[Air-L] Possibly of interest: A paper on validity issues in the use of social network analysis with digital trace data

[Kevin G Crowston]

I'd like to call your attention to a paper that I hope may be of interest and use to members of this research community, entitled "Validity issues in the use of social network analysis with digital trace data", by two of my students, James Howison and Andrea Wiggins, and me. The paper appeared in the Journal of the Association for Information Systems, a journal that I expect few of you regularly follow, so I thought an email might help bridge the gap. 

In the paper we argue that data obtained from online systems, e.g., mailing lists, wiki and the like, are different in important ways from the typical data of SNA, but that the implications of these differences for the validity of research are not always fully appreciated. For example, in contrast to a cross-sectional survey, data from systems often record events that happened over time that have to be collapsed to create a network. However, doing so may create apparent connections in the network that don't occur in the data. We provide a set of suggestions for using such data in SNA studies (and for reviewing papers that use such data). The abstract is included below. 

The official URL for the paper is http://aisel.aisnet.org/jais/vol12/iss12/2/. If you're not a subscriber, I'd be happy to mail you a copy or you can find a preprint at http://crowston.syr.edu/content/validity-issues-use-social-network-analysis-digital-trace-data. Comments on the paper are always welcomed. 


Validity issues in the use of social network analysis with digital trace data. JAIS 12(12) paper 2. 

There is an exciting natural match between social network analysis methods and the growth of data sources produced by social interactions via information technologies, from online communities to corporate information systems. Information Systems researchers have not been slow to embrace this combination of method and data. Such systems increasingly provide "digital trace data" that provide new research opportunities. Yet digital trace data are substantively different from the survey and interview data for which network analysis measures and interpretations were originally developed. This paper examines ten validity issues associated with the combination of data digital trace data and social network analysis methods, with examples from the IS literature, to provide recommendations for improving the validity of research using this combination.


Kevin Crowston
Syracuse University                            		Phone:  +1 (315) 443-1676
School of Information Studies			Fax:    +1 (815) 550-2155
348 Hinds Hall						Web:    http://crowston.syr.edu/
Syracuse, NY   13244-4100   USA




Kevin Crowston
Syracuse University                            		Phone:  +1 (315) 443-1676
School of Information Studies			Fax:    +1 (815) 550-2155
348 Hinds Hall						Web:    http://crowston.syr.edu/
Syracuse, NY   13244-4100   USA