Your Memory For Construction Workers Is Worse Than You Think (Unless You Are One)

I recently read Bamboo People by Mitali Perkins, an eye-opening novel about child soldiers in modern-day Burma. It tells the story of two boys from different ethnic groups: Chiko, a Burmese boy forced into the Army, and Tu Reh, a Karenni boy whose family is driven from their home by Burmese soldiers. When chance events throw the two together, Chiko and Tu Reh get to know each other not as faceless enemies, but as people.

There's quite a bit of social psychology research on group identity, in-groups, and out-groups, but this story actually brought to mind some vision science experiments on a phenomenon called change blindness. The basic idea is that we notice a lot less than we think we do. For example, watch this video from psychologist Dan Simons.

The man who picks up the phone is a different actor wearing different clothes, but people very rarely notice the switch. We’re less observant than we think.

But maybe we're less observant int his case because it's a video. Surely, people would notice changes in real life! And this is where we get another one of my favorite psychology experiments ever. It's explained in this video here.

In this experiment, psychologists posing as visitors to campus asked random pedestrians for directions. Halfway through the conversation, several people carrying a door forced their way between the speakers, and  took advantage of the distraction to substitute a different person as the direction-asker. Surprisingly, about 50% of the pedestrians did not notice the change in conversation partner.

The psychologists noticed that pedestrians closer in age to the direction-askers were the most likely to notice the switch. They guessed that this was because people paid more attention to individuals in their own social group.

To test this hypothesis, they reran the experiment, but this time the direction-askers were dressed as construction workers. And as predicted, the percentage of pedestrians who noticed the change dropped dramatically. It seemed that pedestrians labeled the direction-askers as construction workers and didn't notice any details beyond that.

I find it fascinating that people automatically sort the people they meet into different groups and adjust the amount of attention they pay to them. It's an interesting question to ask when building your characters. What types of people would your character view as part of her social group, and what types of people would your character see without really seeing?

This week, I am also giving away a signed copy of Bamboo People by Mitali Perkins. There are two ways to enter the drawing.

1. Share this post on twitter and leave a comment with your twitter handle.

2. RSS subscribers will find a secret word at the end of this article. To enter the drawing, e-mail liviablackburne at gmail dot com with the secret word in the subject line.

I will draw a winner on Wednesday, May 11 2011.

Hope you enjoyed the post!  To get regular updates on psychology and writing, use the subscribe option in the left sidebar.

Daniel J. Simons, & Daniel T. Levin (1998). Failure to detect changes to people during a real-world interaction PSYCHONOMIC BULLETIN & REVIEW


  1. Love these types of experiments. Tweeted the post. (@hektorkarl)

  2. Oh wow, that's really amazing and eye-opening. Thank you for sharing the videos with us. :)

  3. This is an amazing study. We can learn a lot from this as writers.

  4. Is there another way to test this? I think expectation plays a bigger role than facial recognition here. If it happened to me, I would just assume that I had gone crazy, because why would a construction worker suddenly switch places with another construction worker in the middle of asking for directions?

    As for the film, we've been conditioned by movies for entertainment. Editing that seems natural to us (and mimics the way we actually use our vision) was not always used and allows us to tell more complex and detailed stories. It is part of that conditioning that if we see someone walk off screen then we jump cut to another location and see a similar character walk in, that it is the same character and they got from point A to point B. It saves the (boring) time of following the character from the front door, through the living room, up the stairs, down the hall, and into the bedroom. Plus, in this video the lighting was very different, which would (to me) be a reasonable explanation for why the man and the outfit looked different.

    But it also reminds me of something my professor said in the first film class I ever took (the one that got me to switch my major!). He told us that as film makers we need to give the audience what they expect to hear rather than true to life sound. He told us a story about a shooting where the neighbors thought it was firecrackers, because they had been conditioned by film to expect a gun to sound a certain way.

    Thanks for the post! It was really interesting. :)

  5. Very interesting! In the first video the actors look somewhat alike, so I think that's easier to miss than the second video's live person switch!
    I will be thinking about how we pay attention as I develop and write new characters! @lesliebulion

  6. I just tweeted this via @moonbridgebooks. I know I am terribly unobservant, and purposely don't pay much attention to people I think I'll never see again. Interesting how you relate this to characters in writing.

  7. Thanks, all!

    Fakesteph -- So it's funny. The people who did notice the difference usually finished the conversation anyways. I think they were just too weirded out to do anything else. But afterwards, when asked if they noticed anything funny, they answered that the person had changed.

  8. Interesting. To be fair, the actors have a similar look. I did notice the shirt. With the door experiment, the direction giver doesn't seem to even look at the asker. Maybe it's more of a demonstration of how we don't really see each other.

  9. I live in South Africa, which is a very violent society. We have about 30 thousand murders a year in a country containing about 40 million people.

    So I think it would be interesting to redo this experiment here, for instance.

    I’m sure most people seeing your video would’ve asked themselves if they would’ve noticed the switch. I certainly asked myself that, and I thought that I definitely would’ve noticed.
    I also believe that most of my fellow South Africans would’ve noticed, too, the reason being that when you are approached by any stranger, for whatever reason, you try to make a very thorough assessment of the possible threat he may pose. Certainly, one of the first things you do is a very careful scrutiny of what he looks like, because you’re asking yourself all sorts of questions such as: Where could he possibly come from? What is his probable level of education? Does he look like he is some sort of addict? Does he look as if he should be in jail?

    So that by the time you start talking to him you already have a fairly fixed idea of what kind of person he is, based on a purely visual appraisal of him. That is, there was a fairly tense visual engagement with the person even before you started to interact. Which, I believe, is very different from the environment where the video was made.

  10. The brain usually retains the information that it thinks it really needs. This is to utilize our memory as efficiently as possible. Those details that it thinks it does'nt need are being erased quickly.