Mindful Machine Learning

Abstract. Mindfulness refers to a stance of nonjudgmental awareness of present-moment experiences. A growing body of research suggests that mindfulness may increase cognitive resources, thereby buffering stress. However, existing models have not achieved a consensus on how mindfulness should be operationalized. As the sound measurement of mindfulness is the foundation needed before substantial hypotheses can be supported, we propose a novel way of gauging the psychometric quality of a mindfulness measurement instrument (the Freiburg Mindfulness Inventory; FMI). Specifically, we employed 10 predictive algorithms to scrutinize the measurement quality of the FMI. Our criterion of measurement quality was the degree to which an algorithm separated mindfulness practitioner from nonpractitioners in a sample of N = 276. A high predictive accuracy of class membership can be taken as an indicator of the psychometric quality of the instrument. In sum, two findings are of interest. First, over and above some items of...