Mysuru streaming plan: one test, one cut-off, and the "most creative student" scholarship

Technical objections

• Validity: a verbal reasoning test measures verbal reasoning, not "general learning ability" for a science curriculum. The inference from score to placement is not warranted.
• Standardisation: an ad hoc translation by one lecturer, with no back-translation, no try-out and no reliability estimate, breaks the test's standardisation. Scores are not interpretable.
• Norms: norms from urban students years ago do not describe this population; a cut-off derived from them has no defensible meaning.
• Culture and language bias: the construct and item content favour particular home backgrounds; the score partly measures familiarity, not ability.
• Measurement error: a single 40-minute group test cannot support a high-stakes, irreversible placement. A confidence band, not a cut-off, would be the honest output.

Ethical objections

• Labelling: "slow learner" attached at entry becomes a self-fulfilling prophecy (teacher expectations, peer treatment, the student's self-concept).
• Diagnosis by proxy: low IQ is not a diagnosis; identifying a learning disability requires a multidisciplinary assessment by qualified professionals, not a group test.
• Consequences: placement is effectively irreversible and shapes subject choices and self-belief at 16.
• Consent and transparency: parents are not being told what the test is, what its limits are, or how to appeal.

What is defensible

Diagnostic subject-specific screening for the purpose of support, not segregation: short curriculum-based tests in each subject, error analysis, teacher nomination, work samples, attendance and language background, reviewed after four weeks, with every student remaining in the mainstream class and receiving add-on support. Kiran's profile is a signal for investigation and mentoring, not for a label.

Kiran (IQ 128, 46%, 58% attendance)

• Spearman: high g predicts potential, not achievement. A large gap between g and attainment means the constraint lies outside ability - motivation, attendance, emotional state, study skills, or home circumstances.
• Sternberg: he may be strong analytically and weak in practical and creative application - or strong in all three and simply not doing the work the examination rewards.
• Personality: 58% attendance and no on-time submissions point to low conscientiousness, disengagement or avoidance, not low ability. Conscientiousness is the trait that most consistently predicts academic attainment.
• The honest conclusion: "high IQ, low output" is a puzzle to be investigated, not a student to be dismissed as lazy.

Divya (IQ 98, top of class, prize-winning writer)

• IQ around the average does not cap achievement; her conscientiousness ("relentless") and sustained practice explain a great deal.
• Gardner's multiple intelligences make the point that linguistic-creative strengths are real strengths that a verbal reasoning test may under-sample or that her test anxiety may have depressed.
• The 30-point gap between the two students' IQ scores and their attainment is the strongest evidence in the room that IQ alone predicts very little here.

What the college should do for Kiran

An individual meeting within a week; check for sleep, health, home and mental-health issues; a small number of proximal goals; daily/weekly attendance monitoring with a named mentor; catch-up on missed practicals; a structured assignment schedule with immediate feedback; and a re-check at four weeks. Only if difficulties persist despite this should a formal, professional assessment be considered.

Problems with the downloaded test

• A five-minute unusual-uses task samples divergent thinking, which is a component of creativity, not creativity in writing. The criterion (creative writing) and the predictor do not match.
• Single administration, single scorer, no rubric, no reliability check: scores will reflect the scorer's impressions as much as the student's ability.
• One task at one moment cannot distinguish fluency from genuine originality; a student who writes slowly but brilliantly is penalised.
• Using a foreign, unstandardised instrument as a high-stakes selector is also unfair on students who have not met the task format.

Defensible process

1. Define the criterion first: a portfolio of three pieces (one set in class, two at home), a 45-minute supervised writing task, and a short interview about the student's process.
2. Score with a rubric on the four divergent-thinking dimensions - fluency, flexibility, originality, elaboration - plus craft (structure, voice, diction) and revision quality. Publish the rubric in advance.
3. Use two independent scorers and resolve differences above a set margin by a third reader; report inter-rater agreement.
4. Control for advantage: the supervised piece ensures the work is the student's own; the same conditions and time for all.
5. Include teacher evidence and the student's own statement, weighted, so that a single bad day does not decide the outcome.
6. Record and disclose the process, keep scores on file, and tell applicants how the decision was made and how to seek review.

Say plainly that the lecturer's downloaded task may be kept as one small piece of evidence, but a scholarship of this kind cannot rest on a five-minute score from a single marker.