Bayesian statistical modelling and occupational therapy research

In Japanese occupational therapy, Bayesian statistical modelling attracts attention.

The advantage of Bayesian statistical modelling lies in the flexibility of data analysis.

Occupational therapy research explores complex themes of the human as an occupational being.

Such complicated phenomena are difficult to handle with general statistics and occupational therapy researchers face the problem of reduced accuracy of data analysis.

On the other hand, Bayesian statistical modelling can build a statistical model suitable for complicated phenomena, so it can be expected to raise the quality of occupational therapy research.

Bayesian statistical modelling software includes Stan, JAGS, WinBUGS and others.

My recommendation is Stan.

Stan can flexibly execute hierarchical Bayesian model such as mixed model,  state space model, in addition to the t-test, correlation analysis, χ 2 test, ANOVA, structural equation model, etc.

I think many occupational therapy researchers should utilise Bayesian statistical modelling.