Time spent on math and science homework linked to higher standardized test scores, but not higher grades

Time Spent on Math and Science Homework Linked to Higher Standardized Test Scores, But Not Higher Grades

“The results clearly indicate that Time on homework does not have a significant association with Final grade in these analyses” (61).

The homework debate has continued in the United States for well over one hundred years. How much homework should teachers give? How should teachers assess homework (for accuracy or completion)? What is the purpose of homework? In their paper, “When is Homework Worth the Time?: Evaluating the Association Between Homework and Achievement in High School Science and Math,” authors Adam V. Maltese et al. remark that over the decades homework has been seen as a way to pull students toward academic mastery, toward educational preparation and competitiveness in increasingly globalized marketplaces. They note that, “After the Russians first sent a mission into space, there was a general feeling in America that students were under-prepared, and homework was seen as a tool to improve the educational preparation of students and ensure America’s safety and development” (53).

Studies about the time that students are required to spend on homework are inconclusive. For example, the authors note that the National Assessment of Educational Progress concluded that as of 2003, only 10-12% of students reported having two or more hours of homework per night. The National Center for Educational Statistics argued that there was an increase of students reporting two or more hours of homework each night from 7% in 1980 to 37% in 2002 (53). Although much media coverage has touted the idea that American high school students are overburdened by homework, the authors of this paper argue that, “statistics do not support the notion that a majority of high school students in the U.S. toil away on homework each evening after school…These data indicate that most of the arguments against homework, which appear in the popular media, may originate from a vocal minority” (53).

There are undoubtedly benefits of completing homework. And most previous studies identified a positive relationship between the amount of homework and student academic achievement. For instance, Cooper et al. “conducted a meta analysis of the results from the studies and concluded that the amount of homework completed by students has a significant and positive relationship with student achievement and noted that this relationship is greatest for secondary students” (54).

However, Maltese et al. argue that there are design-flaws with other cited studies. Namely, previous studies tend to link homework to summed assessment scores, such as overall GPAs: “It appears that none of the studies directly evaluated the amount of homework reported by students for specific courses and their grades or achievement results in those classes” (55). So, for this paper, Maltese, Tai, and Fan investigate the following question: “[W]hat is the association between time spent on homework and the academic achievement measure of final course grades and standardized test scores for students in Grade 10 science or mathematics classes?” (55).

Their study uses data collected from the National Education Longitudinal Study (NELS) and the National Center For Educational Statistics (NCES), which collected extensive data on U.S. science and math education. In particular, the authors use NELS and ELS surveys about homework from students enrolled in 10th grade science and math courses. These surveys asked students to report how many hours per week they spent both in school and out of school working on science and math homework. To gain a better understanding of the link between homework and academic achievement in a given course, only students who were enrolled in a single math course or single science course were included in the study (58). (The authors include a lengthy discussion of the (fairly technical) accounting measures taken to ensure the most accurate results for this study. Among other things, they note that student demographic information such as gender, age, and parental education levels, were taken into account.) The authors then looked at final course grades as marked on official transcripts for 10th grade science and math students and as well as standardized exam scores (58).

The Findings

  • “These results clearly indicate that Time on homework does not have a significant association with Final grade in these analyses” (61)
  • In 2003 and also in 2007, approximately 70% of U.S. science teachers and 77% of math teachers reported using homework as part of grading system (66)
  • NELS (National Education Longitudinal Study): surveyed science students completed 83 hours of homework per year
  • NELS: surveyed math students completed 93 hours of homework per year (67)

The ELS surveyed math students completed 150 hours of homework per year (67).

  • Together, the above hours of time spent on science and math homework translate to between 100 and 180 extra 50-minute “classes”
  • Students who report either very low levels of homework or very high levels of homework had the lowest course grades and lowest test scores

There is no consistent link between time on homework and final course grades

  • Overall grades reflect effort and achievement; class participation, attendance, group work, etc. may mitigate effect of homework on overall grade

“[T]here is a strong positive association between time on homework and standardized test scores” (65).

  • Time spent on homework has a stronger positive link to higher standardized test scores than with grades
    • The “style” of homework itself may prepare students for standardized tests
    • In 2007, 81% of math teachers reported that they “Always or Almost Always” assigned homework that involved problems of question sets. Only 16% of science teachers reported the same (67).
  • Students who spent 1-60 minutes per day on math and science homework scored 1.8-2.2 points higher on the standardized test than students who reported doing no homework (61).

Students who spent 61-120 minutes on math and science homework per day scored 2.9-3.0 points higher on the standardized test than students who reported doing no homework (61).

  • Overall findings “indicate a very modest association between homework and achievement” (68)

“[G]iven the size of the achievement gains for the amount of time spent on homework, we think there is under-realized potential in using homework to effectively impact student learning” (68).


Paper Title:  When is Homework Worth the Time?: Evaluating the Association Between Homework and Achievement in High School Science and Math

Authors: Adam V. Maltese (Indiana University), Robert H. Tai (University of Virginia), and Xitao Fan (University of Macau)

Full Paperhttps://eric.ed.gov/?id=EJ995290

Published: High School Journal, Volume 96, No. 1, 2012, Pages 52-72