Re: Tools for Validating
Trevor James has just circulated a (draft) paper for comment "Improving wildlife data quality: guidance on data verification, validation and their application in biological recording".
Question is, do we have enough tools in Recorder to actually carry out all the essential validation tasks efficiently?
We have many tools within Recorder to manipulate our own data but what about incoming data?
True we have a means of influencing verification by adding a new taxon and setting this as the preferred ID but we have little else.
For example I have an Event sent in by a contributor which has the wrong grid reference to the named Location. I need to somehow isolate this Event for a variety of reasons: to stop the Location/grid reference combination from being used/reported, to exclude it from my Location hierarchy system (County then District then Parish - as used by a few LRCs) as wouldn't insert correctly, to prevent wrong reporting.
Because it's someone else's Event + Sample + Taxon occurrences my options for carrying out validation tasks are limited, all I can do is "change" the determination of each taxon occurrence.
What I would like to be able to do is:
1. Mark the Event as invalid
2. Mark the Sample as invalid
3. Mark the Location as invalid AND/OR non-preferred (e.g. whilst the name "LEICESTER" might be legitimate as an historic location, it's not something we want people to use nowadays)
4. Be able to select a command on right clicking the offending Event which allowed me to send an email to the originator which a) detailed what is wrong, b) offers a quick fix on their database via an import
5. Be able to keep track of "quarantined" items (which of course never ever find their way into reports) - perhaps via the Filter - so that they can be later corrected.
People are increasingly looking to LRCs for guidance on location naming. This is not going to be possible while we still have to cope with incoming errors that we cannot fix readily.