Avoid Confirmation Bias and Level Up Your Stats

Sites like Facebook are set up so as to only display content they think we want to see, and because our feeds are generally curated by the friends we have, it’s not likely you’ll witness anything shocking or outside your comfort zone, opinion-wise. And if it were to happen, it’s human nature to dismiss anything we don’t agree with.

This is known as confirmation bias, or our inherent instinct to tune out what runs counter to our worldview and seek out information that validates what we already believe.

The phenomenon, identified by The Social Dilemma, has seemingly permeated every aspect of daily life, but we particularly notice it within our work. As seen within Rightpoint’s Analytics and Insights practice, when we create a data visualization strategy for clients, they have presuppositions of what they want to see or think, and they look for data points or KPIs that confirm those biases. Due to this narrow view of data, these clients are missing broader challenges to their assumptions, which leads to wrong decisions.

Think of this like driving a big rig. If you don’t check your blind spot when you change lanes, you could hit somebody and potentially injure yourself or do damage to your car. Similarly, if we aren’t considering all variables or bringing our inherent biases to the project, the ramifications could be costly for the company or client—not only monetarily, but in time and reputation, as well.

With the proper planning, though, you can avoid confirmation bias and deliver experiences that target a whole range of clients and present surprises along the way:

Catching Issues

Proactivity is necessary to head issues off at the pass, specifically keeping a level head as to how you approach data:

• Lay out all data points, especially those that challenge your own conclusions and beliefs—you will likely fall into the trap of confirmation bias if you don't push yourself

• If enough data points and KPIs are telling you the same thing, then you’re probably on the right track; but if you only check a few, they might be all over the place or not a representative sample

• Partner with your strategic analysts as an integrated team member; they can translate business challenges into actionable analysis supported by the right data and metrics—challenging the status quo at every step

• Avoid drawing absolute conclusions from univariate analysis—which uses only one variable to conduct analysis

Unbiased Partnerships

Our work revolves around this mindset. At Rightpoint, we develop strategic measurement strategies to help clients define metrics and KPIs that matter, with an eye towards avoiding confirmation biases. In line with client goals, we work collaboratively to articulate analytics aspirations into digestible analysis and data visualizations, working as a trusted partner to educate on the importance of mitigating erroneous data.

As an example, our partnership with a large asset management firm led us to support their strategic data and analytics initiatives, which helped combat confirmation bias.  Working with their global marketing team, we built a data strategy incorporating many different data streams into a self-service product, all aligned to a strategic measurement framework. We worked hand-in-hand with the data end users to build a single source of truth for the future. The ability to note myriad sources of data and alignment across an organization ensured we encountered minimal confirmation bias, which are often on display when we look at only one single source of data, or we look at data through too narrow a lens. This way, we ensured we were always making the best decisions for the stakeholders of this company.

Avoiding confirmation bias is possible—all it takes is an openness to accepting data points that may challenge your original hypothesis, plus a willingness to keep digging for the truth. Want to learn more about how you can produce results free from confirmation bias? Get in touch with Rightpoint today.