Thank you for your observant comments on our study. We agree that the terms “accuracy” and “reliability” describe 2 distinct issues; therefore, each stands alone in the context of the scientific method. Accuracy is the degree of closeness of measurements of a quantity to that quantity’s actual value, whereas reliability, related to repeatability and reproducibility, is the degree to which repeated measurements are the same under unchanged conditions. However, we disagree with the notion that using the same statistical method to address 2 issues implies that the 2 issues are being confused. Many times in research studies, several separate questions are answered with the same statistical method (eg, the t test).
In the present study, we used the Bland-Altman method to answer the questions of accuracy and reliability. Several statistical methods may be applicable in a given situation, so we take issue with the idea that using the Bland-Altman method to assess reliability is a “methodologic mistake.” Not only has this method been described by its inventors to answer the question of reliability in numerous publications, but it also does so in a scientifically meaningful manner. In particular, bias represents the average measurement difference between 2 raters. A zero bias indicates that the raters obtain, on average, the same measurements; this corresponds to a good outcome for interrater reliability. In addition, the limits of agreement quantify the variability of the measurement differences. Wide limits of agreement indicate poor interrater reliability, whereas narrow limits of agreement, together with a small bias, indicate good interrater reliability. As such, the Bland-Altman method summarizes interrater reliability on the original scale (millimeters), which is not the case for intraclass correlations.
When we conducted our study, we evaluated some of the more popular digital model types marketed to the orthodontic community today. Whether we like it or not, these new technologies are available and being used by an increasing number of practitioners. Therefore, the new technologies need to be critically appraised with regard to their efficacy and efficiency. In this way, the body of knowledge is advanced in evidence-based orthodontics. When we concluded that dental measurements taken on digital models can be as accurate as, and may be more reproducible and significantly faster than, those taken manually on traditional plaster models, we based these conclusions on solid research data. We made it perfectly transparent how these data were acquired and would welcome any similar studies to determine whether our findings are reproducible. Unless evidence is provided to the contrary, we stand by our conclusions and would not call them misinterpretations or misconceptions.