Educational Measurement, Assessment, and Evaluation
What this domain tests
Educational Measurement, Assessment and Evaluation tests whether you can choose and interpret assessment tools, judge item quality and use results to improve learning. Expect definitions, formula-based item analysis and school-based assessment (SBA) scenarios.
Core rules
- Assessment gathers evidence; evaluation judges it against criteria; measurement assigns numbers.
- Validity: the test measures what it claims. Reliability: consistency across occasions/raters. A test can be reliable but not valid.
- Norm-referenced tests compare students; criterion-referenced tests compare performance with defined outcomes. SBA in Sri Lanka is mainly criterion-referenced and formative.
- Table of specifications: blueprints content by topic and cognitive level so coverage matches objectives.
- Item analysis: difficulty index p = number correct ÷ number attempted; discrimination D = p(upper group) − p(lower group), often using the top and bottom 27%.
- Formative assessment is during learning; summative at the end; diagnostic locates specific difficulties.
Worked example
An item is attempted by 100 students. The top 27% group has 24 correct; the bottom 27% has 9 correct; total correct = 33. p = 33/100 = 0.33, so the item is difficult. D = (24/27) − (9/27) = 0.89 − 0.33 = 0.56, which is good discrimination. Reasoning: a difficult item can still be useful if high achievers answer it more often than low achievers; if D were negative, the item would be suspect.
Traps
- Assuming p = 0.9 is a good item; it is very easy and may not discriminate.
- Confusing validity with reliability, or reliable with fair.
- Using SBA only as a mark bank rather than feedback.
- Treating a low mean as proof the test was valid; the blueprint and item quality matter.