Toward better research designs

Meditation research has usually relied on cross-sectional group comparisons (meditators vs. non-meditators) with either pre/post measurements (often with randomized control groups) or single measurements (with matched control groups). This is a feasible approach for relatively simple research questions (e.g., “Is the meditation group better/superior than the control group in variable X?”). However, group comparisons involving experienced meditators as participants can be expected to often suffer from high heterogeneity because these participants differ in many ways and it might then be difficult to find a control group that matches the experimental group in all important respects. In other words: Control of “nuisance variables,” that is variables other than meditation that might also have an impact on the outcome, is hard to achieve. The ensuing variation in results that can be expected in such cases makes it difficult to detect even pronounced effects. Moreover, there might not be so many experienced meditators who are available for examining a specific research question, thus yielding low sample sizes with low chances of finding an effect. Connected to the first two points, a third drawback with the usual group comparisons is that a single cross-sectional measurement (or even two such measurements) cannot really capture more detailed and specific changes over time that might be postulated by more precise theories of meditation. And fourth, more specific, custom-tailored measurements, as postulated earlier, considering meditators’ personalities and experiences, are hard to make in a group setting. Therefore, we suggest that designs in meditation research at least in part move toward the individual and to repeated measurements. Two ways seem especially appropriate: Single-case experimental designs and designs that make the role of researcher and meditator in principle exchangeable.

< Prev   CONTENTS   Source   Next >