Talk:Gun laws of Australia/Archive 3

Archive 1

Archive 2

Archive 3

This article is or was the subject of a Wiki Education Foundation-supported course assignment. Further details are available on the course page. Student editor(s): Caitlinsmitt (article contribs).

ABS statistics

According to ABS statistics, not only are murders by knife attacks in Australia more common than murders by gun attacks but knife attacks are also more deadly. For the years 2001 to 2015 there was a 1.34 times greater chance of dying if attacked with a knife than of dying if attacked with a gun. The first step of the calculation is made by adding the number of murders and attempted murders, firstly in gun attacks and then in knife attacks to give a total number of attacks for each type of weapon. Then the percentage of attacks in each case resulting in death is calculated and it is on average 1.34 times higher in knife attacks than in gun attacks.

References

The fact that a lengthy explanation of the methodology for this calculation is necessary is evidence that it violates WP:NOR, and is original research. The assertion compiles information from primary sources, which is probably OK on its own, but then draws inferences from that data based on calculations. Why these numbers or these calculations? If a secondary source made the same conclusion then it'd be OK. Felsic2 (talk) 16:29, 30 March 2017 (UTC)

Agreed, we can't go performing anything more than routine calculations, otherwise it'll be OR. Stickee (talk) 02:26, 31 March 2017 (UTC)

The result of the calculation is obvious, correct, and a meaningful reflection of the sources. It is WP:CALC not OR. — Preceding unsigned comment added by Antihypocritic (talk • contribs) 11:27, 2 April 2017 (UTC)

This is reinserted under the heading "Measuring the effects of firearms laws in Australia". It is not NPOV to ignore general rates of homicide or homicide by other weapons when evaluating gun control measures.Antihypocritic (talk) 11:03, 3 April 2017 (UTC)

SYNTH is not numerical summarization, See WP:SYNTHNOT. ABS is a RS.Antihypocritic (talk) 12:01, 3 April 2017 (UTC)

If you're having to perform a numerical analysis, then it's original research. Stickee (talk) 12:03, 3 April 2017 (UTC)

Please see WP:SYNTHNOT and also WP:AVRC Antihypocritic (talk) 12:21, 3 April 2017 (UTC)

One of the reasons that this is original research is that you're making an analysis, or a set of calculations, that no reliable secondary source has made. I could perform some simple calculations comparing gun violence to stock market indices, and it would be OR for the same reason. Wikipedia editors should reflect what the sources say, not create new conclusions on our own. Felsic2 (talk) 23:42, 3 April 2017 (UTC)

If nothing else, the calculations in the paragraph quoted above are self-evidently wrong. It says (with my emphasis here:

... knife attacks are also more deadly. ... there was a 1.34 times greater chance of dying if attacked with a knife than of dying if attacked with a gun. ... adding the number of murders and attempted murders, ... to give a total number of attacks ...

The calculations describe the chance of dying in a murder or attempted murder - but that is not all attacks. An attacker may only seek to injure, not kill; the calculations as described do not include those attacks. Mitch Ames (talk) 13:30, 4 April 2017 (UTC)

Thank you to those who have endeavoured to guide me about this edit. I thought that I was answering the various points as things developed. Obviously I have not done this adequately. CamV8 tried to guide me to this talk page to discuss. Belatedly I am taking that advice. Some things I could have made clearer in response to points and policies that have been referred to by other editors;

CamV8 suggested knife violence was not relevant to an article about gun control. I think I answered that with my summary when I reinserted. I submit that to consider alternative non-firearm homicide methods in a section headed “Measuring the effects of firearms laws in Australia” of an article on gun control must be relevant to the article if it is to be NPOV. Indeed elsewhere in the article there is reference to non-firearm suicide and non-firearm homicide.

CamV8 also cited the WP:RS policy. There is only one source involved and that is the Australian Bureau of Statistics. I submit that it must be taken as a reliable source. The ABS is already the subject of a number of citations in the article.

Other points raised are all to do with whether my edit was Original Research or proper use of a Primary source. I thank Felsic2 for the time taken to give an example of what would be Original Research because I think it reveals what might have been a misunderstanding of my source and what I was doing. My using two citations may have given the false impression that I was taking two sources and making some analysis of the kind Felsic2 mentions in the example.

I reiterate, there is only one source, the ABS. Also I am referring to only one series of ABS statistics, that is; “VICTIMS, Use of weapon in commission of offence.” The fact that I used two citations was because that was needed to see the full time series. One cited report deals with years 2001 to 2009 the other with subsequent years. (There was an error in the citation for later years because it did not cover the 2015 year and that needs to be fixed.) I am not combining material from multiple sources.

I did not draw any interpretation or conclusion from the statistics, the facts in my edit are prima facie. I only performed a calculation to add the numbers, calculate percentages and compare them. I submit that according to Wikipedia policy routine calculations such as these do not count as original research.

Policy is that Primary Sources that have been reputably published may be used in Wikipedia, but only with care, because it is easy to misuse them. I submit that the ABS is a reputable publisher and I have not misused the source.

Finally Mitch Ames made the point that attacks made with the intention only to injure are not included in the statistics I quoted and that therefore the calculations are wrong. Mitch Ames is correct to the extent that the edit should have made it clear that the calculations only concern attacks made with murderous intent.

I look forward to further comments.Antihypocritic (talk) 09:13, 6 April 2017 (UTC)

"There is only one source involved": It doesn't matter if it's one source, if you're making a comparison of two indices/values that isn't performed by the reliable reliable source, it's OR. To rephrase for clarity: we cannot compare, only RSs can. Stickee (talk) 05:27, 7 April 2017 (UTC)

"... the calculations only concern attacks made with murderous intent ..."
That's still not correct. According to Murder (Australian law) a person may commit murder in some cases (eg when intending to inflict grievous bodily harm) even when not intending to kill, ie "murderous intent" is not a prerequisite for murder. In the ABS "murder" statistics we do not know how many of them involved "murderous intent" - some of the murders may have been the result of an intent to cause GBH, not murder. Thus we cannot make any statement about "murderous intent". Mitch Ames (talk) 03:46, 8 April 2017 (UTC)

New edit will simply report ABS statisticsAntihypocritic (talk) 08:39, 18 April 2017 (UTC)

I would advise you to paste your proposal here first, and make the edit later if there is a consensus here. Stickee (talk) 12:01, 18 April 2017 (UTC)

That's not any better. Your still adding across multiple years, then performing comparison of successful to unsuccessful, then creating a percentage, and finally placing the two methods side by side. Every single step is another set of OR that's being conducted. Stickee (talk) 05:43, 19 April 2017 (UTC)

@Antihypocritic: I think that this is a lost cause, but as a good-faith attempt to rescue it... Perhaps if you try to tell us here (not in the article), in a short simple sentence, what point you are trying to make with these edits - then we might be able to find some way of achieving that goal, provided it's not OR or SYN.

For example your first attempt, as quoted at the start of this talk page section, could be summarised as "knife attacks are more deadly than gun attacks; you are more likely to be killed if attacked by a knife than a gun". Now we've already demonstrated that the sources you cite do not support that statement - but if you're actually trying to say something else, and that something else is supported by the references, then if we know what the point was, we'd have a better chance of being able to help you rather than fighting against you.

As I said, it may not be possible, given the data, but it might be worth a try. Mitch Ames (talk) 11:59, 19 April 2017 (UTC)

My most recent edit was; “The Australian Bureau of Statistics publishes data about victims of homicide that includes the type of weapon used. In the years 2001 to 2015 the total number of incidents of murder and attempted murder with a firearm was 1,436 and of those 530 were recorded as murder; that is a percentage of 36.9% where the victim died. For the same 15 year period the total number incidents of murder and attempted murder with a knife was 2399 and of those 1169 were recorded as murder; that is a percentage of 49.73% where the victim died.”

References

Instead of that I could have just mentioned the statistics for the latest 2015 year by stating; “In the year 2015 there were 57 incidents of murder and attempted murder with a firearm and of those incidents 27 were recorded as murder; that is a percentage of 47.37% where the victim died. Also in 2015 there were 109 incidents of murder and attempted murder with a knife and of those incidents 70 were recorded as murder; that is a higher percentage of 64.22% where the victim died.” However to put the information forward in the most accurate non cherry picking form I performed a routine calculation and added across the years. It should be noted that the ABS reports cited each contain data for a number of years.

Felsic2 on 30 March 2017 said; “…….The assertion compiles information from primary sources, which is probably OK on its own, but then draws inferences from that data based on calculations.” My latest edit simply compiled the information by addition as approved by Felsic2 and did not draw inferences. The inference is prima facie there in the ABS (reliable source) statistics.

Felsic2 on 28 March 2017 edited as; “According to ABS statistics, deadly knife attacks in Australia are more common than gun attacks. The percentage of murders by knife for the years 2001 to 2015 is 1.34 times higher than the percentage of gun murders.” Here we have Felsic2 making a comparison of percentages which is what I did in my first edit.

Stickee on 31 March 2017 said; “Agreed, we can't go performing anything more than routine calculations, otherwise it'll be OR.” Surely my latest edit must only be routine calculations. There is only addition and then the calculation of a percentage, I fail to see how this could not be other than routine calculation. It may be helpful if someone could give an explanation of why addition and percentage calculations are not to be considered routine in this particular instance.

Mitch Ames disagrees that ABS sources support his summary of my edit; "knife attacks are more deadly than gun attacks; you are more likely to be killed if attacked by a knife than a gun" and invites me to explain the point I am trying to make. To the contrary, the ABS source does prima facie show that an attack by a person wielding a knife more often results in death than an attack by a person wielding a gun. The points made by Mitch Ames earlier about whether the attacker only intended to injure and thus could not be said to have “murderous intent” and murder including attacks intended to cause grievous bodily harm are common to both knife and gun attacks and common to the charges of murder and attempted murder and so do not alter the import of the statistics. The point is the ABS statistics prima facie show that in Australia an attack by a person wielding a knife more often results in death than an attack by a person wielding a gun. I respectfully suggest that other editors should follow the links given in my last edit and see this for themselves.

So, in the light of the previous comments and edit of Felsic2 and Stickee, I have two questions; why are my additions and percentage calculations not to be considered routine and why is it not permissible to state the result of those calculations when they do nothing other than show what the ABS statistics prima facie are. Antihypocritic (talk) 12:06, 23 April 2017 (UTC)

the ABS source does prima facie show that an attack by a person wielding a knife more often results in death than an attack by a person wielding a gun. — Exactly which cells in those spreadsheets show the total number of all attacks with knife, and all attacks with gun? Ie, including all non-sexual assaults in 2001-2009, and those not related to homicide in 2010-2015. Without those figures, any general claim ("attack by a person wielding a knife/gun") has no basis. Mitch Ames (talk) 12:48, 23 April 2017 (UTC)

why are my additions and percentage calculations not to be considered routine — WP:CALC requires "consensus among editors that the result of the calculation is obvious [and] correct". In this case editors other than you do not appear to agree that the calculations are "obvious", and I have pointed out several times that they are not correct (eg they do not include all attacks/assaults). Mitch Ames (talk) 13:09, 23 April 2017 (UTC)

Thank you Mitch Ames for again taking the time to focus my attention on the issues involved. You are correct in highlighting that the ABS spreadsheets I have cited do not show all attacks with knife or gun. The statistics are only of cases that have been adjudged as murder and attempted murder and reveal that in attacks in those circumstances death is more likely from knife wielders. The statement you have taken issue with was followed immediately in the same paragraph by comments I think show that context but perhaps I should have qualified it.

BUT my most recent deleted edit did not make any such statement. My edit merely mentioned the total numbers of incidents of murder and attempted murder and the percentage relationships of murder to the totals. It was for that reason I said that the result was prima facie in the statistics. I have done nothing other than quote those statistics in summation.

Mitch Ames partially quotes the WP:CALC policy. The full quote of policy is; “Routine calculations do not count as original research, provided there is consensus among editors that the result of the calculation is obvious, correct, and a meaningful reflection of the sources. Basic arithmetic, such as adding numbers, converting units, or calculating a person's age are some examples of routine calculations.” I understand that to mean that editors will act in good faith and not capriciously deny consensus to calculations that are obvious, correct, and a meaningful reflection of the sources.

Mitch Ames claims the calculations are not correct (eg they do not include all attacks/assaults). This is confusion between the summation and the inference Mitch Ames believes I may have previously drawn although my last edit did not state any inference. The WP:CALC policy is as to the correctness of the calculation not the inference and I believe the calculations are correct.

By previous comments and an edit both Felsic2 and Stickee have shown a good faith willingness to accept routine calculations. That is why I asked; “why are my additions and percentage calculations not to be considered routine and why is it not permissible to state the result of those calculations when they do nothing other than show what the ABS statistics prima facie are?”. Surely the result of simply adding a column of numbers and showing the percentage relationship between those numbers in the original ABS tables without drawing any inference must just be basic arithmetic? Antihypocritic (talk) 09:24, 26 April 2017 (UTC)

I have done nothing other than quote those statistics in summation. — You haven't just quoted the statistics, or even summarized them; you've calculated and presented a ratio (percentage who died) based on your selection of values, implying that the ratio has some meaning, even when you don't explicitly declare the meaning. Based on your previous edits and the fact that you haven't given us anything else (even when I explicitly asked) I'm assuming that the point that you're trying to make is that, in general, an attack by a knife is more likely to be fatal than an attack by a gun. If you're trying to make some other point, then please tell us what it is.

So we still have the problem that:

• WP:CALC requires "that the result of the calculation is obvious". — I disagree that the calculation is obvious, and I suspect that others do also. To quote Felsic2 right at the start of this discussion: "The fact that a lengthy explanation of the methodology for this calculation is necessary is evidence that it ... is original research."
  (By contrast, if you had simply added up all of the murders, for example, for each year and said "the total number of murders by gun from 2001 to 2015 was nnnn", that would have been OK, to me at least, as a routine and obvious calculation.)
• WP:CALC requires "that the result of the calculation is ... a meaningful reflection of the sources" — I disagree that result is a meaningful reflection of the sources. You appear to have chosen and calculated with some of the numbers to present a result (the percentages) as a meaningful indicator of something that is not directly supported by the sources. This is contrary to WP:SYN, which says "do not combine different parts of one source to reach or imply a conclusion not explicitly stated by the source".

Mitch Ames (talk) 12:59, 27 April 2017 (UTC)

Mitch Ames still assumes that the point I am trying to make is; “in general, an attack by a knife is more likely to be fatal than an attack by a gun”. But as I most recently stated on 26 April 2017; “The statistics are only of cases that have been adjudged as murder and attempted murder and reveal that in attacks in those circumstances death is more likely from knife wielders.” Not attacks in general. On 23 April I was clearly talking in the context of murder and attempted murder. On 6 April I said; “the edit should have made it clear that the calculations only concern attacks made with murderous intent.” On 4 April Mitch Ames recognised; “The calculations describe the chance of dying in a murder or attempted murder - but that is not all attacks.” So to make it clear, I am not making any point about gun and knife attacks in general but only about those in the context of attacks in the circumstances of murder or attempted murder.

Thanks to Mitch Ames for agreeing that the addition of the numbers is a routine and obvious calculation. But Mitch Ames continues and states; “You appear to have chosen and calculated with some of the numbers to present a result (the percentages) as a meaningful indicator of something that is not directly supported by the sources.” But I have not just used “some” of the numbers, I have added all of the numbers of each of gun and knife attacks before calculating the percentages. Nor have I just made a “selection of values” as Mitch Ames mistakenly claims, I have used all of the values. Look at the numbers for the year 2015 (or of the totals for the years 2001 to 2015). Even without calculating the percentages it is quite clear and obvious (explicit with no room for doubt) that more of the attacks with knives result in murder and fewer of the attacks with guns. Calculating the percentages merely quantifies what is explicit from the source.

WP:SYNNOT says “SYNTH is not explanation.” “SYNTH is when two or more reliably-sourced statements are combined to produce a new thesis that isn't verifiable from the sources. If you're just explaining the same material in a different way, there's no new thesis.” As I said above I am merely quantifying (explaining) what the material is and it is “verifiable from the sources”. Antihypocritic (talk) 09:12, 2 May 2017 (UTC)

Mitch Ames still assumes that the point I am trying to make is; “in general, an attack by a knife is more likely to be fatal than an attack by a gun”. — So what is the point that you are trying to make? I repeat my statement that it would be easier to help you if we knew what point you were trying to make with the inclusion of the percentages. Simply including some numbers in the article without explanation or reason does not help the reader - especially the values for knife deaths (in an article about gun laws). Quoting WP:NOTSTATS, item 3: "Statistics that lack context or explanation can reduce readability and may be confusing; ... articles with statistics should include explanatory text providing context".

I have not just used “some” of the numbers, I have added all of the numbers of each of gun and knife attacks" — You have not added all of the number of gun and knife attacks - you have only included murder and attempted murder, excluding assault etc. Ie, you have selected a specific subset of all attacks with gun/knife.

WP:SYNNOT says “SYNTH is not explanation.” ... I am merely quantifying (explaining) what the material is — WP:SYNNOT#SYNTH is not numerical summarization also says "Summarizations based on statistical methods, however, are original research by synthesis, as they involve the reinterpretation of data, and decisions about which statistical methods ... are appropriate." I maintain that selecting only the values for murder and attempted murder (and excluding assault etc) and (implicitly) comparing percentages computed from them is a decision about which statistical method is appropriate, and thus OR.

We don't seem to be getting any closer to an agreement here. Perhaps you should consider raising an WP:RFC to solicit input from some other editors. Mitch Ames (talk) 12:56, 2 May 2017 (UTC)

The article is “Gun laws in Australia”. My edit was entered under the heading “Measuring the effects of firearms laws in Australia”. It is axiomatic that in assessing the effects of firearm laws on homicide and suicide the deadliness of alternative methods must be relevant. The article already contains references to alternative methods of suicide.

The Australian Bureau of Statistics has published data for the years 2001 to 2015 about victims of “murder” and “attempted murder”. The data includes the type of weapon used. With respect to 2015 the data for firearm and knife is;

2015	Murder	Attempted Murder
Firearm	27	30
Knife	70	39

For the years 2001 to 2015, the data for each year can be added to give totals as;

2001-15	Murder	Attempted Murder
Firearm	530	906
Knife	1169	1230

Of these attacks; that is, those adjudged to be murder or attempted murder, a greater proportion of the attacks with knives result in death and a lesser proportion of the attacks with firearms. These attacks are not all the attacks with knives or firearms but are a subset of the most serious attacks. My point is that the statistics self evidently show that an attacker who uses a knife and intends to murder or to inflict grievous bodily harm will be more likely to kill than an attacker who uses a firearm with the same intent. My edit did no more than add the data for each year and calculate the percentages to show the proportions of attacks resulting in death. I believe that is CALC not OR.

My edit was; “The Australian Bureau of Statistics publishes data about victims of homicide that includes the type of weapon used. In the years 2001 to 2015 the total number of incidents of murder and attempted murder with a firearm was 1,436 and of those 530 were recorded as murder; that is a percentage of 36.9% where the victim died. For the same 15 year period the total number incidents of murder and attempted murder with a knife was 2399 and of those 1169 were recorded as murder; that is a percentage of 49.73% where the victim died.”

I note that other edits under the heading “Measuring the effects of firearms laws in Australia” also quote ABS statistics without any explanation and I am seeking approval for a similar edit.Antihypocritic (talk) 07:42, 13 May 2017 (UTC)

References

"It is axiomatic that in assessing the effects of firearm laws on homicide and suicide the deadliness of alternative methods must be relevant." — Even if it is axiomatic, it does not change the fact that your calculations are OR by SYNTH. If it is axiomatic, then find a reliable source that has considered the deadliness of alternative methods when it assessed the effects of the firearm laws, and report what that source says. Remember that you can't assess the effects of gun laws considering the deadliness of alternative methods, because such assessment by you would clearly be OR.

"The article already contains references to alternative methods of suicide." — The article mentions suicide by alternative methods, but those mentions do not include the level of analysis that you are trying to apply. (They may cite sources that do the analysis, which is OK.)

"... the statistics self evidently show that an attacker who uses a knife and intends to murder or to inflict grievous bodily harm will be more likely to kill than an attacker who uses a firearm with the same intent." — As I've already pointed out, the statistics do not include (for example) attacks that intended harm (including GBH) but did not intend to kill, and that were not fatal. (If you think they do, tell me exactly which cells have the numbers for such assaults.) Without that information you cannot possibly calculate the likelihood of an attack being fatal because you don't know how many such attacks were not fatal.

"... calculate the percentages to show the proportions of attacks resulting in death. I believe that is CALC not OR" — The fact that several editors disagree with you automatically means that it fails to meet the criteria for WP:CALC, which include "consensus among editors that the result of the calculation is obvious, correct, and a meaningful reflection of the sources"; clearly you do not have that consensus.

"... other edits under the heading “Measuring the effects of firearms laws in Australia” also quote ABS statistics without any explanation" — They "quote ABS statistics", they do not apply their own calculations to those statistics.

We are going round in circles here, and nobody else appears to be agreeing with you. If you still think your case has merit, please raise an WP:RFC. With any luck that will bring in some other opinions, and in any case should give as a resolution to the debate. Mitch Ames (talk) 10:16, 13 May 2017 (UTC)

Lack of citation on sentence regarding attacks by Aboriginal peoples

"From the landing of the First Fleet there was conflict with Aborigines. Firearms were used to protect explorers and settlers from Aboriginal attack."

Given that the article already notes that a citation is needed, I would like to suggest that this line be removed unless a credible citation is sourced. It doesn't add to the article, as the information is essentially covered by the statement: "Firearms were also used for hunting, protection of persons and crops, in crime and fighting crime, and in many military engagements."

It is described with more detail in the European settlement section: "From the landing of the First Fleet there was conflict with Aborigines over game, access to fenced land, and spearing of livestock. Firearms were used to protect explorers and settlers from Aboriginal attack. A number of punitive raids and massacres of Aboriginals were carried out in a series of local conflicts. The history of these conflicts is contentious (see History wars)."

This is somewhat more nuanced but without citation it's unclear whether the inclusion of 'Aboriginal attack' as a rationale for guns was of such significance that it warrants particular mention. For example, as a penal colony, was there a greater (perceived or actual) threat from convicts? Also, Australia's first peoples are not referred to as 'Aboriginals'.

These statements are a superficial representation of an incredibly complex aspect of Australian history and introduces a bias favouring British settlers. The page also links to https://en.wikipedia.org/wiki/History_wars which I would also suggest is not an unbiased representation of that debate, with much greater emphasis given to what could be termed the conservative opinion, with the counter opinion included only for the appearance of balance or to be repudiated.

KPKnox (talk) 03:07, 3 October 2017 (UTC)