Testing the compatibility of measurement methods in the analysis of cephalometric radiographs

Introduction. The comparison of the results obtained by different methods has been the subject of many scientific publications over the years. A number of discussions are related to statistical analyses used to test the compatibility of the results obtained by different research techniques. There has been no consensus on the choice of the appropriate statistical method, and so the issue of the compatibility of the results still constitutes the area for further investigation related to the application of methods which would constitute reliable tests indicating a higher precision of their compatibility character. Aim of the study. To test the compatibility of two methods of cephalometric measurements using three comparative methods of diverse sensitivity for the assessment of the compatibility of the tested results. Material and methods. In the first method, cephalometric radiographs were traced and measured manually with trace foil. In the second method, landmarks were identified by a mouse-driven cursor on the screen. Twelve values were measured for randomly selected 14 subjects. Pearson’s correlation coefficient, the Mann-Whitney U test and the Wilcoxon test were used to assess the compatibility character of both methods. Results. Both correlations at the significance level of 0.05 and the Mann-Whitney U test did not show statistically significant differences between both methods. Statistically significant differences were obtained for 9 of 12 measured values in the Wilcoxon test at the same level of significance. Conclusions. Correlations and tests based on the comparison of the mean value are not sufficiently sensitive and they do not indicate statistically significant differences between the results.