Talk:Fatal dog attacks in the United States/Archive 12

Identification of breed(s) involved in incidents
The dog breed(s) or dog type(s) identified in the "Category of Dog" field should match the breed information provided in a news or law enforcement source. In the case that there is a variance (or disagreement) of the breed(s) involved in one or several news sources, then both breeds should be listed in the "Category of Dog" field. For example, if one news source describes the dog as a "mixed breed dog" and another news source describes the dog as a "Mastiff", then the "Category of Dog" field should include both descriptions as: "Mastiff or mixed breed dog". In this case, both news sources should be provided as references - one source to substantiate the breed as a "Mastiff" and the second source to substantiate the breed as a "mixed breed dog". Note: In the case that a dog's description includes multiple (three more) breeds, then by definition it is a mixed breed dog and should be listed as a "mixed breed dog" in the "Category" field; however, the various breeds (if known) can be described in the "Circumstances" field.

When citing a specific study using their statistics on breeds involved should be used. The CDC study specifically said 56% of fatalities were pit bull and 10% rottweiler. Their finding has been removed. Crypticruin (talk) 15:18, 20 March 2019 (UTC) — Preceding unsigned comment added by Jacobm co (talk • contribs)

ANI
For transparency, Nomopbs has made a filing on ANI at WP:ANI, which in part relates to this discussion. PearlSt82 (talk) 13:36, 27 March 2019 (UTC)
 * if you and would be interested in a third opinion after the ANI threads closes and the discussion above is still 3PO eligible (i.e. no other editors have entered the dispute) feel free to ping me and I will offer up something. Best, Barkeep49 (talk) 14:22, 29 March 2019 (UTC)
 * Hi, the ANI was archived with no action. I would still be interested in a third opinion on this. To restate, the issue is: "Should the "Studies" section contain a bulleted list of primary studies, or a summary of the secondary literature in prose format?" Let me know if this is sufficient, or if you'd want me to formally relist on the WP:3PO page. PearlSt82 (talk) 15:03, 10 April 2019 (UTC)
 * Yeah go ahead and relist - I'm busier on Wiki than I was and am not sure I'll be able to get to this as promptly as I'd like. But I've added it to the "list" and will do it if no one else has. Best, Barkeep49 (talk) 15:47, 10 April 2019 (UTC)
 * Thanks, this has been relisted. PearlSt82 (talk) 16:12, 10 April 2019 (UTC)
 * You need to clear up the meaning of 'primary' and 'secondary' study per WP:MEDDEF because you have it wrong. There are no primary studies mentioned. They are all secondary.
 * From what I can see, in the fatal section, secondary studies listed are:
 * Breeds of dogs involved in fatal human attacks in the United States between 1979 and 1998
 * Co-occurrence of potentially preventable factors in 256 dog bite-related fatalities in the United States (2000-2009)
 * Dog-Bite-Related Fatalities -- United States, 1995-1996
 * Fatal dog attacks, 1989-1994
 * Human fatalities resulting from dog attacks in the United States, 1979-2005
 * Traumatic deaths from dog attacks in the United States.
 * Primary studies include:
 * Human Deaths Induced by Dog Bites, United States, 1974-75
 * Mortality, Mauling, and Maiming by Vicious Dogs
 * Studies that do not meet WP:RS include U.S. Dog Bite Fatalities Breeds of Dogs Involved, Age Groups and Other Factors Over a 13-Year Period (2005 to 2017)
 * In the non-fatal section, these should all be removed, as this article is about fatal dog attacks. Dog aggression or Dog bite might be the better place for this information. Its not immediately clear why these two should be separate articles, but thats another discussion. In this section, the only secondary study is A Review of Dog Bites in the United States from 1958 to 2016: Systematic Review of the Peer-Reviewed Literature, the rest are primary studies.
 * Additionally, this source from the AVMA should be used, and priority and weight should be given to the most recent studies, as some of these go back to the 1970s. Since a lot of these come from the plastic surgery field, as opposed to the animal behavior/veterinary field, there might be others that are appropriate as well. PearlSt82 (talk) 17:22, 10 April 2019 (UTC)


 * You are incorrect. 24 of the 25 studies are NOT primary, they are either secondary or tertiary. And the AVMA reference is irrelevant to the topic, and is NOT a study nor a literature review. Using the term "literature review" on the webpage doesn't make it one in fact. According to this writer, it's a lobby pamphlet masquerading as scientific writing. Nomopbs (talk) 00:01, 11 April 2019 (UTC)


 * Please continue the discussion in the correct section. Nomopbs (talk) 00:01, 11 April 2019 (UTC)

Vandalism. Anyone have rollback privs?
IP editor 2601:647:4001:6533:f4fd:3d20:2db0:6055 just made 8 edits to Fatal dog attacks in the United States which seem like vandalism. It looks like WP:vandalism to me. Does anyone have WP:Rollback privileges to revert these 8 changes?
 * In the section "Study: Centers for Disease Control and Prevention 2000: 1979–1998" they added "Golden Retrievers" when those aren't even mentioned in that study. And they modified two sentences so now they are incomplete sentences.
 * In the section "Study: University of Texas Study 1982: 1966–1980" The editor swapped German Shepherds (12) and Dobermans (2) to be German Shepherds (2) and Dobermans (12), and likewise swapped St Bernards (8) and Rottweilers (1), so now all the statistics are wrong.
 * Nomopbs (talk) 22:19, 31 March 2019 (UTC)
 * You don't need rollback privs to remove something that was added by another user. Everyone has access to the WP:UNDO button.  WhatamIdoing (talk) 19:56, 12 April 2019 (UTC)

List of primary studies
I've removed the list of primary studies per WP:PRIMARY. All of these studies are listed and analyzed in this secondary source: https://www.avma.org/KB/Resources/LiteratureReviews/Pages/The-Role-of-Breed-in-Dog-Bite-Risk-and-Prevention.aspx and as Wikipedia is a tertiary source, we need to leave the analysis and listing of primary sources to reliable secondary sources, like the AVMA. This source should be worked into the article and also replace the content for the other 6 listed studies. PearlSt82 (talk) 13:10, 24 March 2019 (UTC)

Response about last night's edits by Nomopbs and this morning's reversions by PearlSt82
My entire work last night was spurred by the article's language "There are a limited number of studies concerning the number of human deaths caused by dogs" and "The most recent study of the epidemiology of fatal dog bites in the United States is... (a 2013 study)", when in fact there are numerous studies and that six-year-old study was no longer "the most recent study".

PearlSt82, you have misunderstood the purpose of my adding the studies in the manner in which I did. It was NOT to list them as a "reference list" (like one might see at the bottom of a study). It was to list them as relevant to the topic of "Fatal dog (and near fatal) attacks in the United States" and/or as a precursor to summarizing relevant bits to put into this article. Removal of my edits from last night constitutes disruptive editing.

Just because you don't like where you think I'm going with the studies is no reason to destroy my work at the outset.

I cannot believe that you argue this ref contains a review of all the studies/citations I added last night, since (a) that AVMA ref was written in 2014 and I included 12 studies written afterwards, and (b) that study isn't (and wasn't) even cited in the article!

I see you put back some text that was there previously (which I had removed last night). I removed it because it lacked citation, was original research WP:NOR, promoted a single POV, and the single citation that was there did NOT support anything said in the entire section. That's why I removed it. I guess you need a detailed explanation, which I'll give you.

Here is the old section I had removed. I could have just flagged it like this (but that would have been ridiculous):


 * There are a limited number of studies concerning the number of human deaths caused by dogs in the United States, and the number of attributed fatalities is difficult to validate or cross-compare with other study results because:
 * studies dependent on surveying news reports may not find all the relevant news items
 * different studies use different data collection methodologies and evaluation criteria
 * breed identification requires a subjective evaluation based on observation of outward physical attributes against imprecise breed standards.

Explanations in more detail:


 * Old section included "There are a limited number of studies concerning the number of human deaths caused by dogs in the United States". Six had been listed, I had included 4 more. I suspect there are even more I didn't include. Therefore, there is NOT a limited number of studies concerning..."


 * Old section included "the number of attributed fatalities is difficult to validate or cross-compare with other study results" which violates 'no original research' policy. It is clear from my reading of multiple studies, that the study authors have had no trouble validating fatalities (including back in the 1900s) nor trouble discussing results of 'other studies' or ignoring them (because it's not their business to cross-compare studies).


 * Old section included "because studies dependent on surveying news reports may not find all the relevant news items" which is false or misleading. Even in the earlier studies (pre-internet), the authors searched and consulted multiple databases in order to locate as much of the information as possible. You have no idea if deaths were missed, or how many or few that may have been, or how it may or may not have changed the conclusions of the study. To mention it is a logical fallacy.


 * Old section included "because different studies use different data collection methodologies and evaluation criteria" is irrelvant. The authors don't seem to have any trouble with it.


 * Old section included "breed identification requires a subjective evaluation based on observation of outward physical attributes against imprecise breed standards" violates "no original research". It is trying to introduce a point of view based on other studies related to shelters and dogs and not related to the studies on fatalities and injuries. One cannot refute information in multiple studies with a broad statement such as this without a citation.

Nomopbs (talk) 16:05, 24 March 2019 (UTC)
 * The old section used this reference as a citation, but it is irrelevant to the topic. It does not support a single sentence or concept in the old section. In fact, that webpage is a list of links to other articles. I checked several which seemed potentially supportive and I didn't find any.

by PearlSt82

 * This is not disruptive editing. I agree the dogsbitelaw link is not a good source. However, the following four sources:


 * AVMA #1 - https://www.avma.org/KB/Resources/LiteratureReviews/Pages/The-Role-of-Breed-in-Dog-Bite-Risk-and-Prevention.aspx
 * AVMA #2 - https://avmajournals.avma.org/doi/abs/10.2460/javma.243.12.1726
 * CDC - https://web.archive.org/web/20151116130816/http://www.cdc.gov/homeandrecreationalsafety/images/dogbreeds-a.pdf
 * American Journal of Sociological Research - http://article.sapub.org/10.5923.j.sociology.20130302.02.html


 * all concur that breed identification is difficult and inaccurate. I would support the removal of the clause "There are a limited number of studies concerning the number of human deaths caused by dogs in the United States" because the AVMA source cites 65 sources, and as you note, there are several recently published studies on top of that. I would propose the first sentence of the paragraph to read "The number of attributed fatalities is difficult to validate or cross-compare with other study results because:" and the bulleted points listed as written in the previous revision.
 * The list of primary sources is still inappropriate regardless of citation. From WP:IRS: "Articles should be based on reliable, third-party, published sources with a reputation for fact-checking and accuracy. This means that we publish the opinions only of reliable authors, and not the opinions of Wikipedians who have read and interpreted primary source material for themselves." - ie we need secondary sources discussing this material. PearlSt82 (talk) 16:48, 24 March 2019 (UTC)

By Nomopbs
I guess that would make your AVMA ref a quaternary source — or in Wikipedia lingo, that would still be a tertiary one, since tertiary seems to include all composite/synth type sources. Nomopbs (talk) 17:19, 24 March 2019 (UTC)
 * WP:PRIMARY in this case would be a victim or victim's family account of a mauling.
 * WP:SECONDARY in this case would be a news report of a mauling.
 * WP:TERTIARY in this case would be a study that is comparing news reports of maulings.

by PearlSt82
Primary sources in this case would include: victim testimony, news reports of the attack, and studies that focus on a specific set of data like the University of Texas Health Science Center study - this only analyzes attacks from a single trauma center. A secondary source would be a study of the studies, like the first AVMA source. See WP:MEDDEF for a further breakdown. Since this is a list article, primary sources are fine for documenting individual incidents, but any analysis of trends in DBRF, claims regarding factors in attacks, etc, need to come from secondary sources. There is obvious overlap between this area and the Dog bite article, so it might be worthwhile to review both and see where content can be merged. PearlSt82 (talk) 17:29, 24 March 2019 (UTC)

By Nomopbs
I agree there may be overlap, and I'm working on both subjects. In the meantime, you have to allow time to actually "get it done" and quit poking bears and let the bear do some work instead of engaging in endless conversations about work in progress. Nomopbs (talk) 17:41, 24 March 2019 (UTC)
 * Discussion and consensus are part of the Wikipedia article building process. Any controversial or lengthy changes likely to be contested should be discussed per WP:BRD. PearlSt82 (talk) 17:51, 24 March 2019 (UTC)
 * Third_opinion has been requested. PearlSt82 (talk) 19:06, 26 March 2019 (UTC)


 * Yes, this was my thinking as well. PearlSt82 (talk) 21:29, 10 April 2019 (UTC)
 * Yes, discourage, not prohibit. In this case, however, there is only 1 primary study out of 25. PearlSt82 erroneously seems to think that almost all of them are primary studies. Nomopbs (talk) 23:32, 10 April 2019 (UTC)
 * No, I believe you have misunderstood how MEDRS defines primary and secondary studies. Those are primary studies. Bondegezou (talk) 06:49, 11 April 2019 (UTC)

Chart of studies showing they are not primary studies (by Nomopbs)
I have a copy of every single one of those studies and have pasted into this table below the language from each study that indicates whether or not it would be categorized per WP:MEDRS as a primary, secondary or tertiary source. There is only one out of these 25 which could possibly be considered a primary source.

On the other hand, the AVMA publication PearlSt82 presents to replace all of these is (a) not a study, it's a summary, and (b) does not have an author. "This peer-reviewed summary has been prepared by the American Veterinary Medical Association Animal Welfare Division. While principally a review of the scientific literature, it may also include information gleaned from proprietary data, legislative and regulatory review, market conditions, and scholarly ethical assessments. It is provided as information and its contents should not be construed as official AVMA policy."

I can't imagine why a summary by a veterinary membership organization should supplant information from a dozen or more medical professionals on the subject of damage or death to the human body. The subject of human fatalities is more of concern to medical personnel than it is to veterinarians. Nomopbs (talk) 18:44, 10 April 2019 (UTC)

Discussion

 * Per WP:RS: "Articles should rely on secondary sources whenever possible. For example, a paper reviewing existing research, a review article, monograph, or textbook is often better than a primary research paper." - the AVMA source is a review article. Many of the studies below are reviews of patients from a single intake/trauma center, or are individual case studies, which would make them primary studies. Animal behavior is absolutely relevant to dog attacks, which is why organizations like the AVMA should be weighted heavily, as they aim to study the causes of dog attacks, not just the damage caused by them. PearlSt82 (talk) 18:50, 10 April 2019 (UTC)


 * A review of hospital records from one's own facility does NOT make a study primary. Nomopbs (talk) 23:36, 10 April 2019 (UTC)


 * The AVMA article is NOT a 'review article'. It doesn't EVALUATE the other studies. Instead, the AVMA summary criticizes the other studies and uses them as examples as if they're 'fake science' in pursuit of its own ideas of promoting the idea that dog 'breeds' cannot be categorized as aggressive or not (irrelevant to the subject at hand) and promoting the viewpoint that breed specific legislation is not a solution to dog bites. Again I state, the AVMA summary is an ANONYMOUSLY authored opinion piece that even the AVMA itself disclaims as "not AVMA policy". Tell me how that could possibly be a reliable source? Nomopbs (talk) 23:51, 10 April 2019 (UTC)


 * I am not the only one whose opinion of the AVMA post is negative in this context, and that it is not even a literature review. At this link, Alexandra Semyonova picks it apart from the title, through each paragraph, each claim, the citations used, the purpose, and through to the conclusion. She also writes: "This document is a lobbying pamphlet against BSL," and "The disclaimer note at the bottom of the [AVMA's] first page of this pamphlet is deceptive. It calls this document ‘peer-reviewed’, a term reserved for scholarly works, usually at submission to scientific journals. The term is misplaced here. This is a political pamphlet; it is not a peer reviewed scholarly article or a review of the scientific literature." Nomopbs (talk) 23:53, 10 April 2019 (UTC)


 * This is most definitely a review article. The entire preamble is as follows: "This peer-reviewed summary has been prepared by the American Veterinary Medical Association Animal Welfare Division. While principally a review of the scientific literature, it may also include information gleaned from proprietary data, legislative and regulatory review, market conditions, and scholarly ethical assessments. It is provided as information and its contents should not be construed as official AVMA policy. Mention of trade names, products, commercial practices or organizations does not imply endorsement by the American Veterinary Medical Association.". The fact that there is no single author is indicative that is in the voice of the organization. Alexandra Semyonova's website is not a reliable source, and her opinion on the matter is not enough to challenge the source. There is of course, always WP:RSN if you're looking to challenge this with more eyes on the matter, but in my opinion, given the fact that this is a review article of 65 studies, per wikipedia policy, it should be given highest weight on matters related to dog bite risk, especially when pertaining to breed. PearlSt82 (talk) 14:17, 11 April 2019 (UTC)


 * The preamble is irrelevant if it doesn't follow through with what it says. It is NOT a literature review. It doesn't talk about what is in the literature. It changes the subject to dog breeds and BSL, completely away from dog bite injuries and dog bite fatalities. And it ends with "its contents should not be construed as official AVMA policy." Enough said. Nomopbs (talk) 16:30, 11 April 2019 (UTC)


 * Alex Semyonova is not being used as an RS for citations on mainspace articles, but just to point out that I'm not the only one with the opinion about that AVMA article. Her writeup is an excellent argument that the AVMA article is "not science", is "not a literature review", and is an opinion piece drawn up to promote "no BSL". Her opinion should be no less of a valid viewpoint than mine or yours. Read it. Nomopbs (talk) 16:30, 11 April 2019 (UTC)


 * Unauthored articles most definitely are NOT scientific literature, studies, etc. and the AVMA disclaims that the writeup is their policy. Therefore your position that the AVMA writeup is the voice of AVMA is faulty logic. It's anonymous because NO ONE WOULD DARE PUT THEIR NAME ON IT and hence put their reputation on the line! Nomopbs (talk) 16:30, 11 April 2019 (UTC)


 * They put the exact same preamble on every one of their literature reviews, which also do not have an author listed. This is without question a scientific review article. PearlSt82 (talk) 16:34, 11 April 2019 (UTC)


 * Note also that the AVMA writeup is a self-published work. Refer to WP:SECONDARYNOTGOOD. Nomopbs (talk) 01:55, 12 April 2019 (UTC)
 * From the WP:SECONDARYNOTGOOD page:
 * It has a reputation for fact-checking and accuracy.
 * It is published by a reputable publishing house, rather than by the author(s).
 * It is "appropriate for the material in question", i.e., the source is directly about the subject, rather than mentioning something unrelated in passing.
 * It is a third-party or independent source, with no significant financial or other conflict of interest.
 * It has a professional structure in place for deciding whether to publish something, such as editorial oversight or peer review processes.
 * The AVMA meets all these with flying colors. They are the accreditation body for veterinary schools in the United States and are the publishers of the Journal of the American Veterinary Medical Association. This is absolutely not a self published source. PearlSt82 (talk) 11:19, 12 April 2019 (UTC)


 * You're joking, right? They hire someone to do a literature review, ask someone else to look at it, remove the author's names, take it as their own, add a cookie-cutter disclaimer, then self-publish it. And to boot, the subject matter is about breed bans and doggie psychology. For gawd sake, man, it claims toy breeds are the most aggressive! (Yet they don't kill humans.) That AVMA writeup is completely irrelevant to the topic of Fatal dog attacks in the United States. Perhaps you didn't notice the evaluation I did of the AVMA writeup. It's in the section below at . Nomopbs (talk) 13:34, 12 April 2019 (UTC)


 * The review discusses and cites numerous studies on dog bite related fatalities. It is completely relevant to the subject of this article. We can certainly take this to RSN if no other editors chime in here. PearlSt82 (talk) 13:43, 12 April 2019 (UTC)


 * I've filed this at RSN here: Reliable_sources/Noticeboard. PearlSt82 (talk) 13:57, 12 April 2019 (UTC)


 * The AVMA ref is about which dogs are aggressive, and is only tangentially related to a discussion about fatalities. Perhaps you should add it to the Dog aggression article. But it doesn't belong here, whether or not it's a RS in general terms. Nomopbs (talk) 16:04, 12 April 2019 (UTC)
 * Fatal dog attacks are mentioned in the following text: "If you consider only the much smaller number of cases that resulted in very severe injuries or fatalities,21,23 pit bull-type dogs are more frequently identified. However this may relate to the popularity of the breed in the victim's community, reporting biases and the dog's treatment by its owner (e.g., use as fighting dogs21). It is worth noting that fatal dog attacks in some areas of Canada are attributed mainly to sled dogs and Siberian Huskies,56 presumably due to the regional prevalence of these breeds. See Table 1 for a summary of breed data related to bite injuries." and "It should also be considered that the incidence of pit bull-type dogs' involvement in severe and fatal attacks may represent high prevalence in neighborhoods that present high risk to the young children who are the most common victim of severe or fatal attacks. And as owners of stigmatized breeds are more likely to have involvement in criminal and/or violent acts46—breed correlations may have the owner's behavior as the underlying causal factor.". This is direct discussion of the issue. I agree that it can, and should also be used in the Dog aggression article. PearlSt82 (talk) 17:18, 12 April 2019 (UTC)
 * Additionally, from WP:MEDPRI: "Primary sources should not be aggregated or presented without context in order to undermine proportionate representation of opinion in a field. If material can be supported by either primary or secondary sources – the secondary sources should be used." and "Wikipedia policies on the neutral point of view and not publishing original research demand that we present prevailing medical or scientific consensus, which can be found in recent, authoritative review articles, in statements and practice guidelines issued by major professional medical or scientific societies (for example, the European Society of Cardiology or the Infectious Disease Society of America) and widely respected governmental and quasi-governmental health authorities (for example, AHRQ, USPSTF, NICE, and WHO), in textbooks, or in some forms of monographs." The AVMA source would directly fall under a review article issued by a major professional veterinary society. PearlSt82 (talk) 19:00, 10 April 2019 (UTC)


 * 24 of the 25 studies ARE secondary sources. Nomopbs (talk) 23:42, 10 April 2019 (UTC)

Evaluation of AVMA writeup
Comparing the list of 65 studies cited in the AVMA writeup to the 25 studies I placed in the Fatal dog attacks in the United States article, there is only one -- ONE SINGLE STUDY! -- that appears on both lists. So, no, the AVMA study is NOT a review of the same literature as relates to dog bite related fatalities and/or injuries and certainly is not a substitute for the 25 studies currently in the article.


 * LOCATION relevance? 32 out of 65 studies are from countries outside of the USA. In Fatal dog attacks in the United States, we're interested in dog bite fatalities (and serious injuries) within the USA. Non-USA writings cited by AVMA: 2, 4, 6, 8, 1, 12, 13, 16, 17, 18, 22, 27, 28, 29, 30, 32, 33, 36, 42, 43, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 63


 * TIMELY relevance? 29 out of 65 are over 20 years old (pre-2000): 1, 2, 3, 4, 5, 6, 7, 8, 14, 16, 19, 21, 22, 35, 36, 38, 39, 40, 41, 43, 49, 50, 54, 57, 60, 61, 62, 63, 64


 * That leaves remaining only 16 studies that are USA-based and within the last 20 years (2000+): 9, 10, 15, 20, 23, 24, 25, 26, 31, 34, 37, 44, 45, 46, 47, 65


 * And since citations 34 & 37 are duplicates (whoops AVMA), that leaves 15.


 * Of these remaining 15 topics, 2 are about fatalities (23, 47), 7 about injuries or general dog bite information (9, 10, 15, 24, 25, 26, 65), and the remaining 6 (20, 31, 34&37, 44, 45, 46) are subjects unrelated to Fatal dog attacks in the United States such as behavioral information, shelter evaluations, and psychology topics like "The Case Against Dog Breed Discrimination By Homeowners' Insurance Companies", "Canine analogs of human personality factors", "Managing stigma of outlaw breeds: A case study of pit bull owners", and "Assessment, management, and prognosis of canine dominated-related aggression". These psychological, behavioral, and fiscal topics do not belong in a list of studies about injuries and deaths.


 * So a possible NINE (9) of the original 65 citations (14%) might be relevant as 'replacements' for the 25 studies currently posted.


 * Meanwhile, five years have passed since the AVMA writeup, and half of the studies I added (12 of 25) are more recent than the AVMA study.

So, no, the AVMA study is NOT a review of the same literature as relates to dog bite related fatalities and/or injuries and certainly is not a substitute for the 25 studies currently in the article Fatal dog attacks in the United States.

Nomopbs (talk) 01:15, 12 April 2019 (UTC)

I agree that the AVMA "study" is flawed. There are plenty of supposition and non-peer-reviewed references. They do not evaluate the dozen or so peer-reviewed published medical studies since 2009 from the US that clearly show, from multiple geographic locations and multiple medical disciplines, that pit bulls are significantly more dangerous, especially to children. I, along with others, have contacted the AVMA regarding this, only to be ignored. The good news is, the American Academy of Pediatrics is currently performing a meta-analysis of the pediatric studies. Should have something this year. Once this is published, things will change. I think how Wikipedia recommends secondary studies only is NOT what the medical professionals would recommend. By posting just one, I agree it could be skewed, but having a large number of studies from multiple geographic locations and medical disciplines that all show the same thing shows the reader the consistency of the data. Dapb567 (talk) 17:41, 14 April 2019 (UTC)


 * I respectfully disagree (not directly w/you but with these studies). What these studies have concluded is simply that there is a population of mixed breed "bully-type" dogs or "medium size dogs with a blocky head" that cause injuries. As the name "pit bull" is not a breed but a broader and somewhat ambiguous classification arbitrarily applied to a large population of mixed breed "bully-type dogs" or "medium size dogs with a blocky head", without valid DNA analysis (which to my understanding these studies fail to include), then any conclusions in these studies related to breed risk are to your point, suppositions and not scientific conclusions (see my comments on this topic along with a few referenced studies under the "This is a mess" section). If talking about the specific breed "American Pit Bull Terrier", then to my knowledge not one medical study contains a valid risk-based conclusion that accurately correlates a risk analysis to the American Pit Bull Terrier by DNA (and DNA is the most important part if trying to correlate risk with a specific breed). The studies I referenced in my comments under the "This is a mess" section below found that the majority of "pit bull type" dogs are in fact mixed breed dogs; furthermore, the "pit bull type" category contains 3 or 4 distinct breeds so again, unless a medical study properly distinguishes and classifies injuries by the 3 or 4 unique "pit bull type" breeds, any conclusion that contains the ambiguous "pit bull" category is at best, questionable (as "pit bull" can include a number of different breeds - both purebred and many mixed breeds). Per the peer-reviewed studies on the topic of breed identification and DNA, if proper DNA analysis was performed in the medical studies, most likely (this is my guess, but my guess is based on the data and conclusions in the many studies on breed identification and canine DNA) the breeds in the medical studies would be identified (by DNA) as: "mastiff/bulldog mix", "labrador/mastiff mix", "cane-corso/bulldog mix", "american staffordshire terrier/labrador mix", "bulldog/labrador mix" (aka "bullador"), and way too many more (over 50 mixes of various combinations of breeds, both "pit bull type" and not "pit bull type") that can resemble a "medium size blocky head dog" or "bully-type dog" to list, all of which are currently getting arbitrarily lumped into the "pit bull" category - it's the difference between identifying a "2 door car" (general type classification that can include many different car models) and a "Honda Civic" (specific breed/model). So again, without valid DNA analysis any risk-based conclusions related to breed are questionable at best and simply cause more confusion around the topic of bite prevention. And before jumping to conclusions about my comments, I'm not a "pit bull advocate" by any means - I consider myself simply an advocate for all dogs. I own 2 dogs - a purebred Dogo Argentino and a Cane-Corso mix - both wonderful dogs and both not considered "pit bulls" (although many people "think" they are something called a "pit bull" because of their appearance). My reason for commenting is to point out that a medical study is only as good/valid as its sources for data - and breed identification without DNA analysis is unreliable (and that's not my opinion, that's a scientific consensus) which will devalue the entire study. We should all focus on reducing dog bite incidents instead of these unscientific "battles" over which breed(s) are more "dangerous" than other breed(s) - that's pure folly and probably the source of a lot of unnecessary controversy on the "Fatal Attacks" pages. Any large/strong dog can cause serious injuries. Michael2468b (talk) 23:25, 14 April 2019 (UTC)
 * Dapb567, you might be interested in Ten simple rules for editing Wikipedia.
 * Michael, everything I've read about DNA testing for dog breeds indicates that it's junk science. I have never yet seen anything that says it's a good idea to use DNA breed identification tests for any purpose more serious than entertainment.  WhatamIdoing (talk) 05:45, 15 April 2019 (UTC)
 * Wow, well if the DNA is "junk science" then other methods must be even worse. I'll agree to disagree, as I don't think that numerous peer-reviewed studies on the subject of breed identification and canine DNA qualify as "junk science". Maybe some of the cheap "home test DNA kits" are less accurate and to your point, are only for entertainment - but comparing "home test DNA kits" to valid scientific DNA testing techniques performed in comprehensive canine DNA studies is not a valid comparison. On the contrary, I'd argue that studies that don't include canine DNA for breed identification are really only valid for "entertainment" purposes as breed identification without DNA testing has error rates of between 40-60% according to several canine DNA studies. Most scientific studies would consider an error rate of 10% unacceptable for a core/critical dataset (typically, they want a 95% confidence interval or better on critical data) - so for the studies that contain breed data without DNA, an error rate averaging 50% (confidence interval of only 50%) for a core/critical dataset (breed) would be considered grossly unacceptable by scientific standards and the scientific community. Michael2468b (talk) 07:11, 15 April 2019 (UTC)


 * This paper discusses identification regarding visual id vs DNA testing - the dog genome was mapped in 2005, described here and DNA testing is described as being a legitimate means of determining parentage by the AKC here, so I don't know if I would label the DNA testing element as junk science. PearlSt82 (talk) 14:12, 15 April 2019 (UTC)


 * In that study, in the section "Methods, DNA Analysis" the study authors wrote: "There were 130 American Kennel Club (AKC) registered purebreds in their database and the laboratory reported “an average of 84% accuracy in the first-generation crossbred dogs of known parentage”..." but AKC has over 200 breeds, so this study was performed with an incomplete database and only 84% accuracy. Then the study authors wrote: "The laboratory had in their database the AKC breed American Staffordshire Terrier but not any breeds identified as Pit Bull or American Pit Bull Terrier. Because of the common ancestry, historical reciprocal registrations, and similar morphology, we used visual identifications of American Staffordshire Terrier, Pit Bull, and American Pit Bull Terrier as matches to the DNA identification of American Staffordshire Terrier." They what?!?!?! And you think that's a reliable study? Or you think that supports Michael2468b's assertion that DNA samples can correctly and 100% identify the breed of dog? Nomopbs (talk) 14:37, 15 April 2019 (UTC)
 * I don't really understand what you're getting at here. I think what they mean by the American Staffordshire Terrier clause is that this study is dealing with AKC purebreeds and that American Staffordshire Terrier is the only "pit bull" type dog recognised as a purebreed by the AKC, but not all pitbulls are American Staffordshire Terrier. I don't really see how that would make this study unreliable - the main conclusions they come to is that visually identified dogs of unknown parentage often conflict when said dogs are DNA tested. There is of course always room for error in a dog of unknown parentage, but my understanding of DNA testing is it should allow for at least the identification of a purebreed, a mixed dog, and some kind of informed background on a mixed dog's parentage. PearlSt82 (talk) 14:45, 15 April 2019 (UTC)
 * I agree, not really sure what nomopbs is getting at - 84% accuracy is still vastly superior compared to obsolete visual breed identification methods which, if I remember correctly, average at best around 50% between all the numerous studies. Also regarding the methodology of matching several breeds (American Pit Bull Terrier, etc.) to the American Staffordshire Terrier breed (a breed commonly classified as a "pit bull type" breed), that methodology serves the purpose of controlling for bias towards those specific breeds (e.g. avoid undercounting their population) - in other words, they applied that methodology to the data to control for bias/error by avoiding the outcome of misidentifying a positive "pit bull type" breed (e.g. American Pit Bull Terrier) as not a "pit bull type breed". Not perfect, but still much better than using uncontrolled data from unscientific visual identification methods. Michael2468b (talk) 14:55, 15 April 2019 (UTC)

Chart comparing primary vs secondary (by Nomopbs)
This is a chart attempting to figure out the logic expressed earlier in the discussion, and what we should do about it.

Both Bondegezou and PearlSt82 interpret that a secondary study means "a study of a study", using WP:MEDRS as their guide. (If this is true, then I highly recommend someone edit WP:MEDRS/WP:MEDDEF and clarify where between primary and secondary should fall a "retrospective study of records of patients not ever personally seen by the doctor doing the study". I still assert such is a secondary source according to the first half of the secondary source definition. But today, let's go with only the second half of the secondary source definition.)

PearlSt82 mentioned elsewhere on this page, in section 'ANI' (an odd place, but whatever), a list of studies he interprets as primary or secondary. Using that list, and using the definition of "a secondary study means a study of a study", I've made an interesting comparison chart (see below).

Using the definition that "a study is only secondary if it's a 'study of a study'" and the idea that primary studies should not be allowed in the Fatal dog attacks in the United States article, then all of the studies (including all of the original 6 studies in the article) need to be removed and the only remaining studies left would be the two new studies "Abstract: A Review of Dog Bites in the United States from 1958 to 2016: Systematic Review of the Peer-Reviewed Literature" and "Dog Bites in the United States from 1971 to 2018: A Systematic Review of the Peer-Reviewed Literature" (which upon closer examination are actually the same study).

Therefore, only one study would remain: Dog Bites in the United States from 1971 to 2018 A Systematic Review of the Peer-Reviewed Literature (Bailey 2018).

Now, if that's how y'all see it, then I'll be happy to remove ALL of the studies (the 25 new and 6 old) and replace it with a writeup of just this one study. Vote on it? (Please add discussion or votes below the chart.)

Nomopbs (talk) 15:46, 12 April 2019 (UTC)

Invite for comment. Nomopbs (talk) 15:47, 12 April 2019 (UTC)

Invite for comment. Nomopbs (talk) 15:48, 12 April 2019 (UTC)


 * WP:MEDRS is clear. We should focus on secondary sources, meaning "studies of studies". A retrospective review of cases is a primary study. MEDRS does allow some use of primary studies, but if we've got two systematic reviews to use, then I would suggest we use those two and, indeed, drop reference to all the primary studies. MEDRS may seem a very restrictive policy, but there are good reasons behind it. Bondegezou (talk) 15:52, 12 April 2019 (UTC)


 * Agreed, with the addition that we should also be using the AVMA source discussed above. PearlSt82 (talk) 17:18, 12 April 2019 (UTC)


 * The AVMA ref is not relevant. Please answer on the scenario whether or not AVMA is allowed into the article. Nomopbs (talk)


 * From the abstract and conclusion, I don't see that the "Abstract: A Review of Dog Bites in the United States from 1958 to 2016: Systematic Review of the Peer-Reviewed Literature" review discusses fatalities. It seems to discuss "severe dog bites" and evaluations breed-specific legislation. However, I don't have full access to the paper. If you do, could you post any sections that are relevant to fatalities? PearlSt82 (talk) 18:10, 12 April 2019 (UTC)


 * Here's a challenge, I'll write up a sample for Bailey if you write up a sample for AVMA. Post below. Nomopbs (talk) 19:07, 12 April 2019 (UTC)
 * I would say the key takeaways from the AVMA source with respect to this article are:
 * Fatalities make up a small portion of total serious bite injuries
 * While pit-bull type dogs are more frequently identified in fatalities, this overrepresentation may be due to the popularity of the breed in the victim's community, reporting biases and the dog's treatment by its owner
 * the incidence of pit bull-type dogs' involvement in severe and fatal attacks may represent high prevalence in neighborhoods that present high risk to the young children who are the most common victim of severe or fatal attacks
 * factors relating to the individual animal (such as training method, sex and neutering status), the target (e.g. owner versus stranger), and the context in which the dog is kept (e.g. urban versus rural) prevent breed from having significant predictive value in its own right
 * These points can then be worked into a prose format. PearlSt82 (talk) 19:16, 12 April 2019 (UTC)


 * Do you see how many times the AVMA "study" (actually article is more appropriate) says "may be due", "may represent" when making up excuses regarding the over-representation of pit bulls?? Pure supposition and opinion. There is no data that supports this. This is not how conclusions should be made. Supposition is not science. Dapb567 (talk) 17:46, 14 April 2019 (UTC)


 * I have some ideas for a re-write. Rather than going with studies followed by information, how about information that's cited by studies? Let me work on a sample. brb Nomopbs (talk) 20:25, 12 April 2019 (UTC)


 * Nomopbs, you are being unnecessarily combative. If you wish to add a source to an article, it is your responsibility to demonstrate that it is reliable and relevant. From what I've seen, many of the sources you've added do not discuss fatal bites. Some of them mention fatal bites as an aside. Only those that discuss fatal bites in more details should be used in this article, because they directly address the main topic. Red Rock Canyon (talk) 19:25, 12 April 2019 (UTC)


 * The fatal (10) and non-fatal (15) studies were in separate sections. Didn't notice that, huh? Your explanations do not excuse removing a large chunk of the page as your first ever edit here. Then you call me combative for expressing my frustration that you stuck your edit fingers into the pie while we're discussing it and instead of just tasting it or screwing up your face in disgust... you simply threw it all out the window? Gone. Nomopbs (talk) 20:25, 12 April 2019 (UTC)

Rather than engaging specifically about any individual source at this time (because IMO we need to have a shared understanding of the page's subject before we work on the content), I will say instead: PearlSt82 is generally on the right track. There are some different systems for classifying sources, and PearlSt82's comments are generally in line with the way those terms are meant by the English Wikipedia, for the purpose of describing scientific and Biomedical information. But, again, this is probably something to be discussed later. WhatamIdoing (talk) 20:14, 12 April 2019 (UTC)


 * Thanks for the link... and the 'outdent'. I'm going to try a re-write into a completely different format. We'll see how that come across. Nomopbs (talk) 20:34, 12 April 2019 (UTC)

Standardization of breed names
Please use these guidelines to help keep breed names accurate, consistent, and organized. Currently, the main page contains a variety of different formats (capitalizations, hyphenations, and sometimes spellings) for breed names. For example, the Labrador Retriever is capitalized in a variety of ways including: Labrador Retriever, Labrador retriever, and labrador retriever. While there are a number of different suggested ways to capitalize and hyphenate breed names, the recommendation of this guideline is to follow the format that the American Kennel Club and United Kennel Club use which is to capitalize all names in a breed name, for example: Labrador Retriever. In the case of mixed breeds, the format is also to capitalize all breed names but not the word "mix", for example: Labrador Retriever-Husky mix. In the case of pit bulls, since "pit bull" is not a proper breed but a dog type (a category of dog consisting of at least 4 breeds as listed on the pit bull Wikipedia page), the proper way to capitalize pit bull is with both names lowercase: pit bull (unless it is the first word in a sentence, then it would be: Pit bull and "bull" should not be capitalized). If one of the proper pit bull breeds is identified, then all of the names in the breed name would be capitalized (like any other proper breed), for example: American Staffordshire Terrier, American Pit Bull Terrier, and Staffordshire Bull Terrier. If the number of dogs involved in an attack is more than one, for consistency it should be listed as: Breed (2). For example, if two German Shepherd dogs are involved, then it would be listed as German Shepherds (2).

Examples of breed name format, as recommended:
 * Labrador Retriever
 * Labrador Retriever mix
 * Labrador Retriever-Husky mix
 * German Shepherd
 * German Shepherd mix
 * German Shepherd-Husky mix
 * Doberman Pinscher
 * American Staffordshire Terrier
 * American Pit Bull Terrier
 * pit bull
 * pit bull mix
 * mixed breed dog
 * pack of dogs

Examples if two dogs of the same or mixed breed are involved in an attack, as recommended:
 * German Shepherds (2)
 * Doberman Pinschers (2)
 * mixed breed dogs (2)
 * pack of dogs (2)
 * pit bulls (2)
 * Doberman Pinschers (2), German Shepherds (2), pit bull mixes (2)


 * Would it be better to put this in some kind of infobox so the autoarchive bot doesnt keep pulling this? PearlSt82 (talk) 22:35, 19 April 2019 (UTC)
 * Yes, I think so since these are guidelines for the main page, but I don't know how to move them into an infobox. Jacobm co (talk) 22:52, 19 April 2019 (UTC)
 * There are several talk namespace template messages - maybe the "notice" one would be appropriate? PearlSt82 (talk) 12:51, 20 April 2019 (UTC)

Standardization of breed names
Please use these guidelines to help keep breed names accurate, consistent, and organized. Currently, the main page contains a variety of different formats (capitalizations, hyphenations, and sometimes spellings) for breed names. For example, the Labrador Retriever is capitalized in a variety of ways including: Labrador Retriever, Labrador retriever, and labrador retriever. While there are a number of different suggested ways to capitalize and hyphenate breed names, the recommendation of this guideline is to follow the format that the American Kennel Club and United Kennel Club use which is to capitalize all names in a breed name, for example: Labrador Retriever. In the case of mixed breeds, the format is also to capitalize all breed names but not the word "mix", for example: Labrador Retriever-Husky mix. In the case of pit bulls, since "pit bull" is not a proper breed but a dog type (a category of dog consisting of at least 4 breeds as listed on the pit bull Wikipedia page), the proper way to capitalize pit bull is with both names lowercase: pit bull (unless it is the first word in a sentence, then it would be: Pit bull and "bull" should not be capitalized). If one of the proper pit bull breeds is identified, then all of the names in the breed name would be capitalized (like any other proper breed), for example: American Staffordshire Terrier, American Pit Bull Terrier, and Staffordshire Bull Terrier. If the number of dogs involved in an attack is more than one, for consistency it should be listed as: Breed (2). For example, if two German Shepherd dogs are involved, then it would be listed as German Shepherds (2).

Examples of breed name format, as recommended:
 * Labrador Retriever
 * Labrador Retriever mix
 * Labrador Retriever-Husky mix
 * German Shepherd
 * German Shepherd mix
 * German Shepherd-Husky mix
 * Doberman Pinscher
 * American Staffordshire Terrier
 * American Pit Bull Terrier
 * pit bull
 * pit bull mix
 * mixed breed dog
 * pack of dogs

Examples if two dogs of the same or mixed breed are involved in an attack, as recommended:
 * German Shepherds (2)
 * Doberman Pinschers (2)
 * mixed breed dogs (2)
 * pack of dogs (2)
 * pit bulls (2)
 * Doberman Pinschers (2), German Shepherds (2), pit bull mixes (2)


 * Would it be better to put this in some kind of infobox so the autoarchive bot doesnt keep pulling this? PearlSt82 (talk) 22:35, 19 April 2019 (UTC)
 * Yes, I think so since these are guidelines for the main page, but I don't know how to move them into an infobox. Jacobm co (talk) 22:52, 19 April 2019 (UTC)
 * There are several talk namespace template messages - maybe the "notice" one would be appropriate? PearlSt82 (talk) 12:51, 20 April 2019 (UTC)

Standardization of breed names
Please use these guidelines to help keep breed names accurate, consistent, and organized. Currently, the main page contains a variety of different formats (capitalizations, hyphenations, and sometimes spellings) for breed names. For example, the Labrador Retriever is capitalized in a variety of ways including: Labrador Retriever, Labrador retriever, and labrador retriever. While there are a number of different suggested ways to capitalize and hyphenate breed names, the recommendation of this guideline is to follow the format that the American Kennel Club and United Kennel Club use which is to capitalize all names in a breed name, for example: Labrador Retriever. In the case of mixed breeds, the format is also to capitalize all breed names but not the word "mix", for example: Labrador Retriever-Husky mix. In the case of pit bulls, since "pit bull" is not a proper breed but a dog type (a category of dog consisting of at least 4 breeds as listed on the pit bull Wikipedia page), the proper way to capitalize pit bull is with both names lowercase: pit bull (unless it is the first word in a sentence, then it would be: Pit bull and "bull" should not be capitalized). If one of the proper pit bull breeds is identified, then all of the names in the breed name would be capitalized (like any other proper breed), for example: American Staffordshire Terrier, American Pit Bull Terrier, and Staffordshire Bull Terrier. If the number of dogs involved in an attack is more than one, for consistency it should be listed as: Breed (2). For example, if two German Shepherd dogs are involved, then it would be listed as German Shepherds (2).

Examples of breed name format, as recommended:
 * Labrador Retriever
 * Labrador Retriever mix
 * Labrador Retriever-Husky mix
 * German Shepherd
 * German Shepherd mix
 * German Shepherd-Husky mix
 * Doberman Pinscher
 * American Staffordshire Terrier
 * American Pit Bull Terrier
 * pit bull
 * pit bull mix
 * mixed breed dog
 * pack of dogs

Examples if two dogs of the same or mixed breed are involved in an attack, as recommended:
 * German Shepherds (2)
 * Doberman Pinschers (2)
 * mixed breed dogs (2)
 * pack of dogs (2)
 * pit bulls (2)
 * Doberman Pinschers (2), German Shepherds (2), pit bull mixes (2)


 * Would it be better to put this in some kind of infobox so the autoarchive bot doesnt keep pulling this? PearlSt82 (talk) 22:35, 19 April 2019 (UTC)
 * Yes, I think so since these are guidelines for the main page, but I don't know how to move them into an infobox. Jacobm co (talk) 22:52, 19 April 2019 (UTC)
 * There are several talk namespace template messages - maybe the "notice" one would be appropriate? PearlSt82 (talk) 12:51, 20 April 2019 (UTC)

Standardization of breed names
Please use these guidelines to help keep breed names accurate, consistent, and organized. Currently, the main page contains a variety of different formats (capitalizations, hyphenations, and sometimes spellings) for breed names. For example, the Labrador Retriever is capitalized in a variety of ways including: Labrador Retriever, Labrador retriever, and labrador retriever. While there are a number of different suggested ways to capitalize and hyphenate breed names, the recommendation of this guideline is to follow the format that the American Kennel Club and United Kennel Club use which is to capitalize all names in a breed name, for example: Labrador Retriever. In the case of mixed breeds, the format is also to capitalize all breed names but not the word "mix", for example: Labrador Retriever-Husky mix. In the case of pit bulls, since "pit bull" is not a proper breed but a dog type (a category of dog consisting of at least 4 breeds as listed on the pit bull Wikipedia page), the proper way to capitalize pit bull is with both names lowercase: pit bull (unless it is the first word in a sentence, then it would be: Pit bull and "bull" should not be capitalized). If one of the proper pit bull breeds is identified, then all of the names in the breed name would be capitalized (like any other proper breed), for example: American Staffordshire Terrier, American Pit Bull Terrier, and Staffordshire Bull Terrier. If the number of dogs involved in an attack is more than one, for consistency it should be listed as: Breed (2). For example, if two German Shepherd dogs are involved, then it would be listed as German Shepherds (2).

Examples of breed name format, as recommended:
 * Labrador Retriever
 * Labrador Retriever mix
 * Labrador Retriever-Husky mix
 * German Shepherd
 * German Shepherd mix
 * German Shepherd-Husky mix
 * Doberman Pinscher
 * American Staffordshire Terrier
 * American Pit Bull Terrier
 * pit bull
 * pit bull mix
 * mixed breed dog
 * pack of dogs

Examples if two dogs of the same or mixed breed are involved in an attack, as recommended:
 * German Shepherds (2)
 * Doberman Pinschers (2)
 * mixed breed dogs (2)
 * pack of dogs (2)
 * pit bulls (2)
 * Doberman Pinschers (2), German Shepherds (2), pit bull mixes (2)


 * Would it be better to put this in some kind of infobox so the autoarchive bot doesnt keep pulling this? PearlSt82 (talk) 22:35, 19 April 2019 (UTC)
 * Yes, I think so since these are guidelines for the main page, but I don't know how to move them into an infobox. Jacobm co (talk) 22:52, 19 April 2019 (UTC)
 * There are several talk namespace template messages - maybe the "notice" one would be appropriate? PearlSt82 (talk) 12:51, 20 April 2019 (UTC)