Selected Findings from PISA 2015

U.S. Performance in Science Literacy

• In 2015, average scores of 15-year-olds in science literacy ranged from 556 in Singapore to 332 in the Dominican Republic. The U.S. average score was 496, which was not measurably different than the OECD average of 493. The U.S. average was lower than 18 education systems, higher than 39, and not measurably different than 12 education systems. It was lower than the average in Massachusetts (529), not measurably different than in North Carolina (502), and higher than in Puerto Rico (403) (table S1).
• Education systems varied in the range in performance between their highest performing students and their lowest performing students. The difference in science literacy scores between the highest performing (90th percentile) and low performing (10th percentile) students in the United States was 258 score points, which was not measurably different than the average difference across the OECD countries (247 score points). It was smaller than in 4 education systems, not measurably different than in 26, and larger than in 39 education systems. The U.S. score difference was not measurably different than the gap in Massachusetts (253) or North Carolina (252), but was larger than the gap in Puerto Rico (226) (table S1b, figure S1a).
• The distribution of students across the seven PISA science literacy proficiency levels in 2015 showed that the percentages of top performing 15-year-old students (those scoring at proficiency levels 5 and above) ranged from 24 percent in Singapore to rounding to 0 percent in 10 education systems. Nearly one in ten U.S. 15-year-olds (9 percent) scored at proficiency levels 5 and above, which was not measurably different than the OECD average (8 percent). The U.S. percentage of students at the top levels was lower than in 14 education systems, higher than in 34, and not measurably different than in 15 education systems. The U.S. percentage at the top levels was lower than in Massachusetts (14 percent), and not measurably different than in North Carolina (9 percent) (figure S2a, table S2b).
• The percentages of low performing 15-year-old students (those scoring below proficiency level 2)—which is considered below the baseline of proficiency by the OECD—ranged from 6 percent in Vietnam to 86 percent in the Dominican Republic. In the United States, one in five 15-year-olds (20 percent) scored below level 2 in science literacy, which was not measurably different than the OECD average (21 percent). The percentage of low performers in the United States was higher than in 21 education systems, lower than in 37, and not measurably different than in 11 education systems. It was higher than the percentage in Massachusetts (12 percent), not measurably different than in North Carolina (18 percent), and lower than in Puerto Rico (55 percent) (figure S2a, table S2b).

U.S. Performance in Reading Literacy

• Average scores in reading literacy ranged from 535 in Singapore to 347 in Kosovo and Lebanon. The U.S. average score was 497, which was not measurably different from the OECD average of 493. The U.S. average was lower than 14 education systems, higher than 42 education systems, and not measurably different than 13 education systems. The U.S. average was lower than the U.S. states Massachusetts (527), not measurably different than North Carolina (500) and higher than Puerto Rico (410) (table R2).
• The distribution of student scores in reading literacy showed that the U.S. score difference between the 90th and 10th percentiles (259 points) was larger than the difference in 30 education systems, not measurably different than in 30 other systems, and smaller than in 9. The U.S. difference between the 90th and 10th percentiles was not measurably different than in Massachusetts (243), North Carolina (251), and Puerto Rico (250) (figure R1a, table R1b).
• Percentages of top performing 15-year-old students (those scoring at level 5 or above) in reading literacy ranged from 18 percent in Singapore and 14 percent in Canada to nearly 0 percent in Thailand, Mexico, Peru and Jordan. In the United States, 10 percent of U.S. 15-year-old students scored at proficiency level 5 or above, which was higher than the OECD average of 8 percent. The U.S. percentage was lower than 7 education systems, higher than 30 education systems, and not measurably different than 17 education systems. The percentage of top performers in reading in the United States overall (8 percent) was higher than the state of Puerto Rico (1 percent), not measurably different than North Carolina (10 percent) and lower than Massachusetts (15 percent) (figure R2a, table R2b).
• In reading literacy, the percentage of 15-year-old students performing below level 2, which is considered a baseline of proficiency by the OECD, ranged from 9 percent in Hong Kong (China) to 79 percent in Algeria. In the United States, 19 percent of U.S. 15-year-old students scored below level 2, which was not measurably different from the OECD average of 20 percent. The U.S. percentage was higher than 13 education systems, lower than 37 education systems, and not measurably different than 19 education systems. The percentage of low performers in reading in the United States overall (19 percent) was higher than the state of Massachusetts (11 percent), but not measurably different than North Carolina (18 percent) and lower than Puerto Rico (17 percent) (figure R2a, table R2b).

U.S. Performance in Mathematics Literacy

• Average scores in mathematics literacy ranged from 564 in Singapore to 328 in Dominican Republic. The U.S. average score was 470, which was lower than the OECD average of 490. The U.S. average was lower than 36 education systems, higher than 28 education systems, and not measurably different than 5 education systems. The U.S. average was lower than the states of Massachusetts (500), not measurably different than North Carolina (471) and higher than Puerto Rico (378) (table M1).
• The difference in mathematics literacy scores between students at the 90th percentile of performance and those at the 10th percentile in the United States was 230 points, which was larger than the difference in 15 education systems, smaller than in 20, and not measurably different than in 34 education systems. The U.S. difference was not measurably different than the difference between the highest and lowest performers in Massachusetts (218) or North Carolina (225) but was larger than in Puerto Rico (197) (figure M1a, table M1b).
• Percentages of top performing 15-year-old students (those scoring at level 5 or above) in mathematics literacy ranged from 35 percent in Singapore to nearly 0 percent in Peru, Mexico, Colombia, Costa Rica and Jordan. In the United States, 6 percent of 15-year-old students scored at proficiency level 5 or above, which was lower than the OECD average of 11 percent. The U.S. percentage was lower than 36 education systems, higher than 23 education systems, and not measurably different than 6 education systems. The percentage of top performers in mathematics in the United States overall (6 percent) was lower than Massachusetts (10 percent) and not measurably different than North Carolina (6 percent) (figure M2a, table M2b).
• In mathematics literacy, the percentage of 15-year-old students performing below level 2, which is considered a baseline of proficiency by the OECD, ranged from 6 percent in Macau-China to 91 percent in Dominican Republic. In the United States, 29 percent of 15-year-old students scored below level 2, which was higher than the OECD average of 23 percent. The U.S. percentage was higher than 35 education systems, lower than 28 education systems, and not measurably different than 6 education systems. The percentage of low performers in mathematics in the United States overall (26 percent) was higher than the states of Connecticut (21 percent) and Massachusetts (18 percent), but not measurably different than Florida (30 percent) (figure M2a, table M2b).

Twelve education systems had higher average scores than the United States in all three subjects. The 12 education systems are: Canada, Estonia, Finland, Germany, Hong Kong-China, Ireland, Japan, Macao-China, New Zealand, Republic of Korea, Singapore, and Slovenia. The U.S. state Massachusetts had higher average scores than the United States in all three subjects (tables S1, R1, M1).

U.S. Performance Over Time

The U.S. average science, reading and mathematics literacy scores in 2015 were not measurably different from average scores in for the following years: 2006, 2009 and 2012 for science; 2003, 2009 and 2012 for reading; 2003 and 2006 for mathematics. PISA 2015 mathematics literacy average scores in 2015 were lower compared to average mathematics literacy scores in 2009 and 2012 (table T1).

Top