I have recently been extracting records for a particular site and when I run either a quick report or a report through the wizard I have found a number of records are returned which are not present in the survey data when checked via observer, date and location. In one instance the taxon observation had a 10 figure grid reference and it was possible to check back - what appears to be happening is the correct taxon appears in the report and an additional taxon is added - in this case the correct taxon was Succisa pratensis and an additional record of Symphoricarpos albus was added.  Other rogue taxa include Lychnis (Silene) flos-cuculi, Silene noctiflora, Sonchus arvensis, Trifolium dubium, Trifolium subterraneum, Valeriana dioica, Veronica hederifolia, Viburnum opulus, Viola riviniana - these appear to be associated with correctly reported taxa - although the original cannot be identified with certainty in all cases - the original records would appear to be Silene dioica, Silene latifolia, Sonchus asper, Trifolium sp?, Trifolium sp?, Urtica dioica, Veronica chamaedrys?, Viola odorata. Since I had carried out the orginal survey it was easy for me to spot the additional taxa and to check the survey data entered into Recorder - however if this hadn't been the case I wouldn't necessarily have spotted it. The question is how common are such reporting errors? Has anyone else experienced this and how does it arise and how can it be prevented? - as clearly it is not very helpful!