Skip Navigation
Highlights From the TIMSS 1999 Video Study of Eighth-Grade Science Teaching

NCES 2006-017
April 2006

What Are the Major Findings From the TIMSS 1999 Video Study of Eighth-Grade Science Teaching?

Differences in Science Content

  • Topics: Within the United States, no differences were found between the percentages of eighth-grade science lessons that addressed topics in earth science, life science, physics, chemistry, or other areas (nature of science, technology, environment and resource issues, science and other disciplines) (figure 3). In contrast, physics and chemistry topics were addressed in more Japanese science lessons than earth science and life science. Within Australia, more lessons focused on physics topics than earth science, life science, and chemistry topics. Life science and physics topics were addressed in more Dutch lessons than chemistry, with too few earth science lessons for reliable estimates.

Figure 3. Percentage distribution of eighth-grade science lessons devoted to life science, earth science, physics, chemistry, and other areas, by country: 1999

Percentage distribution of eighth-grade science lessons devoted to life science, earth science, physics, chemistry, and other areas, by country: 1999

‡Reporting standards not met. Too few cases to be reported.
1AUS=Australia; CZE=Czech Republic; JPN=Japan; NLD=Netherlands; and USA=United States.
NOTE: Detail may not sum to 100 because of rounding. Other areas include: interactions of science, technology, and society, nature of scientific knowledge, and science and mathematics.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Third International Mathematics and Science Study (TIMSS), Video Study, 1999.

Top

  • Types of science knowledge: Different types of knowledge were addressed in the science lessons. Comparing across countries, Czech eighth-grade science lessons allocated a larger percentage of public talk time, on average, to address canonical science knowledge (generally accepted scientific facts, ideas, concepts, or theories) than the lessons of the other four countries (figure 4). Across-country differences also were evident in the percentage of public talk time devoted to procedural and experimental knowledge. Japanese science lessons addressed procedural and experimental knowledge (information about how to do science-related practices such as manipulating materials and performing experimental processes) for a larger average percentage of public talk time (25 percent) compared to the lessons of the other countries (between 11 and 17 percent). Further, Japanese lessons allocated a smaller percentage of public talk time (6 percent) to science-related real-life issues (e.g., societal issues or students’ personal experiences) than lessons in all of the other countries Play Video Clip - Netherlands (Flash) except Australia (14, 15, and 17 percent for the Czech Republic, Netherlands, and United States, respectively; 12 percent for Australia; data not shown).


Figure 4. Average percentage of public talk time in eighth-grade science lessons devoted to canonical knowledge, by country: 1999

Average percentage of public talk time in eighth-grade science lessons devoted to canonical knowledge, by country: 1999
1AUS=Australia; CZE=Czech Republic; JPN=Japan; NLD=Netherlands; and USA=United States.
NOTE: CZE>AUS, JPN, NLD, USA; JPN>USA. Analysis is limited to public talk time. The above category was not applied to non-public segments of the lesson because of the nature of independent work and the limitations of the video methodology. During non-public talk segments, students were typically working independently on a set of tasks that may involve different types of knowledge.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Third International Mathematics and Science Study (TIMSS), Video Study, 1999.

Top

  • Challenge and density of science content: In general, Czech lessons contained more content ideas and more challenging ideas than lessons in the other countries with reliable estimates. Based on expert judgments, Czech teachers included a mix of basic and challenging content in 56 percent of the science lessons and mostly challenging content Play Video Clip - Czech Republic (Flash) in 25 percent of lessons (figure 5). In contrast, 47 to 65 percent of the lessons in the other countries included only basic content. In addition, more eighth-grade science lessons in the Czech Republic contained a high density of 15 or more canonical ideas (26 percent) compared to lessons in Japan (7 percent) (data not shown). Other evidence of the higher level of challenge and density of science content in Czech lessons was the inclusion of more unrepeated science terms during public talk in a lesson (56 per lesson, on average), and more unrepeated highly technical science terms during public talk (33 per lesson, on average) compared to science lessons in the other four countries (data not shown). In addition, more Czech science lessons included theoretical ideas (49 percent) than lessons in Japan and the Netherlands (15 and 19 percent, respectively; data not shown).

Figure 5. Percentage distribution of eighth-grade science lessons that were judged to contain challenging content, basic and challenging content, and basic content, by country: 1999.

Percentage distribution of eighth-grade science lessons that were judged to contain challenging content, basic and challenging content, and basic content, by country: 1999
!Interpret data with caution. Estimate is unstable.
1AUS=Australia; CZE=Czech Republic; JPN=Japan; NLD=Netherlands; and USA=United States.
2Challenging content: CZE>JPN.
3Basic and challenging content: CZE>AUS, JPN, USA.
4Basic content: AUS, JPN, NLD, USA>CZE.
NOTE: Totals may not sum to 100 due to rounding. The level of challenge in the science content could not be determined in 3 percent of Dutch lessons and 1 percent each of Australian, Czech, and U.S. lessons because these lessons did not include publicly-presented canonical ideas.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Third International Mathematics and Science Study (TIMSS), Video Study, 1999.

Top

  • Content coherence: The lessons varied in how closely content ideas and activities were woven together to form a coherent, strongly linked lesson. A higher percentage of Australian and Japanese science lessons focused on engaging students in learning content and doing activities with strong conceptual links (58 and 70 percent, respectively) compared to lessons in the Netherlands and the United States (27 and 30 percent, respectively; figure 6). Twenty-seven percent of U.S. science lessons focused on doing Play Video Clip - Australia (Flash) activities, with no attention to content or only brief mentions of science content terms or ideas and no conceptual links among them.




Figure 6. Percentage distribution of eighth-grade science lessons by focus and strength of conceptual links, by country: 1999

Percentage distribution of eighth-grade science lessons by focus and strength of conceptual links, by country: 1999
‡Reporting standards not met. Too few cases to be reported.
1AUS=Australia; CZE=Czech Republic; JPN=Japan; NLD=Netherlands; and USA=United States.
2Doing activities with no conceptual links: USA>JPN, NLD.
3Learning content with weak or no conceptual links: CZE>JPN; NLD>AUS, JPN.
4Learning content with strong conceptual links: AUS, JPN>NLD, USA; CZE>NLD.
NOTE: Detail may not sum to 100 because of rounding and data not reported.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Third International Mathematics and Science Study (TIMSS), Video Study, 1999.

Top

  • Making connections versus acquiring information: Japanese students were supported in activities that focused on making connections among ideas, experiences, patterns, and explanations in more science lessons (72 percent) than in any other country except Australia. More Australian lessons focused primarily on making connections (58 percent) than did Czech and Dutch science lessons (figure 7). In contrast, students in eighth-grade science lessons within the Czech Republic, the Netherlands, and the United States were more likely to be focused on acquiring knowledge in the form of facts, definitions, or algorithms (72, 73, and 66 percent of Play Video Clip - United States (Flash) lessons, respectively) than on making connections. Further analysis revealed that making connections in Australian and Japanese science lessons was most often accomplished through Play Video Clip - Japan (Flash) an inquiry or inductive approach wherein data were collected and then used to develop new ideas in 43 and 57 percent of lessons, respectively (data not shown).

Figure 7. Percentage distribution of eighth-grade science lessons that developed science content primarily by making connections and by acquiring facts, definitions, and algorithms, by country: 1999

Percentage distribution of eighth-grade science lessons that developed science content primarily by making connections and by acquiring facts, definitions, and algorithms, by country: 1999
1AUS=Australia; CZE=Czech Republic; JPN=Japan; NLD=Netherlands; and USA=United States.
2Making connections: AUS, JPN>CZE, NLD; JPN>USA.
3Acquiring facts, definitions, and algorithms: CZE, NLD>AUS, JPN; USA>JPN.
NOTE: Detail may not sum to 100 because of rounding.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Third International Mathematics and Science Study (TIMSS), Video Study, 1999.

Top

  • Content supported by evidence in the form of first-hand data and phenomena: In Australia and Japan, more eighth-grade science lessons Play Video Clip - United States (Flash) supported every main idea in the lesson with multiple sets of first-hand data and with multiple phenomena than in the science lessons of the other three countries (figure 8).
  • Content supported by evidence in the form of visual representations: In Play Video Clip - Japan (Flash) the Czech Republic and Japan, multiple visual representations supported all the main ideas in science lessons more often than in the Netherlands (figure 8).
  • Content supported by multiple types of evidence: Further analysis showed that more Japanese science lessons (65 percent) supported every main idea with at least one set of first-hand data, at least one phenomenon, and at least one visual representation than lessons in the other countries. Australian lessons were more likely to use all three types of evidence to support all the main ideas (47 percent) than Dutch and U.S. lessons (14 and 18 percent, respectively) (data not shown).

Figure 8. Percentage of eighth-grade science lessons that supported all main ideas with more than one set of first-hand data, phenomena, and visual representations, by country: 1999

Percentage of eighth-grade science lessons that supported all main ideas with more than one set of first-hand data, phenomena, and visual representations, by country: 1999
1AUS=Australia; CZE=Czech Republic; JPN=Japan; NLD=Netherlands; and USA=United States.
2More than one set of first-hand data: AUS, JPN>CZE, NLD, USA.
3More than one phenomenon: AUS, JPN>CZE, NLD, USA.
4More than one visual representation: CZE, JPN>NLD.
SOURCE: U.S. Department of Education, National Center for Education Statistics, Third International Mathematics and Science Study (TIMSS), Video Study, 1999.

Top

  • Content developed with real-life issues: Science ideas were developed using real-life issues and examples in more Czech eighth-grade science lessons than in the lessons of the other countries except Australia (data not shown). In Australia, real-life issues were used to support the development of science ideas in more lessons than in Japan. Within the United States, a larger proportion of instruction time, on average, was spent presenting real-life issues as topic-related sidebars rather than using real-life issues to develop science ideas. The opposite was observed within Czech science lessons.

Top