Do professionals have a way of measuring attention span?
September 17, 2011 11:27 AM   Subscribe

Do professional psychologists have a way to quantitatively measure the attention span of humans?

I saw a cartoon about most people losing their ability to concentrate during a 90-minute lecture in a lecture hall. Movies tend to average about two hours long each --- for a good reason. There may be a bell-curve shaped distribution of how good people are at that pair-matching "CONCENTRATION" game. And, people who have ADHD must have a very low value of... ...... what? Have the pros and the psychologists, the ones who came up with IQ scores all those years ago, do they have a way to measure and quantify how good a person's focus and concentration are? What kind of units can "attention span" be measured in?
posted by shipbreaker to Science & Nature (11 answers total) 8 users marked this as a favorite
 
They have all these different tests that assess aspects of attention span like working memory, "executive function", etc. For example, they set up a computer to show Xs and Os on the screen in rapid sequence. you are supposed to hit the spacebar as fast as you can every time there's an O, but not an X. They measure how much you mess up and how fast you go. Then they add some distracting element and do it again.

They use the towers of hanoi puzzle, they use a bunch of cognitive exercises like "how many fruits can you name? how many fruits can you name in alphabetical order?" They make you read words off a card, like the names of colors, and then they write down the words IN colors, but the words are different. So the word "blue" might be in red ink. Then they'll say, "read the color names, unless the ink is red, in which case say red. And do it quickly." And they monitor how many mistakes you make. They do a lot of working memory tests, and then working memory tests while distracting you. Like, "memorize the following sequence of digits." And then you do some other task (like read the colors off the cards), and then they say, "Ok, what was the string of digits?" And they check how many you can remember at different times after you hear the list. They do visual and auditory learning of the things and compare the difference. They make you look at a card with a crazy arrangement of lines on it. Not a shape you'd be familiar with. Then they hide the card and give you a pencil and paper and tell you to reproduce the shape.

A lot of stuff like that. Then they compare it to corpus of test scores from different people and essentially say how much your attention span (or more precisely, how much various better-defined components of what we'd call attention span) correlates with the average person. That's pretty much the units. They don't give you a score like "Your attention span is 200 millizones" or "your daydreamification factor is actually over 4.9 gamma-spaceouts." They just say, "you are in the 13th percentile of whatever specific test and the 64th percentile of this other test . Taken together, these results indicate blah blah blah."
posted by jeb at 11:48 AM on September 17, 2011 [1 favorite]


Response by poster: I am, of course, hoping to find the units named something like MILLIZONES and GAMMA-SPACEOUTS, and not so much the percentile. If it hasn't been invented or officially declared yet, we should do it here in AskMeFi, and say that we were first, and get famous.
posted by shipbreaker at 11:52 AM on September 17, 2011


hmmm...good question...as far as 'quantitative'...i believe they use PET scanners to detect brain activity and can tell if you are 'attentive' or not. however, things we think we are paying attention to, we rarely are, as we spend a surprising amount of time in a low-level trance state. think of it this way: you play the 'concentration' game. the first time, your brain lights up like a christmas tree...the hundredth time in a row? not so much. but you're still playing the game, doing well, and are 'paying attention' to it. (shopping, watching tv, reading...all trance states)...
it's funny you mention movies along with attention span...there was a discovery recently about the most popular movies (highest grossing, oscar winners...etc) and the length of shots in them...specifically the 1/f distribution...my guess is it shocks you out of your trance and keeps you engaged.
but, (and i'm not a clinical psychologist) i'm sure 'average attention span' is probably just measured in minutes and/or seconds.
(by the way, chimps are MUCH MUCH better than we are at the 'concentration' game...video here)
posted by sexyrobot at 11:54 AM on September 17, 2011


All psychology measures are normed; there is no raw unit for anything once you go beyond one individual battery of tests. IQ is not a set measure; 100 means you're average and 110 means you're one standard deviation above average. Percentile scores are the only ones that have any meaning in neuropsychological measures.
posted by supercres at 12:09 PM on September 17, 2011 [1 favorite]


Sorry; SD of IQ is 15.

There is no objective scoring for this kind of thing; it's all relative to the population, hence percentile measures.
posted by supercres at 12:12 PM on September 17, 2011


You can also track attention via EEG. Here is the home version.
posted by Obscure Reference at 12:28 PM on September 17, 2011


As mentioned above, it's a very complicated set of various measurements.

As far as I know, the term "attention span" is 100% an invention of pop culture, for use in "[x] is killing kids' attention spans" where X used to be TV and is now video games or internet.

The argument used to be that TV was bad because it's "passive"; now the same people are against video games even though they require total active attention all the time in order to succeed.

And no one sits through, for example, an hour long lecture without having one thought like "I wonder what's on tv now" or "ow my foot itches." That doesn't mean they get nothing from the lecture. A more useful question might be, "what is attention and how much do we need to learn/succeed at certain tasks?"
posted by drjimmy11 at 12:31 PM on September 17, 2011


Following up on jeb's answer above, one way that people address is by studying the capacity of working memory -- i.e., how many items can you attend to at any given time? The classic answer to this is the magical number 7 (plus-minus 2), which is why phone numbers have seven digits (###-####). More recent research has shown that the limit of working memory is actually more like four items, which is why you can only track up to four moving targets at a given time [Scholarpedia, demo] -- and in fact, work from my old lab -- both published and unpublished -- has shown that you can only attend to a single change at a given time.

What defines an 'item?' Well, that's the million-dollar question, isn't it? :-) While there are hard limitations to how many items can be in your working memory, you can train yourself to pack item information more efficiently; this is called chunking. If you are used to the three-digit area code of your phone number, for example, the three digits are chunked together and are treated as a single item in your working memory.
posted by tickingclock at 1:17 PM on September 17, 2011 [1 favorite]


In the domain of research psychology, there are lots of performance tasks that have "span" in their name. They typically measure short-term and working memory, which are pretty good stand-ins for the lay notion of "attention span."

Although research talks about these two types of memory as distinct entities, it's difficult to define them in a way that draws a clear line. The term "short-term memory" (or STM) is used in contrast to "long-term memory," which is the kind of memory a layperson is likeliest to think of. STM usually refers to memory for very recent events: in some experiments, participants will read a list of items and, after a few hours, are tested on their recall. Researchers are often curious about things you can do between learning and recall that will affect success rates. Does doing a related task in the meantime improve recall? How does it compare to doing something entirely unrelated?

"Working memory" (WM) overlaps with STM to some extent, but most often refers to the process of task and goal maintenance. Effective WM is enormously important for pretty much everything you do. Virtually all tasks consist of steps that have to be carried out in the correct order. The job of working memory is to store information relevant to the task at hand. A memory cache that is constantly getting emptied and updated is a good metaphor.

There are at least two processes actually going on in WM. Information has to be stored only as long as it remains relevant: when task demands change, holding on to old information becomes distracting and leads to errors. On the other hand, WM has to be able to suppress new information from disrupting what's being stored. As a demonstration, think of how hard it is to remember a new phone number when someone next to you is saying a string of random numbers to distract you. Some computational models and other data suggest that the very limited size of WM may actually be optimal for effective updating.

There are lots of experiments in which subjects monitor the computer screen for numbers appearing one at a time. The task, for example, is to push a button whenever the number you see is the same one you saw last — "2 7 6 4 4.". But now I give you new instructions to monitor for numbers that appeared the time before last. Now you have to think back to two numbers ago, and also suppress the urge to respond according to the old instructions you had. The string "2 7 6 4 4" no longer matches the instructions, but "1 6 3 4 3" does.

Attention is a complex notion that's closely related to memory. The most-used metaphor is due to the work of 19th-century philosopher William James, who compared it to a roving spotlight that makes a part of our perceptions more salient (in some sense that is surprisingly difficult to understand). Attention is entangled with task monitoring ("What am I doing right now?"), error monitoring ("Am I successfully doing what I'm doing right now?"), and memory ("What aspects of what I perceive should I hold on to?"). It is very challenging to disentangle the notions of attention and WM. They always work in tandem. People usually mention them in the same breath, and they're under the same heading at conferences and such. If you study one, you study the other.

So, what are the span tasks used to study working memory? Well, they're all basically tasks that maximally tax WM. What we get out of them is your performance rate at any given difficulty level, and the types of errors you make. As an analogy, consider playing Pac-Man (which is very similar in spirit to many experiments in cognitive psychology): one could record the maximum level you achieve, and also analyze your playing strategies to see if you're very prone to backing yourself into a corner, or eating all four power pellets too soon, or whatever.

There's something called "running span." In the generic version of this task, you're presented with longish strings of letters of numbers. After you see a string, you're asked to recall the last n items from it in the correct order. The question is, since the whole thing is usually too long to remember, where do participants tend to focus their efforts? Does performance suffer across the board when we ask you to recall more items, or do we get good recall for items near the ends, and low recall for items in the middle?

A task called "block span" sees frequent use. In generic form, it's a "Simon Says" type of thing where blocks in a grid will light up in sequence, and the participant is asked to repeat the order of blocks. Experimenters can observe a variety of effects on performance by changing the size of the grid and the length of the sequence, etc.

There's also the n-back task, which is the number task I described above. It gets very difficult very fast for n > 2. Here it is previously on Metafilter, disguised as a game (primarily to make it more palatable for child participants). That particular team of researchers is interested in whether training WM (e.g., using their n-back game) could lead to improvements in performance on IQ and general intelligence tests (the current evidence is mixed, but promising).

There are also tasks like "instruction span," in which participants must remember and carry out increasingly longer and more complicated sequences of arbitrary instructions; "operation span," in which participants maintain items in memory while carrying out longer and longer sequences of simple arithmetic steps; and so on.

Lots of other span tasks exist: you can get a good background by reading the Wikipedia article. What they have in common is that they all attempt to exhaust your mental resources in some way or another, with the end goal of discovering your cognitive limitations. A few of these tasks are normed, so it's possible to compare your performance to others of similar age and gender. Most of these tasks were developed with research in mind and are not seen in clinical or therapeutic use, despite being very simple.

What fascinates researchers about memory, and working memory in particular, is how fundamental it is to our ability to function. It is so difficult to define precisely because it's involved in everything we do in the real world. And as a result, many researchers (such as Suzanne Jäggi and her collaborators from the n-back link above) are looking into the question of whether WM can be trained, and if so, what are the effects of improving WM on the rest of cognition? Would we then get better at problem-solving? Seeing patterns? Making conceptual connections? Learning?
posted by Nomyte at 1:42 PM on September 17, 2011 [9 favorites]


The book Flow referred to some studies that measured the amount of data a person could pay attention to in actual bits, and quantified various activities into how many bits they are; this was used to show that it's practically impossible to focus on more than two conversations simultaneously, that sort of thing. I don't have it right here so I can't reference it more specifically, and I can't speak to how they figured out the bits or not, but if you're interested in attention span that's the book to read.
posted by Nattie at 3:14 PM on September 17, 2011


To add to Nomyte's excellent summary, the "antisaccade task" is often used to measure the control of attention. In a typical version of this task, you stare at the middle of the screen until a stimulus appears on one side of the screen. Your job is to move your eyes away from the stimulus. Your automatic response is to look at the stimulus (this is called a "saccade"). To the extent you are able to pay attention to the task properly, you will be able to avoid saccading and you will "anti-saccade". More information can be had via a google search for "anti-saccade task".
posted by Philosopher Dirtbike at 1:17 AM on September 18, 2011


« Older How do I find my lost iPhone using WiFi?   |   for an inverse MST3K Newer »
This thread is closed to new comments.