Data set |
This dataset contains the responses of science teachers when asked questions referring to the teaching and learning provided to learners as well as their perceptions of the school in which they teach.
The data set has 204 variables and 316 cases.
TIMSS is a cross-national assessment of mathematics and science knowledge of fourth and eighth grades. It is conducted by the International Association for the Evaluation of Educational Achievement (IEA) every four years since 1995. South Africa took part in four cycles of TIMSS namely; in 1995, 1999, 2003 and 2011 with fieldwork being conducted in September of 2011. The Human Sciences Research Council (HSRC) conducted the study in 285 schools and assessed 11969 learners across South Africa. To prevent a cluster of scores at the lower end of the scale and to ensure sufficient variance in the data for robust analyses; TIMSS was administered to grade 9 learners in South Africa. As part of the TIMSS study mathematics and science tests and background questionnaires are administered to an intact grade 9 class within a sampled school. TIMSS uses the results from the tests and questionnaires which are administered to learners, teachers and principals to determine the achievement scores as well as the factors associated with academic success.
More specifically, the database includes the following for each country for which internationally comparable data are available:
298 schools sampled and 285 schools realised providing a realisation of 95.6%.
Department of Basic Education Master list of schools
Department of Basic Education Master list of schools database served as the sampling frame. 298 schools sampled and 285 schools realised providing a realisation of 95.6%.
The TIMSS sampling design is a two-stage stratified cluster design:
The sample was explicitly stratified by:
Each student's sampling weight (TOTWGT) is a composite of six factors: three weighting factors corresponding to the stages of the sampling design (school, class, and student ? WGTFAC1, WGTFAC2, and WGTFAC3), and three adjustment factors for non-participation at each of these stages WGTADJ1, WGTADJ2, and WGTADJ3, as described below. The variables described in this section are included in both the Student Background and Student Achievement files (see next chapter). The meaning and interpretation of the weights in each of these files is the same.
WGTFAC1 School Weighting Factor
This variable is the inverse of the probability of selection for the school where the student is enrolled.
WGTADJ1 School Non-participation Adjustment
This is an adjustment that is applied to WGTFAC1 to account for non-participating schools in the sample. Multiplying WGTFAC1 by WGTADJ1 gives the sampling weight for the school, adjusted for non-participation.
WGTFAC2 Class Weighting Factor
This is the inverse of the probability of selection of the classroom within the school.
WGTADJ2 Classroom Non-participation Adjustment
This is an adjustment that is applied to WGTFAC2 to account for non-participating classrooms or classrooms where student participation was less than 50 percent. Multiplying WGTFAC2 by WGTADJ2 gives the second-stage sampling weight, adjusted for non-participation.
WGTFAC3 Student Weighting Factor
This is the inverse of the probability of selection of an individual student within a sampled classroom. In the usual TIMSS case, where entire classrooms were sampled intact, the value was set to one for all students in the classroom. In a few countries, however, students were sampled within classrooms as a third sampling stage: in these cases the value of WGTFAC3 was greater than one.
WGTADJ3 Student Weighting Adjustment
This is an adjustment applied to the variable WGTFAC3 to account for non-participating students in the sampled classroom. Multiplying WGTFAC3 by WGTADJ3 gives the student-within-classroom sampling weight, adjusted for non-participation.
TOTWGT Total Student Weight
TOTWGT is obtained by multiplying the variables WGTFAC1, WGTADJ1, WGTFAC2, WGTADJ2, WGTFAC3, and WGTADJ3 for each student. The sum of these weights within a sample provides an estimate of the size of the population.
A key property of a sampling weight is that the same population estimates for means and proportions (although not the total or the number of units) will be obtained from any weighting variable that is proportional to the original weight (TOTWGT). For example, the sampling weights for a large country could be divided by a constant to make them smaller, and the weights of a smaller country could be multiplied by a constant to make them bigger. Regardless of which constant is used within a country, the weighted estimates of the means and proportions obtained from each of these proportional transformations of the weights will be exactly the same.
SENWGT Senate Weight
The SENWGT sampling weight is TOTWGT multiplied by 500 divided by the sum of the weights over all students in the target grade in each country. This results in a sample size of 500 in each country. SENWGT may be used in cross-country analyses in which each country should be treated equally. When SENWGT is used as the sampling weight for international estimates, the contribution of each country is the same, regardless of the size of the population.
HOUWGT House Weight
The HOUWGT sampling weight is TOTWGT multiplied by the ratio of the sample size (the number of students, n) in each country divided by the sum of the weights over all students in the target grade. HOUWGT may be used when the actual sample size is required for performing significance tests. Although some statistical computer software packages allow the sample size to be used as the divisor in the computation of standard errors, others will use the sum of the weights, which results in severely deflated standard errors for the statistics if TOTWGT is used as the weighting variable.
HOUWGT is the preferred sampling weight for analyses using such software. Because of the clustering effect in most TIMSS samples, it may also be desirable to apply a correction factor such as a design effect to the HOUWGT variable.
Weight Variables Included in the Student-Teacher Linkage Files
The individual student sampling weights generally should be used when you want to obtain estimates at the student level. The exception is when student and teacher data are to be analyzed together. In this case, a separate set of weights have been computed to account for the fact that a student could have more than one teacher. This set of weights is included in the Student-Teacher Linkage file and is listed below.
TCHWGT
This weight is computed by dividing the sampling weight for the student by the number of teachers that the student has. This weight should be used to obtain estimates regarding students and their teachers.
MATWGT
This weight is computed by dividing the sampling weight for the student by the number of mathematics teachers that the student has. This weight should be used to obtain estimates regarding students and their teachers.
SCIWGT
This weight is computed by dividing the sampling weight for the student by the number of science teachers that the student has. This weight should be used to obtain estimates regarding students and their teachers.
The Student-Teacher Linkage file also includes variables that indicate the number of teachers the student has.
Weight Variables Included in the School Data Files
As described earlier in this chapter, the schools in the TIMSS sample constituted the first stage of sampling and were chosen randomly. However, the school sample was designed to optimize the student sample rather than provide an optimal sample of schools, and is rather small in most countries - about 150 schools at each grade level.
SCHWGT School-level Weight
The school sampling weight SCHWGT is the inverse of the probability of selection of the school, multiplied by its corresponding non-participation adjustment factor. It is the product of WGTFAC1 and WGTADJ1.</p
By accessing the data, you give assurance that
Failure to comply with the above is considered infringement of the intellectual property rights of the HSRC.