Thursday, December 08, 2011

Is Image Memorability Important to eLearning?

I discovered this interesting research report from a tweet that pointed to an article in Scientific mind titled Haunting Scenes.  Read at the end why I found more info at MIT News.
Brent Schlenker
Program Director for DevLearn

I wonder how a silly pic like this, on the right, would rank in a memorability algorithm?


Exciting Research
The interesting news is that the research found "the most memorable photos are those that contain people, followed by static indoor scenes and human-scale objects."  Landscapes seemed to be forgettable.  According to previous research humans can remember thousands of images, but understanding what makes certain images more memorable than others was not known.  This study focused on memorability of an image as a measurable trait.  And they were successful enough to create an algorithm that can predict memorability.  Cool!

Disappointing Ideas Regarding Application of Results
The most frustrating part about the article?
"Such an algorithm could be useful to graphic designers, photo editors, or anyone trying to decide which of their vacation photos to post on Facebook" - Aude Oliva, associate professor of cognitive science and a senior author of the paper.
Seriously, MIT?  That's the best you can come up with?  <insert face in hands>
I understand that Facebook is hot and you need to encourage interest in your research in order to get grants and bla, bla, bla, but I was hoping to hear you say something about LEARNING.  I'm thinking that memory is tied to learning some how, right?

Applying These Results to Learning Design
My apologies to the researchers and to MIT for my pedestrian response to academic research.  But I see your research as being important to learning, education, and training more than Facebook and vacation photos.  Shouldn't we be more excited about the possibility of selecting better images that help learners retain more important information in long term memory? Memory does have a little to do with learning.

My favorite quote from the lead researcher, Phillip Isola:
"Pleasantness and memorability are not the same"
I know, in my experience designing eLearning courses, that making graphics simply looked nice, and pleasant, to the corporate stakeholders was the goal if not a battle you had to fight.  Corporate training developers are often asked to PROVE why they chose the graphic design elements they did.  And many times the selection of images to use are decisions based on the whims of clients, and Sr. Management.  Research like this is extremely helpful to our community in making the case for the design decisions we make.

But more importantly it helps us do our jobs better.  Creating media for eLearning courses is part of the job.  Understanding what makes an image memorable is a BIG DEAL!  

And yes, there is a lot more to the design of instruction than the memorability of the images used in that instruction.  But it seems to me like there is a useful connection here.  That's all I'm really saying.  Perhaps I'm wrong, and I completely missed the point on this one.  Oh well, wouldn't be the first time.

Read the research report yourself and let me know what you think.

---------------------------------
COMMENT:  I don't pay to read Scientific American and so I only had access to 5 paragraphs of the article.  So, did those 5 paragraphs entice me to PAY because I became interested?  No.  A quick google search of the researcher from MIT revealed his findings in other more user-friendly online magazines and the actual complete research paper.  Sorry, Scientific American. I liked the article in MIT NEWS.

2 comments:

Kelly Meeker said...

I've always held that opinion on intuitively - that images that have people are more compelling and memorable - but it's certainly great to have that perspective validated by fancy science!

bschlenker said...

I think a lot of design is known at the instinctual level but until science comes along and proves it, the non-design folks don't believe it :)