A case study of the rise and fall of a research project

A large chunk of my applied research history has been in the field of ‘child pedestrian injury’, which is basically injuries involving child pedestrians and moving motor-vehicles.  This field saw an explosion of research (relatively speaking) in the nineties, so when I got into the field in the 2000s, many of the big questions had been answered.  Nevertheless, I was interested enough in the topic to attempt to stake my own claim, and my participation in this research area is a (somewhat) interesting case study of how research can enter and exit a research program.

My specific research interest was in a particular question: is there a space-time interaction in child pedestrian injury, and if there is, can this information be used to reduce the risk of collision?  By space-time interaction I mean are children who are struck by cars in close geographic proximity also struck at similar times of day.  My hypothesis was, yes, and I explored this idea in a number of different settings.  Here is one of several pieces of evidence I found in support of this idea:


Along the x-axis of this graph is collisions ordered by proximity where ‘1’ is the collision occurring closest in geographical distance, and ’30’ is the collision occurring farthest in geographical distance.  Along the y-axis are correlations between collisions in time of day.  Larger positive values suggest that the times that collisions occur are more similar, and larger negative values suggest times that are less similar.  Values of ‘0’ suggest no correlation.  The key information to take away from the graph is that the closer two collisions are in space, the closer they are in time of day.  The relationship between space and time is not particularly strong, but it is measurable.

It seemed to me that there were two possible explanations for this pattern.  One was that since school operating hours dictate the time children are walking, and children who attend the same school tend to get struck in similar geographic areas, it makes sense that these children are also struck at similar times.  The second explanation is that the timing of motor-vehicle traffic flow is geographically and temporally specific (high in some areas at some times, and low in some areas at some times), and that the risk of collision is influenced by this process of traffic flow.

I applied for and was awarded a grant to study this idea in more detail, and specifically, to investigate how this information could be used for injury prevention.  I wanted to know if changing the times of school operation would reduce children’s exposure to traffic, and in turn, risk of injury.  The idea was to alter school operating times by a few minutes here and there to decouple the walking journeys of children from the periods of highest traffic volume.  The findings are summarized by the paragraph from one paper’s abstract:

Our analysis suggests that it may be possible to achieve an almost 15 percent reduction in the total number of encounters between child pedestrians and motor vehicles during the morning commute by staggering school start times such that the periods of high pedestrian activity are temporally staggered from periods of high motor vehicle activity. Our analysis suggests that small changes in school start times could be sufficient to see noteworthy reductions in pedestrian exposure to traffic.

The analysis I did in this paper was based on some empirical data on traffic and child populations, but also on some assumptions that needed validation.  I sought funding for follow-up research to explore the question more fully, but that funding did not come through.  After some careful deliberation, I decided that the idea had no more traction, and I put it to pasture.

This ended my journey in research on child pedestrian injury, or at least this dimension of it.  As it stands, it is highly unlikely any more work will be done in this specific research area; there is less research in child pedestrian injury these days, and moreover, the idea is complicated enough to discourage any casual forays.  So the question will probably never really be answered with any more certainty than exists in the answers my past research in the area has provided.

Thousands, if not millions, of research projects die this way–an idea germinates, develops for a while, offers some interesting preliminary results, but then is just not compelling enough to be explored to a natural conclusion.  This could be because the impact of the research is low, or the methods are inappropriate, or the idea is just not that important to a large enough audience.  Whatever the reason, at some point one has to surrender to the inevitable, and move on to something else.