In Vision It Is Groups, Rather Than Maps, That Determine How We Perceive the World
Round 1
Reviewer 1 Report
In this paper the author has compared the Boolean Map (BM) Theory to the Gestalt theory if visual dimensions, such as color and form, are processed 9 separately in early vision.
The manuscript is well written, the literature was correctly cited, and the aim was interesting. I appreciate the different predictions based on the two models. However, some points need a minor revision to reach publication level.
The main point is in the analyses section. I suggest adding a “statistical analyses” section in which all analyses performed will be explained and all methods reported. I don’t understand why the authors don’t have performed an overall ANOVA but they have performed a control and critical condition. I understand the difference between conditions, but, at first the overall 5x3 rmANOVA is mandatory. See the reference. In the analyses section the amount of correction need to be detailed (in some cases Bonferroni correction was applied wrongly).
Minor:
Move the description of critical condition from introduction to methods “Stimuli section”.
An example of presentation could help in understanding stimuli, their dimensions, and their position.
In line 211-213 please specify the dimension of stimuli
Based on the procedure paragraph, do the participants perform the task on their PC?
How were interferent conditions handled?
Is the HW adequate for conducting the experiment?
More importantly, how was the distance, the screen size and resolution of the display used for presentation? This is an essential point for the psychophysical dimension of the stimuli since they could potentially change a lot between a 13” 2k screen and a 24”1440x900 screen.
Create a statistical section in which all analyses were detailed
Why did the authors conduct two experiments if the results were similar and no particular difference were given to the observer.
An analysis could be used to compare them. The comparison of the overall results could not provide specific conclusions.
Reference should be formatted in text in a sequential order using numbers in a square bracket [1].
References
Nieuwenhuis, S., Forstmann, B. U., & Wagenmakers, E. J. (2011). Erroneous analyses of interactions in neuroscience: a problem of significance. Nature neuroscience, 14(9), 1105-1107.
Author Response
- The main point is in the analyses section. I suggest adding a “statistical analyses” section in which all analyses performed will be explained and all methods reported. I don’t understand why the authors don’t have performed an overall ANOVA but they have performed a control and critical condition. I understand the difference between conditions, but, at first the overall 5x3 rmANOVA is mandatory. See the reference. In the analyses section the amount of correction need to be detailed (in some cases Bonferroni correction was applied wrongly).
Response to 1. The Critical conditions are the key conditions that are central to rationale of the paper. As we state in the Introduction, the Control conditions were included for completeness. In all of the Critical conditions the displays contained two shapes and in the Control conditions there is only one. The control conditions address one question – whether a color difference included in the display affects the time to count the shapes. Comparisons across the Control and Critical conditions are irrelevant to the rationale of the study.
Although it is of course possible to carry out an ‘omnibus’ ANOVA in the way suggested, this is not mandatory. We ask different questions of the data from the Critical and Control conditions, and we answer these with separate ANOVAs in the ways described. The Bonferroni corrections reported in the paper are the standard family-wise corrections implemented in SPSS.
Minor:
- Move the description of critical condition from introduction to methods “Stimuli section”.
Response to 2. We feel that the rationale of the experiment is best conveyed in terms of the description of the displays hence the displays are described in general terms in the Introduction to the paper.
- An example of presentation could help in understanding stimuli, their dimensions, and their position.
- In line 211-213 please specify the dimension of stimuli
- Based on the procedure paragraph, do the participants perform the task on their PC?
Response to 3 – 5.The physical description of the displays is conveyed in terms of pixels. This is because the experiment was run online and each participant used a different computer. The absolute dimensions of the stimuli would likely vary across participants, but the relative dimensions would be constant. The positions of the items were determined randomly on a trial-by-trial basis.
- How were interferent conditions handled?
Response to 6. The experiment was run online and participants were free to run the experiment on whatever computer they had access to and also they were free to choose when and where to test themselves. We have now provided additional clarification of this on p. 5.
- Is the HW adequate for conducting the experiment?
Response to 7. We apologise but we aren’t sure what is meant by this question.
- More importantly, how was the distance, the screen size and resolution of the display used for presentation? This is an essential point for the psychophysical dimension of the stimuli since they could potentially change a lot between a 13” 2k screen and a 24”1440x900 screen.
Response to 8. This is of course true, but we take it to be irrelevant to the effects reported in the paper. It is quite possible that such inter-participant differences in testing conditions should add noise to the measurements, but the robustness of the findings is clear and transcend any variations across the different testing conditions.
- Create a statistical section in which all analyses were detailed
Response to 9. We prefer to report the statistical results in the manner commensurate with standard practice.
- Why did the authors conduct two experiments if the results were similar and no particular difference were given to the observer.
Response to 10. The two experiments show the same key results but vary according to the display duration. Of course, we could report just one experiment (because the results are so similar) but given concerns about replication we feel that both sets of data establish the generality of the results.
- An analysis could be used to compare them. The comparison of the overall results could not provide specific conclusions.
Response to 11. We do not feel that such an analysis would add substantially to the content of the paper and would possibly act as an unnecessary distraction.
- Reference should be formatted in text in a sequential order using numbers in a square bracket [1].
Response to 12. We have made this stylistic change and thank the reviewer for this set of very helpful comments.
Reviewer 2 Report
This paper aims to test whether visual dimensions, such as color and form, are treated separately. The experiment shows that the performance is better when the color / shape discrimination is compatible than when it is incompatible. This data strongly supports the prediction of grouping theory. It is further concluded that a major constraint in vision is the number and kind of groups recovered, rather than the number of functional mappings used. This is an interesting research paper. There are some suggestions for revision.
1） The abstract of this manuscript is weak. The language organization of the abstract has no hierarchy and lacks logic. Otherwise, the abstract does not make clear what the main work of the paper is and what achievements have been made in the research. Authors should reorganize the language.
2） The motivation is not clear. Please specify the importance of this paper.
3） Please highlight the contributions of this paper.
4） In this paper, BM theory and grouping theory are compared. These two theoretical concepts should be elaborated in detail and further compared.
5） It is strongly recommended that authors add relevant logical descriptions under the figures to make the article more readable.
6） In addition to the comparison between BM theory and grouping theory, it should also be compared with other theories to further confirm the conclusions of this paper.
7） Please explain how to count objects in detail.
8） More technical details should be given.
9） When analyzing the experimental results in detail, it is suggested that authors analyze the role of principle rather than stay at the result level.
10） The experimental results are not convincing. More related experimental results should be given.
11） The literature is out of date and there is no research result published in recent two years. Please discuss more recently published solutions.
Author Response
Response to 1. We have edited the Abstract accordingly.
2） The motivation is not clear. Please specify the importance of this paper.
3） Please highlight the contributions of this paper.
4） In this paper, BM theory and grouping theory are compared. These two theoretical concepts should be elaborated in detail and further compared.
Response to 4. Large tracts of the text clearly spell out in detail the nature of the two accounts.
5） It is strongly recommended that authors add relevant logical descriptions under the figures to make the article more readable.
Response to 5. There was a formatting error in the previous submission that meant that the figure legends had been incorporated into the main text – this has now been rectified.
6） In addition to the comparison between BM theory and grouping theory, it should also be compared with other theories to further confirm the conclusions of this paper.
Response to 6. The paper deals with the key theoretical ideas that address the main thrust of the paper. These are discussed in terms of the most relevant extant theories namely, FIT, BM theory, grouping theory and a dimensional weighting account. Nonetheless, our central aim is to test predictions of BM theory in a novel way.
7） Please explain how to count objects in detail.
Response to 7. Without further specification we aren’t sure what is meant by this.
8） More technical details should be given.
Response to 8. We have provided what we feel are all the relevant technical details in the paper.
9） When analyzing the experimental results in detail, it is suggested that authors analyze the role of principle rather than stay at the result level.
Response to 9. Sorry, we aren’t sure what the referee means by this.
10） The experimental results are not convincing. More related experimental results should be given.
Response to 10. We feel that we have described robust effects that replicate over two experiments. Without further details as to which other experiments we should have enacted it is impossible to comment further. We take it that the data from the experiments address directly questions as posed in the Introduction. The data do adjudicate between BM theory and grouping theory and we build a case (that we hope shows) that the data are compelling.
11） The literature is out of date and there is no research result published in recent two years. Please discuss more recently published solutions.
For clarity and brevity, we focus on a set of key ideas and document those with respect to the most relevant literature. We cite Huang (2020).
Reviewer 3 Report
I have reviewed this manuscript. The overall content of this manuscript is very weak.
- Authors should revised the abstract clearly including Background of this study, Aim, Methods, Results, Conclusion and Significance.
- Authors should re-write the Introduction clearly divide by three paragraphs:
-Background and definition about the Vision.
-Previous published research about this topic with their limitations.
-How to address the previous research literature limitations/ lack/ gap in the present study with new experiment design.
- Brief discussion of methods, goal of this study and hypothesis.
- Authors should write the statistical analysis in method section.
-Authors should revise discussion clearly.
Author Response
Response to 1. We have written the Abstract to better reflect these issues.
- - Authors should re-write the Introduction clearly divide by three paragraphs:
Response to 2. We do not feel that this advice is helpful.
- - Background and definition about the Vision.
Response to 3. Because we are uncertain as to what “the Vision” refers to we felt unable to revise the manuscript accordingly.
- -Previous published research about this topic with their limitations.
Response to 4. This is clearly discussed in the Introduction as it stands – we specify work relevant to BM theory and its limitations.
- -How to address the previous research literature limitations/ lack/ gap in the present study with new experiment design.
- - Brief discussion of methods, goal of this study and hypothesis.
Response to 5 and 6. Everything mentioned here is covered in the Introduction and we are both puzzled and confused why anyone would write this.
- - Authors should write the statistical analysis in method section.
We prefer to integrate the statistical reporting in the Results sections as is more standard practice.
- -Authors should revise discussion clearly.
We do not feel the need to do this and it is not a comment made by the other reviewers.
Round 2
Reviewer 2 Report
All my concerns have been addressed. I recommend this paper for publication.
