In 2017, Ukraine conducted Monitoring Survey of Secondary School Teachers and Principals (by the OECD TALIS 2013 methodology). The executive summary report on the results of the study (Shchudlo S. et al., 2018) contains descriptive statistics of the responses of teachers and school principals to questionnaires. However, such important indicators of OECD report as teachers' feelings self-efficacy or job satisfaction (OECD, 2014) were not calculated. The lack of aggregated variables in the national report complicates the interpretation of the survey results for Ukraine and makes it impossible to investigate its relationships, as well as to compare the results of Ukraine to those of other countries that participated in the TALIS study. In this report, we present the self-efficacy beliefs of Ukrainian teachers and its relation to some other important variables, and compare the results with those of some other countries with similar educational systems.

As indicated in the OECD report, teachers’ self-efficacy is related to important demographic characteristics such as years of experience, gender, educational attainment and teaching level. In the majority of TALIS countries, most teachers report holding beliefs that suggest high levels of self-efficacy. However, in most TALIS countries, male teachers reported lower levels of self-efficacy. In other countries where male teachers show higher self-efficacy the strength of this relationship is weak (only in Japan it is moderate). The OECD report also show that more experienced teachers tend to have higher self-efficacy in most countries. There is a general upward trend by experience five-year intervals, though there appears to be a slight stagnation for teachers with 11-20 years of experience, followed by a spike at 21-25 years. In addition, TALIS data indicate that it is not the number of students that are in a teacher’s class that has the large association with teachers’ self-efficacy. It is important for Ukraine to check this, since there is a great deal of discussion in Ukrainian society about the influence of large classes on the efficiency of teaching.

The relationships of teachers’ self-efficacy with some variables are not highlighted in the OECD report, but their research is important for Ukraine. These variables include the regional compositions of Ukraine and the size of settlements in which schools are located. In particular, it is known that the results of the External Independent Evaluation of school graduates in Ukraine vary considerably by region, and are significantly lower among graduates of rural schools (Ukrainian Center for Educational Quality Assessment, 2018). Therefore, it was important for us to check if relevant data are associated with teachers’ self-efficacy. Finally, comparing the characteristics of Ukrainian teachers with those in other countries provides valuable information on understanding the nature of the variability of these characteristics and their interrelations.

The self-efficacy scale in TALIS 2013 was defined using factor analysis from three scales – efficacy in classroom management (then we will use the abbreviation C), efficacy in instruction (I) and efficacy in student engagement (E), each of which were measured by four items from the 34th block of questions. Whereas different procedures can be used for the estimation of scale characteristics and composite scores, for the analysis of Ukraine's data we decided to apply an approach based on the item response theory (IRT) framework as used in, for example, the scaling of context questionnaire data in PISA using the Partial Credit Model (OECD, 2009). To make sure the similarity of the results obtained by different methods, we have constructed one-dimensional self-efficacy scale for Poland data (3797 cases with valid index) using all 12 items with the help of Partial Credit Model in Winsteps software (Linacre J.M., 2018). The alpha reliability coefficient was 0.873, raw variance explained by measures was 45.3% and correlation between indexes was 0.98. Correlation between index components C, I and E obtained through IRT was in the range of 0.5, exceeding the corresponding values obtained through CFA (OECD, 2014). Given such good concordance of results, we have built a similar self-efficacy scale centered at the average value 10 for 3592 Ukrainian teachers (alpha reliability coefficient is 0.863, raw variance explained by measures is 47.6%, the value of the self-efficacy index for teachers varies from 3.71 to 16.26 logit with a standard deviation 1.67). Ukrainian teachers find it most difficult to “Motivate students who show low interest in school work” (9.52 logit). Instead, it is easier for them to “Provide an alternative explanation, for example, when students are confused” (6.13 logit). Significantly lower levels of teachers’ self-efficacy is observed in student engagement (E) than in instruction (I) or in classroom management (C). We studied in detail the properties of the scale and its dependence on some background characteristics. Most of the detected relationship between teachers’ characteristics and their self-efficacy have been confirmed using multiple linear regressions. To compare the level of self-efficacy of Ukrainian teachers, we have built a common scale using TALIS 2013 data for Poland, Croatia (3617 cases with valid index) and Estonia (3048). These countries once had similar educational traditions and percentage of teachers' answers related to self-efficacy were similar. Among them Poland had the highest index. DIF and DGF analysis was conducted.

The level of self-efficiency of Ukrainian teachers does not depend on the region or size of the settlement, where teachers work. The size of the class, the director's gender and the level of education received also does not significantly affect this indicator. Only the teacher's gender, his age, and his experience as a teacher were significant. Female teachers and teachers with experience of more than 5 years in Ukraine, as in most countries, displayed higher level of self-efficacy. Also, the level of self-efficacy increases with age. Compared to teachers in Poland, Croatia and Estonia, Ukrainian teachers have a significantly lower level of self-efficacy. Differential Item Functioning (DIF) analysis revealed that Estonian teachers better than anyone manages to “Motivate students who show low interest in school work”, “Get students to believe they can do well in school work” and “Help my students value learning”. Teachers of Croatia are best at “Implement alternative instructional strategies in my classroom”. Ukrainian teachers significantly better than anyone can “Use a variety of assessment strategies” and “Provide an alternative explanation, for example, when students are confused”, but significantly worse than anyone can “Make my expectations about student behaviour clear”. Differential Group Functioning (DGF) analysis revealed that all countries are similar in efficacy in classroom management. Ukraine is close to Croatia in all three aspects of self-efficacy. At the same time, the components of the self-efficacy index for Ukraine and Estonia differ most: the level of efficacy in student engagement is higher in Estonia, and the level of efficacy in instruction is higher in Ukraine. We also found items in the questionnaire that were differently perceived by men and women in different countries. The next step is to study the link between self-efficacy of Ukrainian teachers and other indexes and background characteristics.

Ayala R.J. (2009). The Theory and Practice of Item Response Theory. New York, London: The Guilford Press. Bandura A. (1986). Social Foundations of Thought and Action: A Social Cognitive Theory. Prentice Hall, Englewood Cliffs, NJ. Camilli G. (2006). Test fairness. In: Brennan R. (ed) Educational measurement. Westport, CT: ACE Praeger series on higher education. Desa D. (2014). Evaluating Measurement Invariance of TALIS 2013 Complex Scales: Comparison between Continuous and Categorical Multiple-Group Confirmatory Factor Analyses. OECD Education Working Papers, No. 103, OECD Publishing, Paris, https://doi.org/10.1787/5jz2kbbvlb7k-en Linacre J.M. (2018). A User's Guide to WINSTEPS. Rasch-Model Computer Programs. Program Manual 4.3.1. https://www.winsteps.com/a/Winsteps-Manual.pdf Masters G. N., & Wright B. D. (1997). The partial credit model. In: M.J. van de Linden & R.K. Hambleton (Eds.), Handbook of modern item response theory. Berlin: Springer. OECD (2014). TALIS 2013 Results: An International Perspective on Teaching and Learning. https://dx.doi.org/10.1787/9789264196261-en OECD (2014). TALIS 2013 Technical Report. http://www.oecd.org/education/school/TALIS-technical-report-2013.pdf OECD (2009). TALIS 2008 Technical Report. https://www.oecd.org/education/school/44978960.pdf OECD (2009). PISA 2006 Technical Report. OECD, Paris. https://www.oecd.org/pisa/data/42025182.pdf Shchudlo S., Zabolotna O., Lisova T. (2018). Ukrainian Teachers and the Learning Environment. Results of All-Ukrainian Monitoring Survey of Secondary School Teachers and Principals (by the TALIS methodology). Drohobych: Trek LTD (in ukr.) Ukrainian Center for Educational Quality Assessment (2018). Official Report on Conducting in 2018 an External Independent Evaluation of Learning Outcomes Obtained on the Base of Complete General Secondary Education. (in ukr.) http://testportal.gov.ua/wp-content/uploads/2018/08/ZVIT-ZNO_2018-Tom_1.pdf