Social influence on salespeople’s adoption of sales technology: a multilevel analysis

Our theory and empirical models assume a unidirectional causal relationship between the sales manager’s SFA adoption and salespeople’s SFA adoption. To test the possibility of reversed causality we calculated a nonrecursive model with both paths tested simultaneously. This procedure is applicable if there is an additional exogenous variable that correlates with one of the predictors and not with the other (Kline 2005). In principle, any variable that explains variance in the one but not the other latent factor can be used. Organizational identification of the manager (we used one item “When I talk about [organization’s name], I usually say ‘we’ rather than ‘they’”) proved to be such a variable; it was significantly related to sales manager’s SFA adoption but not to salespeople’s SFA adoption. Using AMOS 16.0 we calculated a model with sales manager’s SFA adoption and salespeople’s SFA adoption as latent variables. Finally, we entered organizational identification of the manager as the additional exogenous variable. As predicted, the path from sales manager’s SFA adoption to salespeople’s SFA adoption was significantly larger (ß = .38, p < .01), than the opposite path (ß = .04, n.s.). Finally, a model with a path from sales manager’s SFA adoption to salespeople’s SFA adoption fits much better than one with a path from salespeople’s SFA adoption to sales manager’s SFA adoption (1, Δχ² = 7,97).