Emotion-related impulsivity & our three-factor impulsivity scale
Last revised 7/17/2017
Impulsivity is an interesting aspect of human behavior. It is well known that impulsivity is related to many kinds of psychopathology, and the idea is beginning to be advanced that some facets of impulsivity are fundamental to most types of psychopathology (e.g., Carver, Johnson, & Timpano, 2017). For the last several years, my colleagues and I have been exploring this possibility. The work started with the fact that impulsivity is a very broad concept, with many different facets.
In 2011, we published a factor analysis of a set of measures (some preexisting, some created for that study) bearing on impulsiveness versus self-control (Carver, Johnson, Joormann, Kim, & Nam, 2011). Given the breadth of the impulsivity construct, the number of measures that might be included is very large. Our choices were dictated by several considerations other than the breadth of the impulsiveness construct per se, and those choices clearly did not exhaust the available possibilities.
The factor analysis yielded 3 factors, two of which reflect impulsive reactions to emotions, the third of which reflected impulsiveness that did not obviously have emotions as antecedents. One of the emotion-driven factors reflects mostly overt action, the other reflects mostly mental aspects of impulsivity. These 3 factors were subsequently used in a variety of studies that related them to genetic markers, to early adversity, and to measures of a range of tendencies toward psychopathology, both internalizing and externalizing (Auerbach, Stewart, & Johnson, 2017; Carver, Johnson, & Joormann, 2013; Carver et al., 2011; Carver, Johnson, & Kim, 2016; Carver, LeMoult, Johnson, & Joormann, 2014; Hooper & Carver, 2016; Johnson, Carver, & Joormann, 2013; Johnson, Carver, Mulé, & Joormann, 2013; Johnson, Tharp, Peckham, Carver, & Haase, 2017).
The accumulation of evidence thus far suggests that the two factors that assess reactions to emotion relate to a wide range of problems in behavior. In contrast, the factor that does not reflect emotional antecedents is less reliably related to problems. However, that also tends to make the latter a useful source of information regarding discriminant validity. Thus, we are making available these three indices for other research applications.
Feelings Trigger Action is the label chosen for the factor reflecting overt action in response to emotion. It loaded 3 scales, two of them slightly abbreviated versions of preexisting scales, and one scale created for the project. The preexisting scales were Urgency (from the UPPS, Whiteside & Lynam, 2001) and the Positive Urgency Measure (Cyders et al., 2007). The newly created scale was Reflexive Reactions to Feelings. These scales differ in the extent to which they refer to a particular emotional valence. In part for that reason, there are circumstances in which one might prefer to use one or another of the scales by itself. However, the three also share properties, which can make it useful to use them as an index. Below we provide instructions for creating that index.
Pervasive Influence of Feelings is the label chosen for the factor reflecting mostly cognitive responses to emotion. It loaded several scales: The highest loaders were Negative Generalization (Carver et al., 1988), Sadness Paralysis (created for the project), and Emotions Color Worldview (created for the project). In moving forward, we suggest that an index of those three scales represents the underlying construct well. Again, we recognize that there are circumstances in which one might prefer to use one or another of the scales by itself. Below we provide instructions for creating an index of the three scales.
Lack of Follow Through is the label chosen for the factor that does not obviously relate to emotions. Its strongest loaders were Lack of Perseverance (from the UPPS, Whiteside & Lynam, 2001), the Brief Self-Control Scale (scored in reverse, Tangney, Baumeister, & Boone, 2004), Laziness (from the BICS, Jackson et al, 2010), and Distractibility (newly created). The Laziness scale has a different response format than the other scales, and for that reason we have dropped it from further studies. The Brief Self-Control Scale has cross-loaded on the emotion-relevant factors in other analyses; for that reason we are inclined to regard it as more ambiguous with respect to its place in the 3-factor matrix and will not include it in this factor. We provide instructions below for creating an index of Lack of Perseverance and Distractibility.