An error has been identified in the Office for National Statistics (ONS) online time use survey (OTUS) data, which affects the following releases:
Recoding the raw data meant 10% of the weighted sample was dropped in error. The error was identified while we re-platformed our analysis system so that it was more automated. This includes improved embedded quality assurance, and as a result the risk of the error reoccurring is reduced. We will be updating weights to account for additional sources of bias (non-response bias) in the near future.
Although the new corrected data differs to that which is published, the key messages presented in the articles are unchanged. We will not be revising these with updated figures.