Reliability and Validity of Optoelectronic Method for Biophotonical Measurements

摘要

Reliability and validity of measurements is of utmost importance when assessing measuring capability of instruments developed for research. In order to perform an experiment which is legitimate, used instruments must be both reliable and valid. Reliability estimates the degree of precision of measurement, the extent to which a measurement is internally consistent. Validity is the usefulness of an instrument to perform accurate measurements of quantities it was designed to measure. Statistical analysis for reliability and validity control of low-coherence interferometry method for refractive index measurements of biological fluids is presented. The low-coherence interferometer is sensitive to optical path difference between interfering beams. This difference depends on the refractive index of measured material. To assess the validity and reliability of proposed method for blood measurements, the statistical analysis of the method was performed on several substances with known refractive indices. Analysis of low-coherence interferograms considered the mean distances between fringes. Performed statistical analysis for validity and reliability consisted of Grubb's test for outliers, Shapiro-Wilk test for normal distribution, T-Student test, standard deviation, coefficient of determination and r-Pearson correlation. Overall the tests proved high statistical significance of measurement method with confidence level > 0.0001 of measurement method.