Dear Authors / Dear Editor,

Thank you for considering my comments with care. I believe the revision have improved the clarity and the rigor of this RR, which was rather strong in the first place. I am happy to see the planned study carried out and very intrigued by the results!

Sincerely,

Alex Ten

The authors have addressed all of my comments and I am happy with the current Stage 1 submission.

The review is organized according to PCI RR reviewer guidelines. Some of the comments may not fully fit the section-headers/questions, but I added them just to try to aid the authors with the next version of the manuscript, should they and the editor decide to resubmit.

I) Does the research question (RQ) make sense in light of the theory or applications? Is it clearly defined? Where the proposal includes hypotheses, are the hypotheses capable of answering the research question?

Yes, the RQ makes perfect sense. The title, the abstract, and the introduction suggest that the study aims at elucidating the causal pathway between curiosity and memory. The authors set out to clarify whether the state of curiosity affects memory above and beyond its motivational effect active exploration. This would suggest that curiosity reflects a physiological state in which mnemonic encoding is more powerful or efficient. On the other hand, curiosity might enhance learning — not through priming the memory system for incoming information — but by energizing more thorough exploration. In this case, curiosity would only improve memory to the extent that the learner is able to act on it.

This seems to me like a perfectly valid research question, as I cannot convincingly argue against any of the two alternatives. The novel experimental treatment (passive group) is appropriately introduced to address the research question. However, hypotheses and the corresponding analysis approaches need to be developed a little more (see section III below).

One minor comment is to adjust the prioritization of study goals in the pre-registration (as well as the subsequent report). The manuscript often shifts the spotlight from the novel and interesting research question to replication of old results. It would be easier to assess the proposed hypotheses and analyses if they clearly and saliently addressed the main research question.

II) Is the protocol sufficiently detailed to enable replication by an expert in the field, and to close off sources of undisclosed procedural or analytic flexibility?

The protocol seems sufficiently detailed to enable replication by an expert in the field, but this job could be further simplified. The authors could specify their models in more detail to aid the reviewers’ assessment. What priors are planned to be used and why? What are the likelihood functions? Are Likert-scale variables treated as continuous (if so, why?) or ordinal-categorical variables? What group would be considered the reference group? It would be great to see conventional formal model specifications (and perhaps some DAGs).

III) Is there an exact mapping between the theory, hypotheses, sampling plan (e.g. power analysis, where applicable), preregistered statistical tests, and possible interpretations given different outcomes?

I would like to comment on several potential issues concerning the correspondence between the RQ, proposed hypotheses and methods of analysis.

Hypothesis 2 is simple and straightforward. However, the model addressing the hypothesis includes several variables serving the replication purpose rather than directly testing the hypothesis. I don’t think it is necessarily wrong, but it would be helpful if the authors provided theoretical or logical justifications of including these variables in the model (that addresses hypothesis 2).
Paragraph 1B from section 2.1 of the PCI RR Guide for Reviewers explains that “The inclusion of hypotheses is not required– a Stage 1 RR can instead propose estimation or measurement of phenomena without expecting a specific observation or relationship between variables.”. I believe that Hypothesis 3 of the manuscript, which the authors label as “nonspecific predictions”, should be treated not as a planned hypothesis-testing procedure, but as an exploratory analysis, because there is no hypothesis (i.e., a theoretically derived prediction). The authors do provide possible interpretations of different outcomes, but these seem more like ad hoc rather than theoretically inspired explanations.
An alternative way of treating this problem is to list 3 different hypotheses and explain which modeling outcomes would support / refute each.

Should the authors and the editor find my assessment mistaken, it might help (to avoid this kind of issue in the future) to unambiguously list the hypotheses in one place in the manuscript and connect the modeling / analyses part more directly to that section by stating which results (e.g., a significantly positive coefficient of X on Y) would support or refute a given hypothesis. It would also be helpful to provide an interpretation of all coefficients of variables included in the models.

Power analysis
This is a minor comment. While the research question is clear, valid, and appealing, a several planning decisions seem to subserve the replication of a previous study (Cen et al., 2024), rather than addressing the answering the main question.

One example is that it seems like the reported power analysis was conducted to determine the sample size for a test that would replicate test results from Cen et al. (2024), but given the novel research questions that aim to expand on this work, focusing the power analysis on test replications seems a bit odd. Furthermore, it would be useful to read more details about how and why the specific power analysis technique works, especially given the plans to perform Bayesian parameter estimation (as opposed to null-hypothesis significance testing). The information provided about the power analysis is very limited, but from what is given, I am wondering why did the authors opt to use normal distributions rather than the posteriors (which are not guaranteed to be normal) from the previous study?

For proposals that test hypotheses, have the authors explained precisely which outcomes will confirm or disconfirm their predictions?

Yes.

Is the sample size sufficient to provide informative results?

To be determined by further planning.

Where the proposal involves statistical hypothesis testing, does the sampling plan for each hypothesis propose a realistic and well justified estimate of the effect size?

Probably. However, only a single power analysis has been conducted to determine the sample size for 80% power of a previously observed effect size (see section III-3 above).

Have the authors avoided the common pitfall of relying on conventional null hypothesis significance testing to conclude evidence of absence from null results? Where the authors intend to interpret a negative result as evidence that an effect is absent, have authors proposed an inferential method that is capable of drawing such a conclusion, such as Bayesian hypothesis testing or frequentist equivalence testing?

The main hypothesis (that curiosity enhances memory regardless of the ability to actively explore) can be supported by demonstrating the absence of an interaction between group (active vs passive) and curiosity. This would be tricky to demonstrate with a simple HDI-based test.

Have the authors minimised all discussion of post hoc exploratory analyses, apart from those that must be explained to justify specific design features?

Yes.

Have the authors clearly distinguished work that has already been done (e.g. preliminary studies and data analyses) from work yet to be done?

Yes.

Have the authors prespecified positive controls, manipulation checks or other data quality checks?

Yes, the authors plan to closely replicate the active-exploration group from a recent study by Cen et al. (2024) as a positive control for this study.

However, the authors do not seem to plan any manipulation checks for the novel treatment (passive viewing). They do acknowledge a potential limitation of this manipulation: participants can still exert active control (e.g., eye movements) in the passive-viewing condition, but no plans to address this limitations are provided. Nevertheless, this seems like a very minor issue since participants in the active group are also allowed to control their eye movements.

When proposing positive controls or other data quality checks that rely on inferential testing, have the authors included a statistical sampling plan that is sufficient in terms of statistical power or evidential strength?

See section III-3.

Does the proposed research fall within established ethical norms for its field? Regardless of whether the study has received ethical approval, have the authors adequately considered any ethical risks of the research?

The proposed research does not seem to involve any violations of ethical norms. However, I am no expert in ethics in cogsci research and the authors do not explicitly discuss any ethical considerations in the planned study.

​​​​​