To test the performance of this method, we compared the WRS to observed-to-expected ratios, a common

parametric statistic for physician cost-efficiency analysis, in terms of stability across time and across trimming methodologies.

This reviewer wonders if the significance could have been enhanced by evaluating the data at a higher level of measurement and using a

parametric statistic such as a t-test.

In other words, the statistical tests they used are appropriate for

parametric statistics (means) but not appropriate for non-parametric statistics (medians).

Parametric statistics are based on the theory of a normal distribution.

3)

parametric statistics are robust to modest violations of normality (non-equality of variances, samples from non-normally distributed populations) and thus can be used with non-normal distributions, as long as the normality violations are not excessive.

All data were subjected to homogeneity (Hartley) and normality (Shapiro-Wilk) tests to verify the assumptions of

parametric statistics. Data were subjected to analysis of variance (ANOVA), and means were compared by the Tukey test at 5% probability, using the software SISVAR 5.4 (Ferreira 2011).

The processing steps are briefly explained below as follows: (1) estimate and write: the images were bias-corrected and segmented into GM, WM, and CSF; (2) DARTEL create template: a customized template was created for our study; (3) DARTEL existing template: once the study-specific template was created from the above step, the remaining subjects were registered nonlinearly to this template using DARTEL existing template module; (4) normalize to MNI space; (5) smooth: after normalizing and registering all subjects to MNI space, the resulting images were modulated (without including affine component) and smoothed using a full-width half-maximum (FWHM) of 8 mm; and (6)

parametric statistics: the smoothed images were used for statistical inference.

Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were log transformed for

parametric statistics. Post-hoc analysis was done by the Bonferroni correction.

Data in this distribution are interval/ratio-level data that are usually assessed by

parametric statistics. However, legitimate reasons exist for using nonparametric versus

parametric statistics, particularly with the small sample size and distribution.

It should be noted that many biological early warning systems, built on the speed analysis of the test objects, use

parametric statistics, excluding the nature of the data distribution.

Didactic Engineering is characterized as being a qualitative action research method, and rejects the Classical

Parametric Statistics method, the control case or experimental groups and control groups.