The psychological performance inventory: Is the mental toughness test tough enough?

Mental toughness, stemming in part from Loehr's 1986 classic research, is widely alluded to as a critical variable in the popular media and applied sport psychology. We evaluated the construct validity of responses to Loehr's (1986) Psychological Performance Inventory (PPI) by 263 student-athletes from an elite sports high school. As confirmatory factor analysis yielded a poor model fit and an improper solution for the a priori model, we pursued exploratory factor analyses that resulted in a 5-factor model that fitted the data well. However, further analyses showed that key correlates of mental toughness were more strongly correlated with the factors based on the original structure than factors based on the alternative structure. In conclusion, neither the original PPI nor the subset of PPI items in the better-fitting alternative model was a sound measure of mental toughness, indicating that a good fit is a necessary but not sufficient condition for construct validation. Good instrumentation must be strong in terms of conceptual/theoretical considerations, psychometric properties, and relationships to key correlates hypothesised to be meaningfully related to it.