[Air-L] difference in representativeness: offline or online

[Elke Greifeneder]

  Hi all,

it is great to see iSchool members on that list! And yes, I can confirm: 
we do Internet studies. I am working on my phd at the Berlin School of 
Library and Information Science at the Humboldt-Universität in Berlin 
and my work has a lot do with Internet studies. And... I am stuck.

I am working on how to collect purposeful data in digital library 
evaluations. Sure, there are multiple factors that influence the quality 
of the data. One of them is the use of an appropriate method. Currently, 
researchers of digital libraries mostly do surveys and logfiles. Both 
methods met the user where he is: online. All other methods -  
interviews, focus groups, usability thinking aloud tests etc - are done 
"offline" that means in a local environment (for example in the 
institution who does the study). My question is whether this produces 
representative data? Can I take users which are locally available (for 
example users in my Berlin library) but who may not be the actual users 
of the worldwide accessible digital library? If I want to know what my 
(real) users need in future development and I only ask my Berlin users, 
what is the data's informative value?

I know that some internet anthropologist say that with internet studys 
we no longer can make the distinction between aged and young, masculine 
and female - the old standard criteria for a representative sample - but 
instead that the real differences in online communities lay in the 
cultures: do my users use social networks, are they night- or 
day-workers, what is their social background: would they ask for help 
etc. In that sense being representative means that one has all 
(cultural) user types be present in a sampling. This implies that the 
question is not where the study takes place - offline or online - , it 
matters whether all "user cultures" are present.

Does anyone on that list knows a paper that supports these thoughts or 
can help me with an idea what indicators I could use to prove it?

Thanks for all the great input I already received on that list. Every 
help comment on that topic would be more than welcome!

Best wishes
Elke


-- 
Elke Greifeneder
Lecturer
Berlin School of Library and Information Science
Humboldt-Universität zu Berlin

Associate Editor
Library Hi Tech - Emerald