Skip to main content
. Author manuscript; available in PMC: 2022 Jun 1.
Published in final edited form as: Behav Res Methods. 2021 Jun;53(3):1031–1045. doi: 10.3758/s13428-020-01472-7

Table 3.

Summary of MNLFA DIF Findings and Four-Study Analogue Design

CONSTRUCT Does MNLFA mis-detect DIF for items with no changes in stem or response scale? Results from contrast code comparing Study 1 versus Study 2. Does MNLFA detect changes in the item stem? Results from contrast code comparing Study 1 versus Study 2. DOES MNLFA detect changes in response scale changes (only, not stems)? Results from contrast code comparing Study 3 versus Studies 1–2. DOES MNLFA detect changes in items with both stem and response scale changes? Results from contrast code comparing Study 3 versus Studies 1–2 and from contract code comparing Study 4 versus Studies 1–3,
Items where DIF not expected because item was identical over condition Items where DIF expected because stem changed only Items with DIF expected because of response scale change only, C2: Items with changes in stem and response scale, C1: Subset of Items and ALL had change in stem and response scale,
Alcohol Consequences 0 items of 9 show DIF 3 items out of 9 show DIF 1 item out of 9 shows DIF 2 items out of 9 show DIF 4 items out of 9 show DIF
Drug Consequences 1 item out of 12 show DIF 4 items out of 11 show DIF 0 items out of 12 show DIF 2 items out of 11 show DIF 2 items out of 12 show DIF