Research Design

From 2012-2015, the QIC-ChildRep evaluated two real-world implementations of the QIC Best Practice Model using a rigorous, experimental design. Lawyers for children in Georgia and Washington State were randomly assigned to two groups, an experimental and control. The experimental group received the intervention of two full days of training, plus quarterly follow-up in the form of pod (group) meetings and individual discussions with a resource attorney (coach). These two elements of support were intended to reinforce the attorneys’ retention of the six core skills and to ensure adherence to the intervention model. 

Data were gathered to test whether attorneys practiced according to the QIC Best Practice Model and whether so doing improved safety and permanency outcomes for children. The data cover 4,274 children and 263 attorneys from the two states and 37 different judicial districts. Data sources include state court and child welfare administrative and docket reports; regular attorney surveys, and quarterly pod meeting and coaching session notes. Data from the attorneys cover their personal characteristics and their specific actions in representing a child including activities near the time of the assignment and at six-month intervals. 

Important Differences Between the States

Between Georgia and Washington State, there were some significant differences in the state law and practice. Georgia at the time (the law changed as the study was wrapping up) was technically a best interest state while Washington State was a client-directed state. There were case-to-case variations in both states, however, with some children receiving best interest representation in Washington and some client-direction in Georgia. 

Georgia law required appointed counsel for each child while Washington State law required counsel for the child only when requested by a child over 11 or in the discretion of the court. Thus, the median age of children at appointment of counsel was 6 in Georgia and 11 in Washington State. Again, there was case-by-case and jurisdictional variation.