I think she said 18 or 19 people had died in the last four months, and claimed that 1-2 per year was normal.
I am merely pointing out that 1-2 per year would not be the correct comparison. How many one would expect to die in this period (the winter months remember, when more people tend to die anyway) out of a population of 1800 would depend very much on the demographics of her particular route. Something like 10-12 is probably about right for an average population. In which case 18 would not be statistically significant at all.
One would expect to find a great many similar routes where 18 people had died over the winter just as if one flips a coin ten times it would not be extraordinarily unusual for it to come up heads seven times - and nor would it be proof of anything if it did.
Again - I'm not disputing that there is a problem. There is - and it is evident from the population level staistics which show a very high level of statistical significance. I'm just pointing out that this story is not proof of anything at all, and is not worthy of reporting.
She said in the last four months.
I think she said 18 or 19 people had died in the last four months, and claimed that 1-2 per year was normal.
I am merely pointing out that 1-2 per year would not be the correct comparison. How many one would expect to die in this period (the winter months remember, when more people tend to die anyway) out of a population of 1800 would depend very much on the demographics of her particular route. Something like 10-12 is probably about right for an average population. In which case 18 would not be statistically significant at all.
One would expect to find a great many similar routes where 18 people had died over the winter just as if one flips a coin ten times it would not be extraordinarily unusual for it to come up heads seven times - and nor would it be proof of anything if it did.
Again - I'm not disputing that there is a problem. There is - and it is evident from the population level staistics which show a very high level of statistical significance. I'm just pointing out that this story is not proof of anything at all, and is not worthy of reporting.
Fair enough. But there are a lot of coincidences in the last two years.
Sure - but reporting non-events as significant doesn't help the cause one whit. Just makes us look silly.