Welcome to The Spark, bringing you the world's best behavioral science. 

We’re building Applied Scholar to be an invaluable resource for high-end practitioners working in and around organizations on Talent and People topics. As part of that, we strive to make behavioral science easier to access, easier to understand, and easier to apply.

In this Spark, we’ll go all out on the easier to understand and apply parts. We'll start with a very direct and actionable summary of the research on Fairness and Bias Reduction, then touch on Learning, AI, and Team Performance.  Let’s dig in.

SPOTLIGHT

Fairness and Bias Reduction

Awareness-based unconscious bias training doesn’t reduce bias. It never really has and it’s unlikely that it ever will. Don’t run unconscious bias workshops. Don’t pay for vendors to run these workshops. And don’t make digital learning courses on unconscious bias available to employees. We’ve spent thousands of hours on this topic and all that is quite settled. Email us - just hit reply - if you have any questions on this part of the story.

So that describes what NOT to do. Now let’s talk about what actually works based on our massive research effort, with an eye towards easier to understand and easier to apply. There are four ways to reduce bias.

Modules - Eliminate bias at its cognitive source, aka the “module” level. Simply put, we’re attempting to clear the brain of select biases. This is really hard, takes a lot of expertise, and can do harm if not designed well. For the vast majority of organizations, it’s best to steer clear of this one.

Filters - Add cognitive or environmental “filters” that catch biased beliefs before they turn into biased behavior. The biased beliefs still exist, but they never materialize into actual biased behavior. There are several evidence-based approaches here that can be implemented at scale in most organizations.

Routines - Interrupt automatic and biased thought patterns by activating a less biased process. These can be particularly powerful in moments where bias is likely to occur, as it guides individuals through a more deliberate thought process. Quite scalable too.

Context - Change the context to make it less likely for bias to ever materialize. With this approach, thought patterns or behavior that might be biased are never activated in the first place, by changing the context or environment. 

Here are five examples of how organizations and businesses can use these approaches to solve common challenges related to fairness and bias:

• Embed short reminders right before key talent decisions that focus decision-makers on relevant criteria and give clear guidance (Routines)
• Leverage technology to scan key talent documents for problematic language related to fairness, discrimination, and bias - even better, do this in real-time so it’s immediately corrected (Filters)
• In hiring or evaluations, get super clear on the competencies for the role and make these competencies visible as much as possible throughout the process (mixture of Routines and Context)
• If possible, scrub irrelevant information (age, gender, race, name, etc.) from the hiring process until initial judgements based on competencies can be made (Context)
• If you’re ready for a challenge and to put your high-end skills to the test, explore how you can leverage growth mindset as a way to reduce the rigidity of stereotypes, or how you can leverage counter-stereotypical examples to balance problematic stereotypes - these two are more sophisticated approaches, so proceed with appropriate nuance and caution (Filters)

If you’d like to learn more, we’ve created a full set of videos and resources on the topic of Fairness and Bias Reduction that’s available via the Vault (still on free trial), accessible here.

Taking a step back from this singular topic, we hope this sort of summary also gives you a sense of how incredibly valuable behavioral science research can be when it’s easier to understand and apply. 

 FEATURED RESEARCH 

Let's dig into the research.

Learning in the Age of Algorithmic Video: Continuing with our theme of applying the research, we recently came across an excellent evidence-based summary by Carl Hendrick, an expert in learning and instruction. If your role involves learning or digital content - or you’re simply curious about how YouTube and TikTok are influencing human development - we would highly recommend checking out Carl’s summary. Even if you’re not interested in the particular topic, it might be worth a quick scan to get a sense of what a really good research summary looks like. Carl definitely makes the research easier to access (40+ research links embedded naturally in his writing), easier to understand (he provides short, concise summaries of different viewpoints), and easier to apply (he sprinkles in recommendations throughout and closes with five major points). Well done Carl!

Strange Impacts of Sycophantic AI: James Elfer - the founder of MoreThanNow, a leading organization in applied behavioral science - recently shared a post with links to four research papers highlighting some of the emerging issues with sycophantic AI. For any of us who are influencing how AI is rolled out and adopted within a team or organization, this is worth checking out. The design choices we collectively make on the integration of AI into organizations and businesses are sure to have some widespread - and not always straightforward or positive - outcomes on human behavior. We hope this can be immediately applicable for many of you as the rollout of AI tools across organizations and businesses accelerates.

 PERSPECTIVES TO SHARE

High-quality commentary from the world of applied behavioral science. 

Diversity and Team Performance: There’s a lot of nuance and understandable sensitivity around the topic of diversity and team performance. We recently came across a Podcast episode that did a superb job navigating this topic. The discussion is nuanced, applicable, and recognizes the challenge of following evidence-based practices within the messy real-world conditions inside orgs and businesses. Check out the Podcast here on YouTube, the link will take you right to the start of the relevant discussion.

 LAST BUT NOT LEAST

We want to hear from you.

In our continued nerd quest to read and explore as much research as we can and then serve that up beautifully to you, we’re always interested in hearing from you. If you ever have something to share - perspectives, research, what you find valuable or not valuable - just hit reply on the email and it’ll reach us.

On the “easier to access” front, we’ve also now added a searchable newsletter archive to the Vault. Our vast history of newsletters are now loaded there (all three of them! more to come of course!) and easily accessible for Vault subscribers.

We hope that you enjoyed this edition of the Spark. See you again soon!