How Reliable Are Eyewitness Reports?

based on 25 ratings

Talk It Over

Is seeing believing? How certain are you of your facts when you see something happen? How sure can you be that people are reporting correctly when they tell you what they saw or heard?


  • Access to a television that records or a DVD player
  • 5 or more volunteer test subjects
  • Tape recorder


  1. Record a short scene (10 minutes or less) from a television show. Select a scene that has many characters and lots of action. Or you might use a short segment from a DVD. Try to select a scene that will not be familiar to your subjects.
  2. Write a quiz about the scene. Ask specific questions such as the precise words a particular character used or who was in the scene when something important happened. Also, ask for details such as the color of someone's hair or whether a certain character was taller or older than another.
  3. One at a time, have each of your volunteer subjects view the scene. When it is over, turn on your tape recorder and interview each of your witnesses. Ask your list of questions and record the answers on tape. Do not interrupt, correct, or assist your interviewees in any way.
  4. Later, when you are on your own, listen to the taped answers. Make data tables summarizing the responses from all your subjects. Analyze your data in several ways. You might determine, for example,
    1. The average number of right answers
    2. Which questions evoked the greatest number of wrong answers
    3. The percent accuracy for each question:
          • (the number of people who got it right ÷
      • the total number of people who answered the question) × 100
    4. Which questions produced the widest diversity in your subjects' responses

Stay Safe

Whenever you use human subjects in an experiment, be sure to behave in a courteous manner and treat your subjects with respect. Don't use a violent or obscene recording for this experiment. Don't ask your subjects personal questions. Don't insult or embarrass your subjects in any way.

Go Easy

Get an adult's help with the planning, recording, interviewing, and data analysis.

Go Far

You can expand your project by adding other measures of human perception and memory. Try asking your subjects the same questions again 2, 3, or 5 days after they view the recording. Or test two groups of subjects: interview people in one group immediately after the viewing, but wait 24 hours to interview people in the other group. Try asking some trick questions that lead your subjects to say what you want them to. How effective is such questioning in shaping what people think they saw?

Show Your Results

Your data tables will depend on how you designed and carried out your experiment. Start with a simple table like this:

Question How Many Subjects Got It Right? Percent Accuracy

Write short paragraphs stating your conclusions about how reliable your eyewitnesses were and which details produced the greatest number of witness errors. If possible, show your TV recording as part of your project display. Provide copies of your quiz so viewers of your project can test their observational skills for themselves.

Tips and Tricks

  • The more test subjects you interview, the more reliable your data will be.
  • The success of this project will depend in large part on the questions you ask. State them clearly and make them specific enough to distinguish the eagle-eyed observers from their less attentive peers.
  • Make sure the test conditions are the same for all your subjects. Test them in a quiet room away from distractions. If possible, test them all at the same time of day. Don't allow interruptions during the testing, and don't chat with your subjects before, during, or immediately after the test session.
Add your own comment