Common methodological issues and suggested solutions in bone research

Bone research is a dynamic area of scientific investigation that usually encompasses multidisciplines. Virtually all basic cellular research, clinical research and epidemiologic research rely on statistical concepts and methodology for inference. This paper discusses common issues and suggested solutions concerning the application of statistical thinking in bone research, particularly in clinical and epidemiological investigations. The issues are sample size estimation, biases and confounders, analysis of longitudinal data, categorization of continuous data, selection of significant variables, over-fitting, P-values, false positive finding, confidence interval, and Bayesian inference. It is hoped that by adopting the suggested measures the scientific quality of bone research can improve.