Assessment Validity and Reliability Quiz

Assessment Triangle—coherence fix. In a Grade 10 electricity unit, the LO emphasises planning valid investigations and reasoning from evidence. The test, however, is 20 MCQs on definitions, and the rubric awards marks for neat handwriting. Which TWO changes most directly restore coherence among Cognition–Observation–Interpretation as discussed in the triangle diagram?

• Replace half the MCQs with a task where students design an investigation (variables, predictions with justification) and submit a planning sheet.

• Keep the MCQs but extend the time limit so learners can think more.

• Replace the rubric with analytic criteria on modelling, variable control, and evidence-based reasoning.

• Add a surprise oral quiz on definitions to diversify evidence.

• Award bonus marks for creativity in poster layout to motivate students.

Answer explanation

Q1 (A, C) — Assessment Triangle coherence
Explanation: The LO targets investigation design and evidence-based reasoning (Cognition). (A) changes the Observation to collect the right evidence; (C) aligns Interpretation by scoring the intended constructs. Extra time (B), more definitional quizzing (D), or layout marks (E) don’t repair the cognition–observation–interpretation link.

Validity under misaligned weighting. Your course outcomes prioritise experimental design and scientific argumentation, yet (as in the stacked bar chart shown in class) the final exam counts 60–70% while labs count 10%. Which validity threat is most acute here?

• Content validity

• Construct validity

• Criterion-related validity

• Face validity

Answer explanation

Q2 (A) — Validity under misaligned weighting
Explanation: Overweighting the exam while underweighting labs mis-samples the intended outcomes, threatening content validity—the assessment no longer represents the domain emphasis on design and argumentation.

Reliability remedy. Two markers score the same titration lab report very differently. What is the best first action to improve reliability while preserving validity?

• Run a rater-calibration session using anchor samples and refine the analytic rubric descriptors.

• Increase the number of test items.

• Average the two scores; differences will cancel out.

• Ban open-response tasks in favour of auto-marked items.

• Extend submission time by 24 hours.

Answer explanation

Q3 (A) — Reliability remedy
Explanation: Inter-rater inconsistency is best addressed by calibrating markers with anchors and tightening analytic rubric descriptors. More items (B), averaging (C), or banning open response (D) dodge the reliability issue and risk validity; time (E) is irrelevant.

Acting on the radar chart (student perspectives). In our radar chart, labs scored highest for deep learning but only moderate for fairness; group projects were seen as least fair. Which TWO design changes most likely raise perceived fairness without reducing deep learning?

• Share transparent, criterion-referenced rubrics and exemplars for lab reports; hold a short norming demo with students.

• Replace labs with weekly MCQ quizzes of equal weighting.

• Build individual accountability into group tasks (e.g., individual mini-memos, contribution logs, rotating roles) while keeping collaborative inquiry.

• Curve all practical marks to align with exam distributions.

• Reduce project complexity and remove open-ended components.

Answer explanation

Q4 (A, C) — Raising fairness without losing depth
Explanation: Transparent, criterion-referenced rubrics with exemplars (A) clarify expectations and scoring; individual accountability within group work (C) addresses contribution equity. Weekly MCQs (B), curving (D), or removing open-ended tasks (E) either ignore fairness causes or reduce deep learning.

Cycle alignment (Diagnostic → Formative → Summative → back to Diagnostic). Which sequence places the assessment moves in the correct order for a coherent cycle?

• Pre-instruction probe of prior conceptions → Exit ticket during the lab → Cumulative end-of-unit project/report → Entry check at the start of the next unit

• Exit ticket during the lab → Pre-instruction probe of prior conceptions → Cumulative end-of-unit project/report → Entry check at the start of the next unit

• Pre-instruction probe of prior conceptions → Cumulative end-of-unit project/report → Exit ticket during the lab → Entry check at the start of the next unit

• Entry check at the start of the next unit → Pre-instruction probe of prior conceptions → Exit ticket during the lab → Cumulative end-of-unit project/report

Answer explanation

Q5 (A) — Correct cycle order
Explanation: A coherent cycle is Diagnostic (pre) → Formative (during) → Summative (end) → back to Diagnostic (start of next unit). Option A matches this progression; other orders break the feedback loop.