inferences about the texts or their authors. The course systematically surveys these methods in a logical progression, with a very practical hands-on approach where each technique will be applied in lab sessions using appropriate software, on real texts.

Objectives

This course is aimed to provide a practical foundation to and a working knowledge of the main applied techniques of quantitative text analysis for social science research. The course covers many fundamental issues in quantitative text analysis such as inter-coder agreement, reliability, validation, accuracy, and precision. It also surveys the main techniques such as human coding (classical content analysis), dictionary approaches, classification methods, and scaling models. It also includes systematic consideration of published applications and examples of these methods, from a variety of disciplinary and applied fields, including political science, economics, sociology, media and communications, marketing, finance, social policy, and health policy. Lessons will consist of a mixture of theoretical grounding in content analysis approaches and techniques, with hands on analysis of real texts using content analytic and statistical software.

Prerequisites

Ideally, students in this course will have prior knowledge in the following areas:

An understanding of probability and statistics at the level of Advanced Quantitive Methods. This course is not heavily mathematical or statistical but students without the prerequisite level of quantitative experience will find the material difficult to follow. However, it will be possible to apply all of the methods covered using the WordStat software in all but the the last two sessions of the course, even if the students fail to grasp the full statistical workings of each method.

Willingness and ability to use the WordStat/QDAMiner software, a commercial package de- veloped by Provalis Research. This software will be used for all but the last two lessons, although the R library (see next item) may also be used for this purpose.

Familiarity with the R statistical package. Stata may also be used but the lab sessions will be designed to use R coupled with a customized R library designed by the instructor. This is in development and available from http://github.com/kbenoit/quanteda

Detailed Outline

Meetings

Classes will meet for five sessions, on Mondays. These are:

13 January: Phoenix House, PX, Rm. 201 – 12-4pm

27 January: College Green, Seminar Rm.3, from 11am-1pm and 2pm-3pm

10 February: Phoenix House, PX Rm 201 – 12-4pm

3 March: College Green, Rm.2 – 12-2pm and 3-4pm (NOTE CHANGE FROM 17 FEB)

10 March: 1.16 Foster Place, 12am-4pm

Approximately 2/3 of the time will be devoted to lectures, and the other half will consist of “lab” sessions where we will work through exercises in class. Typically we will have a two hour lecture, followed by a break, followed by lab sessions.

Computer Software

Computer-based exercises will feature prominently in the course, especially in the lab sessions. The use of all software tools will be explained in the sessions, including how to download and install them.

Grading

Grading will be based on a combination of four take-home exercises.

Recommended Texts

There is no really good single textbook that exists to cover computerized or quantitative text analy- sis. While not ideally fitting our core purpose, Krippendorf’s classic Content Analysis — just updated — is the next best thing. The staple book-length reading is therefore:

Krippendorff, K. (2013). Content Analysis: An Introduction to Its Methodology. Sage, Thousand Oaks, CA, 3rd edition.

Another good general reference to content analysis that you might find useful as a supplement is:

Neuendorf, K. A. (2002). The Content Analysis Guidebook. Sage, Thousand Oaks, CA.

Other readings will consist of articles, reproduced in the coursepack (and if possible, available as downloadable pdf files from the links below.

Detailed Course Schedule

Session 1: Introduction and Issues in quantitative text analysis

This topic will introduce the goals and logistics of the course, provide an overview of the topics to be covered, and preview the software to be used. It will also introduce traditional (non-computer assisted) content analysis and distinguish this from computer-assisted methods and quantitative text analysis. We will cover the conceptual foundations of content analysis and quantitative content analysis, discuss the objectives, the approach to knowledge, and the particular view of texts when performing quantitative analysis. Two examples will be discussed (based on the Gebauer et. al. and Schonhardt-Bailey readings). Required Reading:

Krippendorff (2013, Chs. 1, 2)

Grimmer and Stewart (2013) Recommended Reading:

(example) Gebauer, Tang and Baimai (2008)

(example) Schonhardt-Bailey (2008)

Roberts (2000)

Neuendorf (2002, Chs. 1, 2, 3) Lab session:

Exercise 1: Working with Texts in QDAMiner

R code to implement Exercise 1 (using quanteda).

Session 2: Selecting Texts and Features

Textual data comes in many forms. Here we discuss those formats, and talk about text processing preparation of texts. These issues include where to obtain textual data; formatting and working with text files; indexing and meta-data; units of analysis; and definitions of features and measures commonly extracted from texts, including stemming, stop-words, and feature weighting. We also cover research design issues designed to ensure reliability and validity, including how to measure agreement. Required Reading:

Krippendorff (2013, Ch. 3–4, 6, 10)

Jivani (2011)

http://en.wikipedia.org/wiki/Stop_words (and browse lists of stop words)

Manning, Raghavan and Schütze (2008, 117–120) Recommended Reading:

Wikipedia entry on Character encoding, http://en.wikipedia.org/wiki/Text_encoding.

Browse the different text file formats at http://www.fileinfo.com/filetypes/text.

Neuendorf (2002, Chs. 4–7)

Klingemann et al. (2006, Appendixes I–II)

Banerjee et al. (1999) Lab session:

Exercise 2: Extracting features from texts and summarizing them descriptively

Session 3: Quantitative methods for comparing texts

Quantitative methods for comparing texts, such as char- acterizing texts through concordances, co- occurrences, and keywords in context; complexity and readability measures; dissimilarity measures; and an in-depth discussion of text types, tokens, and equivalencies. Required Reading:

Krippendorff (2013, Chs. 4–5, 12, 13)

Choi, Cha and Tappert (2010)

Lowe et al. (2011)

Recommended Reading:

DuBay (2004)

Lab session:

Exercise 3: Exploring and comparing texts.

Session 4: Automated dictionary-based approaches

Automatic dictionary-based methods involve association of pre-defined word lists with particular quantitative values assigned by the researcher for some characteristic of interest. This topic covers the design model behind dictionary construction, including guidelines for testing and refining dic- tionaries. Hand-on work will cover commonly used dictionaries such as LIWC, RID, and the Harvard IV-4, with applications. We will also review a variety of text pre-processing issues and textual data concepts such as word types, tokens, and equivalencies, including word stemming and trimming of words based on term and/or document frequency.

Required Reading:

Neuendorf (2002, Ch. 6)

Laver and Garry (2000)

Recommended Reading:

Pennebaker and Chung (2008)

Assignment:

Exercise 4: Applying dictionary coding using QDAMiner.

Session 5: Document classification, machine learning, and non-parametric scaling

Classification methods permit the automatic classification of texts in a test set following machine learning from a training set. This topic introduces classifiers and explains one of the most popular classifiers, the Naive Bayes model. It covers in depth the nature of category systems, methods for assessing classifer performance, the effects of feature weighting on classification accuracy. The topic also covers validation and reporting methods for classifiers and discusses where these methods are applicable.

Required Reading:

Manning, Raghavan and Schütze (2008, Ch. 13)

Evans et al. (2007)

Laver, Benoit and Garry (2003)

Benoit and Nulty (2013)

Recommended Reading:

An online article by Paul Graham on classifying spam e-mail. http://www.paulgraham.com/spam.html.

Bionicspirit.com, 9 Feb 2012, “How to Build a Naive Bayes Classifier,” [http://bionicspirit. com/blog/2012/02/09/howto-build-naive-bayes-classifier.html.](http://bionicspirit. com/blog/2012/02/09/howto-build-naive-bayes-classifier.html)

Yu, Kaufmann and Diermeier (2008)

Statsoft, “Naive Bayes Classifier Introductory Overview,” http://www.statsoft.com/textbook/naive-bayes-classifier/.

Assignment:

Exercise 5: Classifying and scaling movie reviews and court briefs. Uses QDAMiner/Wordstat to classify textual data from http://www.cs.cornell.edu/People/pabo/movie-review-data/ and from Evans et al. (2007), and Quanteda to scale them.

References

Banerjee, M., M. Capozzoli, L. McSweeney and D. Sinha. 1999. “Beyond Kappa: A Review of Interrater Agreement Measures.” The Canadian Journal of Statistics/La Revue Canadienne de Statistique 27(1):3–23.

Barberá, Pablo. 2013. “Birds of the Same Feather Tweet Together: Bayesian Ideal Point Estimation Using Twitter Data.” http://files.nyu.edu/pba220/public/birds_jan2013.pdf.

Benoit, K. and M. Laver. 2007. “Compared to What? A Comment on ‘A Robust Transformation Procedure for Interpreting Political Text’ by Martin and Vanberg.” Political Analysis 16(1):101– 111.

Benoit, Kenneth, Michael Laver and Slava Mikhaylov. 2009. “Treating Words as Data with Error: Uncertainty in Text Statements of Policy Positions.” American Journal of Political Science 53(2, April):495–513.

Benoit, Kenneth and Paul Nulty. 2013. “Classification Methods for Scaling Latent Political Traits.” Presented at the Annual Meeting of the Midwest Political Science Association, April 11–14, Chicago.

Choi, Seung-Seok, Sung-Hyuk Cha and Charles C. Tappert. 2010. “A Survey of Binary Similarity and Distance Measures.” Journal of Systemics, Cybernetics and Informatics 8(1):43–48.

Clinton, J., S. Jackman and D. Rivers. 2004. “The statistical analysis of roll call voting: A unified approach.” American Journal of Political Science 98(2):355–370.

Däubler, Thomas, Kenneth Benoit, Slava Mikhaylov and Michael Laver. 2012. “Natural Sentences as Valid Units for Coded Political Texts.” British Journal of Political Science 42(4):937–951.

DuBay, William. 2004. The Principles of Readability. Costa Mesa, California. http://www. impact-information.com/impactinfo/readability02.pdf: Impact Information.

Evans, Michael, Wayne McIntosh, Jimmy Lin and Cynthia Cates. 2007. “Recounting the Courts? Applying Automated Content Analysis to Enhance Empirical Legal Research.” Journal of Empirical Legal Studies 4(4, December):1007–1039.

Gebauer, Judith, Ya Tang and Chaiwat Baimai. 2008. “User requirements of mobile technology: results from a content analysis of user reviews.” Information Systems and e-Business Management 6(4):361–384.

Ginsberg, Jeremy, Matthew H Mohebbi, Rajan S Patel, Lynnette Brammer, Mark S Smolinski and Larry Brilliant. 2008. “Detecting influenza epidemics using search engine query data.” Nature 457(7232):1012–1014.

Grimmer, Justin and Brandon M. Stewart. 2013. “Text as Data: The Promise and Pitfalls of Auto- matic Content Analysis Methods for Political Texts.” Political Analysis (In Press).

Jivani, Anjali Ganesh. 2011. “A Comparative Study of Stemming Algorithms.” Int. J. Comp. Tech. Appl. 2(6, Nov-Dec):1930–1938.

Klingemann, Hans-Dieter, Andrea Volkens, Judith Bara, Ian Budge and Michael McDonald. 2006.

Mapping Policy Preferences II: Estimates for Parties, Electors, and Governments in Eastern Europe, European Union and OECD 1990-2003. Oxford: Oxford University Press.

Krippendorff, Klaus. 2013. Content Analysis: An Introduction to Its Methodology. 3rd ed. Thousand Oaks, CA: Sage.

Lampos, Vasileios, Daniel Preotiuc-Pietro and Trevor Cohn. 2013. A user-centric model of voting intention from Social Media. In Proceedings of the 51st Annual Meeting of the Association for Com- putational Linguistics (ACL).

Laver, M. and J. Garry. 2000. “Estimating policy positions from political texts.” American Journal of Political Science 44(3):619–634.

Laver, Michael, Kenneth Benoit and John Garry. 2003. “Estimating the policy positions of political actors using words as data.” American Political Science Review 97(2):311–331.

Lowe, W. 2008. “Understanding Wordscores.” Political Analysis 16(4):356–371.

Lowe, William and Kenneth Benoit. 2013. “Validating Estimates of Latent Traits From Textual Data Using Human Judgment as a Benchmark.” Political Analysis 21(3):298–313.

Lowe, William, Kenneth Benoit, Slava Mikhaylov and Michael Laver. 2011. “Scaling Policy Preferences From Coded Political Texts.” Legislative Studies Quarterly .

Manning, C. D., P. Raghavan and H. Schütze. 2008. Introduction to Information Retrieval. Cambridge University Press.

Martin, L. W. and G. Vanberg. 2007. “A robust transformation procedure for interpreting political text.” Political Analysis 16(1):93–100.

Metaxas, Panagiotis T., Eni Mustafaraj and Daniel Gayo-Avello. 2011. How (not) to predict elections. In Privacy, security, risk and trust (PASSAT), 2011 IEEE third international conference on and 2011 IEEE third international conference on social computing (SocialCom).

Mikhaylov, Slava, Michael Laver and Kenneth Benoit. 2012. “Coder Reliability and Misclassification in the Human Coding of Party Manifestos.” Political Analysis 20(1):78–91.

Neuendorf, K. A. 2002. The Content Analysis Guidebook. Thousand Oaks CA: Sage.

Pennebaker, J. W. and C. K. Chung. 2008. Computerized text analysis of al-Qaeda transcripts. In The Content Analysis Reader, ed. K. Krippendorf and M. A. Bock. Thousand Oaks, CA: Sage.

Roberts, C. W. 2000. “A conceptual framework for quantitative text analysis.” Quality and Quantity 34(3):259–274.

Schonhardt-Bailey, Cheryl. 2008. “The Congressional Debate on Partial-Birth Abortion: Constitutional Gravitas and Moral Passion.” British Journal of Political Science 38:383–410.

Slapin, J. B. and S.-O. Proksch. 2008. “A scaling model for estimating time-series party positions from texts.” American Journal of Political Science 52(3):705–722.

Yu, B., S. Kaufmann and D. Diermeier. 2008. “Classifying Party Affiliation from Political Speech.” Journal of Information Technology and Politics 5(1):33–48.

New Directions in Analyzing Text as Data Workshop (2013)

Tue, 23 Jul 2013 12:03:38 +0000

New Directions in Analyzing Text as Data Workshop (2013)

Schedule: Version 26 September 2013

London School of Economics and Political Science

Friday, September 27, 2012, 8:30 – Saturday, September 28, 2012, 17:00

For registered participants only

Venue: New Academic Building, Lincoln’s Inn Fields, lower level room NAB.LG.03

Part of an annual event jointly sponsored by the Institute for Quantitative Social Science at Harvard University and the Ford Center for Global Citizenship at Northwestern University, this year’s event will take place in London, hosted by the Department of Methodology of the LSE. These conferences are co-organized by Ken Benoit (PI on the QUANTESS [Quantitative Analysis of Textual Data for Social Sciences] project at LSE), Daniel Diermeier (two previous conferences at Kellogg), and Arthur Spirling (“New Directions in Text Analysis” at Harvard). Last year this conference took place at Harvard, and next year is scheduled to take place at Kellogg.

FRIDAY

BREAKFAST, COFFEE AND REGISTRATION (Friday, 8:30-9:00)

OPENING REMARKS: Ken Benoit (9:00-9:15)

1) POLICY ANALYSIS FROM TEXTUAL DATA (9:15-10:45AM)

“Scaling Politically Meaningful Dimensions Using Texts and Votes,” Benjamin Lauderdale.

“The Evolution of the Security Dilemma in US-China Relations,” Erin Baggott.

Discussants: Justin Gross and Will Lowe

COFFEE BREAK (10:45-11:00AM)

2) EFFECTS FROM TEXTS (11:00-12:30PM)

“Judicial Identity and the Written Word: Evidence From Causal Text Analysis,” Michael Gill and Andrew B. Hall.

“Analysis of Causal Topics in Text Data and Time Series with Applications to Presidential Prediction Markets,” Hyun Duk Kim, ChengXiang Zhai, Thomas A. Rietz, Daniel Diermeier, Meichun Hsu, Malu Castellanos, and Carlos Ceja.

Discussants: Andrew Martin and Kevin Quinn

LUNCH (12:30-1:30PM)

3) NEW APPLICATIONS FOR OLD SCHOOL QUANTITATIVE CORPUS LINGUISTICS APPROACHES (1:30-3:00PM)

“Towards the automated measurement of deliberative communication.” Valentin Gold and Katharina Holzinger

“UK Press Portrayal of Migrant Groups: A Public-Facing Application of Quantitative Corpus Linguistic Analysis to Newspaper Texts,” (and more about the overall project here) Scott Blinder and William Allen.

Discussants: Benjamin Lauderdale and Philip Resnik

**COFFEE BREAK **(3:00-3:30PM)

4) TRICK, OR TWEET? (3:30-5:00PM)

“Evolution of Political Discussion on Twitter – Ideology, Issues, and Events”, Pablo Barbara, Rich Bonneau, John Jost, Jonathan Nagler, Josh Tucker

“Predicting and Interpolating State-level polling using Twitter textual data”, Nick Beauchamp

Discussants: Andy Eggers and Noah Smith

19:00-21:30 Participant Dinner (by invitation)****

SATURDAY

COFFEE (Saturday, 9:00-9:30AM)

5) MEDIA FRENZY (Saturday, 9:30-11:00AM)

“Taking Out the Trash: An Empirical Investigation of Strategic Timing of News Releases.” Greg Austin and Joseph Walsh.

“News Media: platform or power broker? A study of political quotes in newspaper content using syntactic analysis.” Wouter van Atteveldt.

Discussants: Arthur Spirling and Cheryl Schonhardt-Bailey

MORE COFFEE (11:00-11:15AM)

6) NAMED, FRAMED, AND TAMED (11:15-12:45)

“Identifying Media Frames and Frame Dynamics within and across Policy Issues,” Amber E. Boydstun, Justin H. Gross, Philip Resnik, Noah A. Smith

“Signals of Doubt: Harnessing the Power of Text-Mining to Understand Media Coverage of Climate Change,” Constantine Boussalis

Discussants: Paul Kellstedt and Justin Farrell.

LUNCH (12:45-13:30PM)

7) AND NOW FOR SOME COMPUTER SCIENCE APPROACHES (13:30-15:00)

“Measuring Ideological Proportions in Political Speeches,” Yanchuan Sim, Brice D. L. Acree, Justin Gross, and Noah A. Smith. (And a political science version of this paper available here.)

“Tracing the Flow of Policy Ideas in Legislatures: A Computational Approach,” John Wilkerson, David Smith, Nick Stramp, Jeremy Dashiell. John will also briefly describe the Poliinformatics project.

Discussants: Hanna Wallach and Ken Younge

COFFEE BREAK (15:00-15:15)

8) TOOLS AND METHODS ROUNDTABLE (15:15-17:00PM)

“Structural topic models for open-ended survey responses,” Brandon M. Stewart.

“Policy Miner: From Oysters to Pearls”, Hossein Rahmani and Christine Arnold.

“Corpus-Based Dictionaries for Sentiment Analysis of Specialized Vocabularies,” Doug Rice and Christopher Zorn.

“PiCloud: A Simple Approach to Cloud Computing,” Ken Younge.

(General audience reaction will replace discussants for this session.)

FOLLOWED BY DISCUSSIONS AT THE PUB

FULL LIST OF PARTICIPANTS

<td width="336"> **Department/Institution** </td>
<td width="336"> Department of Political Science, University of California, Davis </td>
<td width="336"> Department of Methodology, LSE </td>
<td width="336"> Department of Govenment, Harvard University </td>
<td width="336"> Center for Empirical Research in the Law, Washington University </td>
<td width="336"> Department of Government, London School of Economics </td>
<td width="336"> Department of Government, Harvard University </td>
<td width="336"> Department of Methodology, London School of Economics </td>
<td width="336"> Department of Government, Harvard University </td>
<td width="336"> Department of Government, London School of Economics </td>
<td width="336"> School of Law, Pennsylvania State University </td>
<td width="336"> Department of Political Science, University of Maastricht </td>
<td width="336"> Harvard Law School </td>
<td width="336"> Kellogg School of Management, Northwestern University </td>
<td width="336"> Civil and Building Engineering, Loughborough University </td>
<td width="336"> Department of Government, London School of Economics </td>
<td width="336"> Law and Social Sciences, Nottingham University </td>
<td width="336"> Department of Government, Harvard University </td>
<td width="336"> School of Computer Science, University of Massachussets </td>
<td width="336"> Department of Government, London School of Economics </td>
<td width="336"> Department of Methodology, London School of Economics </td>
<td width="336"> Carnegie Mellon University </td>
<td width="336"> Political Science Department, University of Washington </td>
<td width="336"> Department of Politics, New York University </td>
<td width="336"> Department of Political Science, University of Alabama </td>
<td width="336"> Department of Politics, New York University </td>
<td width="336"> Yale University </td>
<td width="336"> Political Science Department, University of Washington </td>
<td width="336"> Chair of International Relations and Conflict Management, Konstanz University </td>
<td width="336"> Department of Methodology, London School of Economics </td>
<td width="336"> Krannert School of Management, Purdue University </td>
<td width="336"> Data Editor, The Economist </td>
<td width="336"> School of Law, UC Berkeley </td>
<td width="336"> Department of Politics and International Relations, University of Leicester </td>
<td width="336"> Department of Government, Harvard University </td>
<td width="336"> University of Milan </td>
<td width="336"> Department of Political Science, Northeastern University </td>
<td width="336"> Department of Political and Cultural Studies at Swansea University </td>
<td width="336"> School of Computer Science, Carnegie Mellon University </td>
<td width="336"> Department of Politics at New York University </td>
<td width="336"> Political Science Department, Texas A&M University </td>
<td width="336"> Department of Methodology, London School of Economics </td>
<td width="336"> University of Maryland Institute for Advanced Computer Studies (UMIACS) </td>
<td width="336"> Department of Politics and International Relations, Oxford University </td>
<td width="336"> Compas, Oxford University </td>
<td width="336"> London School of Economics </td>
<td width="336"> MZES, University of Mannheim </td>
<td width="336"> Harvard Law School </td>
<td width="336"> Computer Science, Sheffield University </td>
<td width="336"> Chair of International Relations and Conflict Management, Konstanz University </td>
<td width="336"> MZES, University of Mannheim </td>
<td width="336"> Compas, Oxford University </td>
<td width="336"> Communication Science Department, VU University of Amsterdam </td>

**Name**

Amber E. Boydstun

Aude Bicquelet

Andrew Hall

Andrew Martin

Andy Eggers

Arthur Spirling

Ben Lauderdale

Brandon M. Stewart

Cheryl Schonhardt-Bailey

Chris Zorn

Christine Arnold

Constantine Boussalis

Daniel Diermeier

Derek Thomson

Derek William Vallès

Ekaterina Kolpinskaya

Erin Baggott

Hanna Wallach

Jack Blumenau

Jin Jian

John Gasper

John Wilkerson

Jonathan Nagler

Joseph Walsh

Joshua Tucker

Justin Farrell

Justin Gross

Katherina Holzinger

Ken Benoit

Ken Younge

Kenneth Cukier

Kevin Quinn

Laura Morales

Michael Gill

Monica Poletti

Nick Beauchamp

Nina Wiesehomeier

Noah Smith

Pablo Barbera

Paul Kellstedt

Paul Nulty

Philip Resnik

Raymond Duch

Scott Blinder

Stefano Pagliari

Thomas Daubler

Travis Coan

Trevor A Cohn

Valentin Gold

Will Lowe

William Allen

Wouter Van Atteveldt

Quantitative Text Analysis 2F, Essex 2013

Mon, 22 Jul 2013 10:16:38 +0000

Quantitative Text Analysis 2F

Essex Summer School 2013

Instructor:Prof Kenneth Benoit, LSE

TAs:Dr. Paul Nulty, LSE;Petra Martina Baumann, U. Salzburg

Course handout

Day 1:Introduction and Issues in Quantitative Text Analysis

slides

exercise 1

Texts you will need for exercise 1 – on the S: drive

Day 2:Textual Data, Units of Analysis, Definitions of Features

slides

demonstration of quanteda corpus creation and term-document matrix creation (in R)

Zipf’s law demo code inStata; inR(you need your own output .csv files)

exercise 2

Day 3:Research Strategies in Quantitative Text Analysis

slides

exercise 3– the dictionary file (.lic) you will need is on the S: drive

Day 4:Quantitative methods for comparing texts

slides

exercise 4

Day 5:Quantitative Content Analysis

slides

exercise 5

Day 6: Automated Dictionary-Based Approaches

slides

Exercise 6

Day 7:Document classification and introduction to machine learning

slides

additional slidesfromTom Mitchell’s Machine Learning textbook

exercise 7

Day 8: Non-parametric scaling models for text

slides

exercise 8; and exercise8.R

Day 9: Parametric Models for Text Scaling

slides

exercise 9

Day 10:Working with Big Text Data: Twitter

slides part 1

slides part 2

exercise 10. This requires the R script phones.R but you will also find a hyperlink to this in the assignment instructions. The two files loaded directly from the website in phones.R are ip.Rdatafor the iPhone-related Tweetsand an.Rdatafor those related to Android phones.

You might also be interested in a large corpus of 1.6 million english tweets available fromhttp://cs.stanford.edu/people/alecmgo/trainingandtestdata.zip (see the general web page athttp://help.sentiment140.com/for-students), if you want to play with a large dataset.

MY560: Unsupervised Text Scaling Methods

Tue, 11 Jun 2013 08:52:19 +0000

Here are the slides and files from my “Unsupervised Text Scaling Methods” workshop in our MY560 series from June 11, 2013.

Slides here

Lab Exercise here

Files for lab exercise: the amicus briefs and the Irish 2010 budget speeches

MY560: Supervised Methods for Classifying and Scaling Texts

Tue, 28 May 2013 12:37:01 +0000

Here are the slides and files from my “Introduction to Supervised Methods for Classifying and Scaling Texts” workshop in our MY560 series from May 28, 2013.

Slides here

Lab Exercise here

Files for lab exercise: Movie Reviews and the amicus briefs

ME104 Linear Regression Analysis, 2012

Mon, 23 Jul 2012 06:25:28 +0000

ME104 Linear Regression Analysis

Professor Kenneth Benoit

London School of Economics and Political Science

Course Handout as pdf

Objectives and Learning Outcomes

This course focuses on building a greater understanding, theoretical underpinning, and tools for applying the linear regression model and its generalizations. With a practical focus, it explores the workings of multiple regression and problems that arise in applying it, as well as going deeper into the theory of inference underlying regression and most other statistical methods. The course also covers new classes of models for binary and count data, emphasizing the need to fit appropriate models to the underlying processes generating the data being explained.

This course is primarily about data analysis and developing a deeper understanding of the generalized linear model. The focus is on practice, and this focus is reflected in the choice of texts and in the emphasis on applied coursework. While this course deals to some degree with the generalized linear model on a mathematical and theoretical level, its main focus is practical, the ability to use the techniques when faced with the need in practical research. Consequently the learning method combines lectures and reading with hands-on statistical programming exercises using real datasets.

The learning outcomes associated with this nine-week course are aimed at students being able to:

Develop a deeper understanding of the linear regression model and its limitations;

Know how to diagnose and apply corrections to some problems with the generalized linear model found in real data; discussed;

Use and understand generalizations of the linear model to binary and count data;

Develop a greater familiarity with a range of techniques and methods through a diverse set of theoretical and applied readings;

Know where to go to learn more about the techniques in this class and those called for that were not covered in this class.

Prerequisites

An introductory course in quantitative methods is a prerequisite for this course, since this course extends rather than introduces linear regression analysis.

The Stata statistical package is used for all exercises, and students in ME104 should be familiar or at least prepared to quickly learn this package.

Logistics

Meetings. Classes will meet over ten days for a combination of lectures and lab sessions. Lectures will take place from 10:00–12:30 each day, with labs following from 13:30-15:00. The lab sessions will consist of guided exercises using Stata to implement practical data analysis.

Teaching assistant. The TA in 2012 is Carolina Plescia, C.Plescia@lse.ac.uk. Carolina will be leading the lab sessions.

Computer software. Stata 12 will be be used for all exercises.

Reading materials consist of:

Kennedy, Peter. 1998. A Guide to Econometrics. 5th ed. Oxford: Blackwell.This text provides a three-level discussion of each topic: first a general discussion, then a technical discussion, and then a very technical discussion. Most students find this quite useful since it permits them to dig as deep as their abilities let them or as their need allows.

Agresti, Alan and Barbara Finlay. 2009. Statistical Methods for the Social Sciences (4th Edition). Prentice Hall.Written by statisticians rather than econometricians, this text provides a very accessible treatment of the methods covered in this course in a traditional textbook format. For those who consider this too basic, I recommend Wonnacott and Wonnacott (below) as an alternative.

Wonnacott, Thomas H. and Ronald J. Wonaccott. 1990. Introductory Statistics. 5th Ed. New York: Wiley. This text is recommended, but not required, as a supplement or alternative to Agresti and Finlay if you should prefer W&W or find the A&F text too basic.

The home page for this course will be https://www.kenbenoit.net/ME104/ and you will be able to find this course handout there, links to exercises, and additional material.

Note that a great resource on Stata and other aspects of applied quantitative methodology can be found on the Department of Methodology’s YouTube channel, http://www.youtube.com/MethodologyLSE.

Detailed Schedule

Day 1 – The Error Term and Properties of Estimators

Required Reading:

Kennedy Ch 1–2, “Introduction” and “Criteria for Estimators"

A&F Review by reading Chs 4–5, “Probability Distributions” and “Statistical Inference: Estimation

Recommended Reading:

W&W Ch 7, “Point Estimation”

Homework: Exercise 1

Day 2 – The Classical Linear Regression Model

Required Reading:

Kennedy Ch 3 “Interval Estimation and Hypothesis Testing”

A&F Ch 19, “Multiple Regression and Correlation”

Recommended Reading:

W&W Chs 11–13, “Fitting a Line”, “Simple Regression”, and “Multiple Regression”

Homework: Exercise 2

Day 3 – Inference, Intervals, and Hypothesis Testing

Required Reading:

Kennedy Ch 3 “The Classical Linear Regression Model”

A&F 6, “Statistical Inference: Significance Tests”

Recommended Reading:

W&W 8, “Confidence Intervals” and Ch. 9 “Hypothesis Testing”

Homework: Exercise 3

Day 4 – Diagnosing problems with the CLRM

Additional materials for lecture: anscombe dataset, anscombe.do

Required Reading:

Kennedy Chs 5-6 “ Kennedy Chs 5–6, “Specification” and “Wrong Regressors, Nonlinearities, and Parameter Inconsistency”

A&F 14, “ 14, “Model Building with Multiple Regression”

Recommended Reading:

Berry, William D. and Stanley Feldman. 1993. “Multiple Regression in Practice.” In Regression Analysis, ed. Michael Lewis-Beck.

Fox, John. 1993. “Regression Diagnostics” In Regression Analysis, ed. Michael Lewis-Beck. London: Sage.

Homework: Exercise 4

Day 5 – Problems with predictors

Additional materials for lecture: hiedata_short.dta, hieexample.do

Short guide to interpreting logged coefficients in regression models

Required Reading:

Kennedy Ch 12, “Multicollinearity”; Kennedy Ch 11.1, “Errors in Variables”

A&F Ch 14

Recommended Reading:

W&W 8, “Confidence Intervals” and Ch. 9 “Hypothesis Testing”

Homework: Exercise 5. Do file for exercise 5 here.

Day 6 – Problems with error assumptions

Additional materials for lecture: dail2002.dta

Required Reading:

Kennedy Ch 7, “Non-zero expected disturbance”; Ch 8, “Non-spherical disturbances”; Ch 11, “Simultaneous Equations”

A&F Ch 16.6, “Structural Equation Models”

Recommended Reading:

Hanushek, Eric A. and John E. Jackson. 1977. Statistical Methods for Social Scientists. San Diego: Academic Press. Ch 8, “Structural Equations.”

Kenneth Benoit and Michael Marsh. 2008. “The Campaign Value of Incumbency: A New Solution to the Puzzle of Less Effective Incumbent Spending.” American Journal of Political Science 52(4): 874-890.

Homework: Exercise 6.

Day 7 – Models for binary data: Binary Logistic Regression

Required Reading:

Kennedy Ch 16, “Qualitative Dependent Variables”

A&F Ch 15.1–15.3, “Logisitic Regression”

Recommended Reading:

Aldrich, John H. and Forrest D. Nelson. 1984. Linear Probability, Logit, and Probit Models. Beverly Hills: Sage.

W&W Ch 18 “Maximum Likelihood Estimation”

Homework: Exercise 7.

Day 8 – Models for categorial and ordinal data

Required Reading:

Kennedy Ch 16.2-16.3

A&F Ch 15.4–15.6

Homework: Exercise 8.

Day 9 – Models for count data

Do file for lecture 9

Required Reading:

Kennedy Ch 16.4, “Count data”

A&F Ch 15.4–15.

Recommended Reading:

King, Gary. Unifying Political Methodology: The Likelihood Theory of Statistical Inference. Cambridge, England and New York: Cambridge University Press, 1989. Reprinted, Ann Arbor: University of Michigan Press, 1998. Chapter 5.7–5.10.

Homework: Exercise 9.

Day 10 – Estimating uncertainty in inferential models

Do file for lecture 10

Required Reading:

King, Gary; Michael Tomz; and Jason Wittenberg. 2000. “Making the Most of Statistical Analyses: Improving Interpretation and Presentation.” American Journal of Political Science 44(2, April): 341- 355.

Recommended Reading:

The Zelig package for R — this will allow you to implement the “CLARIFY” techniques in King, Tomz and Wittenberg (2000)

Kennedy Ch 21, “Robust Estimation”

King, Gary. 1986. “How Not to Lie with Statistics: Avoiding Common Mistakes in Quantitative Political Science.” American Journal of Political Science 30(3):666-687.

Computer-Assisted Text Analysis, Essex 2012

Mon, 09 Jul 2012 13:02:34 +0000

Computer-Assisted Text Analysis

Essex Summer School

Prof Kenneth Benoit, LSE

Dr. Paul Nulty, LSE

NOTE: Krippendorf 3rd ed. text is now available from the summer school office!

Course handout

Day 1:Introduction and Issues in Quantitative Text Analysis

slides

exercise 1, and french texts

Day 2: Textual Data, Sampling, and Working with Texts

slides

exercise 2

Day 3: Descriptive Inference from Texts

slides

exercise 3

Zipf’s law demo code in Stata; in R(you need your own output .csv files)

Day 4: Research Design Issues

slides

exercise 4

Day 5: Thematic analysis

slides

exercise 5

Day 6: Classical content analysis

slides

Exercise 5: In the lab, complete this content analysis exercise:http://www.dpeg.net/MCSphase0B.

Day 7: Automated Dictionary-Based Approaches

slides

exercise 7

Day 8: Text Scaling Models – from Dictionaries to “Word-scoring”

slides; application slides

exercise 8; exercise8.R; budget_2010.csv

Day 9: Classifiers

slides; additional slides fromTom Mitchell’s Machine Learning textbook

exercise 9

Day 10: Parametric Models for Scaling Texts

slides

exercise 10

Job listing: Research Officer in Quantitative Text Analysis

Thu, 29 Dec 2011 11:21:20 +0000

Job opening: Research Officer in Quantitative Text Analysis

Duration: 24 months Start Date: 1 March 2012 or as soon as possible thereafter Salary: £31,998 – £38,737 p.a. incl.

Applications are invited for the post of Research Officer, to assist with a principal research officer for the European Research Council funded grant Quantitative Text Analysis for the Social Sciences (QUANTESS).

The research officer’s general duties and responsibilities will be to work with text and the computer organization, storage, processing, and analysis of text. These tasks will involve a combination of programming, database work, and statistical computation. The texts will be drawn from social, political, legal, and commercial examples, and will have a primarily social science focus. The research officer will be expected to work with existing tools for text analysis, actively participate in the development of new tools, and participate in the application of these tools to the social scientific analysis of textual data.

The successful applicant will be expected to possess advanced skills and experience with computer programming, especially the ability use a language used in text processing such as Python; familiarity with SQL; and experience with the R statistical package or the ability to learn R.

The successful candidate should have a postgraduate degree in computer science, computational linguistics, or possibly a cognate social science discipline and have an interest in text analysis and quantitative linguistics, have a knowledge of social science statistics and have worked in a research environment previously.

To apply for this post please go to www.lse.ac.uk/JobsatLSE and select “Visit the ONLINE RECRUITMENT SYSTEM web page”. If you have any queries about applying on the online system, please call 020 7955 6656 or email hr.jobs@lse.ac.uk quoting reference 1223243.

Closing date for receipt of applications is: 31 January 2012 by 23.59 (UK time).

Correlation causes causation

Sat, 22 Oct 2011 16:23:17 +0000

A brilliant cartoon from Craphound:

Technical questions

Sat, 23 Jul 2011 12:01:08 +0000

Just finished teaching at the Essex Summer School in Data Analysis, Session 1. Got a lot of technical questions from my class…

Technical questions

Computer-Assisted Text Analysis (Essex Summer School)

Fri, 15 Jul 2011 11:05:39 +0000

This concerns the short course I am teaching at the Essex Summer School in Social Science Data Analysis, University of Essex, from 11-22 July 2011.

Course handout (syllabus)

Day 1: Introduction to Computer-Assisted TextAnalysis

Day 2: Textual Data, Sampling, and Working with Texts

Class Assignment 1 and thefiles you need for it.

Day 3: Descriptive Inference from Text

Class Assignment 2

R code solution for Assignment 2

Stata code solution for Assignment 2

Day 4: Research Design issues in textual studies

Day 5: Classical Quantitative Content Analysis

Class Assignment 3

French manifestos for Assignment 3

UK Manifestos for Assignment 3

Day 6:Automated dictionary-based approaches

Class Assignment 4

The fileLaverGarryAJPS.ykdused in Assignment 4.

Day 7: Dictionary Construction; Words as Data Approaches

Class Assignment 5

The file iebudget2010.zipused in Assignment 5.

Day 8: Text Scaling Models I

Class Assignment 6

The fileLBGexample.csvused in Assignment 6.

Assignment 6 solution in Excel and SPSS.

Day 9: Text Scaling Models II

Class Assignment 7

Day 10: Additional Scaling Issues

No lab assignment but we will go through the budget speech example using this code: Assignment_8.r.

EUI Multi-Level Models Course

Mon, 23 May 2011 08:26:21 +0000

An Introduction to Multi-Level Models (Using Stata)

European University Institute, May 23–27, 2011

Professor Kenneth Benoit

Methodology Institute, London School of Economics

https://www.kenbenoit.net/mlm/

Course handout here.

Readings are available from Mark Franklin’s Dropbox account for this course. If you are not yet subscribed, then email mark at Mark.Franklin@EUI.eu.

Day 1: Introduction and Motivation for multi-level models.

Required Reading: Rabe-Hesketh & Skrondal (2008, Chs. 1–2); Stata manual for reshape.

Recommended Reading: Franzese (2005); Gelman (2006); Austin, Goel & van Walraven (2001).

Homework 1 and Homework 1 Answer code.

Day 2: Estimating models with multi-level data.

Required Reading: Continue with Rabe-Hesketh & Skrondal (2008, Chs. 1–2) and Stata [XT] manual.

Recommended Reading: Steenbergen & Jones (2002); Austin, Goel & van Walraven (2001); Snijders & Bosker (1999); Goldstein (2003).

Homework 2: Question 2.3 from p87 of Rabe-Hesketh & Skrondal. Use xtmixed for part 2.3.2. Homework 2 Answer code.

Day 3: Random-intercept models.

Required Reading: Rabe-Hesketh & Skrondal (2008, Ch. 3).

Recommended Reading: Austin, Goel & van Walraven (2001); Snijders & Bosker (1999).

Homework 3: Question 3.2 from pp133-4 of R-H&S, plus:

Compare a “random effects”, “fixed effects” (using the “fe” option to xtreg), and “between effects” regression by running them and discussing the differences on the estimated coefficient on the “deprive” variable, including the student-level covariates as in part 3 of the question.

Homework 3 Answer code.

Day 4: Random-coefficient models.

Required Reading: Rabe-Hesketh & Skrondal (2008, Ch. 4)

Recommended Reading: Austin, Goel & van Walraven (2001); Snijders & Bosker (1999)

Homework 4: No pre-assigned homework today, although we will go through an example together in class that I will post on-line on the morning of Day 5.

Day 5: Extensions of the multi-level model

Random intercept (xtlogit) example from class (.do file).

Required Reading: Continue with Rabe-Hesketh & Skrondal (2008, Chs. 6, 9)

Recommended Reading: McMahon & Heath (1992)

CEU Computerized Text Course Announcements

Wed, 13 Apr 2011 15:32:47 +0000

This post concerns the short course I am teaching at Central European University, Budapest from 14-21 April 2011. Stay tuned to this post for future announcements.

The course handout (syllabus) is available here.

The readings are now available in a zip file of pdfs from /assets/files/CEU_CTA_Benoit_readings.zip.

Day 1: Introduction and Overview

Slides for Day 1.

Class exercise for Day 1 here, and the files you need for it here.

Day 2: Descriptive Inference From Text

Slides for Day 2.

Class exercise for Day 2 here.

Day 3: Classical Content Analysis

Slides for Day 3.

Class exercise for Day 3 here.

Day 4: Dictionary-Based Coding

Slides for Day 4.

Class exercise for Day 4 here.

Day 5: Text As Data (and Wordscores)

Slides for Day 5.

Class exercise for Day 5 here.

Day 6: Text Scaling

Slides for Day 6.

No class exercise for Day 6.

R code from class example.

Principles of Comparative Research Week 7 Outline Guidelines

Mon, 01 Nov 2010 11:22:33 +0000

On Friday, November 12 your five-page (maximum) outline of your research proposal is due.

In brief: This is designed to be a shorter version of your final assignment research proposal, providing a more summary version of the project will propose in fuller form at the end of the term. It will include a research question, an indication of the key causal explanation, a discussion of the key observable implications of that causal explanation, and a justification for the project. To make things even more clear, we have suggested explicit guidelines below.

Note:

YOUR NAME needs to be on all of your written work please!

All short proposals must be submitted through turnitin.com.

This outline will be graded and counts for 20% of the final course grade.

Suggested contents for your outline:

Title.

Overview.

What is the main problem you wish to address? Summarize the project and its motivation in no more than a page. Focus on what general theme the project addresses (e.g. “do campaigns make a difference”) and describe the specific and concrete way that your research will investigate this theme (e.g. to study the association between spending recorded in Irish election campaigns and relative vote share received in the election, at the constituency level by candidates).

What contribution will your study make, even if very small, to our understanding of the problem? this section can be short.

Theory and/or Hypotheses. Every project will be different with regard to this section, with some being more aimed at testing “causal relationships” than others. But in this part, probably between 0.5-1.5 pages depending on what your paper is about, you will describe what you hope to test or establish or discover. For instance, you might be interested in examining how the choice of executive institutions influences post-conflict political systems (a version of the presidential versus parliamentary system debate). Or you might be interested in explaining why in a specific case it appeared that two democracies went to war with another, despite the widespread belief that this never happens.

Data. This section needs to clearly identify what the main unit of analysis is and how you will observe it. This section needs therefore to be both conceptual and practical. Sometimes, these two levels point to different units, for instance: say you are interested in why voters in two-ballot (or “mixed member”) electoral systems split their ticket by voting for different parties on different ballots. Your conceptual unit of analysis, following from your theory about why and how voters split tickets, is at the individual level. But your practical unit of observation may be at the level of aggregated election returns, say party totals for each ballot, in an electoral district. This is because no record is published by the election authorities linking individual votes on the two ballots. Alternatively, if you had access to an election study (a detailed survey of individual respondents), then your unit of observation would be an individual survey respondent.

Methods. Here you need to be very clear as to what methodology you plan to use to analyze your data, linking it to theory. Don’t be overly vague, for instance by stating “I will use qualitative and quantitative methods to analyze my dataset. Instead, state: “I will use macro-statistical methods to determine the association between the level of democracy and the rate of foreign war involvement, at the country-year level. I will also analyze the three cases of the X-Y 1970 conflict, the Y-Z 1973 conflict, and the ABCD 1980 conflict in detail using historical accounts.”

Expected and/or potential findings. Here you should also point out what sort of results from your data analysis would lead you to specific conclusions. Conversely you should also point out what sort of evidence would lead to to reject those conclusions. This acts as a good test that your theory and/or hypotheses are potentially falsifiable. Note that if your paper is more of an investigative report on something that has received little systematic study, then your findings will be more of a descriptive nature and hence the falsifiability criterion will apply less to your project than to others. But it should still be easily possible to describe what you might or might not find following a full investigation of your data.

So you want to get a PhD in Political Science ...

Thu, 28 Oct 2010 06:54:05 +0000

This is painfully funny… if completely, ahem, unfounded.

European Political Science Association

Wed, 20 Oct 2010 15:37:16 +0000

The recently formed European Political Science Association has just issued a call for papers for its 1st Annual General Conference, to be held in Dublin, Ireland from June 16-18, 2011. The conference takes place in the excellent Guinness Storehouse Conference Centre. (The Storehouse, along with the Book of Kells at Trinity College, are the two most visited tourist sites in Ireland.)

Information on the conference and the online proposal system can be found here, and the EPSA website accepts discounted early membership registration and payment here.

EPSA is also on Facebook and Twitter (@europsa) if you fancy those.

Enhanced Assessment Methodology

Sun, 03 Oct 2010 14:27:47 +0000

Next stop: London School of Economics and Political Science

Wed, 22 Sep 2010 14:18:29 +0000

I’ve accepted a position at the London School of Economics and Political Science, where I will be Professor and Head of the Methodology Institute from January 2011.

My workplace from Jan 2011

My pre-first round Hungarian forecast judged the most accurate

Wed, 28 Apr 2010 12:16:42 +0000

You have to be able to read Hungarian, but here is the story:

“A Republikoné és a Századvégé volt a legpontosabb mandátumbecslés”

Don’t read Hungarian? Try http://translate.google.ie/?hl=en&tab=wT#hu|en|.

Hungary's Fidesz will win a Two-Thirds Majority of Seats on April 25

Wed, 21 Apr 2010 06:29:35 +0000

My forecast of the final parliamentary seat distribution for the Hungarian elections is the following:

Fidesz: 264 seats (68.4%)

MSZP: 58 seats (15%)

Jobbik: 48 seats (12.4%)

LMP: 16 seats (4.1%)

I’ve written about this in more detail in a post on the Monkey Cage blog.

Hungary's Fidesz will win a Two-Thirds Majority of Seats on April 25

Wed, 21 Apr 2010 06:29:35 +0000

My forecast of the final parliamentary seat distribution for the Hungarian elections is the following:

Fidesz: 264 seats (68.4%)

MSZP: 58 seats (15%)

Jobbik: 48 seats (12.4%)

LMP: 16 seats (4.1%)

I’ve written about this in more detail in a post on the Monkey Cage blog.

Field seminar B 2009-2010, Spring Weeks 6 and 7

Fri, 19 Feb 2010 16:38:06 +0000

NOTE: Week 6 will be a double-length session, since Week 7 clashes with the Political Economy mini-conference. As a result, there is no field seminar meeting on Mar 12. Instead, the Week 6 (Feb 26) will last from 9-13:00.

Strategic Behavior , Dr. Jon Slapin

Readings:

Thomas Schelling. 1978. Micromotives and Macrobehavior. Chapter 1 & chapter 3.

Timothy Besley. 2006. Principled Agents. Chapter 1 & chapter 2.

Barry Weingast. 1997. The Political Foundations of the Rule of Law. Am_erican Political Science Review_ 97(2).

R. Preston McAfee. Introduction to Economic Analysis. Chap 7.1. (pgs 7-251 to 7-270). Available from www.introecon.com.

Statistical haiku

Fri, 29 Jan 2010 16:54:23 +0000

See Keisuke Hirano’s Haiku page. Here is one of my favorites:

T-stat looks too good.
Use robust standard errors–
significance gone.

Field seminar B 2009-2010, Spring Weeks 1 and 2

Wed, 13 Jan 2010 10:33:36 +0000

Interviewing techniques for academic research, Dr. Jacqueline Hayden

Friday January 22^nd and 29^th

Location: Room 3131 Arts Block (Podcast/Video Recording and playback facility)

Time: 9-11 am and 12-1pm

Note: Room 3131 is booked by the English Dept. from 11-12. I have 2 slots booked from 9-11 am and 12 -1. The best option is to take the 9-11 am slot which will allow us to record a number of interview sessions " then (if necessary) we could utilise the 12 " 1pm slot for self and peer evaluation.

This is a practical workshop designed to develop your techniques for academic interviewing. The focus of the two sessions will be on developing your awareness of the issues relating to the following

Why " when to interview?

Pre- interview preparation

Interviewee identification and selection

Preparing your questions

Interviews as part of a data set " checking veracity of interviewee claims/triangulation

Interviewing strategies and techniques

Peer group and self-evaluation

Format

Your interviews will be recorded for self and peer evaluation " we will review the interviews in the 2nd session between 12 and 1pm on both days.

Making this workshop effective requires a fair amount of advance preparation and input from you. You will get out of it what you put in! The best way to find out what works and doesn’t work in interview scenarios is for you to conduct an interview and be interviewed and then to review the end product.

Prior to the workshop you will need to talk to each other about the following:

Who wants to act as an interviewer and interviewee " (depending on time it is hoped that everyone will play both roles).

Select subject areas where you can play expert interviewee and interviewer. More than one person can interview an ’expert’ " you just need to liaise on whose doing what. You can for example use your own research agenda (or current research proposal) or you can create a ‘mock’ research issue " this has its advantages in that it means you can do something outside your own ‘box’. In practical terms this means that Student X is designated as an expert theorist on say " political parties and Student Y will interview him/her for some part of his research. Be creative about choosing your subjectsâ€¦many of you are already on the way to being experts on specific research areas but you also have first hand knowledge of politics from your own domestic political contexts so remember you can adopt many different ’expert’ hats.

You may be able to encourage senior PhD’s or perhaps even some staff to act as your interviewee " all you need is someone willing to take on an ’expert’ role in some area for 10-15 minutes. There is no harm asking!

Create a running order or agenda for the two sessions

Having done all of the above you need to agree on a class member to prepare a list of topics stating who is the ’expert’ interviewee and who’s interviewing (reminder – there can be more than one interviewer per ’expert’ interviewee). This will allow me to allocate a duration for each interview session.

Essential requirement for participation is that you write up a brief memo " 500-750 words which outlines where interviewing fits into your research agenda. I can’t offer a critique of your interview skills if I don’t understand your objectives so specifically you must

designate the aspect or part of your research that necessitates an interview

outline the information you want to illicit from the interview

how you intend to use the interview material

how you intend to validate/triangulate the interview data

how you intend to store or share the data for other scholars (replication purposes)

how you will evaluate the material you get from the interviewing process

outline a framework for the core topics and questions you will use for your interview

PLEASE EMAIL ME THIS OUTLINE BY WED 20 Jan so that I have time to read them before the workshop (please note I have a very heavy teaching schedule on Thursdays)

Reaction papers

Those wishing to write a reaction paper may use their pre-prepared workshop memo as a starting point for an evaluation of the projected role of interviews in their research.

Getting started

I have emailed you papers from a Symposium on interviewing in political science and a paper based on the experience of interviewing in divided societies. Please find a fuller bibliography below prepared by Adrian Bunea below.

Bibliography

Interviewing techniques " Compiled by Adriana Bunea

Handbook of interview research: context & method. 2001. Ed. By Jaber F. Gubrium, James A. Holstein. Sage Publications.

Interviewing experts. 2009. Ed. Alexander Bogner, Beate Littig and Wolfgang Menz. Palgrave Macmillan.

Elite and specialized interviewing. 2006. LA Dexter, A Ware, M Sanchez-Jankowski. ECPR Press.

Mason, Jennifer. 2002. Qualitative researching, Ch 4 “Qualitative Interviewing”. Sage Publications.

Symposium on Interview Methods in Political Science in Political Science and Politics (2002) 35 (4).

Joanne McEvoy 2006 ‘Elite Interviewing in a Divided Society: Lessons from Northern Ireland’ Politics 26 (3): 184-191.

Craig D. Murray and Judith Sixsmith. 2003. ‘E-mail: A Qualitative Research Medium for Interviewing?’ in _Interviewing _Sage Benchmarks in Social Research Methods Vol. 2. Ed. Nigel Fielding. London: Sage. 128-148.

Field seminar B 2009-2010, Week 10

Sat, 28 Nov 2009 10:02:18 +0000

The Michigan Model, Professor Michael Marsh

This is classically presented in A. Cambell et al The American Voter, which is available via Google online. Critical chapters are 2, 6 and 7. Or look at a redced version, for instance in R. Niemi and H. Weisberg eds, Classics in Electoral Behaviour ch 22 [based on TAV ch 6-7].

A sympathetic but contemporary restatement is M. Lewis-Beck at al, The American Voter Revisited ch 6-7.

For Ireland

Michael Marsh, Party identification in Ireland An insecure anchor for a floating party system’, Electoral Studies, 25, (3), 2006, p489 – 508

available from INES page at [www.tcd.ie/ines]

A recent debate on party identification is Bartle and Bellucci’s edited volume Political Parties and Partisanship, Routledge 2008. See esp:

Chapter 3: Thomassen and Rosema

Chapter 4 B. Grofman et al

Chapter 6 Clark et al

Chapter 11 Green and Schickler – a defence of TAV.

Field Seminar B 2009-2010, Weeks 8-9

Wed, 18 Nov 2009 15:35:16 +0000

Week 8: The impact of referendums (MG)

Lawrence LeDuc, T_he Politics of Global Democracy: referendums in global perspective_. Peterborough Ontario: Broadview Press, 2003

Matt Qvortrup, A_ Comparative Study of Referendums: Government by the People_, 2nd ed (Manchester: Manchester University Press, 2005).

Claes De Vreese (ed), The Dynamics of Referendum Campaigns: an international perspective (Basingstoke: Palgrave Macmillan, 2007)

Maija Setala, Referendums and Democratic Government: normative theory and the analysis of institutions (Basingstoke: Palgrave Macmillan, 1999)

Svensson, Palle (2002), Five Danish Referendums on the European Community and European Union: A Critical Assessment of the Franklin Thesis European Journal of Political Research 41/5: 733"50.

Further reading:

Bowler, Shaun and Todd Donovan (1998), Demanding Choices: Opinion, Voting, and Direct Democracy (Ann Arbor: University of Michigan Press). Overview of the USA experience.

Darcy, R. and Michael Laver (1990), Referendum Dynamics and the Irish Divorce Amendment, Public Opinion Quarterly 54/1: 1-20.

de Vreese, Claes H. (2006), Political Parties in Dire Straits? Consequences of National Referendums for Political Parties, Party Politics 12/5: 581-98.

Lupia, Arthur and Richard Johnston (2001), Are Voters to Blame? Voter Competence and Elite Maneuvers in Referendums, in Matthew Mendelsohn and Andrew Parkin (eds), Referendum Democracy: Citizens, Elites and Deliberation in Referendum Campaigns (Basingstoke: Palgrave). (2001), 191-210.

Marthaler, Sally (2005), The French Referendum on Ratification of the EU Constitutional Treaty, 29 May 2005, Representation 41/3: 230-9.

Michael Gallagher and Pier Vincenzo Uleri (eds), T_he Referendum Experience in Europe._ Basingstoke: Macmillan, 1996.

Week 9: Electoral Behaviour I: The Dependent Variable (MM)

Cees van der Eijk et al, ‘Rethinking the dependent variable in voting behavior: On the measurement and analysis of electoral utilities’, Electoral Studies 2006, 25(3): 424-47

Guy Whitten and Harvey Palmer ‘Heightening Comparativists’ concern for model choice’, American Journal of Political Science, vol 40(1) 1996: 231-60

See also:

Wouter van der Brug et al The Economy and the Vote. Cambridge UnIversity Press, 2008, esp pp. 31-53.

Michael Marsh ‘Stability and change in structure of electoral competition 1989-2002’ in, editor(s)Diane Payne, John Garry and Niamh Hardiman , Irish social and political attitudes, Liverpool, Liverpool University Press, 2006, pp94 – 111, [see http://www.tcd.ie/ines]

Field Seminar B 2009-2010, Weeks 5-6

Wed, 28 Oct 2009 17:20:14 +0000

Week 5 : (30 Oct) Raj Chari: Interest Groups in the EU

Frank Baumgartner and Beth L. Leech. 2001. “Interest Niches and Policy Bandwagons: Patterns of Interest Group Involvement in National Politics.” The Journal of Politics 63(4): 1191-1212.

Burson-Marsteller,2005. The Definitive Guide to Lobbying in European In European Institutions " Based on a survey of the EP, the Council and the Commission (Spring; valuable analysis from more of a practitioner perspective).

David Coen. 1997. ‘The Evolution of the large firm as a political actor in the EU’, JEPP, 4.1 Justin Greenwood. 2007. Interest Representation in the EU 2nd ed. (Palgrave)

Christine Mahoney, 2007, ‘Lobbying Success in the US and the EU’, Journal of Public Policy, Vol 27/1, 44-45

Sonia Mazey and Jeremy Richardson, 2006. ‘Interest Groups and EU Policy Making: Organizational logic and venue-shopping’ in Richardson (ed.) Power and Policy Making in the EU 3rd ed (Routledge)

Silke M. Trommer and Raj Chari. 2006. ‘The Council of Europe: Interest Groups and Ideological Missions?’ West European Politics, Vol. 29, No. 4, 665 " 686, September 2006 (available on my webpage)

Woll C. 2006. ‘Lobbying in the EU: from sui generis to a comparative perspective’ Journal of European Public Policy vol 14. 3, 456-69

Week 6: (6 Nov) Raj Chari: Regulating EU Interest Groups

BertÃ³k, JÃ¡nos. 2008. Lobbyists, Governments and Public Thrust: Building a Legislative Framework for Enhancing Transparency and Accountability in Lobbying. Paris: OECD. http://www.oecd.org/dataoecd/5/41/41074615.pdf

Brinig, Margaret F., Randall G. Holcombe. and Linda Schwartzstein. 1993. “The Regulation of Lobbyists.” Public Choice 77 (2): 377-384.

Chari, Raj, Gary Murphy, and J. W. Hogan. 2007. “Regulating Lobbyists: A Comparative Analysis of the USA, Canada, Germany and the European Union.” The Political Quarterly 78 (3): 422-438.

Dinan, William. 2006. “Learning Lessons? The Registration of Lobbyists at the Scottish Parliament.” Journal of Communication Management 10 (1): 55-66

Greenwood, Justin. 1998. “Regulating Lobbying in the European Union.” Parliamentary Affairs 51(4): 587-599

Hogan, Murphy and Chari, ‘Next Door they have regulation, but not hereâ€¦’, Canadian Political Science Review 2(3), September 2008.

Nolan Committee. 1995. Standards in Public Life, Vol. I, Cm 2850-I HMSO, London, available at: www.archive.official-documents.co.uk/document/cm28/2850/2850.htm

Opheim, Cynthia. 1991. “Explaining the Differences in State Lobby Regulations.” The Western Political Quarterly 44 (2): 405-421

Public Administration Select Committee. 2009. Lobbying: Access and Influence in Whitehall, London: The Stationary Office (available at http://www.publications.parliament.uk/pa/cm200809/cmselect/cmpubadm/36/36i.pdf; along with the Nolan Committee Report above, this will give you a good idea on how the UK has evolved in terms of ideas regarding lobbying laws)

Thomas, Clive S. 1998. “Interest Group Regulation Across the United States: Rationale, Development and Consequences.” Parliamentary Affairs 51 (4): 500-515.

Week 7: (13 Nov) Reading Week (No class)

Field Seminar B, 2009-2010, PhD

Wed, 30 Sep 2009 15:52:21 +0000

The Field Seminar B course handout is available here, although we are still working out the content and some of the topics for later sessions.

The first four weeks look like this:

Week 1: (2 Oct 2009): Ken Benoit " Measuring Policy Positions through Text I

Week 2: (9 Oct) Ken Benoit " Measuring Policy Positions through Text II

Week 3: (16 Oct) Gail McElroy " Legislative Organization: Party Influence

Week 4: (23 Oct) Gail McElroy " Legislative Organization and Behaviour: Beyond Congress

The detailed reading list for these first four weeks is:

Week 1: Measuring Policy Positions through Text I (Kenneth Benoit)

Budge, Ian, Hans-Dieter Klingemann, Andrea Volkens, Judith Bara & Eric Tanenbaum. 2001. Mapping Policy Preferences: Estimates for Parties, Electors, and Governments 1945"1998. Oxford: Oxford University Press. Especially Introduction, Chapters 1-4, Appendix II.

Klingemann, Hans-Dieter, Andrea Volkens, Judith Bara, Ian Budge & Michael McDonald. 2006. Mapping Policy Preferences II: Estimates for Parties, Electors, and Governments in Eastern Europe, European Union and OECD 1990-2003. Oxford: Oxford University Press. Especially Introduction, Chapters 4-6.

Kenneth Benoit, Slava Mikhaylov, and Michael Laver. 2009. “Treating Words as Data with Error: Uncertainty in Text Statements of Policy Positions.” American Journal of Political Science 53(2, April): 495-513.

Exercise:

You must read the manifesto coding instructions, found here (from Appendix II of MMP1 and MMP2).

Code the test manifesto from this document (Appendix II) as an exercise, but only the sections headed WE WILL GROW, PROSPER and ALLIANCES.

Recommended:

Volkens, Andrea. 2001a. Manifesto Research Since 1979. From Reliability to Validity. In Estimating the Policy Positions of Political Actors, ed. Michael Laver. London: Routledge pp. 33"49. ** **

Week 2: Measuring Policy Positions through Text II (Kenneth Benoit)

Jonathan B. Slapin and Sven-Oliver Proksch. 2008. “A Scaling Model for Estimating Time-Series Party Positions from Texts”, American Journal of Political Science 52(3), 705-722.

Laver, Michael, Kenneth Benoit & John Garry. 2003. “Estimating the policy positions of political actors using words as data.”_ American Political Science Review_ 97(2):311"331.

Laver, Michael & John Garry. 2000. “Estimating Policy Positions from Political Texts.“A_merican Journal of Political Science _44(3):619"634.

Recommended:

Budge, Ian and Paul Pennings. 2007. “Do They Work? Validating Computerised Word Frequency estimates against policy series.” Budge and Pennings_. Electoral Studies_ 26: 121-129.

Benoit, Kenneth and Michael Laver. 2007. “Benchmarks for Text Analysis: A Reply to Budge and Pennings."Electoral Studies 26:130"135.

Budge, Ian and Paul Pennings. 2007. “Missing the message and shooting the messenger: Benoit and Laver’s ‘response’.” _Electoral Studies _26: 136-141

See also: http://wordscores.com

Week 3: Legislative Organization: Party Influence (GM)

Gary Cox and Mathew McCubbins. 2005. Setting the Agenda. Cambridge: CUP. Part I, Chapters 3,4,7,10&11.

Keith Krehbiel. 1993. “Where’s the Party?"British Journal of Political Science. 23: 235-266

Lawrence ED, Maltzman F, Smith SS. 2006. Who wins? Party effects in legislative voting _Legislative Studies Quarterly _31 (1) : 33-69

Aldrich, John H. and David W. Rohde. 2001. “The Logic of Conditional Party Government: Revisiting the Electoral Connection.” In Lawrence C. Dodd and Bruce I. Oppenheimer (eds.), Congress Reconsidered 7th ed. Washington, D.C.: CQ Press: 269-292.

Suggested Reading:

Krehbiel, Keith. 1998. Pivotal Politics: A Theory of U.S. Lawmaking. Chicago: University of Chicago Press.

Edward H. Stiglitz and Barry R. Weingast, 2009, “Agenda Control in Congress: Evidence from Cutpoint Estimates and Ideal Point Uncertainty.” Manuscript. Available at http://polisci.stanford.edu/faculty/weingast/

Mayhew, David R. 1974. Congress: The Electoral Connection. New Haven: Yale University Press.

Week 4: Legislative Organization and Behaviour: Beyond Congress (GM)

Simon Hix, Abdul G. Noury and GÃ©rard Roland. Democratic politics in the European Parliament. Cambridge : Cambridge University Press, 2007.

McElroy, Gail and Kenneth Benoit. 2009. “Party Policy and Group Affiliation in the European Parliament.” . Forthcoming BJPS.

Yoshinaka, Antoine, Gail McElroy and Shaun Bowler. Manuscript. Rapporteurs in the European Parliament: Partisans or Honest Brokers?

Stephen A. Meserve , Daniel Pemstein and William T. Bernhard. 2009. Political Ambition and Legislative Behavior in the European Parliament.The Journal of Politics, Vol. 71(3): 1015"1032

Suggested Reading:

McElroy, Gail. 2007. “Legislative Politics as Normal?: Voting Behaviour and Beyond in the European Parliament”. European Union Politics

Update on unlocking iPhones

Wed, 09 Sep 2009 06:55:56 +0000

The best way to unlock modern iPhones using the most recent Apple firmwares is to use RedSn0w. Here is a guide for unlocking the 2G iPhones (such as mine). If you want to unlock a 3G or 3GS, you need to jailbreak the phone first using RedSn0w, and then unlock it using UltraSn0w, the unlocking tool for iPhone 3G.

The general website for unlocking iPhones, from the iPhone Dev-Team, can be found at their blog.

Political Science Impact Ranking to be Published

Fri, 04 Sep 2009 14:29:48 +0000

The paper I wrote last year with Michael Marsh, “A Relative Impact Ranking of Political Studies in Ireland will be published in the 3rd issue of Economic and Social Review. A corrected manuscript can be downloaded by clicking on the title.

Abstract

Against a background of the Irish government’s concerns with Key Performance Indicators (KPIs) and the British government’s wishes for a more quantitative Research Assessment Exercise (RAE), our study conducts a relative impact assessment of the study of politics, government, political science, and international relations in Ireland. Impact is measured as citations from the publications of permanent staff in eight Irish politics departments, based on data compiled in April 2008 from two leading academic indexes “ISI’s Web of Science and Scopus " as well as the now popular Google Scholar. We discuss some of the criticisms that naturally arise in a study of this nature. Then, following similar exercises in other disciplines (e.g. economics), we use the impact measures to compare and rank individual scholars as well as departments. We also explore the extent to which the choice of different indexes, and different measures, influences the results that we obtain. While there are differences, in particular between indexes based purely on articles and those that access books and other material, the results from the different indexes are strongly correlated.

CMP error dataset page now up

Fri, 28 Aug 2009 11:51:08 +0000

I’ve just posted, somewhat late, the full replication materials and error dataset from:

Kenneth Benoit, Slava Mikhaylov, and Michael Laver. 2009. “Treating Words as Data with Error: Uncertainty in Text Statements of Policy Positions.” American Journal of Political Science 53(2, April): 495-513.

The simplest method is to download both the error dataset and the replication code (in R), including the datasets from Adams et al (2006) and from Hix et al (2006), in a single zip file.

Full instructions as well as links to all of the files are available from the Dataverse page. Note that in the published article, the web page is listed as http://www.politics.tcd.ie/cmp/. This link works but simply redirects your browser to the page on this site.

The power of simulation

Mon, 24 Aug 2009 09:57:13 +0000

This is very close to what we do in statistical sciences!

Success through simulation

On Causality

Tue, 02 Jun 2009 13:23:13 +0000

And this from the Matrix Reloaded (the best scene and dialogue of the film by far):

Merovingian: You see, there is only one constant, one universal, it is the only real truth: causality. Action. Reaction. Cause and effect.

Morpheus: Everything begins with choice.

Merovingian: No. Wrong. Choice is an illusion, created between those with power, and those without. Look there, at that woman. My God, just look at her. Affecting everyone around her, so obvious, so bourgeois, so boring. But wait… Watch – you see, I have sent her dessert, a very special dessert. I wrote it myself. It starts so simply, each line of the program creating a new effect, just like poetry. First, a rush… heat… her heart flutters. You can see it, Neo, yes? She does not understand why – is it the wine? No. What is it then, what is the reason? And soon it does not matter, soon the why and the reason are gone, and all that matters is the feeling itself. This is the nature of the universe. We struggle against it, we fight to deny it, but it is of course pretense, it is a lie. Beneath our poised appearance, the truth is we are completely out of control. Causality. There is no escape from it, we are forever slaves to it. Our only hope, our only peace is to understand it, to understand the `why.’ `Why’ is what separates us from them, you from me. `Why’ is the only real social power, without it you are powerless.

How to set proxy settings for R (Mac OSX)

Wed, 14 Jan 2009 17:51:40 +0000

I run R behind a firewall, and found it tricky to set the proxy settings for R so that I could directly install packages, access outside data using load(url()), etc. The solution lies in the http_proxy environment variable. Here is how to set it up:

Create a text file called .Rprofile in your home directory (on Mac OSX, this is ~ or /Users/myuser/). This text file should have the following line in it:

Sys.setenv(http_proxy="http://username:password@tcdproxy.tcd.ie:8080-8243")

substituting your own username, password, and proxy server and port settings of course.

It seems that running this from R before you need it does not work, since I have heard reports that it only works if you run it as the first command in R when you start a new R session. By setting this up in your ~/.Rprofile you avoid having to retype it each time you start R, as well as making sure that this command is run at startup.

PhD Scholarships in Political Science at Trinity College Dublin

Thu, 08 Jan 2009 16:25:54 +0000

Fully-funded studentships are available for study in the PhD programme in Political Science at Trinity College Dublin, paying tuition plus a stipend of up to €15,000 per year.

Application for this financial support is automatically considered as part of the normal admissions process. The Department of Political Science normally admits six students per year, and typically provides financial support in the form of tuition grants, stipends, teaching and research assistantships, and support for postgraduate conference and research travel.

Our four-year doctoral programme is designed to give students the highest level of training to become scholars at the forefront of the discipline. Most of our Ph.D. graduates have gone on to careers in research and teaching at established universities, while others have gone on to careers in government and business.

The Department of Political Science at Trinity ranks among the best political science departments in the world according to independent evaluations of its research and teaching (e.g. in the global ranking of political science departments conducted by Professor Simon Hix at the London School of Economics). Ph.D. researchers also benefit from the Department’s strong links with Institute for International Integration Studies at Trinity College Dublin. This Institute brings together researchers from different disciplines to study international integration. In addition, the Department is involved in the Dublin School of Social Sciences, which provides graduate training in collaboration with University College Dublin.

The doctoral programme is designed to impart substantive knowledge about politics, and to train students in the techniques, including quantitative methods, necessary to recognize, design and implement cutting-edge research. The writing of a dissertation of original research follows a rigorous training in how to design and conduct research.

We seek outstanding and motivated graduates for entry to our doctoral programme in September 2009. Applicants are particularly welcome from students who wish to conduct research in areas in which the Department has particular research strengths: comparative European politics, European Union politics and policies, European law, American politics (especially legislative and judicial politics), Irish politics, African politics, political economy, political institutions, public policies, electoral systems, political parties, legislative politics, public opinion and electoral behaviour, international law, international organisation, conflict in developing countries, political geography, and political theory. The Department also invites applications from students whose research develops innovative quantitative methods, such as the quantitative analysis of political texts, as this is another strong area of research among staff.

Prospective students are encouraged to consult the webpages of members of staff (http://www.tcd.ie/Political_Science/), and to email individual staff members who could be potential supervisors for their Ph.D. dissertations. General enquiries may be directed to the Director of Graduate Studies, Dr. William Phelan (). Further details of the Ph.D. programme can be found at http://www.tcd.ie/Political_Science/postgraduate/phdindex.php

Applications for entry in September 2009 are now being taken, with a closing deadline of April 1st 2009. Students who require an early decision on their entry to the doctoral programme are encouraged to contact the Director of Graduate Studies.

Data Analysts Captivated by R's Power

Thu, 08 Jan 2009 11:36:43 +0000

Here’s a recent article published in the New York Times about R, the freely available statistical package:

“Data Analysts Captivated by R’s Power”

By ASHLEE VANCE

Published: January 7, 2009

A programming language’s versatility and its price (free) makes it a favorite of data analysts. http://www.nytimes.com/2009/01/07/technology/business-computing/07program.html?partner=permalink&exprod=permalink

Data Analysts Captivated by R’s Power

Thu, 08 Jan 2009 11:36:43 +0000

Here’s a recent article published in the New York Times about R, the freely available statistical package:

“Data Analysts Captivated by R’s Power”

By ASHLEE VANCE

Published: January 7, 2009

A programming language’s versatility and its price (free) makes it a favorite of data analysts. http://www.nytimes.com/2009/01/07/technology/business-computing/07program.html?partner=permalink&exprod=permalink

Social scientists use real data

Wed, 07 Jan 2009 16:30:50 +0000

Here’s a good one from http://www.phdcomics.com for all you PhD students:

Spring 2009 course information now posted

Wed, 17 Dec 2008 14:55:53 +0000

I have now added the detailed course outlines to my web page for the courses I am teaching in Spring 2009. These are:

Introduction to Quantitative Research Methods (MSc)

Quantitative Methods II (PhD)

MSc Modules on Electoral, Party Systems

Wed, 03 Dec 2008 15:32:25 +0000

As promised I have now posted the presentations from the weeks in Government Institutions. These are in pdf format so should be readable by everyone. Note that you must be on the local TCD network (or using the VPN or proxy server) to access these files.

Week 5 Electoral Systems:

My overview

Electoral systems and party systems

Electoral systems and voter participation

Electoral systems and womens’ representation

Week 7 Political parties and party systems:

My overview

Party systems

Party families

TCD political science on a par with "best of Britain";

Wed, 15 Oct 2008 13:27:12 +0000

Michael Marsh and I have just finalized our comparative assessment of the research output an impact of political studies scholars in Ireland. This exercise compares citation measures for all permanent, full-time academic staff in eight Irish politics departments (in the Republic and in Northern Ireland). That paper can be found here and earlier discussions during the drafting of this paper feature elsewhere on this blog.

As a benchmark we also compared the Irish departments to the Department of Government at the University of Essex. As Essex is considered by many to be the best politics department in Britain and received the maximum score of 6* in the last RAE, using this department for a reference is setting the bar quite high. In addition, with 25 full-time, permanent staff, Essex’s department is also larger than all of the Irish departments in our study except Queen’s.

We found in our analysis that in per capita terms, Essex was ahead of TCD, but just barely, with 103.5% of TCD’s score. Its score is also higher on articles, but again not by much, at 117.8%. On Google Scholar-based measures, TCD is actually higher per capita than Essex: almost 2 Google cites/year higher than Essex. TCD also had a higher Google citation score than Essex when these figures were produced by divided by the total years of all staff members by the sum of all of their Google citations. Finally, TCD’s total per capita Google citations were also higher by about 22 citations per capita.

Boxplot of overall impact by department, measured as total Google citations

Differences between departmental profiles can be summarized graphically using box plots such as the one above. Each box shows the inter-quartile range of the values for each department, along with a bar for the median value and additional markings outside the box showing some of the extreme values. For comparison, we have also plotted the distribution of Essex’s department of government. While Essex’s median total citations is higher than any Irish department’s, its interquartile range is slightly below TCD’s. The above figure plots this data for the total Google Scholar citations, a figure that takes into account every type of publication. Departments are plotting in increasing order, and the counts shown on the y-axis are logged to reduce skew. In comparison with Essex, TCD has practically the same median value on productivity, with a slightly larger range.

Boxplot of productivity by department, measured as Google Citations/year

Different departments not only have different numbers of staff, they also have very different age profiles. Counting total citations clearly advantages departments with senior staff whose longer careers have provided more opportunity to accumulate citations. In the second box plot, we portray the profiles of departments according to productivity, on a linear scale, according to the Google Scholar citations per year. Essex, shown in gray, has a median that is all but indistinguishable from Trinity’s.

TCD political science on a par with "best of Britain";

Wed, 15 Oct 2008 13:27:12 +0000

Michael Marsh and I have just finalized our comparative assessment of the research output an impact of political studies scholars in Ireland. This exercise compares citation measures for all permanent, full-time academic staff in eight Irish politics departments (in the Republic and in Northern Ireland). That paper can be found here and earlier discussions during the drafting of this paper feature elsewhere on this blog.

As a benchmark we also compared the Irish departments to the Department of Government at the University of Essex. As Essex is considered by many to be the best politics department in Britain and received the maximum score of 6* in the last RAE, using this department for a reference is setting the bar quite high. In addition, with 25 full-time, permanent staff, Essex’s department is also larger than all of the Irish departments in our study except Queen’s.

We found in our analysis that in per capita terms, Essex was ahead of TCD, but just barely, with 103.5% of TCD’s score. Its score is also higher on articles, but again not by much, at 117.8%. On Google Scholar-based measures, TCD is actually higher per capita than Essex: almost 2 Google cites/year higher than Essex. TCD also had a higher Google citation score than Essex when these figures were produced by divided by the total years of all staff members by the sum of all of their Google citations. Finally, TCD’s total per capita Google citations were also higher by about 22 citations per capita.

Boxplot of overall impact by department, measured as total Google citations

Differences between departmental profiles can be summarized graphically using box plots such as the one above. Each box shows the inter-quartile range of the values for each department, along with a bar for the median value and additional markings outside the box showing some of the extreme values. For comparison, we have also plotted the distribution of Essex’s department of government. While Essex’s median total citations is higher than any Irish department’s, its interquartile range is slightly below TCD’s. The above figure plots this data for the total Google Scholar citations, a figure that takes into account every type of publication. Departments are plotting in increasing order, and the counts shown on the y-axis are logged to reduce skew. In comparison with Essex, TCD has practically the same median value on productivity, with a slightly larger range.

Boxplot of productivity by department, measured as Google Citations/year

Different departments not only have different numbers of staff, they also have very different age profiles. Counting total citations clearly advantages departments with senior staff whose longer careers have provided more opportunity to accumulate citations. In the second box plot, we portray the profiles of departments according to productivity, on a linear scale, according to the Google Scholar citations per year. Essex, shown in gray, has a median that is all but indistinguishable from Trinity’s.

Validation of Google Scholar as an Impact Measure for Political Science

Sun, 12 Oct 2008 20:40:06 +0000

One interesting side finding of the citations data from my paper with Michael Marsh was the cross-validation, throuch comparison with other measures, of the results from Google Scholar compared to other bibliographic indexes. Facing the criticism that we ignored books by looking only at ISI and Scopus (article-based) indexes, Michael and I collected data from Google Scholar – which includes books and book chapters – and also collected separate data on book citations from the “Cited Reference Search” feature of the SSCI from ISI/Web of Science. Comparing this to Google Scholar, it seems that Google correlates the best with the total Cited Reference Search data. It also correlated better with all other indexes in the overall pair-wise correlation contest.

Inter-measure Correlations of Citation Counts

This figure shows the relationships between five different sources, taking in each case total citations: Google scholar, ISI, Scopus, ISI-CRS (books only) and ISI-CRS (all material). ISI has been the most widely used source to date despite concerns about its coverage. We can see that its results correlate quite well with those form Scopus (0.73), and also with Google Scholar (0.69) but less well with the full ISI-CRS where the correlation is only 0.52. Most strikingly, it correlates at only 0.25 with the books only citations from ISI-CRS. Scopus provides pretty similar results, correlating modestly with ISI-CRS (total cites) and more strongly with Google Scholar, but also very poorly with the book only measure from ISI-CRS. It is very clear that measures based just on books will give different results to those based just on articles, although article based measures do correlate more highly with total measures than do book only measures, suggesting that a majority of citations are to articles. The two most inclusive measures, Google Scholar and the full ISI-CRS, provide the highest correlation in the whole matrix, 0.84. In general the analysis here seems to “confirm"the worth of Google scholar. Not only does it correlate more highly with ISI and Scopus, but it also matches well with the total CRS citations (books and articles), suggesting that Google provides a valid aggregate measure that includes both books and articles.

PO7003 Field Seminar A in Political Science – 2008-09

Thu, 25 Sep 2008 11:37:24 +0000

Note: The latest version of the course handout is available from http://www.tcd.ie/Political_Science/postgraduate/phdtaught.php. That version lacks the hyperlinks to the readings below, however.

The meeting time and place is Friday 10-13:00, College Green 4, starting 3 October 2008. In practice the actual length of the Seminar will depend on the instructor, and typically will run from 10-12:00 only. While I am the coordinator for the course, the weekly seminars are led by a rotating set of staff from Political Science

Week 1 – 3 Oct 2008 – Ken Benoit - Policy Positions and Policy Spaces

Kenneth Benoit and Michael Laver. 2006. Party Policy in Modern Democracies. London: Routledge, 2006. Wiesehomeier, Nina and Kenneth Benoit.“Presidents, Parties And Policy Competition.“University of Konstanz and Trinity College Dublin manuscript. Version: July 7, 2008.

Week 2 – 10 Oct – Ken Benoit - Electoral System Origins

Kenneth Benoit. 2007. “Electoral Laws as Political Consequences: Explaining the Origins and Change of Electoral Institutions."Annual Review of Political Science 10: 363-90.

Boix, C. 1999. “Setting the rules of the game: The choice of electoral systems in advanced democracies."American Political Science Review 93 (3): 609-624.

Andrews, Josephine and Robert Jackman. 2005. “Strategic Fools: Electoral rule choice under extreme uncertainty."Electoral Studies 24: 65-84.

Colomer, Josep M. (2004). “The Strategy and History of Electoral System Choice.” In Josep M. Colomer (Ed.), Handbook of Electoral System Choice, pp. 3"78. New York: Palgrave Macmillan. (photocopied)

Birch, Sarah, Frances Millard, Marina Popescu, and Kieran Williams (2002). Embodying Democracy: Electoral System Design in Post-Communist Europe. Houndmills, Basingstoke: Palgrave Macmillan. Ch 1. (photocopied)

Benoit, Kenneth and Jacqueline Hayden. “Institutional Change and Persistence: The Evolution of Poland’s Electoral System 1989-2001."_Journal of Politics _66 (May, 2): 396-427.

Week 3 – 17 Oct – Ken Benoit - Electoral System Consequences

Gallagher, Michael and Paul Mitchell. 2005. “Introduction to Electoral Systems.“In Michael Gallagher and Paul Mitchell, eds., The Politics of Electoral Systems. Oxford: Oxford University Press. pp3-24.

Cox, Gary W. 1997. Making votes count: strategic coordination in the world’s electoral systems, Political economy of institutions and decisions. Cambridge, U.K. ; New York: Cambridge University Press.

Duverger, Maurice. 1959. Political parties: Their organization and activity in the modern state. Second English Revised ed. London: Methuen & Co. Book II, Chapter 1, “The Number of Parties.“Also see this related reading.

Duverger, Maurice. 1986. Duverger’s Law: Forty Years Later. In Electoral Laws and Their Political Consequences, edited by B. Grofman and A. Lijphart. New York: Agathon Press.

Gallagher, Michael. 1991. “Proportionality, disproportionality, and electoral systems.” Electoral Studies 10 (1): 33-51.

Ordeshook, Peter, and Olga Shvetsova. 1994. “Ethnic heterogeneity, district magnitude, and the number of parties.” American Journal of Political Science 38 (1): 100-123.

Who should be included in and who should author scholarly impact assessments?

Fri, 30 May 2008 09:34:44 +0000

"Everyone heading for the top / But tell me how far is it from the bottom / Everybody want to go to heaven / But nobody want to die. “ Peter Tosh

Following the initiation of my research project on citations with Michael Marsh, a lively debate has taken place over whether and how scholarly impact should be measured and compared in the study of politics in Ireland. Some of this debate is visible on this website; some of it has occurred in private communications between me, Michael, and many individual scholars; some of it has occurred in meetings of research committees at different universities; other debates have taken place at meetings of the PSAI executive committee and most recently, at a meeting of heads of politics departments in Ireland (May 23, 2008). What we would like to do here is to address a few main questions concerning the exercise.

Q: Is it the prerogative of an individual scholar to “opt out” of the exercise?

A: No. Our data comes from publicly available sources on public_ations, also by definition (and etymology) public, and anyone has the right to analyze these.

First, the research was conducted on the basis of publicly available data, and anyone has the right to analyze these data. Just as no TD could demand to be omitted from a study of legislative activity based on openly published data, individual scholars employed at public universities in Ireland have no basis on which to demand to be omitted from the exercise. We might add here that, much like other public servants, the scholars included in our exercise are all permanent employees with very secure employment, in a higher-education sector that is among the best-paid and best-pensioned in the world. Second, as a profile of research impact in Ireland, the analysis would be invalidated as a comparison of scholarship if those included were selected according to their levels of impact. (Since those with the lowest measured impact have the greatest incentive to be omitted from the exercise, this is exactly what we suspect would happen.) So unless we held a purely randomly drawn lottery to remove individual scholars, allowing (self-)selected opt-outs would invalidate our exercise.

Q: Is it the prerogative of a department to “opt out” of the exercise?

A: No, in the same way that individual permission is not required. Furthermore, we feel there is a significant first-mover advantage to starting this (inevitable) sort of exercise within our own ranks, rather than waiting for it to be imposed by external forces (such as the HEA).

For the same reasons that apply to individuals, our use of the data on research impact from public sources is open for analysis by anyone. Our analysis of scholarly impact at the level of an academic department is simply an aggregation of that data based on publicly available information - usually the department’s own web site listing the permanent members of its staff.

Furthermore, we feel that assessing the scholarly impact of the study of politics in Ireland is a positive step toward highlighting its success, and that aggregating to departments provides a natural basis for this profile to be developed. Selectively excluding departments would undermine the goal of promoting Irish scholarly activities by limiting the scope of the exercise.

Whether departments like it or not, some form of impact assessment seems inevitable in Ireland, and by showing that we can rank ourselves, as well as conduct a lively and healthy debate on how assessment should be performed, from within our ranks, we have a better chance to be on top of this process than having it be thrust on top of us. Citation-based measures are certainly not the only method of assessing impact, but by putting forth bibliometric measures as one method – as does our paper – we have started the debate and focused attention on the strengths and limits of this particular method.

We would also add that numerous comparative bibliometric exercises that are inclusive of all departments in a particular context have already been written or published in other contexts:

Dale, Tony and Shaun Goldfinch. 2005. “Article Citation Rates and Productivity of Australasian Political Science Units 1995-2002.” Australian Journal of Political Science 40(3, Sept): 425-434.

Hix, Simon. 2004. “A global ranking of political science departments.” Political Studies Review 2(3): 293-313.

As well as in Ireland specifically (including economics):

Elgie, Robert and Iain Mcmenamin. 2008. “Journal Publications From Politics Departments In Ireland 2003-2007: An Update Using The Hix Method.” Dublin City University Manuscript.

Rouane, Frances and Richard S. J. Tol. 2007. “Centres of Research Excellence in Economics in the Republic of Ireland.” Economic and Social Review 38(3, Winter): 289-322.

Coup, Tom and Patrick Paul Walsh. 2003. “Quality Based Rankings of Irish Economists 1990-2000.” Economic and Social Review 34(2, Summer/Autumn): 145-149.

Q: Should our review exclude departments outside the Republic of Ireland?

A: This is solely a question of whether including such departments improves our paper, not a political question whereby the limits of the Irish national university system prevent us from comparing scholars or departments from other national university systems.

In fact, we feel that there are sound reasons to wish to compare Irish politics departments to departments in other contexts, since this provides an external benchmark for the scholarly performance of Irish politics departments.

Previous studies (e.g. Hix 2004, Dale and Goldfinch 2005) have compared department from different national systems, and other rankings of insttutions and deparmtents (e.g. the Times Hugher Education Supplement) regularly perform international comparisons. Our study would hardly be unique in this regard.

A sensible question arises with regard to whether any ranking of individual political scientists is meaningful in a selected international context, and here we would tend to agree that either the context should be restricted to a meaningful boundary for inclusion, or broadened to a Europe-wide or fully international focus.

Q: Does the fact that the authors appear at the top of the rankings in their own paper not impugn the credibility of the exercise?

A: It should not, because the rankings are based on completely open, publicly available criteria, and not created, manipulated, transformed, or selectively applied by the authors.

In other words, any other scholar would be able to reproduce these results using the methods which we have carefully and explicitly detailed in our draft paper, and also publicly listed in order that others might also verify that the authors’ citation rates are what they claim. We welcome any scrutiny in this regard.

Turned around, would the report’s conclusions be more credible if the two scholars at the bottom of the rankings had undertaken this exericise? Putting aside the fact that two scholars with zero citations on any ranking would never undertake such an exercise, as long as the method is public and can be replicated, the identity of the authors or position of the authors in the ranking should not be related to the conclusions drawn. In other words, the authors, whether from the bottom, top, or a randomly chosen position in the rankings, should reach exactly the same conclusions.

We do agree however that it would be preferable from many standpoints - certainly from that of not attracting the ire of the bulk of the Irish political science community! - had an independent, third-party commission undertaken this review. But as far as we know, no such commission has been awarded or plans to be tendered for a stidy such as ours any time soon. This is no doubt why political scientists in Australasia and economists in Ireland and the Netherlands (see citations above) have conducted such studies themselves, just as we have.

Ranking of Politics Scholarship in Ireland

Wed, 09 Apr 2008 15:07:43 +0000

Updated: 4 June 2008

The preliminary version of a paper that profiles departments and ranks the top scholars working in Ireland on politics, government, and international relations according to scholarly impact (measured by citations) has now been posted for comments at http://www.tcd.ie/Political_Science/staff/kenneth_benoit/citations/.

The current version of the paper is dated April 18, 2008.

Here is a Dilbert cartoon that offering one apparently common view that may apply to bibliometrics:

There has also been some interesting correspondence regarding our paper. On April 24, 2008 we received this official letter from Professor Jennifer Todd, Chair of the Research and Innovation Committee from UCD’s School of Politics and International Relations. (This letter was also publicly sent to the Political Studies Association of Ireland mailing list.) Our reply was also posted to the PSAI mailing list:

Dear Jennifer,

Thank you for your letter dated 25 April 2008 about the data collection stage of our planned paper, "A Relative Impact Assessment of the Academic Study of Politics In Ireland". As we originally stated quite clearly in our draft paper and on our website, we had posted the data we collected not because we intended for this data to form our final analysis, but rather in the interests of full accountability to those involved, especially for the purpose of catching errors from those most familiar with their own citation data. Having served this purpose, the tables you refer to were removed as planned on April 30, 2008.

As for your offer of assistance, we would be delighted were the PSAI to appoint a subcommittee to assist with additional data collection for our paper. While our present concern is with assessing research impact in early-to-mid 2008, furthermore, we also welcome your suggestion that the bibliometric information we have been gathered be updated on a regular basis. We are ready to cooperate immediately on such a venture, but just in case the PSAI machinery moves too slowly, we have also undertaken independently to gather the additional data you and the SPIRe Research and Innovation Committee suggested would be vital to our analysis. Here we refer to the citations from ISI's "Cited Reference Search."Adding this additional data will mean that our exercise examines four measures in total: ISI's article citations, ISI's more inclusive (but less quality-controlled) "cited reference search,"Scopus's article citations, and Google Scholar's results which tend to include just about everything. If you can think of any more empirically observable, third-party measures of scholarly impact by which scholarly impact could also be compared for politics scholars in Ireland, we would be glad to consider including those as well. The response attached to your letter from the SPIRe Research & Innovation Committee raises a number of other interesting points on which reasonable people might simply disagree, and we think it unlikely that any subcommittee or other formal collective forum will be sufficient to resolve this disagreement. In particular we think it unlikely that any objective, comparative assessment of scholarship can avoid the purported "bias [from] a culture of extensive mutual citation, rather than reliance on original data and sources."The focus of our analysis is on citations of published scholarly work, and this is unlikely to change. Of course, science and scholarship more generally tend to make advances in a cumulative, often adversarial fashion. Just as we are free to compile data from publicly available sources such as ISI, Scopus, and Google Scholar, you and your colleagues are free to offer rejoinders to our paper or present alternative comparative assessments of scholarly impact. Indeed, we look forward to the healthy debate focused on research impact that would result from such an exchange.

Sincerely yours,

Kenneth Benoit Michael Marsh Department of Political Science, Trinity College

On June 4, President of the PSAI (Political Studies Association of Ireland) sent the following e-mail to all members of the PSAI mailing list:

The Heads of Departments of politics in Ireland met in Dublin recently to discuss matters of common interest to the profession. These included CSPE in the leaving cert and the impact of Bologna on politics degrees. The most contentious issue, however, was that of research measurement highlighted by recent research by colleagues in TCD. It was recognised that politics departments compete less with each other than with other disciplines in their internal university battles for resources. Metrics may be an inevitable by product of changing funding structures. Nevertheless, the Heads felt it would be useful if the PSAI could facilitate a consensus on the range of appropriate measures. It was agreed, therefore, that the Association would convene a meeting of nominees from each department to share views on what metrics might be appropriate for use in combination or singularly within each institution. The PSAI may also invite members with a particular expertise in the issue of metrics. The proposed date of the meeting is 18th June.
Neil Collins

How to get an iPhone working on Meteor Ireland

Sat, 05 Apr 2008 21:12:32 +0000

This is my experience with getting an iPhone working with Ireland’s Meteor network. Currently the iPhone is only available from O2, and only with an 18-month service contract. I have been using a simple, pay-as-you-go plan from Meteor and wanted to use that in an iPhone. Here are the steps I took to get it working.

Buy an iPhone. Not so easy to do in Ireland, but you can find them on eBay, buy them from an Apple Store (in the US or say, London), or talk your way into getting an AT&T store to selling you one without requiring you to sign the activation contract in the store. (Tell them it’s for a friend who will activate it Apple-Store-style using iTunes from home.)

Jailbreak and unlock the iPhone. This used to be complicated but now it is ultra-easy thanks to the excellent utility ZiPhone. My phone came with version 1.1.4 of the iPhone software, and I used ZiPhone 3.0 (30th of March 2008). You simply connect your iPhone to your computer with the included USB cable, turn off iTunes once it starts up, and run ZiPhone. (I was using the Mac version, but the Windows version is supposed to work similarly.) I chose the first option, “Unlock, Jailbreak, and Activate”. You get a message telling you to wait 4 mins 30 secs, a bunch of stuff scrolls across your iPhone, it restarts a couple of times, and at the end of the process, you have an activated and unlocked iPhone. ZiPhone also automatically installs a piece of software called Installer.app (3.01 on mine, from Nullriver Software), plus the BSD Subsystem and OpenSSH.

Remove the AT&T SIM card included with the iPhone and replace it with your own SIM card. There is a pin hole you need to push on the top of the phone with a paper clip, and the SIM tray slides out. You simply replace the existing card with your own, and then reinsert. There are plenty of videos showing how to do this on YouTube, for example this one. You can then toss out the AT&T card or keep it for a souvenir. Your iPhone should be working fine with Meteor now. It may ask you to enter the PIN for your SIM card, if you have such a security feature enabled. Just like with any other mobile phone, you need to enter this every time you turn on the phone, although normally the iPhone is just asleep, not turned off. If you want to turn it off you have to hold down the sleep/power button for about 5 seconds and then confirm that you wish to power off.

Get HTTP working through GPRS. This is the step to get everything except Mail working — everything that uses http, that is: Safari, Maps, Stocks, and Weather. I found that YouTube did not work this way, nor did the iTunes, but I think that is because both work only through WiFi. Probably the 3G version will fix this… In the meantime, here are the steps you need. These are specific to Meteor Ireland.

Sign up for a data plan. It is possible to enter the wap settings for EDGE and to use the Meteor proxy server (following similar instructions to the ones from a similar page explaining how to do this with Vodafone Ireland), but this method is punitively expensive! I signed up for Meteor’s most basic plan at â‚¬20/month and paid for a â‚¬9.99 data add-on that gives me 250MB per month.

Phone Customer Care at 1905 and make sure they turn on the isp networking for you. This is the trickiest part since the Customer Care person may think you are talking about wap, which you are not, and since they must request to a technical support person that they activate this on their system for your account. Evidently by default it is not turned on. I had to make about five calls to Customer Care before Meteor got this right.

Enter the EDGE settings for Meteor. These are:

APN: isp.mymeteor.ie

Username: my

Password: meteor

Yes you do need to turn your phone off and turn it back on. Before you do this, be sure to deactivate Wi-Fi if you have been using that, since otherwise you won’t be able to test the EDGE/GPRS function. Once you do that it should work!

Things that still do not work:

Voicemail. Visual voicemail will probably never work, but there are ways to hack the phone so that the voicemail button calls Meteor voicemail. I am working on this too.

Readings for March 7 – Campaign Finance

Wed, 05 Mar 2008 14:08:23 +0000

March 7 – Campaign Finance and the Effects of Spending

Required:

Jacobson, Gary C. 1978. “The Effects of Campaign Spending in Congressional Elections."American Political Science Review 72 (June): 469-91.

Kenneth Benoit and Michael Marsh. “A Fistful Of Euros: Incumbent And Challenger Campaign Spending Effects Under The Single-Transferable Vote Electoral System.” Trinity College manuscript.

Stratmann, Thomas. 2005. “Some talk: Money in politics. A (partial) review of the literature."Public Choice 124:135-56.

Recommended:

Kenneth Benoit and Michael Marsh. “The Campaign Value of Incumbency: A New Solution to the Puzzle of Less Effective Incumbent Spending.“Trinity College manuscript.

Moon, Woojin. 2006. “The Paradox of Less Effective Incumbent Spending: Theory and Tests."British Journal of Political Science 36: 705"721.

Readings for Field Sem. B – Autumn 2007

Mon, 28 Jan 2008 12:19:21 +0000

Week 2: Electoral System Origins

Colomer, Josep M. (2004). “The Strategy and History of Electoral System Choice”. In Josep M. Colomer (Ed.), Handbook of Electoral System Choice, pp. 3-78. New York: Palgrave Macmillan. (photocopied)

Birch, Sarah, Frances Millard, Marina Popescu, and Kieran Williams (2002). Embodying Democracy: Electoral System Design in Post-Communist Europe. Houndmills, Basingstoke: Palgrave Macmillan. Ch 1. (photocopied)

Week 3: Electoral System Consequences

Gallagher, Michael and Paul Mitchell. 2005. “Introduction to Electoral Systems.” In Michael Gallagher and Paul Mitchell, eds., The Politics of Electoral Systems. Oxford: Oxford University Press. pp3-24. (photocopied)

Cox, Gary W. 1997. Making votes count: strategic coordination in the world’s electoral systems. Cambridge: Cambridge University Press.

Duverger, Maurice. 1959. Political parties: Their organization and activity in the modern state. Second English Revised ed. London: Methuen & Co. Book II, Chapter 1, “The Number of Parties.” (photocopied)

Duverger, Maurice. 1986. “Duverger’s Law: Forty Years Later”. In Electoral Laws and Their Political Consequences, edited by B. Grofman and A. Lijphart. New York: Agathon Press. (photocopied)

Gallagher, Michael. 1991. “Proportionality, disproportionality, and electoral systems.” Electoral Studies 10(1): 33-51. (photocopied)

Ordeshook, Peter, and Olga Shvetsova. 1994. “Ethnic heterogeneity, district magnitude, and the number of parties.” American Journal of Political Science 38(1): 100-123. (JSTOR)

Course-related discussions

Thu, 24 Jan 2008 23:56:29 +0000

This will be a place for posting announcements (and comments) related to courses I teach.

New web page

Fri, 23 Nov 2007 21:17:07 +0000

I have now completed moving my old web page to the new server, and it now has a completely new look, feel, and function as well. The use of WordPress allows posts and comments, which I welcome on any topic. Since I am still getting used to the new format myself, I will try to improve the new page over the next several weeks.

Any feedback regarding mistakes or broken links is especially welcome.

Posts | Kenneth Benoit

Quantitative Text Analysis (TCD 2018)

Overview

Objectives

Detailed Course Schedule

Required Reading

Recommended Reading

Resources

Required Reading

Recommended Reading

Resources

Required Reading

Recommended Reading

Resources

Resources

Required Reading

Recommended Reading

Required Reading

Recommended Reading

Resources

Final Assignment

Project Guidelines

Content

Formatting

Deadline

References

Quantitative Text Analysis (TCD 2016)

Meeting times

Short Outline

Objectives

Detailed Course Schedule

Required Reading:

Recommended Reading:

Required Reading:

Recommended Reading:

Required Reading:

Recommended Reading:

Required Reading:

Recommended Reading:

Final Assignment

Project Guidelines

Content

Formatting

Deadline

Data Mining and Statistical Learning

How to install the R package topicmodels on OS X

The Quantitative Analysis of Textual Data (NYU Fall 2014)

How to batch convert pdf files to text

Quantitative Text Analysis 2E, Essex 2014

Quantitative Text Analysis (TCD)

Short Outline

Objectives

Prerequisites

Detailed Outline

Meetings

Computer Software

Grading

Recommended Texts

Detailed Course Schedule

References

New Directions in Analyzing Text as Data Workshop (2013)

FRIDAY

BREAKFAST, COFFEE AND REGISTRATION (Friday, 8:30-9:00)

2) EFFECTS FROM TEXTS (11:00-12:30PM)

LUNCH (12:30-1:30PM)

3) NEW APPLICATIONS FOR OLD SCHOOL QUANTITATIVE CORPUS LINGUISTICS APPROACHES (1:30-3:00PM)

**COFFEE BREAK **(3:00-3:30PM)

4) TRICK, OR TWEET? (3:30-5:00PM)

SATURDAY

COFFEE (Saturday, 9:00-9:30AM)

5) MEDIA FRENZY (Saturday, 9:30-11:00AM)

MORE COFFEE (11:00-11:15AM)

6) NAMED, FRAMED, AND TAMED (11:15-12:45)

LUNCH (12:45-13:30PM)

7) AND NOW FOR SOME COMPUTER SCIENCE APPROACHES (13:30-15:00)

COFFEE BREAK (15:00-15:15)

8) TOOLS AND METHODS ROUNDTABLE (15:15-17:00PM)

Quantitative Text Analysis 2F, Essex 2013

MY560: Unsupervised Text Scaling Methods

MY560: Supervised Methods for Classifying and Scaling Texts

COFFEE BREAK (3:00-3:30PM)