BirdForum is the net's largest birding community dedicated to wild birds and birding, and is absolutely FREE!

Register for an account to take part in lively discussions in the forum, post your pictures in the gallery and more.

Reply to thread

Message: <blockquote data-quote="jurek" data-source="post: 3669484" data-attributes="member: 3357">Your story of Willow Tits is a perfect example of very common mistake coming in publications which used big data: bad results based on little data are published along good results based on good data. And at the end, both are treated as good.I guess the project had lots of valid data points for common species like, say, Robin. For species with little data, like Willow Tit, errors were sometimes sufficent to confuse the picture. But good and bad was published together, and some reader will later pick only Willow Tit and believe nonsense. Unfortunately, many 'assurances' which people believe about big data points are not true:- More data does not mean less mistakes. It simply means propagation of errors to a bigger magnitude.- That most data is true, does not mean that minority of bad data will not spoil results.- That somebody verifies data by eye and removes ones which look bad is not making data much better. Because one is really producing data set based on his pre-conception. When one finds something new in that data, how do we know whether it is real or an error not filtered because was not expected? When one does not find something, maybe it was hidden by mistake? I wanted once to write some article or something to a bird magazine 'what mistakes people make when using big databases of bird records' but never got to it. So this little post must be enough...But coming back to the humble topic, whether a birder can trust ebird when looking for a new bird. My private rule of thumb is to trust the place to go for a new bird only if there are at least 5 records coming from at least 4 different people, and if there is more similar places with the same bird around. Unless I know the birder by name.</blockquote>

Verification