eBird Data: Review Process

"Thank you for using eBird. Your participation is valued, and the observations you submit are an important part of helping us better understand birds and biodiversity in your area and around the world."-Andrew S. Aldrich, Illinois Regional Data Reviewer

Cedar Waxwings (Bombycilla cedrorum), Nov. 12, 2019

On Nov. 15, 2019 I received an email from one of eBird's regional data reviewers.  He was reaching out about a submission I made for 24 Tufted Titmouse (Baeolophus bicolor) on Nov. 12, 2019.  It was nearing sunset when I made the observation.  The light wasn't favorable for identification, but I could clearly see the crests on their heads and see that their bodies were predominantly gray in color.  Tufted Titmice are year-round residents in northeast Illinois, so I made the submission despite the high count alert I received prior to submission.  Fortunately, the eBird reviewers are much more experienced than me.  Mr. Aldrich informed me that while Tufted Titmice are present in this area, it would be highly unusual for them to travel in such a relatively large flock.  Cedar Waxwings, however, are known to travel in groups of this size - especially in the fall when they are looking for berries.  They also have a prominent head crest and mostly gray in color.  What I appreciated most about the email was the explanation Mr. Aldrich provided.  He didn't just ask me to change my submission, but provided the reasoning which will help me discern between these two species in the future.

Up to this point, you may have wondered how eBird ensures the integrity of the data submitted to their site.  After all, anyone and everyone is allowed to participate and there is no formal training or minimum experience level required.  In truth, there is a chance that inaccurate data finds its way into the eBird database.  Media evidence (images or audio) is only optional, not required, and it's easy to mistakenly identify a species or the total number observed.  Fortunately, eBird's alert system helps to mitigate these issues.

As soon as a submission is made, any rare or unusual species or unusually high counts of common species are sent off to a regional data reviewer for confirmation.  As I explained above, the eBird site also alerts the person prior to submission.  If you choose to continue with the entry, eBird prompts the user to attach media.  This is a really helpful feature, as the internal alert system has allowed me to correct a few species prior to making submissions.  On one checklist I was reporting a Golden-crowned Sparrow, but an alert that it was rare for the latitude/longitude provided and the time of year, I continued researching to find that it was actually a Golden-Crowned Kinglet. On another checklist I was reporting a Little Blue Heron and eventually figured out that it was a Great Blue Heron.  On another, I was going to report 6 Marsh Wrens but the alert allowed me to properly identify them as American Tree Sparrows.  Now, this isn't to say that anything flagged as rare or unusual is wrong.  They have been in my case, but I'm still familiarizing myself with the birds in my area and learning how to properly identify them.  Many users successfully report rare birds that are then confirmed based on the media they provide, such as those found on the Illinois Rare Bird Alert list.

Another fail-safe in the eBird database is the ability for other users (not just reviewers) to flag or report birds that have been misidentified.  As I've explained in an earlier post, data is available to all users as soon as it's submitted to eBird.  This means that I could also comb through other eBirder's submissions and report an error.  Others could do the same with mine.  I think this particular design feature is great, especially because there is no other vetting process for data that doesn't trigger an alert.