Reliability and validity of the Self Report Symptom Inventory (SRSI) in youth: An experimental study of feigned and careless responding

This study evaluated the reliability and validity of the Self-Report Symptom Inventory (SRSI) in adolescents, aiming to determine whether the SRSI effectively distinguishes between valid and distorted symptom reports under different instructed response conditions.
Eighty-eight adolescents, aged 12 to 17 years, from the Dutch-speaking general population completed the SRSI twice: once under honest instructions, and once under one of four randomly assigned conditions: honest (n = 24), feigning depression (n = 23), feigning pain (n = 22), or careless responding (n = 19). Internal consistency, test-retest reliability, and discriminative validity were analyzed. Exploratory analyses examined the impact of careless responding on SRSI scores and the utility of consistency items (i.e., positive health items) in differentiating between careless and feigned responding.
The SRSI demonstrated high discriminative validity between honest and feigned responses, with specificity of .90 and sensitivity of .91 at the standard cutoff (> 9). When careless responses were combined with feigned responses – both reflecting distorted symptom presentations – sensitivity decreased to .80. Internal consistency was high for both the genuine (α = .91) and pseudosymptoms (α = .81) scales, and test-retest reliability was high (n = 24; r = .93 and .88, respectively). Among careless responders, approximately 50% scored above the SRSI pseudosymptoms standard cutoff. Exploratory findings suggest that endorsement of multiple consistency items after failing the SRSI may reflect careless rather than feigned responding.
These initial findings support the potential utility of the SRSI as a symptom validity measure for adolescents. The lower specificity at the standard cutoff than that found in experimental feigning studies in adults (.90 vs. ≥ .95) suggests that the cutoff may need to be raised in youth samples. Further validation in clinical populations is necessary before recommending its use in applied settings.