Pain and physical activity for one individual

Aims: there is increasing awareness that for effective patient care we need more than only randomized controlled trials with groups of participants and carefully collected single case (N = 1) data have several important advantages over traditional group-level studies. With the advance technology, collecting relevant from a becoming easier by day, this offers tremendous opportunities understanding how behaviors displayed an individual can be influenced one or key variables. For example, pain experienced influences amount time spent on physical exercise. Method: using publicly available observational data, five models are compared: classical ordinary least squares (OLS) linear regression model; dynamic model (DRM); two-level random-intercepts (2LRI); continuous covariate first-order autoregressive correlation (CAR1); trend (OLST). These compared in terms overall fit statistics, estimates relation between activity (response variable interest) (covariate interest), residual statistics. Results: 2LRI outperforms all other both provides relative extremes provided models. CAR1 OLST demonstrate almost identical performance substantially better OLS – which performs worst DRM. Conclusion: DRM, CAR1, OLST, account serial typically present somewhat different ways under assumptions, perform OLS. Implications these findings observational, quasi-experimental, experimental studies discussed.