Abstract
There is a dearth of research and practice guidelines for understanding how to reconcile employee needs for interaction relative to their need to concentrate in conditions of privacy. There are gaps in the findings and also assumptions underwriting research on 'environment-behaviour' (E-B) relations in the context of understanding the interface between workspace, interaction, privacy and group relations. Based on the identified limitations in previous studies, the research evolved an E-B CIS (Contextual, Integrative and Structural) framework as a basis for rigorously exploring the functional structure of an organisation work setting (i.e., in terms of group membership and required communication patterns) in relation to various workspace parameters (i.e., openness, sitting-visibility, and proximity), unplanned interactions (i.e., chance encounters and spontaneous conversations) and also their associated effects on individual privacy and group relations. The framework afforded three explorative research questions for investigation at both the individual and interpersonal levels applying sociometric techniques. A furniture design company participated in the research with 136 office workers - out of 145 from 6 open-plan offices. The offices were designed with low-panelled workspaces in high density, which allow a high likelihood of interaction and thus posing privacy conflicts. Specifically, the investigation looked at: 1) E-B relations between workspace and interaction, and their associated impacts on concentration privacy, group coordination and member relationships; and 2) the links between concentration privacy and group relations in the context of workspace interactions. The findings are interpreted using two principal constructs: 1) interaction- interdependence balance and 2) workspace-interdependence congruence. The latter is proposed to be a prerequisite of the former, and together offer a pragmatic basis for enhancing group relations whilst controlling for concentration privacy, especially for the workers with a large number of frequently required interactions. Limitations are discussed and future research is also proposed to test the generalisability of the current findings in relation to the two key constructs.