by
Les May
THIS
morning
just after 7o’clock, I watched a Sky News presenter, who for reasons
which completely baffle me was standing outside 10 Downing Street,
ask a hapless government minister ‘when
are schools going to reopen?’,
a
question he could not possibly answer.
At
the moment we do not even know whether the numbers of new infections,
as measured by the figures published at the end of each day, are
fluctuating randomly around some stable figure, indicating a plateau,
or are actually falling as
appears
to be the case.
Using
the published data for the period after 8 April it is possible to
calculate* that there is approximately a 1 in 9
chance
(11%)
that we would get figures like this if the number of new infections
was in fact stable and not really falling. This
is
hardly
evidence that there should be an easing of the so called ‘lockdown’.
It
seems to me utterly irresponsible for the media to constantly use
‘personal
stories’
of the difficulties that people face being cooped up, e.g. with
children in small spaces and no garden, to subtly intimate that the
government should be able to tell us when various restrictions are
going to be eased. You can only have an ‘adult
conversation’
when both sides are willing to behave like adults.
I
suggest
that the sensible thing to do is to keep watching what is happening
in China. You don’t have to believe
the quantitative data, i.e. the figures which are published regarding
infections and deaths; look at the qualitative data, i.e. how and
when China is easing controls on movement and allowing facilities to
reopen. Parts of China which have been living under strict controls
for three months are only now
beginning to reopen. This
may be a clue there about how long some
of our
own restrictions need to be in place.
*
Any
test is complicated by the so called ‘Monday
Effect’.
The test I used is the non-parametric Cox-Stuart test for trend
modified to take this into account. The
figure I give is approximate, but indicates the need for continuing
caution.
(Look
about half way down the page.)
^
The
figures for China may indeed be suspect, but does anyone take the US
figures for testing seriously?
************************************
Posts
No comments:
Post a Comment