Validation of records
Making sure the maps are correct – the validation process
For the Atlas to be THE definitive record of the breeding and wintering birds of Britain and Ireland during 2007-11 we needed to be 100% certain that all submitted records were correct. With 19 million bird records submitted it was inevitable that the odd error could creep in along the way. See, for example, the Crested Tit map that appeared as 'Bird of the Day' which shows two records a considerable distance outside the “normal” range. Rather than mis-identification, it is likely that these were due to an error quoting a grid reference, or simply ticking the wrong species online.
In December 2008 we launched a “validation module” of the Bird Atlas website which enabled Atlas Regional Validators to view all the records submitted in a region and query oddities. We expected that the majority of these would be errors in submitting the data (e.g. selected the wrong species, transposed grid references, count typos) but there could have been some questions about species identification.
If a Regional Organiser queried an observer's record a message to this effect would appear on their Bird Atlas Online Data Home page. Each queried record was labelled according to whether it was the identification (ID), number (No.), breeding status (BS), grid reference (GR) or colony nest count (Nt) that alerted the validator to a possible error. The observer then had the option to delete or edit the record if they agreed with the error, or email the validator to clarify the record, providing a description if required for local or national rarities.
Unlocking the science to reveal the state of nature
David Noble takes a sober look at the latest State of Nature Report.
What effect might annual releases of non-native gamebirds be having on native biodiversity?
Henrietta Pringle reveals the work behind a recent paper on gamebirds and predation