One of the central issues when evaluating treatment effectiveness is the issue of causality. This means that we want to know whether observed improvements in client functioning can be attributed to the treatment. In this respect, an RCT-design (RCT: randomized controlled trial) is generally regarded as the 'gold standard' by means of which alternative explanations for client progress can be ruled out. In an RCT-study, clients are randomly allocated to a treatment group and a control group. The primary goal is to test whether an intervention works by comparing it to the control condition, usually either no intervention or an alternative intervention. In practice, however, it is not always desirable or feasible to conduct an RCT-study. In this presentation, I will demonstrate that a single case study with an interrupted time series design can be an interesting alternative to an RCT-design. Single case studies typically involve repeated, systematic measurement of a dependent variable, usually before, during, and after the active manipulation of an independent variable (e.g., applying an intervention). In a single case design, the client can serve as his/her own control because the client’s functioning before (baseline) and during treatment can be compared. Like RCT’s, single case designs can thus provide a strong basis for establishing causal inference. Moreover, single case studies have clear advantages over group-based RCT-studies since they provide information on treatment progress of individual clients, which is usually the aspect practitioners are most interested in.
In this presentation, I will discuss the most important features of single case designs, the research questions they can address, as well as the most commonly used types of single case designs. Subsequently, I will focus on the different techniques available for the analysis of single case data, and discuss their strengths and limitations. More specifically, I will focus on Simulation Modeling Analysis (SMA; Borckardt et al., 2008), for which a freely downloadable software program is available. This program is easily accessible for practitioners and has several advantages over visual techniques and other types of (statistical) analysis. The application of SMA in practice will be demonstrated. It is shown how SMA can identify changes in level and slope of symptoms across intervention phases, on the basis of which causal inferences regarding the effectiveness of the intervention can be made.
