Almost 10 years ago I fell into likely a vain pursuit of any shred of fact to support the contention that “humans process visual information 60,000 times faster than text.” And despite trying every approach I could conjure, including offering a cash prize, the trail ran cold back to 1982 (all the blogged efforts are corralled by the 60000x tag).
I tend to forget about it.
When a subject line appears in the inbox, maybe once a year, I first speculate that someone found it. I might have to open the checkbox and shell out the promised $60. Once came last week, but it was another researcher in pursuit, who wrote:
Pardon the inbox invasion, but I came across your seasoned blog post about how we supposedly process images 60000X faster. I’ve been trying to find the 3M study too for this book I’m writing. I found the same pdfs and no reference/citation to the original research. (no surprise at all)
I didn’t want anything, only a thank you for doing this sleuthing back in 2012, documenting your investigation and keeping the blog post live in 2022. Wish more people in tech were like you in this respect. Your blog post will be one of my references for this chapter on data visualization.
I assured her that this was hardly an inbox invasion.
But it will likely never die because it just rolls off the screen so easily. It’s has the essence of something…. truthy.
One outcome of this is I keep my antennae open to other such utterances. One came my way a week or so ago, I think it was a link from Stephen Downes OLDaily to a Poynter article (are these things articles? posts? who cares?) about communicating in video– To tell stronger stories, don’t narrate the video. Explain it.
It has an interesting premise- that our tendency to want to narrate a video scene (as its easy to do) might actually compete with the ability for a viewer to understand it. Author Al Tompkins contends that the narration is more effective when it explains the scene. I might quibble with the notion that there is a “truth” in the explanation, but I think it’s more about giving the scene context beyond what is depicted visually.
And hey, we know that humans can process those visuals 60,000 time faster than— hold it.
The part that I am jumping on now is where one of “those” truthy statements are made under the heading “Hang on, a little bit of visual theory”
Ann Marie Seward Barry, associate professor of communication at Boston College, cited a study by the Educational Foundation of the American Association of Advertising Agencies that shows that even when we see images on television, we usually don’t understand them.
Barry writes, “Even when we watch television, we misunderstand approximately 30% of what is shown to us. Our emotional state, our mindset at the time and our experience all seem to conspire against our seeing things as they really are. We go about our lives, however, mostly assuming that what we see really ‘is,’ as if there were no intermediary process.”
https://www.poynter.org/reporting-editing/2021/to-tell-stronger-stories-dont-narrate-the-video-explain-it/
I bristle when a “study” is referenced without any references. But here it is asserted as fact as some kind of universal law that we misunderstand 30% of what we watch on television. Right away I was questioning, for what do we even mean by “television”? The network based tightly programmed flavor of TV I watched in the 1970s and 1980s hardly is the same medium as what we absorb now as video on screens.
It took a few rounds of searching in Google Scholar to locate the source.
Why do web writers not apply the most basic feature of web writing– the humble link? The journal version I found at Sage Viewer Miscomprehension of Televised Communication: Selected Findings (Jacoby and Hpyer, 1982) but of course, PAYWALL. Search, and search again, I found the same title at ResearchGate where I could at least read what looks like as scanned print pages (the date is listed as January 2000, but maybe it’s a reprint as the article has 1982?? It’s not grey as it states it was uploaded by one of the co-authors (Jacoby) in 2014.
Regardless, here I can read the full study. You too.
If I understand correctly, in the early 1980s, they got 200+ participants to sit in a room at a shopping mall to watch on a VHS tape played to a TV two thirty second advertisements or excerpts from news shows. They then answered a 6 item quiz on factual recall.
All of the methodology is explained and not terribly unreasonable, but one is left to wonder how that same 30% result from 1981 is extrapolated to something like a tweeted news clip seen on a smart phone in 2022? One might say the format makes no difference. I am not a media excerpt enough to declare the level of validity, but I would not want to rest my case on this. Would you?
The point makes sense and we ought to question how much people walk away from video content with the same full understanding we put into it when we produce it. And the inferences from Gestalt theory offer some useful consideration points:
Probably the most important Gestalt concept is the theory that in any visual display there is always one element that will be perceived as the object. Everything else is perceived as background. In television, we clutter our screens with information we think will appeal to viewers. In fact, the brain is not able to process so much information.
https://www.poynter.org/reporting-editing/2021/to-tell-stronger-stories-dont-narrate-the-video-explain-it/
Why hello chryon.
The article does give me some pause to think about the relationship of visuals in a video and what is said along with them. But I can’t feel good about this bold proclamation of a 30% miscomprehension rate across the board. And I certainly suspect that the percentage of comprehension across the human board of video content is likely low.
This is quite different from the 60000 times faster claim, as that one lacked any kind of real research to support it. For the 30% misunderstood claim, it’s more of a matter if the research study behind it is even applicable.
Again, I am not on the firmest of grounds with my questions, but I do enjoy the digging into and finding the threads back to original sources, which seems to be just something writers or snazzy presentation makers don’t have time to do.
What say you? Will you stand confidently on that 30% plank?
Featured Image: CC0 image of retro TV on street from Pixnio modified by Alan Levine with superimposing of 30% image from Pixabay by TheDigitalArtist… your guess is as good as mine as to what license that makes, but CC0 seems close enough.
Fascinating! Thanks.
It suggests more possible experiments are needed. Just as you pointed out. So caveat is 30% under the conditions in the original study, and anyone who tries to project, parlay that into another context is making a big leap.
I’m no scientist, but, as for me, I’m always curious about methodology and can this experiment as published be replicated. 30 second adverts, TV screen, random sample in a mall, I think it’s technically feasible to get those inputs together and run another set of people through the study. Compare results. If you can statistically derive the 30 percent on the replicated attempt, then you have at minimum met the criteria being reproducible. And hence, possibly an “effect”. From there don’t know what ELSE you could surmise, other than see if funneling into 1.)online 2.)mobile formats also has effects.