Revisiting ‘less is more’: A failure to replicate the association between increased knowing and decreased liking

I am grateful for the authors for their detailed and thorough responses. I am satisfied with the actions taken and have no more substantive feedback.

All requested changes have been made, and I'm happy to accept the paper as it is!

This is a Stage 2 registered report. After the extensive review of the Stage 1 manuscript, there is little to add here. The report seems to be very well written, and the study conducted according to the protocol. This study is a gemstone of the reforms in our field, highlighting the best practices related to study registration, transparent reporting, and reproducibility via data and material sharing. The discussion is fair and extensive, not overreaching. Here are a few minor points that I think might improve the manuscript: 

-            It would be great to see a deviation report: a section describing the deviation from the preregistered protocol. If there were not deviations and the protocol was followed exactly, that should be written in the deviation report.

-            A note about missing data (or the lack thereof) would be useful.

-            “We conducted additional analyses and again found no indication for variations in liking depending on the number of traits presented.” – It would be good to present some more nuance here. I agree with the authors that the results do not support the less is more effect in a linear fashion, but they found a pairwise difference (4 traits more liked than 8) which was in the expected direction at least. I think it is worth an honorary mention, even though the results overall still don’t support the previous findings.

As far as I can tell, the Stage 2 report does not deviate from the registered procedure (2C, 2D). The tracked-changes manuscript also shows that no earlier section of the paper was changed (2B). As the data were collected as planned, I think that they are perfectly able to test the hypotheses as preregistered (2A), with all major conclusions being being supported by the evidence (2E).

I have also downloaded the data and code from OSF and run it on my machine. I don’t primarily work in R, so apologies if this is mistaken, but I think the code as it stands now doesn’t properly install the medmod package from Github. However, after adding a few lines of code doing just that, the replication package runs perfectly otherwise. I then went ahead and did a spot check of a number of results from the paper: All results that I checked properly reproduced!

I don’t have any additional concerns about the paper, expert perhaps in the discussion section, where the following paragraph seems to make quite far-reaching conclusions that are not directly supported by the data (in fact they could not be, given the claims that are being made). I understand this is typically done in discussion sections but I’d suggest cutting this paragraph altogether, the paper would be stronger without it: “Correcting biases such as the person positivity bias and false consensus effect may be difficult, particularly if they are unconscious. However, provided that these biases do influence our attitudes and behaviors towards others, simply having an awareness of them may be helpful in avoiding the resulting cognitive dissonance and consequent avoidance behavior during early stages of impression formation. To this extent, knowledge of these biases may help foster more meaningful connections between dissimilar individuals, facilitating opportunities to bridge divides between social groups.”

Other than that, I want to congratulate the authors on a strong replication attempt, and I’m happy to accept the paper after a small set of minor revisions found below:

(Very) Minor points: 

In Table one, it should be ‘Signal - X’ and not ‘Signal- X’.

The ‘Holm method’ is named after a person (Sture Holm), so the ‘H’ should be capitalised.

The name Zöe is spelled both Zöe and Zoe at different parts of the paper.

In the CRediT table, Ashleigh’s last name is not present while it is present for all other authors.

“We found no indication of order effects, and reported” should be “and report”

It should be “PhD students at the University of Kent”, not “at University of Kent”

“Consequently, our findings challenged” should be “challenge”

“supported previous research suggesting” should be “support”

“extending on their research” doesn’t quite work, it should be either “extending their research” or “building on their research”

“(n = 153; χ(1) = 305.9, p < .001, h = .67 95% CI [0.57, 0.76]).” Is missing a comma prior to “95%”

“Since we found no support for the effect on perceived similarity, we therefore concluded no support for dissimilarity cascades in impression formation” is a bit ungrammatical, it could be rephrases as either “we therefore concluded ‘no support’” or something like “we therefore concluded that there was no support”.

At several places throughout the paper, you use “Norton et al ‘s” when it should be “Norton et al’s”.

“lower levels of liking(Study 2).” has a space missing prior to (Study 2).

“Interestingly, an identical effect was found for liking, whereby liking increased between traits one and five, but plateaued from the fifth trait onwards” This is one of the few (or the only) passive constructions in the paper, I suggest making it active, i.e., “we found an identical effect for liking”.

“We found the same for each comparison (1 vs. 2 traits, 2 vs. 4 traits, 3 vs. 6 traits, 4 vs. 8 traits and 5 vs 10. traits; χs > 25.6 , ps < .001).” should be “vs.”, not “vs”.

LeBel et al. is referred to as both (2018) and (2019), with the correct citation being (2018), https://journals.sagepub.com/doi/10.1177/2515245918787489.

“Study 2: Perceived similarity and linking by condition” should be “liking by condition”, I assume.

In Table 1, “0.14 [-0.03,0.30]” is missing a space.

“We found that more people thought they like a person more the more they know more about them” should be “would like” or some similar construction.

“; t(329.4) = 1.67, p = .097, d = 0.14, 95% CI [-0.03,0.30]).” is again missing a space at the very end.

“https://doi/org/10.1037/0033-2909.116.1.75” does not work

“Zhu, M., & Feldman. G. (2023). Revisiting the links between numeracy and decision making: Replication Registered Report of Peters et al. (2006) with an extension examining confidence. Collabra: Psychology, 9(1), 77608. https://doi.org/10.1525/collabra.77608 “ there should be no period after “Feldman”.