Policymakers, legislators, and educators in the United States continue to strive to improve K–12 STEM education. Access to high-quality STEM education for all students and students’ strong performance in STEM subjects are necessary for achieving and maintaining the STEM proficiency needed for economic growth, international competitiveness, and scientific literacy (Bush 2019; Committee on STEM Education 2018; NSB 2020). This section presents indicators of U.S. students’ performance in STEM subjects in elementary and secondary school, beginning with performance in mathematics in fourth and eighth grades. Next, it examines mathematics and science performance of U.S. 15-year-olds in an international context. Finally, it examines U.S. performance in computer science, both nationally and internationally.

Although average scores for the nation’s fourth and eighth graders on a national assessment of mathematics improved from 1990 to 2007 (by 27 points for fourth graders and 18 points for eighth graders), scores have remained essentially stagnant since 2007 (Figure K12-1). There were no statistically significant changes in the 2019 mathematics scores on the National Assessment of Educational Progress (NAEP) relative to scores 12 years ago (2007).^{​NAEP is the largest nationally representative and continuing assessment of what America’s students know and can do in various subject areas. The NAEP mathematics assessment is administered to students in fourth and eighth grade every 2 years and students in 12th grade every 4 years. NAEP mathematics assessment results are reported as average scores on a 0–500 scale for fourth and eighth grades and 0–300 for 12th grade. NAEP reports student performance in two ways: scale scores, and student achievement levels. Regarding scale scores, NAEP states that “a statistically significant scale score that is higher or lower in comparison to an earlier assessment year is reliable evidence that student performance has changed” (NAEP 2018). Regarding student achievement levels, the National Assessment Governing Board (NAGB), an independent board that sets policy for NAEP, has developed three achievement levels, which are determined by score ranges that indicate students’ achievement relative to expected achievement for each grade level. These score levels are: basic (partial mastery of knowledge and skills), proficient (solid academic performance at grade level), and advanced (superior academic performance). NAGB suggests that these levels are subject to refinement, and the results should be interpreted with caution. Because these achievement levels are considered provisional, they are not reported here. More information about NAEP scoring is available here: https://nces.ed.gov/nationsreportcard/guides/scores_achv.aspx.} Furthermore, the gap between high- and low-performing students has increased. The lowest-performing students at each grade level (those scoring within the 10th percentile) posted scores in 2019 that were lower than their scores a decade earlier, whereas the highest-performing students in the fourth and eighth grades (those scoring within the 90th percentile) had improved their scores since 2009 (Table SK12-1).

The NAEP 2019 mathematics data show that scores for Black, Hispanic, Native Hawaiian or Pacific Islander, and American Indian or Alaska Native students persistently lag behind the scores of their White and Asian peers. There are potentially many contributing factors to this persistent lag, including lack of access to high-quality STEM instruction and structural and systemic educational and societal inequities that affect students’ educational experiences and performance (Bowman, Comer, and Johns 2018; Hanushek et al. 2020; Pearman 2020; Reardon, Kalogrides, and Shores 2019). In 2019, Black, Hispanic, Native Hawaiian or Pacific Islander, and American Indian or Alaska Native students scored 18–25 points lower than White students in fourth grade and 24–32 points lower in eighth grade (Figure K12-2). Asian students have consistently outperformed all other groups, posting 2019 scores that were 14 points and 21 points higher than those of White students among fourth and eighth graders, respectively.

Score gaps of 23–42 points also exist at both grade levels between students from low- and high-SES families,^{​SES is indicated by a student’s eligibility for the National School Lunch Program (NSLP), with eligible students classified as low-SES students. Student eligibility for a free lunch program is a less-than-perfect measure of SES (Harwell and LeBeau 2010).} students with and without disabilities, and students who are and are not English language learners (Figure K12-3). Male students slightly outscored female students by 3 points in fourth grade in 2019, but there was no difference in scores between males and females in eighth grade. Gaps up to 10 points appear in comparisons by region and school type, with fourth- and eighth-grade students in the Northeast scoring higher than students in the South and West, and students in suburban schools scoring higher than students in city, town, and rural schools (Figure K12-4).^{​The U.S. Census Bureau divides the United States into four regions: Northeast, Midwest, South, and West. For more information, see https://www2.census.gov/geo/pdfs/maps-data/maps/reference/us_regdiv.pdf.}

Looking more deeply into gender differences, the only statistically significant difference between male and female eighth graders’ scores was among Black students, with females outscoring males by 5 points (Table K12-1). Among fourth graders, however, Black male students and Black female students had similar scores, whereas male students outscored female students among White, Asian, and Hispanic students. High-SES students had higher scores than low-SES students within all racial and ethnic groups at both grade levels, with score differences ranging from 15 points to 29 points.

NAEP mathematics assessment results were recently released for students in 12th grade. These results show similar patterns as the results for fourth and eighth graders. The average score in 2019 was not significantly different from the score when the assessment was last administered in 2015 nor from the score in 2005, and scores for Black, Hispanic, and American Indian or Alaska Native students were lower than scores for White and Asian or Pacific Islander students (Table SK12-1).

The NAEP science assessment was administered to fourth, eighth, and 12th graders in 2019, but results were not available in time for inclusion in this volume of Indicators. The results for the 2015 NAEP science assessment showed that the average NAEP science scores for the nation increased 4 points between 2009 and 2015 in both fourth and eighth grades but did not change significantly in 12th grade. More detailed analysis of the 2015 NAEP science scores can be found in Indicators 2018. State-level data on mathematics and science achievement can be found in State Indicators.

U.S. 15-year-olds rank higher internationally in science literacy than they do in mathematics literacy,^{​In PISA, the assessment of science literacy focuses on students’ ability to engage with science-related issues, and with the ideas of science, as a reflective citizen. It requires students to engage in reasoned discourse about science and technology using their knowledge of facts and theories to explain phenomena scientifically. It also requires students to know the standard methodological procedures and patterns of reasoning used in science to evaluate or design scientific inquiries and interpret evidence. In PISA, mathematics literacy is defined as students’ capacity to formulate, employ, and interpret mathematics in a variety of contexts. It includes reasoning mathematically and using mathematical concepts, procedures, facts, and tools to describe, explain, and predict phenomena.} as shown by their performance on the 2018 Program for International Student Assessment (PISA).^{​PISA measures the performance of 15-year-old students in reading, mathematics, and science. PISA technical standards require that students in the sample be from 15 years and 3 months to 16 years and 2 months at the beginning of the testing period; thus, students in the United States were sampled on the basis of age rather than grade level. The assessment aims to measure students’ ability to apply their knowledge to solving problems. PISA is coordinated by OECD and conducted every 3 years. The assessment was first implemented in 2000 and, as of 2018, includes 79 education systems. Education systems include both countries and selected cities within countries. PISA also provides information about students’ proficiency levels. More information about these levels is available here: https://nces.ed.gov/surveys/pisa/2018technotes-6.asp.} In mathematics, U.S. 15-year-olds in 2018 ranked 25th among 37 OECD countries; compared to the 36 other OECD members, the U.S. average in mathematics literacy for 15-year-olds was lower than the average in 24 education systems, higher than in 6, and not measurably different than in 6 (Figure K12-5). The average score in 2018 was lower than the OECD average and did not measurably change since 2003 (Figure K12-6). In comparison, U.S. students fared better in science, ranking in the 7th position out of 37; compared to the 36 other OECD members, the U.S. average score in science literacy was lower than the average in 6 education systems, higher than in 19, and not measurably different than in 11. The U.S. average science score was higher than the OECD average in 2018 and improved by 13 points between 2006 and 2018.^{​Although it appears that PISA mathematics scores declined by 5 points, from 483 points in 2003 to 478 points in 2018, this score difference is not statistically significant, meaning that the scores are similar.}

Japan, South Korea, Estonia, and the Netherlands were the highest-scoring OECD countries in mathematics in 2018, and Estonia and Japan were the highest scoring in science. PISA also tests students in several city-based and non-OECD-country education systems (Figure K12-7). Among these entities, Beijing, Shanghai, Jiangsu, and Zhejiang (B-S-J-Z) in China, Singapore, and Macau (China) were the highest scorers in both mathematics and science. The scores for cities and small countries in Figure K12-7 are not directly comparable to those of the United States, which has a much larger and more heterogenous education system. These scores are presented to provide context for U.S. performance compared to that of some of its global competitors in science and technology businesses and innovation.

Average PISA scores indicate score gaps between male and female students in many countries, including the United States (Figure K12-8). In the United States, male students outperformed female students in mathematics literacy by 9 points, but there was no measurable difference between male and female students’ scores in science. Among OECD countries, on average, male students outscored female students by 5 points in mathematics, and female students outscored male students by 2 points in science. In such countries as Finland, Norway, Iceland, and Israel, female students outscored male students by substantial margins in both mathematics and science.

The Trends in International Mathematics and Science Study (TIMSS) is another international comparative study that measures trends in mathematics and science achievement in fourth and eighth grades every 4 years. TIMSS is designed to align broadly with mathematics and science curricula in the participating education systems and, therefore, to reflect students’ school-based learning. The United States has participated in every administration of TIMSS since its inception in 1995, including the most recent administration in 2019. TIMSS 2019 data were released too late for inclusion in this edition of Indicators; data on U.S performance in 2019 are available here: TIMSS 2019 U.S. Results.

In the 2018 International Computer and Information Literacy Study (ICILS),^{​ICILS is a computer-based international assessment of eighth-grade students’ capacities “to use information communications technologies (ICT) productively for a range of different purposes, in ways that go beyond a basic use of ICT.” First conducted in 2013, ICILS assessed students’ computer and information literacy (CIL) with an emphasis on the use of computers as information seeking, management, and communication tools. Twenty-one education systems around the world participated in ICILS in 2013. The second cycle of ICILS was administered in 2018 and continued to investigate CIL, with the added international optional component to assess students’ computational thinking (CT) abilities, as well as how these relate to school and out-of-school contexts that support learning. The United States participated in ICILS for the first time in 2018, along with 13 other education systems. Among them, 9 education systems, including the United States, participated in the optional component of CT. ICILS is sponsored by the International Association for the Evaluation of Educational Achievement and is conducted in the United States by the National Center for Education Statistics (NCES) in the Institute of Education Sciences of the U.S. Department of Education.} U.S. eighth-grade students’ average score was higher than the international average on computer and information literacy and was not measurably different from the international average on computational thinking (Figure K12-9).^{​The ICILS average score for both computer information literacy and computational thinking assessments did not include the U.S. score because the U.S. sample did not meet the guidelines of a sample participation rate of 85% needed to be included in the average score. The U.S. participation rate of 77% did meet the requirements for score validity and reporting. For more information, see https://nces.ed.gov/surveys/icils/international-requirements.asp.} Computer and information literacy refers to the ability to use computers effectively in everyday life at home, work, and school, whereas computational thinking refers to the use of computers to solve problems and includes such skills as programming. The United States scored 5th among the 14 education systems that participated in the computer and information literacy assessment and 4th among the 9 education systems that participated in the computational thinking assessment. Internationally, Denmark and South Korea were the highest-scoring countries on both assessments. The city of Moscow had the second-highest score in computer and information literacy. U.S. students outperformed their counterparts in France, Luxembourg, Chile, Italy, Uruguay, and Kazakhstan in computer and information literacy and Germany, Portugal, and Luxembourg in computational thinking.

ICILS also provides information about U.S. eighth-grade student scores by sex, race or ethnicity, and school poverty level. Female students outperformed male students in computer and information literacy by 23 points (531 compared to 508); there was no measurable difference in scores between male students and female students in computational thinking (Figure K12-10). Scores differed far more substantially by school poverty level. U.S. eighth-grade students in schools with less than 10% of students eligible for free or reduced-price lunch outscored students in schools with 75% or more students eligible by 88 points in computer and information literacy (564 compared to 476) and 112 points (557 compared to 444) in computational thinking (Figure K12-11).^{​The difference in computational thinking scores rounds to 112 points (112.35) when the more precise scores of 556.79 and 444.44 are compared.}

As with NAEP scores, ICILS scores by race or ethnicity indicate that U.S. Black, Hispanic, Native Hawaiian or Pacific Islander, and American Indian or Alaska Native students posted scores that lag behind those of their White and Asian counterparts in both computer and information literacy and computational thinking (Figure K12-12). Researchers suggest that contributors to these persistent differences include inequitable access to high-quality computer science instruction and to computer or wireless technology, lack of a culturally relevant curriculum, and societal narratives about who is good at computer science (IEA 2020; Margolis et al. 2017; Vakil 2018). ICILS scores ranged from 563 for Asian students to 470 for American Indian or Alaska Native students.