Oct 02 2012

More Inattentional Blindness

If you have never seen the basketball video – take a look at it now.

This is a demonstration of inattentional blindness (or attentional blindness) – when we are focused on one task this interferes with our processing of other information. This is exactly why you should not text while driving, or even talk on the phone while driving.

The cause of this is conceptually simple: our brains have limited processing power, more limited than we would like to think. When we use some of that processing power for one task it is not available for other tasks, even basic tasks like seeing obvious things right in front of our eyes. This concept is called load theory, and researchers have documented numerous ways in which it manifests. A related concept is that of interference – when we perform one task it reduces our performance on other tasks. In fact, the act of multitasking itself causes interference because multitasking requires processing power (it takes brain power to switch among more than one task)  which is taken away from each task.

Interference is probably greater for tasks that are vying for the same parts of the brain. It seems that different areas or modules in the brain participate in multiple networks engaging in different tasks. Placing a processing load on one module for different tasks causes significant interference. Some modules participate in very basic functions, like perception, attention, and memory, and therefore become overloaded very easily.

A recent study has demonstrated a new aspect of this phenomenon. Up to now research demonstrating inattentional blindness has used visual clutter to distract from seeing the target – following the basketball interfered with the ability to detect the gorilla. The new research creates the same effect without visual clutter but instead using visual memory:

Participants in the study were given a visual memory task to complete while the researchers looked at the activity in their brains using functional magnetic resonance imaging. The findings revealed that while the participants were occupied with remembering an image they had just been shown, they failed to notice a flash of light that they were asked to detect, even though there was nothing else in their visual field at the time.

This research suggests that remembering an image uses similar resources to seeing an image – that visual memory and perception are similar in terms of the brain resources that are used (which is in line with previous research).

Assuming the results of this study are reproducible, it extends the implications of inattentional blindness. Not only is texting or using a cell phone distracting while driving, the researchers suggest that trying to visualize directions or remember that image on a navigational GPS system can cause interference and reduce a driver’s ability to detect obstacles in front of them.

There are a few take-home messages from this line of research I would like to emphasize. Our brains do have limited processing power, which means if you want to optimize your performance you really should minimize distractions and forget about multitasking. While driving or performing a similarly important task where the prices of error can be extreme, it is especially important to minimize distractions. If there is someone else in the car, have them navigate, change the music channel, or do anything other than driving so that the driver can use all of their mental resources for the task at hand. If you alone and have to perform a significant task, like figuring out directions, then pull over.

All of this applies even more so to people who have any form of cognitive impairment, including those who have suffered a head injury or are in the early stages of dementia. It’s important to minimize visual and cognitive clutter and to take steps to reduce processing load.

I would also apply this principle to everyday tasks. For example, we use an electronic medical record system at my institution. In my opinion, the interface of this system is incredibly suboptimal – there is a high degree of information clutter on the screen at any one time. I find myself using a large amount of brain processing power sorting through the visual information, most of which I do not need. It’s almost as if the system were designed to be a psychological experiment in which processing load and interference were maximized. This adds to the processing load of actually practicing medicine, which itself is quite high.

In other words, this research provides neuroscientific evidence and justification for some basic principles of programming – to keep output and the user interface as clean as possible, to minimize clutter, and also to keep contrast between different pieces of information high and help draw the user’s attention to where it needs to be while avoiding popups or other distractions. I can attest to the cognitive load of systems that violate these principles.

This line of research also highlights the fact that humans are inherently bad at tasks that require maintaining focused attention for long periods of time, especially when there is a long time between significant events. While driving we are paying attention for unexpected events (like being cut off, or a sudden obstacle in the road) that rarely occur. Every time we let our attention lapse or be distracted by another task we are playing Russian roulette, hoping that nothing unexpected happens at that time, and most of the time it doesn’t. This lulls us into a false sense of security, until the day we get unlucky.


Machines, on the other hand, are excellent at maintaining consistent vigilance. They never tire or become distracted. This is why, in my opinion, self-driving cars have so much potential to reduce car accidents and improve traffic flow. Already Google has cars driving around the city going many miles between the need for driver intervention. Eventually we will likely have fully automated cars, but very soon we will likely have computer assisted driving. A human will still be in control, but an onboard system will help avoid collisions and keep cars in line and at a proper distance from other cars.

The only downside of this I can see is that drivers will become lazy and complacent behind the wheel, comfortable in the knowledge that the onboard system will take over if they lapse. It’s just another way in which we are delegating our cognitive tasks to computers. Perhaps this is not a downside, however. If we can offload a lot of our daily cognitive tasks to computers we can free up human brain power for more important or creative tasks. We may become increasingly dependent on our machines, but also have increasing potential to create and think.

17 responses so far

17 thoughts on “More Inattentional Blindness”

  1. Banana101 says:

    During one of the recent SGTTU’s podcast, I had wondered why no one brought this video when everyone was discussing humans/attention spans and multi-tasking!

    Totally a tangent, but I believe it was the same podcast (or around that time anyways) that questioned the “Hearing Voices” network. You were spot on: It’s totally steeped in anti-psychiatry and anti-medicine. Regardless of which group (I think the podcast mentioned the one out of Australia), they are all steeped in quackery. If you visit their blogs, you’ll see mostly see “Big Pharma” hatred, and news stories about how X medication caused a person to commit a crime. Definite mental illness denialists. I agree, if an MI is not causing one problems, it’s not technically an MI, more or less. But it’s a bit of a stretch to claim such things, in my opinion. They all seem to be die-hard Dr. Szasz and/or Dr. Peter Breggin fans as well –Scientology’s CCHR co-founders. I run a blog which often discredits Breggin (and his cronies). They’re the new charlatans, and all of these groups think you can manage MI’s with *their* books, supplements, what have you. They are also also affiliated in one way or another. How this happened, well, I’m still searching for. Their methods? Reckless at the very least.

    I just wanted to fill you in. Great blog and podcast–keep up the awesome work!

  2. Ori Vandewalle says:

    I totally support the idea of self-driving cars, but I believe their widespread implementation is far off. The problem is that while accidents due to human error happen all the time, we don’t tend to think this is all that big a deal. If we implement self-driving cars, however, and there is a fatal accident that can be traced back to the self-driving bits, there will be a giant backlash against the idea of these cars even if, statistically speaking, they’re safer than human drivers.

  3. That is why I think we will go through a period of computer assisted driving. There will still be a human behind the wheel, but a computer driver will help avoid accidents and can take over completely if necessary (a driver falls asleep, for example). This will probably dramatically decrease highway fatalities, which will go a long way toward public acceptance.

  4. superdave says:

    Prediction: Computer assisted driving will be seen as important to saving lives as handwashing was in the early 20th century.

  5. Mr_Hunnicutt says:

    I think another hurdle will be ironing out the liability when there is an accident. Who is at fault? What kind of maintenance will these systems require to keep them safe and how does that factor in the cost/benefit analysis?
    Perhaps we might first see this technology implemented in buses or taxis, where there is already a dedicated support staff in place to address such issues.

  6. Kris says:

    This is why I get upset when people suggest listing to audio books while driving. The evidence is pretty clear on how dangerous having a phone conversation while driving can be. In my experience, the attention required to follow an audio book where you’re often required to visualize complex scenes and consider complex emotional states of multiple characters utterly dwarfs the attention required to follow a phone call.

    This study seems to back up that impression, and it makes me wish all the more that people would discourage rather than encourage listening to audio books while driving until we have definitive evidence.

  7. Fred Cunningham says:

    The first steps are already in the market place. Auto parking, blind spot warnings, collision avoidance, electronic stability control, and lane drifting alerts. The trick is to get them in a car I can afford.

  8. ferrousbueller says:

    I definitely noticed the gorilla, and I counted the correct number of passes, 16 (the people who made the video missed one near the end). I’ll be sure to present this evidence of my multitasking aptitude to the police officer next time I get pulled over for texting while driving. Thanks, Dr. Novella!

  9. ferrousbueller says:

    Upon closer inspection, I made the mistake. There are only 15 passes. What looks like a quick give-and-go near the end is actually the tall guy stealing a pass from the girl. Poor sportsmanship, if you ask me.

  10. ConspicuousCarl says:


    But do you notice which card changes color in this video?


  11. MikeB says:

    I use the inattentional blindness material to teach new writing students about the necessity of establishing reading periods without distractions: too many students think they are reading when they are simply skimming because their ipods are attached to their heads and their roommate is chatting on the cellphone.

    But the basketball/gorilla video is so notorious that it doesn’t work anymore: many people have already heard of it. Are there other new, unexpected video tests out there?

  12. SARA says:

    #Kris – I agree with you on the audio book thing. I listened to an audio book on a 13 hour trip. And I was unaware of anything during entire 100 mile stretches but what happened in the book. And while that kind of thing happens without the book, it’s not usually quite that dramatic a time frame.

    I also found that when I hit cities I had to turn it off or I would miss all my exits and become a dangerous driver.

    It is a great way to pass time on a long trip, but I in the end the importance of safety should outweigh the convenience of passing time.

  13. How are audio books any different than listening to the radio, which have been standard equipment on cars for half a century? I get just as absorbed in a given episode of This American Life as any audio book.

    This morning, as I was making a left turn off a major street onto a side street, a driver was pulling out of said side street into main street devoting full attention to gazing into their iPhone. Maybe not too stupid to breathe, but definitely too inattentive to drive. Forget computer controlled cars, how about a court orders this person to take public transportation?

  14. ccbowers says:

    “This is why I get upset when people suggest listing to audio books while driving. The evidence is pretty clear on how dangerous having a phone conversation while driving can be. In my experience, the attention required to follow an audio book where you’re often required to visualize complex scenes and consider complex emotional states of multiple characters utterly dwarfs the attention required to follow a phone call.”

    There is no doubt that cell phones are not unique in their ability to divert attention from driving, but I still think that a phone conversation is unlike other examples (e.g. radio, ebooks, w/in car conversations, billboards, etc.) in the amount of attention they require.

    If I am speaking to a person on a cell phone, I am much more compelled to maintain my attention to that person since I am aware that the other person requires certain cues to communicate that I am listening, and I am aware that they cannot see my face or the road for cues on why I am not responding. It also doesn’t help that the voice quality of cell phones is pretty poor (relative to a corded phone- I know, whats that? or close conversation), which makes extra concentration necessary.

    When the communication is one-way (like the radio, podcast, or ebook) I do not feel as compelled to maintain attention to that, since I know it is not relying on me in any way. With an ebook or podcast, you don’t even have to worry about missing anything since you can rewind. Of course, ideally everyone would be driving with vigilance and no distractions, but if you are driving for several hours on a highway your mind will divert its attention with daydreaming which I am not sure is much better

  15. BillyJoe7 says:


    Then you missed a truly mind blowing experience. 😉

  16. locutusbrg says:

    I think we use the same software for records.

  17. Jared Olsen says:

    Steve, I’m sure you’re not suggesting we refrain from listening to, say, the SGU while driving??!

    “It’s almost as if the system were designed to be a psychological experiment in which processing load and interference were maximized.”

    -How do you know it’s not? 😉

Leave a Reply