The aim of this interactive R shiny app is to show how nonlinear associations can be estimated with a statistical model using only simple extensions of the basic linear model. Moreover the shiny app allows for comparisons between fractional polynomials, B-splines and natural splines. Parameters and coefficients of each method can be changed and the resulting nonlinear response function is visualized and dynamically updated. In this way, one can explore how the coefficients affect the functional form of the association between an explanatory variable and a continuous outcome variable.
This interactive learning tool addresses all those interested in the estimation of nonlinear associations. It was developed in association with the STRATOS initiative and partly funded by the Austrian Science Funds (FWF).
Link to the shiny app