Area under the Curve as an Alternative to Latent Growth Curve Modeling When Assessing the Effects of Predictor Variables on Repeated Measures of a Continuous Dependent Variable

Round 1

Reviewer 1 Report

REVIEW

Title of the paper: Area Under the Curve as an alternative to Latent Growth Curve Modeling when assessing the effects of predictor variables on repeated measures of a continuous dependent variable

Manuscript Number: stats-2409755

General conclusion: Major Revision.

 

Comments

After carefully reading the proposed paper, this paper contains an interesting proposal; my overall impression is that the manuscript presents some results that could be useful in practice. I have a good opinion about this work and recommend its acceptance after addressing the following aspects:

My comments are:

1.     Some keywords should be added.

2.     The Abstract is very general. It is necessary to mention a brief description of the content of the manuscript in a clear and concise manner so that the reader can understand the content of the manuscript.

3.     In Figure 1, where is the caption of the figure?. The caption of the figure should be added. In addition, the quality of the figure is very poor, this figure should be saved in eps format.

4.     Equation 1 is numbered twice.

5.     The first letter of each word in all sections, subsections, and subsubsections should be in a uniform format as either a capital letter or a small letter throughout the manuscript.

6.     The author writes some words in a bold format and then the same word not bold, for example see Table 3 and Table 4.

7.     The quality of figures 2 and 3 is very poor, these figures should be saved in eps format.

8.     The algorithm which is used in Monte Carlo simulations should be added to the paper.

9.     In table 1, the value of kurtosis is negative, what is the meaning of this?

10.  In table 1, who the MSE and Bias are computed?

11.  In references, the DOI should be added in all references, and the reference format must be unique.

 

 

 

Moderate editing of English language is required.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

Dear colleague,

Congratulations on the article. It is exciting!

Only one change would be necessary - at least, the version that reached me does not respect the model imposed by the magazine editors. Please use that template.

I wish you success!

Author Response

Reviewer 2

Comments and Suggestions for Authors

Dear colleague,

Congratulations on the article. It is exciting!

Only one change would be necessary - at least, the version that reached me does not respect the model imposed by the magazine editors. Please use that template.

I wish you success!

Response: Thank you very much. I believe that the latest version of the manuscript is now in the correct format.

Reviewer 3 Report

An alternative regression-type model for explaining the effects of covariates on a continuous outcome in a longitudinal framework (with repeated measures) is proposed. The method hinges on the radically new idea of using areas under the outcome-time curve (AUC) as a new outcome, and then using this AUC as the outcome in a standard regression framework. To be fair, the authors are not actually proposing this (has already been done elsewhere), they are merely proposing two different ways to measure teh AUC. Some simulation studies are then carried out to compare with latent growth curve models (LGCMs), which are routinely used for these kinds of data.

Major Points
------------
(i) This is a radical idea, which took me some time to absosrb.... I think Sec 2 should explain pictorially by giving a simple example, how the AUC actually has anything to do with the outcome! This is not obvious to me... Please show (state?) how larger values of the outcome result in larger values of AUC, etc. (I understand the author is NOT proposing this for the 1st time, but still, some "selling" of this idea should still take place.)

(ii) Monte Carlo Simulations: The author does one study for the LGCM and one study for the AUC, separately. But what one really wants to see is if AUC is better at estimating the effects of covariates that are actually present in LGCM! This is after all what is being sold, correct? Please discuss/do this, if it makes sense; else state why this is impossible or would not be a fair comparison.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Thanks for the good response to all the comments.

Minor editing of English language required

Author Response

Thank you kindly for your comments. I went through the manuscript again and made additional edits to ensure the English language is correct. 

Reviewer 3 Report

The bolding in Table 5 needs to be revised, a few entries are incorrectly bolded (they are not the highest).

Author Response

Thank you very much for your comment. I went over table 5 and changed all of the wrongly bolded entries.