A comprehensive, but not-exhaustive index where I list and expand upon documented biases that have been shown to affect data collection, reasoning, decision-making and working with others. It also refers to practical examples and mitigation strategies.

The world is built on foundations of sand: unstable, shifting, and more fragile than we like to admit. People rely on shortcuts, assumptions, and impressions to move through a complex world, and those heuristics can be useful because they help us act without stopping to analyze every detail.

The tradeoff is a deviation between our model of the world and reality. It affects the way we collect information, interpret patterns, make decisions, and work with others.

The motivation to compile this list is self-improvement. An idea becomes easier to spot once it has a name, a shape, and a familiar pattern. This is the first step towards awareness. The goal is not to fully eliminate biases, but to mitigate and reduce its effect, as an effort to get closer to truth.

⚠️ Important: This is intended as a personal guide rather than a rigid framework to impose or expect others. There is a real danger in overcorrecting the factors, or setting expectations of others when it comes to processes that are, by nature, personal and intimate. Please make a sensible use out of this resource.

Notes:

  1. Everyone has made biased decisions. It’s important to remember that it’s a common pattern of how we are wired.
  2. Examples may be slightly uncomfortable to read because they reveal habits or faults that we may have observed in us or others.

Reasoning

Bias Explanation Example Mitigation
Confirmation bias People seek and favor information that supports an existing belief while discounting contradictory evidence. A manager who doubts a candidate pays more attention to mistakes than to signs of capability. Actively search for disconfirming evidence and consider the opposite.
Availability bias Judgments about frequency or importance are shaped by how easily examples come to mind. After hearing about a public project failure, a team overestimates the chance of the same failure happening to them. Check patterns against data instead of relying only on vivid examples.
Fluency bias Information feels more trustworthy when it is easier to process or more polished. A polished presentation seems more convincing than a rougher but more accurate analysis. Do not confuse ease with truth; verify claims against evidence.
Recency bias Recent information is given too much weight. One difficult meeting overshadows months of strong collaboration. Review a longer timeline before forming a judgment.
Negativity bias Negative events or impressions carry more weight than positive ones. One critical comment overshadows several strong pieces of feedback. Weigh positive and negative evidence together through structured reflection.
Authority bias Extra weight is given to a claim because it comes from someone with status or power. A team accepts a senior leader’s opinion without enough challenge. Ask for evidence, not just rank, and make disagreement easier.

Decision-making

Bias Explanation Example Mitigation
Sunk-cost fallacy People keep investing in a bad plan because they have already spent time, money, or effort on it. You’ve spent 40 hours on a website redesign that is clearly not working, but you keep pushing forward because “I’ve already invested too much to stop.” Ask: “If I were starting fresh today, would I still choose this?” Set exit criteria in advance, and schedule review points where past investment is ignored on purpose.
Framing effect People make different choices depending on how the same information is presented, such as as a gain or a loss. A treatment described as “90% survival” feels more attractive than one described as “10% mortality,” even though both mean the same thing. Rephrase the same option in at least two ways, especially gain and loss form. Compare the raw numbers before deciding.
Loss aversion Losses tend to feel more important than equally sized gains. You refuse to sell an item or leave a bad investment because the pain of taking the loss feels worse than the possible benefit of moving on. Evaluate decisions in forward-looking terms: “What is the best move from here?” Use pre-set rules, like maximum loss limits or decision checklists, to reduce emotion-driven sticking.
Outcome bias The quality of a decision is judged mainly by its result instead of the reasoning behind it. A rushed hiring decision is later praised because the hire performs well. Evaluate the process used at the time, not only the final outcome.
Self-serving bias Success is attributed to internal qualities, while failure is attributed to outside circumstances. Someone credits a promotion to talent but blames rejection on politics or timing. Use the same standard when reviewing wins and setbacks.
Self-handicapping Obstacles are created or emphasized in advance so failure feels less threatening. A student says they barely studied so poor results can be blamed on lack of preparation. Focus on honest preparation and reflection instead of self-protection.

Working & Communicating with others

Bias Explanation Example Mitigation
False consensus bias People overestimate how much others share their beliefs, preferences, or behavior. A manager assumes everyone prefers spontaneous meetings because they do. Ask for perspectives directly instead of assuming agreement.
Correspondence bias Another person’s behavior is explained mainly through character, while situational factors are downplayed. A colleague misses a deadline and is judged as careless instead of overloaded. Consider what context may be missing before making character judgments.
Actor-observer effect People explain their own behavior through circumstances but explain others’ behavior through personality. We blame traffic for our lateness but call someone else disorganized for being late. Apply the same explanatory standard to yourself and others.
Halo effect One positive trait shapes a broader judgment of a person or product. Someone articulate or attractive is also assumed to be more capable. Evaluate across separate dimensions instead of relying on one impression.
Charisma bias Confidence, charm, or presence is mistaken for competence or integrity. A persuasive speaker is treated as the strongest thinker despite weaker ideas. Separate delivery from substance by using explicit criteria.
Communication style bias People are judged more by how they communicate than by the value of what they say - it also varies according to culture. An explicit communicator is seen as sharper, while a implicit communicator is treated as vague. Follow-up, focus on receiving clarity, usefulness, and context rather than assuming someone is incapable.
Groupthink The desire for harmony suppresses dissent and realistic evaluation of alternatives. A team quickly aligns around the first acceptable plan because no one wants to challenge the group. Introduce a process that intentionally invites generation and comparison of tradeoffs between multiple options. Examples: bringing outside views, RFCs, play the role of devil’s advocate, psychological safety.
Language bias People are judged unfairly based on their accent, dialect, grammar, word choice, or communication style rather than the value of what they are saying. A candidate is seen as less capable because they speak with a non-native accent, even though their ideas are strong and well supported. Focus on substance over style, and separate communication preferences from actual competence.
Identity-based assumptions People make judgments about behavior, competence, or fit based on stereotypical expectations associated with gender, nationality, political stance or club association. A woman is described as “abrasive” for being direct, while the same behavior in a man is seen as confident leadership. Replace assumptions with evidence, and examine whether the same behavior is being interpreted differently depending on who displays it.

Sampling & data collection

Bias Explanation Example Mitigation
Self-selection / nonresponse bias People who choose to participate, or who actually complete the study, may differ systematically from those who do not. People with strong opinions are more likely to answer a survey, while less engaged participants ignore it. Compare respondents with the target population when possible and design outreach to reduce uneven participation.
Survivorship bias Conclusions are drawn from visible successes while failed or missing cases are ignored. A founder studies only successful startups and assumes their habits explain success. Look for what is missing from the sample before generalizing.
Social desirability bias People give answers they believe are more socially acceptable, respectable, or morally appropriate rather than fully honest ones. A participant underreports prejudiced attitudes or unhealthy habits because they want to appear more acceptable to the researcher. Use anonymous responses, neutral wording, and indirect questioning where possible.
Demand characteristics Participants pick up cues about what a study seems to expect and adjust their answers or behavior accordingly. A participant guesses the hypothesis of the study and responds in a way that confirms what they think the researcher wants to find. Reduce obvious cues, standardize instructions, and separate the study purpose from the participant experience where possible. Participants with previous experience with academic research are more likely to pick up the constructs studied.
Inconsistent design Poorly structured or inconsistently applied research methods distort the data before interpretation begins. Different participants receive slightly different instructions, making their responses difficult to compare fairly. Standardize procedures, wording, and conditions across participants as much as possible.
Sampling bias Some members of a population are systematically more likely to be included in a sample than others. A company surveys only weekday in-store customers and treats the results as representative of all customers. Use more representative sampling methods and ask who may be excluded before drawing conclusions.