Prediction Intervals using Classical and Inverse Regression for the Simple Linear Calibration Problem with Lack of Fit (2010)

Location: NASA Langley Research Center
Role: Lead Developer
Languages Used: R
Problem: Because no instrument is exact, it is necessary to know the differences between what is measured and the actual data value. It is also necessary to know the consequences of ignoring or estimating that difference.
Solution: To address these issues, researchers constructed a simulation study that fit a linear model to data generated by a quadratic model. My task was to write the code that would conduct the experiment.