JCE Assessment: Validity, Reliability, and Item Quality
Core concepts
JCE Assessment concerns how evidence of candidate performance is designed, scored, and judged. Candidates must command: validity (does the task measure the syllabus objective?), reliability (would repeated measurement agree?), fairness and bias, discrimination, facility (difficulty), marking scheme design, and the balance of formative versus summative purposes. Know the difference between holistic and analytic marking, and between internal assessment moderated by the board and externally marked components.
Traps in questions
Assessment items often hide a single conceptual error. Look for: (a) calling an assessment reliable because it is valid, or vice versa; (b) confusing difficulty with discrimination, since an easy item can still discriminate well; (c) assuming a high mean indicates a good test rather than a possibly too-easy one; (d) accepting a marking scheme that awards marks for style when the objective is accuracy; and (e) treating internal assessment as unmoderated.
Rule of thumb
Use the "objective first" test: state the syllabus objective, then ask whether the task and the mark scheme both point at it. If either drifts, validity is compromised regardless of how consistent the marking is.
Worked example
A practical task asks candidates to titrate and record results. The mark scheme awards 6 marks for accuracy of the final value and 4 for recorded observations. If the syllabus objective is safe technique, the weighting is misaligned: technique should carry substantial credit. The correct critique names the validity problem (construct misalignment) and proposes reweighting, not simply increasing the total marks.