Solomon Messing

Research interests: images, race, politics, networks, technology, machine learning, computational social science.

Here is my CV, last updated May 2011.

And my recently launched blog, which is part of the r-bloggers network. 

Contact

email: [last-name] AT stanford DOT edu             (this prevents spam from clever RegEx email harvesters)

450 Serra Mall, Building 210, Stanford, CA 94305

Current projects

Skin Complexion, Race and Politics

Analysis of images in attack ads in the 2008 Presidential Campaign advertisements (with Ethan Plaut and Maria Jabon).  "Bias in the Flesh: Attack Ads in the 2008 Presidential Campaign" APSA poster presentation. ICA paper available here. 

"Experimental evidence of the impact of skin complexion on electoral preference" (with Shanto Iyengar, Kyu Hahn, Jeremy Bailenson), APSA paper available here.

Analysis of skin complexion and electoral outcomes in 2010 (with Shanto Iyengar).

Social Media and Political Communication

"An Era of Social Media Effects? How Social Media Change the Way We Consume News and Reduce Partisan Selective Exposure" (with Sean Westwood) - Web experiments show that the mere presence of social endorsements changes the way people select news content, reducing the impact of news source in favor of social cues. Paper presented at the Midwest Political Science Association annual conference in Chicago, IL, 2011, currently under double-blind review.  Draft available here.

Social Network Analysis

Labs for Social Network Analysis using R (with Dan McFarland, Mike Nowak, and Sean Westwood) - A few of us in the Comm Department took on the task of adapting Dan McFarland's labs for Social Network Analysis to R (most labs previously relied on UCINet). My main contributions are lab 5 (two-mode networks and visualization) and the peer influence portion of lab 7 (peer influence and QAP) - now complete.

"The source of deliberative opinion change: Reconsidering the roles of knowledge, group polarization, peer influence and persuasion" (with Sean Westwood).  Using network data, shows that directed interactions of high discourse quality between participants is more predictive of final attitudes than previously understood phenomena including individual-level knowledge gain or group-level polarization.  MPSA paper available here.

Software

Some R packages I've worked on:

ImageMetrics  R package to analyze image data, designed to facilitate analysis described in “Bias in the flesh” (see above). Package vignette available here.

wvioplot  An R package to produce violin plots with weighted data, which allows the visualization of the distribution of a quantitative variable along with its quartiles and a measure of central tendency. A violin plot is a combination box plot and kernel density plot.

triads with Sean Westwood, Mike Nowak, and Dan McFarland. An R package to determine node-level triad type membership for the 16 types of triads that occurr in a directed network. Based on SAS code developed by James Moody.

NetCluster with Mike Nowak,  Sean Westwood and Dan McFarland. An R package to facilitate network clustering and evaluation of cluster configurations. 

NetData with  Sean Westwood, Mike Nowak and Dan McFarland. This package contains all data needed for Dan McFarland's SNA R labs.

Also:

Open IAT/bIAT (with Sean Westwood) - Web application that implements the Implicit Associations Test and brief Implicit Associations Test, based on JavaScript, PHP, and MySQL.