Broad content sampling used to establish relative standing is characteristic of norm-referenced testing.
Explaining how motivations produce a turning point requires analysis, a higher-order process than simply recalling a definition.
Restricted-response essays limit the content and/or scope of the answer, while extended-response essays give wide freedom.
Criterion-referenced testing judges mastery of specified learning outcomes regardless of how other students perform.
Norm-referenced testing compares individuals with other members of the group, which suits selecting the top performers.
A final examination that judges course outcomes and informs program decisions is summative: product-oriented and given at the end.
Formative assessment addresses immediate instructional problems during learning, so calling this diagnostic would be inappropriate.
Difficulties are first handled through formative assessment during teaching; if they persist, deeper diagnostic assessment follows.
Summative assessment evaluates outcomes at the end of a program or course.
Assessment as learning involves students monitoring their own progress, setting targets and adjusting study strategies (self-assessment).
Self-assessment is the classic example of assessment as learning, where learners reflect on and regulate their own learning.
Ipsative assessment compares a learner's current performance with his or her own earlier performance rather than with others.
Diagnostic assessment investigates the underlying causes of persistent learning difficulties that ordinary corrective teaching has not resolved.
Formative assessment monitors learning while instruction is in progress and feeds back into adjusting the next lesson.
Measurement only assigns numbers to performance; it does not interpret what the numbers mean.
Evaluation combines several kinds of evidence to make a value judgment about the success of a program.
Measurement records quantity (scores), while assessment goes further by interpreting performance data to judge the quantity and quality of learning.
Placement assessment is given before instruction to check prerequisite skills and assign learners to suitable groups or levels.
Cronbach's alpha estimates internal consistency among items measuring the same construct.
Content validity is the extent to which test items represent the full syllabus or domain.
Concurrent validity compares a new measure with an established criterion measured at approximately the same time.
Predictive validity is demonstrated when a test predicts a future criterion.
Construct validity concerns whether a test actually measures the theoretical construct it claims to measure.
Inter-rater reliability is consistency between different raters or judges.
Test-retest reliability checks stability of scores over repeated administrations.