Data Pre‐ and Post‐processing
R. Russell Rhinehart · 2016
Data processing means that the user decided to discard data, adjust data, or create data. In general, these sorts of manipulation are counter to the scientific method and honesty in investigation, but often data processing can be justified. When experimental conditions are changed, continuous processes move through a transient on their way to a steady state. There are diverse opinions related to pre- and post-processing of data. Pre- and post-processing of data is essential to correct and eliminate errors. Steady-state models are frequently used to analyze and design processes and products. If the regression model is used to impute the missing response data values, then there is zero deviation between the imputed data values and the modeled values, which means that imputation has no impact on the sum of squared deviations (SSD). Using the regression model to impute values for missing data is equivalent to excluding those values from the regression.