Cognitive Bias Susceptibility Methodology

The Cognitive Bias Test measures susceptibility to seven well-studied cognitive biases — anchoring, hindsight bias, overconfidence, belief bias, base-rate neglect, sunk cost and outcome bias — with short decision tasks, then compares those results with twelve self-rating statements to give a Bias Blind Spot. The output is a profile, not a single “bias score”: seven separate 0–100 scores, two plain averages shown on a map, an archetype named by written rules, and three habits matched to the archetype and the highest scores.

The choice of these seven biases follows Teovanović, Knežević and Stankov (2015), who built scales for the same seven and found them only weakly correlated with each other. This tool uses LifeByLogic's own questions, not theirs. It is an LBL-original exploratory instrument: it has not been normed or validated, and its scores are not percentiles.

Cognitive biases are not curiosities of the laboratory. They have been documented in medical diagnosis, legal judgment, financial decisions and policy reasoning, and they show up in ordinary decisions about jobs, relationships and money. Tversky and Kahneman's 1974 paper in Science argued that biases are systematic byproducts of mental shortcuts (heuristics) that work well most of the time but produce predictable errors in specific contexts.

Knowing which biases showed up in your answers can suggest where to add a check: writing down your own estimate before hearing someone else's, starting a probability judgment from the base rate, or deciding on a stopping point before you invest. The tool reports a pattern rather than one composite because two people with the same average can differ completely in which biases moved them.

Each task follows a format from published heuristics-and-biases experiments; every question, scenario and statement is LifeByLogic's own. The tool does not use or reproduce items from the Comprehensive Assessment of Rational Thinking (Stanovich, West and Toplak, 2016), the Adult Decision-Making Competence battery, the Cognitive Reflection Test, the Heuristics-and-Biases Inventory or any thinking-dispositions scale.

In October 2026 each item was checked against the published items it most resembles. Items that followed a published one too closely were rewritten: two belief-bias syllogisms close to well-known textbook items (walking whales, and living things that need water) and a base-rate problem that used the numbers and structure of Tversky and Kahneman's taxi-cab problem. Items that stated wrong, outdated or ambiguous facts were replaced or corrected, among them a births-per-day figure, a store count, a flight count that varies by season, a disputed city-population comparison, a changing heritage-site count, two ambiguous history questions, a planet's “day”, and the setup, dates or outcome wording of all four hindsight scenarios. The sunk-cost scenarios were also rebuilt as pairs that differ only in money already spent, because the earlier pair compared continuing a language you had already half-learned with starting a new one, which a forward-looking person could rationally prefer.

The scoring follows the published designs where they exist and departs from them where a single person must take every condition; each departure is listed in the limitations below.

Step one: seven task scores and seven self-ratings. Each bias is scored from its own task on a 0–100 scale by the formulas in the table below, and each has a self-rating from the twelve statements. Bars are colored by the test's own cut-points: below 33 low, 33–65 mid, 66 and above elevated.

Step two: the map and the Bias Blind Spot. The map shows two plain averages with equal weights: the three tasks scored against a correct answer (belief bias, base-rate neglect and overconfidence) across, and the four tasks scored by how far an irrelevant detail moved the answer (anchoring, hindsight, sunk cost and outcome bias) up. The grouping is by how each task is scored, not a factor structure estimated from data. The Bias Blind Spot is the mean absolute gap between task scores and self-ratings; its direction is self-flattering when biases with a gap above +5 outnumber those below −5 by two or more, self-critical in the reverse case, and balanced otherwise.

Step three: archetype and habits. The archetype is the first of these rules that fits: The Calibrated when all seven scores are below 35; The Certain when overconfidence and belief bias are both 50 or above; The Escalator when sunk cost and outcome bias are both 50 or above; The Anchored when anchoring is 50 or above and the highest of the seven; The Contrarian when anchoring is below 25 and base-rate neglect is 50 or above; otherwise The Variable. Each archetype's text describes only what its rule guarantees. Three habits follow: an archetype defined by one bias gets that bias's first two habits, one defined by two biases gets the first habit for each, and the remaining places go to the highest-scoring other biases, one habit each (The Calibrated and The Variable get one habit for each of their three highest-scoring biases; ties follow the order of the bars). Each habit card names its source and says whether it was tested in experiments or is a reasoning check.

The paid report. After the result, the page offers The Bias Tripwire, a paid PDF report. Only the seven task scores and the seven self-ratings go to checkout; the gaps, the Bias Blind Spot, the map's two averages, the archetype and the three habits are worked out again from them by the rules above. The report adds a tripwire, an if-then line for the bias with the highest task score (a tie goes to the bias listed first, and when every score is 0, to the bias you rated yourself most prone to). Its check is written down when your self-rating on that bias ran 15 or more points below your task score, and asked as a question in the moment otherwise. The tripwire is the report's own design, not a validated measure, and the report compares you with no one, since the test has no norms.

The table below gives each variable, what it captures, how it is measured, where the task format comes from, and how it is scored.

There are none. The test has not been given to a reference sample, so its scores are not percentiles, there is no average to compare a score with, and the color bands and Bias Blind Spot bands are the test's own cut-points. The research cited documents each effect in groups of people; it does not provide norms for these questions.

Reference age range: adults 18 and over. The tool's brief format is meant for general adult readers and does not require statistical training.

Few items per bias. Each score rests on four to eight questions, so it is a noisy estimate for one person. This test's reliability has not been measured; in a larger battery, Teovanović and colleagues (2015) found acceptable internal consistency for most of the same biases but not for hindsight bias.

Fixed items. A retake uses the same questions, so it partly measures your memory of them, and scores also move with mood, time of day and fatigue.

Within-person designs. Anchoring, sunk cost and outcome bias are classically measured by comparing groups. Here one person sees every condition, so some people will notice the comparison and answer consistently, which tends to make these scores conservative.

Two kinds of scale. Self-ratings and task scores come from different formulas that were never calibrated against each other, so the Bias Blind Spot is directional rather than precise. Classic blind-spot studies compare how people rate themselves with how they rate others; this test compares self-ratings with task answers instead.

The map is a summary. Studies that analyze many biases together report different structures, from largely separate biases (Teovanović and colleagues, 2015) to a weak general factor mostly explained by numeracy and actively open-minded thinking (Erceg, Galić and Bubić, 2022); see Berthet (2021) for a review. The two averages are a summary of your answers, not a validated measure of two abilities.

What is not measured. Hundreds of biases have been described; this tool covers seven. Confirmation bias, framing effects, the planning fallacy and the Dunning–Kruger effect are not measured, and response times are not recorded. Most of the experiments behind these task formats used Western, educated samples, and the test has not been checked across cultures.

Finally: cognitive bias susceptibility is one input into decision quality, not the only one. Domain expertise, situational factors, time pressure, emotional state and motivation all shape actual decisions. The tool is a frame for self-reflection, not a measure of overall judgment quality.

No independent review of this tool is recorded. The October 2026 repair, with its item-by-item rights check, scoring checks and claim fixes, was released after review by the author, Abiot Y. Derbie, PhD, a cognitive neuroscientist. See our About page for the site's reviewers.

• v2.2 (October 17, 2026) — Outcome bias now counts a rating of 0 as 0; before, a 0 was read as 50, so a score with a 0 rating could differ from the one its formula gives. A balanced Bias Blind Spot no longer says your self-ratings tracked your answers when the gaps were not small: when the average gap is 15 points or more, the message gives it and names the two biases with the widest gaps. The result now offers The Bias Tripwire, a paid PDF report, after the three habits.
• v2.1 (October 14, 2026) — Rights check and scoring repair. Items that followed published ones too closely, or stated wrong or outdated facts, were rewritten. Overconfidence now reads 0 when confidence matches accuracy (it read 20 before); anchoring and hindsight average signed shifts instead of counting only movement toward the anchor or outcome; base-rate neglect measures the pull toward the evidence-only answer instead of any distance from the correct answer; the sunk-cost pairs now differ only in money already spent; outcome-bias cases are shown apart without labels; the map shows two plain averages instead of weights described as factor loadings; archetypes follow the written rules above; habit sources were corrected. Saved results from before this date used the earlier scoring for these scores and the map, so they are not directly comparable with later ones. This page was rewritten to describe the tool as it is.
• v2.0 (June 2026) — Provenance update: item labels and framing changed to remove references to specific published instruments. The methodology page published at that time described tasks the tool does not use (a reflection test, conjunction and confirmation tasks, dispositions scales and a 0–10 scale); it was replaced in v2.1.
• v1.0 (May 2, 2026) — Initial public release.

• Arkes, H. R., & Blumer, C. (1985). The psychology of sunk cost. Organizational Behavior and Human Decision Processes, 35(1), 124–140.
• Baron, J., & Hershey, J. C. (1988). Outcome bias in decision evaluation. Journal of Personality and Social Psychology, 54(4), 569–579.
• Berthet, V. (2021). The measurement of individual differences in cognitive biases: A review and improvement. Frontiers in Psychology, 12, 630177.
• Erceg, N., Galić, Z., & Bubić, A. (2022). Normative responding on cognitive bias tasks: Some evidence for a weak rationality factor that is mostly explained by numeracy and actively open-minded thinking. Intelligence, 90, 101619.
• Evans, J. St. B. T., Barston, J. L., & Pollard, P. (1983). On the conflict between logic and belief in syllogistic reasoning. Memory & Cognition, 11(3), 295–306.
• Fischhoff, B. (1975). Hindsight ≠ foresight: The effect of outcome knowledge on judgment under uncertainty. Journal of Experimental Psychology: Human Perception and Performance, 1(3), 288–299.
• Fischhoff, B., & Beyth, R. (1975). “I knew it would happen”: Remembered probabilities of once-future things. Organizational Behavior and Human Performance, 13(1), 1–16.
• Jacowitz, K. E., & Kahneman, D. (1995). Measures of anchoring in estimation tasks. Personality and Social Psychology Bulletin, 21(11), 1161–1166.
• Kahneman, D., & Tversky, A. (1973). On the psychology of prediction. Psychological Review, 80(4), 237–251.
• Lichtenstein, S., & Fischhoff, B. (1977). Do those who know more also know more about how much they know? Organizational Behavior and Human Performance, 20(2), 159–183.
• Pronin, E., Lin, D. Y., & Ross, L. (2002). The bias blind spot: Perceptions of bias in self versus others. Personality and Social Psychology Bulletin, 28(3), 369–381.
• Scopelliti, I., Morewedge, C. K., McCormick, E., Min, H. L., Lebrecht, S., & Kassam, K. S. (2015). Bias blind spot: Structure, measurement, and consequences. Management Science, 61(10), 2468–2486.
• Stanovich, K. E., & West, R. F. (2008). On the relative independence of thinking biases and cognitive ability. Journal of Personality and Social Psychology, 94(4), 672–695.
• Stanovich, K. E., West, R. F., & Toplak, M. E. (2016). The Rationality Quotient: Toward a Test of Rational Thinking. Cambridge, MA: MIT Press.
• Teovanović, P., Knežević, G., & Stankov, L. (2015). Individual differences in cognitive biases: Evidence against one-factor theory of rationality. Intelligence, 50, 75–86.
• Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185(4157), 1124–1131.

Terms used on this page, defined in the LifeByLogic glossary:

Browse the full glossary →

• Take the Cognitive Bias Test — the tool itself.
• Behavior Lab hub — sister tools.
• About the author — Abiot Y. Derbie, PhD.
• LifeByLogic editorial policy — how all our methodology is sourced, reviewed, and disclosed.