Psychometric Properties of the Spanish Full and Short Forms of the System Usability Scale (SUS): Detecting the Effect of Negatively Worded Items

Abstract

This study examined the SUS’s psychometric properties with the Spanish population considering the plausible method effects associated with negatively worded items. A short form consisting of either direct or reversed items was also examined. Participants were 1321 Spaniards who completed the SUS. Confirmatory analyses showed that the SUS was a valid measure with a one-factor structure when method errors associated with negatively worded items were considered (CFI = .932, TLI = .898; RMSEA = .055, CI 90% = 0.047, 0.062), and shown evidence of reliability (Cronbach’s alpha = .76). The short version with only positively worded items also showed to be a valid (CFI = .973, TLI = .946; RMSEA = .057, CI 90% = .041, .075) and reliable measure (Cronbach alpha = .77) to assess usability. This is the first study to clarify the effect of negatively worded items in the structure of the SUS and propose a short version of the SUS to be used with Spaniards when a brief version is preferred to test usability.