[Stoves] RESEARCH PROPOSAL

[Crispin Pemberton-Pigott]

Dear Gloria

>Two communities will be selected and
>the results to be compared and analysed ...

I am just completing a survey of 36 homes in Maputo (well...other people are
walking around, I do the math) and in that case we looked at the same homes
with and without an improved stove.

There are SO many variables it is a nightmare enough just to visit the same
places, let along trying to draw much from the effect of large studies.
Kirk Smith wrote a document a few years ago on how many homes to sample and
how to process them. It is really a good read so I think you should start
there.  Perhaps Dana at CEIHD can locate the document in question.

Whatever it says there is a good example of what people are going and how
accurate the results are.

What is not giving very good results is doing tests of performance where the
test and reporting does not include enough information to remove some
variables and unintended consequences.  For example, if you test with an
improved stove, do people simmer much longer (until the wage earner comes
home) because the fuel is saved and basically free?  People get used to
buying or using a certain amount of fuel and will sometimes save nothing at
all if the stove allows them to cook longer!

Things like that.  It can be really hard to show what the improvement is in
fuel saving.  With air quality it is rather easier because all you have to
show is what people breathe.  However, if people are cooking much longer
because the fuel lasts longer, you must capture that deviation from earlier
practice.  If the pollution is 1/2 and they cook twice as long, you are no
better off (are you?).

Regards
Crispin 
in Matsapha, Swaziland