Data entry is really cranking along now! At present, we have 198 verified, 144 unverified, and around 30 entries to be compared and added to the verified list (more on this below). This adds up to around 375 total entries in the system at present – nearly double of what we had at the launch of the project. Good work, everyone!
As you may recall, we have a two-step data entry system. The first, or original entry, is compared against a duplicate entry (verification entry) typed independently by another individual (hence the multiple spreadsheets available for download). Although this is time-consuming for all involved, the extra level of confidence in our data is completely worth it. If the data are completely identical between both independent data entries, the data immediately go onto the Public Data Sheet. If there is a discrepancy between the original entry and the verification entry, we have to go back to the original paper and find the source of the confusion. At present, I would estimate that around 50 percent of all entries need to have at least some detail or another resolved before they can be checked off the list. These problems range from quite minor to rather major. Resolving the problems takes time and delays the posting of the data (sometimes by a few days, as we don’t always have time to immediately track down the relevant references).
As I’ve worked on comparing original vs. verification entries, I’ve noticed a number of common errors that we all (including me) make. In an effort to eliminate delays caused by such errors, I thought I would highlight typical problems:
- Failure to identify the side (left or right) that an element belongs to, or incorrectly assigning it to a side. In general, go by what the data table says. If the data table doesn’t indicate the side the element belongs to (left or right), but the text does, indicate this in a note to me in the email. This way I can more rapidly identify the source of the problem when comparing with other data entries. Make sure you type the entry into the correct column!
- Re-identifying taxonomic assignments. I know and you probably know that many specimens once called Trachodon are now lumped into Edmontosaurus. But, please use the paper’s original taxonomy. If they call the specimen Trachodon annectens, use this name. If it really bugs you, just send a note along with your data entry. We will sort it all out at the end, before data analysis.
- Incomplete entry of data. In large monographs (lengthy, detailed descriptions of specimens), data tables may be scattered across several widely separated pages. Make sure you check out the whole publication, to ensure that you find all of the relevant measurements. Sometimes, measurements are hidden away in the text, rather than appearing in convenient table form.
- Using an out-of-date spreadsheet. The Verification List is updated nearly daily, and the New Data Entry spreadsheet has been updated once since the project launch. Make sure you are using the latest versions!
- Forgetting to put your name in the appropriate column. We want to make sure you get credit for your hard work, so please remember to enter your full name on the data sheet!
The scientific success of the ODP, and the acceptance of our project within the scientific community, will hinge largely on a clear, accurate, and useful database. Working together, we’re making this a reality!