I'd be interested in hearing from folks about any initiatives or emerging needs on their campuses to do social media analysis in the social sciences. I am familiar with efforts at SUNY Stony Brook, Penn State, and Ohio State, but these are all at the graduate level. Anyone trying to crack this nut at the undergraduate level?
The following sites and articles may be of interest to you. I have done virtual ethnography (as an undergrad student who promoted the use of technology) and used a software for data analysis (MAXQDA) which I also used with a student whom I was mentoring through her senior thesis. I found focus on the thesis and questions to be a key issue when using larger data sets and approached undergrad work much like focus group research. The software analyzes key word usage which helps to eliminate useless data as well as to identify newly discovered questions which should be considered.
Thanks for the thoughtful reply. The paper you link to is a valuable read, thanks. The work we have undergrads doing is very similar (but not in as much depth) as this work. What tool did they use to collect their Tweets? What tools did they use to do the analysis? I notice the Air Force sponsored their work. How did they make that connection?
I am familiar with W5 and their work in Online Ethnography - thanks for that link. Our faculty that lead this effort are expert in this type of work. Two have their Ph.D's in this space. That said, a good link in this area is always welcome.
I am familiar with MAXQDA. We use Atlas.ti extensively here. We've found these tools a "bridge too far" for the typical social science undergraduate. They still do (and need to do) some assignments where they code text manually to learn about this methodology. Tools like IBM's Modeler, R, and Rapid Miner can be setup to do this type of "first pass" analysis allowing students to concentrate on the "discover". The other problem with MAXQDA, Atlas.ti, etc. is cost. How do you deploy enough copies so 300 undergrads a semester can use it? IBM gifted us Modeler and our VIDIA environment makes R, Rapid Miner, Orange, etc. available in unlimited numbers for free.
What campus are you on? What other assignments/tools do you use for your undergrads? Would you be willing to share assignments? Would you like access to VIDIA for doing this type of work? Thanks again for the post. It was full of useful stuff and much appreciated.
Unfortunately, the paper does not specify what software they used though they did specify the type. There is a good listing of open source software at http://www.predictiveanalyticstoday.com/top-30-software-for-text-an... Some of these look much too complicated for undergraduates, though others don’t appear to be so. They would have to be tested to assure that they are appropriate.
Re: Air Force Connection: the AF has a program that sponsors research through grants. Info is at http://www.wpafb.af.mil/library/factsheets/factsheet.asp?id=8127 They do sponsor undergrad research and details are available at the website.
I am at Plattsburgh and unfortunately the campus has been very slow to adapt and move into the tech world. When I want to try software, it is usually purchased by me. While the campus is improving somewhat it is still very behind where it should be. Most technology is geared toward science or business and there is no social science technology on campus. When I researched the different software, I found MAXQDA to be the most user friendly, but I have found Australian and European software to be more developed with the user in mind versus US technology which is often to “techie” for the average user. I would use it in an undergrad class if available; however, as you say the cost of these can be significant.
Because of the cost and lack of availability I am hampered in creating assignments in which qualitative analysis software can be used. I was able to use it with the student I was mentoring because I had the software myself. The student was interested in learning the effect of school consolidation on a small rural town where she was raised. I was familiar with her research question and she did focus groups (town residents and teachers) and interviews (of education administrators) and examined the views of specific groups for similarities and differences. It was set up so that it could be expanded and used as the basis of her Master’s Thesis. If you have anyone with an interest in this type of research I would be happy to share.
I am not familiar with VIDIA and would love to learn more about it by accessing it. Thanks and please accept my apologies for the “quickly put together” response. I am getting ready for a trip and teaching this summer and this week is somewhat busy.