ABSTRACT
In most, if not all, scientific experiments and analyses, the researcher must summarize, interpret and/or visualize the data to gain insights, to search for patterns, and to make inferences. The main goal of smoothing from an EDA point of view is to obtain some insights into how data are related to one another and to search for patterns or structure. Loess is a locally weighted regression procedure for fitting a regression curve by smoothing the dependent variable as a function of the independent variable. It is well known in regression analysis that checking the assumptions made about the residuals is important. This is equally true when applying any smoothing technique, and people can use similar diagnostic plots. Regression splines use a piecewise polynomial fit to model the relationship between a predictor variable and a response variable. These spline models can provide a way to adequately model the relationship between the variables, while avoiding negative consequences.
