15 Types of Cognitive Bias to Avoid in UX Research

Cognitive biases impact our day-to-day on a personal level, but within our work environments, as well. They effect our judgement of others and can impact perceptions. This is detrimental to UX design and UX research that needs to be valuable and accurate to deliver on successful user-centric projects. Understanding cognitive biases and how to avoid them throughout your UX research process is imperative if you’re aiming to deliver a high-quality finished product to your clients.

In this article, I’ll highlight 15 cognitive biases UX researchers regularly come across. Read on for 15 types of cognitive bias to avoid in UX research:

Confirmation Bias

Confirmation bias refers to the tendency UX researchers can have to interpret data so that it specifically supports their pre-existing beliefs, effectively ignoring other valuable insights. This is also the most common type of bias in research. The problem this presents is that the final product is not designed with the end user in mind. Researchers need to be careful not to discount data that doesn’t support their assumptions, but to review all data equally.

False-Consensus Bias

False-consensus bias has to do with a researcher viewing their own opinions and behaviours as normal, or common-place, and the opinions and behaviours of others as abnormal. This is similar to confirmation bias in the sense that researchers make assumptions about end users and do not conduct thorough research to support their final design. In some cases, UX researchers with false-consensus bias may not even conduct research as they assume their end users share their outlooks. It’s important for UX researchers to follow due process and conduct thorough research irrespective of their own opinions and behaviours.

Recency Bias

Recency bias is when UX researchers place more value on events that have occurred in recent times, over those that are historic. Because it’s easier to remember our most recent experiences these UX researchers will place more value on their most recent encounters and rely more heavily on them when making design decisions. The easiest way to overcome this type of bias is to ensure that detailed notes are taken of every interview or interaction, for easy reference later on in the research process.

Primacy Bias

Primacy bias effectively translates to “first impressions count”. In some cases, UX researchers tend to remember the first, or most memorable, impressions in the interview stage, and disregard all other encounters. As with recency bias, it is important to keep details notes of every interview so that the data can be thoroughly reviewed, and primacy bias can be avoided.

Anchoring Bias

Referred to as “anchors”, anchoring bias is a cognitive bias wherein individuals tend to rely on their first impressions and make comparisons with new information to that original experience. For example, when conducting research, the UX designer comes across app subscription A at a high price point. So, when the next app subscription, subscription B, is brought up at a lower price point, the UX designer will “anchor” subscription A as the relevant source of information and determine that subscription B is cheap and possibly lower quality. UX researchers can avoid this type of cognitive bias by referencing multiple sources when making key decisions.

Peak-End Bias

Peak-end bias references the tendency UX researchers have to value highly emotional moments, or “peaks”, as well as the last moments, or “end”, of any interaction or experience. To this end, researchers who experience a positive “peak” and a positive “end” will view the overall experience as positive (and the opposite is true if the “peak” and “end” were deemed negative). While it is difficult to reframe intense emotional moments, thorough note-taking can assist when revisiting these interactions.

Implicit Bias

Also known as unconscious bias, implicit bias is our tendency to have a collection of attitudes or preconceived stereotypes which we associate with people unintentionally. In UX research, this can affect research, as UX researchers may conduct interviews within limited parameters, based on demographics or geographics (as an example). It’s important to determine parameters prior to setting out to collect research, as UX researchers can take the time to determine inclusive parameters.

Hindsight Bias

Having hindsight bias refers to some UX researchers’ tendency to overestimate just how predictable events that have happened in the past were. This further influences these UX researchers into believing that they can accurately depict future outcomes, based on this same notion. UX researchers can avoid this cognitive bias by framing interview questions that bring to light more of the individuals past behaviour (that is objective) or avoid framing interview questions that put individuals in the hindsight mindset.‍

‍

Social Desirability Bias

Social desirability bias links to our innate need as humans to feel accepted within our community. Because of this, social desirability as a cognitive bias refers to the tendency interviewees have in answering their questions in a way that makes them fit in or be looked upon favourably. This results in skewed data that can mislead the design team. It is therefore imperative to communicate the confidentiality of interviewees’ answers and to request honesty throughout the interview process, or else conduct observational research instead.

Serial Position Bias

Serial position bias refers to UX researchers’ ability to recall the beginning or end of a list or sequence more accurately than that of the middle. For example, in a list of fetures a designer is more likely to remember those listed closer to the top or the bottom and forget, or struggle to remember, those listed in the middle. One way to overcome serial position bias is to reference key information at the beginning and the end of the interview, or the user interface experience.

Clustering Illusion Bias

Clustering illusion bias refers to the tendency UX researchers can have in analysing data and isolating false patterns that aren’t there. In some instances, random data may present in coincidental groups, such as having occurred in a short time frame, and UX researchers will see this information as trend clusters. One way to avoid clustering illusions is to ensure your interview questions are not leading. Another is to incorporate quantitative data sources into your UX research.

Framing Bias

Framing bias has to do with the tendency people may have to interpret information in a certain way, dependent on how that information is displayed (and not based on the facts presented). UX researchers may reach different conclusions, and make certain judgements, if that same information is presented to them in two different ways. Avoid this cognitive bias by reframing your interview questions to be neutral and open-ended.

Sunk-Cost Bias

Sunk-cost bias refers to the tendency of continuing with a task or project when time, effort, and other resources (such as funding) have been invested, irrespective of whether the cost of that task or project outweighs the benefits gained. By way of example, a UX product design team may utilise resources to build out a feature prototype only to discover that that feature does not provide value to the users. Instead of abandoning the feature, the product design team pushes forward with it. While these situations are never great to realise, the important thing is to not waste any further resources on the item.

Transparency Bias

Transparency bias is the tendency to overestimate how well people understand other people’s internal thoughts and feelings, as well as how well other people understand their own internal thoughts and feelings. The trouble with the illusion of transparency is that it can lead to miscommunication. In the instance of conducting interviews, participants may overestimate how much interviewers are able to glean from their body language and thus do not feel the need to clarify their answers. UX researchers need to incorporate affirmative feedback and take time to clarify any points throughout the interview process for in-depth insights.

Fundamental Attribution Bias

Fundamental attribution bias is when an individual tends to attribute another individual’s decision-making and actions to their character, while they attribute their own behaviours to external factors that they deem are out of their control. Not knowing an individual’s situation and how that may impact their behaviour is what leads us to make judgements of that individual. UX researchers can reframe their own judgements in these scenarios by considering their own actions whenever they have been in a similar situation to better understand and appreciate their interviewee’s response to any given question.

Understanding these 15 types of cognitive biases to avoid in UX research will help you to identify and avoid them to ensure your UX research remains unbiased and of value. Preparation is always key. At the end of the day, the most important thing is to deliver a high-performing product to your clients, and cognitive biases are just one element of UX research that could impact on delivering a quality result.