Autism Information Progression and the Impact of Misinformation on Autism Knowledge, Awareness and Stigmatization

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is a very interesting and important topic, and its originality is demonstrated by the literature review, and the relatively small number of previous studies meeting the criteria for inclusion in the systematic review. The report is clear and well-written. The methodology is largely clearly stated but in my opinion would be improved by providing some more details as follows:

Reasons for excluding studies from the review if they included any participants who were not US residents. There may be sound reasoning for this, but the choice was not explained, and was not clear from the title or abstract.

How were the survey questionnaires completed, e.g. solely online? By sending email links? Were any measures taken, or considered, to verify that the survey participants had the connection to autism that they claimed to have? Such measures may have been considered but ruled out as impractical or impossible, however, this should maybe be explained briefly, as there is concern about the numbers of 'impostor' participants in autism research, particularly when online methods are used to collect data e.g. Pellicano, E., Adams, D., Crane, L., Hollingue, C., Allen, C., Almendinger, K., Botha, M., Haar, T., Kapp, S. K., & Wheeley, E. (2024). Letter to the Editor: A possible threat to data integrity for online qualitative autism research. Autism, 28(3), 786-792. 

Also, if participants were paid or compensated in some way for their time, e.g. through Amazon Mechanical Turk, this should probably be stated for clarity.

Overall, however, I enjoyed reading this research and think it makes a valuable contribution to the field.

Author Response

Comment 1: Reasons for excluding studies from the review if they included any participants who were not US residents. There may be sound reasoning for this, but the choice was not explained, and was not clear from the title or abstract.

Response 1: Thank you for the comment. The exclusion of studies involving non-U.S. residents in the systematic review phase was intentional to ensure methodological consistency across the study. Since Phase 2 of the study involved a survey conducted exclusively with participants residing in the United States, restricting the systematic review to studies with U.S.-based populations helped maintain comparability between findings across both phases. This approach minimized potential variability arising from differing geographic, cultural, or healthcare system contexts, thereby strengthening the overall coherence of the study. The inclusion criteria have been updated to reflect the reasoning. Please refer to lines 283 to 287.

Comment 2: How were the survey questionnaires completed, e.g. solely online? By sending email links? Were any measures taken, or considered, to verify that the survey participants had the connection to autism that they claimed to have? Such measures may have been considered but ruled out as impractical or impossible, however, this should maybe be explained briefly, as there is concern about the numbers of 'impostor' participants in autism research, particularly when online methods are used to collect data e.g. Pellicano, E., Adams, D., Crane, L., Hollingue, C., Allen, C., Almendinger, K., Botha, M., Haar, T., Kapp, S. K., & Wheeley, E. (2024). Letter to the Editor: A possible threat to data integrity for online qualitative autism research. Autism, 28(3), 786-792. 

Response 2: Thank you for this important comment. The survey questionnaires were completed entirely online using Qualtrics. Participants were recruited through two primary channels: Amazon Mechanical Turk (MTurk) and the UNT Kristin Farmer Autism Center. For MTurk recruitment, a study listing was created that included a description of the survey and a link to the Qualtrics questionnaire. The description clearly specified that only individuals diagnosed with autism or family members or caregivers of autistic individuals were eligible to participate. To enhance data quality, several qualification criteria were applied, including a minimum of 5,000 completed Human Intelligence Tasks (HITs) and an approval rate of at least 90%. Participation was also restricted to individuals located in the United States. In parallel, the director of the UNT Kristin Farmer Autism Center disseminated a study flyer to their clients, which included autistic individuals and their family members. The flyer provided a brief description of the study along with the survey link, allowing interested individuals to participate voluntarily. The manuscript is updated to include this information. Please refer to lines from 378 to 384.

Regarding verification of participants’ connection to autism, no external validation measures were implemented, as such approaches were deemed impractical in an anonymous online study. However, the survey included a self-report question asking participants to specify their relationship to autism (e.g., self-diagnosed, or family member). We acknowledge the concern regarding potential “impostor” participants in online research, and this limitation has now been noted in the manuscript. Kindly refer to lines from 840 to 844. The use of MTurk qualification criteria and targeted recruitment through a reputable autism center were intended to help mitigate this risk while maintaining feasibility.

Comment 3: Also, if participants were paid or compensated in some way for their time, e.g. through Amazon Mechanical Turk, this should probably be stated for clarity.

Response 3: Thank you for the review. Yes, the participants were compensated for completing the survey. The manuscript is updated to include this information. Please refer to lines 384, 385 and 386.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors,

Thank you for the opportunity to review this manuscript Autism Information Progression and the Impact of Misinformation on Autism Knowledge, Awareness and Stigmatization, submitted to Behavioral Sciences journal.

Your research provides important contribution to understanding how misinformation and social factors shape public perceptions of autism.

To further strengthen your manuscript, I recommend the following:

- The KAB model is introduced but not sufficiently justified or integrated with autism and misinformation literature.

- The conceptual relationships among knowledge, awareness, stigma, misinformation, cultural beliefs, and social norms need clearer theoretical grounding.

- The quality assessment of studies is missing.

- Instead of presenting the 23 articles primarily in tables and isolated summaries, integrate them into a narrative that explicitly answers RQ1: How has information evolved over time?.

- Table 7. Address sample diversity and limitations, survey participants were 84.28% White. Given that your systematic review highlights significant disparities for Latino and other ethnic communities, the survey findings may not accurately reflect the "sociocultural influences" you intended to study.

- For H5 (Awareness negatively correlates with stigma), you reported a correlation of 0.166 (p=0.018) but stated it was not supported. Typically, p < 0.05 is considered significant. Please clarify your threshold or explain if the effect size was considered too weak to support the hypothesis.

- Provide the full path coefficients and indirect effect statistics for the mediation analysis (H6) to allow readers to verify the lack of a significant relationship between stigma and awareness.

- Use the 2000–2025 timeframe to show a clear "before and after" regarding specific milestones, such as the shift from the schizophrenia model to the spectrum disorder model mentioned in your introduction.

I encourage the authors to revise the manuscript thoroughly and resubmit.

Sincerely,

Reviewer

Author Response

Comment 1: The KAB model is introduced but not sufficiently justified or integrated with autism and misinformation literature.

Response 1: Thank you for the review. The manuscript is updated to integrate KAB with autism and related constructs. Please refer to lines 132 to 160.

Comment 2: The conceptual relationships among knowledge, awareness, stigma, misinformation, cultural beliefs, and social norms need clearer theoretical grounding.

Response 2: Thank you for the review. The hypotheses are updated with supporting theoretical grounding. Please refer to lines 189 to 201; 219 to 223 and lines 237 to 248.

Comment 3: The quality assessment of studies is missing.

Response 3: Thank you for the review. The manuscript is updated to include a brief description of the quality assessment of the studies. The manuscript also includes a table that summarizes the quality assessment of each study included in the research. Please refer to lines 323 to 336 as well as table 2.

Comment 4: Instead of presenting the 23 articles primarily in tables and isolated summaries, integrate them into a narrative that explicitly answers RQ1: How has information evolved over time?.

Response 4: Thank you for the review. Table 2, 3, & 4 has been removed from the manuscript. Furthermore, a consolidated narrative has been added to the “Research Questions” section that addresses RQ1 on how autism information has evolved over time. Pease refer to lines 571 to 585. Thank you for the meaningful review.

Comment 5: Table 7. Address sample diversity and limitations, survey participants were 84.28% White. Given that your systematic review highlights significant disparities for Latino and other ethnic communities, the survey findings may not accurately reflect the "sociocultural influences" you intended to study.

Response 5: Thank you for this important observation. We agree that the racial and ethnic composition of the survey sample limits the representativeness of the findings. We have revised the manuscript to explicitly acknowledge that a greater percentage of survey respondents identified as White, which may constrain the generalizability of the results to more diverse populations. We also clarified that this limitation is particularly relevant given the disparities identified in the systematic review among Latino and other minority communities regarding autism knowledge, stigma, and access to services. In addition, we noted that differential participation across racial and ethnic groups may reflect sociocultural differences in perceptions of autism, stigma, and willingness to engage in autism-related research. Communities in which autism is more stigmatized or viewed negatively may have been less likely to participate, which could have contributed to the overrepresentation of White respondents. Please refer to lines 822–833.

Comment 6: For H5 (Awareness negatively correlates with stigma), you reported a correlation of 0.166 (p=0.018) but stated it was not supported. Typically, p < 0.05 is considered significant. Please clarify your threshold or explain if the effect size was considered too weak to support the hypothesis.

Response 6: Thank you for highlighting this point. Although the Pearson correlation between awareness and stigma was statistically significant (r = .166, p = .018), the observed association was positive rather than negative, which is contrary to the hypothesized direction. Because H5 predicted a negative relationship between autism awareness and stigma, the hypothesis was not supported. The manuscript has been updated to reflect this observation. Please refer to lines 547 to 551.

Comment 7: Provide the full path coefficients and indirect effect statistics for the mediation analysis (H6) to allow readers to verify the lack of a significant relationship between stigma and awareness.

Response 7: Thank you for your input. We have revised the manuscript to report the available path coefficients, standard errors, confidence intervals, and model fit statistics from the mediation analysis. The results indicate that awareness did not significantly predict stigma when controlling for knowledge, and therefore the hypothesized mediating effect was not supported. Kindly refer to lines 553 to 563 as well as Table 7.

Comment 8: Use the 2000–2025 timeframe to show a clear "before and after" regarding specific milestones, such as the shift from the schizophrenia model to the spectrum disorder model mentioned in your introduction.

Response 8: Thank you for this helpful suggestion. We revised the introduction section to explicitly organize autism information across the 2000–2025 timeframe and highlight major milestones in the evolution of autism-related information. Please refer to lines 93 to 114.

Reviewer 3 Report

Comments and Suggestions for Authors

Overall, the manuscript addresses a relevant and timely topic, particularly given the growing presence of autism-related misinformation and its possible implications for knowledge, awareness, and stigmatization. A clear strength of the paper is its attempt to combine a systematic review with an empirical survey in a single design, and to place sociocultural factors and misinformation within the same conceptual framework. At the same time, the manuscript in its current form has several theoretical, methodological, and interpretive weaknesses that should be addressed before it can be considered for publication. Here are some of my major suggestions which need to be considered:

• The adaptation of the Knowledge–Attitude–Behavior (KAB) model is introduced as the foundation of the study, but the conceptual justification is not sufficiently convincing. In particular, the manuscript treats autism awareness as the “attitude” component and stigmatization as the “behavior” component, yet this mapping is asserted rather than fully argued. Because the logic of the hypotheses depends on this adaptation, the paper would benefit from a more rigorous explanation of why awareness should be treated as attitude and why stigma is framed as behavior rather than as an attitudinal or evaluative construct.
• The hypotheses also need tighter alignment with the theoretical discussion. For example, the paper hypothesizes that awareness should reduce stigma, yet later findings do not support that expectation. This would have been less problematic if the introduction had engaged more critically with the possibility that awareness alone may not be sufficient to alter stigmatizing beliefs. Similarly, the hypothesis that misinformation is directly related to knowledge and awareness is conceptually ambiguous, especially because the observed results suggest a positive relation between misinformation and knowledge. That raises concerns about whether “knowledge” is being treated as objective knowledge, perceived knowledge, or general familiarity.
• The systematic review does not yet appear methodologically robust enough to function as a strong systematic review. The inclusion criteria are highly restrictive, especially the limitation to U.S.-based studies involving only autistic individuals and family members. This substantially narrows the evidence base while the manuscript still makes broad claims about the “progression” of autism information over time. In addition, although the manuscript states that the GRADE framework was used to evaluate strength and certainty of evidence, those assessments are not meaningfully presented in the results. There is no clear risk-of-bias table, no quality appraisal summary, and no transparent integration of those ratings into the synthesis.
• Some items were adapted from the ASK-Q, but the manuscript states that original nominal-response items were converted into five-point Likert items. This is a substantial change in measurement format and requires much stronger justification and validation than is currently provided. In addition, several constructs appear to have been measured with newly developed items reviewed by an expert and evaluated primarily through Cronbach’s alpha. For a study that introduces modified and newly constructed scales, this is not sufficient evidence of validity. Exploratory or confirmatory factor analysis (EFA/CFA) would be much more appropriate. If it is not applicable, please state that and give us a reason why.
• The sample imposes serious limits on generalizability. Participants were recruited through MTurk and one university autism center, and the manuscript reports that the sample was predominantly male, largely young adults, and mostly White. In addition, the target population consists of autistic individuals, family members, and caregivers rather than the general public. This makes the sample appropriate for a specific subgroup perspective, but not for broader conclusions about autism knowledge, awareness, and stigma in the wider population. That limitation should be more explicitly acknowledged and integrated into the interpretation.
• The analytic strategy seems too basic for the complexity of the proposed model. The manuscript relies mainly on descriptive statistics, correlations, regressions, and a mediation analysis, yet the conceptual framework implies multiple interrelated pathways. A more coherent analytic approach, such as a structural model or at least a more systematic multivariable framework, would be preferable.
• The manuscript reports that greater knowledge is associated with greater awareness, which is theoretically consistent. At the same time, it also reports a positive association between misinformation and knowledge, which is conceptually counterintuitive with KAB model and deserves much more scrutiny. This raises the possibility that the knowledge variable is not capturing objective knowledge in a straightforward way, or that item wording and scoring may need to be checked carefully.
• There are also apparent inconsistencies in the reporting and interpretation of the hypothesis tests. For example, the manuscript states that autism awareness did not significantly influence stigma, but it also reports a correlation coefficient with a p-value of .018. That is a statistical inconsistency, or at minimum a wording problem, because a significant association is being described as non-significant. Likewise, the reported correlation between misinformation and awareness is given as 0.73 with a non-significant p-value of .180, which strongly suggests a typographical, coding, or reporting error. These issues must be corrected before the results can be confidently interpreted.
• The main weakness of the discussion is that it remains too descriptive and not analytical enough. The most surprising findings, especially the positive association between misinformation and knowledge and the unclear role of awareness in reducing stigma, are not unpacked in sufficient depth. Instead of carefully considering whether these findings may reflect theoretical complexity, measurement problems, sampling features, or statistical issues, the discussion moves rather quickly to practical recommendations. That weakens the scientific contribution of the paper.
• The limitations section is relevant, but it should be more deeply integrated into the interpretation of the findings. The manuscript acknowledges the modest sample size, MTurk recruitment, cross-sectional design, and self-report measures, but it does not fully spell out the implications of these limitations. In particular, causal language should be avoided throughout the discussion, since the cross-sectional design does not support causal inference.
• The discussion would also be strengthened by more clearly separating what is supported by the systematic review from what is supported by the survey. At several points, the manuscript blends both sources of evidence into a single conclusion, even though the evidentiary strength and methodological role of the two phases are different.

Author Response

Comment 1: The adaptation of the Knowledge–Attitude–Behavior (KAB) model is introduced as the foundation of the study, but the conceptual justification is not sufficiently convincing. In particular, the manuscript treats autism awareness as the “attitude” component and stigmatization as the “behavior” component, yet this mapping is asserted rather than fully argued. Because the logic of the hypotheses depends on this adaptation, the paper would benefit from a more rigorous explanation of why awareness should be treated as attitude and why stigma is framed as behavior rather than as an attitudinal or evaluative construct.

Response 1: Thank you so much for your input. The manuscript has been updated to include justification on why autism awareness is treated as attitude and why autism stigma is considered as behavior. Please refer to lines 147 through 152.

Comment 2: The hypotheses also need tighter alignment with the theoretical discussion. For example, the paper hypothesizes that awareness should reduce stigma, yet later findings do not support that expectation. This would have been less problematic if the introduction had engaged more critically with the possibility that awareness alone may not be sufficient to alter stigmatizing beliefs. Similarly, the hypothesis that misinformation is directly related to knowledge and awareness is conceptually ambiguous, especially because the observed results suggest a positive relation between misinformation and knowledge. That raises concerns about whether “knowledge” is being treated as objective knowledge, perceived knowledge, or general familiarity.

Response 2: Thank you for the input. The review has been addressed in the manuscript. Please refer to lines 65 to 70 which address the relation between awareness and stigma. Also, please refer to lines 228 to 235 which includes the statements that support hypothesis 4. Also, some of the hypotheses are updated with supporting theoretical grounding to better support the hypothesis.

Comment 3: The systematic review does not yet appear methodologically robust enough to function as a strong systematic review. The inclusion criteria are highly restrictive, especially the limitation to U.S.-based studies involving only autistic individuals and family members. This substantially narrows the evidence base while the manuscript still makes broad claims about the “progression” of autism information over time. In addition, although the manuscript states that the GRADE framework was used to evaluate strength and certainty of evidence, those assessments are not meaningfully presented in the results. There is no clear risk-of-bias table, no quality appraisal summary, and no transparent integration of those ratings into the synthesis.

Response 3: Thank you for this valuable comment. We have strengthened the manuscript in several ways. First, we clarified that the restriction to U.S.-based studies was intentional to ensure consistency with Phase 2 of the study, which surveyed participants exclusively in the United States. Please refer to lines 283 to 287, where the inclusion criteria have been revised. Also, we have added a limitation stating the generalizability of the systematic review phase. Please refer to lines 807 to 813. Second, we expanded the presentation of methodological quality by more clearly reporting the GRADE assessments and risk-of-bias findings in the Data Collection and Analysis section (section 3.1) through both tabular and narrative summaries. Please refer to lines 323–336 for the narrative and also please refer to Table 2 for the quality assessment of studies included in the systematic review.           

Comment 4: Some items were adapted from the ASK-Q, but the manuscript states that original nominal-response items were converted into five-point Likert items. This is a substantial change in measurement format and requires much stronger justification and validation than is currently provided. In addition, several constructs appear to have been measured with newly developed items reviewed by an expert and evaluated primarily through Cronbach’s alpha. For a study that introduces modified and newly constructed scales, this is not sufficient evidence of validity. Exploratory or confirmatory factor analysis (EFA/CFA) would be much more appropriate. If it is not applicable, please state that and give us a reason why.

Response 4: Thank you for this important comment. We agree that converting nominal-response items to five-point Likert scales represents a meaningful modification of the original ASK-Q format and requires stronger justification. In response, we have revised the Data Collection and Analysis section to clarify that Likert scaling was used to capture gradations in endorsement and to provide greater response variability for correlational analyses. We also acknowledge that this modification may alter the original psychometric properties of the instrument and that findings should therefore be interpreted cautiously. Please refer to lines 348 and 353. Regarding construct validation, we agree that EFA/CFA would provide stronger evidence of validity. However, given the modest sample size relative to the number of measured items and the exploratory nature of the study, full factor-analytic validation was not considered statistically stable. We have now stated this limitation clearly in the manuscript and framed the reported Cronbach’s alpha estimates as preliminary evidence of internal consistency rather than comprehensive validation. Please refer to lines 364 to 370 and lines 844 to 850.

Comment 5: The sample imposes serious limits on generalizability. Participants were recruited through MTurk and one university autism center, and the manuscript reports that the sample was predominantly male, largely young adults, and mostly White. In addition, the target population consists of autistic individuals, family members, and caregivers rather than the general public. This makes the sample appropriate for a specific subgroup perspective, but not for broader conclusions about autism knowledge, awareness, and stigma in the wider population. That limitation should be more explicitly acknowledged and integrated into the interpretation.

Response 5: Thank you for the input. We have added this limitation to the manuscript. Please refer to lines 835 to 840.

Comment 6: The analytic strategy seems too basic for the complexity of the proposed model. The manuscript relies mainly on descriptive statistics, correlations, regressions, and a mediation analysis, yet the conceptual framework implies multiple interrelated pathways. A more coherent analytic approach, such as a structural model or at least a more systematic multivariable framework, would be preferable.

Response 6: We thank you for this insightful suggestion. We agree that the conceptual model implies complex, interrelated pathways that could be optimally tested using structural equation modeling (SEM). However, given the exploratory nature of this study, the cross-sectional design, and sample size considerations, SEM was not considered appropriate for the present analysis. Instead, we employed a stepwise analytic strategy including descriptive statistics, correlations, multiple regression, and mediation analysis to examine key relationships among variables. We have now clarified and strengthened the justification for this analytic approach in the Data Collection and Analysis section (please refer to lines 388 through 393) and acknowledged in the Discussion and Conclusion section that future research with larger samples should employ SEM or other latent variable modeling techniques to more comprehensively test the proposed framework (please refer to lines 881 to 883).

Comment 7: The manuscript reports that greater knowledge is associated with greater awareness, which is theoretically consistent. At the same time, it also reports a positive association between misinformation and knowledge, which is conceptually counterintuitive with KAB model and deserves much more scrutiny. This raises the possibility that the knowledge variable is not capturing objective knowledge in a straightforward way, or that item wording and scoring may need to be checked carefully.

Response 7: Thank you for this insightful comment. We agree that the positive association between misinformation and the knowledge variable requires closer consideration. In response, we have revised the several sections of the manuscript to more carefully interpret this unexpected finding. Please refer to lines 228 to 235; 632 to 634. Specifically in the Discussion and Conclusion section, we now note that the knowledge measure may reflect not only objective factual accuracy, but also general familiarity or exposure to autism-related information. Individuals with greater exposure may acquire both accurate information and misconceptions simultaneously. We also acknowledge the possibility that item wording, scoring, or construction operationalization may have influenced this result.  Please refer to lines 768 to 779.

Comment 8: There are also apparent inconsistencies in the reporting and interpretation of the hypothesis tests. For example, the manuscript states that autism awareness did not significantly influence stigma, but it also reports a correlation coefficient with a p-value of .018. That is a statistical inconsistency, or at minimum a wording problem, because a significant association is being described as non-significant. Likewise, the reported correlation between misinformation and awareness is given as 0.73 with a non-significant p-value of .180, which strongly suggests a typographical, coding, or reporting error. These issues must be corrected before the results can be confidently interpreted.

Response 8: Thank you for this important observation. We carefully reviewed the statistical reporting and interpretations throughout the manuscript and made the necessary corrections. Regarding Hypothesis 5, the hypothesis proposed a negative correlation between autism awareness and autism stigma. However, Pearson correlation analysis indicated a statistically significant positive relationship between awareness and stigma (r = .166, p = .018). Because the observed relationship was opposite to the hypothesized direction, H5 was revised as not supported rather than non-significant. Please refer to lines 547–551. Regarding Hypothesis 4, we appreciate the identification of the typographical error in the reported correlation between misinformation and autism awareness. The correct correlation coefficient is r = .073 (previously reported as r = .73), with p = .180. This has been corrected in the revised manuscript (line 544). Additionally, all statistical values, tables, and narrative interpretations were rechecked to ensure consistency and accuracy throughout the manuscript.

Comment 9: The main weakness of the discussion is that it remains too descriptive and not analytical enough. The most surprising findings, especially the positive association between misinformation and knowledge and the unclear role of awareness in reducing stigma, are not unpacked in sufficient depth. Instead of carefully considering whether these findings may reflect theoretical complexity, measurement problems, sampling features, or statistical issues, the discussion moves rather quickly to practical recommendations. That weakens the scientific contribution of the paper.

Response 9: Thank you for this valuable comment. We agree that the original Discussion and Conclusion section required deeper analytical interpretation of the more unexpected findings. The Discussion and Conclusion section has been revised to provide a more analytical interpretation of the unexpected findings. Specifically, we expanded the discussion of the positive association between misinformation and knowledge and considered potential theoretical and measurement-related explanations (lines 768–779). In addition, we strengthened the discussion of the relationship between autism awareness and stigma by further examining its implications and possible underlying factors (lines 757–767).

Comment 10: The limitations section is relevant, but it should be more deeply integrated into the interpretation of the findings. The manuscript acknowledges the modest sample size, MTurk recruitment, cross-sectional design, and self-report measures, but it does not fully spell out the implications of these limitations. In particular, causal language should be avoided throughout the discussion, since the cross-sectional design does not support causal inference.

Response 10: Thank you for your valuable suggestion. The limitations section has been substantially revised to better integrate the study’s limitations into the interpretation of the findings. Please refer to lines 835–866. In addition, we carefully reviewed the Discussion and Conclusion section and revised instances of causal language where appropriate to ensure that the findings are interpreted in terms of associations rather than causation.

Comment 11: The discussion would also be strengthened by more clearly separating what is supported by the systematic review from what is supported by the survey. At several points, the manuscript blends both sources of evidence into a single conclusion, even though the evidentiary strength and methodological role of the two phases are different.

Response 11: Thank you for your valuable comment. To address your concern, we have revised selected sections of the Discussion and Conclusion to provide a clearer distinction between findings from the systematic review and those from the survey phase. Specifically, the second and third paragraphs of the Discussion and Conclusion section relate to the systematic review findings, whereas paragraphs four through seven focus on the survey phase findings. In addition, where appropriate, we have explicitly identified the relevant study phase to clarify the source of evidence and strengthen the interpretation of the results.

Reviewer 4 Report

Comments and Suggestions for Authors

Autism Information Progression and the Impact of Misinformation on Autism Knowledge, Awareness and Stigmatization

Thank you for the opportunity to review your research. While I think your manuscript needs some work, I do think the topic is interesting and with some effort your study could add to the understanding of autism and information progression. My comments:

1. You should know that I no longer think we can treat the ASD population as one population. Some of my comments reflect this. This is a bias of mine.
2. Phase one of your study provides important insight into different ethnic/racial groups and is a strength of your study. I would like to see more recent studies on genetic testing or comments about the changes in genetic testing (see my comment below)
3. In 2012-2013 genetic testing was not as common as it is now. I would reference how assessable overall genetic testing was at the time of the studies you cite. This does not dimmish the ethnic/racial differences but is an important point particularly if there are differences in private insurance vs. public healthcare (e.g. Medicaid). Additionally, before 2019 only limited genetic testing was recommended. It was not until 2019 that exome sequencing was recommended for neurodevelopmeal disorders. Lastly, early developmental delays and more severe symptoms are associated with higher levels of genetic Moreno-De-Luca et al. (2020). Only one of the studies you cited looked at genetic testing after 2019 (e.g. Zebolsky et al., 2020) and the data from this study was likely pre-2019. At a minimum I suggest adding a comment about this in the Discussion.
4. The Knowledge-Attitude-Behavior (KAB) model treats behavior in a linear fashion (e.g. human behavior as a rational linear process). This has been one of the criticisms of the KAB model. Could your findings that increased knowledge and awareness didn’t lead to decrease stigma be partly related to the KAB model which informed your hypotheses? I think your results highlight the oversimplistic assumptions of the KAB model. My thoughts about this:
   1. You mention a list of factors that are also related to stigma such as unmet needs, parent report of autism severity or number of autistic children in the family. What is the relationship between these factors and stigma? Does increased knowledge address these factors? Could these factors be the reason you didn’t find that increased knowledge didn’t eliminate stigma? If that is the case, just providing more information will not decrease sigma related to the other factors you did not examine. You should at a minimum discuss the issue in your limitations.
   2. I think you are overstating the conclusions from the regression analysis of the influence of cultural beliefs and social norms on knowledge, awareness and misinformation (lines 482-495). It felt like you were trying to fit the KAB model when it didn’t fit well.
   3. I wonder if misinformation could actually lower sigma. You assume that misinformation might increase sigma but some misinformation such as the misinformation around an “intact mind” in non-verbal autistic individuals or misinformation about “special talents” of autistic individuals could lower sigma.
5. I am particularly concern about the relationship between stigma and severity of disability. There are two issues here. Frist the ASD population is extremely heterogeneous. I suggest you mention the extreme heterogeneity of the population. Secondly, I think we can assume that the autistic individuals who participated had the cognitive and language abilities to participate and thus do not represent those autistic individuals who don’t. Parent participants might represent autistic children or adults with more severe disability, but you don’t give any information about this (you may not have this information), so I don’t know if this is a factor. If your study mostly represents the subpopulation of the spectrum that has lower support needs that is fine, but this would also be a limitation and should be mentioned.
6. Overall my two biggest concerns about your study is the use of the KAB model and that your study may not apply to those autistic individuals with cogitative and language abilities. As mentioned above the parents who responded may be caring more severely disabled autistic children or adults, but I can tell if this is the case. If it is the case, then you have surveyed 2 distinct populations within the spectrum included in your results. Stigma may be very different for these populations and the factors driving stigma may also be different. If you have data that indicates this, I suggest you look at the differences in responses between parents and autistic individuals who took the survey. If you do not have the data I suggest mentioning this in the limitations.
7. Lately, I have included some comments in the manuscript some overlap with the comments here. Please ignore the different colors of highlighting in the manuscript they just reflect my inability to change the color back to yellow.
8. I hope you find my comments helpful. 

Comments for author File: Comments.pdf

Author Response

Comment 1: Phase one of your study provides important insight into different ethnic/racial groups and is a strength of your study. I would like to see more recent studies on genetic testing or comments about the changes in genetic testing (see my comment below).

Response 1: Thank you for your valuable input. We have updated the manuscript to include a brief description of advancement of genetic testing in the Introduction section. Please refer to lines 106 to 111.

Comment 2: In 2012-2013 genetic testing was not as common as it is now. I would reference how assessable overall genetic testing was at the time of the studies you cite. This does not dimmish the ethnic/racial differences but is an important point particularly if there are differences in private insurance vs. public healthcare (e.g. Medicaid). Additionally, before 2019 only limited genetic testing was recommended. It was not until 2019 that exome sequencing was recommended for neurodevelopmental disorders. Lastly, early developmental delays and more severe symptoms are associated with higher levels of genetic Moreno-De-Luca et al. (2020). Only one of the studies you cited looked at genetic testing after 2019 (e.g. Zebolsky et al., 2020) and the data from this study was likely pre-2019. At a minimum I suggest adding a comment about this in the Discussion.

Response 2: Thank you for your valuable input. We have revised the Discussion and Conclusion section of the manuscript to acknowledge advances in genetic testing and to clarify that findings from earlier studies should be interpreted within the historical context of testing availability, affordability, and clinical recommendations at that time. We also added text emphasizing the need for continued provider education, equitable access to evolving genetic services, and improved communication between professionals and families. Please refer to lines 675 to 686.

Comment 3: The Knowledge-Attitude-Behavior (KAB) model treats behavior in a linear fashion (e.g. human behavior as a rational linear process). This has been one of the criticisms of the KAB model. Could your findings that increased knowledge and awareness didn’t lead to decrease stigma be partly related to the KAB model which informed your hypotheses? I think your results highlight the oversimplistic assumptions of the KAB model. My thoughts about this:

Response 3: Thank you for your valuable comment. We agree with your observation that the KAB model has been critiqued for its linear and overly simplified assumption that human behavior follows a rational, sequential process. We acknowledge that this limitation is relevant to the interpretation of our findings. Hence, we have revised the manuscript to explicitly incorporate this criticism of the KAB framework (lines 132–135). In addition, we have updated the Discussion and Conclusion section to show that our results do not fully support a simple linear interpretation of the KAB model, particularly given that increased knowledge and awareness were not associated with reduced stigma. Please refer to lines 750–757.

1. Comment 3.1: You mention a list of factors that are also related to stigma such as unmet needs, parent report of autism severity or number of autistic children in the family. What is the relationship between these factors and stigma? Does increased knowledge address these factors? Could these factors be the reason you didn’t find that increased knowledge didn’t eliminate stigma? If that is the case, just providing more information will not decrease sigma related to the other factors you did not examine. You should at a minimum discuss the issue in your limitations.

Response 3.1: Thank you for your review. We agree and acknowledge your important point that several factors associated with stigma such as unmet care needs, parent-reported autism severity, and caregiving demands are not directly addressed by knowledge-based interventions. We agree that increased knowledge alone is unlikely to modify these broader contextual and experiential factors, which may therefore help explain why we did not find that knowledge and awareness were associated with reduced stigma. Based on your suggestion, we have expanded the limitations section to explicitly discuss this issue (lines 813–821), emphasizing that informational approaches alone may be insufficient unless complemented by interventions that address practical support needs and lived family experiences.

2. Comment 3.2: I think you are overstating the conclusions from the regression analysis of the influence of cultural beliefs and social norms on knowledge, awareness and misinformation (lines 482-495). It felt like you were trying to fit the KAB model when it didn’t fit well.

Response 3.2: Thank you for this insightful comment. We really appreciate the concern regarding the interpretation of the regression findings and the potential overstatement of the results. We agree that the effect sizes, particularly the low R² values for some of the models, indicate that cultural beliefs and social norms explain only a modest proportion of the variance in autism knowledge, awareness, and misinformation. So, we have revised the relevant section to avoid overinterpreting and tone down these associations and to ensure that the findings are not framed as strongly supporting a KAB-type linear relationship. Please refer to lines 646 to 657.

3. Comment 3.3: I wonder if misinformation could actually lower sigma. You assume that misinformation might increase sigma but some misinformation such as the misinformation around an “intact mind” in non-verbal autistic individuals or misinformation about “special talents” of autistic individuals could lower sigma.

Response 3.3: Thank you for this suggestion. In this study, we initially assumed that increased awareness would be associated with reduced stigma; however, we acknowledge that this relationship may be more complex, and that some forms of misinformation may in fact reduce certain aspects of stigma. We have therefore included your suggestion in the manuscript as a direction for future research. Please refer to lines 874 to 881.

Comment 4: I am particularly concern about the relationship between stigma and severity of disability. There are two issues here. Frist the ASD population is extremely heterogeneous. I suggest you mention the extreme heterogeneity of the population. Secondly, I think we can assume that the autistic individuals who participated had the cognitive and language abilities to participate and thus do not represent those autistic individuals who don’t. Parent participants might represent autistic children or adults with more severe disability, but you don’t give any information about this (you may not have this information), so I don’t know if this is a factor. If your study mostly represents the subpopulation of the spectrum that has lower support needs that is fine, but this would also be a limitation and should be mentioned.

Response 4: Thank you for this insightful comment. We agree that autism is highly heterogeneous. We have already acknowledged this in the Introduction (lines 47–51). In addition, we have revised the limitations section to note that our findings may more strongly reflect the experiences of autistic individuals who are able to independently complete surveys, as well as family members of autistic individuals (lines 850–855)

Comment 5: Overall my two biggest concerns about your study is the use of the KAB model and that your study may not apply to those autistic individuals with cogitative and language abilities. As mentioned above the parents who responded may be caring more severely disabled autistic children or adults, but I can tell if this is the case. If it is the case, then you have surveyed 2 distinct populations within the spectrum included in your results. Stigma may be very different for these populations and the factors driving stigma may also be different. If you have data that indicates this, I suggest you look at the differences in responses between parents and autistic individuals who took the survey. If you do not have the data I suggest mentioning this in the limitations.

Response 5: Thank you for this meaningful comment. We understand your concern. Since we do not have the necessary data, we have included this as a limitation in the manuscript. Please refer to lines 850 to 864.

Comment 6: Lately, I have included some comments in the manuscript some overlap with the comments here. Please ignore the different colors of highlighting in the manuscript they just reflect my inability to change the color back to yellow.

Response 6: Thank you for these thoughtful comments. We have addressed the additional comments that were highlighted in the manuscript.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors, 

Thank you for addressing our concerns.

I do not have any other comments, 

Regards

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have adequately addressed the reviewers’ suggestions and revised the manuscript accordingly. In its current form, the manuscript is suitable for publication, and I recommend its acceptance.

     

Reviewer 4 Report

Comments and Suggestions for Authors

Thank you for addressing my concerns. Your article is much improved and reads very well.