Dr. Damian Sendler Developmental Cognitive Neuroscience
Damian Sendler: An exciting new scientific field has the potential to improve the lives of children through better understanding of brain and cognitive development. We can learn a lot more about how the brain develops and how it functions by conducting large-scale longitudinal neuroimaging studies. If you've ever wondered what questions a study like the HEALthy Brain and Cognitive Development (HBCD) Study might be able to answer, this manuscript is for you. The data should not be used to advance narratives about the development of children from marginalized communities that are de-contextualized and scientifically questionable. A framework for analyzing and interpreting HBCD data will be developed, organized, and built using sampling, cultural context, measurement, and theories from developmental science. With the help of the HBCD's large-scale and collaborative nature, we hope to help the scientific community think critically about the ways in which their work has perpetuated narratives of systemic injustice in developmental science.
Damian Jacob Sendler: Increased attention has been given to America's many systemic injustices over the next two decades, as evidenced by events such as the COVID-19 pandemic and the massive worldwide protests in support of the Black Lives Matter movement. Introspection is warranted in all areas of life, including academia, despite the focus of the past year on healthcare and the criminal justice system. It is our hope that this collaborative effort will provide developmental cognitive neuroscience researchers with an opportunity to reflect on the implications of the study's findings, especially with regard to future HEALthy Brain and Child Development (HBCD) research.
Dr. Sendler: The 21st century is an exciting time for human development research and advancement. This field of study has rapidly built a foundation of knowledge that informs fundamental questions about the drivers of brain change and cognitive development in humans over the last several decades. In tandem with this theoretical progress, multiple safe methods have been developed to investigate genetic, brain and behavioral factors beginning in the womb. NIDA at the National Institutes of Health (NIH) has been working with world-class scientists across the United States to design a prospective examination of brain, cognitive, socioemotional, and behavioral development beginning prenatally and continuing into childhood. Findings from this research could have significant ramifications for children's long-term emotional well-being, particularly for those who have been exposed to drugs or other stressors while still in the womb or shortly after birth.
It is impossible to overestimate the importance of this research. Approximately 7500 pregnant women from across the United States are expected to participate in the study. Brain and cognitive development will be correlated with a wide range of factors including prenatal and postnatal substance exposures to drugs and alcohol (opioids and alcohol), as well as family stress and socioeconomic status, as well as toxic metals in the environment (e.g., lead). Because it will last a decade, the most comprehensive data collection of its kind on early life will be gathered. HBCD will complement an existing study that examines the development of approximately 11,500 children from the age of nine through late adolescence in a similar longitudinal fashion (Adolescent Brain Cognitive Development [ABCD] Study). Despite the fact that the focus of this article is on the upcoming HBCD study, we expect that many of the issues raised here will also apply to the ABCD study. In Garavan et al. (2018) and Simmons et al. (2021) there is a thorough discussion of sampling and design procedures for ABCD.
Even though the overall design of the HBCD study is similar to that of the ABCD study, the goal of the HBCD study is to shed light on how early life exposures increase the risk of negative long-term outcomes in the domains of psychological, behavioral, and neural development.. This is an epidemiological cohort study by definition. Medical and public health researchers frequently use this study design (Carlson and Morrison, 2009; Thiese, 2014), but cognitive scientists and psychologists do not. In order to get the most out of this large project, it is critical to establish early on the limitations and advantages that this research design has to offer. Observational studies have strengths and weaknesses, as well as issues that must be taken into account when reporting data from these studies. Data collection has a social and economic context that must be considered, as well as biases that may be introduced during the analysis of the data when drawing conclusions about the development of children in different parts of the United States. We'll also talk about how these data can be misapplied and distorted to feed prejudices and myths about the development of children in underserved communities. Lastly, we will propose a framework for interpreting HBCD data through the lenses of conscious sampling, cultural context, measurement, and interdisciplinary development science theory
Observational studies of large cohorts are used to investigate correlations or associations between the variables that are being studied. Researchers in the HBCD study used early life experiences (such as stress, adversity, and substance use) as well as tests of physical health, brain development, and cognitive performance to determine the variables to be examined in the study. Even though this study design may seem like a causal framework, with early exposures acting as the cause and measures of brain and behavior being interpreted as the outcomes, observational studies don't have this specificity. Confounding or missing variables are common in complex correlational relationships. If a mother is predisposed to drug use genetically or environmentally during pregnancy, these same factors may influence her child's postnatal development as well.
The true experimental design includes random sampling, manipulation of the independent variable, and measurement of the dependent variable as part of the experiment itself. It is the study's ability to identify these correlational relationships in a large representative cohort that makes it so valuable to the scientific community. The findings from HBCD can help narrow the experimental space for testing hypotheses about the causal relationship between two associated variables. In this way, HBCD data can provide strong evidence for risk and protective factors that must be tested in an experimental setting. The fact that many of the connections that are discovered can't be tested on humans in an ethical manner is noteworthy. It is unethical to administer a teratogen to a pregnant woman in order to study the long-term effects on her child's development. As a result, animal models are becoming increasingly valuable. An experiment based on HBCD findings cannot be designed because human subjects cannot be tested for specific associations, but rather because of the implications for the interpretation of such results, namely the risk of being misinterpreted as reflecting causation.
The conventional reporting and interpretation of observational cohort studies is of the utmost importance. There are variables known as "risk factors" that increase a person's chances of experiencing a certain outcome, while "protective factors" work to mitigate the effects of risk. Keeping in mind that risk can be both relative and absolute is essential. Comparing a subsample of a population that has had an experience or exposure of interest to another that is otherwise identical in terms of confounding variables but has not had this specific exposure provides information on relative risk in medical epidemiology studies. In contrast, the risk of a specific outcome for any single child who has been exposed to a defined exposure is known as absolute risk. According to a relative risk assessment, children who are exposed to teratogens in their early lives are more than three times as likely to suffer from neurological disorders as children who do not face these challenges. As long as there are fewer children with diagnosed neurodevelopmental disorders in the exposure sample than there are children in the total exposure sample, the risk to any one child in that group may still be very low.
Parametric statistics are frequently used in the psychological sciences to document differences among groups (i.e., relying solely on statistical significance or a p value). It is possible to detect small effects by simply increasing the total sample size. As a result of the influence of sample size, the tendency to emphasize statistical significance can be problematic and misleading (Zuo et al., 2019). Because of the large sample size, this is an especially serious problem for HBCD. As a result, even small and possibly insignificant effects may be statistically valid. If the measured differences are actually significant, regardless of the sample size, another critical metric is effect size (Flora, 2020, Serdar et al., 2021, Sullivan and Feinn, 2012). To summarize, when reporting and interpreting results, it is critical to include effect size estimates and both relative and absolute risk scores in addition to statistically significant findings in order to better understand the real-world impact of differences between groups.
We'd like to say one final thing about what large epidemiological studies mean when they talk about "outcome" variables. At the end of the day, these outcomes are typically clinically, socially, or economically relevant variables. Anxiety disorders in adolescents or heart disease in middle age are two examples. In order to evaluate variables like cortical volume or attention span, they must fall outside the normal range of variability. This makes it difficult to evaluate them as outcomes in and of themselves. Scientific interpretations of data may be most difficult when results link experiences and exposures to outcomes that fall within the normal range of variability. In the context of clinically relevant outcomes, such as disease or disorder, absolute and relative risk make a lot of sense.
In part, this is due to the fact that human development does not follow a linear path toward adulthood. Current research suggests that human development is an adaptation to the ecological niche of a child over short timescales, according to the available evidence (Johnson et al., 2015, Werchan and Amso, 2017). Given a limited supply of resources, any organism is likely to prioritize adaptation and survival at the expense of other aspects of its biology. There is no such thing as a "good" or "bad" brain in the context of normal variability. There is only one kind of human variation. Aside from these sub-groups, the distribution of participants into these groups is not completely random. Low-income, Black and Brown, and immigrant communities are disproportionately affected by these kinds of experiences and exposures, without a doubt because of the country's systemic social injustices (Burger and Gochfeld, 2011, French et al., 2020, Williams et al., 2016). So these 'risk factors' are much more likely to be associated with race, ethnicity, and income, increasing the likelihood of finding correlations between variables in an observational cohort study. Because of this, these statistically significant connections could reinforce prejudicial interpretations of marginalized communities' lived experiences and strengthen the systems and structures that support social injustice even further. [page needed] If you're looking at the predictive value of one variable (e.g., race) while also controlling for another (e.g., income), you may have a problem. Mathematically plausible results, however, are most likely the result of de-contextualized interpretations that are significantly less valid and reliable than they appear. On top of that, these analyses tend to ignore a wide range of sociocultural and policy-based lived experiences of American children and families. While this may be the case, there is a tendency to interpret the relationship between experience and outcome as a fixed maturational process and through the lens of common scientific biases, which can be problematic. To avoid the pitfalls of biased interpretations and maximize the usefulness of this potentially invaluable data set, we review common scientific biases in the following sections.
Sample size, generalizability, normativity, and theory building are all examples of interrelated scientific biases (see Hruschka et al., 2018). It is our firm belief that any bias in science renders the science inaccurate and invalid, and we want to make that clear at the outset. We'll talk about everything from threats to measurement validity to internal validity biases (sampling, measurement, normativity). This article's context is social justice, but the quality of science is also at stake, as we want to make sure the reader knows. A lack of attention to these biases could jeopardize the usefulness of the information obtained, which could have a negative impact. We're not advocating that all studies use large, representative samples; there are many good reasons to use smaller, more specialized samples in certain studies. When it comes to theory-building, we encourage researchers to think about how their findings may or may not generalize to a larger population, consider the experiences of populations that are not included in the sample, and most importantly, think about how the context of humanity and the cultural context may adaptively shape development.
HBCD has the potential to recruit a representative sample of American families and shift developmental cognitive neuroscience research towards a more equitable standard of research. Efforts to achieve this standard should not stop at the sampling strategy. W.E.I.R.D. is an acronym for White, affluent families, which are disproportionately represented in developmental science samples (Western, Educated, Industrialized, Rich and Democratic). Scientific biases have been shaped by the lack of diversity in previous research. Developmental journals' sampling methods and interpretative limitations were examined in a review of five of the most high-profile journals in the field (Bornstein et al., 2013). Some 25% to more than 2/3 of the articles examined in this review failed to adequately report on race and ethnicity. Of the published articles, 41.4% omitted or reported ethnic and racial information, including those that only reported the samples were "predominantly white" or "about half minority."
Unbiased research involving human subjects relies on random and representative sampling. We recommend Bonevski et al. (2014) for a comprehensive discussion of various efforts and successes in recruiting underrepresented groups. Representative sampling requires substantial time and resources, as well as local research-community partnerships, according to the authors. Establishing these kinds of collaborations is fraught with difficulties. In the United States, unethical medical practices have a long history of being used on people of color, particularly those of African descent (see Scharf et al., 2010; Washington, 2006). Academic institutions and research are therefore more or less trusted or distrusted based on how much people believe in these things. Even more complicated are the relationships between academia and the communities it serves and the people who live there. Before deciding how to increase the number of people who participate in research, research sites must consider these issues through the eyes and voices of the people they serve. By forming partnerships, we can (a) build genuine trust, (b) improve science communication, (c) lower the barriers to participation that community members face (e.g. time and money), and (d) commit to learning about the actual needs of these communities and, if appropriate, offering services within their developmental science expertise. It's true that a partnership is a two-way street. In the following sections, we propose a theoretical framework that emphasizes the importance of individual and community identities in a way that helps us understand the difficulties in recruiting and participating in research from underrepresented populations.
For example, current practices in big data and open science often focus solely on measurements of internal validity and strict experimental control to make inferences about generalizability that oftentimes ignore cultural context or sociodemographic factors (Hruschka et al., 2018). Direct replication is frequently the focus of efforts to alleviate the replicability crisis, but generalizability is rarely considered. Applied outside of academia, findings with limited generalizability can lead to biased news coverage, policies, and diagnoses. Overgeneralization can be avoided by paying attention to the surrounding cultural context and personal experiences.
Damian Jacob Markiewicz Sendler: To ensure good internal and external validity, the appropriateness of the measure used in the assessment of the child's lived experiences must be considered. Using iPads to conduct cognitive testing on a child who has never been exposed to interactive digital media, for example, may result in results that are skewed. There is a danger in relying on home questionnaires that ask only about books in the home but not about community resources like public libraries, community centers, and so forth. Isolating a child in a room with a stranger, according to Bronfenbrenner (1974), is one of the most unnatural experiences a child can have, and it is likely to be perceived differently by children of different ages and backgrounds. It is important to take into account the family's overall level of comfort with the testing environment and researchers prior to data collection to avoid bias.
To date, many of the measures in our field are normed or primarily tested on a homogenous sample of white middle-class children, because most of the developmental literature has excluded participants from racial/ethnic minority populations (Syed et al., 2018). We must take into account how these measures will affect other American communities. HBCD raises concerns about the accuracy of current measures because of the high proportion of non-white and low-income families that may be affected by early drug exposure and adversity. An example of this would be the case in which a person's developmental trajectory may be affected not by their early life experiences but by the validity of the measures used to evaluate their development and performance. Indeed, the lack of measurement invariance is a well-known source of measurement bias. The degree to which a psychometric test measures the same construct across different measurements or populations is known as measurement invariance (Putnick and Bornstein, 2016, Wang et al., 2018). HBCD is particularly concerned about measurement noninvariance. Differential item functioning is a form of noninvariance that can occur between groups (DIF). Gender, language, education level, race, and other subgroup characteristics, in addition to the study's main construct (Gibbons et al. 2011, Martinková et al. 2017), are assessed by these psychometric test items. A lack of confidence in the validity of the group differences results. Using measures that have been normed for one group (e.g., affluent White children) but administering them to another (e.g., lower-income and/or non-White children) group is a real risk for studies (Lewis et al., 2012). A comprehensive review of measurement invariance issues in developmental science, as well as best practices for statistically testing and reporting measurement invariance, is provided by Putnik and Bornstein (2016). Statistical methods for identifying specific types of DIF and methods for dealing with DIF when it occurs during data analysis are discussed in detail in papers written by Andrich and Hagquist (Andrich and Hagquist, 2012, Hagquist, 2019, Hagquist and Andrich, 2017). These questions should be addressed at the outset of any HBCD study planning. For outsiders, these issues become ensconced in a false sense of security once data are collected and made widely available.
As a result, HBCD data collection sites and future data users should think carefully about measurement biases. According to HBCD, one of the most important objectives is to learn and record how children's development changes with age and exposure to different environments. Over time and across different study sites, HBCD is designed to use the same measures. Finally, when deciding on which measures to include, it is critical to make sure that a representative sample was used to standardize them and to take into account how closely the standardization sample's demographics match up with those of the current test sample. Community partnerships and liaisons, like sampling bias, can be a tremendous resource for understanding the lived experiences and challenges of communities as well as evaluating the appropriateness of measures and data interpretation in context.
Damian Sendler
By 'othering' important aspects of non-dominant environmental realities, the normativity bias obscures important aspects of dominant group experiences through overgeneralized interpretations of data derived from narrow sampling and measurement that lacks broad validity. Debt models contribute to biases in development narratives because they focus on children in non-dominant cultures, rather than contextual demands that are relevant to the child, instead of focusing on their development. harshness and predictability are two key dimensions that shape development and can be used to characterize environmental experiences that both positively and negatively impact an organism, for example (Belsky et al., 2012, Ellis et al., 2009). Hardness is defined as how often external factors (e.g., the frequency of housing relocations) cause disability and death; predictability is the degree of environmental stochastic variation in environmental conditions that govern development (Ellis et al., 2009). While high levels of harshness and unpredictability can lead to stressful situations for children, the effects of stress can be different for each child. Adaptive cognitive development occurs when the stressor can be avoided or additional resources can be accessed. Deficit models fail to take into account this stress-adapted trajectory. Deficit models create erroneous narratives about communities that are already underserved, further marginalizing the children who live there. It is possible that children from different populations are subjected to different levels of harshness and predictability that influence their development in a positive and negative manner.
Damien Sendler: Interventions based on theories that are based solely on the experiences of children from homes that do not experience different levels of adversity and stressors are rendered ineffective. Researchers' theoretical frameworks may be based on an incomplete understanding of critical spectra (i.e., harshness, unpredictability) even when they consider variation in experience, as the vast majority of research relies on the experiences of culturally, racially, and socioeconomically homogenous samples (Bornstein et al., 2013, Syed et al., 2018). While theory building is essential, it is only possible if children's lived experiences are accurately and consistently conceptualized. "Interpretative power," or the ability to recognize and account for an individual's experiences and behaviors within the context of their cultural context, is a critical skill for scientists to acquire" (Brady et al., 2018). As demonstrated by critical race theory, this interpretative power approach emphasizes the importance of identity and experience (see Delgado and Stefancic, 2017). An interdisciplinary approach to developmental science interpretation will be introduced in the following section, which takes into account an individual's cultural context as well as the environmental exposures that accompany it.
As distinct organisms occupying various ecological niches, infants, children, and adults are conceptualized as such in ecological accounts of development (Bronfenbrenner, 1974, Werchan and Amso, 2017). How does an organism's current stage of ontogeny and the context within which it is developing benefit from a particular variation in development? The aforementioned deficit models portray early life adversity as detrimental to development and capture the long-term "consequences" of chronic early-life stress (Rifkin-Graboi et al., 2021). Longitudinal research suggests that context influences how stress-developed cognitive functions perform over time (Hackman et al., 2015). (Nederhof et al., 2014, Nederhof and Schmidt, 2012, Young et al., 2018). It is possible to identify situations in which individual differences in brain structure and cognitive functioning may be beneficial to the well-being of the child during times of adversity, as evidenced by research that takes an alternative approach to the subject (Frankenhuis and Nettle, 2020). As a reminder, we're not saying that adversity is always a good thing. Instead, we propose that one can acknowledge the need to alleviate early life stressors and adversities while also not erroneously judging development in response to the needs of survival in difficult contexts as itself "not good. ". The data on how children respond to stress and adversity must instead be interpreted with objectivity.
For example, consider a data set that shows a correlation between socioeconomic status and executive function. Our understanding of growth is shaped by three alternative explanations, none of which are mutually exclusive. First and foremost, living in a low-income household is detrimental to a child's brain development. This interpretation is based on a deficit model. Alternatively, if we could measure every child on the planet, children in low-income households are the norm, and children in high-income American households are bucking that trend (Amso et al., 2019). As a third aspect of brain development, adaptation to one's own life experiences is essential. An abundance of conventionally enriching opportunities and activities that engage cortical resources may be available to high-income children in this situation Cortical surface area in children who live in under-resourced communities is smaller because they are exposed to fewer of these traditionally defined enriching experiences. However, this doesn't mean they'll have different long-term outcomes because of this. To begin with, all of these children have equally "good" brain development, which means that their cognitive abilities have been shaped to meet the demands of their current environments (Amso, 2020). This means that any child, given the opportunity to participate in a variety of traditionally enriching community programs, would do the same. Current research in this area is hindered because of the overreliance on normative ideology. To put it bluntly, the implicit, scientifically questionable assumption is that data from predominantly white, wealthy American children can define what is normal and good. A history of postcolonial systems and structures that define who and what is valuable in American society is the source of this assumption, not science.
Damian Jacob Sendler
If we go back to our earlier example, bias occurs when the scientific community is compelled to interpret adaptation as a deficit on the part of the lower income group, which may or may not be the case. This outcome, of course, conforms to and reinforces preexisting stereotypes. Conversely, it is less common to view better socioemotional regulation in children with a history of poor maternal caregiving sensitivity as an advantage. Children who have previously had insensitive caregiving experiences tend to perform better on memory tasks with emotional content (Rifkin-Graboi et al., 2021). Children who have experienced poor maternal sensitivity appear to be better at managing socioemotional information in memory than those who have a better socioeconomic status. On a measured timescale, the brain's ability to adapt to this is exactly what it should be, regardless of the circumstances in which it developed.
As a matter of fact, we argue that neither a child with a high income or a low level of caregiving or emotional memory has a clearly superior or inferior brain development. It is impossible to compare children who have had different experiences without using normativity bias. It is important to note that these practices do not provide an opportunity to understand limitations or adaptations in the face of unexplored environmental experiences, alternative types of environmental enrichment, or context relevant cognition (e.g., exhibiting atypical abilities to plan and represent goals at a cognitively "adult" age). The interpretation of developmental trajectories is bolstered by research outside of the field of developmental neuroscience. We can turn to economics and critical race studies for help with this.
We can enhance our interpretive strength by, for example, drawing on economics research. The field of developmental neuroscience may have a blind spot, according to recent economics research. The influence of a family's socioeconomic status on the brain has been overemphasized by developmental neuroscientists when studying how experience shapes the brain (SES). So it tends to disregard the bigger picture, which is the ecological context of the local community (Amso and Lynn, 2017).
As evidenced by findings from the ABCD study (Taylor et al., 2020) and the Opportunity Insights Project (Chetty and Hendren, 2018), children's immediate and long-term outcomes are influenced by environmental contexts outside of their families. Neighborhood poverty has been linked to differences in prefrontal and hippocampal volumes, according to Taylor et al (2020). ABCD data does not tell us if these effects are the result of maturation and adaptation, or if they are permanent or reversible. In any case, this is an important question that must be answered by HBCD. What stage of development do the findings we've uncovered represent? If a child grows up in a high-opportunity neighborhood, regardless of his or her family's socioeconomic status, he or she will be more likely to have intergenerational mobility and a higher adult income (Chetty et al., 2016, Chetty and Hendren, 2018). They found that children adapt to an enriching environment if they spend time in a more affluent neighborhood. For Taylor et al(2020) .'s findings to be interpreted as an adaptive response to neighborhood poverty that is reversible, rather than a deficit in brain development that is set in stone at an early age, the economic data must be taken into consideration (Amso, 2020).
Community resources, social policies, criminal justice policies and neighborhood opportunities all have an impact on cognitive development independent of or in conjunction with the household context or specific pre-and postnatal exposures. HBCD needs to further investigate this issue. As an example, there are data on people who are considered to have "beaten the odds" of adversity, or who are considered to be "resilient" (Ellis et al., 2017, Masten, 2014). We contend, however, that the odds themselves differ according to the individual's socio-political identities. Critical race studies can shed light on these odds by focusing on intersectionality. Kimberlé Crenshaw (Crenshaw, 1991) coined the term "intersectionality" to describe the idea that each person carries multiple political and social identities (e.g., race, gender), that these identities intersect, and that they can be a powerful force in creating opportunities for both oppression and advantages. As previously mentioned, racial and gender differences moderate the impact of neighborhoods on intergenerational mobility. As a result of contextual influences of systemic racism, Economists have found that Black males do not get the same benefit from high opportunity neighborhoods as other groups (Chetty et al., 2014).
Furthermore, Crenshaw argues that our identities are compounded and are both shaped by and actively shaped by our experiences. However, in actual scientific practice, these identifiers are frequently treated as random variables. For example, racism and socioeconomic inequality are inextricably linked to the exposures that children from marginalized communities often experience (Amso and Lynn, 2017, Frankenhuis and Nettle, 2020). HBCD research and theory development must take into account these intersections. Rather than simply classifying these social and political identities, we challenge scientists to investigate the ways in which they work together to shape development.
Researchers will need to examine how the challenges of opioid exposure, for example, are not isolated, but also how socio-political identities like race or residential status (e.g., residing in a rural or more urban area) further compound and complicate this particular childhood stressor/adversity. Understanding the link between prenatal opioid use and later child outcomes requires an understanding of how structural systems differentially offer treatment versus incarceration based on socio-political identities for opioid use. In one predominantly White community, parents struggling with opioid addiction are offered treatment, while in neighboring predominantly Black communities, parents may be sentenced to prison, depriving their children of a caregiver. This allows for a different interpretation of data (Hansen et al., 2020, James and Jordan, 2018, Saloner and Cook, 2013, Santoro and Santoro, 2018). We can get a clearer picture of the developmental challenges that children in the United States face when we combine these facts with what we already know about the importance of mechanisms like maternal/parental modulation. Children from jurisdictions that don't offer recovery services can be "othered" if we try to understand how they adjust. It is possible to develop a theory of how these conditions can shape children by examining the diversity of lived experiences.
ABCD and the upcoming HBCD studies offer unique opportunities to gain insight into the conditions and factors that influence the development of children. Researchers should consider the following methodological checkpoints to minimize bias throughout the course of these studies.
Researchers should be aware of sampling and measurement biases in the early stages of protocol development and participant recruitment. As a way to partner with local communities, better understand their lived experiences, increase trust, evaluate sampling and measurement appropriateness, and make necessary protocol adjustments, researchers can use community members as liaisons or even as members of the research team. An immediate red flag in a study's design may be the possibility that local sites may alter testing details in a way that affects the ability to combine data from all sites. We have two points to address this issue. For starters, the study may already have a number of biases due to the reasons outlined in this document. This problem can be solved in two ways. The first is by creating an advisory panel comprised of local sites that can consider protocol adjustment requests from the HBCD board of directors. The panel will be able to make scientifically sound decisions about whether or not the requested shift is appropriate and justifiable..
Researchers should be aware of biases that influence measurement and theory construction when analyzing and interpreting data. We find that avoiding sampling and measurement bias reduces the likelihood of incorrect and de-contextualized data interpretations significantly at the beginning of the process. We encourage researchers to look for measurement invariance and adopt a developmental systems and interdisciplinary approach in order to combat these issues and promote a more equitable and inclusive developmental science. We recognize that this may necessitate time and effort that some HBCD members may not be able to devote to their own research. Including an external ethics board and an external committee composed of relevant social and cultural scientists from historically excluded groups will go a long way to ensure that culturally relevant and data-appropriate measurements and nuanced interpretations are used.