Questions: Psychometric Testing and Assessment Instruments

Question 1

A clinician is screening for depression in a cancer ward, where untreated depression significantly worsens medical outcomes. Which adjustment to the PHQ-9 cut score best fits this context?

• A. Raise the cut score to increase specificity, reducing unnecessary referrals
• B. Lower the cut score to increase sensitivity, catching more true cases even at the cost of false positives
• C. Use the standard cut score regardless of context — that is what standardization means
• D. Eliminate the cut score and rely on clinician judgment alone

Answer: (B)

Cut score selection is a context-dependent tradeoff, not a fixed property of the instrument. In a cancer ward, missing a true case (false negative) has severe consequences, so higher sensitivity — catching more true cases — is the priority even if it means more false positives. Raising the cut score would improve specificity (fewer false positives) but at the cost of missing more true cases, which is the wrong tradeoff here. Option C reflects a common misconception: standardization means the instrument and administration are consistent, not that cut scores are context-free.

Question 2

A student scores 68 on an intelligence test. The test has a standard error of measurement (SEM) of 6 points. The cutoff for an eligibility decision is 70. What is the most defensible interpretation?

• A. The student scores below the cutoff and is clearly ineligible
• B. The student's true score is approximately 68 ± 6, so the score range overlaps the cutoff — the decision requires professional judgment, not mechanical cutoff application
• C. The SEM is irrelevant once a score is obtained; the observed score is the best estimate
• D. The student should be retested until the score stabilizes above or below the cutoff

Answer: (B)

The SEM converts a single score into an interpretable range: 68 ± 6 means the student's true score likely falls between 62 and 74, which spans the cutoff of 70. Treating 68 as a precise, definitive value ignores the inherent measurement error in any psychometric instrument. Option A applies the cutoff mechanically — the very error this concept warns against. Clinicians must communicate scores as estimates with uncertainty ranges and apply professional judgment, especially when scores fall near decision thresholds.

Question 3

A higher cut score on a diagnostic instrument typically improves its usefulness for clinical assessment.

• T. True
• F. False

Answer: False

This is false. Raising the cut score increases specificity (fewer false positives) but decreases sensitivity (more missed true cases). Whether this improves usefulness depends entirely on the clinical purpose. For mass screening where false positives trigger costly or burdensome interventions, higher specificity may be preferred. For conditions where missing a true case is dangerous, higher sensitivity (lower cut score) is preferable. There is no universally 'better' cut score — only an appropriate one for a given context.

Question 4

An instrument validated on a predominantly college-educated, English-speaking Western adult sample may misclassify symptoms in elderly patients with limited education, even if the instrument itself is technically sound.

• T. True
• F. False

Answer: True

True. Psychometric instruments are validated relative to a normative sample — the comparison group that provides the reference distributions. When you use an instrument with a patient whose characteristics differ substantially from that normative sample, you are applying a ruler calibrated on different people. The instrument's technical reliability and construct validity still hold for its original population, but those properties may not generalize. Matching instrument norms to the patient population is a core clinical selection criterion.

Question 5

Why is there no single universally correct cut score for a clinical screening instrument, and what should a clinician consider when selecting one?

Model answer: Because every cut score represents a tradeoff between sensitivity (catching true cases) and specificity (excluding true non-cases), and the optimal tradeoff depends on the clinical stakes. A clinician should consider the consequences of false positives (unnecessary treatment, stigma, cost) versus false negatives (missed diagnosis, untreated illness), the prevalence of the condition in the population being screened, and the downstream resources available for follow-up.

Cut score selection is applied decision theory. Lowering the cut catches more cases but produces more false alarms; raising it reduces false alarms but misses more cases. Neither is universally superior — the right cut depends on what errors are more costly in a specific clinical context. Competent practice requires understanding this tradeoff explicitly rather than defaulting to published cut scores as if they were context-free truths.