One of the great things about eBird is the data quality standard and review of unusual bird reports.
When submitting a checklist to eBird you are asked to verify unusual sightings. Perhaps there was a typo and you accidentally entered a high number or wrong species. However, don't just mark the checkbox verifying the entry is correct as you intended--give a reason.
You see, every time the automatic filters indicate you saw an unusually high number or unusual species, a real person reviews your list. Rather than waiting for the Reviewer to send you an email asking for more details, provide the details in the comments for any species that eBird flags as unusual. If you don't provide comments in your checklist and then don't respond to the eBird Reviewer's email request for more information, Reviewers have no choice but to invalidate your record.
Now it may be that the number you saw was just barely over the threshold, or the species is generally rare but your bird is a known rarity. In this case the Reviewer will simply verify your sighting from within eBird and your record gets accepted to the pubic database.
Whenever possible, the Reviewer is a local expert on the birds of the area. If they deem that your sighting is unusual enough, they will want more details. Plumage, behavior, habitat, songs and calls all help verify a locally rare species. Provide as much information as you can. Of course, digital photos are ubiquitous these days. Even a blurry photo from your camera-phone can help establish the identity of a species.
On the other hand, perhaps the species is expected, but in much lower numbers than you reported. Explain any reason for unusually high numbers.
Now the filter settings aren't perfect. The Reviewer is able to change the settings to more accurately reflect reality. If you think a species or high number is being flagged too sensitively, add a note to that effect to your comments as the Reviewer will read it. On the other hand, if you entered a bird you thought rare, but the filter accepted it without challenge, then you might also drop an email line to your Reviewer.
Reviewers are assigned by counties, though most reviewers are responsible for more than one county. They can help you with ID questions, status and distribution. They are a good resource for you to learn the local birds. They may suggest that you saw a more common bird. But remember, no one can change your records but you.
So, what happens if the Reviewer doesn't overturn the automatic filters and accept your report?
Firstly, the acceptance or not of a reported bird is not a reflection of you as a person or as a skilled birder. It doesn't mean you didn't actually see the bird you reported. It does mean that the bird was unexpected and you provided insufficient documentation to sway the decision. I have invalidated my own records when entering lists of birds from years long past because of lack of written details in my notes about an unlikely species.
Secondly, your list of species is always available to you, whether correctly identified or not, whether the Reviewer validated it or not. Remember, all records stay in the system and can be reviewed again. Invalidated records can be accepted and accepted birds can be reviewed and invalidated later.
What will happen though, is that your invalidated record will not appear on the public eBird maps and bar charts for abundance and frequency. Your records are available, though, to researchers, along with the Reviewer's comments, so they can make their own decisions.
I hope this provides some insight into the data quality of eBird and how providing comments on your checklist can save time and effort and help a Reviewer come to a good decision about flagged records.
More information:
http://ebird.org/content/ebird/news/a-comment-about-comments
http://ebird.org/content/ebird/news/ebird-data-quality
http://ebird.org/content/wi/news/theimportanceofcomments