Evaluation¶

Version: v4 · 2026-09-28 · History
Prime #: 1357
Domain group: Humanities
Origin domain: Philosophy
Subdomain: Evaluation Theory → Philosophy
Also from: Education & Pedagogy, Engineering & Design (beyond software), Law & Governance, Computer Science & Software Engineering
Aliases: Criterion Based Evaluation
Related primes: Comparison, Verification, Validation, Measurement, Evidence, Decision, Cognitive Appraisal, Joint vs. Separate Evaluation

Core Idea¶

Evaluation applies a criterion-bearing frame to a bounded object, interprets relevant observations or features against that frame, and produces a result that counts as a verdict, score, rank, grade, priority, or action-guiding judgment. The evaluator may be a person, group, institution, or rule-governed procedure. The criteria may be explicit in a rubric or specification, or reconstructable from repeated judgments, but without some standard of relevance there is only description.

The abstraction is the relation among object, criteria, observations, and evaluative result. It survives changes of substrate: a teacher grades an essay, a panel assesses a proposal, a classifier scores an input, a court weighs conduct against a standard, and an engineer judges a design against several goals. Their local criteria differ, but the same operation turns features into a purpose-relative judgment.

How would you explain it like I'm…

The Measuring-Stick Check

When your teacher looks at your drawing and checks if it has a sun, a house, and your name, that's evaluation. She has a list of what counts as good, she looks at your drawing, and then she says "great job" or "add your name." Just saying "I see a house" is not evaluation — you need the list of what counts.

Judging by a Rule

Evaluation is judging something by comparing it to a standard. There are four parts: the thing being judged (like a science project), the rules for judging (the rubric), what you notice about the thing (it has a clear hypothesis but no graph), and the result (a B, or "add a graph"). If there is no standard at all, you are only describing something or saying whether you like it. Teachers, judges at a talent show, and even computer programs all evaluate in this same way.

Criteria-Based Judgment

Evaluation is the act of measuring something against a set of criteria and producing a verdict, score, rank, or decision. It always involves four roles: an object (an essay, a bridge design, a job applicant), a criterion frame (a rubric, a safety code, a threshold), the observations of the object that the criteria make relevant, and a result that guides what happens next. The evaluator can be a person, a committee, an institution, or an algorithm. The key contrast is with description: "this essay is 800 words" is description, but "this essay meets the length requirement" is evaluation, because a standard decides which features matter. Grades, diagnoses, risk ratings, and predicted classes all play the same structural role even though they mean different things.

Evaluation is a relation among four roles: a bounded object, a criterion-bearing frame, the observations of the object that the frame makes relevant, and an evaluative result (verdict, score, rank, grade, priority, or action-guiding judgment). The frame does real work: it selects which features matter and maps them to an outcome. Criteria may be explicit—rubric, specification, objective function, threshold, reference class—or implicit and reconstructable from a pattern of repeated judgments. The evaluator may be an individual, a panel, an institution, or an algorithm. A teacher grading, a court applying a legal standard, a classifier scoring an input, and an engineer reviewing a design against safety goals are literal instances of the same structure, not metaphors. What doesn't transfer between them is the local meaning of the criteria and the verdict; a pass mark, a risk tier, and a predicted class play the same structural role but are not interchangeable outputs. Without a standard, you have description or reaction, not evaluation.

Broad Use¶

Education: apply learning criteria to work and produce feedback, grades, or mastery judgments.
Science and medicine: assess evidence quality, intervention effects, diagnoses, or competing explanations.
Engineering and design: judge an artifact against performance, safety, usability, and maintainability criteria.
Law and governance: apply rules and purposes to conduct, policies, or claims and issue a reasoned judgment.
Organizations: score candidates, projects, risks, suppliers, and programs for selection or improvement.
Computation: a rule-governed evaluator maps features and criteria to classifications, scores, or rankings.

Clarity¶

Evaluation separates observing from judging. Measurement can report a value without saying whether it is good enough. Comparison can reveal a relation without making it decision-relevant. Evaluation declares—or at least makes recoverable—the criterion that converts an observation into a judgment.

It also separates the evaluation from its later uses. A verdict can inform a decision without itself choosing a course of action, and it can be persisted as a review without the persisted artifact being identical to the evaluating act.

Manages Complexity¶

Evaluation compresses many features and standards into an inspectable result. Rubrics, scorecards, test suites, review protocols, and multi-criteria models make the mapping explicit, allowing disagreements to be located in the observations, weights, thresholds, or criteria rather than hidden inside a global impression.

The compression is lossy. A score or verdict cannot preserve every relevant feature, and changing criteria can change the result without changing the object. Good evaluation therefore keeps enough of the frame and warrant visible to audit the compression.

Abstract Reasoning¶

For object (x), criterion frame (C), and relevant observations (O(x)), an evaluation procedure (E) produces result (r=E(x,O(x),C)). Changing (C) can change ® while (x) remains fixed; changing observations can change confidence without changing the governing criterion; changing the procedure can alter how multiple criteria are combined.

This grammar prompts four questions: what exactly is being evaluated, against which criteria, from which observations, and with what result type? If any role is missing, an apparent evaluation may instead be measurement, description, comparison, or unsupported preference.

Knowledge Transfer¶

A grading rubric teaches the design reviewer to expose criteria and weights. A software test suite teaches the policy evaluator to distinguish individual checks from the aggregate verdict. Peer review teaches the classifier designer that a result without an inspectable warrant may be difficult to contest even when it is reproducible. What transfers is not the local standard but the object-criterion-observation-result structure and the audit of how those roles are connected.

Example¶

A committee evaluates three proposals. It observes feasibility, expected benefit, cost, and risk; defines criteria and weights; interprets each proposal's evidence under that frame; and produces a ranked recommendation. The ranking is not identical to the raw measurements, the pairwise comparisons, or the final funding decision. It is the evaluative result produced by applying the declared frame.

Relationships to Other Abstractions¶

Current abstraction Evaluation Prime

Parents (1) — outgoing typed relations

Evaluation presupposes Comparison Prime

Evaluation presupposes Comparison because judging an object requires placing its relevant features and a criterion or reference in a shared frame.

Children (74) — abstractions linked here by typed relations

Analysis of algorithms Domain-specific is a kind of Evaluation

Analysis of algorithms is a strict kind of Evaluation: In computer science, the analysis of algorithms is the process of finding the computational complexity of algorithms—the amount of time, storage, or other resources needed to execute them.
Array-access analysis Domain-specific is a kind of Evaluation

Array-access analysis is a domain-specific kind of Evaluation: In computer science, array-access analysis is a compiler analysis approach used to decide the read and write access patterns to elements or portions of arrays.
Binomial test Domain-specific is a kind of Evaluation

Binomial test is a strict kind of Evaluation: Binomial test is an exact test of the statistical significance of deviations from a theoretically expected distribution of observations into two categories using sample data.
Breusch–Pagan test Domain-specific is a kind of Evaluation

Breusch–Pagan test is a strict kind of Evaluation: its frozen identity entails the parent's defining structure while adding domain-specific restrictions.
Capability Maturity Model Domain-specific is a kind of Evaluation

It remains its own entry because its identity is fixed by the organization and assessment scope, five maturity levels and their ordering, key process areas or goals, institutionalization and evidence criteria, appraisal method, current and target level, improvement actions, measurement feedback, distinction between process maturity and output performance and relationship to CMMI.

▸ Show 69 more

Considered purchase Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the buyer and stakeholders, product or service, stakes and reversibility, perceived risks, information search, alternatives and criteria, decision roles, timeline and postpurchase consequences.
Context-Based Sustainability Domain-specific is a kind of Evaluation
The bounded object is an accountable organization or other social system, and the procedure evaluates its measured impact on a named vital capital.
Cover test Domain-specific is a kind of Evaluation
Cover test is a strict kind of Evaluation: A cover test or cover-uncover test is an objective determination of the presence and amount of ocular deviation.
Cumulative Prospect Theory Domain-specific is a kind of Evaluation
CPT instantiates **Evaluation** by applying a criterion-bearing frame to prospects and producing a ranking.
Decision Quality Domain-specific is a kind of Evaluation
Decision Quality assesses a choice process against purpose-bearing criteria.
Doctrine of inherency Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the governing jurisdiction and current doctrine, patent claim and each limitation, single prior-art reference and date, express teaching, asserted unstated feature, inevitability rather than possibility, evidentiary support and burden, enablement and recognition issues and anticipation versus obviousness role.
Domination Analysis Domain-specific is a kind of Evaluation
Domination analysis evaluates an output by guaranteed feasible-solution rank.
Dynamic application security testing Domain-specific is a kind of Evaluation
Dynamic application security testing is a strict kind of Evaluation: Because the tool is implementing a dynamic testing method, it cannot cover 100% of the source code of the application and then, the application itself.
Engineering Critical Assessment Domain-specific is a kind of Evaluation
ECA specializes Evaluation to a specified flaw, failure-mechanics criterion and bounded integrity judgment.
Equine drug testing Domain-specific is a kind of Evaluation
The bounded object is an attributed specimen from a horse in regulated competition.
Exploratory data analysis Domain-specific is a kind of Evaluation
Exploratory data analysis is a domain-specific kind of Evaluation: Exploratory data analysis denotes approach of analyzing data sets in statistics within statistics.
Face validity Domain-specific is a kind of Evaluation
Face validity is a strict kind of Evaluation: its frozen identity entails the parent's defining structure while adding domain-specific restrictions.
Forecast skill Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the forecast and predictand, cases and matching, lead time and spatial scale, deterministic or probabilistic form, verification data, score orientation and decomposition, reference forecast, skill-score formula, uncertainty and missing cases.
Form, fit and function Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the reference and candidate item, controlled specification and configuration, form attributes and tolerances, fit interfaces and installation envelope, function performance and operating conditions, verification evidence, interchangeability decision, excluded material process reliability or certification attributes and change-control consequences.
Full employment Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the economy population and period, working-age labor force and participation, employment and unemployment definitions, frictional structural and cyclical components, potential-output or inflation criterion, estimated full-employment unemployment rate, underemployment and distribution, measurement uncertainty and policy-theory convention.
Hannan–Quinn information criterion Domain-specific is a kind of Evaluation
Hannan–Quinn information criterion is a domain-specific kind of Evaluation: In statistics, the Hannan–Quinn information criterion (HQC) is a criterion for model selection.
Hicklin Test Domain-specific is a kind of Evaluation
Hicklin Test is a strict kind of Evaluation: its frozen identity entails the parent's defining structure while adding domain-specific restrictions.
High-stakes testing Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the assessed construct and test instrument, test taker and other accountable units, score scale and reliability, decision threshold or classification rule, attached benefits sanctions or access consequences, intended use and validity evidence, fairness accommodations and disparate impact, preparation and incentive effects, appeals retesting and error costs and distinction between assessment information and consequential policy.
Highly accelerated stress audit Domain-specific is a kind of Evaluation
Highly accelerated stress audit is a strict kind of Evaluation: HASA (highly accelerated stress audit) is a proven test method developed to find manufacturing/production process induced defects in electronics and electro-mechanical assemblies before those products are released to market.
Holistic grading Domain-specific is a kind of Evaluation
Holistic grading is a strict kind of Evaluation: Holistic grading or holistic scoring, in standards-based education, is an approach to scoring essays using a simple grading structure that bases a grade on a paper's overall quality.
Hostile work environment Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the jurisdiction and governing law, employment relationship, protected characteristic, unwelcome discriminatory conduct, subjective perception, objective severity or pervasiveness, frequency and context, workplace effect, actor and employer-liability rules and available defenses.
Machine-Learning Learning Curve Domain-specific is a kind of Evaluation
**`prime:evaluation` — proposed strict subsumption parent.** The curve repeatedly evaluates model states or fits under a shared metric and produces an action-guiding diagnosis; the child adds ML exposure and split-specific obligations.
MAGIC criteria Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the claim and intended audience, measured magnitude, articulated pattern, population and context of generality, prior beliefs and interestingness, design and inferential credibility, tradeoffs among dimensions and supporting evidence.
Manning criteria Domain-specific is a kind of Evaluation
Manning criteria is a domain-specific kind of Evaluation: The Manning criteria are a symptom-based clinical decision rule for identifying irritable bowel syndrome from the relationship among abdominal pain, defecation, stool frequency, stool consistency, mucus, and distension.
Mean integrated squared error Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the unknown function or density f, random sample and estimator f_n, integration domain and measure, pointwise error, squared L2 norm, expectation over samples, integrated variance and integrated squared bias decomposition, finite-sample MISE and AMISE, bandwidth or complexity choice and integrability assumptions.
MUSHRA Domain-specific is a kind of Evaluation
MUSHRA scores bounded audio outputs against a visible original and calibrated hidden/anchor controls under a listener-quality criterion.
Nimber Domain-specific is a kind of Evaluation
Nimber instantiates Evaluation because it maps each eligible game position to a contextually meaningful value that predicts outcome class and composes across independent subgames.
Optimality criterion Domain-specific is a kind of Evaluation
The accepted reference-grade review places Optimality criterion under Evaluation because the child instantiates or depends on the parent's broader structure while retaining its own constitutive identity.
Outliers ratio Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the video dataset and subjective experiment, objective metric and mapping, mean opinion score and dispersion, tolerance multiplier and interval, paired cases, outlier rule, denominator and uncertainty.
Performance Appraisal Domain-specific is a kind of Evaluation
Performance Appraisal strictly instantiates **Evaluation**.
Ronchi test Domain-specific is a kind of Evaluation
Ronchi test is a strict kind of Evaluation: An optical figure test that samples light returned from a mirror through a periodic grating and compares the resulting band pattern with that predicted for the desired surface and test geometry.
Scoring Rule Domain-specific is a kind of Evaluation
**`evaluation`.** A scoring rule applies a criterion-bearing map to a bounded forecast and realized evidence, producing an interpretable score.
Siegel–Tukey Test Domain-specific is a kind of Evaluation
The bounded object is the relative dispersion of two independent populations as represented by their samples.
SMART Criteria Domain-specific is a kind of Evaluation
Every applied SMART review is an Evaluation using a declared five-prompt objective-writing criterion frame.
Social experiment Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the social question and hypothesis, participants or units, setting and time horizon, manipulated or naturally varied condition, comparison or counterfactual, outcomes and participant perspectives, assignment and confounding, observation and analysis plan, consent oversight and risk, external validity and interpretation limits.
Software inspection Domain-specific is a kind of Evaluation
A software inspection bounds a software work product as its object; uses an artifact-specific approval or defect standard as its criterion frame; assigns trained reviewers and a defined procedure to examine relevant features; and maps those observations into recorded defect findings that license rework, follow-up, and eventual closure.
Software Review Domain-specific is a kind of Evaluation
Software Review is a strict kind of Evaluation: A criterion-bearing examination of a software work product by designated stakeholders to identify defects, assess status or compliance, and produce comments, approval, or action decisions.
Standard-Setting Study Domain-specific is a kind of Evaluation
Standard-Setting Study is a strict kind of Evaluation: Standard-setting study is an official research study conducted by an organization that sponsors tests to determine a cutscore for the test.
Subgroup analysis Domain-specific is a kind of Evaluation
Subgroup analysis is a strict kind of Evaluation: Subgroup analysis refers to repeating the analysis of a study within subgroups of subjects defined by a subgrouping variable.
Subjective validation Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the person and prior beliefs, statement prediction interpretation or pair of events, ambiguity or base-rate weakness, personal relevance and emotional salience, selective matching and memory, perceived accuracy or causal relation, ignored mismatches and alternative explanations, reinforcement loop and experimental comparison or calibration.
Test (assessment) Domain-specific is a kind of Evaluation
Test (assessment) is a strict kind of Evaluation: An examination (exam or evaluation), or test, is an educational assessment intended to measure a test-taker's knowledge, skill, aptitude, physical fitness, or classification in many other topics (e.g., beliefs).
Test Drive Domain-specific is a kind of Evaluation
A Test Drive strictly specializes **Evaluation**.
Test (law) Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the jurisdiction and date, legal issue, authoritative source, elements or factors, logical relation, burden and standard of proof, evidence mapping, exceptions, remedy and precedential status.
Total revenue test Domain-specific is a kind of Evaluation
Total revenue test is a strict kind of Evaluation: In economics, the total revenue test is a means for determining whether demand is elastic or inelastic.
Transfer pricing Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the related entities and ownership or control relation, jurisdictions and tax periods, accurately delineated controlled transaction, functions assets and risks, contractual and actual conduct, transfer-pricing method and tested party, comparables and adjustments, arm’s-length price or range, documentation, corresponding and secondary adjustments and uncertainty or dispute mechanism.
Two-alternative forced choice Domain-specific is a kind of Evaluation
Two-alternative forced choice is a strict kind of Evaluation: Two-alternative forced choice (2AFC) is a method for measuring the sensitivity of a person or animal to some particular sensory input, stimulus, through that observer's pattern of choices and response times to two versions of the sensory input.
Typical versus Maximum Performance Domain-specific is a kind of Evaluation
criterion and observation frame determine what performance verdict means.
Ultimate Fact Domain-specific is a kind of Evaluation
**Evaluation** is the broader abstraction this entry instantiates.
Unit testing Domain-specific is a kind of Evaluation
Unit testing is a strict kind of Evaluation: Checking a bounded software unit against expected behavior with repeatable test cases.
User analysis Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the product and decision scope, intended and excluded user populations, recruitment and sampling, goals and workflows, knowledge skills accessibility and constraints, physical social and technical context, evidence methods, segmentation and uncertainty, derived requirements or personas and traceability to design choices.
Visual field test Domain-specific is a kind of Evaluation
Visual field test is a domain-specific kind of Evaluation: A perimetry examination mapping visual sensitivity across a patient's field of view.
Warranting theory Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the target and observer, identity or trait claim, information cue and its source, target control or manipulability, warranting value, perceived credibility, cross-channel or offline corroboration, competing deception and impression-management explanations, context and observer differences and resulting judgment.
Widely applicable information criterion Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the observed data and pointwise likelihood units, posterior distribution and draws, log pointwise predictive density, variance penalty and effective parameter count, factor of minus two convention, standard error, model comparison and asymptotic relation and diagnostics for unstable contributions.
Yau's conjecture on the first eigenvalue Domain-specific is a kind of Evaluation
It remains its own entry because its identity is fixed by the unit ambient sphere and dimension, closed embedded minimal hypersurface, induced metric, Laplace–Beltrami operator and eigenvalue convention, first nonzero eigenvalue, asserted equality to n, coordinate-function eigenmodes, proved special cases and open general status.
Cognitive Appraisal Prime is a kind of Evaluation
Cognitive Appraisal is the organism-centered species of Evaluation whose object is a situation and whose criteria are goal significance and coping capacity.
Effectiveness Prime is a kind of Evaluation
Effectiveness is a strict kind of Evaluation: it evaluates whether a bounded acting system produces a declared intended result under stated conditions.
Peer Review Prime is a kind of Evaluation
The accepted reference-grade review places Peer Review under Evaluation because the child instantiates or depends on the parent's broader structure while retaining its own constitutive identity.
Perfection Prime is a kind of Evaluation
Perfection is a strict kind of Evaluation: Perfection is the limiting status assigned when an object, act, or state fully satisfies the controlling ideal, completeness condition, or excellence criterion and admits no relevant defect under that standard.
Truth value Prime is a kind of Evaluation
The accepted reference-grade review places Truth value under Evaluation because the child instantiates or depends on the parent's broader structure while retaining its own constitutive identity.
Utility Prime is a kind of Evaluation
The accepted reference-grade review places Utility under Evaluation because the child instantiates or depends on the parent's broader structure while retaining its own constitutive identity.
Verification Prime is a kind of Evaluation
Verification is the strict species of Evaluation whose criterion is a stated specification and whose defined checking procedure yields evidence and a conformance verdict.
Collection development Domain-specific includes Evaluation
Every collection-development practice contains service-relative Evaluation of resources or holdings against collection purpose and criteria.
Expectancy Disconfirmation Domain-specific includes Evaluation
Expectancy Disconfirmation contains Evaluation because the signed expectation-performance gap becomes a criterion-bearing satisfaction judgment rather than remaining a descriptive residual.
Natural Capital Domain-specific presupposes Evaluation
Natural capital requires a criterion-bearing evaluation that treats ecological condition and future service capacity as an asset rather than unpriced background.
Peak-end rule Domain-specific presupposes Evaluation
Peak-End Rule presupposes Evaluation because it specifies how an extended experience is converted into a retrospective verdict that guides repeat-or-avoid choice.
Review Domain-specific includes Evaluation
A Review strictly contains an Evaluation and adds attribution, warrant, persistence, addressability, and documentary affordances around it.
Substantial truth Domain-specific includes Evaluation
The defamation truth rule contains a criterion-bearing evaluation of a challenged imputation against supported facts that yields a legally meaningful truth or falsity assessment.

Condition / exception Strict for the bounded admitted defamation identity. A materially different or factually unresolved comparison remains an application with a negative or triable result; it is not an exception to the evaluative structure.
Joint vs. Separate Evaluation Prime presupposes Evaluation
Joint-vs-Separate Evaluation presupposes Evaluation because it is a mode effect defined over procedures that map criteria and attributes into verdicts.
News Values Domain-specific is a decomposition of Evaluation
Removing journalism-specific factors from News Values leaves Evaluation's object–criteria–observations–result mapping: candidate events are interpreted against a weighted frame to produce scores and coverage verdicts.

Hierarchy path (1) — routes to 1 parentless root

Evaluation → presupposes Comparison

Not to Be Confused With¶

Comparison: places items in a shared frame to read a relation; evaluation uses such a relation to produce a criterion-bearing judgment.
Measurement: maps an attribute to a scale; evaluation interprets one or more measurements or qualitative observations against a standard.
Verification: the narrower evaluation that tests conformance to a stated specification by a defined procedure.
Validation: asks whether an artifact serves its intended real-world purpose, a further specialization built on verification.
Decision: commits to a course of action; an evaluation may inform that commitment without making it.
Review: a persisted, attributable artifact containing an evaluation, verdict, and warrant; evaluation itself need not be published or recorded.
Bare selection: differential retention can occur without a criterion-bearing result. It becomes evaluation only when a rule or process maps candidates against a reference into an evaluative output.