Assessment and Measurement in Health Settings
Core concepts
Master psychometric fundamentals: reliability (internal consistency, test-retest, interrater) versus validity (content, construct, criterion-including sensitivity and specificity), standard error of measurement, and the difference between screening and diagnostic instruments. Know common health psychology measures by construct (e.g., PHQ-9 for depression severity, GAD-7 for anxiety, pain and disability inventories, sleep and fatigue scales, adherence self-report) without memorizing proprietary items. Understand behavioral assessment methods-self-monitoring, functional analysis (antecedents, behavior, consequences), ecological momentary assessment-and the limits of self-report (recall bias, social desirability). Appreciate measurement invariance and cultural adaptation when instruments cross populations.
Traps
High reliability does not imply validity; questions exploit this. A sensitive screener yields few false negatives but many false positives-know which error a stem prioritizes. Base rates change positive predictive value dramatically, so a "good" test in a low-prevalence setting produces many false positives. Watch for confusion between criterion validity and construct validity, and between cutoff scores as screening thresholds versus diagnostic certainty. Also be alert to floor/ceiling effects and to using a measure outside its validated population.
Worked example
A depression screener with 90% sensitivity and 60% specificity used in a low-prevalence primary care sample: most positives will be false positives. Rule of thumb: PPV depends on prevalence; sensitivity alone does not justify a diagnosis.
Study tip
When a stem names a test statistic, ask "which psychometric property is being tested?" before choosing.