Decision-Making Glossary for Comparing Two Options

Anchored score

A rating tied to observable descriptions at defined points, such as what 1, 5, and 10 mean, rather than an unexplained feeling. See anchored scores in a weighted matrix.

Constraint

A condition an option must meet to remain eligible. A true constraint is checked before preferences are scored. Use constraints to choose between two options.

Criterion

One distinct consequence or quality used to compare options. Useful criteria can change the decision and do not duplicate another criterion. Choose decision criteria that can change the result.

Decision rule

A prewritten condition that explains how evidence becomes a choice, including what must be true and what downside is acceptable. Build a decision rule for two options.

Decision tree

A map of choices, uncertain events, later choices, and endpoints used when a decision unfolds in stages. Draw a decision tree for two options.

Expected value

A probability-weighted average of possible outcomes. It does not show whether a particular downside is affordable or acceptable. Compare two options with expected value.

Exit cost

The money, time, coordination, reputation, and recovery work required to stop or reverse an option. Compare reversible and irreversible decisions.

False dilemma

A question that presents two options as exhaustive when other legitimate choices, combinations, or inaction exist. Write a fair either-or question.

Hard gate

A pass-or-fail requirement that cannot be compensated for by high scores elsewhere. Check disqualifying constraints first.

Margin of sampling error

An estimate of random sampling uncertainty under a probability-sampling design. It should not be attached casually to an open volunteer poll. Read an online poll without overclaiming.

Opportunity cost

The value of the best realistic alternative use of the same scarce resource that a choice prevents. Compare opportunity cost between two choices.

Premortem

A structured exercise that assumes an option failed and works backward to identify causes, warning signs, and protections before commitment. Run a premortem before a major decision.

Probability sample

A sample selected through a process in which population members have a known chance of selection, supporting defined statistical inference when implemented well. Compare sample size and representativeness.

Representativeness

The degree to which the people and conditions in a sample support conclusions about the population named in the claim. Understand representativeness in online polls.

Reversibility

The degree to which a choice can be changed without unacceptable cost or lasting consequence. Use reversibility to set research effort.

Satisficing

Choosing an option that meets a thoughtful acceptance threshold instead of continuing to search for the theoretical best. Set a good-enough decision rule.

Scenario analysis

A comparison of options across several coherent sets of plausible conditions to reveal robustness and switching assumptions. Compare two options across scenarios.

Sensitivity analysis

A test of how the result changes when an uncertain input, score, weight, or assumption changes by a realistic amount. Test a weighted decision result.

Status quo bias

A tendency to prefer the current or default option beyond what a balanced comparison of future consequences supports. Compare staying and changing fairly.

Sunk cost

A past investment that cannot be recovered. It can provide learning but is not a future benefit created by continuing. Make a decision without chasing sunk costs.

Switching point

The value of an assumption at which the preferred option changes, showing which estimate drives the choice. Find the assumption that changes the choice.

Value of information

The benefit of learning something before acting, limited by whether the information can change the choice and improve the outcome. Decide what is worth learning before acting.

Weighted matrix

A comparison that multiplies criterion scores by importance weights and adds the results, with assumptions and sensitivity kept visible. Build a weighted decision matrix.