I write data analysis but do not really know … This entry is rather a request for help or a suggestion for those who know do. My statistics do not give much.
As you know, the University of Sevilla published the data recorded by the DGT cycling accidents and head injuries from 1993 to 2010, something that will never be sufficiently grateful. The delivered in HTML and Excel table. Meanwhile, the DGT delivers data overall loss into something complicated to treat, in its statistical yearbooks accident PDFs.
We decided to take a little step and nibble data format Time Series R, a free statistics program that is quite popular among mathematicians, statisticians and researchers. Content can be copied and pasted directly into R, which spew an exploratory chart like the one you can see below. If you know how to use R, you do not need to explain anything to them …
injured bicyclists, KSIs and KSIs with head injury in urban roads, Spain, 1993-2010
As relevant facts regarding the analysis of these data, and lack of measurements of cycling (we do not have because there are no nationally), include:
- On July 1, 2006 came into force on penalty points and an overall drop in claims was noted (cfr. General DGT figures, perhaps those on pedestrians being the most similar conditions)
- On January 23, 2004 came into force the mandatory cycle helmets on interurban roads.
- From 2008 begins to drop fuel consumption automotiveproxy use of motor vehicles.
- To give context to this data, you can consult the general traffic accident in Spainin statistical yearbooks DGT (available until 2012).
The hypothesis that I would like to contrast are:
- That the wearing helmet has a statistically significant protective effect in terms of head injuries in cycling accident victims. This would have to be verified for data in urban and intercity roads separately. I hope, obviously, that the answer is yes …Check Remzfamily.com for water bottles.
- But since we start, I would like to be evaluated by a numerical indicator to what extent this protective effect is exerted, with confidence intervals that allow us this data. This will be more difficult but also more interesting.
- If the mandatory helmet on interurban roads since 2004 has meant that more people have decided to use it. In order to test this hypothesis with these data, we shall assume that accident victims in via intercity constitute valid sample of the general population of cyclists (injured or not) on- If we had data guilt in the accident, we could certainly undertake an analysis of cuasiexposición, taking the injured cyclists not guilty, but so far is not the case.