Neural Network Modeling for Organizational Psychology

Introduction

Within a previous chapter of this book we discussed the usefulness of ANN models in order to describe the processes of emergence which characterize most aspects of Organization Science. The conclusions reached in this chapter can be thus summarized:

• 1.

  The emergence processes are of the foremost importance in all domains of Economics and Organization Science, so as to make their study and modeling an unavoidable necessity;

• 2.

  In principle some kinds of ANN models can be used to describe and model emergence processes; however, the most suitable models appear to be the ones based on recurrent networks;

• 3.

  A correct and efficient use of recurrent networks in modeling emergence processes requires a specific and high mathematical as well statistical expertise, a circumstance which, up to recent times, hindered their application to the above quoted domains.

In the present chapter we will shortly review some concrete examples of applications of ANN models to Organizational Psychology. Our aim will be, firstly, to assess the successes and failures of these models, and, in the second place, to understand how they could be related to other models, like agent models, best known within Social Sciences and, in particular, Organization Psychology. This understanding could open the way to an easier and friendly use of recurrent ANN models. Our arguments will be based on a distinction between two kinds of emergence processes: the ones related to the emergence of individual constructs and the others to the emergence of collective constructs. In the former case the presence of suitable individual and contextual features gives rise to the emergence of suitable individual attitudes of wholistic nature. In the latter case the features of single individuals belonging to a group, and reciprocally interacting, give rise to the occurrence of collective features and phenomena. It is to be underlined that both kinds of emergence are of equal importance (and often concomitant) within Organizational Psychology.

In recent times there has been within Organizational Psychology an increase of the interest in the use of computational models (see, for instance, Weinhardt and Vancouver, 2012). This interest concerned also ANN models (even if agent-based models still seem to be more popular among organizational psychologists). However, despite this circumstance, so far we didn’t observe an explosive increase of the number of papers dealing with applications of ANNs to Organizational Psychology. A possible explanation of this lack of papers could be based on the fact that the use of ANN models requires a lot of mathematical and software expertise, often scarce among organizational psychologists. While this chapter does not pretend to help in filling this cultural gap, notwithstanding we will preliminarily introduce some rudimentary notions about the ANN models so far used in Organizational Psychology. Therefore the aim of the following section is to allow the readers to understand at least the general concepts underlying these models.