Topic: Data Quality

I have just looked at the records for Desmidiaceae on the Atlas.
There are 48 records from the Encyclopedia of Life and not one has a location or date and all have an identification verification status of "unconfirmed".
This is a mockery, an insult to all those who strive to supply data of the highest quality and explains why the reputation of the Atlas is so poor in certain circles.

Please can we move away from this obsession with numbers, place the emphasis on data quality and apply some minimum standards before datasets are added to the Atlas.



Re: Data Quality

Hi Christine,

Thank you for your comments. The records from EoL are images, they don't have any location or date information, and I am slowly removing them as we get better replacements. I do wonder whether we should just delete them all and not have so many images on the Atlas.

I spend a great deal of time responding to concerns about errors in the data on the NBN Atlas, I do place an emphasis on data quality and I note your concerns about having images without location and date information.

Many thanks,

Best wishes, Sophia

Sophia Ratcliffe
NBN Atlas Data Manager


Re: Data Quality

Hi Sophie,
This was not meant as a personal criticism, more a statement of despair and frutration.
We appreciate the efforts you have made to improve data quality and the submission process.
Data quality is the responsibility of the members submitting datasets, therefore the onus is on us to ensure that standards are maintained.

Until it can be demonstrated that the data on the Atlas is of high quality, you will not convince those who have a low opinion of the Atlas otherwise. So perhaps the PR machine can be directed towards quality rather than quantity.

Best wishes



Re: Data Quality

Better still, ditch EOL altogether. It detracts from data quality in other areas. For example by showing a picture of the wrong species (try Micropeza lateralis which shows an image of Micropeza grallatrix - a south mediterranean species)
How many other EOL pictures are wrong?
These images on EOL were most certainly not assigned by workers with any kind of international reputation or even a rudimentary knowledge.
Isn't there a thread somewhere else where we grumbled about this EOL junk?
Substitute with NHM Scratchpads


Re: Data Quality

Hi Darwyn,
Thanks for your comments, we will remove all the EoL images from the NBN Atlas.

Thanks, Sophie

Sophia Ratcliffe
NBN Atlas Data Manager