Cyberaggression in the workplace : construct development, operationalization, and measurement

Abstract

Previous research on aggression and violence in the workplace has focused on factors in an organizationally and physically bounded context. Over the last two decades, as organizations have adopted information and communication technologies to support work processes, a new form of workplace aggression has emerged--symbolic aggression using email.

Cyberaggression, defined as aggressive or hostile behaviours that are either perceived in received email communications, or enacted in sent email communications, is a recent and understudied form of symbolic aggression in organizations. While the enactment of this type of aggression presupposes access to organizational information and communications technologies the unique nature of the computer mediated context in cyberaggression differentiates the phenomenon, and subsequently the construct, from other workplace aggression constructs.

Cumulatively, this research (a) developed measures for both source and target cyberaggression, (b) investigated and confirmed the dimensionality of the cyberaggression construct, and (c) investigated the relationships between cyberaggression and individual and situational predictors, and individual level psychological, somatic health, and behavioural outcomes.

Using a mixed-methods approach three studies were conducted in series. The first study employed a phenomenological approach using a critical incident methodology in order to understand cyberaggression as a social process. The results identified a theoretical and empirical model and several potential measures of cyberaggression.

The second study utilized a survey methodology administered to a sample of individuals who used email at work to develop scales for measuring cyberaggression and investigating the dimensionality of the cyberaggression construct. Exploratory factor analysis suggested that cyberaggression is a multi-dimensional construct that consists of both perceived and enacted behaviours that are both source and target specific and that are empirically differentiable.

The third and final study, also using a survey methodology administered to a large sample of working individuals, was designed (a) to validate the measurement instruments, (b) to confirm the dimensionality and construct validity of cyberaggression, and (c) to identify and model several antecedent and consequent variables related to cyberaggression. The measures were consistent and reliable across two samples used within the study. Structural Equation Modeling using CFA and Latent Variable Path Analysis were used to develop and test a model of cyberaggression. While the resulting structural models exceeded the minimum thresholds for good model fit, there were mixed results in terms of the number of supported hypotheses.