UI Design Newsletter – July, 2009

Are you like them?

One of HFI's favorite buttons passed out at courses and trade shows is "Know thy users for they are not you."

In the context of typical user interface design, this makes sense. There are many variables that make other people different than the designer: computer experience, domain experience, vocabulary – the usual run-down of "cognitive differences". Therefore we conduct usability tests to insure we design for "them" and not "me".

Nowadays we're moving to the next phase of design competency. We know we can design to insure people "can do" the task. Our next challenge is to influence people to "want to do" a given action. This is the science of persuasive design. But are we aware of some pitfalls in designing persuasive messages? Are we testing the right things? Do we know the tricky points?

Men are man-like, women are woman-like

Well, it turns out that gender can alter your findings.

Two marketing professors, Wheeler (from Stanford) and Berger (from Wharton), wondered how much influence gender had when people made choices when answering preference questions presented on the Web.

Priming the primeval prime

As humans go, we're pretty much suckers for being influenced by what we see and do. This is the priming effect in short. And therein lies the story.

Here's how the priming effect works. In one early example, subjects had to unscramble sentences. (Sounds like an editorial job I had once.) Some subjects got words associated with a stereotype of being elderly, like "wrinkled". Other subjects got more neutral words.

Guess what? After concluding the study, subjects who unscrambled the elderly sentences walked more slowly to the elevator than the subjects with the neutral (control) sentences! They became more elderly in their gait.

This sounds like unconscious influences. And that's exactly how psychologists treat it: an unconscious influence. They call it "priming". I call it priming the primal unconscious.

In the current research, Wheeler and Berger wondered if men and women may end up responding differently to the same prime.

This means, of course, that other researchers would have to start being more careful when interpreting findings. But it also means that you, the persuasion designer, might have to check whether your persuasive messages have the same or different effects on your various audiences.

For example, prior research shows that when shopping for clothes women tend to select "possibility-driven" choices (or, "see what's out there"), and men tend to select "purpose-driven" (or "get a necessary item"). (Notice the word "tend", please.)

So now we have predictable outcomes from a known event. Shopping causes men to behave one way, and women another way.

What if we used shopping as a "prime" or an unconscious influence? Would men and women carry out a second, unrelated task, in the same manner as shopping, when primed to think about shopping first?

I call this the "primal prime". (Shopping is a primal event.)  

Doing the test

Here's the "unrelated task". The authors compared men vs. women on how they made decisions on a web questionnaire. Participants chose between two options across eleven difference scenarios. Here's one example.

Pretend you're a participant in a usability test.

"You are driving across the country to help out your friend. How would you prefer to do it?"

A. Do it in the fewest days possible to save time.
B. Take a longer time but make an event out of it, stopping along the way.

Pick a choice. No fair putting it off.

Will your answer be influenced?

Yes, your answer probably has already been influenced. This is because you just read about "shopping" and we already know that shopping creates a shopping attitude.

Therefore, some men among you probably became more purpose-driven and chose option A. And some women among you probably became more possibility-driven and chose B.

Whereas, if I had not mentioned anything about shopping prior to you seeing that question, men and women would be equally attracted to both A and B.

This is what the researchers found, as follows.

Researching the prime mover

To address this issue, the researchers systematically primed their participants in one of two ways. Prior to answering the 11 questions, all the participants were randomly assigned to write about one of two different topics.

Half were asked "Imagine you are shopping for clothes. What would the experience be like?" (This question was already known to get different responses from men vs. women.) When answering the 11 questions, participants receiving this prime were influenced along gender lines we discussed.

The other half were asked about the geography of their home state. Participants primed by writing about geography showed no gender differences in their choice of responses in the 11 questions.

What this means for you

The authors conclude that "the same prime leads to different effects". That is, participants were primed with shopping. And men and women answered unrelated questions using their shopping frame of mind.

Do your website users have a "shopping frame of mind" when visiting your site?

Maybe they just came from another site, like Amazon, and carry over the shopping "prime" they received. In this research above, women tended to choose possibility-driven choices, and men tended to choose purpose-driven choices.

Can the same thing happen to your site visitors? Probably so.

Test for both user experience and usability

So the question remains, how do you support both of these interaction styles?

Our typical "usability" context tends to be "purpose-driven". We conduct usability tests by asking participants to complete a task. What could be more purpose-driven than that? But – typically we expect all our participants to have that same attitude.

However, this research tells us that an alternative attitude could be driving your site visitor. Perhaps they are "possibility-driven". Can your site support this alternative?

How do you test your success in support of "browsing" or "exploring" or other ways of "window-shopping"?

Perhaps it's time to include alternative test questions, such as...

1. "Show me how you would explore this site?"
2. "Share your thoughts on what interests you or perhaps bores you about this site?"
3. "Compare your experience here with window shopping or browsing a store for possibilities."

You get the idea. The same prime, like shopping, can have different effects on different people. As we've seen, the contrast of men and women is just one example of different approaches to the shopping prime.

Is it time to go beyond just "purpose-driven" usability?

Can we start including persuasive design with "possibility-driven" user-experience?

References

Wheeler, S. Christian and Berger, John. When the same prime leads to different effects. Journal of Consumer Research. 34(3), 357-368, 2007.

If you are designing for older adults, it is not just about making the interface simpler and less dependent on technical knowledge. Since they explore in a "rigid" way, I would think it is most important to provide strong feedback for errors. So the error management needs to teach the older adult by indicating clearly that an error or failure has occurred, and then steer the older adult towards a better solution.