Will Lowe

Welcome. I am a political methodologist specializing in statistical text analysis with applications to, frankly, anything that seems interesting. Most recently that’s ideal point estimation from text, central banking, legislative conflict, and policing. But not together, obviously. Some of that stuff is below and in the trusty CV.

Starting this semester I’ll be Lecturer in the Department of Politics at Princeton, helping teach POL 345 (Intro to Quantitative Social Science) and workshops for the graduate Research Methods Sequence. The rest of the time I’m still a Senior Research Specialist in Q-APS, the Program in Quantitative and Analytical Politics. (That’s where we store our methods people.) Nobody knows what a Senior Research Specialist does, and I’m not about to let on.

Each summer I teach POL 245 (Visualizing Data) at the Freshman Scholars Institute where a select cohort of first generation and low income students meet an awful lot of causal inference, statistics, and coding over six weeks. This year the University wrote about us. We’re ‘Friday’.

You can sometimes find me answering questions on stats.stackexchange.com, writing code on GitHub, or wittering on Twitter. There’s a blog here too. Alternatively you can ponder some biographical stuff. Here’s a picture of how I used to look and still do in my head.


In Preparation

W. Lowe and J. Mummolo ‘The bias is built in: How administrative records preclude meaningful causal assessments of racially biased policing’. Paper prepared for the American Political Science Association meeting, September 2018.

N. Baerg, D. Duell, and W. Lowe ‘Inflation expectations as public opinion?’

W. Lowe. Putting it all on the line: Some unified theory for text scaling (formerly ‘Scaling things we can count’ and ‘There’s (basically) only one way to do: Some unifying theory for text scaling models’). Paper prepared for the American Political Science Association meeting September 2013, Chicago, and endlessly revised.

W. Lowe. Measurement Models for Event Data. Paper prepared for the European Political Science Association meeting June 2013, Barcelona.

Journal Articles

S.-O. Proksch, W. Lowe, and J. Wäckerle and S. Soroka (in press) ‘Multilingual sentiment analysis: A new approach to measuring conflict in parliamentary speeches’. Legislative Studies Quarterly.

N. Baerg and W. Lowe (2018) A textual Taylor Rule: Estimating central bank preferences combining topic and scaling methodsPolitical Science and Research Methods.

Y. Theocharis and W. Lowe (2016) Does Facebook increase political participation? Evidence from a field experiment. Information, Communication and Society 19(2): 1465-1486.

J.-H. Meyer-Sahling, W. Lowe, and C. van Stolk (2016) Silent professionalization: EU integration and the professional socialization of public officials in Central and Eastern Europe. European Union Politics. 17(1) 162-183 → data.

Y. Theocharis, W. Lowe, J. W. van Deth, and G. M. García-Albacete. (2015) Using Twitter to mobilise protest action: Transnational online mobilisation patterns and action repertoires in the Occupy Wall Street, Indignados and Aganaktismenoi movements. Information, Communication and Society. 18(2) 202-220→ preprint

W. Lowe and K. R. Benoit (2013) Validating estimates of latent traits from textual data using human judgment as a benchmarkPolitical Analysis. 21(3): 298-313 → data, video from NCRM (if you like that kind of thing).

S. C. Carey, N. J. Mitchell, W. Lowe (2013) States, the security sector, and the monopoly on violence: A new database on pro-government militiasJournal of Peace Research 50(2): 249-258 → datapreprint

K. Benoit, M. Laver, W. Lowe, and S. Mikhaylov (2012) How to scale coded text units without bias: A response to GemenisElectoral Studies 31(3): 605-608.

W. Lowe, K. Benoit, S. Mikhaylov, and M. Laver (2011) Scaling policy positions from coded units of political texts. Legislative Studies Quarterly 36(1): 123-155. → data

J. Sullivan and W. Lowe (2010) Chen Shui-bian: On independenceChina Quarterly 203: 619-638.

W. Lowe (2008) Understanding WordscoresPolitical Analysis 16(4): 356-371. → data

E. Jenne, S. Saideman, W. Lowe (2007) Secessionism as a bargaining postureJournal of Peace Research 44(5): 537-556. → data

W. Lowe (2004) Content analysis and its place in the (methodological) scheme of thingsQualitative Methods 2(1): 25-27.

G. King and W. Lowe (2003) An automated information extraction tool for international conflict data with performance as good as human coders: A rare events evaluation design’International Organization 57: 617-642. → data

J. Bryson and W. Lowe (1997) Commentary on D. Ballard, M. M. Hayhoe, P. K. Pook and R. P. N. Rao, Deictic codes for the embodiment of cognitionBehavioral and Brain Sciences 20:743-744.

Edited Collections

J. Bullinaria and W. Lowe, editors (2002) Connectionist Models of Cognition and Perception, World Scientific

Book Chapters

O. Bekou, W. Lowe, and D. Birkett (2016). Fostering cooperation through technology-driven tools. In O. Bekou & D. Birkett (Eds.), Cooperation and the International Criminal Court (pp. 396–416). Brill.

J.-H. Meyer-Sahling, W. Lowe, C. van Stolk (2012) ‘Attitudes of public officials towards models of bureaucracy in Central and Eastern Europe’ In Executive Politics in Times of Crisis, M. Lodge and K. Wegrich (Eds.), Palgrave Macmillan. pp.99-117.

W. Lowe (2010) Identifying your accompanist in Close Engagements with Artificial Companions, Y. Wilks (Ed.), Benjamins. pp.95-99

W. Lowe (2006) ‘New methods for conflict data’ in Programming for Peace: Computer-Aided Methods for International Conflict Resolution and Prevention, R. Trappl (Ed.) Springer pp.321-334

W. Lowe (2003) Uncertainty and Sentence completion test in M. S. Lewis-Beck, A. Bryman and T. F. Liao (Eds.) Encyclopedia of Research Methods in the Social Sciences, Sage.

W. Lowe (2003) ‘Rare Events Research’. Encyclopedia of Social Measurement, Elsevier.

Peer-reviewed Conference Papers

(Not the ordinary unreviewed “send us an abstract an we’ll decide” kind like APSA, MPSA, EPSA etc.)

N. R. Baerg, W. Lowe, S. Ponzetto, H. Stuckenschmidt, and C. Zirn. (2014). Estimating central bank preferences. Paper presented at Workshop on Language Technologies and  Computational Social Science, ACL 2014, Baltimore, Maryland.

W. Lowe, O. Bekou, and E. Hunter (2008) ‘Markup and metadata in international criminal law’. eSocial Science Conference, Manchester.

J. Dowdall, W. Lowe, J. Ellman, F. Rinaldi and M. Hess (2004) The role of multiword terminology in knowledge management in Proceedings of the Fourth International Conference on Language Resources and Evaluation.

W. Lowe and G. King (2003) Some statistical methods for evaluating information extraction systems in Proceedings of the Tenth Conference of the European Association for Computational Linguistics pp.19-26.

W. Lowe (2001) Towards a theory of semantic space in J. D. Moore and K. Stenning (Eds.) Proceedings of the Twenty-first Annual Meeting of the Cognitive Science Society LEA pp.576-581

J. Bryson, W. Lowe and L. A. Stein (2000) ‘Hypothesis testing for complex agents’ in A. M. Meystel and E. R. Messina (Eds.) Proceedings of NIST Workshop on Performance Metrics for Intelligent Systems National Institute of Standards and Technology Special Publication 970. pp.233-240.

W. Lowe (2000) What is the dimensionality of human semantic space? in R. M. French and J. P. Sougné (Eds.) Proceedings of the 6th Neural Computation and Psychology Workshop Springer pp.303-311.

W. Lowe and S. McDonald (2000) The direct route: Mediated priming in semantic space in L. R. Gleitman and J. K. Joshi (Eds.) Proceedings of the 22nd Annual Conference of the Cognitive Science Society LEA pp.806-811.

S. McDonald and W. Lowe (1998) Modelling functional priming and the Associative boost in M. A. Gernsbacher and S. J. Derry (Eds.) Proceedings of the 20th Annual Conference of the Cognitive Science Society LEA pp.675-680

W. Lowe (1997) Meaning and the mental lexicon in M. E. Pollack (Ed.) Proceedings of the 15th International Joint Conference on Artificial Intelligence, Morgan Kaufmann pp.1092-1097. (7M bitmapped, sorry)

W. Lowe (1997) ‘Semantic representation and priming in a self-organizing lexicon’ in J. A. Bullinaria, G. Houghton and D. W. Glasspool (Eds.) Proceedings of the 4th Neural Computation and Psychology Workshop Springer pp.227-239


If you think ‘word embeddings’ and neural networks are new and cool, here’s thesis about them, from back when bootcut jeans were a thing:

W. Lowe (2001) Topographic Maps of Semantic Space PhD Thesis, Institute for Adaptive and Neural Computation, Division of Informatics, University of Edinburgh.

W. Lowe (2003) Content Analysis Software: A Review Technical Report for the Identity Project, Weatherhead Center for International Affairs, Harvard University.

R. C. Shillcock, S. McDonald, P. Hipwell and W. Lowe (1998) Distinctiveness of spoken word context predicts lexical decision time. Centre for Cognitive Science Research Paper EUCCS-RP-1998-4


R Packages


cbn contains replication materials for Caliskan, Bryson and Narayanan (2017), adds some new bootstrap statistics, and allows you to do similar analysis on your own word vectors. You can find it here: https://conjugateprior.github.io/cbn


qss.student bundles the problem sets from Imai (2016) Quantitative Social Science. It is designed for your students to install if you are teaching with this book. You can find it here: https://conjugateprior.github.io/qss.student


twfy is a light wrapper around TheyWorkForYou‘s API. You can find it here: https://conjugateprior.github.io/twfy.


Events is for manipulating event data of the kind generated by KEDS/Tabari. Go to the events homepage.


Austin is an R package for doing things with words. Right now it allows you to scale texts in the style of Wordscores and Wordfish. Go to the austin homepage.


Resha is an R package re-implementing Harun Reşit Zafer’s stemming tool for Turkish. He says it’s “less aggressive than Snowball”. You can find it here: https://github.com/conjugateprior/Resha.

Java Applications


Yoshikoder is a cross-platform multilingual content analysis program.
Go to the Yoshikoder homepage. Github pages are here


JFreq counts words, quickly. If you have a lot of documents that need to be preprocessed and turned into a word frequency matrix in a hurry without filling up your disk, this might be the software for you. Go to the JFreq homepage.


The YKConverter is a utility that tries to extract the text from documents in various formats (HTML, Word, PDF, Powerpoint, Excel, encoded text) and save it as UTF-8 encoded plain text. Go to the YKConverter homepage.


The Re-encoder takes a folder full of text files in one file encoding and switches them into another one. It needs a better name. Go to the Re-encoder homepage.


Content Analysis in Python

brief demonstration of how easy it is to do basic content analysis in python. Not really software but not really a tutorial either. Perhaps it will be useful or inspiring to someone.

There’s more bits of python code on the blog.

Third Party Software


VBPro is Mark Miller’s classic free content analysis software. I am simply hosting the latest version and cannot answer questions about it. Please address questions to Mark: mmarkmiller [at] mac.com.

Download vbpro.zip (for Windows only).


You can email me at wlowe [at] princeton [dot] edu.

If it’s sensitive, or you just like to play with fun toys, then you can use this:

PGP: public key
PGP fingerprint: 5D5B 2B00 382C E647 8B68  87D0 0AC2 609A 0570 BA8A

You can also send me things if you address them to

Department of Politics
Princeton University
314 Fisher Hall
New Jersey 08544

Here’s a map.