Selective attention and sensitivity to auditory disturbances in a virtually-real Classroom: Comparison of adults with and without AD(H)D

Reviewer #1 (Public review):

Summary:

This is an interesting study on AD(H)D. The authors combine a variety of neural and physiological metrics to study attention in a VR classroom setting. The manuscript is well written and the results are interesting, ranging from an effect of group (AD(H)D vs. control) on metrics such as envelope tracking, to multivariate regression analyses considering alpha-power, gaze, TRF, ERPs, and behaviour simultaneously. I find the first part of the results clear and strong. The multivariate analyses in Tables 1 and 2 are good ideas, but I think they would benefit from additional clarification. Overall, I think that the methodological approach is useful in itself. The rest is interesting in that it informs us on which metrics are sensitive to group effects and correlated with each other. I think this might be one interesting way forward. Indeed, much more work is needed to clarify how these results change with different stimuli and tasks. So, I see this as an interesting first step into a more naturalistic measurement of speech attention.

Strengths:

I praise the authors for this interesting attempt to tackle a challenging topic with naturalistic experiments and metrics. I think the results broadly make sense and they contribute to a complex literature that is far from being linear and cohesive.

Weaknesses:

Nonetheless, I have a few comments that I hope will help the authors improve the manuscript. Some aspects should be clearer, some methodological steps were unclear (missing details on filters), and others were carried out in a way that doesn't convince me and might be problematic (e.g., re-filtering). I also suggested areas where the authors might find some improvements, such as deriving distinct markers for the overall envelope reconstruction and its change over time, which could solve some of the issues reported in the discussion (e.g., the lack of correlation with TRF metrics).

I also have some concerns regarding reproducibility. Many details are imprecise or missing. And I did not find any comments on data and code sharing. A clarification would be appreciated on that point for sure.

There are some minor issues, typically caused by some imprecisions in the write-up. There are a few issues that could change things though (e.g., re-filtering; the worrying regularisation optimisation choices), and there I'll have to see the authors' reply to determine whether those are major issues or not. Figures should also be improved (e.g., Figure 4B is missing the ticks).

Reviewer #2 (Public review):

Summary:

While selective attention is a crucial ability of human beings, previous studies on selective attention are primarily conducted in a strictly controlled context, leaving a notable gap in underlying the complexity and dynamic nature of selective attention in a naturalistic context. This issue is particularly important for classroom learning in individuals with ADHD, as selecting the target and ignoring the distractions are pretty difficult for them but are the prerequisites of effective learning. The authors of this study have addressed this challenge using a well-motivated study. I believe the findings of this study will be a nice addition to the fields of both cognitive neuroscience and educational neuroscience.

Strengths:

To achieve the purpose of setting up a naturalistic context, the authors have based their study on a novel Virtual Reality platform. This is clever as it is usually difficult to perform such a study in a real classroom. Moreover, various techniques such as brain imaging, eye-tracking, and physiological measurement are combined to collect multi-level data. They found that, different from the controls, individuals with ADHD had higher neural responses to the irrelevant rather than the target sounds, and reduced speech tracking of the teacher. Additionally, the power of alpha-oscillations and frequency of gaze shifts away from the teacher are found to be associated with ADHD symptoms. These results provide new insights into the mechanism of selective attention among ADHD populations.

Weaknesses:

It is worth noting that nowadays there have been some studies trying to do so in the real classroom, and thus the authors should acknowledge the difference between the virtual and real classroom context and foresee the potential future changes.

The approach of combining multi-level data has the advantage of obtaining reliable results, but also raises significant difficulty for the readers to understand the main results.

An appraisal of whether the authors achieved their aims, and whether the results support their conclusions.

As expected, individuals with ADHD showed anomalous patterns of neural responses, and eye-tracking patterns, compared to the controls. But there are also some similarities between groups such as the amount of time paying attention to teachers, etc. In general, their conclusions are supported.

A discussion of the likely impact of the work on the field, and the utility of the methods and data to the community, would highlight the contributions of the work.

The findings are an extension of previous efforts in understanding selective attention in the naturalistic context. The findings of this study are particularly helpful in inspiring teacher's practice and advancing the research of educational neuroscience. This study demonstrates, again, that it is important to understand the complexity of cognitive processes in the naturalistic context.

Reviewer #3 (Public review):

Summary:

The authors conducted a well-designed experiment, incorporating VR classroom scenes and background sound events, with both control and ADHD participants. They employed multiple neurophysiological measures, such as EEG, eye movements, and skin conductance, to investigate the mechanistic underpinnings of paying attention in class and the disruptive effects of background noise.

The results revealed that individuals with ADHD exhibited heightened sensory responses to irrelevant sounds and reduced tracking of the teacher's speech. Overall, this manuscript presented an ecologically valid paradigm for assessing neurophysiological responses in both control and ADHD groups. The analyses were comprehensive and clear, making the study potentially valuable for the application of detecting attentional deficits.

Strengths:

• The VR learning paradigm is well-designed and ecologically valid.

• The neurophysiological metrics and analyses are comprehensive, and two physiological markers are identified capable of diagnosing ADHD.

• This research provides a valuable dataset that could serve as a benchmark for future studies on attention deficits.

Weaknesses:

• Several results are null results, i.e., no significant differences were found between ADHD and control populations.

• Although the paradigm is well-designed and ecologically valid, the specific contributions or insights from the results remain unclear.

• Lack of information regarding code and data availability.

Author response:

We are glad that the reviewers found our work to be interesting and appreciate its contribution to enhancing ecological validity of attention research. We also agree that much more work is needed to solidify this approach, and that some of the results should be considered “exploratory” at this point, but appreciate the recognition of the novelty and scientific potential of the approach introduced here.

We will address the reviewers’ specific comments in a revised version of the paper, and highlight the main points here:

· We agree that the use of multiple different neurophysiological measures is both an advantage and a disadvantage, and that the abundance of results can make it difficult to tell a “simple” story. In our revision, we will make an effort to clarify what (in our opinion) are the most important results and provide readers with a more cohesive narrative.

· Important additional discussion points raised by the reviewers, which will be discussed in a revised version are a) the similarities and differences between virtual and real classrooms; b) the utility of the methods and data to the community and c) the implication of these results for educational neuroscience and ADHD research.

· In the revision, we will also clarify several methodological aspects of the data analysis, as per the reviewers’ requests.

· After final publication, the data will be made available for other researchers to use.