Making sure you’re using the newest filters & an update

We are nearing a major advance in the Michigan eBird filters: the creation of about 40 new filters. Here are the existing 26 filters for MI as described in our last blogpost:

ebirdfiltermap04042014

and here are the ones that will result from our upgrade:

ebird filter07232014

This change will result in greatly fine-tuned filters for each county/group of counties, streamlining the data entry system and minimizing annoying unnecessary flagging of records. But in order to fully take advantage of this upgrade, it is important to make sure you’re not inadvertently using older, outdated versions of the filters. Here’s how it happens: If you are using BirdLog to enter your data, during data entry it asks you to tell it where you are birding with this dialog:

photo

When you use “Choose a Recent Location” or “Create Offline Checklist,” the app accesses a cached version of the filters from the device’s hard drive. This way the device doesn’t have to waste time communicating with the cell tower or WiFi connection. This makes the process much faster and more convenient, no doubt, but you get an outdated version of the county checklist for which to enter your data (assuming changes have been made to that filter since you last used it)! In order to avoid this, simply use one of the other categories: “Choose a Location From Map”, “Create a New Personal Location”, “Choose a Nearby Hotspot”, “Choose Hotspots by City”, or “Choose a Nearby Personal Location.” Any of these will force your device to use the cell tower or WiFi to download and display the current filter, refreshing it with the current version. *Note: this issue does not apply to data entry on ebird.org, which automatically accesses the internet to generate its checklist.*

As these new splits are implemented over the next several months, Team eBird Michigan will begin the process of tweaking each filter to most accurately represent that county’s nuances of bird distribution and abundance. As those of you with local expertise notice improper species or counts flagging, we would appreciate your suggestions for fixing the filter. Please put these comments *in the species comments section of the species that flagged*. All users are still asked to submit photos and detailed descriptions even when you think the filter is too strict. But these notes are very helpful to us reviewers who are attempting to get things just right, and often don’t have time to bird each county at the level required to understand each peculiarity of bird distribution there. Here are two examples of helpful suggestions, the first of summering Redheads at St. Clair Flats, St. Clair County, which flagged (but shouldn’t, and won’t after the split):

ebird1

and the second of post-nesting Acadian Flycatchers in Kent County, where we needed to allow for more than 2 individuals after July 10 (but were only allowing for 2 because singing behavior drops so precipitously during late summer).

ebird2

Both of these suggestions will be used to correct the local filter, and eventually all of Michigan’s filters will very tightly track the actual abundance level of each species in each area. So, please keep the species comments coming, and we hope you’ll enjoy the rollout of these souped-up new filters over the next many months!

Advertisements

One thought on “Making sure you’re using the newest filters & an update

  1. following our email conversation this weekend, I landed here in surfer mode. we remain puzzled by filters. filters seem to be based on historical observations – whether eBird only or many others as in literature is unclear. we know most of the literature is geographically smaller than either data sets or geographic distribution from eBird, but the filters make MANY assumptions and require the observer to vigorously defend against the filter (apparently whether the filter is valid or not). so it seems that the filters skew the data? do the filters skew the data even more than the data might be skewed by “observer error”? we wonder if eBird will eventually offer two portals to the database: the raw raw data and the reviewed data? we postulate that many of the filtered entries – especially those that are rejected – will eventually prove true. for example might be cardinals in New England. when we were but children, cardinals were rare even in northern CT and a cardinal considered a wildly absurd observation in rural Worcester Co MA. but now cardinals nest here in central ME 200 miles N and even 20 miles inland from ocean moderated weather. does filtered eBird pick up the advancing front of new species arrivals / range extensions or because of the filters does recognition delay by some years? same argument for CBC.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s