Dear Authors,

I have carefully reviewed the revised version of your manuscript, and I must say that it was a pleasure to do so. I am impressed by the thoroughness with which you addressed the feedback provided during the review process. This reflects a strong commitment to improving the manuscript, while also maintaining transparency in managing the changes made to the sections approved during Stage 1. As you mentioned in your response letter, the revisions were primarily minor, but it is evident that you approached them with due care, identifying and resolving any issues with clarity and precision.

The authors have addressed my comments in a convincing manner, and I do not have additional comments. I appreciate the changes they made to the manuscript, particularly reagrding implications for the IAT and the open acknowledgement of the relatively poor model fit. I enjoyed being a part of this review process!

I have finally had the opportunity to review Stage 2 of the article titled “Implicit Ideologies: Do Right-Wing Authoritarianism and Social Dominance Orientation Predict Implicit Attitudes?”.

Firstly, I want to state that the sections revised in Stage 1 have not been modified, and the hypotheses declared then are those reported in the current version of the manuscript. My revisions will therefore primarily focus on unrevised sections.

According to the method, you tested 24 models (12 tasks for the two independent variables). This is also evident from the R script where you estimated the models using a for loop. Therefore, it is unclear why the fit indices are reported for only two models. I strongly recommend clarifying this point.

From the script, it is also evident that you tested the measurement model for RWA and SDO. For clarity, I would report the results of these models, as they provide important details on the quality of the measurements, which could be relevant for contextualizing the results of the structural models. For example, the lower predictiveness of SDO might be due to measurement issues of the construct.

Additionally, you used dynamic cut-offs. As reported in the cited paper, this method can be applied to measurement models in the SEM framework but not to structural models as you did. Moreover, the method is uncommon, and reading the results without contextual information makes understanding difficult. For instance, what are the implications of choosing a cut-off with a low level of misspecification versus a high level? I strongly suggest providing more information on this method to improve reader comprehension. To further refine the comprehensibility of your results, I recommend incorporating the B-H Critical Value present in Table 6 into Table 3.

This comment bridges both the method and results sections. Therefore, I consider it a secondary addition, although it is highly relevant. The question I asked myself while reading the results is: how do implicit and explicit measures correlate for the same pairs of words? This could provide further details on whether explicit and implicit measures capture the same or a similar latent construct.

In conclusion, I would like to thank the authors for the excellent work they have done on this research project, which provides relevant results. While the manuscript is robust in its current form, I believe that addressing the points raised will enhance further its clarity and rigor.