As readers of this blog and colleagues familiar with my approach to rhetoric know, I argue consistently that rhetoricians should be as familiar with statistics and behavioral research methods as behavioral economists, data scientists, and human-computer interaction researchers.
Too often, rhetoric as a discipline doesn’t embrace findings that should be shaping our scholarship.
When I work on print or digital media design, I adopt the same methods as most user interface / user experience experts. The UI/UX world embraces A/B testing a bit too enthusiastically, while rhetoricians ignore or dismiss data-driven A/B testing.
What is A/B testing? In its simplest form, an A/B test involves presenting two randomly selected audiences with two different designs. For example, you might present a single page in two different typeface pallets and then measure how long readers engage with the page. You can also test how well a smaller sample from each test group remembers the text. Other common tests include asking how much the text was trusted as authoritative.
One of the most common A/B tests involves using two sets of button or link designs on a website to determine how the colors, sizes, and placements of buttons influence readers. Changing the text on a button might result in significant user behavior changes.
Design choices that lead people to make choices they’d otherwise avoid are known as “dark patterns.” The name is attributed to UI/UX researcher Harry Bignull, though designers have long engaged in the practice is misleading audiences through design.
Rhetorical scholars used to believe in the power of reason and logic. Today, most of my colleagues understand that people make impulsive decisions and then try to rationalize those choices. I realize it is uncomfortable to admit that all people, including educated researchers, are not models of rational consumers. Economists admitted there was no rational consumer, and we need to do the same.
The dark masters of A/B testing seek to implement the most “effective” UI/UX designs based on what benefits their employers. Government agencies, non-profits, and corporations use dark patterns to “nudge” user choices.
Students majoring in rhetoric or technical communication should be required to take UI/UX courses that introduce A/B testing methodologies.
As teachers, we should discuss dark patterns and A/B testing in our courses. We should strive to give students an awareness of subtly manipulative design. The dark patterns are rarely as obvious as Comic Sans vs. Times Roman or a large blinking red button vs. a pleasing blue text link.
Data generated by thousands of A/B test design impressions guide today’s web designers. These designers care less about aesthetics than past generations. A visually incoherent design might produce the desired A/B choice, leading to websites that seem shockingly bad.
I recently wanted to cancel a subscription service. The link for cancelling the service was almost too small to read, in a light color on a shaded background. By comparison, the “Renew Now!” and “I Don’t Miss Out!” buttons were impossible to ignore. It’s likely many people give up trying to cancel their subscriptions when confronted with such a visual mess.
I realize there are arguments for choices we want people to make. Maybe you would choose to defend making some choices more likely than others. I understand that making beneficial choices the passive default can benefit society.
Rhetoric traditionally overlaps with philosophy. Designers seem to be in need of moral philosophy. Instead of asking, “Can I lead people to make this choice?” the first question should be, “Should I be leading people to this choice?”