Overcoming Confirmation Bias

When we need to make decisions it is temping to interpret, favour, and recall information that conforms to pre-existing beliefs or hypotheses. This is universally known as ‘confirmation bias’ and if not left in check leads to faulty decision making due to the overemphasis on factors that conform to pre-existing beliefs. Additionally, those who succumb to confirmation bias tend to disregard or minimize information that is not consistent with their pre-existing beliefs or agenda, and over-emphasize information that appears to validate it.

Overcoming bias is important, and even those who succumb to the fallacy of “priding themselves” in their “decision making” may have all kinds of different biases that are oblivious to them. It’s important to develop self awareness to keep this in check, and it’s only possible if we first recognize that having bias is an inherent attribute that can negatively impact the quality of our decisions.

To help overcome confirmation bias, it’s important to:

  • Seek out diverse opinions: Be aware of tendencies to seek confirmation. It’s better to focus on validating the strength of your decisions with competent people who think differently and have diverse sets of backgrounds. Avoid tendencies of solely seeking out validation from those within your own thought-bubble. Once you’ve decided to seek out diverse opinion, it’s important that you are also open to them.
  • Actively work to influence the culture to promote, open dialogue, disagreement, and critical thought: Having critical thinkers who understand how their biases can negatively impact their decisions is one thing, but influencing the entire culture is another thing all together which helps multiply its effectiveness systemically across the organization.
  • Be objective and self aware: Objective data and measurement provides good data points in which to base decisions off of, however it’s also easy to dismiss otherwise objective data when it doesn’t align with your pre-conceived ideas. Just as it’s important to be objective, it’s important to take a step back and assess and validate if you are actually being objective or further succumbing to confirmation bias under the guise of being objective.
  • Actively consider the other side of the argument: Invite people to contradict your position and don’t fear disagreement. Of course, people who disagree with you may also have confirmation bias, so it’s important that all discussions are rational and predicated on the idea that the purpose is to move froward with decisions in the most sound and responsible way.

As we move along the path of significantly reducing confirmation bias, it’s important to remain rational, objective, self-aware, and to not introduce new paralysis into the decision making process due to the heavier burden of reducing bias from the equation.

The key is to work smartly and to get to the place where removing bias is inherently part of the decision making method. Don’t disregard the need for time-boxing decisions. Identify the strength of decisions weighted by what we know and what we don’t know. The higher the significance of the decision, the higher the bar should be established in terms of validating and ensuring that bias has been minimized as much as possible.

People who are highly skilled decision makers know their strengths and they also know when their strengths could impede objective decision making. They are also highly self-aware which when combined with their other strengths allows them to effectively understand things objectively and make good decisions. One may be carefully skilled at debate, articulation, and crafting speech to influence outcomes, but these attributes on their own don’t make for an effective decision maker without competence and self-awareness.


As Enterprise and Solution Architects have the ability to steer technology direction to help organizations meet strategic outcomes, getting a hold on confirmation bias is paramount. Additional consideration regarding our decisions needs to be had and we need to be methodological about it.

Hyper-focus is when someone has a vision or desired target state and proceeds to be hyper-focused on this singular outcome or idea. Hyper-focus on single outcomes such as a successful product launches or business directives are fine as long as adequate attention is given to risks and constraints. However, hyper-focus on outcomes that are predicated on self-serving ones agenda or hypotheses lead to weak and faulty decision making.

Imagine EA recommendations for moving ERP systems away from monolith on-premise solutions towards cloud integration and PaaS models. One EA may consistently consume and relay information that only validates his or her belief that this is bad due to excessive operational costs, lack of data ownership, jurisdiction of data issues, change of vendor, and inherent cloud security risks. The more balanced EA approach is to consider those things, understand the weight and impact, understand where these issues can be mitigated, and rationalize the advantages of a cloud approach over the long term. However, if left solely to the devices of the first EA, the narrative pertaining to PaaS models would make it, unfairly, look quite bleak with the resulting directives having quite a negative impact on the organization over time.

In addition to the ideas presented above, the following positive-patterns, specific to technology and process, can help ensure confirmation bias is minimized within EA disciplines:

  • Don’t sugar coat information based on what you believe people want to hear. It’s important to be as accurate and precise as possible. Sugar coating information is a form of confirmation bias as you are assuming you know what the other person wants to hear and are tailoring your response towards that.
  • Create an effective ARB gating process which includes the most competent thinkers across the discipline. Be wary of “the elders” approach where you limit questions at the gating process to a small handful of approved people. This is akin to always asking the same people to challenge or help assess the quality of your decisions which can reduce diversity of thought and create thought bubbles.
  • Don’t seek validation. Objectively seek out critical questions and the correct path forward.
  • Don’t assume that what worked in the past will work in the future, and be wary about making decisions based on this presumption.
  • Avoid getting stuck on legacy or expensive ‘technology stacks’ largely premised on the relationship with the technology vendor.
  • Avoid decisions on technology that are largely based on an architect or a decision makers familiarity with that technology. These decisions may not inherently be in the best interests of the organization and should be scrutinized further. Decisions predicated on the organization’s capability to operate and deliver should be the focus instead.
  • Appreciate when given information that is contrary to your beliefs or position and proceed accordingly with an objective evaluation.
  • Decisions often need to be made even when we don’t have all of the facts. Ensure that any decision made is caveated with the weight of the decision given what we know and don’t know. Be objective and avoid rash choices which may solely be predicated on our own confirmation bias.

Post a Comment

Your email address will not be published. Required fields are marked *