Counter-intuitive research reported in the New York Times raises some questions about how we design self-help materials.

A study to be published this year in the journal Psychological Science, led by Dr. Kornell, shows how strong this effect [The brain automatically associates perceptual fluency, or ease of storage, with retrieval fluency, ease of recall] can be. Participants studied a list of words printed in fonts of varying sizes and judged how likely they would be to remember them on a later test. Sure enough, they were most confident that they’d remember the words in large print, rating font size (ease of processing) as more likely to sustain memory even than repeated practice. They got it exactly backward. On real tests, font size made no difference and practice paid off, the study found [bold added].

And there is similar error in perceptions of the implications of font familiarity.

. . .[T]he researchers conducted a large experiment involving 222 students at a public school in Chesterland, Ohio. One group had all its supplementary study materials, in English, history and science courses, reset in an unusual font, like Monotype Corsiva. The others studied as before. After the lessons were completed, the researchers evaluated the classes’ relevant tests and found that those students who’d been squinting at the stranger typefaces did significantly better than the others in all the classes — particularly in physics.

“The reason that the unusual fonts are effective is that it causes us to think more deeply about the material,” a co-author of the study, Daniel M. Oppenheimer, a psychologist at Princeton, wrote in an e-mail. “But we are capable of thinking deeply without being subjected to unusual fonts. Think of it this way, you can’t skim material in a hard to read font, so putting text in a hard-to-read font will force you to read more carefully.”

I do not think this means that we should redesign all our websites to be as hard to understand as possible.  At a minimum, small and unfamiliar fonts will scare readers off, and even if they try to read, understand, and remember, they may start with more anxiety.

At a minimum, we should not be swayed by this research until we see it replicated with our target constituencies.

But the idea that you want people to work when engaging material is a very valuable one, and we should consider how to build sites that facilitate user investment in the process.  The trick, of course, is not scarring people away.  I think the way is to design interactive systems in which there is work required, but also immediate feedback.  Self-tests do that.  Asking people to repeat something aloud might do that (and be very helpful for court preparation.)

Ideas, please share.

P.S. Kate Bladow, in the comment below, links to an important study: “Jakob Nielsen has shown that low-literacy users don’t scan,” but rather read slowly and carefully.