Expert Article Library

How Credible Are Eyewitnesses?

By Joseph E. Badger  Email:

Author’s note: The following article was originally published under the title "The Credibility of Eyewitnesses," Law and Order Magazine, December, 1994. Presented here with permission of the author.

Cryptic messages emit from police radios:

"Signal-10 traffic, Car 29. On US 41 at County Road 250N, a 10-50 P.I., possible ten-zero."

"Unit 34, Code R at Elmhurst and Vine, Code 3."

"Car 18, see the man on I-64 at mile post 45 eastbound, 10-50, Code F."

Police officers, or anyone with a scanner, knows that out there someplace there is a bad wreck with injuries, perhaps a fatality. You respond.

Upon arrival you see cars all on and off the road and debris strewn all over the highway. Although it's two o'clock in the morning on a normally deserted stretch of road, onlookers seemingly have come out of the woodwork. Dozens of people are milling around.

You rush to victims lying in the street, call for ambulances and wreckers, and begin the investigation process. What about witnesses?

Whatever you do, don't approach someone and ask, "Did you witness this?" The word "witness" turns people off. They immediately think "I don't want to get involved" or "If I say I'm a witness, I'll have to miss work to testify at a trial or something."

It's better to ask of the group in general, "What happened here?" or "Which way did this car come from?" Someone is likely to chime in with an answer. Don't act like you're God's Gift to Accident Investigation or give the impression you already have all the answers. Play dumb. Make the people in the crowd feel as though they are the smart ones and that you merely want to learn from them.

Once you have a person who is willing to talk to you, don't say "Give me a written statement." Sounds too formal and demanding. Since you are obviously busy, hand them a pad and pen and ask, "While I take some pictures, just write down what you saw." Then go on about your business. When the person is finished, thank them and say something like, "I'll go over this later. In case I can't make something out, give me a phone number where I can reach you if I have any questions." And add, "Oh, put your name down at the bottom so I'll know who to call." If you say, "Sign this," you are putting them on the defensive. With the statement in their handwriting, it can be concluded later that the statement is indeed theirs. If they sign it, fine; but don't make an issue of it; you want them on your side.

Many officers don't want to mess with getting statements in writing; rather they simply listen to what witnesses say and perhaps take some notes. That can come back to haunt you months later. You quote some witness in your narrative and when it comes time for trial the witness says, "That's not what I said." Get it in writing.

Are witness statements, oral or written, reliable? How many times have you had three witnesses who give three different versions of how a single accident occurred? Is someone lying or merely mistaken? Does it sound like these people must have witnessed three different accidents?

Some witnesses try to be too helpful. They will tell you what they think they saw, not what actually happened. Or they hear a noise, look up, and see cars spinning around - then quickly theorize what must have happened - and that's the scenario they explain to you. Other passersby may tell you what they think you want to hear, so don't ask "How fast was the red car going?" Just ask them what they saw. If you use the term "fast," they may think you mean "speeding," and invent all sorts of numbers. "What was that car doing just before they hit?" should elicit a better answer.

After you have accumulated the statements, made all your measurements, taken photographs, and filled in the blanks on your crash report, you get the scene cleared and go back on patrol or head off on another assignment. Maybe there is still a lot more investigating to be done, or perhaps you need to do some reconstruction. This involves vehicle damage analyses, preparing scale diagrams, mathematical computations, and so on.

As you're putting the pieces back together, some things the "witnesses" told you may make little sense. You do the best you can. Months later, when the prosecutor (in criminal cases) or plaintiff and defense attorneys (in civil cases) start taking depositions, you discover that the witnesses or people involved in the accident, have changed their stories. Before being deposed, those people often get to read your accident report and/or other witnesses' accounts. Witnesses often discuss their views of the event with relatives. They tell their friends about "the bad wreck out on the highway." They read newspaper stories and watch TV coverage of the incident. Sometimes, after seeing or reading about an accident, witnesses you never heard of call the station to report what they saw.

How good are witnesses? How credible are they? Do witnesses really remember what they saw? What affects a witness' memory?

According to a report in the St. Petersburg Times, dateline Los Angeles, "Those who study memory are becoming increasing unwilling to trust it." Researcher Henry Roediger, at the American Psychological Association, said that experiments with college students at Rice University in Houston indicate that it's quite easy to introduce false memories. "People confidently remember events that never happened to them," he said.

Still quoting from the St. Petersburg Times article, "When [Roediger] showed subjects a film and then later a written version of the same story with minor changes, they failed to notice the discrepancies. Later, when asked what they'd seen on film, they reported the version they'd read."

If a witness sees an accident and later reads someone else's account of the same accident, will that have an effect of the first witness' recollection of the event? California psychiatrist Lenore Terr says that terrifying incidents are particularly susceptible to memory mistakes because the horror and confusion interferes with the memory process.

"When a witness on the stand says, 'I really, really remember this,' it's compelling testimony," Roediger said. "What we're seeing is that people say, 'I really, really remember' something that never happened."

How reliable, then, are traffic accident witnesses?

While driving through a small town in southern Indiana several years ago, I found myself behind an 18-wheeler. The trucker obviously knew I was behind him and he was traveling about 29 miles an hour in a 30-mph zone. In front of him was a car (which I couldn't see). Suddenly, another car pulled from a side street up ahead. I heard an exceptionally loud screech of tires; and a moment later: SMASH! The tractor-semitrailer came to a stop; I pulled in behind him, turned on my emergency lights, got out and walked around to see two cars tangled in the intersection. Who had the best view of this entire circumstance? Who should make the best witness? Of course, the trucker who was sitting high in the air looking down at the whole intersection. He had to have seen everything unfold smack dab in front of him. I asked him to pull over and stand by. He obliged.

Fortunately, no one was hurt. After taking some measurements and snapping a few pictures, I got the intersection cleared and went over to my witness. "Boy, you were right there," I said. "What'd you see?" He proceeded to tell me that he was just driving along and "This lady just came right out." I asked, "What about the car in front of you?" "She didn't do nuthin'," he said, "she just slammed into the other one." Hmm. Didn't do nothing? Four of the nicest, juiciest skidmarks you've ever seen led right into the impact area and I'd heard the tires slide from at least one hundred feet away. Not wanting to put words in the witness' mouth, I asked, "Didn't she do anything to avoid the accident?" "Nope, she never swerved or hit her brakes or nuthin'," he said. Some witness.

How trustworthy are the accounts given by drivers involved in traffic accidents? Once in that rare while you might have a motorist tell you "Gee, I flat blew that stop sign; I just didn't see it." But often, drivers make very self-serving statements. "Yeah, well I might not have come to a complete stop... but that guy was speeding." Generally, it isn't that the other car was necessarily speeding. It's that when you get crashed into and the side of your car caves in, the hard impact feels as though the other vehicle must have been going fast. Getting hit at 30 mph probably feels like 130 mph at the time.

Both drivers, whose vehicles collide nearly head-on at the crest of a hill, invariably say that as they came over the hill the other car was in their lane. On some narrow county roads that is sometimes the case. That is, both cars are left-of-center. However, even when one car drifts across the imaginary center line, that driver states the other car came over into his or her lane. Did it? In these cases, you've got to go with the physical evidence. Sadly but typically, there often isn't any. No tire marks, no gouges. Just debris scattered everywhere. We have all worked accidents where, by the time we get there at least, there no physical evidence remains. That is especially true in those instances when the Evidence Eradication Team has hosed off the road. Or when a well-meaning wrecker driver sweeps everything away.

When you really need a witness, where are they? And when you've got one, how reliable is their description of what happened? How good are lay people at determining the speeds of vehicles? Occasionally they opine, "He was traveling at a high rate of speed." What does that mean? It is very subjective and depends a lot on a person's vantage point. If you're driving along at the speed limit, say, 40 mph, and meet several cars coming the other way, all of which are going around 40 mph, you might conclude they were "probably doing the speed limit." If, on the other hand, you step off a curb and get "dusted" by one of those cars, then it was probably traveling at a "high rate of speed." You don't admit that you were so dumb or inattentive to have stepped out in front of a car ... so it had to be speeding.

People see (or think they see) strange things when it comes to traffic accidents. There's always that phantom car that only one witness out of five happened to notice. Did four people just not see it or was it really ever there? Was the phantom car merely imagined? Some people rationalize. If they see one thing happen, then something else just had to have happened to cause it.

When you hear something often enough, you tend to believe it. Rumors abound after many accidents. Workers at a factory, who hear that one of their co-workers was involved in a car crash, circulate all manner of tales. Some stories are compilations based on several different versions. "Well I heard he was going 90 miles an hour." This developed after someone hears such comments as (1) "From the damage I saw, he had to have been speeding," (2) "That guy? He always drives fast," (3) "I'll bet he was late for work," and (4) "He paid a speeding fine last week."

What about that damage? I was recently in a salvage yard, checking out a severely deformed car when a curious wrecker driver came over to have a look. After seeing how badly the front of the car was crunched in he asked, "Mercy, how fast was that car going?" "Zero," I said, "the car was parked and an out-of-control truck ran into it." Just because you have a lot of damage to a particular vehicle doesn't necessarily equate to the speed of that vehicle. Witnesses at an accident scene note such things as damage and relate it to speed. Some attorneys simply hold up an enlarged photograph of an apparently demolished car to a jury and offer, "Just look at that damage!" and let the jurors mentally compute how fast it surely must have been going. Damage can be misleading. The outer skin of today's cars can crumple, wrinkle, tear, bend, buckle, and twist out of shape. It doesn't take much force to mess up sheet metal. Hoods, for example, are wont to warp in most frontal collisions. Doors cave in effortlessly when nudged in the side by another car. And it doesn't take much to sheer off the entire top of a car when it underrides a semitrailer. A witness may see only the damage itself and then speculate that the car must have been speeding.

Witnesses often change their stories after they've had time to reflect back on the event. They may make one assertion to police at the scene, then revise their statement after talking with family, friends, fellow workers, attorneys, or insurance agents.

Unscrupulous interviewers, usually an advocate for one side or another, can easily sway witnesses by the way they ask the questions. "Wouldn't you say that the red car was going over the speed limit?" "We talked with Mr. and Mrs. Doe and they said the car ran the red light, what do you say?"

Then there's that time-lapse memory thing. A lawyer calls in a witness two years after an accident. The witness hadn't even thought about the accident until they received a subpoena. How good is their memory? Even if they read their own statement made at the time, is their memory precisely refreshed? Once a person makes up his or her mind that what they said is exactly what they saw, it is almost impossible to convince them otherwise... even if they don't really remember the events at all. Some witnesses are better at articulating a circumstance, are more convincing - even if they're wrong. A practiced liar may persuade you to believe his or her story is factual; whereas, an undereducated but honest individual may tell it exactly the way it was but might not sound too convincing. A polished motivational speaker is apt to come on sounding very truthful even if he or she is mistaken.

Because witnesses may seem credible--and they may truly believe what they are saying is gospel -- doesn't mean they are absolutely correct. Credibility and honesty may have nothing to do with it if their memories slip a bit. Since false memories are often easily induced, and at other times people simply forget a lot, beware of what witnesses say - and get it in writing.