Talk:List of countries by traffic-related death rate/Archive 1

scope

The article was misleading previously because it only referred to mortality rates in developed countries, not developing countries. Developing countries actually have much higher mortality rates for road traffic injuries (RTI) than developed ones.

I tried to clarify this, but perhaps the article should be renamed. Additionally creating a map showing RTI mortality rates by region based on WHO data that can be found here:

http://www.who.int/violence_injury_prevention/publications/road_traffic/world_report/en/

Would be great, but I don't know how to make such a map in Wikipedia.

Uptownnow (talk) 18:21, 20 August 2008 (UTC)

Russia

Russia is not in the list. It would probably be first around 25 per 100000... and China is missing too. And India. Those 3 countries together must have well over 100000 road casualties per year--213.159.118.174 (talk) 10:16, 15 October 2008 (UTC)

Total death toll

It would be interesting to have figures on worldwide death tolls. The rate is probably around half a million deaths per year or a little less, and the cumulated total around the same as WWI casualties, a glorious mileston for the automotive industry. Would be interesting to know.--213.159.118.174 (talk) 10:21, 15 October 2008 (UTC)

New Discussion

A discussion has been started at Wikipedia talk:WikiProject Countries/Lists of countries which could affect the inclusion criteria and title of this and other lists of countries. Editors are invited to participate. Pfainuk talk 12:48, 17 December 2008 (UTC)

title

since we only have OECD countries here, I changed the page title. this addreses some of the issues raised above. Heroeswithmetaphors (talk) 08:17, 5 May 2009 (UTC)

More countries added, please move back. --78.108.106.253 (talk) 02:53, 27 September 2009 (UTC)

Agree we should move back.

More south American data

Bolivia is in here: http://www.paho.org/English/AD/DPC/NC/violence-graphs.htm#levels Doc James (talk · contribs · email) 22:33, 25 October 2009 (UTC)

Saudi Arabia has one of the highest rates in the world

Ansari S, Akhdar F, Mandoorah M, Moutaery K (2000). "Causes and effects of road traffic accidents in Saudi Arabia". Public Health. 114 (1): 37–9. doi:10.1038/sj.ph.1900610. PMID 10787024. {{cite journal}}: Unknown parameter |month= ignored (help)CS1 maint: multiple names: authors list (link) Doc James (talk · contribs · email) 23:09, 25 October 2009 (UTC)

Data

I have found the mother load of data. This document by the WHO in 2009 contains all information on page 251-257. Doc James (talk · contribs · email) 23:36, 25 October 2009 (UTC)

Found an error in the data. There is currently two South Korea entries but no North Korea entry. /94.193.242.248 (talk) 21:47, 29 October 2009 (UTC)

Fixed. /Lokal_Profil 03:11, 31 October 2009 (UTC)

Discrepancy: list vs the color-coded map

Uruguay is listed as 12 deaths per 100,000 population in the list, an orange which is far from the bright yellow "less than 5" color it's been given on the map. I'm not sure which was updated more recently (the map or the list), or even if out-of-date data is the cause of this discrepancy. —Preceding unsigned comment added by 216.188.254.2 (talk) 07:48, 20 November 2009 (UTC)

Ireland's colour on the map is totally out of line with the statistics too. The map needs to be updated 14:23, 19 August 2012 (UTC)

World logo

What value does this add to the article? --John (talk) 02:35, 1 December 2009 (UTC)

It keeps all the lines balanced and thus makes the list look good. There is no requirement for them all to be flags btw.Doc James (talk · contribs · email) 02:38, 1 December 2009 (UTC)

Seriously? Well, at least you are being honest. I suggest a read of WP:MOSICON, and consideration of recoding the table, perhaps with a blank image, so we avoid making it look like we have invented a flag for our planet. All the other entries are flags, so this would be the natural assumption. It is one to be avoided. --John (talk) 04:30, 1 December 2009 (UTC)

In many cases it serves as a good placeholder, as James says above. ←Baseball Bugs ^{What's up, Doc?} carrots→ 05:09, 1 December 2009 (UTC)

The preferred placeholder for situations like this is {{noflag}}. — Andrwsc (talk · contribs) 07:46, 1 December 2009 (UTC)

Icon is no good.

The icon at the top of the data columns is unexplained. What is SUPPOSED to do is rank the countries on the basis of the data in that column. Since it seems to ignore decimal points, it gives a totally wrong listing.Kdammers (talk) 05:39, 10 August 2010 (UTC)

Re-order

The article content does not match its title. The main content, the table of death rates for various countries, is not a list of countries by death rate. It is a list of countries by alphabetical order of country name, which includes information on each country's death rate. It will be much more useful to re-order this table so that the countries are presented in descending or, preferably, ascending order of death rate. This will give the reader all the information currently presented plus the worldwide contextual rank of each country by traffic-related death rate. Right now, the reader must scrutinise each and every entry and hunt down the lowest or highest or third-lowest or whatever other number is of interest, because there's no direct way to search for this information. While it is true that re-ordering the list will take away the alphabetical order, there is a direct way of searching for any particular country: the reader simply types the country's name into the browser search box and is taken directly to it within the list. Meanwhile, the meat-and-potatoes informational content is presented usefully in a ranked list rather than randomly in alphabetical order. —Scheinwerfermann ^T·_C18:46, 19 August 2011 (UTC)

The table is already sortable by clicking on the arrows at the top of each column. Unfortunately, according to Help:Sorting#Sorting tables by default, there is no way to make it sort a particular column by default (ie without the user clicking something first).  Stepho  talk  21:42, 19 August 2011 (UTC)

Oh, lookit there! Thanks much. More useful than any particular default sort would be an instruction telling the reader to click the non-intuitive little double-arrow-box dingus to sort by that column's heading. —Scheinwerfermann ^T·_C22:18, 19 August 2011 (UTC)

Yeah, kind of sad that it can't pre-sort them. I guess users just have to learn how to sort for themselves.  Stepho  talk  09:43, 20 August 2011 (UTC)

Denominator

Why per 100,000 capita? I always want to see per 10^(3*n) probably due to having been brought up with SI units. After all, would anyone suggest that the column by billion travel km should be per 100 million travel km? I guess the original reason is that when one organisation is drawing up statistics they think that E.g. 5 per hundred thousand is more understandable than 0.5 per million, but that's a problem of education, and becomes completely irrelevant when one is looking at a wide range of values. — Preceding unsigned comment added by Zsalya (talk • contribs) 17:38, 16 October 2011 (UTC)

Character Sort

The rate per 1 billion miles sorts as character, not numeric. So, '13' comes before '5'. Is there a way to fix this? Might be a major problem across the famous Wikipedia.--161.7.117.246 (talk) 21:56, 30 November 2011 (UTC)

All the 'N/A's were screwing up the sorting. I removed them and it performs much better now.  Stepho  talk  05:29, 1 December 2011 (UTC)

Oops, I mistakenly deleted the references and categories at the bottom. Chrome does weird things when I'm not looking :( Luckily an anonymous editor restored them.  Stepho  talk  20:47, 1 December 2011 (UTC)

Fatalities per vehicles

I have calculated and added fatalities per 100,000 vehicles, for countries that were missing it, using data from wikipedia's vehicles per capita page ([[1]]). I divided the "fatalities per 100,000" by "vehicles per 1000" and multiplied the result by 1000. For countries that already had this data I did not change anything. But most of the time the existing values correlated well with the calculated ones even though they were probably based on different years. The spreadsheet I used for calculation can be found at Traffic related deaths per 100,000 vehicles I have not cited or referenced it properly. Yet I think this is very important and revealing data. The fatalities per capita does not show enough information. In a country that has fewer cars the chance of having an accident is obviously smaller. What is usually more interesting is the risk of driving a vehicle. The fatalities could be vehicle passengers or pedestrians. The statistic gives an approximation of how dangerous each vehicle is. Or assuming an average of one driver per vehicle how dangerous each driver is. So I think it is a good estimate of how safe it is to drive in that country. Road fatalities per 1bln km is also a very useful statistic that shows how safe are the roads in the country. But is much more difficult to get. Also some different countries have different average travelling distances. So it might not necessarily show the general risk of driving in that country. To be honest, the fatality rate seems to be too high, almost unrealistic in some African countries in the bottom of the list. One explanation could be that the number of vehicles is significantly under-reported, since all of those countries have very low number of vehicles, Togo has 2 per 100,000 people. Another explanation could be that most of the vehicles in those countries are buses that are driving through very dense population, so any single crash could involve many passenger and pedestrian fatalities. — Preceding unsigned comment added by Tiraniyaya son (talk • contribs) 09:51, 16 December 2011 (UTC)

One possible explaination is that these countries tend to use utes (pickup trucks) and then overload them with 10-20 people - top heavy and no seat belts. Similar to your scenario for buses, one accident can claim many lives - even worse between two vehicles. I have no reference for this but I'd believe the official data.  Stepho  talk  10:02, 16 December 2011 (UTC)

That makes sense. Just found this article from less than a month ago Football team in Togo involved in fatal bus crash. The Togo is the country with the worst statistics. The bust was carrying the national football team. Six people died.--Tiraniyaya son (talk) 10:55, 16 December 2011 (UTC)

According to those calculations Togo has 14,050 and Ehtiopia has 11,666 fatalities/100,000 vehicles. Which seems to be unrealistically high. That would make using a vehicle on a regular basis similar to playing Russian roulette. On the other hand, it would easily explain what happened to the Togo national team. If statistics are true what happened to them wasn't an accident but something expected. --Tiraniyaya son (talk) 16:43, 16 December 2011 (UTC)

So the best way to proceed would be to try to find the real stats on fatality rate. If none are available then we could use the best available estimates of numbers of vehicles and fatalities and calculate the fatality rate from there.--Tiraniyaya son (talk) 16:43, 16 December 2011 (UTC)

I also just noticed that WHO provides the latest data not only on the fatalities but also on the number of registered vehicles and even the distribution of fatalities, i.e. whether they were passengers or pedestrians. The data set can be found here Mortality, Distribution of road traffic deaths by road user. It is worth noting that the fatality numbers are "Number of road traffic deaths predicted using negative binomial regression". So they could potentially be overestimated. The number of vehicles on the other hand is the number of registered vehicles, so it could be underestimated. This could explain unrealistically high fatalitys rate per vehicle.--Tiraniyaya son (talk) 16:43, 16 December 2011 (UTC)

I thought about adding the distribution of the fatalities by the type of the road user, i.e. passenger or pedestrian or a cyclist, but then I thought that they do not make sense unless you also have data of how many people actually exist under that type. Knowing how many cyclists die only makes sense if you know how many cyclists were there.--Tiraniyaya son (talk) 00:38, 19 December 2011 (UTC)

Same for male female fatality ratios.

I just noticed that while the number for Ethiopisa more or less matches, the number for Togo doesn't. The number of registered vehicles is 48234, which divided by the population of Togo and multiplied by 1000 gives around 7, not 2 like it says in the World Bank data. So there is some inconsistency already. It might be worth to go through all the data one by one. And I was wrong, WHO does not get the vehicles numbers from IRF, it gets directly from countries. It uses IFR for road density data, and that is why I got confused. So I guess we should use the larger number and maybe add it to the vehicles per capita article first and then use it from there.--Tiraniyaya son (talk) 18:20, 16 December 2011 (UTC)

The discrepancy between the WHO data and the IFR data could largely be explained by the fact that IFR does not count 2-wheel vehicles while WHO does. I am going to update the table with the fatality rates based on the number of vehicles from the WHO database. It also makes sense since it is usually the main source of fatality numbers. In the cases where there is already existing data from some other source I will leave the existing data.--Tiraniyaya son (talk) 00:38, 19 December 2011 (UTC)

I recalculated fatalities per vehicle with the data from WHO. In many countries it gave big improvements because WHO data includes motorcycles, which in many african countries can be more than 70% of vehicles. But in one case it made even worse, particularly in the case of Central African Republic which appeared to have 23,980.1 fatalities per 100,000 vehicles, which clearly does not make sense. The obvious explanation is that the number of vehicles is underreported. WHO report mentions 5834 vehicles with 77% of them motorcycles. IFR says that that they have 0 vehicles per 1000 people, which I guess means that it is less than 1 vehicle per 1000 people, or if they round to the closes digit, less than 0.5 vehicles per 1000 people. The population of CAR was 4342735, which would make IFR number even much lower than the WHO number. By the way, I have not noticed this before because initially I got the data directly from Wikipedia's page for vehicle numbers, which gives 4 vehicles/1000 people for CAR, but it is not clear where this data is coming from since World Bank lists the IFR data.--Tiraniyaya son (talk) 16:00, 19 December 2011 (UTC)

By the way, report has two numbers for the number of road deaths, one is the reported number and the the other is an estimate. The fatalities per capita uses the estimate and it is right in doing so. Because most of the time the many fatalities can go unreported, particularly in the African countries. In fact, report says that only 660,000 deaths are reported while the total estimate for the report is 1,230,000. But for the fatalities per vehicle it is probably better to use the reported deaths number, because the number of registered vehicles could be underreported too. So possibly the errors would cancel each other and would be closer to the real value of fatalities per vehicle. I think a bad estimate is better than no estimate in this case. So I am going to recalculate the rates using this method.--Tiraniyaya son (talk) 16:00, 19 December 2011 (UTC)

A very good review of the fatality rates is: Estimating global road fatalities. Unfortunately it is too old, published in 2000 and most data in it is even older. However it is still has a lot of points worth mentioning. The fatality rate for Ethiopia is based on numbers from 1996 and is 1951/100,000vehicles, which is much more realistic. It also notes that the fatality rates for Africa and Latin America were growing and were expected to grow in future, even if slower, while the fatality rates for Asia and Europe were expected to fall. The authors note that both the casuality data an the number of vehicles data can be wrong: Some countries impose a de-registration fee so few motor vehicles are removed from the official registers while in many countries, owners try to avoid registering vehicles because of the associated fees. It also cites another study, The PHARE multi-country road safety study, to note that "The recent PHARE report highlighted the difference between the number of registered vehicles reported by national experts and the IRF, with most countries having more vehicles registered than reported by the IRF". Since IRF data was an underestimate 11 years ago, it might still be so. Unfortunately they did not repeat this study with new data.--Tiraniyaya son (talk) 18:20, 16 December 2011 (UTC)

In that report Central African Republic also has the highest number of fatalities per vehicle at 3,390/100,000. But it could be an overestimate too.

I've added an asterisk to the numbers that were calculated, to indicate that they may not be as reliable. ToastyKen (talk) 04:35, 26 April 2014 (UTC)

Source year

Some data is based on two different years. I'm not sure whether to specify it as a range or just note the latest year. For now I'm going to put in the range.

Venezuela

Nothing on Venezuela? The map says the death toll there is high. — Preceding unsigned comment added by 167.206.244.10 (talk) 01:39, 26 February 2018 (UTC)

Death toll in egypt

In the 'Road fatalities per capita' image, Egypt is shown in dark red. However, in the table, the 'Road fatalities per 100,000 inhabitants per year' for Egypt is listed as only '8.1'. Which is correct? 130.89.101.28 (talk) 07:09, 20 January 2011 (UTC)

Furthermore, the source cited says "Reported road traffic fatalities (2007): 12 295d (70% males, 30% females)"

Page 106 of http://whqlibdoc.who.int/publications/2009/9789241563840_eng.pdf - that is the second reference in the article.

By my math, 12295/75,500,000 = x/100,000 seems to indicate that x = 16.2. Which is not 8.1 or >40. 75.61.128.56 (talk) 00:46, 13 December 2011 (UTC)

Death toll in Estonia

I'm sorry but the whole article has very old information. In 2010 death toll in traffic accidents in Estonia was 78 thus making death rate by 100 000 inhabitants 5,8. source is here. I am telling this because the number in the article is 15,3 or something like that, which is almost 3 times higher than it is now and it is also equivalent to some banana republic in Africa. You might as well rename the article "List of countries by traffic-related death rate long time ago, so information is a bit off for some countries". We need to find new information for ALL the countries. --Lollerman1 2 (talk) 12:00, 14 February 2011 (UTC)

Ethiopia

The highest estimate for Ethiopia's road fatality rate I've found so far is 1900/100,000vehicles from here: Ethiopia has the highest per capita rate of car fatalities in the world. The source is not clear as with most other estimates. There is also this article Khat, driver impairment and road traffic injuries: a view from Ethiopia from WHO. It's from 2010 and it cites Ethiopian National Road Safety Coordination Office (which apparently does not have a website) with an estimate of 1140/100,000 vehicles. But the article also mentions that the real figure may be higher due to underreporting. It also mentions the legal status of drug Khat as a possible contributor to the fatality rate.--Tiraniyaya son (talk) 16:43, 16 December 2011 (UTC)

Death toll in Spain

Data is wrong, here official sources from the spanish authorities. Mortality 2010, 1730 people. Divided over 47 Million inhabitants we obtain 3,669 deaths per 100.000 thousand inhabitants. Could someone please correct? I am unfortunately a beginner. http://www.dgt.es/was6/portal/contenidos/documentos/seguridad_vial/estadistica/accidentes_24horas/resumen_anual_siniestralidad/resumen_siniestralidad031.pdf — Preceding unsigned comment added by Molinasang (talk • contribs) 19:01, 21 April 2011 (UTC)

Number of cars per capita

Number of cars per capita should appear in this table if deaths per capita is included. this information is vital for interpreting per capita deaths.

Also, do the deaths include pedestrians that were hit, or just people inside a car? This gives insight into how safe driving is (and how careful pedestrians are). could be expressed as percentage of collisions where pedestrian dies/is injured over total collisions, or percentage of pedestrian deaths over total. Dazalc (talk) 04:19, 15 August 2012 (UTC)

Number of cars per capita

Number of cars per capita should appear in this table if deaths per capita is included. this information is vital for interpreting per capita deaths.

Also, do the deaths include pedestrians that were hit, or just people inside a car? This gives insight into how safe driving is (and how careful pedestrians are). could be expressed as percentage of collisions where pedestrian dies/is injured over total collisions, or percentage of pedestrian deaths over total. Dazalc (talk) 04:19, 15 August 2012 (UTC)

Updated Data

The WHO has put out the latest update http://www.who.int/violence_injury_prevention/road_safety_status/2013/en/index.html 131.217.33.146 (talk) 01:46, 18 March 2013 (UTC)

Feel free to start editing the article. Wikipedia allows anyone to this.  Stepho  talk  05:49, 18 March 2013 (UTC)

Poland

Data for Poland is certainly wrong and the factor and deathtoll don't match comparing to population. According to Polish wikipedia there were 3582 deaths in 2012.

http://pl.wikipedia.org/wiki/Wypadek_drogowy#W_Polsce — Preceding unsigned comment added by 83.18.153.146 (talk) 09:58, 24 September 2013 (UTC)

Moreover, http://www.economist.com/blogs/easternapproaches/2013/06/polish-driving which is given as a source, mentions not 1400 in 2012, but 4200 in 2011. Seems that data for Poland are completely wrong. — Preceding unsigned comment added by 85.221.157.117 (talk) 17:36, 9 January 2014 (UTC)

Calculations

A lot of these numbers are just far off! I checked 4 random nations, and 3 of them were not correct according to the sources linked. Could this be fixed as soon as possible, people are using these numbers, and believe in them. — Preceding unsigned comment added by 193.91.226.152 (talk) 11:49, 1 November 2013 (UTC)

Please feel free to start editing. Don't worry too much about formatting, syntax, etc - one of us (probably me) can tidy it up afterwards.  Stepho  talk  21:43, 1 November 2013 (UTC)

Give Rankings?

Is there a way to add a column to the matrix that gives the ranking (i.e. 1,2,3...) of a country, regardless of which column is used to sort? Don't know enough about wiki formatting to do it myself, but when you are 150 spots down it can be hard to tell... — Preceding unsigned comment added by 129.170.26.127 (talk) 16:52, 26 November 2013 (UTC)

The is no automated way to do that. We could add a column by hand but it would then need to be carefully maintained for the rest of eternity and sooner or later the order will change and nobody will type in the new numbers. Much better to just let the reader click on the different sort columns (including ranking) as they wish.  Stepho  talk  15:00, 27 November 2013 (UTC)

Data is off

I found following data, and significant differences in the data presented here (appearantly, china has more deaths due to traffic which seems to make sense since its a lot larger):

• china 220 000
• india 196 000
• russia 35 000
• brazil 35000
• egypt 31000

see http://www.washingtonpost.com/blogs/worldviews/wp/2013/01/18/a-surprising-map-of-countries-that-have-the-most-traffic-deaths/

Please update the figures 109.133.103.157 (talk) 14:52, 3 January 2014 (UTC)

The numbers you just mentioned are not from reliable sources. The Chinese and Indian numbers are from a reader comment, not from the Washington Post article itself and are therefore unverified and untrustworthy (and another reader pointed to them as wrong). The numbers for Russia, Brazil and Egypt don't seem to be anywhere on that page at all. We need solid, reliable and verifiable facts before we can change anything.  Stepho  talk  23:39, 3 January 2014 (UTC)

Well, the numbers regarding China in the table don't add up, anyway. If it is 20.5 per 100,000 inhabitants (a number I have seen elsewhere as well) & there is a population of 1.3 billion we have 266,500 fatalities, not the 68,000 listed.

Here: http://apps.who.int/gho/data/node.main.A997?lang=en you can find a number of 275,983 for China in 2010. (with the same 20.5 per 100000, BTW) bossel (talk) 17:59, 1 April 2014 (UTC)

Justifiability of using private vehicles

Shouldn't we include a section on whether it is hard to justify of private people using their own (private) vehicles for people transport, if the death rate is above say 15 deaths per 100 000 inhabitants per year, and when traffic rules are disregarded, ... by them (ie India)? After all, these vehicles could be easily denied access from city centers, improving transport efficiency and not hindering the transportation of cargo (see Car free movement, List_of_countries_by_traffic-related_death_rate ). KVDP (talk) 15:25, 3 January 2014 (UTC)

That would be WP:ORIGINALRESEARCH. We need solid, reliable sources that support your claim. Note: for a claim of this magnitude, the opinion piece from the Washington Post mentioned above is not considered reliable enough. A reference with evidence of its claims is required.  Stepho  talk  00:00, 4 January 2014 (UTC)

Why a table is not WP:NOR

Why a table is not WP:NOR and why the reverts 02:50, 30 June 2014 and 18:05, 19 July 2014 by user User:Massyparcer were wrong (and especially his repeatedly used "explanation" in the change summary: "inventing your own criteria by mixing WP:WHO and IRTAD is a violation of WP:NOR - It clearly says that you cannot mix sources A and B to derive conclusion C. Neither can you invent your own criteria out of the air."):

See (read!) in general:

• Wikipedia:OR#Synthesis_of_published_material
• Wikipedia:These are not original research
• Wikipedia:What SYNTH is not

Especially:

• Wikipedia:OR#Routine_calculations
• Wikipedia:What_SYNTH_is_not#SYNTH_is_not_an_advocacy_tool citation: "If someone doesn't like what was said, and they therefore cry SYNTH, others almost certainly will be right to cry foul. Virtually anything can be shoehorned into a broad reading of SYNTH, but the vast majority of it shouldn't be."
• Wikipedia:What_SYNTH_is_not#SYNTH_is_not_mere_juxtaposition
• Wikipedia:What_SYNTH_is_not#SYNTH_is_not_ubiquitous citation: "If your understanding of SYNTH includes all instances of reading a table, because reading a table requires "synthesizing" the entry in the table with the label of what the table is, your understanding of SYNTH is wrong. "
• Wikipedia:What_SYNTH_is_not#SYNTH_is_not_presumed citation: "If you want to revert something on the grounds that it's SYNTH, you should be able to explain what new thesis is being introduced and why it's not verified by the sources. You don't have to put the whole explanation in the edit summary, but if someone asks on the talk page, you should have something better ready than "Of course it's SYNTH. You prove it isn't." The burden of proof is light: just explaining what new assertion is made will do, and then it's up to the other editor to show that your reading is unreasonable. But in any disagreement, the initial burden of proof is on the person making the claim, and the claim that something is SYNTH is no exception."
• Wikipedia:What_SYNTH_is_not#SYNTH_is_not_a_catch-all "If there's something bugging you about an edit, but you're not sure what, why not use SYNTH? After all, everything under the sun can be shoehorned into a broad-enough reading of SYNTH. Well, because it isn't SYNTH. It's shoehorning. To claim SYNTH, you should be able to explain what new claim was made, and what sort of additional research a source would have to do in order to support the claim."
• Wikipedia:What_SYNTH_is_not#SYNTH_is_not_a_policy "It's part of a policy: no original research. If a putative SYNTH doesn't constitute original research, then it doesn't constitute SYNTH. The section points out that synthesis can and often does constitute original research. It does not follow that all synthesis constitutes original research."
• Wikipedia:What_SYNTH_is_not#SYNTH_is_not_just_any_synthesis citation: "SYNTH is original research by synthesis, not synthesis per se. ... It seems clear to me that "synthesis of published work" was assumed to be part of the legitimate role of Wikipedia."
• Wikipedia:What_SYNTH_is_not#SYNTH_is_not_primarily_point-by-point
• Wikipedia:What_SYNTH_is_not#SYNTH_is_not_numerical_summarization citation: "Treatment of numeric data is an encyclopedic issue: summarization by sum, average, etc. are necessary expedients, and should not be confused with original research."

-- ZH8000 (talk) 13:20, 4 July 2014 (UTC)

To User:Massyparcer: About the most primitive aspects of percentage calculation

The WHO report in table A2 on page 248 states that the reported number of road traffic deaths of Republic of Korea to be 5505 in 2010. They adjust this number due to different reporting and counting methods among the many countries to an estimated number of road traffic deaths of 6784 (point estimate).

So we can immodestly claim that 5505+1279 equalises with 6784. So far so good. And we also can solidly claim that this is an increase of exactly 23.2334242 per cent. In other words, the estimated number is exactly 123.2334242 percentages as large as the base figure (5505).

Now, if you take another base, a more up-to-date number, such as 4762, then you can either simply add the previous adjustment of 1279 (= 6784-5505 !) as a rough, but only linear adjustment. Or, if you like so, you can alternatively, as a slight better approach, add the same percentage as before, namely 23.24%, so that then you get more recent base + previous percentual adjustment = more recent point estimate, or in numbers: 4762 + (0.2324 * 4762) = 5868.6888 = 5869.

So the difference between the linear and the percentual approach is just 172, or in other words, just 13.5% less than the linear adjustment. ;-))

So much about the 101 of the 101 about percentage calculations.

And now, my dear User:Massyparcer, please go and enter the correct numbers. – Thank you so much!
-- ZH8000 (talk) 16:33, 20 February 2015 (UTC)

Adjusted figures – why and how!

• This list of countries by traffic-related death rate is solely and reasonably based on the WHO report. The most recent report is currently available from 2013.

Actually, it only shows the numbers from the year 2010 by all the countries – and nothing else.
As long as there is no new WHO report available, and in order to stay compliant with the most recent WHO report, more recent figures shall only be added/changed, if the more recent figures follow the same eventual adjustment patterns which have been applied by the WHO report itself!

Otherwise you fail the fundamental wikipedia policy about WP:OR. In other words: Either you do not change the 2010 figures at all, or then you precisely follow the same pattern applied by WHO.

The pages 48-51 of the WHO report clearly describes why there are ajdustments necessary to correct the plain raw numbers issued by the respective countries: in order to reflect the different reporting and counting methods among the many countries.

Just two examples about different reporting and counting methods are:

• e.g. "a death after how many days since accident event is still counted as a road fatality?"
• or e.g. "to compensate for underreporting in some countries"

How do I find out which countries need an adjustment? Go to table A2 on pp. 242-255 and check the difference between reported numbers of road traffic deaths and point estimate.

And therefore, you shall, either, just simpy apply the same absolute difference to the more recent numbers you have available (linear approach), or then add the same percentual adjustment (percentual approach, see previous post) in order to stay compliant with the WHO report.

You, as an author of this arcticle, are not entitled to assume different counting methods than the ones WHO is applying.

Therefore, any change not complaint with the WHO report pattern will be considered WP:OR and shall consequently be undone.

-- ZH8000 (talk) 21:50, 6 March 2015 (UTC)

I'm not sure you read methodology carefully, sir. Only a part of the data are actual estimates to compensate different reporting approaches or underreporting. But mostly it's a result of projection of old data. Report (WHO 2015 report is out btw) states that for countries with traffic-related death registration (Group 1): "Countries where the latest death registration data submitted to WHO is earlier than 2012, but not earlier than 2005. Deaths in year 2013 were estimated based on a projection of the most recent death registration data using the trends in reported surveillance data: this category contains 54 countries." (p. 71). So when actual data are aligned to the trend line, the results are close. Otherwise the estimates are very distant from the actual numbers. E.g. WHO estimate for Belarus which you reverted to are based on a report submitted to WHO in 2011 hence contain statistics max up to 2010 or even earlier. So they just did linear regression for some year range, for example 2004-2010 and got a projection for 2013. But we had a rapid decline in traffic-related deaths so the projection is not even close for 2013 (1282 estimated vs 894 reported) and even more distant for 2015 (658 reported). The same for many other countries in the report (e.g. exactly the same for Lithuania). So you either should allow us to set right numbers for our countries or add another column with official reports and state that WHO estimates are also projections that turned to be seriously off in some cases. — Preceding unsigned comment added by 178.122.94.117 (talk) 14:18, 18 January 2016 (UTC)

On p. 71 I find nothing but the statistical data of Benin. So please tell me where I can find your cited source.

Secondly, if there is a new report, go ahead and update the data – for all countries of course. But please use the column point estimate of table A2!!

Thirdly, pls do not forget to sign your talk contributions. -- ZH8000 (talk)

It is to be found on pages 48-51!!

You misinterpret the text too far. Belarus belongs to group 1 (Countries with good death registration data). Belarus provided data including 2010 (latest reported year!). So there were no estimates about Belarus' figures. The adjusted figures made in column reported numbers of road traffic deaths and point estimate in table 2 are solely due to in order to reflect the different reporting and counting methods among the many countries. So no! Your interpretation is wrong. The adjustements made to Belarus' data are not projected, but based on factual data. But adjustments were made to reflect the different reporting and counting methods, of course.

This is indeed a very good, prototypical description, why it is so difficult to make seriously comparable lists among different countries. And actually the only way to do it. Therefore to arbitrarily update the figures by locally published data, but neglecting WHO's adjustment method, is futile and counter-constructive. I will therefore re-adjust Belarus' and other changed data. -- ZH8000 (talk) 10:44, 26 January 2016 (UTC)

Hi, ZH8000. Yeah, I should create account I guess as I have been more and more active on wikipedia lately. 2015 report is out indeed and you can find it at WHO site. As for methodology, please refer to Methodology section. 178.122.70.245 (talk) 22:15, 26 January 2016 (UTC)

Revert of Belarus Figures

To 178.122.115.174: Sorry if I reverted correct Belarus figures. I did not intend to do so. Mea culpa.Seligne (talk) 23:33, 6 June 2015 (UTC)

Additional column 2014

The User:Rudolph Davis added an additional column "Road fatalities per 1,000,000 inhabitants (2014)" on the basis of the source (besides, without citing it).

Even though I understand the intention (availability of more current data) and I do acknowledge his effort, I do not think it is a very useful task, rather the opposite.

Why?

• It does not use a (single) world-wide accepted and valid source(s).
• It will always be and stay (more than) incomplete (thoroughly proven by the given example and source).
• The significant advantage of a world-wide acknowledged organisation, such as the UN, taking care of it and providing harmonising work will be fully lost:

Advantages:

• WP:NPOV
• Harmonizing the different methods to measure and collect the information among the many countries and institutes, in order to make them really comparable!
• Seriosity: Un has the ability and competence to do so.

• As many other trials to list (competitive) ranking lists here on WP sadly examplify, the attempt to provide the necessary NPOV and harmonizing work (in this list's case: impossible! ... besides being [[WP:OR]) by the WP community thoroughly fails in most cases (I don't know any where it does not fail), given the many indissoluble disputes they generate.

Finally, it should be (IMO) not the intention to provide a ranking list, BUT useful, seriously "pre-processed" information. This cannot be fulfilled by such a column about "the most current data available", even though, as far as I understand it, it can be tempting (eventually motivated by chauvinistic wishes?).

Therefore, I will reverse Davis' work again, for now. But of course, the discussion about such a change can go on (and eventually lead to its acceptance, however without my support).

Feel free to provide your opinions on this issue. -- ZH8000 (talk) 12:14, 31 August 2015 (UTC)

None of the data is correct by the WHO page cited, often by several figures.

131.252.69.92 (talk) 02:20, 23 April 2017 (UTC)

Can you give an example? Some of the sources count things differently (eg some only count vehicle occupant deaths without pedestrian deaths and others count both). The numbers have been normalised to put all figures on an approximately equal basis.  Stepho  talk  09:13, 23 April 2017 (UTC)

The article claims that there are 4984.3 fatalities per 100,000 cars in Ethiopia. 5% of drivers dying per year strikes me as too high. The actual source for that statistic says it's 25.3 per 100,000 cars. The Democratic Republic of the Congo is even more extreme: 6,405.4! But the source says 33.2. Look for yourself: http://www.who.int/violence_injury_prevention/road_safety_status/2015/TableA2.pdf

The data in this article is gobbledygook. How could anyone think that Guinea kills of 9.5% of its driver per year? Does that really sound plausible to anyone? 131.252.69.92 (talk) 01:12, 26 April 2017 (UTC)

For Ethiopia, the 25.3 is for every 100,000 inhabitants - for which the source and the table agree. The 4984.3 figure is for every 100,000 vehicles. This figures shows that they have only a few cars and really bad drivers - not surprising for a third world country. The fatality rate is probably magnified by overloading the vehicles with too many passengers (think of trucks with 30 people standing in the cargo area - I have personally rode one of these in China) and accidents involving driving into crowds - also not surprising in third world countries.  Stepho  talk  23:07, 26 April 2017 (UTC)

An other issue is under-reporting: counting accurately how many people have been killed in the 30 days after an accident is not so easy. I believe that for that reason the WHO applies some correction factor. Today, it is written there are 385.7 fatalities per 100,000 cars in Ethiopia. This assumes around 7 millions cars in the country. — Preceding unsigned comment added by 88.136.217.216 (talk) 14:44, 15 December 2019 (UTC)

Missing countries

moved from my talk page. -- ZH8000 (talk) 13:50, 23 March 2018 (UTC)

ZH8000, I noticed that several of your edits in Jan 2016 removed a number of countries from the table in this article. I'm sure this was accidental as I know how difficult these edits are. The following list is all the countries that appear in the WHO tables but don't appear in the table tho I haven't checked all of them to those edits:

Brunei, Burundi, North Korea, Djibouti, Haiti, South Sudan, Syria, Venezuala

Someone a little while ago accused me of an error in the world map, which is at the beginning of the article, over Venezuala and I didn't think to check this table at the time, but it could be the reason for the query. Since I don't have data for all the columns at hand it would be much easier for you to do this, but I could make a start if you don't have the time. Chris55 (talk) 18:16, 14 March 2018 (UTC)

@Chris55: Hi, I am sorry for my late response. I see your point. In Jan 2016 I just updated the list according the WHO 2015 report. Somehow I forgot Djibouti (I just added it). The rest is simply not listed. So I have no data for them. -- ZH8000 (talk) 13:50, 23 March 2018 (UTC)

Left/right traffic

Why is left/right traffic included as a column? Presumably this is to encourage the reader to infer a relationship with the numbers? If there is a connection (with RS) this should be stated, otherwise there are probably more relevant stats that could be included, such as % car ownership, urban speed limits, etc. Paulbrock (talk) 10:08, 25 November 2017 (UTC)

I wondering the same thing. Why not to rely on any source? — Preceding unsigned comment added by 86.67.188.170 (talk) 23:37, 23 April 2018 (UTC)

I think it should be deleted. The table is already rather wide, and I don't see how it adds any relevant information. Probably more relevant would be something like per capita income - which I'm not suggesting we do, but is more relevant than left/right. Adpete (talk) 00:19, 29 October 2018 (UTC)

I've removed it. Adpete (talk) 02:03, 29 October 2018 (UTC)

Map out of sync

Hi, The map is completely out of sync with the table. Please correct. Thanks, Yann (talk) 16:32, 23 October 2018 (UTC)

UK

The UK seems to be missing entirely. Nothing for "Brit" or "England" or anything. — Preceding unsigned comment added by 86.137.49.220 (talk)

It's under "United Kingdom". MPS1992 (talk) 16:39, 4 May 2019 (UTC)

Hong Kong

ZH8000, I noticed you deleted Hong Kong, citing the reason "HK is not a country". But Honk Kong is listed as a separate country in most of the popular indexes that also appear on Wikipedia (Human Development Index, World Happiness Report, Democracy Index etc). Adrian Tofei (talk) 23:27, 8 May 2019 (UTC)