163 General Discussion 7 Ethica (2023), m-Path (Mestdagh et al., 2023), and Quenza (2023). Recent years have seen many EMA applications become available that continuously innovate to improve the user-friendliness for both the participant and their clinical setting (the iamYu app is another good example; Lichtwarck-Aschoff & Otten, 2023). It is, therefore, no surprise that EMA is frequently used to examine intervention effectiveness, with both nomothetic designs (e.g., randomized controlled trials; Kramer et al., 2014) or idiographic designs (e.g., multiple baseline or ABAB phase studies; Gosens et al., 2024). Furthermore, research questions involving moderation or mediation can also be answered with EMA data. Of course, participants need to invest more time in EMA than in a pre-posttest design, which does entail some burden. However, our feasibility study showed that most participants found this extra burden manageable. Moreover, this extra effort is outweighed by the added benefits of EMA, like participants gaining self-insight and providing research and practice with accurate reflections of change processes. In sum, the technology is available, the method is feasible, and there are no research questions that involve change for which a pre-posttest design is better suited than an EMA design. EMA mitigates many measurement issues of pre-posttest designs, but there is certainly room for improvement. Change will occur at more timescales than EMA can ever hope to capture alone. Employing an EMA design forces the researcher to make choices about the sampling frequency and duration, finding a balance between measuring the construct of interest at the appropriate timescale without overburdening the participant. One can adopt very dense protocols with many surveys per day spanning one or several weeks, or one can employ once-per-day diary protocols that can span months. One way or the other, EMA will miss out on dynamics at faster or slower timescales. With the once-per-day frequency adopted in Chapters 4, 5, and 6 we could not capture change within days (e.g. affect dynamics), but EMA can be complemented by mixing it with other data sources. I have four suggestions to do so. First, routine outcome monitoring surveys are administered in clinical practice once per several months. These assessments capture slower change processes than what EMA typically captures. Relating change on both timescales to each other will enrich the picture of the overall change process, for example by contrasting fine-grained EMA data with the slower change processes as captured with routine outcome monitoring at a more abstract level during that same timeframe. Second, wearables hold great promise for monitoring physiological changes in heart rate, skin conductance, bodily temperature, or movement. The great advantage of wearables is that they 1) are feasible for people with intellectual disabilities (de Looff et al., 2022), 2) collect data passively, evoking no burden on participants other than putting
RkJQdWJsaXNoZXIy MTk4NDMw