[Stoves] report with dissapointing results from cleaner cookstoves-new take

[Andrew Heggie]

On 13 December 2016 at 22:25, Andreatta, Dale A.
<dandreatta at sealimited.com> wrote:
> The drum I've been banging on lately is the question of whether the cook loves the stove.  It's been a few days since I looked at the original article, but I believe the picture showed a batch load stove with a very small fuel chamber, with a pile of large wood in the background.  If the user has to break the wood into small pieces, and/or remove the pot to add the fuel, they probably won't use the stove.  I've not looked at the original paper in the Lancet, but it would be interesting to know if they tracked how much the stove was used.  I suspect it wasn't used very often.

Dale I suspect the pictures in the article where stock pictures,
pickked by the BBC reporter.

The study shows whilst there were stove failures ( mostly battery
related as they had fans but more failures of the solar panels which
charged the battery) they were repaired promptly and utilisation seems
to have been high for at least one meal per day (It looks like two
stoves were provided per household with a charger battery and lamp).
Of course if the stoves were not used as much as they should have been
this could affect the result also.
>
> As others have pointed out, there appears to be a lot of background smoke from various sources.  I recall data from the article that said particulate levels were still pretty high even when using LPG, which produces virtually no particles.  It might be possible they were never going to get a measurable health effect, no matter how clean the stove.

And that seems to be the case here
>
> Finally, I'll suggest there could be something funny with the statistics.  I'm not an expert, but I think the basics of statistical analysis goes like this:
>
> Step 1, set up a hypothesis to measure:  The stove makes a difference.
> Step 2, this determines the null hypothesis:  The stove doesn't make a difference.
> Step 3, pick a confidence level:  Let's say 90%.
> Step 4, do the experiment and determine how confident you are that the hypothesis is correct.
> Step 5, if greater than 90% confident, report that the stove makes a difference.
> Step 6, if 89% or less confident the stove makes a difference, report that the stove makes no difference.
>
> I question the last part.  I think this is why we are constantly see conflicting information about "Eggs are bad for you", then "No, eggs aren't bad for you".
>


The report does not criticise the cleaner stoves, it just suggests in
the situation they were deployed there was no measurable impact on
respiratory illness.

AJH