Definition

Single system research design is a term that is associated with program evaluation. This form of research design is highly similar to the experimental designs that are taught in a typical research methods book. There are several differences between single-system research design and experimental design.

The first is the context. Single system research design is associated with program evaluation while experimental design is related to hard and social science research. Another difference is how rigorous each research method is. Generally, single system research design is not as rigorous as experimental design.

Among the reasons for this lack of rigor is sampling. Single system research design is intended to assess how well a program is doing. Therefore, the sample size is limited to the number of people participating in the program. Often there is no sample and all participants are also in the study. The sample size is usually larger for experimental design, but again this depends on the context.

Another difference is random assignment. With single system research design random assignment is not possible which means that there is a lack of independence. Sometimes this lack of independence can also be a problem in experimental design as well but not always.

Since the sampling and the lack of independence are problems this leads to problems with external and internal validity as well for single system research design. It is often not possible to generalize due to sample size or to assess cause and effect due to the limitations of single-system research designs.

Finally, because of these issues, it is often unnecessary to use inferential statistics. If no sampling is done there are no inferences to make about the population. However, for single system research design the purpose is to determine the health and state of the program and not to draw strong conclusions. The goal is intervention rather than the development of theory.

To put it simply, single-system research design is not concerned with being strongly scientific in the traditional sense like experimental design. The goal of this approach is to assess programs and not necessarily to publish data that would withstand the scrutiny of the peer review process

Common Steps

There are several steps involved with single system research design. First, an outcome measure or variable needs to be selected and measured several times. Whatever outcome measure is selected it must be reliably measured and must vary with time.

The results of this measurement must then be graphed to see if the program has had any potential influence on the outcome i.e. a before and after effect. However this depends, there are times when an outcome measure is measured to determine if there is a problem first. For example, looking at reading rates and seeing if they ever fall below a certain threshold to justify the implementation of a program.

Whatever the case, the graphed data is used to make decisions. Again the focus is on that specific outcome measure and not generalizing or making strong cause and effect claims.

Measuring Over Time

The actual design of a single system research design is the same as in experimental design and includes some of the following and more…

• AB
• ABA
• BAB
• ABAB
• Some include a C or a different intervention

These design formats are explained in most research design textbooks. The “A” represents the state of the program without an intervention. The “B” represents measurement during the intervention. The goal is to see a difference in the graphs when the intervention is present to provide evidence that the program is working

How Single System Research Design is Used

To consolidate this information into one place. Single system research design is used for the following.

• Formative-to adjust and enhance an existing program
• Summative-Appraise results and outcomes of a program
• Quality assurance-Check compliance with regulations ie audits of behaviors
  
  The new one is quality assurance this is similar to formative or summative with the difference of a focus on compliance with external standards.

Conclusion

Single system research design is for program evaluators and not really researchers. The difference is what is focused upon. Since it has limited scope, single system research design normally does not meet the standards of science. Despite, this, it is still a tool for assessing the strength and quality of a program.