ACE-Expert flashcards
ACE-Expert · Build recall with a focused review session.
33 cards. Read the term and the definition here. The deck at the top of this page is optional.
- Test blueprint
- A documented content outline specifying domain areas, item counts, and cognitive-level targets that every operational exam form must match to ensure content validity.
- Subject matter expert panel
- A group of qualified practitioners recruited against explicit criteria to write, review, or set standards for exam content, with documented conflict-of-interest disclosure.
- Governance authority
- The designated body holding decision rights over exam content, scoring, and standards, with documented accountability and version-controlled specifications.
- Technical manual
- A program document recording blueprint rationale, item development procedures, psychometric methods, and standard-setting evidence to support exam defensibility.
- Role separation
- A governance principle keeping item writing, review, and standard-setting roles distinct for the same content to reduce bias and conflict.
- Version control
- A documented system tracking changes to exam specifications and items, including a change log and re-validation of affected content after edits.
- Item difficulty (p-value)
- The proportion of examinees answering an item correctly; values near 0.0 or 1.0 yield little information about candidate ability.
- Point-biserial correlation
- An item-total correlation indicating how well an item discriminates between high and low performers; negative or near-zero values flag defective items.
- Distractor analysis
- Examination of how often each incorrect option is chosen; effective distractors attract some examinees while high performers rarely select them.
- Single correct key
- A well-constructed item property where exactly one option is unambiguously correct and no length, grammar, or format cues reveal it.
- Cognitive level
- The classification of an item by the mental process required (recall, application, analysis), matched to blueprint targets during item writing.
- Item pretesting
- Administering new items in a non-scored capacity to gather difficulty and discrimination data before they enter the operational item pool.
- Sensitivity review
- A fairness screen examining items for cultural bias, stereotyping, or construct-irrelevant content that could disadvantage subgroups before operational use.
- Classical test theory
- A measurement framework stating observed score equals true score plus error, with reliability defined as true-score variance divided by observed-score variance.
- Cronbach's alpha
- An internal-consistency reliability estimate based on inter-item covariance, commonly used for polytomously scored items and scale scores.
- KR-20
- The Kuder-Richardson formula 20, an internal-consistency reliability estimate applicable to dichotomously scored (right/wrong) items.
- Standard error of measurement
- An index of score precision computed as the standard deviation times the square root of one minus reliability, used to build confidence bands.
- Item response theory
- A modeling framework estimating item difficulty, discrimination, and guessing parameters to describe the probability of a correct response across ability levels.
- Test information function
- An IRT-derived curve summarizing measurement precision across the ability range, showing where the test best discriminates among examinees.
- Common-item equating
- A procedure placing exam forms on a common score scale using a stable set of anchor items shared across administrations.
- Differential item functioning
- A statistical analysis detecting items that behave differently across matched subgroups, flagging them for review rather than automatic removal.
- Modified Angoff method
- A standard-setting procedure where panelists estimate the probability a minimally competent examinee answers each item correctly, and ratings are summed to a cut score.
- Bookmark method
- A standard-setting approach ordering items by difficulty where panelists place a bookmark at the point a minimally competent examinee would master content.
- Contrasting-groups method
- A standard-setting technique deriving a cut score from the score distributions of groups known to be competent versus known to be incompetent.
- Hofstee method
- A standard-setting procedure constraining the cut score using panelist-specified minimum and maximum acceptable pass rates and cut scores.
- Cut score validation
- The process of gathering external evidence, such as job performance or criterion-group outcomes, to support the defensibility of an established passing standard.
- Item exposure control
- A security mechanism limiting how frequently an item is administered, retiring or rotating items that become overexposed to candidates.
- Fairness review
- A screening process for construct-irrelevant content, cultural bias, and disadvantageous language, supplemented by DIF statistics and expert judgment.
- Security incident response
- Documented procedures addressing suspected item compromise, breaches, and score invalidation, including investigation and remediation steps.
- Test accommodations
- Modifications such as extended time or alternate formats provided to eligible candidates while preserving the construct the exam intends to measure.
- Kirkpatrick model
- A program evaluation framework assessing four levels: reaction, learning, behavior, and results, used to judge training and program effectiveness.
- PDSA cycle
- A continuous improvement method (Plan-Do-Study-Act) structuring iterative testing and refinement of exam processes and outcomes.
- Psychometric trend monitoring
- Ongoing tracking of difficulty, discrimination, reliability, and pass rates across administrations to detect drift and trigger corrective action.