When we are analyzing a crash, how should we as accident reconstructionists approach what eyewitnesses and involved drivers say about what happened? For some reason, I feel obligated to say the obvious - yes, witnesses are poor estimators of speed and distance. We all know that. But today, I want to encourage you not to use that as an excuse to dismiss everything that a witness says. There's much that could be said about that - and I'll probably say it at some point - but here I'll confine it to simply saying that, in my own practice, I try to separate out the quantitative estimates witnesses give from the qualitative story they tell. There may be important information to be gleaned from witnesses, particularly related to the events leading up to a crash for which there may be no physical evidence.
I say all that by way of introduction. What I'm actually thinking about today is those cases where an accident reconstructionist is presented with two different stories from eyewitnesses or involved drivers about how a crash occurred. For those of us who typically practice accident reconstruction in a legal context, our role is to bring physics and physical evidence into the conversation and to inform our clients or a jury what these can tell us about a crash. My proposal in relationship to competing stories by witnesses is simply this...eyewitnesses and involved drivers sometimes make testable statements - testable with physics and physical evidence. One valuable role an accident reconstructionists can play is to test statements from witnesses and involved drivers and offer an opinion about which of the two stories is more consistent with physics and the evidence. Of course, it's possible that both stories are inaccurate and physically impossible. I don't deny that possibility. I do want to push you, though, not to come at witness testimony with a dismissive attitude. This will only close you off to potentially valuable information about a crash. See the statements witnesses make as severable. Their quantitative estimates are severable from their qualitative statements. Two parts of their story are severable. Most witnesses aren't going to be all right or all wrong. Figure out what you can actually test in their statements as an accident reconstructionists. Test that. Then, leave the rest alone and let the jury decide.
Update (02/18/17) - A couple of the commenters on this article offered extensive thoughts on this topic, which I thought were insightful and helpful. They come at it from a different perspective than I did, and I thought what they said was worth sharing. Thank you, John Daily and Paul Beavis, for offering your thoughts, which I'm including below for the readers edification!
John Daily (Adjunct Faculty at the Institute of Pulice Technology and Management): "Another point to be made about witness testimony in a crash situation is how much time the witness had to observe the event about which he was testifying. For example, let's say a driver is in a line of vehicles all traveling about the same speed. Suddenly, one of the vehicles swerves left and immediately has a head-on with another vehicle. The observer driver is likely able to give a reasonably accurate and, if corroborated by other drivers in the same line, precise estimate of the speed of the line of vehicles immediately prior to the crash. Clearly, the reason is that the observer had the time to see and evaluate the event. On the other hand, if the observer is in a different position, where he can observe the anomaly for only a second or two, his testimony with regard to vehicle speed will be less accurate.
Finally, as interviewers, we have to ask ourselves if the witness is being honest with us. An uninvolved witness mat have more veracity than an involved witness who would possibly benefit from being untruthful. The classic case is the rollover crash where the live guy says the dead guy was driving, regardless of the physical evidence or the testimony of uninvolved witnesses.
Paul Beavis (Forensic Collision Investigator at Northumbria Police): "The difficulty with dealing with witness evidence is often the way their evidence has been obtained. Having taken so many witness statements myself as traffic officer it was so easy to inadvertently and completely without comprehension lead their evidence, ultimately the witness evidence statement became my own translation of their own evidence.
Now as a forensic collision investigator I see the complete and inherent pitfalls with statements of evidence being taken by Police officers. In the modern day world my advice would be that their evidence is recorded verbally and later transcribed, clearly the use of open questions and where applicable closed questions is fundamental to their evidence given, thereby reducing any source of misrepresentation by the statement taker. When reconstructing a collision and comparing the witness evidence to the physical evidence often traits of Police officers can be found within the body of the witness evidence.
When evidence from a witness is obtained via video recording procedures (often due to some level of vulnerability and so on) the evidence offered is untainted by those leading the interview, a typed transcript follows which is not composed of 20 plus paragraphs of the interviewing officer's summaries.
When reconstructing I generally find that the witness evidence is in close agreement with the physical evidence, but often within the body of the report the witness evidence needs to be assisted to link conclusively with the physical evidence. Importantly, for a set phase of the crash sequence, stating what the witness recalls, then the interpretation of the associated physical evidence and by providing a verifiable or suggestive evidence based link between the two can offer some reliability of the witness evidence to the court. As always science doesn't lie but misrepresentation of the identified physical evidence and / or misrepresentation of witness evidence, whether done consciously or not, is always misleading to the court."