Photo: PSU
Oregonian reporter Jeff Mapes has the scoop on an interesting Portland State University study:
“Jennifer Dill, an urban studies professor, wants to recruit a wide range of cyclists — and lots of them — to tote global positioning system units on their bikes for a 10-day period. She said researchers hope to learn whether cyclists go out of their way to use bike lanes and bicycle-friendly streets and to see how travel habits are affected by such factors as the weather.
Cyclists interested in participating in the study can contact Dill through this e-mail address: bikegps@pdx.edu. She said she’s looking for bicyclists throughout the Portland metropolitan area, although they have to live on the Oregon side of the Columbia River.”
And the last sentence makes a strange reference to cyclists breaking laws:
“Dill also noted that while the GPS unit will track what streets cyclists travel on, it can’t determine whether they’re following traffic laws.”
Hmmm. Not sure what that’s all about.
Sounds like a neat bit of research. Can’t wait to see the results. Read the full article here.
Rats, I thought I’d scooped you Jonathan – but it looks like we posted at about the same time. I truely hope she gets a good response.
I suspect the bit about traffic laws is just to reassure participating riders that if you don’t come to a full stop at every stop sign, you won’t be ticketed as a result of your participation. How many people would volunteer if they thought their minor traffic violations would be monitored? I don’t think there’s a pedestrian, biker, or motorist alive who follows the letter of the law 100% of the time. And if there is, he probably isn’t someone I’d want to know!
maybe it’s so cyclists feel safe in participating (as in the cops won’t give them tickets based on gps data). that is my glass half-full take on it.
The study is invalid before it even starts. The subjects will know what is being studied and may alter their routes to skew the results.
Of course there is going to be a degree error arising from participants knowingly skewing their commute routes. However,this problem occurs almost always in some degree when it involves knowing participants. It is important to keep in mind that this isn’t the study of an abstract social phenomenon, it is finding commute patterns, I think it will be fairly enlightening. I’m sure the Professor has accounted for such problems with the research data and will divulge any problems. I think this study has the potential to be entirely valid, even though it runs the risk of not being as such. I do agree that it is slightly problematic Kat. Maybe they should just stock 175 cyclist to see their routes, although that may skew (scare!) their regular commute routes into something entirely different. HA! Be well all!
I think saying “it can’t determine whether they’re following traffic laws” may just be letting people know they don’t have to worry about that and can ride like they normally would.
Yes, I agree, it appears they do not want the results to be jaded one way or the other by riders changing their routes and styles, due to the fear of being called out for it.
::sniff:: Well, too bad Santa’s in Tulsa, Oklahoma, NOT PDX.
Its not clear to me why subjects would skew the data? I asked to sign up for the study and I’m not going to change my route. How would I benefit from creating a misperception about my commuting habits?
Further, people are always bemoaning the lack of data and this study is an opportunity to have more information. Its not going to be perfect, nothing ever is.
I agree – having advertised the study as one of routes will cause some skewing. Choices made be participants may be seen a policy drivers, or cases that they’d like to see portrayed in the study.
Advertising the study a speed study – or time on bike – or anything else that would distract the participant from the real object would be better. But at the same time this is not a random sample, we are being asked to self-select, so some error is to be expected.
I think a bigger problem in a study like this isn’t that the participants know that they are being studied, but rather the difficulty in getting a representative sample. How can you know if you have an appropriate cross section? There are various techniques that can be used to mitigate the problem, but it’s still a major problem in sampling any population. You also have other potential problem. What happens if you get a spell of dry (or wet) weather during the study duration? Anyway, this isn’t a trivial exercise.
right, seems like you could replicate a blind experiment by just following us around from behind like the cops often do. they already have gps in their cars from what I understand – not that it seems to help them rush out of starbucks in time to prevent crime 😉
or, you could just put a chip in new bikes and then get the data after a few months…but, we are not professors.
everyday, We are generating data that only We know about. glad to see academia beginning to catch up. maybe she can put a gps on sam adam’s bike and see how busy he is?
they are controlling for weather(it seems) by doing it over springtime where the weather would only dissuade a few. they are also striving to get “all different types” of cyclists – but, obviously, 100 of them/us would not qualify as a statistically relevant sample given all of the cyclists in portland. and, of course, the routes that most of us take are heavily predicted by bike lanes, bike routes and low traffic areas. pretty much expect the data to look like any of the bike there maps that I am sure prof dill looked at as part of the lit. review.
maybe, if you did 100 messengers in downtown, you would have a valid study. but, hey, she is getting paid to think about this stuff, so goes to show you how smart we are!!