Instead, it may be useful to focus on the influence of performance on MCS. In the studies we reviewed, "performance" was defined as the dependent variable and research sought to identify independent variables that produced variations in this construct! There were no instances of performance being studied as an independent variable, which would recognize the possibility of feedback loops. In the wider management accounting literature, performance has been included as an independent variable in explaining some characteristics of the design, use and usefulness of control systems [Chenhall, 2003]. In future MCS- strategy research, performance might be regarded as a contingent variable influencing the design or use of a particular control system (although from a contingency perspective, this could present problems of assessing "proper fit" [Langfield-Smith, 1997]). For example, it is not inconceivable (but intuitively appealing) that poor performance might influence the ways in which MCS are designed and used [Langfield-Smith, 1997].