Measuring CEO Responsible Leadership: Development and Validation of a Linguistic-Based Instrument

We present a new, linguistic-based measure of CEO responsible leadership (RL). Its theoretical foundations rely on Upper Echelons Theory and its measure is based on a key word dictionary capturing RL from a linguistics approach. This new instrument was developed and validated, by following established procedures in a series of four studies. In Study 1, we develop our theoretically derived baseline dictionary, which we subsequently validate for convergent and discriminant validity against an established measure based on human coding (Study 2) and against other dictionaries (Study 3). In Study 4, we apply techniques from machine learning to a text corpus of a set of responsible and irresponsible leaders: First, we inductively extend our dictionary with algorithm-suggested words used by responsible leaders in our corpus. Second, we use these techniques to examine the prediction accuracy of our dictionary. We find that our dictionary can accurately distinguish between responsible and irresponsible leaders and that machine-generated words increase the prediction accuracy.