Interwoven: Achievement in reading, science, and math

This is the sixth entry in the Education Leader’s Guide to Reading Growth, a 7-part series about the key factors that affect reading success—and how reading success affects overall student success.

Struggles in science? Low math scores? Maybe your students need more reading practice.

At first, “reading practice” may seem an odd answer to questions about science and math achievement, but upon closer examination, the interwoven nature of achievement is revealed.

As discussed in the first post in this series, research has shown that students with strong reading skills are much more likely to graduate high school on time and enroll in college. As we’re about to see, multiple studies have also shown that students with strong reading skills are more likely to perform well in science and math, too.

Higher reading achievement, higher science achievement

The connection between reading achievement and science achievement can perhaps be most clearly seen in PISA results, where it’s consistent both geographically and temporally. The Programme for International Student Assessment (PISA) is the world’s largest assessment of teenage students, assessing students in dozens of countries every three years.

One researcher analyzed three different sets of PISA scores, representing more than 800,000 students in more than 50 countries.¹ For the 2000 data set, there was a statistically significant correlation between reading and science achievement in all 43 countries examined, with correlations ranging from 0.675 to 0.916. The average for all countries was 0.840. The United States came with an above-average correlation at 0.884.

For the 2003 data set, the results were similar. The correlation between reading and science achievement was statistically significant in all 41 countries examined, ranging from 0.599 to 0.892, with an average correlation of 0.805. Once again, it was higher in the United States, where the correlation was 0.813.

The correlation appeared once more in the 2006 data set, where all 56 countries had a significant correlation between reading and science achievement, ranging from 0.603 to 0.902. The average correlation (0.819) was higher than it had been in 2003 (0.805), which is notable, as several countries that year piloted a new version of the science assessment that was specifically designed to reduce the reading difficulty or reading load of the questions while retaining the science content. (A correlation is not available for the United States, as there was an error in the country’s 2006 testing and its scores were not included in the analysis.)

In the PISA’s own analysis of the 2006 tests, they found the correlation between the regular PISA science and PISA reading assessments was higher still, at 0.83. The new test—the one designed with a lower reading load—had a smaller, but still quite notable, correlation of 0.73. In comparison, the correlation between general interest in science and student performance was 0.13, and the correlation between enjoyment of science and performance was 0.19.² Even with the revised test, the relationship between reading skill and science achievement was much closer than the one between interest and enjoyment of science and science achievement.

Another study dug deeper into the relationship between reading achievement and science achievement to look more closely at how reading difficulty affects scores. This one analyzed the Progress in International Reading Literacy Study (PIRLS) and the Trends in International Mathematics and Science Study (TIMSS) achievement scores of 185,475 fourth-grade students in 34 countries.³

The researchers categorized questions on the TIMSS assessment by their reading difficulty or reading demand level: low (32% of all questions), medium (38%), or high (30%). They also split students into three equal groups according to their PIRLS scores: an upper tercile, a middle tercile, and a lower tercile.

In all countries, students in the upper tercile of reading achievement answered more science questions correctly, on average, than students in the other two terciles. Students in the lowest reading tercile averaged the fewest number of correct science answers.

This pattern was true regardless of the level of reading difficulty—and the average percent correct for each group did not drop precipitously as the reading demand increased. Students in the upper reading tercile answered 66% of the low reading demand questions correctly and 62% of the high reading demands, a decrease of only 4%. For both the middle and the lower reading terciles, the decrease in average percent correct between low reading demand and high reading demand was only 6%.

In both cases, the connection between reading achievement and science achievement seems to be more than just an artifact of reading difficulty of the science questions, as it remains high even when reading difficulty decreases.

The author who analyzed the PISA scores felt there wasn’t necessarily a direct cause-and-effect relationship between reading and science—that reading comprehension did not directly cause science proficiency, nor did science proficiency cause reading comprehension. Rather, she preferred the explanation that it’s the products of extensive reading practice (e.g., background knowledge, reading strategies, and general vocabulary) that drive higher science proficiency.¹

However, it’s not just that students who are better readers tend to have higher science knowledge. A third study examined these two variables—reading skill and science knowledge—and found some quite surprising results.

This study looked at more than 1,600 high school students (grades 9–12) from suburban, rural, and urban schools. Researchers used separate assessments to measure students’ reading skills and science knowledge. They then compared those to four different measures of science achievement: score on state science test, grade in the science course, score on a multiple-choice science comprehension test, and score on an open-ended science comprehension test.

Across all age groups, both reading skill and science knowledge were significant predictors of all four science achievement measures. However, a further regression analysis found reading skill had a larger effect than science knowledge on three of the four science achievement measures.⁴

The data revealed that reading skill actually helped students compensate for gaps in science knowledge for most measures of science achievement. In fact, for all four achievement measures, students with higher reading skills but lower science knowledge scored as high or higher than students with lower reading skills but higher science knowledge.

Furthermore, while the effect of reading skill was noteworthy for students with lower science knowledge, it was even greater for students with higher science knowledge. It significantly enhanced their performance on all four measures of science achievement. The authors recommend reading—specifically books and science articles—to help students gain more science knowledge.

Higher reading achievement, higher math achievement

In comparison to science—even the hard sciences—math seems like it should involve more numbers and fewer words, so does reading skill still matter?

Several studies say it matters quite a lot.

Returning to the TIMSS and PIRLS analyses, this time comparing reading and math, a familiar pattern emerges.³ Once again, the researchers categorized questions on the TIMSS assessment by their reading difficulty. For the math questions, 35% fell into the low category, 35% in medium, and 31% in high.

Looking at the three reading terciles, the top third of readers answered more math questions correctly than the middle tercile, which in turn had a higher percent correct than those in the lowest tercile. This pattern held across all three levels of reading difficulty in all 34 countries.

The authors noted that, unlike with science, it seemed students with lower reading skills were additionally disadvantaged on the math questions that required more reading. The performance of students in the upper tercile of reading skill changed very little as the reading demand increased—this group averaged 66% correct on both low reading demand and high reading demand questions, dropping slightly to 63% with medium reading demand questions.

There was more variation with the middle reading tercile, whose average percent correct decreased by 3% between low reading demand and high reading demand questions. For the lowest reading tercile, the decrease was even larger at 6%. Further analysis found the inter-tercile difference for high reading demand questions was significantly different from the inter-tercile difference for low reading demand. Based on this data, reading difficulty does play a role in math achievement—but even at the lowest levels of reading difficulty, the connection between high reading achievement and math achievement persists.

Another study of fourth-grade students suggested that “reading may be a necessary and important component in overall math competence and should not be overlooked in drawing conclusions about mathematics skills.”⁵ An analysis showed that reading performance was highly correlated with the two main components of mathematics, computation and applications. Students who performed well in reading tended to perform well in mathematics; students who were not proficient in reading did not perform well on math measures.

A different study followed students from fourth to seventh grades to see how reading achievement and math achievement interacted over time. The results showed that not only were reading and math achievement highly corrected in fourth grade, but that there was a tendency for students with higher initial reading scores to have higher mathematics growth rates over time. Students with higher growth rates in reading also tended to have higher growth rates in mathematics.⁶

The authors concluded that “a solid foundation in reading may facilitate gains in more than just reading and would directly affect the growth of mathematics.”

Looking at these researchers’ recommendations—that reading is a good way to gain science knowledge and that reading may directly affect math growth—it seems that increases in reading practice may affect much more than just reading scores. In fact, students may even see benefits beyond improved school achievement: Studies have shown that reading practice can also enhance students’ empathy for others, self-confidence as readers, motivation to read throughout their lives, and positive attitudes toward reading.⁷

So if you have students who are struggling with low achievement in science or math—or if you have advanced students looking to further boost their science or math performance—then you may want weave in high-quality reading practice as a key element in your intervention or enrichment programs.

What are the tools you need to start infusing high-quality reading practice throughout your educational programs or initiatives? In our last entry in this series, we provide a clear checklist of the must-have resources and must-take action steps for supporting high-quality reading practice.

To read the next post in this series, click the banner below.

References

¹ Cromley, J. G. (2009). Reading achievement and science proficiency: International comparisons from the Programme on International Student Assessment. Reading Psychology, 30(89), 89-118.

² Organization for Economic Cooperation and Development (OECD). (2007). PISA 2006: Science Competencies for Tomorrow’s World, Volume 1: Analysis. Paris: OECD Publications.

³ Martin, M. O., & Mullis, I. V. S. (Eds.). (2013). TIMSS and PIRLS 2011: Relationships among reading, mathematics, and science achievement at the fourth grade—implications for early learning. Chestnut Hill, MA: Boston College.

⁴ O’Reilly, T., & McNamara, D. S. (2007). The impact of science knowledge, reading skill, and reading strategy knowledge on more traditional “high-stakes” measures of high school students’ science achievement. American Educational Research Journal, 44(1), 161-196.

⁵ Thurber, R. S., Shinn, M. R., & Smolkowski, K. (2002). What is measured in mathematics tests? Construct validity of curriculum-based mathematics measures. School Psychology Review, 31(4), 498-513.

⁶ Shin, T., Davison, M. L., Long, J. D., Chan, C. K., & Heistad, D. (2013). Exploring gains in reading and mathematics achievement among regular and exceptional students using growth curve modeling. Learning and Individual Differences, 23(1), 92-100.

⁷ International Reading Association. (2014). Leisure reading: A joint position statement of the International Reading Association, the Canadian Children’s Book Centre, and the National Council of Teachers of English. Retrieved from http://literacyworldwide.org/docs/default-source/where-we-stand/leisure-reading-position-statement.pdf