Factors associated with U.S. adults' willingness to allow teenagers to play tackle football.

Little is known about the individual factors, such as knowledge and attitudes (i.e., football safety knowledge, football attitudes), related to adults' willingness to allow adolescents to participate in tackle football. To address this gap, this study examined the extent to which football safety knowledge and attitudes toward head injury risk are associated with adults' willingness to allow teenage boys to play high school tackle football. Data were obtained from an internet-based survey of a nationally representative sample of U.S. adults aged 18 to 93 years (n = 1,018). We conducted multilevel linear regression modelling to examine independent effects of the football safety knowledge- and attitude-based predictors. Our analyses revealed that knowledge of football safety measures, along with four of the five attitude-based variables were significantly associated with adults' willingness to allow teenagers to participate in tackle football, over and above demographic factors. This study provides the first nationally representative examination of willingness to allow tackle football participation while extending our understanding of the gap between policy, public perception, and behavior present in U.S. high school football. These results point to promising directions for stakeholders aiming to increase tackle football participation as an increased understanding of the factors associated with participation may help inform effective policymaking, intervention design, and parental decision making.

Introduction

In 2019 the Centers for Disease Control and Prevention reported that sport participation remains low with 46% of youth not participating on any sports team in the past year [1]. A range of national initiatives from organizations such as the U.S. Department of Health and Human Services [2], Aspen Institute [3], and U.S. Olympic Committee [4] have been implemented in the past decade to help take advantage of the public health opportunity provided by youth sport. Despite the promise of these initiatives, segments of the population are disproportionately affected [5]. For example, youth from low-income households are up to 23.5% less likely to have engaged in any sport activity during the past year compared to children from households with annual income greater than $100,000 [5]. In the context of youth football, USA Football, the national governing body over amateur football in the U.S., started the Football Development Model aiming to make the game safer through reduced contact and age-appropriate coaching [6]. Albeit relatively new, these initiatives signal preliminary momentum for participation in youth sports, such as tackle football to develop as a national public health priority.

Although driven primarily by male participation, football remains the second most popular high school sport among boys and girls in the U.S., and it is the most popular boys’ sport by approximately 400,000 participants [7]. Part of this popularity stems from distinctive features of football in that it is accessible to both rural and urban populations [8], inclusive of a wide range of body types [9], and incorporates large team sizes. For these reasons, along with the thoroughly engrained nature of football in American culture, high school tackle football provides a fertile ground for nationally representative investigation of the factors associated with participatory decisions that may have scalable implications for youth public health outcomes [10].

While there are positive aspects associated with tackle football, the landscape in the U.S. is shifting. Nine of the past ten years have seen declines in high school tackle football participation [7]. These declines may be related to increased awareness of the relatively high head impact frequencies and magnitudes in tackle football along with initial work suggesting that earlier participation in tackle football may be associated with worse later-in-life cognitive and behavioral outcomes [11, 12]. These significant strides in brain injury research and increased societal awareness have coincided with regulative changes related to tackling, equipment innovations, practice structural recommendations, and game rules at various levels [11-14]. Notably, there are six states with active or pending legislation limiting the amount of allowable weekly full-contact practice time and/or banning tackle football before a certain age [15]. These moves are historic because aside from concussion legislation, the youth football environment has typically been regulated by a complex sport governance system with minimal involvement by state legislatures [16]. However, while these regulative changes targeting increased safety and national initiatives emphasizing participation are promising for the future of football participation, little is known about the factors driving football participatory decisions at the individual level (e.g., knowledge, attitudes).

To our knowledge, there has not been a nationally representative assessment of U.S. adults’ willingness to allow tackle football participation; nor has there been an assessment of the relationship between individual-level behavioral constructs (e.g., knowledge, attitudes) and willingness to allow tackle football participation. Previous research has explored individuals’ willingness to allow children to play football, but those studies used convenience samples of college students and respondents from only one state [17, 18]. This is a significant limitation in the current literature because youth sport participatory outcomes have been shown to differ significantly across demographic groups [5]. Chrisman et al. [19] explored support for tackling restrictions in youth football, a related topic, but supporting tackling restrictions is not as bold or influential of a parenting decision as willingness to allow teenage boys to play tackle football. Past research has also argued for greater empirical attention on adults in the youth sport context (i.e., coaches, parents, etc.) as they have been shown to play a key role in determining the youth motivational climate and sustained youth sport participation [20-22]. Consequently, the present study was warranted in that it addressed both the underrepresented topic of youth football participatory decisions in adults and incorporated a nationally representative sample. The research question guiding the present study was: After controlling for demographic influences, to what extent are knowledge of football safety measures and attitudes toward head injury risk associated with adults’ willingness to allow teenage boys to play high school tackle football? Our two primary hypotheses were that (1) respondents with greater knowledge of football safety measures would be more willing to allow teenage boys to participate in high school tackle football and (2) respondents with greater worry about head injury risk (i.e., attitude) would be less likely to allow teenage boys to participate in high school tackle football.

Methods

Sample

This was a prospective, cross-sectional survey study. Data were obtained from an internet-based survey of a nationally representative sample of U.S. adults aged 18 to 93 years (n = 1,018). Respondents were members of Ipsos KnowledgePanel, a large probability-based online panel in the U.S., with approximately 60,000 members. The KnowledgePanel sample was developed on a foundation of address-based sampling with a statistically valid representation of the U.S. population as well as many under-researched and often harder-to-reach populations [23]. Ipsos [23] addresses self-selection bias by choosing respondents through the Delivery Sequence File of the United States Postal Service, which is a single sampling frame covering everybody in the U.S. with a postal address. Importantly, a random sample of households are sent a mail invitation to join the panel; people cannot simply volunteer. The response rate was not available but KnowledgePanel takes steps to address non-response bias (see additional information on the KnowledgePanel sampling methods by accessing their methodology report [23]). Before joining KnowledgePanel, participants were provided with written informed consent followed by a standard email invitation to the survey. This study was deemed to be exempt from full Institutional Review Board approval from the authors’ university because the research posed no more than minimal risk to participants.

Survey outcomes

Respondents were guided through demographic questions (See S1 Appendix for list of demographic items) followed by sport-specific football safety knowledge and football attitudinal items (See S2 Appendix) and sets of questions on other topics outside of the scope of the present study. The outcome variable for this study was willingness of U.S. adults to allow teenage boys play tackle football. To assess this outcome, participants responded to the following statement: “As a parent of a teenage boy, or if I was a parent of a teenage boy, I would let him play tackle football if he wanted to.” Participants responded using a 5-point Likert scale ranging from “Strongly Disagree” to “Strongly Agree.” A 5-point Likert scale was selected to provide greater measurement variance within respondents compared to binary yes/no responses. The sample was normally distributed across the outcome variable with a mean score of 3.01 (SD = 1.15; Range: 1–5), median of 3, skewness of -0.34, and kurtosis of -0.76.

There were six primary independent variables of interest used for this project that were based on knowledge and attitudes. These predictors included one knowledge of football safety measures composite variable and five attitudinal variables (i.e., worry about head injuries, belief the media exaggerates the problem of concussions, belief rule changes in recent years have made football safer, belief concussions are a serious problem, and perceived prevalence of concussion).

We calculated a cumulative score from ten true/false questions related to various football safety measures to assess knowledge of football safety measures. These items covered measures such as tackling guidelines, equipment innovations, practice structural recommendations, and game rules at various levels. Respondents were coded with a 0 or 1 depending on whether they answered the question correctly or not, and a cumulative score out of 10 was calculated.

This composite score was calculated by producing a mean score from three worry-related items. Each of the three initial questions had a question stem of “Indicate how much you worry about each statement for your own teenage boy (or how much you would worry if you had a teenage boy)” and covered worry about (a) concussion, (b) long-term brain health, and (c) frequency of head injuries from playing tackle football in high school (see S3 Appendix for specific survey items that were used in the analysis). The scale for each of these worry-based questions was a 4-point Likert scale assessing the extent of worry and ranged from “I don’t worry at all” to “I worry a lot.” These questions were a separate set of items from the knowledge and belief-based questions.

To assess this variable, respondents were provided with the following statement: “The media exaggerated the problem of concussions in football.” A 5-point Likert scale with response options ranging from “Strongly Disagree” to “Strongly Agree” was used.

To assess this variable, respondents were provided with the following statement: “The rule changes implemented for football games in recent years have made football a safer sport.” A 5-point Likert scale with response options ranging from “Strongly Disagree” to “Strongly Agree” was provided.

To assess this variable, respondents were provided with the following statement: “Concussions in football are a serious problem.” A 5-point Likert scale with response options ranging from “Strongly Disagree” to “Strongly Agree” was used.

To operationalize the attitude construct from an additional lens, we utilized respondents’ perceived prevalence of concussion in high school football. This item was used in past research exploring parents’ support for tackling restrictions in youth football [19]. We assessed this variable with a survey item utilizing a 0–100 scale asking respondents: “How many players out of 100 do you estimate would get a concussion during 1 season of high school football?”

As effectiveness and awareness of the vast number of football safety measures remains largely unexplored [14], we approached this construct in an exploratory manner. The items for football safety knowledge were developed by a panel of experts in public health policy, neuroscience, football coaching, and sport psychology (see S2 Appendix for all sport-based survey items). Additionally, the true/false structure of this assessment was similar to past literature assessing the knowledge construct within the KAB model [24-26]. Regarding the worry-based items, three items were used to create the composite score representing worry about head injuries. These items had a Cronbach’s alpha = 0.96 which indicated adequate internal reliability.

Demographic items included age, gender, education, household income, and parent status (see S1 Appendix for raw demographic survey items). Gender and education utilized standard demographic categorization established by Ipsos KnowledgePanel. Gender was dichotomized as male or female. Education was recategorized into those who had attended college (i.e., with or without obtaining a degree) versus those who had not attended college. Household income was recategorized into households earning less than versus more than $75,000. Parent status was utilized as a dichotomous variable based on whether respondents had a child of any age.

Knowledge-Attitude-Behavior (KAB) model

To guide the current study, we utilized the KAB health behavior model that has been commonly employed in numerous health promotion settings to understand how social constructs relate to health behaviors [27-30]. Some of these contexts include, but are not limited to, assessments of concussion management behaviors, doping prevention, cardiac patients, and perceptions of elder abuse [30-33]. Despite the limited research into underlying psychosocial constructs present in youth football participatory decisions, past research guided by the KAB model supported the importance of targeting both the knowledge of, and attitudes toward behavior when aiming to develop effective strategies for improving desired health-related behavior [29, 30, 32].

Data analysis procedure

The present study conceptualized respondents as nested within their state of residence or Level-2 variable in our multilevel analysis because states provide stability and meaning to the structure of much of U.S. society through their established regulative, normative, and cultural-cognitive elements. Not only are states the nesting variable of interest in national high school sport participation data [7], but they are also the governance structure that guides competition (e.g., state associations, state legislatures). Inclusion of this nesting variable allowed us to account for the variability of willingness to allow football participation that was due to the respondents from the same geographical area being more similar to each other than respondents from different geographical areas [34].

For descriptive statistics, frequencies and percentages were computed for each categorical variable, and means and standard deviations were calculated for continuous variables. A general population weight was provided by Ipsos and applied to the data to account for different sample selection probabilities. Listwise deletion was used for handling missing data as any incomplete observations were dropped from the analysis (n = 80), which reduced the sample size to 938. Multilevel linear regression models were used to examine independent effects of the predictors on willingness to allow teenagers’ participation in tackle football. We accounted for clustering of respondents within states to avoid inappropriate estimates of standard errors for the model parameters and thus misleading significance tests [34, 35]. Beginning with the null model, a series of sequential models of increasing complexity were performed following the Hox, Moerbeek, and Van de Schoot [34] recommendation for multilevel modeling. The empty model (Model 1) examined the extent to which the respondents’ state explained variation in their willingness to allow football participation. Model 2, a random intercept model, evaluated whether demographic fixed effects (e.g., gender, education) and parent status significantly predicted the outcome. Finally, we assessed whether the addition of the six football safety knowledge and football attitude predictors (Model 3) significantly improved model fit compared to Model 2. Analyses were performed in R 4.0.3 [36] using the lme4 package [37], full maximum likelihood was used for the estimation of these models, and the level of statistical significance was set to alpha = 0.05.

Results

Descriptive statistics

Table 1 shows the descriptive statistics for the nationally representative weighted sample of U.S. adults (n = 938) with respondents included from all 50 states and the District of Columbia. Respondents were evenly split between male and female (51.3% female), and over half of the sample attended college (62.5%).

Table 1

Weighted sample characteristics (n = 938).

Variable	No. (%) or Mean ± SD
Gender
Male	466 (50)
Female	472 (50)
Age	47.8 ± 17.8
Education
≤ High School	320 (34)
≥ College	618 (66)
Household Income
$0 to $74,999	412 (44)
$75,000 +	526 (56)
Parent
No	568 (61)
Yes	370 (39)
Knowledge of football safety measures	3.07 ± 2.4
Worry about head injuries	2.93 ± 1.0
Belief media exaggerates the concussion problem in football	2.25 ± 1.0
Belief rule changes in recent years have made football safer	3.21 ± 0.8
Belief concussions in football are a serious problem	4.24 ± 0.9
Estimate of how many players out of 100 would get a concussion during 1 season of high school football	25.59 ± 25.1
Willingness to allow teenage boys to play football	3.01 ± 1.2

Model results

The state variance estimate in the null model was 0.003; the state residual variance estimate in the null model was 1.32. Thus, the intraclass correlation (ICC) was 0.0023 which means that 0.23% of variance in respondents’ willingness to allow teenagers’ football participation was attributed to between-state differences. Despite the ICC being low, we proceeded with multilevel modeling because of the increased likelihood of overstating findings (i.e., increased Type-I errors) due to failing to account for clustering effects of non-independent data [38]. Additionally, best practice today is to account for the clustering effect [39]. Single-level regression analyses were also conducted, and the results did not differ in any significant way.

Table 2 presents the results of the multilevel linear regression analysis of U.S. adults’ willingness to allow teenagers to participate in tackle football. The multi-level model including the demographic and control variables (i.e., Model 2) produced a significantly better fit than the null model based on the Chi-square test of deviance (χ2(5) = 27.63; p < .001). Additionally, the multilevel model with the six football safety knowledge and football attitude variables (i.e., Model 3) produced a significantly better fit to the data than Model 2 (χ2(7) = 358.07; p < .001). The R2 for Model 2 (i.e., demographic model) was 0.029, whereas the R2 for the final model (Model 3) was 0.340. Lower Akaike information criterion (AIC) and Bayesian information criterion (BIC) values from Model 3 (AIC: 2612.2; BIC: 2680) compared to Model 1 (AIC: 2975.9; BIC: 2990.4) and Model 2 (AIC: 2958.3; BIC: 2997.0) also indicated that Model 3 was the best fitting model. Notably, since the difference in AIC and BIC is greater than 10, there is “very strong” evidence for the more complex Model 3 [40, 41].

Table 2

Willingness to allow high school tackle football participation regression models.

	Model 1		Model 2		Model 3
Variable	β Est (SE)	p value	β Est (SE)	p value	β Est (SE)	p value
Fixed effects
Intercept (constant)	3.02 (0.04)	<0.001	3.35 (0.15)	< .001*	1.62 (0.26)	< .001*
Gender
Male			Reference		Reference
Female			-.36 (0.07)	< .001*	-0.03 (0.07)	.613
Age			-0.002 (0.002)	.309	-.0003 (0.002)	.902
Education
≤ Graduated high school			Reference	-	Reference	-
≥ Attended college			-0.11 (0.08)	.188	-0.07 (0.07)	.279
Household income
$0 to $74,999			Reference	-	Reference	-
$75,000 +			0.06 (0.08)	.465	-0.002 (0.07)	.977
Parent
No			Reference	-	Reference	-
Yes			0.008 (0.08)	.923	0.08 (0.07)	.212
Knowledge of safety measures					0.04 (0.01)	.002*
Worry about head injuries					-0.27 (0.03)	< .001*
Belief media exaggerates the concussion problem					0.38 (0.03)	< .001*
Belief recent rule changes have made football safer					0.37 (0.04)	< .001*
Belief concussions in football are a serious problem					0.02 (0.04)	.607
Estimate of how many players out of 100 would get a concussion during 1 season of high school football					-0.002 (0.001)	.053
Random effect
State residual variance (SD)	1.32 (1.15)		1.28 (1.13)		0.87 (0.93)
Model testing: χ2 (df)			27.63 (5)	<0.001*	358.07 (6)	<0.001*

Note: Outcome variable: willingness to allow high school football participation; β Est = standardized beta coefficients; SE = standard error. Each model was compared to the immediate previous model.

*p < .05.

No significant demographic contributions were maintained in the final model (Model 3). Although females (Est = -.36; SE = .07; p = < .001) were less willing to allow teenage boys to play high school tackle football compared to males in the demographic model (Model 2), this influence was no longer statistically significant once knowledge and attitude variables were added. Notably, being a parent (Est = .08; SE = .07; p = .212) was not associated with willingness to allow teenage boys to play tackle football.

Four of our primary predictors of interest achieved significance and aligned with our a priori hypotheses. Knowledge of football safety measures (Est = .04; SE = .01; p = .002), worry about head injuries (Est = -.27; SE = .03; p < .001), belief the media exaggerates the concussion problem (Est = .38; SE = .03; p < .001), and belief that rule changes in recent years have made football safer (Est = .37; SE = .04; p < .001) were all associated with the outcome. Perceived prevalence of concussion (Est = -.002; SE = .001; p = .053) was not a significant predictor. Contrary to our hypotheses, belief that concussions are a serious problem (Est = .02; SE = .04; p = .554) was not statistically significant.

A sensitivity analysis was conducted by running the same final model on a sample that included only parents and the pattern of results was highly similar. The only difference in any collected variables was that education (Est = -.25; SE = .11; p = .019) became a significant predictor.

Discussion

While substantial national initiatives have begun to make youth sports a public health priority, little is known about the association between adults’ football safety knowledge and football attitudes and willingness to allow adolescent boys to participate in tackle football. Few studies have examined this gap in tackle football and those that have were limited by utilization of convenience samples [17, 18] or the outcome of support for age-based tackling restrictions [19], which is a rule change rather than the more generalizable public health decision of willingness to allow participation. To address this gap in the literature, along with calls for greater empirical attention to be placed on the role of adults in youth sport participatory research [21, 22], we examined the association between adults’ football safety knowledge and football attitudes about issues specific to tackle football and willingness to allow teenage boys to play tackle football, over and above the contribution of demographic factors.

There were three key findings from this study. First, worry about head injury risk and perceived prevalence of concussion were negatively associated with adults’ willingness to allow tackle football participation. Second, belief that rule changes in recent years have made football safer and that the media has exaggerated the problem of concussions in football were both positively associated with the outcome of interest. Third, adults’ knowledge of football safety measures was associated with support for teenage boys playing high school tackle football. This study provides the first nationally representative examination of willingness to allow tackle football participation while extending our understanding of the gap between football safety policy, public perception, and behavior present in U.S. high school football.

In line with the KAB model which proposes that attitudes play an important role in projecting behavior [27, 29, 30], the current study found support for our attitude-based hypothesis that respondents with greater worry about head injury risk would be less likely to allow teenage boys to participate in high school tackle football. Within this hypothesis, three of our five attitude-based variables significantly predicted the outcome of interest, and a fourth attitude-based predictor had a p-value of 0.053. Among these four attitude-based predictors, the two that were negatively associated with willingness were worry about head injuries and perceived concussion prevalence. Adult worry about head injuries is particularly important because the parent-initiated motivational climate plays a critical role in determining youth athletes’ motivation [20-22]. Our finding of a link between adults’ attitudes about head injury risk and willingness to allow football participation provides additional support for the importance of this parent-initiated motivational climate. If adults’ hold more negative attitudes toward head injury risk, then they were less likely to allow participation. Therefore, in hopes of increasing engagement and adherence in football, we may want to place a significant emphasis on addressing the attitudes of prominent adults (e.g., parents, coaches, teachers) toward the actual, rather than perceived, risks and rewards of football participation.

Perceived concussion prevalence was another predictor of decreased willingness to allow tackle football participation. This estimate is noteworthy because it points to the importance of education-based interventions that aim to inform the public on accurate risk involved in football participation. The mean score of perceived concussion prevalence in our study was 25.6 per 100 athletes, which is a largely overestimated assessment of concussion risk in a representative sample of U.S. adults. This is concerning because past research suggested the expected prevalence was approximately 5 per 100 athletes [42] and more current tackle football research published in 2021 supported a rate of 6.99 concussions per 10,000 athletic exposures [43], both of which are well below the 25.6 per 100 athletes estimate seen in this study. It is possible that the overinflated concussion estimate is related to the increased media attention concussions have garnered over the past 15 years [18, 19], which is likely to influence attitudes regarding football participation. Future youth football campaigns may aim to include an educational component targeting the spread of accurate concussion risk information to allow parents to make decisions based on the most accurate data possible.

Our second key finding was that both belief that rule changes in recent years have made football safer and that the media has exaggerated the problem of concussions in football were significantly associated with increased willingness to allow tackle football participation. As there have been a myriad of modifications to tackle football at all levels over the past decade, it is promising for the future of football to see that there is an association between perception of rule changes and willingness to allow participation. While comprehensive evaluation of most rule changes remains scarce [14], identification of this association bolsters the idea of educating the public on rule changes and corresponding injury risk reductions as a strategic option for increasing football participation. The other positively associated attitude predictor was belief that the media has exaggerated the problem of concussions in football. This was in line with our hypothesis and further strengthens the premise of targeting various types of adult attitudes as a strategy to increase youth football participation. Unfortunately, since fear, tragedy, and misinformation spread much faster and more widely than accurate news [44, 45], disseminating information about football safety improvements in the media is likely to continue to be challenging. Taken together, these two significant findings about rule changes and media exaggeration point to the importance of football stakeholders continuing to emphasize evaluation and improvement in their communication strategies to the public. We do not envision the challenges of navigating the media and disseminating accurate safety information going away any time soon as game rule and practice structure guideline changes are likely to become even more important as researchers continue to learn more about the implications of subconcussive head impacts on individuals’ long-term brain health [11, 46–48].

The third key finding from the present study was that greater knowledge of football safety measures was associated with increased willingness to support tackle football participation. This is consistent with the KAB model and supports our hypothesis that respondents who are more familiar with football safety policies are more willing to support tackle football participation. While this finding is significant, the low mean score of 3.07 out of 10 also indicates that the average adult may not have good knowledge of football safety measures. Related to the low knowledge score, past literature found a gap between policy mechanisms (i.e., safety measures) and behavioral outcomes [49] which may be present in the football context as well. In other words, just because safety policies, such as constantly evolving concussion legislation or game rule changes, are enacted that does not mean that stakeholders are aware of them or that they will succeed with their intended impact on behavioral outcomes. Tackle football provides a fitting example of this complex problem as the sport has seen a tremendous amount of evolution in safety policy over the past two decades [14]. However, according to Macy [14], their efficacy often remains unclear and the relatively low knowledge scores may point to a need for increased education of stakeholders (i.e., parents). Policymakers may be well served to invest in efforts designed to educate their target populations of their proposed policies. If they can educate more people on their policies, then our results suggest more people may be willing to allow participation in football, which is one of the primary intents behind the football safety policies.

While our primary aims were related to psychological predictors, the demographic findings warrant brief mention as one finding deviated from past research that found that females were more likely to support tackling restrictions [19]. In our sample, none of the demographic factors were significant in the final model when psychosocial predictors were added to the model. Future studies may aim to examine these demographic predictors more thoroughly, but the explanatory mechanisms are beyond the scope of this study.

While there were substantial findings with practical implications derived from this study, there were also significant limitations. First, we cannot infer causation since this was a cross-sectional assessment. Second, the sample was comprised of adults rather than only parents. It is possible that the study findings would have been different if the national sample was made up of only parents who may have been able to think more realistically about decisions for their children rather than their theoretical children. However, our sensitivity analysis with a subsample of parents produced the same pattern of findings as the full sample. Third, the football safety knowledge quiz was utilized as a proxy measure, may have been too difficult as an assessment tool for comprehensive knowledge of football safety measures. Fourth, while the results from this study point to football safety knowledge and attitudes toward injury risk being important factors related to the outcome variable, there are likely other critical factors, such as normative beliefs, at play. Fifth, a 1-point Likert scale difference in willingness to allow participation is not as interpretable as the difference between yes and no, but it does show greater variance in opinions. Lastly, since the survey item asked specifically about boys, rather than all children, the results are not generalizable to all genders of tackle football participants. Despite this limitation, we anticipate these findings pointing to future research directions targeting adults’ involvement in youth football participatory outcomes for participants of all genders. Future youth football research may be well served to utilize additional health behavior theories that examine additional constructs.

The results point to the importance of educating parents with accurate safety information about the risks associated with high school football participation. In a broader sense, stakeholders may want to target adults’ knowledge and attitudes in their efforts to improve participatory behaviors. Doing so may help unlock all the powerful individual benefits of sport participation [3, 50–54], while creating potentially large-scale public health payoffs that can be supported by public health policy in youth football [10].

Demographic response options.

(DOC)

Click here for additional data file.

All sport based survey items (Part A).

(DOCX)

Click here for additional data file.

Survey items used in the analysis.

(DOCX)

Click here for additional data file.

Likert item distribution.

(DOCX)

Click here for additional data file.

Raw data file.

(CSV)

Click here for additional data file.

11 May 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Jacob Resch, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf.

2. Please provide additional details regarding participant consent. In the Methods section, please ensure that you have specified (1) whether consent was informed and (2) what type you obtained (for instance, written or verbal). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

3. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ.

4. Please include your full ethics statement in the ‘Methods’ section of your manuscript file. In your statement, please include the full name of the IRB or ethics committee who approved or waived your study, as well as whether or not you obtained informed written or verbal consent. If consent was waived for your study, please include this information in your statement as well.

5. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Lines 18-21: Consider breaking this into two sentences, with the 2nd sentence starting as "Little is known...". This will be more direct for readers. Line 20: Please specify knowledge and attitudes here (i.e., football safety knowledge and football attitudes) for clarity.

Line 24: Normally I do not make formatting edits, but believe PLOS One does publish online ahead of print. So, I am providing them where they stick out to provide a cleaner copy for readers. Specifically here, please remove the line break and paragraph indentation for the abstract.

Line 26: Similar comment to above, please consider referring to "knowledge" and "attitudes" here and throughout as "football safety knowledge" or similar for reader understanding. I believe this will help comprehension throughout.

INTRODUCTION: Overall this section provides sound background, rationale, and importance of this study. However, the flow and paragraph use is somewhat confusing, and may be a matter of field/discipline preferences. Edits are provided below to help provide funneling flow.

Lines 40-41: Consider revising to "...CDC reported that only 26% of youth met this recommendation."

Line 42: Consider specifying the year(s) the 46% statistic comes from. Mentioning because we are in a COVID-era and physical activity, specifically sport, has been impacted.

Lines 47-49: Similar comment; please specify year so it's clear to readers.

Line 59: Typo in this sentence at this point; believe changing "have developed" to "have been implemented..." will work.

Lines 78-83: Consider inserting a couple sentences specifically addressing the concerns of repetitive head impacts/concussion occurring in youth sports. For example, there was initial work suggesting football participation before 12 years old was associated with worse late-life cognitive and behavioral outcomes (Stamm et al. 2015; Neurology). Further, there are relatively high head impact magnitudes (e.g. linear and rotational acceleration) and head impact rates (e.g. number of head impacts players experience in a typical practice or competition) experienced in youth tackle football. Lastly, speaking to the national legislation rather than 1 state may provide more implications. For example, there are currently 6 states with active/pending legislation for minimizing contact practice times and/or banning tackle football before a certain age (see https://brainlaw. com/youth-tackle-football/). I believe this will enhance the rationale for "why" this study is needed.

Lines 86-89: Consider adjusting the flow of this introduction so that your research question comes in the final paragraph of the introduction. As is, the research question is provided, and then further rationale behind it is provided. It may be easier for reviewers to be funneled from the overarching context, the underlying rationale, and then finally to the specific research question as is common in many fields.

Lines105-122: This information is a bit long and unnecessary to the majority of readers, and also may be more appropriate in the methods section (since this is justification for a methodological decision). The authors should consider summarizing it into ~2 sentences stating (paraphrasing) that they used a KAB model which has been commonly employed in numerous health promotion applications to understand how unique social constructs relate to health behaviors. As a reader and reviewer, this is much more direct and gets the same point across. In the limitations, you can also acknowledge this decision was used and the mixed evidence surrounding its general applicability.

Lines 123-127: Please remove this information as nothing new is added with this text that is not already mentioned in the intro/methods already.

Lines 127-130: Consider moving these hypotheses right after your specific aims/research question for improved flow.

METHODS: The methods overall are well-written. The reordering of paragraphs and some further details are warranted. Lastly, greater details and justification for the statistical modeling are warranted due to points raised below.

Line 133: Please specify this was a "prospective, cross-sectional survey study." This will be useful information for readers, as well as future research systematically reviewing work on this topic.

Lines 136-137: Please 1) move this text to the data analysis section to better inform readers this was performed, and 2) make sure the language surrounding population weighting matches the text on lines 236-238.

Lines 156-157: Please also add the median score here for further support of the Gaussian distribution.

Lines 150: Please consider: 1) changing the "measures" here to "Survey Outcomes". 2) provide any supporting evidence for the development, reliability, validity of this survey for addressing this research question. As is, it is (I believe) touched upon within each predictor's paragraph by specifying a Cronbach's alpha and other text, but believe a standalone paragraph with all this information (likely at the end of the survey outcome section) would be more appropriate. 3) state the order in which all the survey items/sections were administrated so it could be replicated in the future. It would be beneficial to have a couple lines of text explicitly stating the order and components of the survey as the "Survey Outcomes" opener, and then provide the survey item paragraphs in the order they were presented to respondents.

Lines 154-155: Can the author please provide their rationale for using a 5-point Likert scale here rather than a binary yes/no for this outcome variable? I foresee this being a point of criticism for interpreting the findings and their implications For example, what do the beta-values presented in Table 2 associating with a 1-point Likert change in willing to allow tackle football participation mean for stakeholders? What is considered an important change? 1 or more-Likert point, any change?

Line 181: Why was the "worry about head injuries" category the only one examined on a 4-point Likert?

Line 176: As mentioned, I assume this is for reliability/agreement assessed by the authors for these items, but without specific detail it is unclear. A paragraph targeting the survey reliability/validity would be fruitful.

Line 217: the authors mention "full maximum likelihood here" but also mention R and the lme4 package. Mentioning because lme4 used restricted maximum likelihood by default, and want to confirm accurate reporting. Additionally, this sentence should be moved down towards the end of the paragraph as it is out of place when describing descriptive statistics in the next paragraph.

Lines 207-215: Please move this paragraph into the "Data Analysis Procedure" section as it is not relevant to the outcome measures.

Lines 201-206: Please confirm the dichotomization of these outcomes was through survey respones (i.e. binary response) rather than data processing by the authors. If data processing by the authors, please provide the exact response options for these so this survey could be replicated. The authors should consider modifying the S1 appendix to include the entire survey contents rather than just the predictor/outcome variables for full transparency.

Line 233: Please revise "p<0.05" to "alpha=0.05" because your threshold is a set value and the p-value threshold is set by your alpha value.

RESULTS: Findings are accurately reported in relation to the models ran. However, concerns are raised about the model appropriateness given the negligible variance attributed to the state levels. Further details are provided below.

Lines 243-250: The authors report the increased likelihood of Type I error as justification for using the multi-level model. However, the ICC value of 0.0023 is considerably smaller than those reported by the referenced simulation study, whose minimum ICC value used was 0.01 (over 4x larger). Without data supporting this decision to include such a small ICC and variance associated with the state level (0.003), it does not appear mulit-level modeling is necessary here. This may also increase your likelihood for Type II error. Please consider and revise if necessary.

Table 1. Please describe how fractions of a frequency are being derived, specifically for the demographic counts. Is this due to the national weighting applied? Please confirm this level of decimal precision is needed, otherwise, please round to nearest whole number for frequencies to avoid confusion.

Table 1. Though some of the predictors/outcomes appear to be normally distributed, it may be optimal to also report their median and range, or the frequency (%) for each Likert category for full transparency.

Table 2. Please remove the "note" section (leave abbreviations) in this table, and instead change the table title to "Willingness to Allow High School Tackle Football Participation Regression Models.

DISCUSSION: The discussion section provides accurate interpretation of the findings observed. However, the noted statistical model concerns have the potential to alter the findings. Thus only broad review is provided here as a 2nd round of review is warranted.

Lines 327-328: The reference indicating 5 concussions per 100 athletes is an outdated (published in 2000) sport epidemiology report. Sport concussion prevalence is considerably higher now, though is not presented as # per athlete as this fails to account for athlete-exposure. Please see publication by Chandran et al. 2021 in the American Journal of Sports Medicine for more current concussion rates.

Line 336: Here and throughout, the directionality of the outcome variable association should be provided (e.g. ".... were significantly associated with [increase? decrease?] levels of football participation willingness."

Lines 359-360: The authors should also consider that the survey used may have been too difficult and thus not an accurate measure of football safety knowledge. As a concussion and head impact biomechanics researcher, I would have only gotten 8/10 items correct.

Lines 398-403: Consider either removing or adding this note text to where it is referenced. Superscripted notes are often ignored or missed by readers, and thus should be placed above where applicable.

Reviewer #2: General Feedback:

Thank you for the opportunity to review this manuscript. It was very well written, easy to follow, and is an important contribution to the literature. I have suggested only minor revisions as detailed below.

Specific Feedback:

Introduction –

Very well written; clear and easy to follow. The introduction develops the important narrative necessary for background understanding of the present study.

Methods –

Would the authors please report the number of individuals that were contacted by Ipsos to complete the survey so that the reader can gauge the response rate? If that information is not available, please say so.

The authors should also provide details on the length of the survey (e.g., was it just the questions for this study or were there other items included for other studies?).

Further, would the authors please report whether there were key demographic factors that were associated with missing data on the received surveys? For example, were older individuals less likely to complete the instrument?

Why were the worry-based items graded on a 4-point Likert scale when the belief-based items were rated on 5-point Likert scales?

Would the authors please clarify if the education level of “attended college (62.5%)” refers only to those who completed a degree program, or if it includes those who may have attended some college without obtaining a post-secondary degree?

Why was household income dichotomized at $75k? Was this a median split?

Results –

For transparency’s sake, I would encourage the authors to also report the frequency of each Likert scale response to all questions, either in Table 1 or in supplemental materials. I understand this is asking for some tedious work, but feel it may be useful for yours and others’ future work that may utilize similar questionnaires. I apologize if this has already been done; I was unable to successfully open the link to the appendix in the review pdf file.

I’m assuming the beta values in Table 2 are standardized. Would you please make this clear in the table note and in the main text where “Est = “ is presented?

Discussion –

The authors note that one possible explanation for overestimated concussion risk is that it might be associated with increased media attention. Were the associations between independent variables evaluated? Specifically, does perception of greater media exaggeration associate with lower perceived concussion risk in your sample? On a related note, did the authors consider interaction effects among independent variables (e.g., parentage and worries, for example)?

Thank you for addressing the concern regarding only asking about boys’ participation in football in your note.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Reviewer Feedback_PLOS ONE_willingness to allow teens to play football.docx

Click here for additional data file.

30 May 2022

Thank you so much for taking the time to review this manuscript. We have carefully reviewed each and every one of your comments and addressed them line by line in the 'Response to Reviewer Comments' document and table. Thank you for making this manuscript better through your thoughtful comments. If you'd like any additional revisions, clarifications, or are unhappy with any of our edits/responses please let us know. Thank you again for your time!

Submitted filename: Response to reviewer comments.docx

Click here for additional data file.

13 Jul 2022

14 Jul 2022

Thank you for the comments. Per reviewer and editor comments, we have added the data set to the submission and made edits to align the narrative, research question, results, and discussion. Please see detailed adjustments in either the Response to Reviewer Comments document or the Revised Manuscript with Track Changes document.

Submitted filename: Response to reviewer comments.docx

Click here for additional data file.

5 Aug 2022

Factors associated with U.S. adults' willingness to allow teenagers to play tackle football

PONE-D-21-23107R2

Dear Dr. Kercher,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Jacob Resch, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

11 Aug 2022

PONE-D-21-23107R2

Factors associated with U.S. adults' willingness to allow teenagers to play tackle football

Dear Dr. Kercher:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Jacob Resch

Academic Editor

PLOS ONE