🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
availability_heuristic
The availability heuristic is a mental shortcut where people estimate the likelihood of events based on how easily examples come to mind. Events that are vivid, recent, or emotionally charged are overweighted, while statistically more common but less memorable events are underweighted. This leads to systematic distortions in risk assessment and probability judgments.
After watching extensive news coverage of a plane crash, a traveler decides to drive 800 miles instead of flying, even though driving is statistically far more dangerous per mile traveled.
After her city experiences a highly publicized shark attack, a woman who has swum in the ocean for decades refuses to go back in the water, convinced that shark attacks are now a common and ever-present danger.
Following intense media coverage of a string of home burglaries in a neighboring city, a man buys an expensive home security system and starts double-locking his doors, even though local crime statistics show his neighborhood has one of the lowest burglary rates in the region.
Binary (yes/no) questions an LLM must answer to identify this aspect:
Does the reasoning rely on examples that are easily recalled or recently encountered?
Type: binaryAre vivid, dramatic, or emotionally salient examples given disproportionate weight?
Type: binaryWould a systematic data review produce a different conclusion?
Type: binaryThe availability heuristic is a mental shortcut where people estimate the likelihood of events based on how easily examples come to mind. Events that are vivid, recent, or emotionally charged are overweighted, while statistically more common but less memorable events are underweighted. This leads to systematic distortions in risk assessment and probability judgments.
The brain uses cognitive ease as a proxy for frequency or probability. Information that is easily retrieved from memory feels more representative of reality, even when retrieval ease is driven by recency or emotional salience rather than actual prevalence.
Seek out base-rate statistics rather than relying on memorable anecdotes. Ask yourself whether your estimate would change if you had never heard the vivid example that comes to mind.
Media coverage of rare but dramatic events (terrorism, shark attacks) leads the public to vastly overestimate their likelihood, while common killers like heart disease receive less fear-driven attention.
Drawing broad conclusions from limited, unrepresentative, or anecdotal evidence.
Ignoring general statistical base rates in favor of specific individual-case info.
Asserting a small first step will inevitably lead to extreme negative consequences.
Use these tools to detect, analyze, or train this aspect.