Assessment
This domain tests whether you can select, design, and interpret assessments, and use results to adjust instruction.
Purpose and timing
Formative assessment happens during learning, stays low-stakes, and produces feedback; summative assessment comes at the end and supports a judgment or grade. Diagnostic assessment precedes instruction. Assessment for learning (formative) and assessment of learning (summative) are not the same as "a quiz" and "a test."
Quality criteria
Validity means the instrument measures the intended construct; content validity is strengthened by a table of specifications. Reliability means consistency across occasions and scorers, improved by more items and clear criteria. Reliability is necessary but not sufficient for validity. Fairness means eliminating construct-irrelevant barriers and providing only documented accommodations.
Formats and feedback
Match the format to the verb: selected response for recall and discrimination, constructed response for explanation, performance tasks and portfolios for complex skills. Analytic rubrics give criterion-by-criterion feedback; holistic rubrics give one overall judgment. Effective feedback is specific, timely, and tied to criteria rather than to the person.
Interpreting scores
Norm-referenced scores (percentile rank, stanine, NCE) locate a student in a group; criterion-referenced scores report mastery of defined content. Grade-equivalent scores are widely misread and should not be treated as placement decisions. Standard error of measurement reminds you that any score is a band, not a point.
Worked example
After a lesson on photosynthesis, a teacher gives a five-item exit ticket. Most students reverse inputs and outputs. The teacher reteaches the next day with a card sort, then rechecks. Reasoning: the exit ticket was formative-low-stakes and used to change instruction-not a summative grade.
Traps that cost marks
Calling every graded task summative; equating reliability with validity; saying a percentile rank means percent correct; using norm-referenced tests to certify mastery; writing rubric criteria that are vague or unrelated to the objective.