Reply #27: Looks like unprofessional cherry-picking to me

[muriel_volestrangler]

Here are the figures I think I get for those 8 cities, for weeks 1-22 of 2011:

15
18
8
16
11
14
18
9
9
8
9
15
8
5
11
11
18
17
11
13
11
8

Amazingly, those 4 weeks they chose before March 19th are the 9+9+8+9=35. The 10 weeks after add up to 120. Of course, if they'd chosen 5 weeks before the date, they'd had to have included that inconvenient '18' which throws their claim out.

FWIW, here are the same figures for 2010:

21
11
12
13
14
15
10
10
11
16
7
17
12
15
13
14
10
9
10
15
19
9


I don't think there's any way this can be called 'statistically significant'. The figures vary a lot from one week to the next (and you'd want information on any infectious diseases - there are flu figures, but not split out by age), and they give no reason whatsoever for picking 4 weeks and 10 weeks. It looks dishonest, to me.

(On edit: a few figures for cities are 'U'='unavailable' - I interpolated a figure for those. The 9+9+8+9 figures do not include any interpolated figures.)