A model is developed by which evaluation of social work services can study a cross-section of factors which affect the performance of that service. It demonstrates how the boundaries of an evaluation can be defined without developing a biased pattern of accountability.
