Reject (Paper is not of sufficient quality or novelty to be published in this Journal)
In this paper, the authors propose a new test procedure for testing agreement under a bivariate normal set up. They compare this new test procedure with other three known test procedures via simulation, and include an empirical application with DCLHb data of 299 patients collected by two different methods.
In my opinion, the contribution of the paper is very limited and the quality of this paper needs to be further improved.
The introduction is very similar to previous work of one of the authors (P. Yimprayoon) and others authors (Lin et al. 2002; Sinha, 2012). The authors don’t overview the main advances in this field, they don’t cite the relevant papers related, and they don’t determine clearly the specific question, hypothesis or aim of their study.
In Material and Methods section, the new test procedure is described directly with the equations (16)-(27). The authors should widen the description of that procedure and include the corresponding proofs.
In Power Computation section, the authors justify a new fourth method (the new method proposed), against other three known methods, with a simulation study only based on small samples sizes, and where the new method not always outperforms the other three methods. In addition, the authors don’t include larger samples sizes in the simulation study, for example, n=300 (which is the sample size of the application included in the paper).
In Numerical Results section, the authors don’t test that the data follow a bivariate normal distribution, as a previous step to consider the dataset adequate for this study. Finally, the authors “accept” the null hypothesis H0 for all test procedures, but null hypothesis are never accepted - null hypothesis can be o cannot be rejected at a significance level.