[Air-l] Statistics Canada - General Social Survey: The Internet and the way we spend our time

[E.J.Helsper at lse.ac.uk]

I think many of us have struggled to find an appropriate way to measure time spent online or (heavy) internet use. Although diaries seem a good solution often people only report one activity when in fact they are doing a number of activities. Or in the case of the PCs when they have 7 windows open at the same time and they are dipping in and out of the internet they do not consider themselves to be actually using the internet. Each of us has different definitions of what is meant by 'using the internet' and what is ‘real’ contact with others. Perhaps people are now more often in touch with friends and family through email and chat, but perhaps what they say is different? Is this a quantity versus quality debate? 

Often people seem confused about emailing, was this included as an internet activity in this study? For many people it technically and practically is not. It is probably also difficult to separate work from personal use for many people, since personal emails are written in parallel to work emails and vacations are booked at work and work is done at home all using the internet or PCs. Extremely hard to measure therefore and frustrating to researchers. 

Does anyone know if there have been extensive comparative studies on the effects of different methodologies to measure internet or media use? What are the differences when you use interviews, diaries, or a survey? How does this correspond to observation of online behaviour? I know that the research has been done but I am not sure if something has been published recently in relation to this topic. I would be interested in compiling the available texts on this subject both for teaching and to inform my own work.

One more note. There will always be measurement error, that's the nature of social statistics, but that's also why interpretation is so important and has to be done with caution. In the case of the Canadian survey or at least this report I think it is not measurement errors that are the problem but perhaps an over interpretation of the statistics in terms of attributing causality and strong effects. 

If anyone has good references on measuring media use time (especially if they are focussed on the internet) I would be interested in receiving these.

Best wishes,

Ellen J. Helsper
Tutorial Fellow
Media and Communications Department
London School of Economics and Political Science
Houghton Street, London WC2A 2AE
E.J.Helsper at lse.ac.uk