Advancing Methods for Studying Complex Systems: Moving Beyond Linear Thinking

摘要

Life is complicated. Some variant of this adage is uttered by everyone who walks this earth, but scientific methods have not always matched the complexity of their subject. Our study designs are largely reductionist.1 Our experiments deduce causality by manipulating variables and measuring the results. Our conceptual models are often linear; we assume that outcomes (dependent variables) are products of the factors (independent variables) we manipulate. This linear thinking dominates medical research. Industry and government spend billions of dollars on clinical trials. The results form the basis for regulating the nation's drugs and inform our assumptions about the effect size of interventions and the probability that they will occur. For decades, the RCT has been the gold standard in the hierarchy of evidence.2 But does this information tell the whole story?