Screening college students for alcohol problems: psychometric assessment of the SASSI-2. Substance Abuse Subtle Screening Inventory.

Abstract

This study examined the psychometric properties of the Substance Abuse Subtle Screening Inventory (SASSI-2) with college students. The first part of the study examined the test-retest reliability of the SASSI-2 at 2-week (n = 55) and 4-week (n = 47) intervals with students recruited from introductory psychology courses. The second part used a new sample of 164 college students to assess the internal consistency of the SASSI-2 subscales, the relationship of the SASSI-2 with other common alcohol screening instruments, and its reactivity to response biases. Participants completed the SASSI-2 under standard instructions, or instructions to minimize any alcohol problems they might have (fake good), or to exaggerate any alcohol problems they might have (fake bad). They then completed the CAGE, MAC and MAST under those measures' standard instructions. The 2-week and 4-week test-retest reliability of the SASSI-2 subscales was moderate to high, but the test-retest reliability for overall classification as "chemically dependent" or "non-chemically dependent" was moderate at 2 weeks and relatively poor at 4 weeks. Internal consistencies for the SASSI-2 subscales ranged from poor to strong (Cronbach alphas ranging from .11 to .93), and the SASSI-2 drinker classification status correlated moderately with drinker classifications based on the CAGE, MAC and MAST. There was no difference in the proportion of participants classified as "chemically dependent" on the SASSI-2 when instructed to "fake good" and under standard instructions, but college students scored as "chemically dependent" more frequently when instructed to "fake bad." When the SASSI-2 is used with college students, its reliability and validity are comparable to other instruments that currently exist in the public domain. The vulnerability of the SASSI-2 to intentional faking warrants further investigation.