How Diversity Reduces Digital Identity Bias

TL;DR

Diverse product teams help expose assumptions that can exclude people from digital identity systems. Designers should account for unintended users and differences in age, disability, income, handedness, device quality, and infrastructure, because unconscious biases and narrow ideas about the user base can create serious barriers to authentication and participation.

Transcript

Yeah. It's already on the road again. All right. Here we go. Dun, dun, dun. This, this morning is brought to you by Peet's, maker of fine caffeinated products. All right. Hi. I'm Janelle Allen, and I'm on the board of directors for Identity Pro, the industry association of identity professionals. Uh, we, we train and mentor and bring forward identi... Read More

Key Insights

• Unconscious bias is difficult to perceive from within because the brain treats familiar assumptions as ordinary. Like an optical illusion that becomes visible after a hint, a hidden mental pattern may become obvious once another person reveals it, and then become difficult to ignore.
• The framing effect is a shared bias in which positively phrased information is more appealing than equivalent negatively phrased information. The presentation contrasts a condom described as 95 percent effective with one described as 5 percent ineffective to show how wording influences preference.
• The peak-end rule is a memory bias that emphasizes the emotion surrounding an event and how the experience concludes. A difficult experience that ends happily may be remembered positively, while a happy experience with a bad ending may be framed negatively.
• Counterfactual reasoning is the process behind thoughts such as what someone should, could, or would have done differently. Regret can reshape later decisions, and experiments described in the presentation indicate that rhesus monkeys also anticipate alternatives and modify their subsequent behavior.
• Bias is a useful feature of cognition because rules of thumb make rapid decisions and processing possible. Problems arise when those shortcuts reinforce existing beliefs, encourage selective interpretation of evidence, or contribute to echo chambers supported by social networks.
• Error management bias is a protective tendency to overestimate danger from threatening objects. In a test involving three objects approaching at the same speed, people judged the spider as moving faster, illustrating how fear can alter perception in ways connected to fight-or-flight behavior.
• Digital identity users are more diverse than the people who design their experiences. Systems created for average office workers may also be used by children, teenagers, retirees, disabled people, caregivers, and other unintended users whose needs were absent from the original product assumptions.
• Economic and infrastructure constraints directly affect access to digital products. Two-thirds of the world lives on less than ten dollars a day, some communities receive electricity only two to three hours daily, and older smartphones may remain in use after passing through multiple owners.

Questions & Answers

Q: Why does the digital identity industry need diverse teams?

The digital identity industry needs diverse teams because designers may not recognize assumptions rooted in their own experiences. People with differing backgrounds and viewpoints can reveal blind spots involving age, disability, income, location, language, orientation, devices, or infrastructure. Addressing those blind spots can prevent exclusion and help identity products give forgotten individuals greater access, voice, and power.

Q: How does unconscious bias affect digital identity design?

Unconscious bias affects design by encouraging creators to imagine users who resemble themselves or fit an assumed average. That narrow picture can shape authentication methods, device requirements, interfaces, and support processes. Because people often cannot see their own biases, products may unintentionally exclude users whose ages, abilities, resources, or living conditions were never considered during development.

Q: What is the framing effect described in the presentation?

The framing effect is a bias that makes people respond differently to equivalent information depending on how it is phrased. The presentation illustrates it by comparing a condom described as 95 percent effective with one described as 5 percent ineffective. People generally prefer the positive framing, even though the two descriptions communicate the same underlying effectiveness.

Q: What is the peak-end rule and how does it shape memory?

The peak-end rule is a bias in which memory is shaped strongly by the emotion of an experience and by its ending. A long, slightly painful event may be remembered positively when it ends happily. Conversely, an otherwise happy event may be remembered negatively when its outcome is bad. The presentation says rhesus monkeys also demonstrate this pattern.

Q: What are unintended users of identity products?

Unintended users are people who use a product even though its creators did not anticipate them during design. A system intended for typical office workers or adults might be used by a three-year-old on an iPad, a teenager entering a consumer service, a retiree, or someone using an older device. Their presence challenges narrow assumptions about users.

Q: How should age be considered in authentication design?

Authentication design should recognize that users can include young children, teenagers, workers, retirees, and people over 65. Children may already authenticate to devices, while a terms-of-use statement asking users to certify that they are 18 does not necessarily prevent teenagers from entering a product. Designers therefore need to evaluate real behavior across age groups.

Q: Why must digital identity products account for disability?

Digital identity products must account for disability because ability can change over a person's lifetime through severe medical incidents, loss of sight, loss of hearing, or becoming bedridden. Product creators may also support elderly relatives or work alongside affected colleagues. Treating disability as remote from ordinary users ignores circumstances that can touch employees, consumers, and caregivers directly.

Q: How do income and infrastructure affect identity access?

Income and infrastructure determine which devices people can obtain, charge, and continue using. The presentation states that two-thirds of the world lives on less than ten dollars a day and that some places have electricity for only two to three hours daily. Expensive smartphone requirements can exclude many users, while recycled devices may continue serving later owners outside the original target market.

Summary & Key Takeaways

• Unconscious biases are mental shortcuts that support quick decisions, but people often cannot recognize their own assumptions without a revealing hint. Humans and non-human primates share patterns such as the framing effect, peak-end rule, and counterfactual reasoning, suggesting that bias is a normal feature of cognition rather than a simple defect.

• Bias becomes dangerous when product creators assume users resemble themselves. Digital identity systems may serve entire organizational or consumer populations, including people whose ages, abilities, locations, orientations, languages, or living conditions were never considered. Teams with differing perspectives can identify blind spots before those assumptions become barriers embedded in products.

• Inclusive identity design must consider children, teenagers, retirees, disabled people, caregivers, low-income users, and people with limited electricity or older devices. These groups may be intended or unintended users. Building with their circumstances in mind can expand access, avoid preventable failures, and give previously forgotten individuals greater voice and power.