Comparison of Methods for Addressing Outliers in Exploratory Factor Analysis and Impact on Accuracy of Determining the Number of Factors
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe article aims to provide the reader with a specific guide to using Exploratory factor analysis. And could be pretty good if it was written better.
First, the test of the article needs correction. For example, in line 151 it is written "Recently, authors have proposed the use of an alternative network based approach", but here the author is one, not several. Which authors are we talking about? In general, I believe that the citation should be checked and corrected. It is not possible to write "Velicer [42] ". This author has initials. You need to be respectful of those you quote.
Second, you should equally make (or not make where the word "Where" is in lines 52 and 67, for example) indents. In the text, punctuation marks are missing in many places where they are needed. Spaces do not meet any standards.
The formulas are written mathematically incorrectly. They should be rewritten (see how mathematicians write them). When an index is written, it is always necessary to indicate what values ​​it can acquire, etc.
The tables are not very informative as they are presented. They are very sparse and difficult to read.
Figure 3 is completely uninformative and, moreover, beyond the scope of the display.
The article should be thoroughly corrected.
Author Response
- First, the test of the article needs correction. For example, in line 151 it is written "Recently, authors have proposed the use of an alternative network based approach", but here the author is one, not several. Which authors are we talking about? In general, I believe that the citation should be checked and corrected. It is not possible to write "Velicer [42] ". This author has initials. You need to be respectful of those you quote.
Response
We appreciate the reviewer’s comments and agree that the manuscript needed a thorough editing. This has been done with an eye toward addressing punctuation issues, typos, and writing style. We have highlighted these changes in the revised document. In addition, we have attempted to address each of the specific comments raised by the reviewer.
- Second, you should equally make (or not make where the word "Where" is in lines 52 and 67, for example) indents. In the text, punctuation marks are missing in many places where they are needed. Spaces do not meet any standards.
Response
We appreciate the reviewer’s comments and have given the text a thorough editing in order to identify and address typos, punctuation and spacing errors, and other formatting issues.
- The formulas are written mathematically incorrectly. They should be rewritten (see how mathematicians write them). When an index is written, it is always necessary to indicate what values ​​it can acquire, etc.
Response
We agree with the reviewer that the equations should match the format of those in statistical and mathematical journals. To that end, we have edited them so that they match the format used in the STATS journal. We have also made sure to include definitions of subscripts, as recommended by the reviewer. These changes have been highlighted throughout the revised manuscript.
- The tables are not very informative as they are presented. They are very sparse and difficult to read.
Response
We appreciate the reviewer’s comment and agree that the tables needed improvement in order to make them easier to read. To this end, we have used bold in order to highlight accuracy rates that are 0.90 in each table. We hope that this change, along with improved formatting, make the tables easier to read. If not, we are very open to suggestions from the reviewers and editor in how we might improve the readability of our tables.
- Figure 3 is completely uninformative and, moreover, beyond the scope of the display.
Response
Both reviewers recommended changes to Figure 3, which we agree was necessary. Therefore, we have replaced the barchart with boxplots, per the recommendation of Reviewer 2. We believe that this has improved the manuscript and hope that the reviewers agree.
- The article should be thoroughly corrected.
Response
We agree that the manuscript needed corrections and edits and have endeavored to enact these in the revision.
Reviewer 2 Report
Comments and Suggestions for AuthorsI congratulate the author to the well-designed study.
1)
I have one proposal: as each of the conditions (number of contaminated variables, proportion of contaminated data, etc. ) are repeated 1000 times, it would be nice to see graphical displays (e.g. box plots) instead of only averages in figure 3. It may be very interesting to the reader also to see the variance of the methods. The refrence line and the mean can also be included in the boxplot.
2)
The empirical example is also using simulated data. It would be nice to have one or two real examples, may be one with a small and one with a medium sample size to demonstrate the methods.
Author Response
I congratulate the author to the well-designed study.
Response
We thank the reviewer very much for their kind words.
1) I have one proposal: as each of the conditions (number of contaminated variables, proportion of contaminated data, etc. ) are repeated 1000 times, it would be nice to see graphical displays (e.g. box plots) instead of only averages in figure 3. It may be very interesting to the reader also to see the variance of the methods. The refrence line and the mean can also be included in the boxplot.
Response
We agree with the reviewer that a boxplot would provide more information than did the barcharts. Thus, we have taken the reviewer’s suggestion and included boxplots to display the results. We believe that this graph improves the display of these results.
2) The empirical example is also using simulated data. It would be nice to have one or two real examples, may be one with a small and one with a medium sample size to demonstrate the methods.
Response
We agree with the reviewer’s suggestion and have included results for the actual data upon which the original simulation was based. We have used the original sample of 260 individuals and then an example using a separate sample of 107 different people. The Empirical Example section has been completely rewritten to reflect these changes. We appreciate this suggestion by the reviewer and believe that these changes have improved the quality of the manuscript and hope that the reviewers and editor agree.
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsI am not satisfied by punctuation singing here.
For editor opinion
