Dynamic systems theory and the process of adolescent developmental change

Abstract

How do humans learn, change, and grow? This broad question has driven research in the disciplines of education, human development, and psychology for the past century. Microdevelopmental research studies report that periods of high variability are a typical precursor to gains in learning and development; Dynamic Systems Theory (DST) draws on chaos and complexity theories to explain this phenomenon as a normal, and indeed necessary, part of emergent change processes. This study explores the accuracy and applicability of DST predictions when applied to a long-term therapeutic change process. This study analyzed archival behavioral rating scores on a randomized group of 31 high-risk adolescent males during their stays at a residential treatment center. These participants were scored every hour of every waking day for periods ranging from five to fifty-eight weeks. This data set provided a rare example of microdevelopmental measurements spread out over a macrodevelopmental period. The guiding research problem for this study was: are the observed behavioral fluctuations among these adolescent participants consistent with the features predicted by DST? Q-sorts of score graphs combined with a case-by-case qualitative analysis supported this proposition. In addition, fuzzy set analysis was used to find necessary and sufficient causes for these participants' positive or negative discharges from treatment. For the participants in this study, the change process usually displayed the bifurcation patterns predicted by DST. The complex causal combination of percentage of weeks displaying variance and time required to make adequate progress was almost always necessary and sufficient to predict the type of bifurcation and discharge status, as was the causal combination of percentage of weeks displaying variance and estimated size of the Zone of Proximal Development. DST predicts that the farther and faster a system changes, the more variance it will display. This was evident; cases making larger amounts of progress usually required more time and displayed more variance. Rapid gain was generally accompanied by more variance, while slow gain almost always displayed less variance. This study provides support for some of the predictions of Dynamic Systems Theory and suggests a few modifications, especially regarding cases where the change effort fails.