Abstract

New visual testing devices that do everything from precise optical measurements to subjective measuring procedures are being introduced at an accelerating rate. In all cases, empirical data must be compared to show the relationship between measurements from different instruments. These comparisons typically use two procedures: correlation and regression. A correlation coefficient demonstrates how reliably measurements from one instrument can be used to replace those of another by providing an index of the strength of the relationship between the two. If the relationship between two instruments is nonlinear, it may be necessary to transform one or both sets of data to linearize their relationship .When this is done, a correlation coefficient is an appropriate and essential statistical test for comparing the data from two instruments.

© 2000 Optical Society of America