I think you want time of day and road type and rural / urban if you're doing a survey.

So numerator and denominator and then if the sample size is large enough it can be analysed a variety of ways but time of day and location would be of value in any analysis

PM rchung about the statistical questions, and you'll have an answer in the time it take him to type it.

(I'm not a statistician, but a survey of about 1,000 respondents yields an error margin of about 3%, if done correctly)

---

----------------------------------
"Go yell at an M&M"

"Or more immigrants riding to/from work during rush hours?"

exactly!

"About 1 in 6 riders killed were wearing a helmet."

now we're talking. this is why we can't use stats that are generalized.

still, we get these threads weekly. so the odds aren't nothing. i'd like to quantify the odds, so that we can present to ourselves, to other athletes, to prospective athletes, what is the real likelihood of death from road riding.

i promise you that if you talk to a lot of triathletes this issue will come up. somebody will say that he or she does not ride on the road any longer, or is considering stopping. this is a threat to the well being of our industry, and i suspect the threat isn't cycling death, but rumors of cycling death.

but i don't think lew kidder would be responsive to this argument today, as his wife, my longtime friend, died day before yesterday from being struck by a car while riding. likewise bridget dawson's family. so we cannot blithely say you're more likely to be struck by lightning. because we knew these people, know their grieving families, i think we need to apply more care to the project.

---

Dan Empfield
aka Slowman

This is all great stuff, and the metrics - whatever they end up being - will certainly be fascinating to breakdown and ponder.

I'm no engineer, but am a bit of an analyst and it seems like overlaying all the data onto a map would be the most telling and be actionable data. For instance, when my hometown bike coalition started mapping car vs bike incidents and ranking them by severity it became very obvious which intersections and roads were the most dangerous to bikes. The city and county governments responded to the bike coalition's advocacy and as a result we now have upgraded intersections at the "death intersections" (where the probability of death was unreasonably high) and we now have colored bike lanes (the entire bike lane is filled in with green or yellow paint) and rumble strips, etc. It has been a game changer for bikes, yet we still have had numerous cyclists hit and killed in the past year.

Despite the incredible work of our local bike coalition and the responsiveness by city and county government to improve bike safety, sometimes the improvements can't be made fast enough. Bridget Dawson was struck and killed on a stretch of road that had a very wide shoulder, was well traveled by bikes, and DOES have a higher than acceptable rate of car vs bike incidents, but the road improvements haven't been made yet (still in the planning stages with the County). Perhaps if the rumble stripping or upgrading the class of the bike lane had already been done, the accident might not have occurred.

Regardless of what metrics emerge, I hope the ultimate goal would be to arm local bike advocacy groups (whether formal or informal groups) with the data so they can lobby their local governments for road safety improvements. If you generated these statistics, overlaid on a map, and said that my county's weighted rate of incidents were higher than the national average it would give us tangible reason to insist that the government focus on road safety improvements. On the flip side, if you said that my county's weighted rate of incidents were lower than the national average, or we were in the top 10 safest bike counties nation-wide, we'd have reason to celebrate and could serve as an example to other communities on how they could improve their safety. Being able to compare safety metrics before the improvements versus after the improvements is the proof that the changes/improvements worked.

It is really easy to be horrified by the accidents that get posted here regularly, especially when they happen in your own backyard and on routes you travel regularly. And it is really easy to generalize the drivers of the cars as complete morons (which they are). Hell, the lady who mowed down Bridget Dawson claimed to be praying (had her eyes closed) when she drifted into the shoulder. Now we all need to be afraid of praying drivers! We really can't change individual behaviors without constant mass communication - such as "don't pray and drive" and "don't drink and drive" and ..... on and on. That is already happening all around us. But we can try and change road conditions. And I think that is where our power is. Get us data, get us lots of advocates, and insist that the government create a safe traveling experience for bikes and cars alike.

Not to derail the topic, but risk assessment is a constant for us athletes, and especially for female athletes. Every solo run I take I am risking rape and murder. Every outdoor ride I take I am risking a car strike. Swimming is pretty safe, but there are some pool denizens that creep me out (like the underwater guy in the corner). Unless I swim in the ocean and get eaten by a shark (happened to a gal at my local beach - Avila), which is pretty deadly. I decided a long time ago that I wouldn't live in a bubble (aka pain cave) and would instead be an advocate for safety. Through that effort, I learned that the government needs data if they are going to be a partner is safety. So, yes, please more data and statistics. Rock on, Dan.

---

Hillary Trout
San Luis Obispo, CA

Your trip is short. Make the most of it.
https://www.slogoing.net/

This site could be useful, too: https://bikemaps.org . It tracks and maps collisions and problems. It can be slow to load.

LOL

---

http://www.jt10000.com/

We can be careful, and respectful. My heart goes out to the families of these incidents.

However, what you are doing is the same thing that states do to promote their lotteries. The law of large numbers allows us to make the following predictions with near certainty about the next 12 months:

• one or more triathletes will die in a bike accident; and
• one or more people will win the Power Ball lottery.

The law of large numbers also says the following, also with near certainty:

• neither of the above will be you.

We all know someone who lives near to us, or works in the same field, or to which we might have some vague connection, and who has won the lottery. Because it is a smaller cohort, we all know someone directly or through one or two connections, who was killed in a bike accident. These things bias our perception of reward/risk.


I'm one of the more ardent (or strident, depending on your point of view and my mood) supporters of riding on the road. If I were to die in a bike accident tomorrow, some here would take that as confirmation that their risk perception is accurate, when in fact it would merely be irony.


The number of hours, or miles, or riders, is so big, and the number of fatalities among triathletes is so small, as to make the risk very, very small. Throw out the deaths that were entirely preventable by the rider (riding in sketchy conditions, riding drunk, riding through intersections illegally, riding without a helmet, etc.), and the risk is made vanishingly small.

---

----------------------------------
"Go yell at an M&M"

"overlaying all the data onto a map would be the most telling and be actionable data."

when i was walking the dogs this morning, and thinking all this over, that very thought came to me as well. we already do have mapping technology on slowtwitch (go on our home page and scroll down to the "map of everything"). what we have now is just the beginning of what i want that map of everything to do. i want specific data sorts to be available to readers via a pulldown menu, so, bike fitters who have met a certain floor standard; or where triathlon bike fatalities have occurred since whatever the inception date we choose. when you click on the incident icon you get a pop up of all the relevant data about that incident.

---

Dan Empfield
aka Slowman

http://www.cityam.com/...-are-there-each-year

heat maps, hour of day, location, severity - self reporting would get you a long way. I suspect the problem is outside of urban areas the geographical spread of cyclists is so thin that its difficult to see patterns in incidents

I think this is well said. Looking at the NTSB data for 2014 (just published in 2016), the actual risk is 0.23 deaths per 100,000 people.

I'm not sure how you could use the site here to get meaningful data. Defining the actual number of rider hours/miles per year would help in knowing what the denominator is to more accurately calculate the risk, but how to do it and how to use social media/ST/internet sites?

Education about how to ride appropriately and safely is likely the best thing that can be done, IMO. The highest death rate is older males and the second risk population are adolescents. Quite a different population/skill set etc. Being a good role model/parent and wearing a helmet and making your child wear one is a good start. We're not going to change risk behavior on a website, but education is key. Safer roads would help, but we all make choices. I know I am not stopping riding outside.



Here is a link to the 2014 data on cycling fatality-cycling information starts on page 152
https://crashstats.nhtsa.dot.gov/#/DocumentTypeList/12

http://www.pedbikeinfo.org/...tsheet_crash.cfm#No1

and a nice article about 'fear mongering' related to cycling.
http://bfw.org/...g-a-bicycle-is-safe/

And simply having a helmet on reduces the risk by 50%.
http://www.bhsi.org/stats.htm

The data suggest that the actual injury rate over the past 20 years is going down and the death rate had been unchanged year over year.



Quick Facts

• Bicyclist deaths in 2001: 732
• Bicyclist deaths in 2014: 726 (NHTSA Traffic Safety Facts)
• Reduction in bicyclist deaths between 2001 and 2011: 7.5 percent
• Bicyclist injuries in 2001: 45,000
• Bicyclist injuries in 2014: 50,000 (NHTSA Traffic Safety Facts)
• Increase in bicyclist injuries between 2001 and 2014: 11.1 percent
• The total cost of bicyclist injury and death is over $4 billion per year (National Safety Council).

Highlights of US statistics available from the US Department of Transportation: Traffic Safety Facts - 2014 Data (released in June, 2016 and still the most recent). Their data has only deaths and injuries related to car crashes.

• 726 bicyclists died on US roads in 2014 (743 in 2013, 726 in 2012. 1,003 back in 1975)



• 50,000 bicyclists were injured in traffic 1n 2014, up from 48,000 in 2013 (52,000 in 2010)





---



• Bicyclists 14 and under killed in 2014: 50 (7% of total. Injured: 6,000



• Bicyclist deaths represented 2.2% of all 2013 traffic fatalities. (2.3% in 2013)



• Average age of a bicyclist killed on US roads: 45 (36 in 2002)



• Males killed 87%. Males injured 83%.
  
  
• Nearly one fifth (24%) of the cyclists killed had been drinking. (Blood alcohol over .01 g/dl)



• Fatal crashes typically were urban (71%) and not at intersections (56%).
• Fatalities were about half during daylight hours.

Who is getting killed in bicycling crashes?
A detailed breakdown of the age, gender, and location of bicycle crash victims is available from The National Highway Traffic Safety Administration. Some of the more noteworthy trends and numbers are:

• The average age of bicyclists killed in crashes with motor vehicles continues to increase, climbing to 45 years old in 2014, up from 39 in 2004, 32 in 1998, and 24 in 1988.
• 88 percent of those killed were male.
• 71 percent of bicyclist fatalities occurred in urban areas.
• 20 percent of bicyclist fatalities occurred between 6 and 8:59 p.m.
• 19 percent of bicyclists killed had blood alcohol concentrations of 0.08 g/dL or higher.
• In 35 percent of the crashes, either the driver or the bicyclist had blood alcohol concentrations of 0.08 g/dL or higher.
• California (128), Florida (139), and Texas (50) lead the nation in the number of bicyclist fatalities.
• Just two states, Rhode Island and Vermont, reported no fatalities in 2014.
• 
  

Causes of injury
According to the 2012 National Survey on Bicyclist and Pedestrian Attitudes and Behaviors, nearly a third of all injuries are caused when bicyclists are struck by cars.
Six most Frequent Sources of Injury Percent Hit by car 29 Fell 17 Roadway/walkway not in good repair 13 Rider error/not paying attention 13 Crashed/collision 7 Dog ran out 4

How can bicycle-related injuries and deaths be prevented? Effective InterventionsEffective interventions to reduce injuries and fatalities to bicyclists include the following: Bicycle helmets
Bicycle helmets reduce the risk of head and brain injuries in the event of a crash.5 All bicyclists, regardless of age, can help protect themselves by wearing properly fitted bicycle helmets every time they ride.
Bicycle helmet laws for children
These laws are effective for increasing helmet use and reducing crash-related injuries and deaths among children.6
Bicycle helmet laws for adults
These laws increase helmet use among adults.6
Promising Interventions
Interventions that have shown promise for reducing injuries and fatalities to bicyclists include the following:
Active lighting and rider visibility

• Fluorescent clothing can make bicyclists visible from further away than regular clothing during the daytime.6
• Retro-reflective clothing can make bicyclists more visible at night.6
• Active lighting can include front white lights, rear red lights, or other lighting on the bicycle or bicyclist. This lighting may improve the visibility of bicyclists.6

Roadway engineering measures
Information about roadway engineering measures, like bike lanes, that can improve safety for bicyclists is available from The Pedestrian and Bicycle Information Center.



2014 NATIONAL STATISTICS-US Dept of Transportation
PedalcyclistsKilledorInjuredandFatalityandInjuryRatesper100,000Population,byAgeandSex-by cause

Failuretoyieldrightofway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
196
27.0
Failuretoobeytrafficsigns,signals,orofficer . . . . . . . . . . . . . . . . . . . . . . .
70
9.6
Notvisible(darkclothing,nolighting,etc.). . . . . . . . . . . . . . . . . . . . . . . . . .
61
8.4
Undertheinfluenceofalcohol,drugs,ormedication . . . . . . . . . . . . . . . . . .
49
6.7
Wrong-wayriding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
35
4.8
Makingimproperturn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
34
4.7
Dartingorrunningintoroad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
28
3.9
Impropercrossingofroadwayorintersection . . . . . . . . . . . . . . . . . . . . . . .
24
3.3
Operatingwithoutrequiredequipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
22
3.0
Ridingonwrongsideoftheroad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
22
3.0
Improperorerraticlanechanging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13
1.8
Failingtohavelightsonwhenrequired . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11
1.5
Inattentive(talking,eating,etc.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10
1.4
Failuretokeepinproperlaneorrunningoffroad . . . . . . . . . . . . . . . . . . . .
8
1.1
Makingimproperentryorexitfromtrafficway . . . . . . . . . . . . . . . . . . . . . . .
7
1.0


Some older data here:
Cyclist fatalities occurred more frequently in urban areas (66%), at nonintersection locations (67%), between the hours of 5 p.m. and 9 p.m. (30%), and during the months of June, July, and August (36%). (NHTSA, 2004)
89% of fatal bike crashes in NYC occurred at or within 25 feet of intersections. ("Bicyclist Fatalities and Serious Injuries in New York City", PDF, NYC government, 2005)
33% of cycling fatalies were at intersections. (Treehugger, 2009)
The most common bike-car collision is a driver failing to yield at a stop sign. (Bicycling Life, 1995)
Study shows the most common kinds of bike-car collisions. (Bicycling Life, 1995)
Riding the wrong way (against traffic) is three times as dangerous as riding the right way, and for kids, it's seven times as dangerous. (Bicycling Life, 1985-89)
Riding on the sidewalk is several more times more dangerous than riding in the street. (William Moritz, 1998) Another study says it's twice as dangerous. (Bicycling Life, 1985-89)
Most deaths on major roads. Fifty-seven percent of bicycle deaths in 1999 occurred on major roads, and 37 percent occurred on local roads. (6)
Streets with bike lanes have a significantly lower crash rate then either major or minor streets without any bicycle facilities (38 and 56% respectively). (William Moritz, 1998)
Streets with bike lanes are safer than those without. Article also has information about the safety of bike paths. (BicyclingInfo.org, 2004)
Texas leads cycling deaths. Texas ranks 14th in number of cyclist fatalities per capita. (5)
Four states lead cycling deaths. Four states (California, Florida, New York, and Texas) accounted for 43% of bicycle deaths in 1999. (6)
Austin stats. Statistics specific to Austin are listed below.


And Austin, Tx used to be a pretty dangerous place to ride. (old data)
Stats specific to Austin
A good collection of statistics regarding car-bike collisions in Austin is available on the bicycle section of the City's Safe Communities page. Here are some excerpts from their report (percentages have been rounded):

• An average of 2.4 cyclists dies each year in Austin. (From 1980-96, the fewest in any one year was 0, and the most was 7.)
• Of the 41 cyclists killed between 1980-1996, 39 were struck by motor vehicles. (>95%)
• 80% of the victims were age 39 or under.
• From 1990-96, 12 cyclists were killed, but another 1,222 were injured. Of those injured, around 15% suffered incapacitating injuries (i.e., needed help leaving the scene of the crash).
• Surprisingly, from 1994-96, nearly 78% of cyclist injuries occurred during daylight hours, which a huge chunk of the total (42%) occurring during just the four-hour period of 4-8pm.
• However, although nighttime collisions are more rare, they're also more dangerous. While only 22% of all injuries occurred at dusk or at night, 46% of fatal and incapacitating injuries did.
• From 1994-96, 57% of bicycle injuries either occurred at intersections or were intersection-related. Another 16% occurred at commercial or residential driveways. Since driveways are really a kind of intersection, ~73% occurred at some kind of intersection.
• Bicyclists were judged to be in violation of the law in 62% of crashes from 1994-96.
• Of the 180 cases in which the police noted motorist-related crash factors, citations were issued to motorists only 35% of the time. (1994-96) For example, in the 39 cases in 1998 where drivers caused or contributed to bike collisions by failing to yield the right of way, citations were issued in only 14 cases.
• While most collisions occurred on city streets, those occurring on the frontage road of I-35 or on county roads were about twice as likely to be fatal or incapacitating (probably due to the increased speed of automobiles on these roads).

Another link to helpful suggestions about avoiding different types of collisions.
http://bicyclesafe.com/


It's always about choices and education about actual risks. I feel bad when I see or hear people who have been involved in bike tradegies, and as a medical professional myself, I hear about it a lot-because people know I bike.

Seems to me there are some obvious take home messages that we should personally always do, and I'm sure most here already do these things.
Wear a helmet and make your child.
Don't drink and cycle.
Be extra cautious between 6 and 9pm on a weekend or 3-6pm on a weekday-or choose very safe roads as this appears to be a high risk time of day.
Ride in numbers-safely, following the rules of the road.
Wear bright clothing.
Ride on safe streets with bike lanes.

And do not worry about the risk as much as people would have you believe. The risk is low, but the choice is always our own.

Dale

I think that confirms what you might intuitively think.

• Morning and afternoon rush hours
• I suspect lunchtime is another factor - the run up to it and a drop between 2 and 3
• Weekends would be different
• Dusk and Dawn - riding in to or away from low sitting sun
• Lorries turning left (right in US)
• riders and drivers dont look at whats coming

http://www.rospa.com/road-safety/advice/pedal-cyclists/facts-figures/

Going back to university I vaguely remember my sampling techniques course (1999!). You can

either
- do a census, ie count the population (both deaths and number of triathlons)
--> due to limited data availability likely associates with substantial measurement error

or
- sample a subset that you can count well and then extrapolate to the whole population
---> this is what I suggest. Less cumbersome

Which subset to choose?
- professional triathletes (you should find all deaths, the total #, and have a good estimate about miles p.a. they do)
- you could also go for region maybe?
- slowtwitchers
- deaths during triathlon races
- maybe corroborating with professional cyclists as a cross check
- combining different subsets can further improve the quality of the estimate

As death is likely a rare event you need to ensure to have enough events so its not subject to randomness. I really don't have signifcance thresholds ready but a guesstimate would be a dozen or more events needed to be able to extrapolate reasonably.

Hope that contributes somehow to start to get going.

Roberto

classic heat map. john snow is your man on this.

another sad fact about stats, though, is that if you want them to be meaningful they have to be complete. the more fields you add, you more likely each observation is to be incomplete. example: i was trying to calculate death rates at a clinic i worked at. i wanted gender, age, cause of death and date of death. if a patient was missing gender on their file, i had to throw the whole thing out.

so data quality is key, and if you're going to do the map, you've got gather solid and consistent location data for all accidents . . .

---

____________________________________
https://lshtm.academia.edu/MikeCallaghan

http://howtobeswiss.blogspot.ch/

Thank you klehner and dtoce - good info and perspective.

---

http://www.jt10000.com/

Here's my +1

Redneck in a Dodge Ram tried to run me off the road on purpose. In a rural area.

A road near a lake on a warm summer day. Boat trailers. Lots of vehicles, usually large vehicles if they are towing, with boats. The boat trailers are often very wide. This one had its right wheels in the bike lane which forced me all the way to the right and I hit a guard rail.

Agreed about additional danger during rush hour (never ride in rush hour), and early morning / late evening with the sun on the horizon.

my thinking is to track the following fields:

activity type: in a race, in training.
nature of accident: collision with car, no car involved.

if no car involved, then the following =>
cause of crash: cyclist lost control; collision with another cyclist; collision with animal (dog, wildlife, etc.); road impairment.

if car involved, then the following =>
driver behavior: alcohol, texting, other distraction, no-fault
driver history: alcohol, license suspension, etc.
vehicle behavior: drifted into cyclist, turned into cyclist, ran stop sign or light, no-fault
nature of accident; vehicle was traveling: parallel to, perpendicular to, cyclist
time of day: morning, daytime, dusk
nature of roadway: bike had a discrete riding space, bike and vehicle shared same space
vehicle density: rural road, semi-rural road, urban road
demographics of cyclist (age, gender, etc.)
experience level of cyclist: beginner, recreational, expert
cyclist equipment: blinking rear light?

what am i missing? mind, either we can track these for substantially ALL fatalities or they will be meaningless fields. so please don't ask for fields not obtainable for every cyclist.

what i imagine doing is setting up some type of database. google doc or something. something online probably. maybe we build something custom and host it here, i don't know, we do host a lot of dbases here. then we just starting listing all the fatalities we know of, and crowdsource the effort among everybody on our own social networks, to find EVERY cycling fatality in triathlon over 3 years.

i can't imagine there's more than several dozen deaths in training and racing in triathlon in north america over 3 years. maybe more. maybe over 100. i just don't know. maybe it's hundreds. i don't think we'll know unless we begin and then see.

what do you think? achievable?

---

Dan Empfield
aka Slowman

It sounds like you are looking for positive predictive value and liklihood ratios. If given a certain set of circumstances, then how likely is it the next event will actually occur. My thought is to be looking for those things that are common in accidents. From what others have reported- drinking, no helmet, urban area, riding at dawn or dusk, specific age, experience. If each of these items are present at different amounts bike deaths, then a likelihood ratio could be developed to say that if you have 3 or more risk factors you have a 1:100 (example) change of being hit and killed. This is similar to heart attacks- high blood pressure, stroke history, diabetes, high cholesterol are all risk factors that point to an increased risk. Realize that there are risk factors that we would need to identify that may not be bike related at all and just intrinsic to a person- age, sex, weight, temper, etc. In the medical field, when you are looking for the best predictor of an event happening to someone is whether they've experienced that event in the past.

"sounds like you are looking for positive predictive value and liklihood ratios"

yes. i'd like to establish a general risk. what is your risk? what is my risk? as a mean? and then look at what causes deviations from the mean, for example, what kind of road i ride? the nature of the road. when i ride THAT kind of road, and so on.

---

Dan Empfield
aka Slowman

How do you define which deceased cyclists are eligible to be included this database? Someone who has competed in a triathlon in the past year? Dan, I think you meet that definition currently because I remember reading about your off-road tri, but I think I've also read that there have been years when you haven't done any races but still considered yourself a triathlete.

Also does semi-rural = suburban?

"What is the definition of a triathlete for this database? Someone who has competed in a triathlon in the past year?"

no. someone who is part of the triathlon community whether or not they've raced in the prescribed time period. i think there are roughly 500,000 triathletes in north america, but i suspect only half of them have raced in the last year. i haven't raced in about a year, but i've spent a lot of time riding my bike on the roads. if i die riding tomorrow, which i might, i should be a part of the cohort.

so, someone who:

1. has raced triathlons competitively;
2. remains part of the triathlon training/racing ecosystem.

that established, i welcome any insight as to how to identify the cohort and identify the cohort size (beyond what i already do).

---

Dan Empfield
aka Slowman

Dan
I would point out that this data is pretty much collected by several agencies already for deaths, so the only unique information would be for triathletes, racing or training, that resulted in a fatality. And if an accident results in death, it would be quite challenging, if not infringing, to try to find the data needed to populate the required fields ideally. And would the families involved want information about accidents that resulted in the loss of a loved one even published on a website?

"I would point out that this data is pretty much collected by several agencies already for deaths, so the only unique information would be for triathletes, racing or training, that resulted in a fatality."

as was pointed out earlier in the thread, only 1 in 6 bicycle fatalities were wearing a helmet. does that sound like relevance for us? i would argue there are almost no stats that you and i can reliably use.

"would the families involved want information about accidents that resulted in the loss of a loved one even published on a website?"

two answers. first, you never know unless you ask, do you? second, i wouldn't assume i'd publish the details of each accident individually, rather the data in the aggregate.

some data might not be reasonably collectable. in which case, we tried, but that becomes an unreliable field. nothing ventured, nothing gained. we have other, reliable fields.

---

Dan Empfield
aka Slowman

Patient identifiable information in clinical research is basically a no no

I'm not sure how a private survey sits.

I can in theory tell you that bob was killed athe 7.06am as the sun rose in Feb on an East facing road being stuck from the rear

How bobs family feels about that is not part of the process though they may not be happy

A more palatable way of doing this might be to simply have an online form that collects the data

It is aggregated up and effectively annonymised off line - an exception might be if x number of people were killed on same stretch of road ( i think the likelihood of this is low)

The data is published as a list of factors common to fatalities and serious injuries specific to the trip community

If this is what you intend to do. I apologise as I misunderstood your explanation

I would be curious to know how many are accidents involving solo cyclists and how many involved cyclists riding with others. My gut tells me i know which one is more prevalent but would like to see if the facts back that up.

"A more palatable way of doing this might be to simply have an online form that collects the data"

that's what i had in mind. but i do imagine we'll need to make public the top line, the name, so that people know whether their entry has already been made.

can you imagine an issue here? the subject is no longer living. we're not asking for medical info in any case. USAT has pretty much already done something similar, just with cardiac events. some of these people were still living, some not.

but i don't know what i don't know, and can't see what i'm blind to. hence my interest in what you all counsel.

just, i hope nobody else just posts, "there is a lot of data on this that the government already has compiled." no, there isn't, as i think we've already properly concluded if anyone reads the ENTIRE thread.

i am in touch with folks from certain bicycle advocacy groups and i do think i'll ask them about methodology.

---

Dan Empfield
aka Slowman