Latest published prevalence rate of 1.18% was on 24 Dec, relating to the week finishing 1 week before that. These are based on a sample of the population.
PCR result @ 60,916 cases on 4 Jan (published on 5 Jan) and are dominated by people with some concern over symptoms.
Yes, you are quite right about date matching. I was taking the latest available data all round. If one goes back to the last ONS prevalence estimate, 'cases'/PCR tests were around 6%. This would make my figure of false positives 84.3%, or a true:false ratio of 1:5.4. This is still hardly a "small possibility" as ONS state.
Put simply, people who have symptoms are much more likely to test positive, than a random sample of people.
And people who are scared by the government and mass media propaganda are much more likely to go for tests anyway. They will be looking for the slightest symptoms - symptoms not subject to control and that could be due to a variety of causes.
The basic point to be made here is that it is not good enough to base public policy on a mass testing programme without any measure or open scrutiny of how accurate that may be. In the inexcusable absence of such validation, it is essential to make some attempt at quantifying it. You, quite rightly, pointed out inconsistency in dates, which I have corrected. There may be other inconsistencies, but it is a start that can be refined. And one does have to start somewhere. No scrutiny is not an option.
But still, as I previously mentioned, even if these are all true cases, it does not necessarily mean that we should be fearful. From both sides of the argument, one can take a more positive view than the government.
I believe the ONS sampling theory is correct, so will give a reasonable assessment of the infection rate across the population.
Currently reported to be about 2.1%. This will be valid for the week commencing perhaps 10 days ago. That means about 1.32million infected in that week.
The daily confirmed cases was about 50k cases/day in the same time period, so 350K people in that week.
This is about 1/4 of the infection survey data!
There are plenty of possible explanations : asymptomatic or mild symptoms, masking by other illness, people wanting to avoid being asked to isolate, no access to test etc.
So I'm afraid there is no denying it, the daily confirmed cases vastly under-report the real number of actual cases. ( by a factor of about 4 )
I believe the ONS sampling theory is correct, so will give a reasonable assessment of the infection rate across the population.
Currently reported to be about 2.1%. This will be valid for the week commencing perhaps 10 days ago. That means about 1.32million infected in that week.
The daily confirmed cases was about 50k cases/day in the same time period, so 350K people in that week.
This is about 1/4 of the infection survey data!
There are plenty of possible explanations : asymptomatic or mild symptoms, masking by other illness, people wanting to avoid being asked to isolate, no access to test etc.
So I'm afraid there is no denying it, the daily confirmed cases vastly under-report the real number of actual cases. ( by a factor of about 4 )
A reported 50k/350k daily cases is 14%, which is 6.8x the estimated ONS estimate. How is this an under-report?
Don't understand your query.
50k was the approximate daily confirmed cases number, which equates to 350k cases per week.
That 350k/week is much less than the ONS estimate of 1.3million/week - Hence an under-report
Your question compared cases/day with cases/week, which leads to a factor of 7, as you say.
Don't understand your query.
50k was the approximate daily confirmed cases number, which equates to 350k cases per week.
That 350k/week is much less than the ONS estimate of 1.3million/week - Hence an under-report
Your question compared cases/day with cases/week, which leads to a factor of 7, as you say.
Sorry, I took your 350,000 as being the number of tests. After all, it makes no sense to add up 'cases' over a nominal number of days and compare with the total estimated cases in the country.
We need to compare percentages. That is ONS prevalence versus 'cases'/PCR tests. Here are the actual figures to compare:
What we see here is that, since mid-October, the positive cases returned by tests are 6-7 times what the ONS estimates. As test numbers increase and apply to a more typical cross-section of the population, we would expect the gap between prediction and measurement to close. What I take from this is that there is a significant amount of errors coming from the tests.
