© 2001 International Association for the Evaluation of Educational Achievement (IEA)
What Are the Roles of Homework and Assessment?
The amount of time students spend on homework assignments is an important consideration in examining their opportunity to learn mathematics. Exhibit 6.16 presents the index of teachers emphasis on mathematics homework (emh). Students in the high category had teachers who reported giving relatively long homework assignments (more than 30 minutes) on a relatively frequent basis (at least once or twice a week). Those in the low category had teachers who gave short assignments (less than 30 minutes) relatively infrequently (less than once a week or never). The medium level includes all other combinations of responses. Details from teachers reports about the length and frequency of their homework assignments are found in the reference section in Exhibit R3.11.
The results show substantial variation across countries and Benchmarking entities in the emphasis placed on homework. Together with Italy, Singapore, and the Russian Federation among the comparison countries, the Academy School District had more than half its students in the high category. For the remaining Benchmarking participants, the majority of students were in the medium category. Very few students were in the low category. One notable exception is Japan (34 percent in the low category), where students were more likely to spend extra time in tutoring and special schools than doing homework.(4) There was little relationship between the amount of homework assigned and students performance. Again, lower-performing students may need more homework assignments for remedial reasons.
Since problem-solving activities will potentially be more bene€cial if they can be extended to out-of-class-situations and stretched over a longer time, TIMSS asked teachers how often they assigned homework based on projects and investigations. The data in Exhibit R3.12 in the reference section show that most students (82 percent on average internationally) had teachers that never or rarely gave such homework. Even though teachers in some of the Benchmarking entities reported giving project-based homework more frequently than did teachers internationally, such assignments did not appear to be made very often. The Benchmarking entities where approximately one-third or more of the students were given projects to do as homework at least sometimes were Connecticut, Massachusetts, Oregon, South Carolina, the Jersey City Public Schools, the Miami-Dade County Public Schools, Montgomery County, and the Project SMART Consortium.
One theme in recommendations for educational reform is to make assessment a continuous process that relies on a variety of methods and sources of data, rather than on a few high-stakes tests. Exhibit 6.17 shows teachers reports about the weight given to various types of assessment. Teachers in the United States as a whole and in most of the Benchmarking jurisdictions reported placing less weight on informal assessment approaches than did teachers internationally. On average internationally, the most emphasis was placed on students responses in class, which were given quite a lot or a great deal of weight for 77 percent of the students. The next heaviest weight internationally was given to teacher-made tests requiring explanations (67 percent of students on average) and to observations of students (64 percent). While the use of teacher-made tests requiring explanations was similar to the international average in many Benchmarking jurisdictions, students responses in class and observations of students were given less weight in the United States as a whole and in most Benchmarking entities (generally for about half the students or less). Exceptions included Jersey City and Miami-Dade, as well as Chicago to some extent.
Internationally, the least weight reportedly was given to external standardized tests, teacher-made objective tests, and projects or practical exercises. On average across countries, about two-€fths of the students (from 37 to 42 percent) had mathematics teachers who reported giving quite a lot or a great deal of weight to such assessments. Across the Benchmarking entities, generally even less weight than internationally was given to external standardized tests. The jurisdictions more similar to the international average were Michigan, North Carolina, Texas, the Academy School District, and Jersey City.
As shown in Exhibit R3.13 in the reference section, eighth-grade students reported substantial variation in the frequency of testing in mathematics class. On average internationally, students were split about in half, with 57 percent reporting having a quiz or test in class almost always or pretty often and 43 percent reporting such testing only once in a while or never. At least three-fourths of the students reported frequent testing in Belgium (Flemish), Canada, the Russian Federation, and the United States. Across the Benchmarking jurisdictions about 80 to 90 percent of the students reported frequent testing. In contrast, about half or more reported infrequent testing in the Czech Republic, Hong Kong, Italy, Japan, and Korea. Within participating entities, there was a tendency for the most frequent testing to be associated with lower-achieving students. One could argue that these students can least afford time diverted from their instructional program. However, teachers may provide shorter lessons and follow-up quizzes for lower-achieving students to monitor their grasp of the subject matter more closely.
TIMSS 1999 is a project of the International
Boston College, Lynch School of Education