M.Phil. / Family Businesses are the most prevalent form of business throughout the world; they are highly pertinent in developing economies (Cavusgil, Kandemir & Kim, 2004). In South Africa it is estimated that 84 percent of all businesses in the formal sector are family-owned (Balshaw, 2004). Even though family businesses represent a dominant form of economic organisation, there is neglect in terms of research, particularly with respect to development of theories that have applied value (Chrisman, Chua & Steier, 2003). The key focus of this study is to provide research evidence for the verification of the predictive model Resulting Commitment Behaviours (hereafter referred to as RCBs) and the relationship(s) of the antecedent’s: Individual Characteristics, Familial Cultural Socialisation and Contextual Factors might have with an employed family member’s RCBs.

Escalation of Commitment (EoC) behaviour occurs when a Decision Making Unit (DMU), such as an individual or group, continues with a course of action despite receiving negative feedback about it. Much research exists, within multiple disciplines, which attempts to explain why DMUs continue with failing courses of action. To date however, there has been very little critical inquiry of such research. Using a historical research approach, this thesis reviews and critically assesses all existing EoC behaviour research and concludes that a number of serious issues exist. These include the use of multiple labels by authors to describe the phenomenon; the considerable uncertainty that exists regarding which DMUs are subject to EoC behaviour; the existence of multiple, concurrent definitions for each ‘theory label’ and important EoC behaviour concepts, such as escalation, DMU, resource, success, failure and commitment, not being adequately defined. It is contended that these and other issues exist primarily because of the scope of the phenomenon and the resultant high quantity and complexity of research; all of which impair research technique. However, independent, pre-existing research technique issues are also proposed as reasons. Ultimately, it is argued that the state of EoC behaviour research is poor. It is considered that the mere recognition of the issues raised in this thesis will assist in the improvement of the research. Yet this aspect in isolation is deemed inadequate. In response, a prescriptive technique is developed which is bifurcated between resolutely defining the important concepts related to EoC behaviour research and creating an ‘integrated framework’ which includes all existing EoC behaviour determinants from all research disciplines. The proposed framework also identifies a number of new potential determinants of EoC behaviour, including the Autoepistemic Sunk Cost Effect (ASCE), the age of the DMU and anthropomorphic revenge motives. It is suggested that these two prescriptive responses will also promote focussed future EoC behaviour research, designated in the thesis as research direction. This thesis contributes to existing knowledge by not only recognising research issues that have not previously been acknowledged but also by prescribing for these issues through a complete concept exploration, coupled with a complete collective framework.