Using Single-Case Experimental Design in Psychology Research
Last updated: Aug 15, 2021
I am so excited to be writing my first Stats-of-1 blog post and I am writing about one of my favorite topics: psychology research. I am a clinical psychologist and I often focus on treatment outcome research. Single case experimental design (SCED) is an extremely powerful tool in this research; however, it is often under-utilized. The power of this design is that it allows us to draw causal inferences between an intervention and change in an emotion or behavior.
Psychology typically privileges RCTs as the gold standard, but there are some weaknesses with these designs. For example, by assessing outcomes pre/post, we can’t definitely say that symptom change is a result of the intervention. Our psychological interventions are often 12 - 16 sessions long and a lot of other things can happen in that time frame. People get new jobs, the weather changes, they might start exercising. All of these things can also affect their mental health symptoms. If we only assess symptoms pre/post we can’t be certain that any change is due to treatment as opposed to something else that changed in their lives. On the other hand, with SCED we are gathering data much more frequently from our participants, so we can make stronger causal inferences on outcomes like these (e.g., emotion-related outcomes) that can vary widely depending on when they’re measured.
Below is the link to my dissertation paper. I used SCED to examine the effects of two skills often taught in therapy (mindfulness and behavior change) on dysregulated anger. SCED was the ideal methodology for several reasons:
- Psychological treatments typically consist of multiple skills that are taught to patients. However, some skills are likely more helpful than others and identifying the most effective skills will allow us to make briefer, more powerful interventions. Examining the effects of specific skills using SCED allows us to begin identifying the skills that are the most helpful to patients.
- There is not a lot of research examining these skills on anger specifically so it doesn’t make sense to do a large RCT right away. If this study showed that these skills weren’t helpful we wouldn’t have wasted nearly as many resources as if we started with a RCT.
- It was my dissertation! I didn’t have time (or money) to do a multi-year RCT but I still wanted to conduct a high quality treatment study. SCED allowed me to do that.
I won’t spoil the article for you but I will highlight some key points about how SCED was used in this study.
- The overall goal of the study was to see whether two skills (mindfulness and behavior change) were helpful in reducing dysregulated anger in patients with emotional disorders (e.g., anxiety, mood, related disorders).
- This was a combined series SCED; it used a multiple baseline and a phase change design. If you’re interested in learning more about theses designs, check out these resources:
- For the multiple baseline, patients were randomized to stay in the baseline phase for two or four weeks. During this time they completed symptom measures with no intervention applied. Randomizing the length of the baseline allowed us to ensure change happened when and only when the treatment was introduced.
- Patients were then randomized to receive mindfulness or behavior change skills first (this was their first phase change). Randomizing skill order allowed us to control for order effects.
- The next phase change came after the first skill; once patients completed four sessions of their first skill they received four sessions of the second skill.
What happened? You’ll have to read and find out!
My dissertation paper: sciencedirect.com/science/article/pii/S0005789419301492