Calibration method choice by comparison of model basis functions to the theoretical instrumental response function

Abstract

Sorting through the large array of calibration methods available for first and second order calibration is often a daunting task for initiates into the field of chemometrics. Justifying the selected method as the most appropriate one is even more difficult. Presented here is a justification for calibration method selection based on matching the model employed in the calibration method with the instrumental response function. This is applied to the disparate types of nonlinearities found in both first and second order calibration. Matching the calibration method to the instrumental response function is employed to parse the decision making process for choosing between branches in the first order parsimony tree. The different types of nonlinearities present in second order data and their implications on calibration model selection are discussed.