Comparative indices of the education quality from the opinions of teachers and principals in TALIS 2018

Abstract From the 61 indicators presented in the two TALIS 2018 Reports, we constructed indices on the education quality, based on the teachers’ and principals’ opinions, by country/economy, for the 10 themes of these Reports. The indices result from the association between the OECD perspective on indicators’ influence on the education quality and whether each country’s score was statistically above, below, or not different from the OECD average. We found that (a) teachers from OECD-members have a positive opinion only in three themes and the principals have a better opinion than teachers; (b) teachers and principals from OECD-partners have a positive or neutral opinion in all themes; (c) professionals with a strong positive opinion on the learning environment and working conditions are mostly from Asia; (d) professionals with a clearly negative opinion are mostly from Europe; and (e) there are only two significant correlations between these indices and PISA 2018 results in the three domains simultaneously: (i) teachers’ opinions index about Student behaviour and classroom management (negative correlation); (ii) and principals’ opinions index about Learning environment (positive correlation).


Introduction
Several large-scale international studies have been conducted with the aim of analysing education systems. Most of them, such as the Programme for International Student Assessment (PISA), the Trends in International Mathematics and Science Study (TIMSS), and the Progress in International Reading Literacy Study (PIRLS) collect data from students, and the Teaching and Learning International Survey (TALIS), done under the responsibility of the Organisation for Economic Cooperation and Development (OECD), collect data from teachers and school leaders.
According to Sellar and Lingard (2014), the organisations that carry out these international studies, namely the OECD, have a strong influence on the governance of education in the different countries. This influence is growing because, in the neoliberal agenda, education is seen as an asset that ensures future job opportunities at the individual level and competitiveness in a global market (Seddon et al., 2013).
TALIS has been held three times, in 2018, 2013, 2018, and its main goal is "to generate internationally comparable information relevant to developing and implementing policies focused on school leaders, teachers and teaching, with an emphasis on those aspects that affect student learning" (OECD, 2019, p. 19). This goal is broken down into several themes: teachers' instructional practices, school leadership, teachers' professional practices, teacher education and initial preparation, teacher feedback and development, school climate, job satisfaction, teacher human resource issues and stakeholder relations, teacher self-efficacy (OECD, 2019). In 2018, two crosscutting themes were added to this list: innovation, equity, and diversity (Albergaria-Almeida et al., 2015;Fernández-Díaz et al., 2016;Veletic & Olsen, 2021). Concerning the governance in education, it is intended that each country will learn from the approaches of other countries with which they can compare themselves in economic and cultural terms and improve the quality of their education.
In TALIS, questionnaires are addressed to teachers and school principals seeking their perspectives on the state of education in their own countries, their views on education and the organisations, in which they work, and the successes and challenges they face (Sørensen & Robertson, 2020;Zakariya, 2020). The results are therefore based on their opinions, perceptions, beliefs, and descriptions of their activities. The OECD (2019OECD ( , 2020b cautions that this information is subjective and therefore may differ from objectively collected data, but considers it as important, because recognises "that teachers are central to the teaching process and school leadership is critical to enhancing education quality" (OECD, 2019, p. 83). In fact, in its various Reports (2010Reports ( , 2014Reports ( , 2019Reports ( , 2020aReports ( , 2020b, the OECD emphasises that teachers are at the heart of the school system and that the most diverse research has confirmed that their quality is essential in promoting student learning. This position is supported by several studies and research (e.g., Hanushek & Woessmann, 2020;Komatsu & Rappleye, 2017;Leigh, 2010;Salo et al., 2012;Scheerens, 2010).
Reliable comparisons of TALIS data across different countries/economies, and thus across cultures, are vital to contribute to evidence-based policy-making, as well as to promote equity and effectiveness of education policies. The TALIS study involves the participation of numerous countries/economies and He and Kubacka (2015) wrote, about TALIS 2013, that many scales in TALIS "have not reached the level of comparability that allows direct comparisons of scale scores" (p. 18). However, in TALIS 2018, the OECD develops and refines the information collected from the teachers and principals in order to establish valid, reliable, and comparable cross-sectional indicators (Ainley & Carstens, 2018). To make these issues more visible, the "framework also includes descriptions of some current limitations, for example, with respect to the valid, reliable and comparable measurement of teaching beliefs" (Ainley & Carstens, 2018, p. 10).
The sample size of TALIS has varied according to the number of participating countries/economies. In the case of TALIS 2018, focused on lower secondary education, data are collected from at least 200 schools per country and 20 teachers and one school leader are involved in each school. The response rate must be 75% of the selected schools, and a school is considered to have responded if 50% of the selected teachers in that school answer the questionnaire. The number of countries/economies has been increasing in the various editions of TALIS: 24 in 2008TALIS: 24 in , 38 in 2013TALIS: 24 in , and 48 in 2018 In the 2018 edition, more than240000teachers and13000principals have participated.

Purpose and objectives
In almost all countries/economies participating in TALIS 2018, reports were published with the corresponding results and studies analysing these results. In some cases, their results are placed in parallel with those of other countries. However, we did not find any study with a comparative analysis of the results of all countries/economies. This is our purpose. To this end, we created indices based on the teachers' opinions and school principals' opinions quantified in the indicators of the OECD reports (OECD, 2019, 2020b) on TALIS 2018. Based on them, we defined two objectives: (a) to compare the results of all countries/economies participating in TALIS 2018 and (b) compare these results with those obtained by countries/economies in PISA 2018, in the Reading, Mathematics, and Science domains.

Methodology
The OECD published two Reports on TALIS 2018 (OECD, 2019(OECD, , 2020b. The TALIS 2018 Results (Volume I): Teachers and school leaders as lifelong learners (V1) and the TALIS 2018 Results (Volume II): Teachers and school leaders as valued professionals (V2). In these Reports, the OECD wrote that the following five "pillars" support the profession of teacher and school leader: • Knowledge and skills base.
• Career opportunities and working regulation.
• Peer regulation and collaborative culture.
• Responsibility and autonomy.
• Prestige and standing.
The focus of V1 is the first pillar. It looks at the demography of professionals, the learning environment and socio-cultural composition of classes, how teachers and principals acquired their knowledge and skills in initial and continuous training, how teachers apply them in the classroom, and how principals develop their leadership. In a simplified way, V1 deals with the learning environment (OECD, 2019). V2 explores the other four pillars, analysing teachers' and principals' autonomy and responsibility, working and peer regulation, collaborative culture, status, stability, and career opportunities for professionals; it focuses on the working conditions and wellbeing of professionals (OECD, 2020b).
In these Reports, the OECD unfolds the analysis of these five pillars into 30 "objectives" (13 in V1 and 17 in V2) and 69 "policy pointers" (26 in V1 and 43 in V2), whose parameters are quantified in 61 indicators, 50 aimed at teachers, and 11 at principals. These indicators are grouped into 10 themes, 6 in V1 and 4 in V2. For each theme, there is a Table showing each country/economy's score on five to seven indicators, and where, based on a 95% confidence interval, each number is highlighted with one of the three colours indicating whether it is statistically above (green), below (grey), or not different from the OECD average (white). Table 1 shows the themes of each Volume, the corresponding pillars, and our designations for each theme, and Figure 1, which we bring from TALIS 2018 Results (Volume I): Teachers and school leaders as lifelong learners (OECD, 2019, p. 32) shows the scores of the first countries/economies of the Table corresponding to the theme KS2-Teaching in multicultural and multilingual settings where we can see the three colours highlighting the scores of different countries/economies in each indicator.
As we said above, in this paper, we make a comparison between the results of all countries/ economies that participate in TALIS 2018 through indices based on the teachers' opinions and on the principals' opinions and compare these results with those obtained by countries/economies in PISA 2018, in the Reading, Mathematics, and Science domains.
Supporting these indices are a classification of indicators and a classification of countries/ economies. Countries/economies are classified into three groups for each indicator according to the colours that the OECD assigns to their scores and that we mentioned above. That is: (a) in one group are those countries/economies whose score is statistically significantly higher than the OECD average; (b) in another group are those whose score is not statistically different from that average; and (c) in a third group are those whose score is significantly lower than that average.
The OECD, in their Reports on TALIS 2018 (OECD, 2019(OECD, , 2020b, and Schleicher, in their publications on TALIS (Fernández-Díaz et al., 2016, analyse the indicators under the five pillars and relate them to the learning environment inside and outside the classroom, the accountability, autonomy, and careers of teachers and school principals, and refer to the potential of each indicator as factors for improving the quality of education. They consider that each indicator can have a positive or a negative influence on the quality of education. If the action expressed by the indicator promotes the improvement of this quality, it will be a factor with a positive influence, if it causes a regression in this quality, it will be a factor with a negative influence. They also consider that there are indicators that have no influence, positive or negative, on the quality of education. Using this OECD perspective on indicators, we divide them into three groups: (1) the indicators that are in favour of improving the quality of education; (2) those that are unfavourable to this improvement; and (3) those that are neutral about such improvement. Tables 2 and 3 show the indicators addressed to teachers (which we have labelled T1 to T50) of V1 and V2, respectively. Table 4 shows the indicators addressed to the principals (which we have labelled P1 to P11) of both Volumes V1 and V2. In these Tables, each indicator is framed in a theme and marked with a symbol: ⇗ if it is favourable to the quality of education; ⇩ if it is unfavourable to that quality; or • if it is neutral.
For example, T1-% of teachers for whom the "use of ICT for teaching" has been included in their formal education or training is an indicator labelled with ⇗ (favourable to improve the quality of education) because the OECD (2019) wrote that the use of ICT "can be considered an expression of innovation" (p. 29) and without training in that area, it will not be possible to do so. On the other hand, following this OECD perspective, the indicator T4-% of teachers reporting a high level of need for professional development in ICT skills for teaching is labelled with ⇩ (unfavourable to the improvement of the quality of education) because it tells us that if teachers in a given country have a high level of that need it is because their conditions for innovating in this area are limited. Finally, T6-% of teachers teaching in classes with more than 10% of students whose first language is different from the language of instruction is labelled with • (neutral to the quality of education) because the socialcultural characteristics of a class are not indicative of more or less quality of education.
Therefore, regarding the quality of education, the position of each country/economy in each indicator results from the crossing of the two classifications that we defined, that is, it depends on the score (above, below or not different from the average of the OECD countries) that it has in that indicator and the type of symbol that characterizes the indicator: • it is positive for a country to have a score above the OECD average on an indicator labelled and to have a score below the OECD average on an indicator labelled ⇩; so we marked the country with + in these indicators;↗  • it is negative for a country to have a score below the OECD average on an indicator labelled ⇩ and to have a score above the OECD average on an indicator labelled ↗; so we marked the country with-in these indicators; • it is neutral for a country to have any level of score on an indicator labelled • and to have a score not different of the OECD average on an indicator labelled ↗ or ⇩; so we marked the country with = in these indicators.
Based on Figure 1, Figure 2 shows the classification of the indicators (in blue), all addressed to teachers, and the marks (in red) for four countries/economies after using the above-mentioned rules: (a) the first indicator is neutral (labelled •), so all countries are marked with = ; (b) the second indicator is positive (labelled ↗), so as the scores of Alberta (Canada) and Australia are green they are marked with +, and as the scores of Austria and Belgium are grey we marked them with-; (c) the third and fourth indicators are positive (labelled ↗), so the settings are similar to the second indicator, except for Australia (marked with = ) because its score is white.
(d) the fifth indicator is negative (labelled ⇩), so as the scores of Alberta (Canada) and Australia and Belgium are grey they are marked with +, and Austria is marked with = because its score is white; (e) the sixth indicator is positive (labelled ↗), so Alberta (Canada) is marked with = because its score is white and the other countries are marked with + because their scores are green.
Regarding teachers' indicators, for each country/economy, we calculated the difference between positive (+) and negative (-) indicators and divided it by the number of indicators in each theme, in  the total of each Volume, or in the total of the two Volumes (V1&V2), depending on the case. Finally, we multiplied the number obtained by 100 and rounded it up to the nearest whole number. We did the same for principals, but only for the total of V1, V2, and V1&V2, as there are a few indicators (8 in V1 and 3 in V2). We have thus obtained what we call the "teachers' opinions index" (TOI) and the "principals' opinions index" (POI) on the potential for the education quality in lower secondary education in each country/economy, for each theme, for the total of each Volume, or for the total of the two Volumes.
For example (see, Figure 2), as theme KS2 has six indicators, for: • Alberta (Canada), the difference between positive (+) and negative (-) indicators is 4 (= 4-0), so its TOI' value is 67 (= 4/6 × 100); • Australia, the difference between positive (+) and negative (-) indicators is 3 (= 3-0), so its TOI' value is 50 (= 3/6 × 100); • For Austria, the difference between positive (+) and negative (-) indicators is −2 (= 1-3), so its TOI' value is-33 (=-2/6 × 100); • Belgium, the difference between positive (+) and negative (-) indicators is −1 (= 2-3), so TOI' value is -17 (=-1/6 × 100). The way the indices are calculated tells us that: (a) a country/economy will be in a good position to improve the quality of education if the index is positive and the higher the index the better the position of that country; (b) a country/economy will be in a weak position to improve the quality of education if the index is negative and the lower the index the worse the position of that country; and (c) a country/economy with an index equal to zero is in a neutral position regarding the improvement of the quality of education.
It is important to say that these indices do not represent any administrative analysis of the education in each country, because the figures in the Tables of the TALIS Reports only translate the opinions and perceptions of teachers and principals about the learning environment in schools and the working conditions of these professionals. That is why we call them indices of teachers' or principals' opinions.
Using these indices, we ranked the countries, by theme, by Volume, and in total of the two TALIS Reports for teachers, and only by Volume and in total for principals, and analysed the results with a particular look at the quartiles of these rankings.
In all analyses in which we used inferential statistics we considered the 95% confidence interval as in the TALIS 2018 Reports (OECD, 2019, 2020b). Table 5 shows the TOI values in the total of Volume I (V1), in the total of Volume II (V2), and in the total of Volumes I and II (V1&V2). The first column (OECD) tells us whether each country/economy is an OECD-country (Member) or an OECD-partner (Partner) and the last column (T-B, which means "Top minus Bottom") gives us the difference between the number of times a given country's index is higher than the third quartile (then belonging to the group we call Top) and the number of times it is lower than the first quartile (then belonging to what we call the Bottom group) in the 10 themes of V1&V2. For example, Kazakhstan has a difference of 7 (= 7-0), because its index is higher than the third quartile in seven themes and is lower than the first quartile in no theme, while France has a difference of-5 (= 1-6) as its index is higher than the third quartile in one theme and is lower than the first quartile in six themes.

Teachers' opinions in TALIS 2018
In Table 5, we also can see, for each column, the average for OECD-members, the average for OECD-partners, the average for all countries/economies, and the first quartile, median, and third quartile values. This table is arranged in decreasing order by the values in column V1&V2, and then by the values in column T-B. The countries in the Top group are separated from the rest by a dashed line as the countries in the Bottom group of the list. Note that the French Community of Belgium is at the first row of the list just because it has no data in V1 and in V1&V2; it does not belong to the Top Group and is separated from it by a solid line.
Regarding the averages, it should be noted that the higher the average, the better the general opinion of the teachers about the theme in question in their schools and the lower it is, the worse is that opinion. We can see that: (1) The averages of V1 (on learning environment), V2 (on working conditions), and V1&V2 are very different in OECD-members and OECD-partners. The differences between them are statistically significant, which tells us that the teachers of the OECD-members have a worse opinion of their schools than the OECD-partners teachers, both in terms of learning environment and working conditions.
(2) The OECD-members average is similar in V1, V2, and V1&V2, which means that the teachers of these countries, in general, do not make big differences between learning environment and working conditions. The position of the teachers from OECD-partners countries is similar.
(3) The United Arab Emirates has the highest value in V1 (86) and Viet Nam in V2 (90), but Kazakhstan has the highest value in V1&V2 (66). France has the lowest value in V1 (-72), and Portugal has the lowest values in V2 and V1&V2 (-67 and-60). These figures show that their teachers have, respectively, the best and the worst opinion about the learning environment and the working conditions in their schools.
(4) The first six countries/economies in Top group are OECD-partners and the last eight countries/economies in Bottom group are OECD-members, which confirms that the teachers of the OECD-members have a very worse opinion of their schools than the OECD-partners teachers. Note that all these six Top group countries have all or most of their territory in Asia 1 and all of these eight Bottom group countries except Japan are European.
(5) There are 12 countries/economies (seven from Europe), none belonging to the Top or Bottom groups, where the V1 average is positive and the V2 average is negative or vice versa. The teachers of these countries have a favourable opinion of the learning environment in their schools but consider that their working conditions are not good or vice versa. The highest differences (77, 70, and 63) belong, respectively, to Czech Republic, Slovak Republic, and Norway (European and OECD-members), whose teachers have a very bad opinion on the learning environment (-34, -41, and -34) and a very good opinion on the working conditions (43, 29, and 29). With a good opinion on the learning environment and a bad opinion on the working conditions, the highest differences belong to Turkey (42 resulting from 28 and-14, respectively) and Chile (40 resulting from 7 and-33, respectively), both OECDmembers. This seems to indicate that in the opinion of teachers, in OECD-partners there is a better balance between the learning environment and working conditions than in OECDmembers.
(6) Looking at the column T-B, we can see that the figures confirm the V1&V2 index. All the countries/economies will a difference ≥ 4 belong to Top group and all the countries/economies with a difference ≤ -4 belong to Bottom group. The country (Kazakhstan) with the best difference (7) also has the best V1&V2 index value, and the country (Portugal) with the worse difference (-7) also has the worse V1&V2 index value.
Tables 6 and 7 present the values of the TOI for each theme and total of V1 and V2 and the T-B values for V1 and V2, respectively. The countries are decreasingly ordered according to the TOI of the total of each Volume (columns V1 and V2) followed by the order in T-B in both Tables. They also present the average for OECD-members, the average for OECD-partners, the average for all countries/economies, and the first quartile, median, and third quartile values. Note that the Belgium French Community does not appear in Table 6 for the reason said above. Both Tables also show the Top and Bottom groups separated from the other countries by a dashed line.  In Table 6, which shows the results about the themes related to learning environment and the pillar Knowledge and skills base, we can see that: (1) In all themes, the OECD-members average is less than the OECD-partners average, and the differences are statistically significant in KS1, KS2, and KS4.
(2) Most of the countries in the Top group are OECD-partners and only England is totally included in Europe. All the countries in the Bottom group are OECD-members and all of them except Japan are European.
(3) For most of the themes, the average opinion of teachers is positive. Only the theme KS6-Socio-demographic and experience profiles of teachers and school leaders have an average less than zero (note that this theme only includes three indicators, one being neutral). As this data is real, not depending on opinions, they show that maybe the teacher group should be, in general, renewed.
(4) The highest Total average (13.5) is for KS4-Student behaviour and classroom management, which seems to tell us that, in general, the teachers consider themselves prepared to lead with these kinds of problems.
(5) Almost the highest values in each theme belong to non-European countries and almost the lowest values in each theme belong to European countries. In Table 7, which shows the results about the themes related to working conditions and well-being and the four last pillars, we can see that: (1) In all themes, the OECD-members average is less than the OECD-partners average, and the difference is statistically significant in CP1.
(2) Only in theme CP3-Teachers' and school leaders' satisfaction with their work, the Total overall average is less than zero. This tells us that, in general, teachers are not satisfied with their work although they score positively the other themes related to working conditions and well-being.
(3) Most of the 10 countries belonging to the Top group are OECD-partners and five of them (the Netherlands, Estonia, the Czech Republic, Latvia, and Romania) are totally included in Europe. Ten (out of 12) countries/economies belonging to the Bottom group are OECDmembers, and nine of them are totally included in Europe.
(4) The highest value in CP1 and CP3 belong to non-European countries, in CP2 there are five countries (two totally in Europe) with the highest value and the Netherlands has the highest value in CP4. The lowest value in each theme belongs to European countries/economies. Looking at Tables 6 and 7 at the same time, we see that: (1) There is no country that belongs to the Top group in V1 and to the Bottom group in V2 or vice versa and there is a statistically significant positive linear parametric correlation (Pearson coefficient), with r = 0.491 and p < 0.001, between the values of V1 and V2 indices, which means that the evaluation that the teachers do on the classroom environment is similar to their evaluation on the working conditions.
(2) There are five countries/economies belonging to the Top group in both lists: Kazakhstan, Russian Federation, Singapore, Shanghai (China), and Viet Nam, all of them have all or most of their territory in Asia. There are five countries, all European, belonging to the Bottom group in both lists: Austria, Italy, France, Spain, and Portugal. This confirms that concerning both the learning environment and working conditions, the teachers with the most favourable opinion about what is happening in their schools are Asian and those with the most unfavourable opinion are European.
From a perspective of improving the quality of education, it seems that lower secondary teachers from non-European regions consider the learning environment and the working conditions and well-being in their schools to be significantly more positive than those in Europe.

Principals' opinions in TALIS 2018
We did the same analysis for principals' opinions. Table 8 shows the POI for the total of V1, V2, and V1&V2, with the countries/economies already decreasingly ordered by the values in the column V1&V2, followed by the order in the column V1 and the order in the column V2. As there are only 11 indicators and there are no indicators from the themes KS2, KS4, CP1, and CP4, we did not present the difference Top-Bottom. However, it should be noted that all the countries/economies have a difference between-1 and 1. As before, there is a dashed line separating the Top and the Bottom groups from the other countries and the French Community of Belgium is in the first row but does not belong to the Top Group.
We can see that:  (1) The OECD-members, OECD-partners, and Total averages are above zero in the three columns. None of the differences between these averages in each column is statistically significant, but all V1 averages are higher than V2 averages and these differences are all statistically significant. This means that, in general, the principles are more satisfied with the learning environment than with their working conditions.
(2) There are six (out of 10) OECD-members in the Top group and in the Bottom group there is the same number of OECD-members and OECD-partners.
(3) Although there is no country belonging to the Top group in V1 and to the Bottom group in V2 or vice versa, there is no statistically significant linear parametric correlation (Pearson coefficient), with r = 0.146 and p = 0.321, between the values of V1 and V2 indices, which means that the evaluation that the principals do on the classroom environment and on the working conditions are not similar.
(4) Half of the countries in the Top group are European, and only three countries in the Bottom group are totally European. This seems to show that European principals are far more positive than teachers about what is happening in their schools.
Comparing the Tables, 6, 7, and 8, we verify that: (1) There are three countries/economies, Kazakhstan, Singapore, and Shanghai (China), belonging to the three Top groups simultaneously. This means that their teachers and principals have a general common opinion about the learning environment and the working conditions in their schools-positive to the improvement of the education quality.
(2) Only Portugal belongs to three Bottom groups simultaneously, what means that their teachers and principals agree in very a negative opinion about the learning environment and the working conditions in their schools.
(3) Although there are several differences of opinion between teachers and principals in each country, considering the total number of all countries, there are no statistically significant differences between the average opinions of teachers and principals concerning the V1, V2, and V1&V2.
In summary, we can say that, in general, the opinions of teachers and principals from the countries that participated in TALIS 2018 about all themes are positive, but the teachers' opinion is worse than the principals' opinions. We also can say that, in general, the opinion of the teachers from OECD-partners countries is better than the opinion of the teachers from OECD-members countries, concerning the learning environment and the working conditions, while the opinion of the principals from OECD-partners countries is similar to the opinion of the principals from OECD-members countries.

Correlations between PISA 2018 results and the opinions of teachers and principals in TALIS 2018
We crossed the opinions of teachers and principals with the PISA 2018 results in the three domains -Reading, Mathematics, and Science-looking for linear parametric correlations (Pearson coefficient) between the results of the countries/economies in this study and their TOI and POI in each theme and Volume totals (see Figures 3 and 4). In these figures, we highlighted the statistically significant correlations in two shades of light blue and indicated the corresponding r and p.
Between the 13 TOI and 3 POI, we found only one TOI and one POI with statistically significant correlation with the scores in the three PISA domains: Reading, Mathematics, and Science. In KS4 -Student behaviour and classroom management, the TOI has the following significant correlations with PISA 2018 results by country/economy: (a) Reading, r = -0.363 (p = 0.023); These correlations are negative, which means that the countries whose teachers consider that they have no problems with student behaviour and classroom management are the ones with the worst results in PISA 2018, showing, perhaps, that teachers do not do a proper evaluation of their action in this field. It seems that they are overestimating their knowledge of how to act in situations of student misbehaviour and overvaluing their actions in classroom management. This result is even more alarming as this theme has the highest average of all themes in Volume I (see , Table 6 and comment (10)).
In the total of V1, the POI has the following significant correlations with PISA 2018 results by country/economy: These correlations are positive, which means that, for the topics related to the learning environment, the countries whose principals consider that the quality of education in their schools is good are the ones that obtain better results in PISA 2018. This seems to indicate that principals have a good grasp of what is important in the learning environment for improving the student success.
There are other statistically significant correlations between the POI and TOI indices and the PISA 2018 results, all negative, but not with the three domains simultaneously. TOI-KS6-Sociodemographic and experience profiles of teachers and school leaders have significant correlations with two domains:  The following indices have significant correlations only with Mathematics: (i) TOI-KS1-Information and communication technology (ICT) for teaching, r = -0.344 (p = 0.030); and (j) POI-V2, r = -0.369 (p = 0.019).
From all these statistically significant correlations with the PISA 2018 results in one, two, or the three domains, only the correlation between these three domains and POI-V1 is positive. There is no significant correlation with a specific CP theme, and all the significant correlations with KS themes (KS1, KS4, and KS6) are with TOI and negative, which seems to confirm that teachers are overestimating their knowledge of how to act in situations related with classroom environment.

Conclusions
This comparative analysis is about the results of TALIS 2018. It is therefore based on the perceptions and opinions of teachers and principals about the learning environment and working conditions in lower secondary schools in the 48 countries/economies that participated in that study. To this end, we constructed an index for teachers (TOI) and another for principals (POI), concerning each theme of the two Reports (V1 and V2) of OECD on TALIS 2018, and V1, V2, and V1&V2 totals, to indicate whether their opinions were more or less favourable to improving the quality of education as seen by the OECD (2019, 2020b) and Schleicher (2016Schleicher ( , 2020, and compared the scores of all countries/economies in that indices. We also compared the times that the TOI and the POI of each country/economy are higher than the third quartile (belongs to the Top group) and lower than the first quartile (belongs to the Bottom group) by themes and V1, V2, and V1&V2 totals.
The 50 indicators for teachers and 11 indicators for principals across the 10 themes defined in the TALIS Reports tell us that, on average, while OECD-partners have a positive opinion in all 13 themes and Volume totals, except one, OECD-members that have a negative opinion in eight, and the differences between their indices averages are statistically significant in seven. Looking at regions, we verify that most of the countries whose teachers and principals have a clear positive opinion on the learning environment and working conditions in their schools are non-European, particularly from Asia: Kazakhstan, Shanghai (China), Viet Nam, the United Arab Emirates, and Russian Federation. Most of those countries whose teachers and principals have a clearly negative opinion are European, particularly from Western Europe: Portugal, Spain, France, and Belgium. The difference between the number of times each country belongs to the Top and Bottom group in all themes confirms all these results.
These results suggest that education professionals in the so-called more developed countries consider that what is happening in their lower secondary schools is indicative that there is a problem with the quality of education that can make it difficult to improve or even cause a setback. This is alarming particularly for various European countries that their lower secondary teachers have the most negative opinion of what happens in their schools in relation to the learning environment, including their training, and to their working conditions and their wellbeing. Although the opinion of European principals is not as negative as that of teachers, it is not generally favourable as regards both the learning environment and working conditions. We also found that there is no significant relationship between these indices and the results of PISA 2018. There are only two statistically significant correlations between the results of these indices and the results of the three domains of PISA 2018: a negative one referring to teachers in a Student behaviour and classroom management theme and a positive one referring to principals and related to the totality of the classroom environment indicators. These correlations seem to show that principals have a good perception of what is important in the learning environment to promote student success and that teachers overvalue their skills in classroom management and student behaviour.
TALIS 2018 data collection was conducted before to the COVID-19, so it is important to examine, in a future TALIS, how teachers are being trained, in-service and pre-service, to teach in these uncertain times. What do teachers' responses to COVID-19 indicate about the development of teacher reflexivity in initial and continuous teacher education? What professional learning needs for teachers were highlighted during the Covid-19 crisis? These are some questions for future investigations.