[[alternative]]Cross-national analysis of 8-grade students’ achievement and background variables in science,and investigate the differences in science lessons activities of different characteristic classes in Taiwan: finding from TIMSS 2003

[[abstract]]This study uses the database of TIMSS 2003 to analyse the relationship between 8th graders’ science achievement and background variables and to investigate the differences in science lessons activities of different characteristic classes in Taiwan. The study included two parts. Part Ⅰ:there were 47 nations involved in TIMSS 2003, correlation analyses was used to find out the relationship between 8th graders’ science achievement and background variables, depend on the different unit of analysis,correlation analyses were performed at two levels: (1)within-country analysis(individual student being the unit of analysis), 47 correlation analyses were conducted at the student level for each of the participating countries for TIMSS 2003 to examine the relationship between students’ science achievement and background variables.(2)between-country analysis(country being the unit of analysis),correlation analysis at the country level were conducted for the aggregated data, the mean science achievement scores and the mean of the background variables were computed for each country, then the correlation coefficients for relevant variables were computed at the country level to find the relationship between countries’ performing and their characteristics. Part Ⅱ: Chi-square test was conducted to find the difference in science lessons activities of different characteristic classes in Taiwan. This part of the study investigates three kinds of different characteristic classes: (1)Classes with high and low performance; (2)Classes with high and low agreement of the statement“I would like to take more science in school”;(3)Classes with high performance but wouldn’t like to take more science in school and with low performance but would like to take more science in school. The findings are as follow: 1. High performance students had more books in their home, higher self-academic expectations, higher self-confidence in learning science and higher valuing science in comparision with low performance students. 2. High performance countries’ students had more books in their home, tended to think themselves don’t do well in science, learned things slowly in science, but thought science isn’t so difficult for them, and had lower valuing science in comparision with low performance countries. 3. The analysis results of science lessons activities of high and low performance classes were similar to the results of the analyses of classes with high performance but wouldn’t like to take more science in school and with low performance but would like to take more science in school. Classes with high performance and high performance but wouldn’t like to take more science in school had more teacher-centered/evaluation activities like teachers’ giving lectures, finding solutions of problems, homeworks, and quizs or tests, less student-center activities like experiments, group learning, explanation and presentation results of experiments and investigation. 4. Except for“how often do you begin your homework in class?”item, classes with high agreement of the statement“I would like to take more science in school”answered other 13 science lessons activities with“Every or almost every lesson”、“ about half the lessons”more and“ some lessons”、“ never”less in comparision with classes with low agreement of the statement.