Does learning occur differently with physical or digital print?
July 10, 2012 at 4:20 am 4 comments
I’m skeptical about this claim: That your brain interprets text in books differently than text in digital form. One argument in support of the claim is an observation (not much data) that we have to re-read digital information more often than print information before we remember it, but doesn’t offer a theory for why that should be true. I find this second claim a bit more plausible: That our memories rely on contextual information, and physical books provide us more cues to support recalling what we read. I wonder, though, if we might not be able to provide more contextual cues through the interface. I’ve started reading the “Our Choice” app on my iPad, and there are lots of cues in that book to provide a sense of “place” (what page you’re on, what pages are around you, what chapter you’re in).
But without stronger evidence that there is a difference, I’m going to keep reading on my iPad and Kindle (well, once I get a new Kindle — my Kindle’s screen died somewhere during my trip to Venice this last weekend).
In other words, the human brain uses location to recall the words it reads, which helps reinforce the information. To trigger a memory, the brain might recall whether it read the information at the top, middle, or bottom of the page, remember a corresponding picture on the page, or even a page number — essentially creating a mental bookmark to cue recall of the information.
“Anyone who has read an e-book can attest that the page provides fewer spatial landmarks than print,” Changizi continues. “In a sense, the page is scrolled without incident, infinite and limitless, which can be dizzying. On the other hand, printed books give physical reference points, which can be particularly helpful in recalling how far along in the book we are, something that’s more challenging to assess on an e-book.”
Entry filed under: Uncategorized. Tags: ebooks, HCI.
1.
alanone1 | July 10, 2012 at 5:52 am
There needs to be a contemporary study of much more depth than this.
At PARC when we first started to synthesize recognizable images on a bit map screen — using the best technology we could put together at that time — we quickly noticed that screens were not as readable for text as paper, and the better a text reader — and rememberer of textual information — one was, the more the disparity was noticeable.
This motivated a fair amount of experiments — and perusals into the literature of how the human eye looks at various kinds of images — over a number of years to try to understand how the variety of parameters in a textual image trade off for skilled readers.
A few years after the initial experiments we got one of just three existing super high resolution SRI eye-trackers, and this — plus a lot of previous work by Derek Fender and others — supplied a few more pieces of the puzzle.
For a really good reader (and these are rare), pretty much everything counts to some extent: resolution, contrast ratio, absolute adaptation, quality of typography and most especially frame rate (if not 0).
The latter affects most readers to the largest degree, with contrast ratio (trading off against adaptation) in second place. There are several neural and cognitive responses to image and brightness change. There is the phi phenomenon (3-4/second), there is flicker fusion (12-75/second), and there is peripheral warning syndrome (well over 100 per second).
Most of our retina except for the 2-3 degree wide fovea is legally blind, as far as acuity, but is extremely sensitive to a change of brightness out of the corner of the eye. This causes the eyeball to aim (do a saccade) at the source of the change (and to bring the head around if needed) to get the high acuity of the fovea to access more detail — which is then woven into the waking dream we like to call reality.
Part of learning to read fluently is learning to bring the saccading of the eyes into a highly controlled scanning, and to suppress the normal tendency to pop all over the place to build up a larger picture of what is going on. (Parents can help their children learn to read better by moving their finger along the words as they are read aloud.)
Tracking eyes that are reading off a screen which is flashing at all — even if the flashing is well beyond the normal consciously flicker fusion rates — will reveal that random saccades are being instigated to the sides of the screen. For good readers, this disturbs the continuity that helps them read at very high speed with very high comprehension and memory.
I haven’t been following any of this for about 20 years or more, and the refresh rate has gone up considerably since then. At some point it should get high enough to get past the “peripheral warning syndrome”. I have a feeling it isn’t there yet.
Subjectively, the e-book readers (with 0 frame rate) seem more readable than the iPad with a retinal display. Neither seem to be as readable as paper (the e-books still have some contrast ratio and absolute accommodation problems).
(I should perhaps mention that I’m one of those highly fluent remembering readers that are bothered by various aspects of displays even today.)
But it would be nice to see a comprehensive study that would get us up to date from the 70s ….
Cheers,
Alan
2.
Leigh Ann | July 10, 2012 at 8:46 am
Like Alan, I am also one of the highly fluent remembering readers and I too am bothered reading on a screen. I still print out papers I want to read in depth and prefer to read them on the paper than on the computer screen.
While I agree that there may be interference from the type of display that is presenting the information, I also agree with Mark about the impact on cognition that the different reading media present.
When sitting down to read a book or paper in addition to the contextual clues that create opportunities for retrieval, there is also at the same time less “noise” in the environment. Ads on the side of the page, the clock on the bottom of the screen, email messages that pop up into your window, all of these distract your brain from forming the pathways most likely to be activated when attempting to retrieve the content. Especially if the content is new and for a domain in which we are a novice.
The novice is storing that information with very little sophistication in terms of conceptual connections and therefore most likely will have a low retrieval probability (as there are not many nearby nodes to be activated to prompt retrieval later). With an increase in distractors and less focus the connections are even weaker and therefore harder to activate in the future.
I’m wondering if there is a way to really sus this out with some lexical decision tasks with priming either from a printed book, from the same text but on a screen with no distractions, and the same text on a screen with what you might expect of a web page displaying something to be read.
3.
Dennis J Frailey | July 11, 2012 at 7:06 am
I too much prefer paper to a screen. It may be related to how I was brought up but I find both the physical layout of a book and the comfort of a printed page to be much more conducive to learning and retaining what I’ve read. But this is just anecdotal evidence from one data point!
4.
Cecily | July 13, 2012 at 11:47 am
A lot of printed text, especially technical or educational material, also has images and diagrams to anchor the reading experience and help people remember what they read. There still isn’t a great way to easily do that consistently in digital, and I think that makes e-texts harder to read. For example, my church magazines have digital format but usually it is just plain text with minimal HTML markup, while the print version has numerous images, and certain parts of the magazine are always in certain places in the print version so we can find them easily.