TALIS 2013 Data
Logistic regression analysis – Enables estimation of the relationship between one or more independent variables (or predictors) on categorical dependent (or predicted) variables with two categories (binary logistic regression) or more categories (multinomial logistic regression).
Regression analysis was carried out for each country separately, as prior analysis showed noticeable differences in regression coefficients between countries.
Multinomial logistic regression – Compares multiple groups through a combination of binary logistic regressions. To calculate logistic regressions, three transformations of data take place: from probability to odds, from odds to log odds, and from log odds to odds ratios.
Multiple linear regression analyses – Computed using the IEA IDB Analyzer version 3.1.8.
Regression analyses – Computed for some indices with:
- The software STATA using population weights
- BRR methodology with Fay’s adjustment for variance estimation
General notes
- No adjustments were made to correct for the multiple analyses, increasing the likelihood that a relationship will be considered significant simply by chance.
- Because cross-sectional survey data were used for the analyses, no causal conclusions could be drawn.
All secondary analyses in TALIS 2013 should be computed using the IEA IDB Analyzer version 3.1.8 or higher to account for the complex sample design of TALIS. Population weights and BRR methodology with Fay’s adjustment for variance estimations must be used.
Data archives are created and made available in SPSS (standard use), SAS, and CSV format.
Two data files exist, one each for teachers and principals per ISCED Level 2 (as well as ISCED 1 and 3 for those countries who participated in these options). Furthermore, a teacher, principal, and mathematics teacher data file is available for those countries participating in the TALIS-PISA Link.
Data files are available on the OECD TALIS website for each population (ISCED 1, ISCED 2, ISCED 3, TALIS-PISA Link) and each participating country or sub-national entity. A combined international public-use teacher file and a combined international public-use principal file are also available. For the TALIS-PISA Link, a mathematics teacher data file is available in addition to the teacher and principal files for those countries that participated in the link option.
All public-use data files are available for download at the OECD TALIS website.
Special permission is needed to obtain data from Iceland.