Helping Teachers Learn to Individualize Instruction

Cook and colleagues (2015) analyzed data from the School and Staffing Survey (SASS) that suggests that fully 41 percent of new elementary school teachers and 44 percent of new secondary school teachers said they were unprepared or only somewhat prepared to differentiate instruction. By way of comparison, the proportions of teachers feeling unprepared or only somewhat prepared to teach their subject matter were just 21 percent and 14 percent for new elementary and secondary school teachers, respectively. Is there some way to help teachers learn how to do this better?

While a majority of American public school teachers receive some form of professional development each year, most of this development is neither very intensive (less than 8 hours per year) nor very effective. For example, Jacob and Lefgren (2004) studied teacher professional development as part of an accountability reform in Chicago, which provided additional training to all teachers in schools with test scores below a specified cutoff. The existence of the cutoff created the opportunity for an unusually rigorous regression discontinuity study, which suggests that the relatively modest in-service training that was provided had no statistically significant impact on student learning outcomes. This finding is consistent with the generally pessimistic assessment of most existing teacher professional development efforts by the U.S. Department of Education's What Works Clearinghouse.

On the other hand, there is some suggestive evidence that more intensive professional development could improve teacher effectiveness. For example, Allen and colleagues (2011) conducted a year-long teacher coaching intervention called My Teaching Partner. Secondary school teachers videotaped themselves delivering a lesson and submitted the videotapes to trained teaching consultants, who viewed the videos and then provided 20- to 30-minute coaching sessions roughly twice a month covering what worked well and what could be improved. The randomized controlled trial found increases in student learning of 0.22 standard deviations. While the cost per teacher of this intervention initially seems high (about $3,700), perhaps a more useful metric is to compare the cost per student from a gain of this magnitude to other candidate interventions. By that standard this intervention is remarkably cost effective.

Under the right conditions, principals can also coach teachers to help them improve their instruction. For example, working in the Houston public schools, Fryer (2017) studied the effect of providing principals with 300 hours of training over 2 years to teach them how to help teachers improve instruction, carried out in a mix of elementary, middle, and high schools. This included training principals on lesson planning and use of data-driven instruction, and having them observe classroom lessons and carry out one-on-one coaching with teachers (in-person analogs in some ways to the online coaching sessions in My Teaching Partner). The intervention improved student test scores in reading and math by between 0.10 and 0.19 standard deviations. While the cost per principal is not trivial, the cost per student for these student-level learning gains is on the order of $10. There would seem to be enormous value in learning more about how to optimally carry out professional development strategies that help teachers better individualize instruction, among other key teaching tasks, and how this compares to other candidate strategies.

Tutoring

Another way to deal with this heterogeneity in student academic needs is through tutoring. For example, Bloom (1984) summarizes a series of randomized controlled trials (RCTs) with K-through-8 students, in which they are taught new subjects (such as cartography and probability), which found that when students were assigned to 1:1 or small-group (3:1) tutoring, test scores increased by fully 2 standard deviations when compared with regular classroom instruction alone. Tutoring also generated large increases in time-on-task, improved student attitudes and interest, increased the amount of feedback and correction between student and instructor (a key characteristic of effective teaching), and ensured that all students received this attention—including those students who were struggling in school (Bloom, 1984). While teachers in regular classrooms tend to focus their attention on students performing at the top third of the achievement distribution, tutoring allows for specialized instruction—focusing on the particular academic needs of each student or a small group of students. In this way, the mismatch between what is taught and what the students need is alleviated.

The challenge for education policy has been that such intensive small-group instruction is very costly, so in practice students lack access to appropriate, high-quality targeted tutoring, especially if their parents cannot afford to procure it for their children. Whereas many public school systems use federal funding from the Title I program (intended to provide resources to help low-income students) to support tutoring programs, most of those programs involve hiring former teachers at salaries that are close to what full-time teachers currently make, resulting in a high hourly cost. The result is that students who are far from grade level and in need of extra help and remediation often get only a very modest amount of tutoring help per week, not nearly enough to bring them up to grade level. As a consequence, they continue to sit in classrooms hearing instruction that is mismatched with what they need.

The “two sigma problem,” as articulated by Bloom (1984), is to identify lower-cost instructional alternatives that are as effective as tutoring. One possible approach is to rethink who is required to deliver small-group instruction; that is, whether we can reduce the labor costs associated with this instructional model by hiring nontraditional instructors who are less costly than traditional teachers. There are encouraging signs of promise for this approach from several RCTs carried out in the Chicago Public Schools (Cook et al., 2015). An RCT with 2,700 male high school students from disadvantaged communities showed that 1 year of tutoring, at a cost of between $2,500 and $3,800 per participant, increased math achievement test scores by 0.19 to 0.31 standard deviations, increased math grades by 0.5 standard deviations, and reduced course failures in math by one-half, in addition to reducing failures in non-math courses. Fryer (2014) found similarly large impacts from this type of tutoring, which relies on instructors who have fewer credentials than regular public school teachers (and hence lower salaries), in the Houston Public Schools.

The approach remains somewhat expensive although if implemented as part of a public school system's regular operations it might not need to provide tutoring to every student every year, since the goal is primarily to help students who have fallen behind so they can re-engage with grade-level classroom instruction. Other challenges beyond cost remain. It may require incorporating tutoring into the school day itself, which raises a number of logistical challenges relative to after-school tutoring.⁶ How to measure and monitor the quality of these tutoring sessions is not yet well understood. And, relatedly, whether tutoring can be delivered on a large scale remains an open question, although this is a generic challenge for most social policies or programs (Banerjee et al., 2017; Davis et al., 2017).

Technology

In principle, some of the challenges of tutoring associated with cost and scale could be addressed by making greater use of technology, though how to achieve that in practice is not yet fully understood. In a developing country context, Banerjee and colleagues (2007) studied the effects of giving elementary school children in India access to 2 hours per week on a computer to play math games that emphasized basic math competencies and were tailored to their level of math skill. The result was gains in math test scores at the end of the academic year on the order of 0.35 to 0.47 standard deviation, although 1 year later these effects had faded to 0.10 standard deviations. Similarly, Muralidharan and colleagues (2019) studied the effects of a technology-assisted after-school instruction program for middle school students in India and found after 4.5 months that there were gains of 0.37 standard deviation in math and 0.23 standard deviation in language (the equivalent of reading/English scores in the United States).

In the United States and other developed countries, Escueta and colleagues (2017) identified 29 RCTs of computer-assisted learning programs, which help tailor the level of instruction to student needs. Two-thirds of these RCTs found beneficial effects on learning, particularly in mathematics, with effect sizes on the order of 0.2 to 0.6 standard deviation. These programs can personalize instruction for each student working on the computer, while also generating data that helps teachers understand where students are and what skills or concepts they are struggling with. In addition to helping personalize instruction, both computer-assisted learning and tutoring have the potential to improve time-on-task by reducing the chances that disruptive behavior by other students in the class impede the instruction.⁷

The advantages of technology are several. Not only is it easier to scale up an automated instructional program than human instruction, but also technology-based systems are increasingly able to adapt to user input, making them more dynamic and responsive to students' learning. Further, because the skills that make people good instructors may be in short supply, as programs increase in size the average quality of instructors is likely to decline unless programs raise wages and hence program costs (Davis et al., 2017). Finally, because today's adolescents are “digital natives,” computer-assisted learning may be a natural adjustment for many.

However, research on implementing such automated systems is sorely lacking, and given the centrality of social experiences to students during this period of life it is important to investigate the advantages and disadvantages of placing students in potentially isolating computer-assisted learning environments. Teachers and technology might best be viewed as complements, not substitutes. For example, teachers may be able to help keep students motivated to continue with computer-assisted instruction by connecting the content to the specific interests of each student, or by helping ensure students' needs are well-matched to the specific computer program's design.

Tracking

In principle, an alternative approach is to reduce the heterogeneity in academic levels and needs within a regular classroom setting by sorting youth together based on how they are performing academically—that is, tracking.⁸ Some of the strongest empirical evidence about the effects of this practice come from developing countries. For example, a large-scale RCT in Kenya found benefits (Duflo et al., 2011), with learning outcomes higher in “tracked” schools for students in both the top and bottom halves of the achievement distribution. The study suggests the benefits of reducing academic mismatch through better-targeted instruction are potentially large enough to outweigh adverse peer effects from having lower-achieving classmates, at least for initially low-performing students.

These encouraging findings for internal tracking in the Kenyan context do not necessarily mean this is the right approach in the U.S. context. One concern with tracking is the potential for tracked students to get stuck in lower academic levels without a means to catch up and take on more challenging work. Moreover, a recent study of mathematics tracking shows that in the United States, high school students in the highest level math courses had a significantly higher math self-concept compared to students in lower level math courses. This pattern held when tracking occurred on a course-by-course basis (versus tracking between schools, which is less common in the United States). The authors concluded that “when students are grouped only for certain courses, they observe the grouping process on an everyday basis and are thus constantly reminded of the relative status of their track” (Chmielewski et al., 2013, p. 948). Tracking may have the adverse effect of widening the gap among students. A student who would want to move to a higher track may find that the students in the higher track have already moved to an even greater command of the content area and vocabulary needed to perform in that area than the struggling student has had an opportunity to access. This is especially concerning in the U.S. context where we find a greater proportion of traditionally underserved students in lower tracks. Another concern is that if teachers prefer to work with higher-achieving students, all else equal, grouping students by academic level (tracking) may expose struggling students to relatively less effective teachers. It could also limit the possibilities for social network equity, since tracking can segregate students in ways that deprive them of social capital and other relational resources (Gamoran and Mare, 1989). Ultimately, whether tracking generates net benefits in the United States remains unclear.

Credentialing

At the very least, the U.S. education sector needs to recognize that college may not be the right fit for all students and needs to find other ways to help students succeed in the labor market—including by providing them with the necessary credentials to succeed. When high schools were initially formed in the late 19th century and postsecondary plans were more circumscribed, a high school diploma signaled to employers that its holder was ready for a job in the surrounding community, largely factory jobs or apprenticeships. However, today, the kinds of skills needed to navigate the job market have broadened markedly, the pathways from high school to jobs that pay a living wage are more heterogeneous, and a greater number of individuals are using high school to prepare for college. A single high school diploma, and the curriculum that goes with it, does not provide the kind of credential that most students need to navigate the job market, which particularly disadvantages low-income students and students who are not planning to attend college.

As noted above, secondary schools today tend to deemphasize vocational education as a purpose-driven pathway through high school meant to prepare youth for direct entry into the workforce (Cass, 2018). Given that only 36 percent of adults between the ages of 25 and 44 in the United States today have a 4-year college degree (U.S. Census Bureau, 2017), it is essential that meaningful high school–based vocational training, which leads to the certification of occupation-specific skills and licensing, be developed to open the door to jobs that would otherwise be inaccessible. Obtaining certification through credentials and licensing offers adolescents opportunities to access valuable and secure labor market positions. Further, credentials and licensure, which ensure skills and knowledge, have been shown to reduce racial inequities in pay for Blacks (Blair and Chung, 2018).

There have been calls for national workforce credentialing systems, many of which emphasize the role of community colleges, to address the growing need for mid-skilled employees (ACT, 2011). Consistent with Porter and Rivkin (2012), who argued that the connections between skills and the economy is local, the Abell Foundation conducted a study of the mid-level job market and high school graduates in Baltimore, Maryland. To understand how to connect high school graduates to mid-skilled jobs that do not require a college degree, the Abell Foundation identified the “best prospect” jobs for youth and matched them with training and certification programs at Baltimore community colleges (Hopkins, 2015). These “best prospect” jobs are mid-skilled occupations that pay a living wage, have future mobility, and require little postsecondary training—often even less than an associates' degree. They found that although there are continuing education and certification programs aligned with the “best prospect” jobs, most high school graduates were unaware of them. Although the data are limited, there was a consistent positive impact on earnings 2 years post-credentialing, especially for those in health care jobs (Hopkins, 2015). Similar to what Porter and Rivkin (2012) found, there are more “best prospect” jobs than qualified employees in Baltimore. Credentialing and the curricula that undergird credentials would ensure that youth today (and in the future) do not emerge from secondary education with a diffuse and unhelpful “one-size-fits-none” high school diploma, ill-prepared to succeed in a changing job market.

While the best way to differentiate educational opportunities for adolescents still remains unclear, there is undoubtedly real value in achieving this goal where possible. Taken together, additional research to better understand how to differentiate these opportunities and more policy attention to ensuring that differential and responsive educational opportunities for all adolescents are implemented could yield high returns.

Decision-Making Skills

To take advantage of the choices and opportunities available to them, adolescents must have ample opportunities to practice sound decision making (Eccles, 2007; Steinberg et al., 2009). The ability to make sound decisions is enhanced by the neurobiological development that occurs during adolescence, resulting in increased “processing speed” and reaction times. Further, cognitive automaticity—the ability to carry out cognitive processes quickly and without effort or intention—which results from myelization and synaptic pruning (see Chapter 2), decreases the cognitive load of everyday decision making, freeing cognitive capacity that can be devoted to more complex thinking (Petanjek et al., 2011; Rakic et al., 1994). This means that as youth mature they are better able to see abstract connections among ideas, and engage in thinking and other cognitive activities with more flexibility and feedback utilization (Vera-Estay et al., 2015).

To make good decisions about schoolwork, as well as in other areas, youth need to practice their emerging abstract reasoning and good decision-making skills and to learn from their mistakes (Byrnes et al., 1999; Fan et al., 2012; Halpern-Felsher and Cauffman, 2001; Steinberg, 2005).

Youth are capable of exercising good decision-making skills, but research also shows that some contexts facilitate mature decision making better than others. Sound decision making can easily be disrupted during adolescence by emotional arousal (Casey et al., 2001; see also Chapter 2). Youth make good decisions when those decisions are not emotionally charged, when they have had prior practice making similar decisions, when the people they admire make similar decisions, and when an immediate “reward” is not offset by a consequence well into the future (Steinberg, 2005).

Practical Knowledge

In addition to these dispositions and skill sets related to decision making, youth need practical knowledge about how to function in a complex social world. With the traditional emphasis on grades and test scores, an additional area of competence that has lagged is the type of practical knowledge that will help adolescents to establish independent adult lives and flourish. Social science research has helped us understand that there are important inequities in youths' exposure to such knowledge outside of the school setting. Lower income youth, for example, are significantly less likely to have basic savings accounts than their more advantaged peers (Friedline et al., 2011). Whereas financial literacy is a primary focus of policies and programs directed at the developing world (Clark et al., 2018), they have diminished in focus in the United States.

For most purposes, adolescents reach “legal” adulthood at the age of 18, and genuine independence in adulthood requires youth to be financially literate. They need the skills to “read, analyze, manage and communicate about personal financial conditions that affect material wellbeing” (Vitt et al., 2000, p. 2). Youth need to be able to understand how to develop and live by a budget, enter contracts for housing leases and mortgages, understand how to negotiate health, auto, and housing insurance, make financial plans, and otherwise engage in economic life (Johnson and Sherraden, 2007; Nussbaum and Sen, 1993; Nussbaum, 2011). Poor financial decisions during this critical transition period can result in high levels of debt and derail youths' ability to meet their basic needs as they continue to mature (Lyons, 2004; Norvilitis et al., 2003). Despite the need to develop these skills to successfully transition into adulthood, schools and other agencies have decreased their emphasis on financial literacy and competency (Johnson and Sherraden, 2007). This is especially concerning considering that adolescence is a time when marketers, retailers, and credit card companies are increasingly targeting youth, enticed by their more than $211 billion in spending power (FONA, 2014).

Many successful programs for improving practical knowledge and financial skills build on “asset theory” (Beverly et al., 2008; Karimli et al., 2015), which is focused more on developing assets rather than income because of the sustaining power of assets in times of income volatility. For adolescents, secondary school experiences that provide “experiential” learning opportunities have been found to yield the most gains in knowledge, intent, and practice (Amagir et al., 2018). Amagir and colleagues (2018) did not find differences in program effectiveness among programs teaching financial literacy as stand-alone offerings and those programs that were integrated into existing curricula in secondary schools. For young adults, participating in financial education programs in college led to increased understanding of financial concepts, the intention to engage in responsible credit card use, the use of budgeting and planning, and fewer compulsive spending decisions (Anderson and Card, 2015; Borden et al., 2008; Bowen and Jones, 2006; Maurer and Lee, 2011).

Adaptability

More than in decades past, adolescents need practice to develop skills in thinking critically and analytically. Whereas in the past, it was only important for a small proportion of our population to become life-long learners, the current and emerging economy requires workers who have developed skill sets applicable to many types of jobs and who are also prepared for continuous learning and problem solving. Throughout their working lives, these skills, dispositions, and ways of engaging the world enable youth to create opportunities and pathways into adulthood.

Relatedly, today's increasingly knowledge-based economy, which yields innovation and entrepreneurship, requires a “mindset” of learning, malleability, and expectation for growth and improvement (Blackwell et al., 2007; Dweck, 2006; Yeager and Dweck, 2012). When young people come to recognize that intelligence is malleable and continues to develop, they are more likely to respond to mistakes and failures as opportunities to learn rather than risks to avoid. These are the kinds of mindsets that succeed in a knowledge-based economy. Adolescent mindsets do not develop in isolation. Schools, teachers, and experiences that encourage opportunities for students to embrace challenges and learn from them support the development of a “growth mindset.”

Psychosocial Skills

A growing understanding of the importance of psychosocial skills, especially in the labor market, highlights the value of having the education sector include them in its focus. There are numerous established approaches for supporting adolescent psychosocial skill development in secondary schools. The select strategies described next have shown some promise, although the evidence base regarding their effectiveness and scalability is still in its infancy.

Project-based Learning. One of the potential ways to support psychosocial development in secondary schools is through project-based learning (PBL). This method differs from regular classroom instruction in that instead of relying on a standard classroom lecture format, students are asked to work on a problem-based project over a sustained period of time. The potential benefits, especially in middle and high school, include leading students to see connections between the classroom and the world, increasing students' ownership of their work, and developing critical thinking skills (Friedlaender and Darling-Hammond, 2007). This is related to the core idea first proposed many years ago by, among others, John Dewey: “Learning by doing.” The potential advantages of PBL relative to standard classroom instruction are in developing additional skills related to creativity, organization, and group collaboration that may be particularly relevant for how the modern economy works and—ideally—to engage students in projects they find interesting (Blumenfeld et al., 1991). This may also help address concerns about the motivation levels of high school students in the United States (National Research Council, 2003).

The effectiveness of PBL remains unclear. Existing studies of PBL implementations tend to focus on comparing student test scores in a district before versus after PBL practices are adopted (Marx et al., 2004; Rivet and Krajcik, 2004). Learning more about the effectiveness of this and other strategies to improve psychosocial skills is an important priority for research.

PBL is not currently widespread in American schools, perhaps because it asks a great deal of teachers that differs from what they learned to do in university or on the job.⁹ Presumably, most teachers already have some lesson plans in place, and given the relatively limited diffusion of PBL most teachers were likely taught in a regular classroom setting themselves. Learning how to plan projects and manage students during this process will require additional work and perhaps training and assistance as well, as will learning how to assess the work product and skills of individual students if they work in groups. More research is needed on the effectiveness of PBL and possibilities for implementation and scalability.

Provision of Feedback. Development of psychosocial skills can also be infused into regular classroom settings, for example by changing how classroom teachers provide feedback to students. Because adolescents tend to be vigilant to feedback, educators have the occasion to harness everyday mistakes as learning opportunities. When mistakes occur in class, on assignments, or during exams, adolescents often draw cues from authority figures on how to respond. If they are not given a chance to reexamine their work or perhaps even a second chance to resubmit their work, it signals that mistakes are evaluative rather than formative (Smeding et al., 2013; Yeager et al., 2013). In other words, without feedback assessments appear to be provided only to sort high-ability from low-ability students, which conveys to students that they are unlikely to be able to learn and improve. On the other hand, when adolescents are provided feedback on mistakes along with assurance that educators recognize their potential and expect improvement, they experience greater motivation to exert effort and embrace challenges.

Promoting Belief in Malleability. Other experimental demonstrations with high school and college students also show it is possible to promote the idea that ability is malleable rather than fixed; students with whom this has been done show better academic outcomes than those in control groups, especially if they all had low levels of previous academic performance (Paunesku et al., 2015; Yeager et al., 2016a, 2016b). The experimental demonstrations reviewed suggest that educational institutions and experiences can be designed, tailored, and facilitated within local contexts to genuinely engage students in ways that encourage learning-growth mindsets.

For example, in a study of ninth graders across 10 high schools in the United States, Yeager and colleagues (2016a) found that a mindset intervention using design thinking improved core course grades for previously low-achieving students and increased the learning-oriented attitudes and beliefs of both low and high performers. As discussed above, mindset interventions are designed to change how students interpret and approach challenges and increase their resilience by exposing and affecting their central beliefs about education and school. Growth mindset interventions seek to explain to students that intelligence can grow with hard work, reframing their struggle with learning as an opportunity to grow rather than a lack of ability. The two-session mindset program used by Yeager and colleagues (2016a) was created through an iterative, user-centered design process.

Promoting Development of Future Identities. Developing their own possible selves and future identities helps young people ascribe meaning to difficult tasks such as schoolwork. As a result, those who have the opportunity to explore and develop the possibilities and pathways for their futures tend to demonstrate more effective self-regulation than those who do not. Experimental studies demonstrate programs that effectively help students to develop motivating future identities with positive consequences for academic trajectories, whether facilitated by trained professionals, by teachers, or even by near-peers (Destin et al., 2018; Horowitz et al., 2018; Oyserman et al., 2006).

Mindfulness and Related Practices. Current evidence suggests favorable academic and psychological outcomes in secondary schools implementing mindfulness practices (Raes et al., 2014; Zenner et al., 2014). Informed by current findings in neuroscience, recommendations for secondary schools include training teens in executive functioning skills (e.g., working memory tasks), providing opportunities for mindfulness activities such as meditation, and deepening self-regulation skills through sequenced, active, focused, and explicit (SAFE) socio-emotional learning programs (Steinberg, 2014). Teacher preparation programs can also serve a critical role in exposing future educators to relevant studies of adolescent brain development so that they are equipped to understand the efficacy of these practices in their classrooms. Furthermore, the same practices may be appropriate and beneficial for learning in certain university classrooms.

Cultivating Meta-Cognition. Finally, meta-cognition itself is a critically important skill for helping youth recognize when their automatic responses, which may be adaptive in most settings, are maladaptive and therefore should be “slowed down” so they can devote more mental energy to conscious reasoning. Several RCTs in Chicago show that it is possible to improve high school graduation rates and reduce rates of violence involvement through such programs (Heller et al., 2017). Similar findings have been found for young men in developing-country contexts (see, e.g., Blattman et al., 2017). Since programs with this focus rely on providers with skills that may be in short supply, it can be a challenge to deliver them effectively at scale. But assuming this problem can be solved, such interventions could be widely scaled, because they require a limited number of contact hours—typically between 10 and 30 total.

Advocates of socio-emotional learning programs suggest, although they are critical, socio-emotional learning programs are often overlooked in secondary schools (AEI/Brookings Working Group on Poverty and Opportunity, 2015; Belfield et al., 2015; Cervone and Cushman, 2015; DePaoli et al., 2017). Given their documented impact on the risk of violence involvement, and the fact that injuries (including from violence) are the leading cause of death to adolescents, incorporating this type of instruction into the health curriculum offered by high schools may yield positive results.

School Start Times

Natural neurological changes in circadian rhythms during adolescence due to hormonal fluctuations cause a misalignment between natural wake times and school schedules (Kirby et al., 2011; Wright et al., 2012; see also Chapter 3). The neurobiological changes lead to wakefulness later in the day and later in the evening, resulting in an accumulation of “sleep debt” and less optimal cognitive functioning during school, especially in the morning hours. Multiple studies document an association between sleep loss and negative academic capacity and performance (Dewald et al., 2010; Shochat et al., 2014). Advanced cognitive capacities for planning and decision making, for which adolescence is a period of rapid growth, may also be impaired by sleep loss (Beebe, 2011; Owens, 2014).

The American Academy of Pediatrics (2014) issued a policy statement in 2014 calling for later school start times as an effective solution for addressing poor sleep hygiene among adolescents. Specifically, it recommended that secondary schools start at 8:30 a.m. or later. While starting the school day later in the morning will not address all the factors contributing to insufficient sleep during adolescence, several studies show that teens get more sleep when school starts later (Owens et al., 2010; Wahlstrom et al., 2014). Moreover, later school start times seem to be associated with improvements in academic achievement (Carrell et al., 2011) and in behavior and mental health, including decreases in teen driving accidents, fewer depressive symptoms, and increased motivation (Danner and Phillips, 2008; Owens et al., 2010; Wahlstrom et al., 2014). Positive outcomes for youth pay off at the societal level: a macroeconomic modeling study found that delaying school start times resulted in significant economic gains at the national level over a relatively short period of time (Hafner et al., 2017).

School districts that have attempted to change the start times have encountered concerns about the impact of doing so on after-school activities (e.g., athletics, employment, caring for younger siblings), student access to public resources (e.g., public library), teacher schedules (e.g., less time with their own families), and the family schedules of students themselves.¹⁰ For school districts, the potential impact on transportation costs is often the primary argument for maintaining current, early start times.¹¹ Given the importance of sleep for adolescents, there would be great value in more research to understand ways to mitigate the costs of later start times, and school district–wide policy changes in school start times provide a naturalistic experimental opportunity to examine associations between adolescent sleep, health, and educational metrics such as absenteeism, grade point average (GPA), and behavioral reports. At the very least, school systems should fully consider the benefits of sleep for adolescents in their planning.

Physical Activity and Nutrition in Schools

Adolescents gain much of their adult height and weight during puberty, and the speed and intensity of growth during this period increases an adolescent's need for energy, protein, and certain micronutrients (Institute of Medicine, 2007). When those needs are not met, adolescents experience slower growth rates, later sexual maturation, lower bone mass, and low body reserves of micronutrients, which can have lifelong health implications (Story et al., 2002). Similarly, adolescence is a critical period for bone mass development, making adequate intake of calcium crucial (Heaney et al., 2000). Between the ages of 9 and 17, adolescents gain about 45 percent of the adult skeleton (Institute of Medicine, 2007; Weaver and Heaney, 2006), and by age 20, they have gained 90 percent of their lifetime bone mass (Santos et al., 2017).

Not only has research shown that physical activity has a host of positive physical and mental health benefits for adolescents, such as reducing symptoms of depression and anxiety and improving mood (Institute of Medicine, 2013), but research has also shown a positive association between physical activity and academic achievement for adolescents (Esteban-Cornejo et al., 2015; Institute of Medicine, 2013; Kristjánsson et al., 2010; Rasberry et al., 2011). Physical activity in youth has also been associated with positive psychosocial traits, including self-efficacy, self-concept and self-worth, social behaviors, and goal-orientation (Institute of Medicine, 2013).

These critical nutritional and physical needs during adolescence call for special environmental supports to encourage healthy physical development. The school day presents numerous opportunities to improve the nutrition and physical activity levels of adolescents. Several federal initiatives aim to improve the nutrition and, to a lesser extent, the physical activity of U.S. adolescents. Most notably, the Healthy, Hunger-Free Kids Act of 2010 (HHFKA) directed the U.S. Department of Agriculture (USDA) to update the nutrition standards for the federal government's core childhood nutrition programs: The National School Lunch Program (NSLP), the School Breakfast Program, and the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), among others. The National School Lunch Program serves 30 million children every school day, many of whom are low-income (U.S. Department of Agriculture, Food and Nutrition Service, 2017). The updated nutrition standards for this program, implemented in 2010 and revised in 2019, require school cafeterias to serve meals that align with the Dietary Guidelines for Americans and regulate the sale of unhealthy foods throughout the school. The goal is to increase the consumption of healthy foods and create a healthy food environment for students. A study of students at 12 middle schools in a low-income, urban school districts found that students consumed more fruit and ate more of their vegetables and entrees after changes to the National School Lunch Program were implemented in their cafeterias (Schwartz et al., 2015).

The Community Preventive Services Task Force (2016) found that meal and snack interventions are effective at increasing fruit and vegetable consumption and reducing or maintaining the rate of obesity and overweight among students (Community Preventive Services Task Force, 2016). These interventions include school meal policies, programs that provide fresh fruits and vegetables to students during lunches and snacks, healthy food marketing initiatives, and nutrition education opportunities that empower students to make healthier choices. In addition, the task force found that combining these interventions into multicomponent interventions was also effective at reducing or maintaining the rate of obesity and overweight in schools.

Federal- and state-level policies also affect adolescents' access to and engagement with physical activity, most clearly through regulations on schools. The majority of states mandate some amount of physical education for students in middle and high schools (80% and 86%, respectively), although the quality and quantity of physical education provided typically fails to meet recommended standards for daily physical activity: only 27 percent of adolescents in the United States met the Centers for Disease Control and Prevention's recommendation of 60 minutes or more of physical activity per day every day of the week in 2016¹² (Carlson et al., 2013; Subcommittee of the President's Council on Fitness, Sports and Nutrition, 2012).

The Physical Activity Guidelines for Americans Midcourse Report Subcommittee of the President's Council on Fitness, Sports, and Nutrition found evidence that multicomponent physical activity interventions in schools can increase physical activity during school hours. These multicomponent interventions, which have been shown to increase physical activity during school hours, typically combine enhanced physical education with other strategies such as health education, classroom physical activity, social marketing initiatives, active transportation to school, and physical environment improvements, among others. These interventions are effective at increasing physical activity during the school day, and when combined with community- and family-based interventions they show strong evidence of increasing physical activity for adolescents outside of school (Subcommittee of the President's Council on Fitness, Sports and Nutrition, 2012).

Trauma-Informed Practices and Mental Health Services

A third example of how recognizing the wholeness of adolescents may change the way schools operate is the recognition that many students, particularly from economically disadvantaged communities, come to school after having experienced significant distress and trauma outside of school, whether due to natural disasters, neighborhood violence, or other adversity. These experiences can affect not only students' health but also their academic achievement. For example, students (including adolescents) who have had a homicide occur within their neighborhood in the recent past exhibit test scores that are one-half to two-thirds of a standard deviation lower than the scores of other teens (Sharkey, 2010). Even youth from advantaged homes and backgrounds come to school with significant unmet mental health needs (Luthar et al., 2013). Among a nationally representative sample of adolescents, fully 40 percent experienced a mental illness over a 12-month period (Kessler et al., 2012). While the mental health services received by youth are most likely to be delivered in the school setting, overall only 45 percent of students with psychiatric disorders receive mental health care (Costello et al., 2014).

The unmet mental health needs of U.S. adolescents pose a key challenge to their ability to succeed in school, something that should be as significant a concern to the education sector as it is to the health sector. There are limits to the role schools can play in resolving this issue, of course, and providing mental health services in the school setting is just one way to help students. Although schools themselves cannot do it all, they can create opportunities for identification and treatment by providing mental health services (Green et al., 2013), although the quality and magnitude of services matters (Paschall and Bersamin, 2018). (See also the discussion of “School-based Health Centers” in Chapter 7). Finally, unmet mental health needs among U.S. adolescents pose a key challenge to their ability to succeed in school, something that should be of as great a concern to the education sector as it is to the health sector.

Safe and Supportive School Environments

Not all youth feel safe and supported while at school. For example, as discussed in the following chapter, weight-based victimization is widespread in schools, and can lead to poorer emotional health outcomes for students of all weight statuses (Buchianneri et al., 2013; Calzo et al., 2012; Puhl et al., 2011; Pont et al., 2017).¹³ In addition, lesbian, gay, bisexual, transgender, queer, and questioning (LGBTQ) students report high rates of discriminatory bullying based on their sexual orientation and their gender identity. A number of negative health behaviors and outcomes for LGBTQ youth have been linked to such negative experiences at school (Russell et al., 2011).

During the past decade, a body of research has identified a number of policies, programs, and practices that help to create a safe and supportive school climate for LGBTQ students and indeed for all students. These include inclusive, enumerated policies; professional development on LGBTQ issues for educators and other professionals in the school setting; LGBTQ-related resources; and the presence of Genders and Sexualities Alliance clubs (GSAs, or Gay-Straight Alliances) in schools (see, e.g., Russell et al., 2010; Walker and Shinn, 2002). Policies that specifically identify sexual orientation and gender identity or gender expression as protected statuses in schools are associated with positive school experiences and health for all youth, including LGBTQ youth (Russell et al., 2010; Hatzenbuehler, 2011). Professional development for educators and other school personnel related to LGBTQ issues can equip them with skills to support all students (Hatzenbuehler, 2011; Kosciw et al., 2016). In addition, there are multiple personal and academic benefits when sexual orientation and gender identity are incorporated into the school curriculum (Black et al., 2012; Snapp et al., 2015). Relatedly, the presence of GSA clubs has been linked with individual student well-being, as well as with positive overall school climate (Ioverno et al., 2016; Poteat et al., 2015; Walls et al., 2009).

Wraparound Services

Addressing the integrated needs of adolescents is critical, especially for youth from socioeconomically disadvantaged backgrounds. While adolescents raised in households toward the top of the socioeconomic hierarchy are likely to have access to high-quality schools, safe and supportive communities, and well-functioning social systems, it is clear that adolescents from socioeconomically disadvantaged backgrounds face special challenges in accessing a well-functioning institutional ecosystem (Jackson, 2019). To this end, schools may come to provide services more commonly provided by other social institutions. These services might include free meals, laundry services, or health care. One of the best-known examples of schools operating in this “wraparound” mode is the Harlem Children's Zone, at the heart of which is a set of well-funded charter schools that aim to create a “pipeline of support” through the provision of food and health care (Harlem Children's Zone 2009, 2010; Tough, 2009).¹⁴ It is still too early to provide a full assessment of the effects of this pipeline approach, although Harlem Children's Zone has shown positive results, particularly with respect to test scores (Hanson 2013; Page and Stone, 2010). Dobbie and Fryer (2011, 2013), for example, examine the effects of the Harlem Children's Zone on student outcomes and find that there are sizable impacts not only on test scores in the short term, but also on longer-term outcomes such as increased college attendance and reduced teen fertility (for females) and incarceration (for males).

Disciplinary Practices

Pressing boundaries, taking risks, and seeking autonomy and independence are hallmark features of adolescence (Steinberg, 2005), yet they often lead to misbehavior and rule violations at schools, especially when coupled with adolescents' general sensitivity to cues of respect or disrespect (Hohnen and Murphy, 2016). These tendencies can have serious consequences if they lead to formal remediation or disciplinary processes. For this reason, it is wise to frame remediation and disciplinary processes in a way that conveys greater respect for adolescents and more effectively encourages and supports academic and behavioral improvement (Okonofua et al., 2016). For example, sending a formal suspension letter to a student emphasizing the positive goals of the suspension is more likely to lead to a successful return to school than sending a purely disciplinary message (Brady et al., 2017; Brady et al., 2018).

A starkly punitive disciplinary orientation is problematic for all youth, but it is all the more harmful when it is administered with an unequal hand. Unfortunately, youth of color, LGBTQ youth, and youth with disabilities are more likely to experience harsh and punitive discipline, even for similar infractions, than other students (Fabelo et al., 2011; Losen and Gillespie, 2012). Having disciplinary policies in schools that are developmentally appropriate may be particularly important given evidence that students of color are disciplined more often and receive more punitive discipline than White students, as noted in Chapter 4. Black students are also more likely than White students to be suspended from school and receive in-school detentions that exclude them from the classroom (Fabelo et al., 2011; Losen and Gillespie, 2012). However, there is no evidence that Black youth, LGBTQ youth, or youth with disabilities violate rules more often or behave worse than other students (Losen and Skiba, 2010; Skiba et al., 2015). The disparities are greatest for infractions that require subjective judgment, such as insubordination, defiance, and disruption (Fabelo et al., 2011; also see Chapter 4 and Box 9-2).

Conflict Prevention and Intervention

In the pluralistic communities our nation sometimes idealizes, diversity is a source of strength and enrichment. However, it is sometimes a source of tension and conflict as well. As a result, contemporary secondary schools are well-advised to undertake systematic efforts to prevent and resolve conflicts among their students (Gregory et al., 2014; Gregory et al., 2017). Conditions that foster conflict prevention include supportive student-educator relationships, academic rigor, culturally responsive teaching, bias-free and respectful classrooms and schools, and opportunities for learning and correcting behavior. Equity-driven principles of conflict intervention include inquiring about the cause of the conflict, problem solving, recognizing student and family voices in identifying causes and solutions, and reintegrating students after conflicts. Finally, a multitiered system of supports that match supports with student needs can contribute to both conflict prevention and intervention. These changes may be enhanced by proactive acknowledgement of racial and cultural inequalities in discipline (Carter et al., 2017; Gregory et al., 2014; Gregory et al., 2017).

Staff and Curriculum

While school populations are increasingly diverse ethnically, racially, and economically, the education workforce is largely White. Indeed, 82 percent of the U.S. education workforce is White. Similarly, only 20 percent of school principals are people of color. In contrast, 51 percent of U.S. school children are of color, and this proportion will increase in the coming decades (Child Trends, 2018).

Because adolescence is a time during which young people are honing their identities, it is important for schools to create learning environments that affirm the cultural backgrounds of all of their students. Mounting evidence suggests benefits for students when they have teachers who share their racial background (Dee, 2004; Egalite et al., 2015; Gershenson et al., 2017); this finding has even been replicated for community college students (Fairlie et al., 2014). For Black students in particular, having just one Black teacher in elementary school is positively related to high school completion and college aspirations (Gershenon et al., 2017). Having a Black teacher is also related to a reduced likelihood of exclusionary discipline such as suspension or expulsion for Black students at all grade levels (Lindsay and Hart, 2017).

The increasing divergence between the demographic backgrounds of secondary school teachers and their students makes the risks of a “hidden curriculum”—one not explicit in textbooks but embedded in the attitudes and behaviors of school staff—even more acute. The concern is that schools may remain places that are fundamentally driven by a single cultural narrative, since the cultural aspects of schooling also matter for students (Carter, 2013). Schools are important contexts for socializing youth, sending messages about appropriate behavior, who is or is not considered “smart” or “successful,” and how to get along with peers and adults (Warikoo and Carter, 2009).

Some studies suggest that more culturally diverse curricula may benefit student outcomes during adolescence. A recent study by Dee and Penner (2017) found that students of various racial/ethnic backgrounds who were assigned to take an ethnic studies course in eighth grade because their GPA was below a certain threshold had increased attendance, a higher GPA, and more credits earned in ninth grade. Similarly, students who took courses in Mexican American studies had higher rates of high school graduation and higher passing rates on state standardized tests compared to students who did not participate in the courses; this was true despite course participants starting off with lower ninth- and tenth-grade GPAs than the nonparticipants (Cabrera et al., 2014). Thus, opportunities for diverse content and exposure to the perspectives of nondominant groups in the curriculum may be beneficial for all adolescents. However, more research is necessary to determine whether there may be any unanticipated effects.

Addressing Stereotypes

As students become more aware of their social identities, including their race, class, gender, and sexuality, schools have an opportunity to positively shape the meaning of those identities rather than reinforcing negative connotations (Umaña-Taylor et al., 2014). However, institutions and educators can sometimes convey, whether intentionally or unintentionally, impressions that students from racial-ethnic minority groups and students from lower-socioeconomic backgrounds have lower academic abilities. Teacher behavior may also reinforce stereotypes that girls have less potential to excel in science and math than boys do. These negative messages can impair students' actual performance and discourage them from pursuing endeavors where they might otherwise be likely to succeed (Murphy et al., 2007). On the other hand, when learning environments positively embrace students' social identities and allow them to consider how their backgrounds may help them to succeed, their sensitivity to social status can serve as a support for motivation and achievement.

Biases held against students from particular social groups can be especially consequential when held by teachers and other school personnel (Tenenbaum and Ruck, 2007). Based on a meta-analysis, teachers hold differential expectations for students based on students' racial background. On average, they have the highest expectations for Asian students and hold the lowest expectations for Black and Latinx students (Tenenbaum and Ruck, 2007). Teachers judge Black and Latinx students to be putting less effort into their studies, compared to White students, even though the students themselves report similar effort (Kozlowski, 2015). Further, teachers have more positive interactions with and make more positive referrals for Whites and Asians, compared to Black and Latinx youth (Tenenbaum and Ruck, 2007). When shown ambiguous facial expressions, teachers interpreted them as expressing anger and made more hostile attributions about those expressions when they were expressed by Black youth than when they were expressed by White youth (Halberstadt et al., 2018).

Biases and stereotypes are accentuated in contexts where an outgroup is easily identified, where there is little or no opportunity for individual interactions, and where there is mixed or confirming evidence regarding a negative stereotype, such as when there are demographic gaps in achievement or tracking that marginalizes an ethnic group (Banaji and Greenwald, 2013; Tajfel and Bruner, 1981). Large secondary schools make it more difficult for teachers to connect with students individually and avoid relying on pre-existing biases and stereotypes. Based on a review of evidence on school size, researchers recommend that secondary schools limit enrollments to between 600 and 1,000 students; the lower end of this range is recommended in particular for schools serving a large proportion of students from communities with limited economic resources (Leithwood and Jantzi, 2009). However, contrary to what students need, 30 percent of high schools have 1,000 or more students (U.S. Department of Education, 2017).

An RCT of small high schools in New York City carried out by the Manpower Demonstration Research Corporation, which exploited randomized admission lotteries to follow students, found that attendance at the small schools increased high school graduation rates by 9.4 percentage points overall (Unterman, 2014). The effects in this study were even larger for Black males, equal to a 12.2 percentage point improvement in graduation rates. The data thereby suggest that small schools may be particularly helpful for adolescents who suffer from stereotypes in larger school settings. Small high schools were also found to increase postsecondary enrollment rates by 8 percentage points¹⁵ although some of the evidence regarding student outcomes in small schools is mixed (Barrow et al., 2015; Wyse et al., 2008).