I'd keep these ideas handy for a choice that deserves a second look. They offer a vocabulary for noticing which clues are getting your attention.
A heuristic is a rule of thumb for making a judgment. A bias is a systematic tendency for that judgment to lean in a particular direction. A useful heuristic can produce a bias in some circumstances, but the terms aren't interchangeable. Framing and hindsight, for example, describe effects without identifying a particular shortcut behind them.
The everyday scenes below are invented illustrations, not reproductions of the experiments.
1. Anchoring: a starting number pulls your estimate
In Amos Tversky and Daniel Kahneman's 1974 paper, people given different numerical starting points arrived at different estimates. Those estimates leaned toward the starting values.
The researchers described adjustment and anchoring as a heuristic: begin with a number, then adjust it. The potential bias is that the resulting estimate remains tilted toward that starting point.
An everyday example
Imagine spotting a brass reading lamp at a flea market. Before seeing the tag, you think you'd pay about $30. Then you turn the tag over: $95. Suddenly, offering $60 feels restrained. It's so much lower than the asking price.
But it's also twice your original figure, and you haven't learned anything else about the lamp. In this illustration, the seller's number has become the reference point for what counts as reasonable.
Your first guess isn't automatically correct, either. Perhaps the lamp is unusually well made, or comparable ones cost more than you expected. Those would be reasons to reconsider. The useful distinction is between changing your estimate because of relevant information and changing it because a number is now in view.
A question to ask
What would my estimate have been before I saw that number?
When practical, jot down a provisional figure before opening a listing or reading someone else's estimate. Then compare the figures and consider what would justify the difference: condition, size, usefulness, or comparable prices. Your own number is a starting hypothesis, not a verdict.
2. Availability: easy-to-recall examples stand in for a count
The availability heuristic uses how readily examples come to mind to judge frequency or likelihood. Tversky and Kahneman (1974) described judgments of this kind in their research tasks.
The possible bias is a mismatch: an example's ease of recall receives more weight than its actual frequency warrants. Remembering something readily and knowing how often it happens are different sources of information.
An everyday example
Your book club is choosing its next title. Someone says, “Can we please pick a shorter one? We always choose enormous books.” Everyone remembers the 700-page family saga that required three months and two extensions.
Then someone checks the group's last ten selections. Two were unusually long. The other eight were ordinary-length novels, including several that the group finished comfortably.
The marathon read is an excellent answer to “Which book felt longest?” It isn't a count of how often the club chooses long books. In this made-up scene, a memorable example has filled in for the record.
The group might still prefer shorter selections. That's a preference it can state directly, without needing the claim that every recent choice was a doorstopper.
A question to ask
Am I counting, or remembering?
If a relevant record is close at hand—the reading list, a calendar, or a set of receipts—look at it. If there isn't one, describe the limits of what you know: “I remember two frustrating waits” is a more precise claim than “That café is always slow.”
3. Representativeness: resemblance overshadows the base rate
Representativeness means judging how likely something is by how much it resembles a typical case. One bias it can produce is base-rate neglect: overlooking how common the possibilities were before considering the particular description.
In the tasks discussed by Tversky and Kahneman (1974), people sometimes judged likelihood by resemblance while overlooking prior frequency. The distinction matters because a convincing match doesn't, on its own, tell you how common a category is.
An everyday example
Suppose a craft seller has a box containing 100 watercolor greeting cards. She tells you that 90 are printed reproductions and ten are hand-painted originals.
You pick up a card with loose blue washes and an uneven-looking brushstroke. It looks exactly like a watercolor painting, so “original” feels like the obvious category.
But the reproductions also show watercolor paintings. The painted appearance is shared by both categories, and printed cards account for most of the box. If you focus only on how well the card matches your idea of an original, you've left out relevant starting information.
The totals don't settle the question, either. A label identifying this particular card as an original would add evidence. The point is to consider the starting proportions alongside details that actually distinguish the possibilities.
A question to ask
Before I focus on the resemblance, how common is each possibility?
Here, you can also replace a guess with a direct question: “Is this one hand-painted or printed?” The base rate helps you assess your assumption; asking the seller can resolve the uncertainty.
4. Framing: equivalent facts arrive in different wording
A framing effect concerns how a choice is described. In Tversky and Kahneman's 1981 research, paired hypothetical decision problems presented mathematically equivalent outcomes in terms of gains or losses. Respondents' choices differed between those presentations.
These were particular decision problems presented to particular student groups. The findings don't establish that every wording change will alter a choice.
An everyday example
For an imaginary bookbinding workshop, suppose an organizer can accurately describe the same completion record in either of these ways:
- “Nine in ten participants finish their journal during the three-hour class.”
- “One in ten participants leaves the three-hour class with an unfinished journal.”
Both statements give the same completion rate. One directs attention toward finishing; the other directs it toward leaving with work still to do. Consider how each description strikes you before deciding whether the class sounds appealing.
This illustration isn't the gain/loss choice task from the paper. It offers an everyday way to examine what a description puts in the foreground.
A question to ask
How does the same information sound from the other side?
Restate the fact while keeping the numbers, group, and time period fixed. “Nine finish during class” isn't equivalent to “nine eventually finish at home.” If your rewrite changes the facts, you've changed the comparison.
Then return to what matters for your choice. For this workshop, that might include whether you want the instruction, whether the pace suits you, and whether you would enjoy finishing the project afterward. A completion rate is relevant information, but it isn't the whole decision.
5. Hindsight: a known outcome starts to seem predictable
Hindsight bias describes how an outcome, once known, can seem more predictable than it did beforehand. In Baruch Fischhoff's 1975 experiments, knowing that an outcome had occurred increased its perceived prior likelihood. Participants could also overestimate how much they would have known without that information.
These were laboratory judgment findings. They don't establish that every person experiences the effect to the same degree, or explain any particular person's retrospective comment.
An everyday example
Imagine finishing a mystery novel. At the halfway point, you had three plausible suspects. After the reveal, the clues fit together beautifully: the missing letter, the unusual errand, the conveniently timed phone call.
Now the ending seems obvious. You might even remember having been more certain than you were.
For a playful comparison on your next mystery, write your prediction on a bookmark halfway through. Include a second possibility if you're torn. At the end, you'll have a record of your earlier judgment to compare with what the reveal makes persuasive.
A question to ask
What did I actually know when I made the choice?
The same distinction is useful when reviewing an ordinary plan. If an outing was spoiled by rain, separate the forecast available when you chose it from the weather you know arrived later.
A disappointing result alone doesn't establish that the reasoning was poor. Likewise, a lucky result doesn't establish that it was sound. Reviewing a decision fairly means considering the information available at the time—not silently adding the ending.
What the research can tell us
These findings concern specific tasks and conditions. Their direction and size shouldn't be assumed to carry over unchanged to shopping, book clubs, or any other everyday setting. Tversky and Kahneman's 1974 account also recognized that heuristics can be useful; a shortcut isn't inherently a mistake.
Later evidence adds detail for two of the patterns here. In the Many Labs replication project, Richard Klein and colleagues (2014) repeated a gain/loss framing task and four numerical anchoring tasks adapted from later anchoring research. Their aggregate results supported the tested framing and anchoring effects, with substantial variation for some effects, including anchoring.
The anchoring tasks weren't a direct reproduction of the 1974 paper's example, and the project doesn't verify every version of these ideas. For availability, representativeness, and hindsight, this article cites the original work rather than replication evidence. The everyday examples and accompanying questions are ways to examine the concepts, not findings from those studies.
Sources
- Tversky, A., and Kahneman, D. (1974). Judgment under Uncertainty: Heuristics and Biases. Science, 185, 1124–1131.
- Tversky, A., and Kahneman, D. (1981). The Framing of Decisions and the Psychology of Choice. Science, 211, 453–458.
- 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.
- Klein, R. A., et al. (2014). Investigating Variation in Replicability: A Many Labs Replication Project. Social Psychology, 45(3), 142–152.