Peer Reviewed Study Questions IIHS Red Light Camera Report

No question that bias/self interest is playing a role in their studies.

The insureance industry would not be biased, would it?

That's why there are independent rapports carefully redacted by RLC vendor. Otherwise results of study can't be trusted. Like here:

Red light camera vendor American Traffic Solutions (ATS) is furious that police in Kansas City, Missouri released a report last week evaluating photo enforcement intersections without giving the company a chance to modify the data. In many cities, ATS plays a key role in authoring such studies, but in this case the Kansas City police department went its own way, arriving at conclusions consistent with other, independent research on the topic.

"The results of the study are very interesting," the Kansas City report stated. "Accidents went up at some locations and down in others without any real clear patterns."
...
Despite the mixed safety results, the cameras over the study period issued 197,494 tickets worth $19,749,400. As a result, the city police commission ordered the police special operations division to re-write the report under the supervision of ATS.

source: http://www.thenewspaper.com/news/37/3703.asp

link to report in article

and thenewspaper.com isn't biased? That's why there's three sides to every story, yours, mine, and somewhere between, the facts.

No matter your, or my view, we can always find a report or statement to support our position. It's up to us to weigh the information being presented and draw our own conclusions.

Actually thenewspaper.com is quoting directly from said rapport (Page 2, Executive summary, chapter to) not commenting on it. I know that for you IIHS and ATS are the only reliable sources of information but it doesn't make them trustworthy. I know you are blindly supporting RLCs but that doesn't change fact, that IIHS and RLC companies are flatly lying with data interpretation. But in those "studies" you believe without doubt.

My biggest contention is, there are no standards to evaluate the information. It seems like every group has their own way of interpreting the records. If everyone would follow the same formula, maybe then we could get useful data.

If you are talking about the IIHS report, the answer is no. Did I read the article and the quote from thenewspaper.com, yes.

However, my comment wasn't directed at the report or its contents, it was to the remark about bias.

@twix, you have identified the major problem in dealing with articles, reports, and studies on Automated Traffic Enforcement. People can not agree on the correct way to do a study (to start with) and then often see in the results what they want to see.

As Box Car wrote above, it is up to each individual to read the material (in addition to the article discussing it) and draw his/her own conclusions.

I'm not talking about reading articles, and taking information that's spoon fed to draw conclusions. I'm talking about taking raw data, and using standards to formulate answers. It wouldn't matter who looked at the data, what city, what intersection, etc. It would all be processed in the same way, so meaningful answers would be apparent.

Do you understand what I'm saying now? And I certainly don't need you to reiterate anything boxcar is saying. I comprehend just fine.

I first commented on the post that said:

I read your original comment as having several points

• standards [for] evaluat[ing] the information
• [same] way of interpreting the records
• follow[ing] the same formula[s]

I thought that your points were on target, and said so (and still think so according to the way I interpreted them).

I also agreed with the comments that Box Car made and “reiterated” them because most people only have access to the articles that report the results of studies.

I did not realize that you were assuming that one could “[take] raw data, and [use] standards to formulate answers” and that these “standards” would apply to any intersection in any city. So, once raw data was processed, “meaningful answers would be apparent”.

Even so, given that standardized analysis was possible, the way that most people would be made aware of results would be through articles reporting on a study. Depending on the author, results can be presented in a biased way. So Box Car's comments still apply.

There are several reasons why I am doubtful that one could take raw data from any intersection and pass that data through a set of formulas and reach a meaningful answer. To be able to do so would require that all of the variables remained constant over the study period. However, consider:

• consistency of yellow light timing
• changes in approach speed limits
• changes in sight lines
• changes in condition of roadway surface
• number and type of businesses along roadway
• signage on RLC
• population growth (or loss)
• new alternative routes since RLC installation
• nearby road changes impacting traffic at this intersection
• changes in type of traffic at this intersection

Other factors are how the “raw data” is determined. Which “accident reports” are considered? Are the accident reports reviewed to determine whether the accident was related to the intersection itself?

How does any analysis deal with the change (and in my opinion, increase) in “distracted drivers” (due to cellphone talking and texting)?

How would an analysis that tried to get at new economic benefit deal with the fact that vehicles are getting safer but more costly to repair?

Still, I agree that any efforts to work by the same standards – as you suggested – would be very helpful.

From the Abstract:

Because a previous IIHS study on RLCs was found to use flawed research methods, as well as to incorrectly report findings, the current IIHS RLC analysis is reviewed for adherence to scientific methods. Our review reveals the 2011 IIHS study is logically flawed and violates basic scientific research methods that are required for a study’s findings to be valid. It has neither internal nor external validity. More importantly, the IIHS did not fully explain the results of its analysis. Correctly interpreting its model’s results actually shows that cities using RLCs had an estimated higher rate of red light running fatalities, specifically 25%, than cities that did not use RLCs in the period “after” cameras were used. Further, the IIHS study was only able to make statements suggesting favorable results from the use of RLCs due to the biased selection of sampled cities.

-----------

Here's the Florida Health Review Study for those who care to look at their methods:

http://hsc.usf.edu/NR/rdonlyres/5FD239D2-785E-4EA6-82C0-603E...

Thank you. I nominate you to come up with the standards.