The function Y = A * X + B must be fitted to the dataset AVLTIME_NLREG.txt. First, the experimental data will be checked for potential outliers that should be discarded because they are outside the "Prediction Band" (PB) with a 95.4% confidence level. If such points exist and the user agrees, a new dataset will be created, preserving the original file and excluding the points located outside the PB (the process is not automatic and the final decision on excluding experimental points rests with the user, not LAB Fit).
Note: The approach proposed by the LAB Fit software is not Chauvenet's criterion, but a Studentized residual test using "Prediction Band", which maintains a fixed 95.4% confidence level while adjusting the critical factor for the sample's degrees of freedom. Experimental points located outside the 95.4% prediction band still have a small probability of being legitimate and should be critically evaluated before exclusion in order to avoid artificial bias.
AVLTIME_NLREG.txt has 48 points. Point number 48 falls outside the "Prediction Band" (at a 95.4% confidence level). Therefore, after an analysis by the user, it could be discarded in a new file called Untitled.txt, as will be shown in the following.