This listing will grow over time, but here are a few resources that I’ve particularly enjoyed:

Crowston, K., Allen E. E., & Heckman R. (In Press).  Using natural language processing for qualitative data analysis. International Journal of Social Research Methodology.

This paper examines the potential of natural language processing for qualitative data analysis. It compares traditional or computer assisted hand coding with nlp techniques. I did some work on this topic, but I feel like I didn’t get deep enough into the nlp end. Crowston et al did get further into the nlp, and, as a result, they ended with the opposite conclusion than I did. They found nlp to be more time consuming than traditional analysis, where I found it to be less. I think that the “moral of the story” is that the more time you put into text analysis, the more you get out of it. Working on the analysis was frustrating, because I had to constantly redefine my goal and recalibrate the distance to it. A a full scale analysis of qualitative data can be very time consuming and requires a great deal of disparate expertise- definitely a good time for humility.

Schwarz, Norbert. 1996. Cognition and Communication: Judgmental biases, research methods and the logic of conversation. Hillsdale, NJ: Erlbaum.

This is a particularly important reading for people working in the field of survey research. It provides an introduction to each of Grice’s maxims of conversation and then relates them to survey questioning. Although there is research on many of the elements of Schwarz’s argument, including research that shows the measurable shifts in answers to specific questions when they are ordered differently, or the research that suggests that more concise questions yield more useful responses, Schwarz puts this research into the wider context of conversation analysis, thus stringing together what are always thought to be separate communicative events into a single communicative event and adjusting our understanding and reasoning accordingly. Although this paper is radical in the field of survey research, it is not radical for a sociolinguist. It’s a good example of how sociolinguistic methodology is relevant to survey research.

Various authors. 2011. 75^th Anniversary Issue. Public Opinion Quarterly. 7(5).

This whole journal edition is devoted to the past, present and future of survey research, and the pieces were written by the best known authorities in the field. Mick Couper, the leading authority on web survey design, wrote ‘The Future Modes of Data Collection.” Nora Cate Schaeffer, the leading authority on question writer (who has collaborated in conversation analysis studies of interviewer and respondent interaction) wrote ‘Questions for Surveys: Current Trends and Future Directions.” This issue comes at a time when the field of survey research is undergoing a transition. The national and a regional AAPOR conferences in 2012 focused on the frontiers of public opinion research. There has been a steep increase in the amount and decrease in the quality of surveys that potential respondents receive, changing the cost benefit equation and inherent relationship between the surveyor and the surveyed. There have also been concurrent, rapid technical changes and changing expectations from customers and respondents.

Pew Research Center report on the Kony 2012 phenomenon

This report is a real inspiration to me, because of its seamless integration of findings from different methodologies. They integrate reliable quantitative analysis with a social media analysis, which pulls the report through into a cohesive and compelling narrative.

Lou, Jia. 2007. Revitalizing Chinatown into a Heterotopia. Space and Culture. Vol. 10. No. 2, 170-194.

This paper is very different from the others I mentioned above. I’m including it, because it demonstrates an entirely different kind of analysis from any of the other sources I’ve mentioned, a visual semiotic analysis. Read through this paper and think ‘what is data?’ Lou strays very far from any kind of traditional numeric dataset, with great success.

Panel: The promises and perils of big data. by Rachel King

A few gems from this article:

“The more you zoom in, the more pathologies you find,” Hammerbacher said.

“It’s hard for me to think of a case where measuring more makes you less certain,” Hammerbacher asserted.

Sentiment Analysis: How Companies Now Listen To The Web. by Doug Henscehn

This is an especially well written introduction to sentiment analysis and its strengths and weaknesses.

http://www.cs.cornell.edu/home/llee/omsa/omsa-published.pdf

This is a really great publication by Bo Pang and Lillian Lee about Opinion Mining and Sentiment Analysis. It was written it 2008, which should make it ancient history. But their work is solid enough to make it a really great starter reference or guidepost.

I’ve added to links to some particularly good reads to the main page of my blog, on the right hand column. Feel free to suggest more!