This message has been sent through the MASSOBS discussion list.
Remember, clicking 'reply' sends your message to the list.
-----------------------------------------------------------------------

Response from Bill Bytheway ([log in to unmask]) below:

I have known about Mass-Observation for a long time, perhaps since the
1970s, but my active engagement with it as a researcher began in March
2001. I commissioned a directive on birthdays which was issued in June
2002. I am interested in how birthdays contribute to a sense of age and
ageing, and how people react to this, personally, within families and
groups of friends and, more generally, as part of contemporary social life.
I could go on at length.

It is clear from the current discussion that tick boxes touch a raw nerve.
My life as a researcher started as a statistician and, for ten years or so,
I designed questionnaires and became 'expert' in using tick boxes to
generate data for a purpose. I understood how 'good' random samples were
needed if we were to generalise effectively and that social inequalities
for example can and should be exposed with authoritative statistical
evidence. So tick boxes have a place in research.

The word 'mass' loosely implies 'population', and the uneasy historical
relationship between M-O and opinion polls is reflected in the discussion.
At one extreme there is the sceptical view that opinion polling can be made
to prove anything, and that at the other it provides a welcome opportunity
for 'the people' to have their say. So the collected views of M-Os may be
interpreted as representing (or purporting to represent) the views of 'the
mass' and, in that context, it is important to assess and discuss
how 'representative' they are.

What makes a panel 'representative'? The scientific ideal is that the panel
should be recruited through strict random sampling procedures. A more
realistic alternative is that the panel should have a 'proper' balance
between the proportions in various basic categories: that they are the same
as in the wider population. Either way, tick boxes are needed if the claim
that the panel is representative is to be defended.

A third alternative is to ensure that certain basic categories are
represented adequately rather than in the right proportions. In other words
the aim is to represent the diversity of the population, ensuring that
there are 'enough' rather than the 'right' numbers in specific categories.
For example, I have described the representativeness of those who responded
to the Birthdays directive in the following way:

Information is available regarding the occupations of the respondents and
where they live. This demonstrates that most are associated with middle
class occupations and most live in southern England. That said however,
there are many exceptions: a typist, HGV driver, carpenter, cleaner,
artist, labourer, florist, waitress, bed and breakfast proprietor and
postmistress. And respondents include people living in such places as the
Shetland Islands, West Wales, Jersey and Northern Ireland, and even one in
Paris and one in Kalgoorlie, Australia.

In short, in writing this, I was attempting to indicate two things: first,
that I was not embarrassed to acknowledge that there are proportionately
more M-Os in middle class occupations, and proportionately more living in
southern England than in other parts of the UK and, secondly, that I am
pleased to register that there are many exceptions, including at least one
living in my neck of the woods (West Wales). The great thing about this
approach is that tick boxes aren't essential. The background information
already available, plus the mini-biogs, is more than adequate.

Dorothy, in her 21st July message, indicated that many researchers want
more information about 'social class, ethnic origin, etc.' Obviously she
has to listen to these views since the Archive needs funds and currently
relies heavily upon the active involvement of social researchers. But this
seems to me like pressure to make M-O more like other sources of social
research data, collecting the same kind of information, etc. etc. and I
would argue that the strength of M-O is that it's so different. Moreover,
we need only reflect on the problems the census has had over the years
with 'measuring' social class and ethnicity, to realise that 'collecting'
such information is not easy. It seems obvious to me that, by asking M-Os
to tick more boxes, we could easily loose some who are already volunteering
information that is so much more telling.

Dorothy says the next directive will be discussing this and I for one will
be very interested to see what correspondents have to say about it.

--------------------------------------------------------------------------------
To leave this list email [log in to unmask]
Alternatively, send the following command to [log in to unmask]

leave massobs
--