You asked

​​I note that you have performed some more black-box modelling of CV19 rates of infection etc. and presented the results here:

Obviously as you or your statisticians are aware, such modelled outputs, are subject to issues regarding GIGO.

Since the data appears to be being misrepresented by news outlets as factual rather than what they actually are i.e. "guesstimates", it is important that a more balanced view of the data is formed, represented and publicised.

Would you mind providing me with the original primary data that was used in the modelling? Can you confirm if it is based on factual case data or "guesstimates"?

The following data are specifically requested:

  1. Number of daily positive CV19 cases confirmed by positive tests by date.

  2. Number of daily CV19 tests conducted by date.

All data for the longest time series available, ideally every day from 1 March to the present day.

We said

Thank you for your enquiry.

To produce our positivity rates we use a Bayesian multi-level regression post-stratification (MRP) model. This allows us to reduce potential bias in the sample, in order to be more representative of the overall population we adjust for age, sex and region.

All estimates are subject to uncertainty, given that a sample is only part of the wider population. We therefore present 95% credible intervals with the Bayesian regression model estimates. A credible interval gives an indication of the uncertainty of an estimate from data analysis. 95% credible intervals are calculated so that there is a 95% probability of the true value lying in the interval.

Our research partners have published a pre-print article with more detail about the methodology behind the modelling, which can be found here.

Unfortunately, we do not hold the a table showing tests conducted daily, by date, and which of those are positive results. In order to fulfil this request, we would need to use a high level of statistical skill and/or judgement in order to create a bespoke analysis or table. Under the Freedom of Information Act 2000, Public Authorities are not obligated to create information in order to respond to requests. We therefore consider this to be information not held.

Access to the Coronavirus Infection Survey data are held in our Secure Research Service, which can be accessed by accredited researchers with approved research projects. The process for accessing the Secure Research Service can be found here.

Please also find a link to our methodology paper for further details on the modelling used here.