Tuesday, 21 May 2013

The 2012 Teschke Study - When Scientists Get It Wrong

See that 'cycle track' data point in the above graph? That's what they call an outlier. It's just way out on its own there. I wonder why? Well, John Forester saw the study and he figured it out:
"...what we actually have before us is the result when total ignorance of traffic engineering combines with bicycle advocacy ideology."
"In the much more impressive cycle-track issue, the authors proclaimed enormous crash reduction without informing the readers of the two relevant facts. First, that their data came from only one installation. Second, that that installation was not along a typical city street but in the only situation in which a plain cycle track could possibly be safe, a place without crossing or turning movements by motorists, cyclists, or pedestrians"
Mr. Forester was mistaken on one point - there was more than one cycle track in the study, but all had minimal or no cross streets, so are not at all useful in predicting the results of cycle track installations on standard roadways

The Teschke study (‘Route Infrastructure and the Risk of Injuries to Bicyclists: A Case-Crossover Study’ is yet another example (like the many so-called ‘studies’ done by John Pucher, Ralph Buehler, Anne Lusk, Conor Reynolds and other health and wellness advocates) of how cycling advocacy within the scientific community can create unscientific research that is marred by expectation bias. Dr. Teschke and her associates are involved in public health research and while cycling does have overwhelming health benefits, neither Dr. Teschke or any of the people involved in the study appear to be specialists in transportation injuries or collisions.

In fact, Teschke, Reynolds, Lusk, Harris et al seem to have built much of their careers around support for bicycle infrastructure that most respected researchers find to be dangerous. These folks are essentially paint and path advocates whose advocacy undermines their attempts at science.

The study is based on self-reporting, which is prone to volunteer or referral bias, and nonrespondent bias. Also, like the infamous Lusk study of 2011, it is based on a route comparison, and these have proven to be very prone to selection bias. I would urge cyclists to be very wary of this study, as 85% of studies done by those who ARE experts in the field of transportation find that cycling infrastructure INCREASES the risk of injury, especially at intersections. Almost every study finds that the increases in injuries at intersections more than counteract any safety gains between road junctions.

See the following link for details:

No comments:

Post a Comment