- Hauer, Ezra. Observational before/after studies in road safety. Estimating the effect of highway and traffic engineering measures on road safety. 1997.
- Elvik, Rune, Truls Vaa, Alena Hoye, and Michael Sorensen, eds. The handbook of road safety measures. Emerald Group Publishing, 2009.
- World Health Organization. Global status report on road safety 2015. World Health Organization, 2015.
- Ogden, Kenneth Wade. Safer roads: a guide to road safety engineering. 1996.
- Lord, Dominique, and Fred Mannering. "The statistical analysis of crash-frequency data: a review and assessment of methodological alternatives." Transportation research part A: policy and practice 44, no. 5 (2010): 291-305.
There are so many aspects that could be investigated - the problem is that you have not specified what general research subject that you will represent: civil engineering, traffic planning, traveller behaviour?
Describe your domain in science, and we can possibly react.
One more point/question: Why worry so much about the subject before you even have started to explore the field? Whichever field you happen to pick, or be given, there will certainly be enough for you to add on to what is already there - that's how science works.
One very good way to find something to dig into is to read a lot on the subject, broadly first, but also deeply where you feel comfortable.
Look up good surveys that are well-written and make notes on how subfields are connected - what types of studies that have been done, and which ones are better than others. Doing that will mean that you will perhaps find a style of study and writing that you like more than others, so that you can find your own way to frame your work. (Yes, as an author you will develop a style - the question if it will be a good one! :-))
Obey all traffic rules in your area. Impose heavy penalties for anyone "bending" the rules. Show respect for the safety of others especially the most vulnerable, like children and the elderly. Give awards for those observing the rules.
Awards to individuals? Why? What? How? We don't have time to observe such things. What I have heard, though, is that there are prices awarded to communities and countries that are meticulous in their attempts to save as many lives in traffic as possible. Obviously, Sweden wins; the legacy of Volvo's work on safety is also part of the story.
you could explore cultural barriers to improving road safety - it will allow you to compare road safety strategy implementation across different countries and cultures e.g Europe (low fatalities) v Asia (high fatalities)
A key question in many developing countries is whether total trafic fatalities, typically on an upward trend, will ever turn down as they did in a number of OECD countries (for instance, they reached maxima in 18 such countries within the 5 years of 1970-1974). A theory of the existence of this maximum is developped in the enclosed document (slightly corrected as compared to the version already on ResearchGate). The new theory makes it possible, by tracking road transport intensity of GDP, to make a forecast of the turning point and avoid panick forecasts such as those of the World Health Orzanization (WHO). Having worked a lot on Algeria (see other enclosed document), I would have liked to apply it to that country but did not have the time to do this yet. There are a number of other countries, such as Lybia perhaps, where the question can also be asked and tests done. A sub-question is whether passenger-km intensity of GDP and freight-km intensity of GDP should be distinguished or whether they can be kept together as done in AJD 149, which starts the ball rolling on this.
Where are you located and where do you want your work to be effective?
In most developed countries the questions are vastly different than the most pressing questions in a developing nation. In the USA, for instance, drivers have not improved their safety orientation for 40 years because the vehicles are more crash-forgiving. Foolish behavior resulting in a crash that would have been fatal or injury-producing 35 years ago now can result in a totally ruined vehicle from which the driver and occupants walk away. Speed limits have increased not because of better driving behavior, but because the vehicles are more crash-worthy.
I also work in Africa. In Ghana it is a very different situation. Poorly constructed and maintained roads cause many crashes. Pedestrian-vehicle congestion is an ongoing cause of casualties and high pediatric pedestrian fatalities. Assumptions of illiteracy among commercial drivers are often false because these drivers simply cannot afford, or understand the benefits of, corrective eye glasses and optical care. Yet these "legally blind" drivers fill the roadways anyway and can cause terrible crashes.
Make sure that your work has application potential in a specific place and with a specific population. There are too few scientists examining traffic safety as a major issue in public health, economic development, and mental health (for anyone who has tried to commute in urban Accra!).
Thank you to all, for your response and references.
@Marc J.I. Gaudry. Libya as a developing suffers from deficiencies in data records, for example. "Vehicle kilometer traveled ” is an important indicator in road safety, but this indicator in Libya is still not available. The question is, how can i measure that indicator in a country like Libya?
No country has data on Vehicle-km. They are always derived from sales of road fuels (often linked to fuel tax receipts) and other information, sometimes by very complex procedures, such as that developed by Laurence Jaeger for France, which required two years of work:
Jaeger, L. (1997, 1999). L’évaluation du risque dans le système des transports routiers par le développement du modèle TAG. Thèse de Doctorat de Sciences Économiques, Faculté des Sciences Économiques et de Gestion, Université Louis Pasteur, Strasbourg, 1997. Soutenue en décembre 1998. Publication AJD-32, Agora Jules Dupuit, Université de Montréal, 365 p., en 1999 . Downloadable from theAgora Jules Dupuit site.
Also, in constructing series, remember that in regression analysis the absolute level of variables is of little import if one uses elasticities.
Take a representative sample of privste vehicles, commercial carriers, and taxis and pay them for odometer readings fo one month. Then extrapolate to get an estimate...at least for registered vehicles.
I do not have good sources in English, the reason being that the footwork done to construct series, say on road safety outcomes and on vehicle-mileage, is usually only summarized in appendices of theses despite the fact that generation of the database is typically 90% of the total work required for a thesis. Still, I want to encourage you to pursue, and to build your database for Lybia. Most excellent empirical theses use data from home countries of students because they have better knowledge and access to their national data than to data sources in the (foreign) universities where they may be presenting the work and defending the thesis. Having directed theses on road safety in a number of countries (Canada, Norway, Sweden, Germany, France, Spain), I would make the following guidance rules:
A. Decide where you are going. It is a good idea to aim high, for instance for monthly data and not only for yearly data. Assume these data are on road safety outcomes (accidents by category; victims by category) and kilometrage by vehicle category. You can always aggregate them later if other explanatory variables are not available on the lowest reasonable period basis (months, as in Algeria), or cannot be transformed into monthly values with some work and information on their variability.
B. Go to sources on road outcomes. Do not hesitate to visit the civil servants who generate the data and to tell them what you are trying to do. Often they (and even more their female secretaries) will be delighted to help you and will tell you how they do things and give you handwritten tables of data you need. In some countries this is tricky because data on say road fatalities may be controlled by the army (Ukraine, Turkey to some extent) and considered as a state secret. So you have to visit the soldiers and bring them on your side: they can realize that road fatalities are not shameful statistics...
C. The same holds for the fuels sales that you need to generate (with other info) the kilometrage by vehicle category. They may be deliveries, and not sales, and the data calendars might be unexpected (lunar calenders, for instance in some parts of Canada). Often the data generated by the tax authority are not completely clean and you have to clean them with their help. In Quebec, for instance, the fuel volume statistics for road vehicles contained volumes for locomotives and for big trucks working on dam site in the North that had to be removed to obtain volumes consumed on the road, The derivation of kilometrage by vehicle type from various sources is a huge maximum likelihood game.
D. Committes. If you can, invite your sources to seminars to tell them where you are and what your results are. I built committes with police chiefs, doctors, etc. and they will teach you a lot. When you ask for help, most are delighted to respond.
Evaluating the innovative concept of consumer-driven safety modifications to traffic flow, design and regulation/enforcement would be an excellent dissertation project.
- Hauer, Ezra. Observational before/after studies in road safety. Estimating the effect of highway and traffic engineering measures on road safety. 1997.
- Elvik, Rune, Truls Vaa, Alena Hoye, and Michael Sorensen, eds. The handbook of road safety measures. Emerald Group Publishing, 2009.
- World Health Organization. Global status report on road safety 2015. World Health Organization, 2015.
- Ogden, Kenneth Wade. Safer roads: a guide to road safety engineering. 1996.
- Lord, Dominique, and Fred Mannering. "The statistical analysis of crash-frequency data: a review and assessment of methodological alternatives." Transportation research part A: policy and practice 44, no. 5 (2010): 291-305.
Concerning bibliography, you will find a very extensive one covering both aggregate and discrete road accident models and their historical origins, in the enclosed AJD 143 working paper which appeared in 4 parts in RETREC in 2013.
First of all you should red Ezra Hauer's book from cover to cover. If you know GIS it is really helpful. Working with governmental agencies you may be able to find data in the open literature on traffic crashes, their locations, their causes, etc. and make a comprehensive study of pertinent variables and their correlates. If you know the intersections where the crashes occurred, the road and weather conditions, driver characteristics, etc. you have a wealth of data to do meaningful analysis. Refer to the following: