To highlight dilemmas between construct validity and reliability
Aim
To highlight dilemmas between construct validity and reliability.
Evidence and reflection
Focus on a specific objective for pupil attainment; for instance, one associated with children’s mathematical understanding and computational competence with number. Consider how the competence and understanding of pupils across the country could be assessed with regard to the selected objective. Focus this, perhaps by imagining some individual children whom you know in different schools, or by discussion with teachers or student teachers working in different schools. Do you think your assessment method could reflect what is really involved in understanding and competence (construct validity), and yet be administered in standard and consistent ways by teachers wherever or whoever they are (reliability)?
Extension
Consider any test materials with which you are familiar. How do you feel that the test’s designers have tried to resolve the validity/reliability dilemma? What compromises have they made? Do you think it is possible to devise assessment with high validity and high reliability for all subjects?
