Applying Generalizability Theory to Differentiate Between Trait and State in the Interpersonal Mindfulness Scale (IMS)

Abstract

The Interpersonal Mindfulness Scale (IMS) is a recently developed measure to assess mindfulness in the context of interpersonal interactions. The IMS showed promising psychometric properties and is considered to be a trait measure, but its temporal reliability and ability to distinguish trait from state have not been rigorously examined using appropriate methods. Generalizability theory (G-theory) is increasingly used to differentiate between trait and state aspects and identify sources of error affecting a measure. This trait/state distinction is important because it is required for evaluation of long-term effects of mindfulness-based training, especially those targeting interpersonal relationships. Responses of 116 participants who completed the scale at three occasions, with 1-month intervals, were analyzed using G-theory and person by item by occasion longitudinal observation design. The 27-item IMS demonstrated strong reliability and generalizability of scores across persons and occasions in measuring interpersonal mindfulness as a trait with G-coefficients of 0.84-0.91. Individual subscales were less reliable. Further investigation combining the most stable items into a trait-only measure did not result in a more reliable measure, nor did combining the least stable items into a state-only measure. This study supported good psychometric properties of the full IMS as a measure of a trait with the total scores generalizable across people and occasions. These findings suggest that the IMS total trait scores can be used to reliably evaluate long-term effects of interventions aiming to enhance interpersonal mindfulness.