Learn it → Apply it → Save as Skill
The same facts, worded two ways, can produce two different decisions. Whether information is presented as a gain or a loss, as a percentage or a count, changes what people choose, even when the underlying numbers never move.
The Dr. Wallace’s Favorite Mental Models is designed to turn useful ways of thinking into reusable tools. Each model gives you a fast visual overview, deeper interactive learning, practical application exercises and an AI-ready Markdown skill you can save for future use.
Instead of merely reading about mental models, you can gradually build your own AI skills library: a collection of portable reasoning skills that can help ChatGPT, Claude, Grok and other AI assistants apply the mental models you value.
When you describe an outcome as a gain instead of a loss, or as a percentage instead of a count, you change the reference point people judge it against. Preferences that felt obvious in one frame often reverse in the other.
Important: both bars describe the exact same outcome. Only the wording changed, and for most people, so did the felt weight of the choice.
The model is most useful whenever a choice, a message, or a policy has to be described in words, and those words carry a reference point whether you intend it or not.
Framing analysis separates the wording of a statement from the facts underneath it, then checks whether your preference depends on the wording.
Notice the reference point already in use: gain versus loss, relative versus absolute, default versus opt-in.
Restate the same facts using the opposite frame, such as gains as losses or survival as mortality.
Check whether your preference changes between frames. If it does, the wording, not the facts, may be driving it.
Decide based on the underlying facts, and choose the frame you use with others on purpose, not by default.
Open a use case to see what the model changes.
Before testing whether "90% fat-free" outperforms "10% fat," recognize that the product has not changed. Decide whether the persuasive frame still leaves the buyer with an accurate picture, not just a flattering one.
Present survival and mortality framing together so the patient sees the same facts from both angles, rather than only the frame that nudges toward one option. Informed consent depends on it.
Frame a concession as an avoided loss for the other side ("you keep X") as well as a gain, since loss-framed language often carries more psychological weight for the person receiving it.
Describing a measure in terms of lives saved versus lives lost changes public support even when the underlying numbers match exactly. Expect that, and decide deliberately which frame to lead with.
Notice whether "you saved $40" versus "you spent $360 instead of $400" is being used to nudge a renewal or upgrade decision, and check the plain total either way.
"10% behind schedule" and "90% complete" describe the same project. Choose the frame that gives the clearest, most decision-useful picture, not the most flattering one.
This is the classic study that established the effect, still widely cited and independently replicated.
Expand for the origin of the finding and how it relates to other biases already in this library.
Amos Tversky and Daniel Kahneman published "The Framing of Decisions and the Psychology of Choice" in Science in 1981, using the disease-response problem above to show that logically identical options produced different majority choices depending on wording. It built directly on their earlier prospect theory (1979), which models people as evaluating outcomes relative to a reference point and weighting losses more heavily than equivalent gains.
Yes. Related forms include relative-versus-absolute framing (a "50% increase in risk" versus "1 in 2,000 rather than 1 in 3,000"), attribute framing (90% lean versus 10% fat), and goal framing (what you gain by acting versus what you lose by not acting). All share the same underlying mechanism: the wording sets a reference point the audience compares against.
No, though the two are often confused. Anchoring biases a numeric estimate toward whatever reference number was mentioned first. Framing changes which choice looks better by describing the same information as a gain or a loss, not by planting a number. A single message can use both at once, which is worth checking for separately.
The original medical framing study found trained physicians also shifted their treatment recommendations based on survival-versus-mortality wording, not only patients.
Almost all communication requires some frame. It becomes a problem only when the wording hides or distorts the underlying facts, not simply because a frame was chosen.
A rewrite can accidentally change the actual claim, not just its wording. Confirm the new phrasing still describes the same fact before using it.
A reported 72/28 or similar split illustrates the direction and typical size of the effect in that specific study, not a rate that will recur in every context.
Always describing costs and never savings, or always describing risk and never safety, quietly biases your own analysis before anyone else even sees it.
A truthful statement can be framed multiple honest ways. The ethical concern is accuracy and disclosure, not whether any wording choice occurred at all.
a decision needs informed consent, such as medical, legal or financial choices. Present both frames, or the plain numbers, rather than the frame that nudges toward one outcome.
precision matters more than persuasion, such as engineering specs or a financial audit. Go to absolute figures rather than a chosen frame.
the audience already distrusts persuasive wording. Heavy framing can backfire and cost credibility; plain, consistent language may work better.
This browser tool structures your thinking. It does not send your information anywhere or perform AI analysis.
Save as Skill converts the model into portable Markdown instructions that teach an AI when and how to apply the framing effect, what questions to ask, what evidence to seek, what mistakes to avoid and how to structure its analysis.
Save the skills you find useful and organize them into your own reusable library. The Markdown is intentionally model-neutral, so it can be used wherever an AI system accepts uploaded files, project knowledge, custom instructions or reusable text context. Platform capabilities differ, so this is a portable skill file rather than a claim of one-click installation.
Not inherently. Every statement requires some frame; there is no neutral, wording-free way to describe most facts. The concern is only when a frame hides or distorts the underlying information, or exploits a bias the audience cannot see past.
The original disease-framing result has proven strong and robust to precise wording changes in later analysis. But broader reviews of framing research find effect sizes vary a great deal with context, stakes and the population studied. Treat the direction of the effect as far more reliable than any single reported percentage.
Not necessarily. The neutral numeric statement is the most useful baseline for checking whether a frame is changing your judgment, but choosing which numbers to report, and to what precision, is itself a kind of framing. The goal is deliberate, disclosed framing, not the illusion of a frame-free fact.