2 min read Usability testing

Unpack misunderstanding

Unpack misunderstanding

If the user misunderstood something, that's a sure sign of a usability issue. But it takes some vigilance to notice it.

Examples:

  • she read the label of a menu item and said it meant something else
  • she explained her options in a radio group but she was wrong about some items
  • she tells what she expects to happen upon clicking a link, but that's not what would happen
  • she explains how the feature works and that's just not true

Once I notice an incorrect interpretation I want to unpack that and understand why she misunderstood it. To do that I must hear her thinking process. What leads her to that false conclusion?

Think-aloud does a marvelous job at uncovering the steps that mislead her. But if think-aloud doesn't help I can still probe to expose more of her thinking and mental models.

The two techniques that come in handy are mirroring and asking for the reason directly.

DO mirror her words or actions

  • Participant: "I don't want to hit 'Publish' just yet, because I want to schedule my post to a later date. 'Publish' would send it out immediately."
  • Researcher: [Issue: she didn't realize that scheduling will come after she hits 'Publish'. Let me mirror, so that she elaborates on her mental model] "So 'Publish' would send the post out immediately..."
  • Participant: "Exactly. There is some software when scheduling is in the next step but then the button is not 'Publish' but 'Publish...' indicating that there are more steps. Or 'Schedule and publish'"
  • Researcher: [Clear. Her mental model is that upcoming steps would be either stated explicitly or indicated with '...']

DO ask for the reason directly

  • Participant: "I don't want to hit 'Publish' just yet, because I want to schedule my post to a later date. 'Publish' would send it out immediately."
  • Researcher: "What leads you to think that?"
  • Participant: "Well, there is some software when scheduling is in the next step but then the button is not 'Publish' but 'Publish...' indicating that there are more steps. Or 'Schedule and publish'"

Published by:

Zsombor Varnagy-Toth

You might also like...

May
22
Filter fakers with bait responses

Filter fakers with bait responses

If there is a monetary incentive for participation, there will be people who try to fake their way into the
1 min read
May
04
Populate prototype with realistic content - part 6, empty demo accounts

Populate prototype with realistic content - part 6, empty demo accounts

Realistic content is not only for prototypes - it's for live systems too. The best is if we do the
1 min read
Apr
27
Populate prototype with realistic content - part 5, repetitive data

Populate prototype with realistic content - part 5, repetitive data

Populating UIs with realistic data is not hard. It's boring. That's why many Designers would copy-paste their way out of
1 min read
Apr
21
Populate prototype with realistic content - part 4, inconsistent data

Populate prototype with realistic content - part 4, inconsistent data

On usability tests users will try to make sense of every little bit of the prototype. They will spot irregularities
1 min read
Apr
13
Populate prototype with realistic content - part 3, unrealistic data

Populate prototype with realistic content - part 3, unrealistic data

Unrealistic data can easily derail our participant's thinking, wasting precious minutes from a usability session. Unrealistic data Dashboards, charts, and
1 min read

Member discussion