[Air-L] Longitudinal qualitative analysis of CMC

James Howison jhowison at syr.edu
Thu Jan 31 08:23:38 PST 2008


Hi,

Our research group has been studying dynamics of free and open source  
software communities, which stand somewhere between work teams and  
online communities (or perhaps they are both, not sure of your  
definition :).  Our approach is outlined in this NSF grant:

<http://floss.syr.edu/proposals/2005hsd.pdf>

Methodologically, we've used content analysis of online archives,  
using hybrid developed schemas (ie theory + induction) which are then  
applied in a fairly positivist way.  We've also used social network  
analysis based on these archives.  We're building up to Computational  
Natural Language Processing approaches for large scale studies.

We're still working up our longitudinal approaches, so far we've  
primarily worked through periodization, identifying relevant periods  
(such as beginning, v1.0 and recent, or end for failing teams).  You  
can see that approach in our study of decision making:

Heckman, R., Crowston, K., Li, Q., Allen, E., Eseryel, U. Y., Howison,  
J., and Wei, K. (2006). Emergent decision-making practices in  
technology-supported self-organizing distributed teams. In  
International Conference on Information Systems (ICIS 2006).
<http://floss.syr.edu/publications/Heckman2006Emergent_Decision-making_Practices_in_Technology-supported_self-organizing_distributed_teams.pdf 
 >

More recently we've begun to look at small periods, such as months,  
and creating time-series for the prevalence of particular behaviors  
(such as group maintenance behaviors), then comparing those time- 
series to time-series of effectiveness measures.  One big issue is  
whether to think in terms of calendar time or event time.  Our teams  
are 'part-time' for participants, so one can't assume constant effort;  
this means one has to normalize observations of behavior in some way,  
so event time really starts to make sense.  We're exploring inter- 
release periods as the 'natural periodization' for our teams, but  
other communities may have different 'zeitgebers'.  You can see a  
monthly approach in our SNA analysis work (other papers cited from  
this one):

Wiggins, Andrea and James Howison and Kevin Crowston (2008) Social  
dynamics of FLOSS team communication across channels, Submitted to  
'Fourth International Conference on Open Source Software (IFIP 2.13)'
<http://floss.syr.edu/StudyP/DSNAWigginsIFIP.pdf>

Another big challenge, for those seeking to quantify change over time  
in these communities, is how to deal with the auto-correlation in time- 
series, through techniques such as ARIMA regression models, which is  
pretty important if you intend to use regression with the time-series  
as variables.

For my dissertation I am examining organizational change by looking at  
the genre of documents produced by free and open source teams over  
time, and comparing time-series of relative genre use to effectiveness  
measures over time. Finally I'm writing narratives of how those genres  
(and genre-systems) emerged and why they, and their relative use,  
change over time.

<http://james.howison.name/pubs/howison_proposal_abstract.pdf>

There's also this excellent paper which examines the emergence of  
participation in a FLOSS project:

von Krogh, G., Spaeth, S.  Lakhani, K. R. 2003, ‘Community, joining,  
and specialization in open source software innovation: a case study’,  
Research Policy 32(7), 1217–1241.

For a larger-scale, non content analysis approach, you might find this  
study interesting:
Christley, S.  Madey, G. 2007, Global and temporal analysis of social  
positions at sourceforge.net, in ‘The Third International Conference  
on Open Source Systems (OSS 2007), IFIP WG 2.13’, Limerick, Ireland.

<http://www.nd.edu/~oss/Papers/oss2007_temporal.pdf>

Finally, Hala Annabi, now at Ohio University, also studying group  
learning over time in the Apache project.

--J





On Jan 31, 2008, at 8:38 AM, Ulrike Pfeil wrote:

> Hello everybody,
> I am interested in doing longitudinal qualitative analysis of  
> asynchronous online communities in order to analyse the development  
> of the online community over time (e.g. looking at how people  
> develop a sense of community, how communication and interaction  
> patterns change over time, how people take on or abandon certain  
> roles within the community etc.).
>
> However, I am having a hard time finding methodological guidance on  
> how to go about it (the longitudinal part). Does anybody have  
> experience with longitudinal qualitative analysis of CMC or can  
> recommend literature that might help me? Any help would be greatly  
> appreciated!
>
> Kind regards,
>
> Uli
> -- 
> Psssst! Schon vom neuen GMX MultiMessenger gehört?
> Der kann`s mit allen: http://www.gmx.net/de/go/multimessenger
> _______________________________________________
> The Air-L at listserv.aoir.org mailing list
> is provided by the Association of Internet Researchers http://aoir.org
> Subscribe, change options or unsubscribe at: http://listserv.aoir.org/listinfo.cgi/air-l-aoir.org
>
> Join the Association of Internet Researchers:
> http://www.aoir.org/
>




More information about the Air-L mailing list