Willem Landman has launched an interesting new effect measure in his PhD project together with Stefan Bogaerts and Marinus Spreen; the Typicality of Level Change. Read in the article below – based on an interview with Willem Landman – about the effect measures used to evaluate interventions by social workers and what Experience Sampling could mean in such research.
Single Case Research
Social workers are constantly learning to place others in their autonomy. That’s their job. In practice, there is an increasing need to work evidence-based; Are the interventions used effective and if so, how effective are they exactly? A Single Case Design is a suitable research design to evaluate the interventions and actions of social workers. The interventions concern individuals or small groups of people, so methods with standardized results do not provide sufficient insight into the client in question. After all, the mean of a group doesn’t necessarily say something about an individual. Single Case Design studies are also are also low in cost and require only a low number of participants. Therefore they are often used to build up the first evidence for a specific intervention, which can then be expanded. Moreover, it is often impossible to immediately have a sufficient number of cases for a Randomized Control Trial for studying the effectiveness of interventions used by social workers. In addition, there are ethical reasons to not work with control groups when researching the effectiveness of interventions on addictions and sexual offences for example. However, there are limitations when using the available statistical measures in Single Case Studies, such as the Nonoverlap of All Pairs (NAP) and Tau-U. For example, it can be determined whether an alcohol addiction of a client decreases, but not to what extent this addiction decreases; Has the intervention halved the number of alcoholic beverage intakes per week or has alcohol intake decreased to zero, compared to before the interventions? So the current statistical measures can show there is a significant difference from the baseline, but it still says little about how big the difference is. And it is precisely this difference that we would like to know.
“Where the effect measures NAP and the Tau-U can show whether an intervention works, the Typicality of Level Change can show how well an intervention works”
Willem Landman and his co-researchers Stefan Bogaerts and Marinus Spreen have introduced the Typicality of Level Change (TLC) as an addition to the existing effect measures (NAP and Tau-U). A statistical effect measure that does map the size of the difference between the baseline (before the intervention) and after the intervention. This allows social workers to more specifically tailor what they want to achieve with a client and how they approach it. In the publication “Typicality of Level Change (TLC) as an Additional Effect Measure to NAP and Tau-U in Single Case Research” you can read how exactly this effect measure is applied.
To get a good picture of the baseline in a Single Case Design, it is important that there are at least five measuring points that precede the intervention. In practice, however, it is very intensive for a social worker to measure situations that give an indication of the baseline. As a social worker, you also don’t want to wait too long with an intervention. Therefore, there is a limitation in how many measurements can be made in advance. Measuring five data points can thus be difficult to achieve. Although the TLC is a very suitable effect measure to still be able to say a lot about the effectiveness of the intervention even though there are not much baseline data points, Experience Sampling could be useful to reduce the problem of a limited number of data points. Willem Landman is currently doing research on a case where forensic psychologists treat a client’s compulsive sexual behavior. The data is gathered by a smartwatch the client is wearing. The smartwatch collects data on heart rate variability and skin conductance data as indicators of craving. This data is collected at multiple times of the day during multiple phases (before, during and after treatment). Such a measurement procedure opens up the possibility to enhance Single Case Designs with Experience Sampling, when the amount of measurements is more extensive than in the earlier described alcohol addiction case; more than five measuring points. This would allow more types of effect measures to be applied, as Standard Deviation will be more reliable and regression models can also be performed. Yet, when the researchers do not have the luxury to include a larger number of measurement points in a research design, evaluating interventions is still very much possible in single case designs, and help determine the effectiveness of daily routine.
In short, the TLC of Willem Landman and colleagues is a good additional effect measure for evaluating interventions and actions of social workers and other practitioners, who in practice still have to deal with a limited number of measuring points prior to the intervention. However, it is interesting to further investigate the possibilities of Experience Sampling as a solution to this limitation.