Tweets, Sources & Links

Social media has been an important part of the Puppycide Database Project since our inception. We aren't concerned with going "viral" or getting a bazillion followers; but we are interested in making sure that the public has free and easy access to all of our research, that sources and volunteers can contact the Project and to make sure that our process of data collection is transparent. Part of accomplishing this goal is a small application our volunteers developed which automatically generates a message on the @PuppycideDB Twitter account. Every time anyone contributes to our database, the information in that contribution becomes immediately available for public review.

Of course, Twitter limits messages to 140 characters, which isn't a lot of text. The first paragraph in this post is 682 characters. So we can't post each record to Twitter or even a summary of the incident covered by the record. We decided to include a short message explaining a new record was created in our database, the name of the law enforcement agency or agencies involved, and a link to the source provided along with the records.

The result usually looks something like this:

What exacty is "a link to the source provided along with the records" ? This isn't the same thing as a link to the record itself. If you click the link in the Tweet above, you'll see that it navigates your browser to a website called archive.org that backs up copies of websites (many news organizations remove or modify articles posted online, or move articles to a private archive system, so we recommend using archive.org to ensure that reference links don't become broken).

PuppycideDB records themselves are also available online through our database search utility. Here is what the webpage containing the database record covering the Tweet above:

PuppycideDB entry for Harris County Precinct 5 killing of Sadie the dog

So why do we include a link to the source of the record as opposed to the record itself? For technical reasons, there is a delay of a few hours between when a record is created in our database and when the webpage for that record becomes available. This delay is very important - for example, it helps to ensure that our website can't be taken offline by someone creating lots of fake records. However, our Tweets go out immediately. If we were to include a link to the record web page itself in each Tweet, the links would not work for the first few hours after its creation. Twitter is constantly directing their users toward new Tweets, so this approach just wouldn't work unless we delayed the Tweets also (which would cause other problems, such as sending out a large batch of simultaneous Tweets each day).

This approach has worked well since it has been set up, and has helped bring attention to police shootings of pets. However, recently, our automated Tweets started having trouble:

A large number of our automated Tweets were being created with broken links. The resulting messages clearly made it look like something was messed up:

Example of a broken link in a Twitter message

Instead of a clickable link, these busted Tweets had a "http://" at the end, followed by nothing. What's the deal here?

Within the Puppycide Database Project's new record submission form, we ask for (and strongly encourage) volunteers to provide a link with each submission. Here is how the issue is phrased on the form:

Puppycide Database Project source URI field in new record submission form

Notice how the field is automatically populated with that "http://" prefix? This is where the mysteriously blank "http://" in the Tweets were coming from - volunteers submitting new records without a link to a source, and our programming not checking to make sure that Tweets with missing links would be handled differently than Tweets with working links. We have now updated our programming to get rid of those dangling "http://" prefixes in automated Tweets.

Its worth pointing out that there are many fields in the new record submission form that are required for submission. Fields like date, law enforcement agency, number of animals involved, and many others are all required. So why don't we also require a link to a source?

This question was the source of no small amount of discussion when the form was first built. And in fact, PuppycideDB always welcomes feedback on our approach to collecting information, so feel free to contact us and let us know about any ideas you might have on this issue. Ultimately, we opted to keep these field optional because so many of the sources that we have for records are not available online.

Puppycide Database Project records are collected in a number of ways. When we first began, nearly all of our records were collected from press accounts of police shootings of dogs that were available online - often from multiple online sources. PuppycideDB has compiled thousands of records using these types of sources. The longer our investigation continues, the more we have become aware of the fact that the press only publicizes a small number of police violence toward animals. We still have work to do before we can say for certain how wide this disparity is, and how that disparity changes based on factors such as what kind of animal was involved and where the incident occurred. However, we do now that such a disparity exists by acquiring complete use of force reports for a significant number of police departments and looking for press reports of use of force incidents involving animals. Interested readers can review PuppycideDB records for the Los Angeles Police Department to get an idea of how wide this disparity can be.

Offline sources of information are important, and often more reliable than online sources of information. In addition to use of force reports, PuppycideDB also relies on court documents - for example, depositions from use of force lawsuits brought by pet owners against municipalities - as well as eyewitness interviews provided to us by pet owners, neighbors and passersby. Including information gleaned from these types of records improves the quality and breadth of PuppycideDB research, however it may not be possible to link to such information, at least not immediately. A large part of the Puppycide Database Project that many people are unfamiliar with is our growing reference library, in which we attempt to consolidate these types of research sources that are relevant to police violence involving animals. When we are able to add references to our own library, we can update existing records to include a link to the library in our records. When we first created the LAPD records discussed earlier in this post, we were unable to provide a direct source ink. Eventually, we obtained copies of and subsequently uploaded 10 years of LAPD police involved shooting reports to our reference library. This allowed us to update all of our database records relying on these sources to use the new link in our library.

To sum things up: the Puppycide Database Project relies on a variety of sources to research and document incidents of police violence toward animals. For our volunteers, researchers and sources: when creating a new record, please do your best to include a link to a reputable source documenting the incident described in your contribution. If such a link is unavailable, that's OK! PuppycideDB is here to assist with researching these types of incidents. If reliable and publicly available documentation of each time a police officer killed an animal existed, there would be no need for the Puppycide Database Project. For our readers, social media followers and researchers interested in the Puppycide Database Project: be sure to check out our database search page for more information about incidents described in our Twitter feed. Also, don't forget that entire copies of the database including all of our data are available absolutely free of charge, either upon request, or through our GitHub page. And to everyone: thank you for your interest in the Puppycide Database Project! With your help we can better understand the causes of police violence involving animals and how to reduce the rates of this type of violence as much as possible.