This chapter presents analysis on the design of 71 upper secondary certificates from 38 education systems. The chapter identifies three models of upper secondary certificates, reflecting the main ways upper secondary certificates function across OECD countries, and discusses their use across general and vocational programmes. The chapter discusses how different models support the principles for upper secondary certification identified by this report: relevance, credibility, fairness and manageability. It also presents new analysis on the relationship between exams and student outcomes using Programme for International Student Assessment (PISA) 2022 data. It finishes by suggesting future topics for research on assessment and certification.
The Theory and Practice of Upper Secondary Certification
Abstract
Upper secondary certification has two main functions. Upper secondary certificates need to certify what students have achieved and to communicate a student’s readiness for their next steps such as higher education, training and work, by providing information that is credible and relevant for their future destinations (see Chapter 1 ‘Upper secondary certification has two key functions for students’). This report has identified four principles of upper secondary certification that are key to realising these functions: relevance, credibility, fairness and manageability. These principles focus on public trust in upper secondary certification and alignment with the curricula and values of education systems (see Chapter 2 ‘Towards principles for upper secondary certification’).
Certificates are commonly made up of multiple assessment components
Across the upper secondary certificates analysed for this report, most upper secondary certificates (57 out of 71) include at least two components (Table 6.1). Components refer to the individual activities and assessments students are required to undertake to achieve an overall certificate. As discussed in Chapter 3, a single approach for assessing students cannot be expected to assess the full range of knowledge, skills and understanding that upper secondary certificates seek to cover. For example, practical activities provide an opportunity to assess language, sport, artistic and technology/fabrication skills, while extended response items can assess students’ higher order cognitive skills like linking ideas and critically engaging with texts. Having multiple assessment components within a single certificate increases the likelihood students are assessed in a range of ways. Given the breadth of learning outcomes in most countries’ curricula and the range of skills young people need for life after school, having multiple assessment components within upper secondary certificate is a valuable policy approach.
Source: Transitions in Upper Secondary Education Certificate and Assessment mapping exercise, data collected across 2024 and 2025.
Chapters 3, 4 and 5 have each analysed how assessment tasks are used, the conditions and scheduling of assessment and the responsibilities for setting and marking within upper secondary certificates. This chapter identifies trends where combinations of these categories come together in overall upper secondary certificates.
Most ‘activities with unseen questions/tasks’ are set or marked externally
The ‘activities with unseen questions/tasks’ category encompasses traditional, paper-based, tests or ‘examinations’ and their digitised formats (as discussed in Chapter 2). Across the 71 certificates analysed, there are 84 individual assessment components that are ‘activities with unseen questions/tasks’. The majority of these assessment components are externally set and/or assessed. These assessment components are also frequently delivered under strictly controlled exam conditions and they often take place at the end of the programme of learning (see Figure 6.1). Throughout the rest of this chapter, assessment components that are externally set and/or assessed ‘activities with unseen questions/tasks’ are labelled as ‘external exams’.
Note: Across the 71 certificates analysed, there are 84 individual assessment components categorised as ‘activities with unseen questions/tasks’.
Source: Transitions in Upper Secondary Education Certificate and Assessment mapping exercise, data collected across 2024 and 2025.
Internal assessment uses a wide range of assessment tasks
Assessment components that are both internally set and assessed tend to not be associated with any one format. As shown in Figure 6.2, internally set and assessed components are delivered as projects and portfolios, practical activities and performances and activities with unseen questions/tasks at approximately similar levels. The format of internally set and assessed components tend to most frequently not be specified at the national level. In this case teachers and schools can determine the tasks themselves and so, in practice, this category could also include a wide range of assessment tasks. Throughout the rest of this chapter, assessment components that are internally set and assessed, of any format, are labelled as ‘internal assessment’.
Note: Across the 71 certificates analysed, there are 68 individual assessment components that are internally set and assessed.
Source: Transitions in Upper Secondary Education Certificate and Assessment mapping exercise, data collected across 2024 and 2025.
With the above identification of ‘external exams’ (assessment components which are activities with unseen questions/tasks and are externally set and/or assessed) and ‘internal assessment’ (assessment that is internally set and assessed, of any format), upper secondary certificates can be categorised into three models.
Model 1: Certificates with external exams and no internal assessment
Model 2: Certificates with internal assessment and no external exams
Model 3: Certificates with external exams and internal assessment
The classification of individual certificates according to these models can be seen in Table 6.2.
Notes: In Finland, the ‘intermediate’ category is used for both assessment setting and marking, because representatives of working life are involved in assessment alongside teachers. However, it is categorised here as a Model 2 certificate as there is no external assessment and teachers are the main assessors. Likewise, for France’s Certificat d'aptitude professionnelle (professional aptitude certificate, or CAP) (vocational upper secondary programmes), the ‘intermediate’ category is used. However, as teachers are the main assessors, it is categorised here as a Model 2 certificate.
Source: Transitions in Upper Secondary Education Certificate and Assessment mapping exercise, data collected across 2024 and 2025.
Model 1: Certificates with external exams and no internal assessment
Very few certificates include just externally-assessed components
Only ten certificates analysed (out of 71) use external exams but do not feature internal assessment. This model is almost exclusively used for certifying general education programmes (see Figure 6.3) and is frequently used when certification is strongly tied to higher education selection. For example, the Matura or Matriculation Examination Certificates in Croatia, Finland and Lithuania fall into this model and are specifically designed to signal students’ readiness for higher education, making success in external assessment a key factor for transitions to higher education. External exams serve a particularly important function in general upper secondary programmes, where transitions to higher education are often the explicit goal. The combination of standardised assessment (all students sit the same paper), external marking (to support consistent marking standards and approach) and highly controlled conditions (to prevent malpractice) aims to facilitate comparisons of student performance. The externality and standardisation of exams promote credibility within certification, enabling upper secondary certificates to play a central role in higher education selection.
Number of certificates used for general and vocational programmes by certificate model
Note: Upper secondary education programmes can be categorised as general or vocational (UNESCO Institute for Statistics, 2012[1]). Some certificates are used for both general and vocational programmes e.g. the Italy’s Diploma d'istruzione secondaria di secondo grado and Türkiye’s Lise Diploması are certificates achieved by students in general upper secondary programmes and vocational upper secondary programmes.
Source:Transitions in Upper Secondary Education Certificate and Assessment mapping exercise, data collected across 2024 and 2025.
Model 2: Certificates with internal assessment but no external exams
Entirely internally-assessed certificates tend to be used for vocational programmes
A significant minority of certificates rely solely on internal assessment (19 out of 71). While some certificates for general education programmes fit this model (e.g. in Japan, Korea and Poland), this model is predominantly used for vocational programmes (see Figure 6.3). Of the 34 certificates which are used for vocational education programmes, 14 of these certificates are entirely internally‑assessed (Figure 6.3).
Differences in the design of upper secondary certificates used for general and vocational programmes may be linked to their purposes and intended post-school pathways. Vocational upper secondary education aims to both prepare young people for higher education and develop labour market-relevant skills and knowledge to support transitions to work directly after school (UNESCO Institute for Statistics, 2012[1]). Reflecting these aims, only three quarters (77%) of young people in vocational education are enrolled in programmes that provide direct access to higher education, while this is case for almost all young people in general education (91%) (OECD, 2025[2]). The variations in intended post-school pathways helps to explain differences in certificate design: upper secondary programmes and certificates designed as pathways to work likely have less pressure to provide external credibility and more pressure to provide valid assessment of occupational skills and competencies, which for practical reasons are often best assessed internally.
Central policymaking still shapes the design of internal assessment
Even when assessment components are internally set and assessed, central policies still have an important role in determining how internal assessment is delivered and contributes to an overall certificate. Central policy can influence internal assessment in the following ways:
Determining the contribution of internal assessment for overall certification. Typically, the weighting of internal assessment when calculating an overall grade, is pre-determined. For example, in France, internal ‘continuous’ assessment is weighted 40% and external ‘final’ assessment is weighted 60% in the Baccalauréat.
Defining grades required to pass. There tends to be a standard policy within a single certificate for what constitutes a passing grade. However, there are some exceptions to this. In Poland, schools are allowed to set their own thresholds for what constitutes a passing grade (e.g. 50%, 10/20 or a Grade C).
Defining the assessment tasks. For example, in Austria, Estonia, Slovenia and Sweden students do an internally-assessed component where the format is pre-determined as a project or practical activity. This is known as the extended essay or diploma thesis in Austria1, practical work or subject research in Estonia, the seminar paper in Slovenia and the upper secondary school diploma project in Sweden.
Defining scheduling. Certain internal assessment components may have fixed submission dates. The student theses or final projects that exist in Austria, Estonia, Slovenia and Sweden, for example, tend to be assessed at the end of the programme of learning.
These examples show the different ways in which central authorities can still have influence over assessment, even when that assessment is internally set and assessed. Countries seeking to standardise the delivery of internal assessment and, perhaps, enhance the reliability of internal assessment, have a range of assessment policy levers to explore.
Internally-assessed certificates necessitate separate higher education entrance exams
In countries where upper secondary certification for general education does not include external exams, there tends to be separate higher education entrance tests. This is particularly the case for general education certificates that devolve high levels of responsibility to schools by not specifying the tasks, conditions or scheduling of assessments. In countries like, Poland, Spain and Türkiye, an upper secondary certificate alone does not give access to higher education institutions. In other countries, including Japan and Korea, upper secondary certification may technically provide eligibility for higher education, but external exams are commonly used for competitive selection to institutions and esteemed programmes. Separate higher education entrance exams are intended to facilitate fair comparison of students. When an entire upper secondary certificate is based on internal assessment, experiences suggest that achieving consistency across different classrooms, regions and schools is highly challenging if impossible to achieve (Chapters 3 and 5 extensively discuss this issue).
Using separate higher education entrance exams can have equity implications. Since the entrance tests are outside the main school system, assessed content may not reflect the upper secondary curriculum and may not align with classroom teaching, learning and assessment. Consequently, schools tend to have a limited role – or no formal role at all – in preparing their students for these tests. To respond to this challenge, students and their families may rely on private tutoring to help prepare for the specific content and format of the entrance test. In Greece, for example, families devote a significant proportion of their household income to private afternoon schools in order to prepare for the higher education entrance exam, the Panhellenic examination (OECD, 2018[3]). Similar reliance on shadow education for higher education entrance can be seen in Korea (OECD, 2022[4]) and Japan (OECD, 2018[5]) may mean that students from families who are less able to afford private tutoring or afternoon schools may be less able to compete equally against their peers. While private tuition exists in systems both and without separate higher education tests, its implications for equity may be greater in systems with separate entrance tests. Higher education entrance tests do not take into account student performance throughout the year, so capable students with less access to shadow education do not have the contributions from teacher-assessed grades to offset any relative disadvantaged they have when compared with peers who are able to access more shadow education for exam preparation.
The existence of higher education entrance exams can also hamper efforts to reform upper secondary certification when there is a gap between the reality of higher education entrance exams and the aspirations of the curriculum. This was seen in Japan in the late 1990s and early 2000s, when entrance exams prioritising routine cognitive skills compromised the implementation of curricula that aimed to take a more holistic and skills-based approach (OECD, 2018[5]). Likewise, in Korea, the format of higher education entrance exams creates resistance to schools implementing a more diverse range of assessment formats for upper secondary certification. In Korea, short-answer or multi-choice test items are widely used for teacher-based assessment in school as these formats are seen to be most aligned with the assessment approach of the College Scholastic Ability Test (CSAT or Suneung). Schools are being instructed to use more open-ended question items and essay-based assessment alongside implementing the new National Framework for the Elementary and Secondary Curriculum (2022) (Korean Ministry of Education, 2022[6]). However, this reform risks being seen as taking time away from teaching, learning and assessment approaches that are more aligned to the high-stakes Suneung (Kwon, Lee and Shin, 2017[7]). These examples show the importance of aligning higher education entrance exams with upper secondary education reforms.
Model 3: Certificates with external exams and internal assessment
Most certificates include both external exams and internal assessment
The majority of certificates use a combination of external exams and internal assessment (42 out of 71 analysed for this report). This model is most common among general education certificates and represents approximately half of vocational certificates (see Figure 6.3). This mixed approach addresses competing demands: external components provide reliability through standardisation, comparability and trust in marking procedures, while internal assessment enables more valid assessments of a diverse range of skills and contextualisation to local contexts.
Countries use different types of external assessments beyond written exams
Model 3 certificates include diverse assessment tasks that are marked and set externally. Overall, among the 71 certificates analysed for this report, 25 use external assessments that are delivered (or may be delivered) as projects, portfolios, practical activities or performances. In some cases, this is owing to ‘exams’ with a practical activity or performance element, depending on the subject (e.g. external exams in Denmark, France, Netherlands, Norway and Slovenia). Table 6.3 groups external assessment beyond written exams into three categories: those that may be delivered as practical or oral exams or include practical or oral components, those that are delivered either as projects and portfolios and those that may be delivered in a range of ways, depending on the context. External assessment beyond exams can be seen as another approach for addressing competing demands as the externality of assessment supports credibility and fair marking, while the diverse tasks support relevance and fairness through increased validity and assessment of a wider range of skills, which may suit a wider range of student strengths.
Note: All components in this table come from Model 3 certificates, except for the assessment components from Ireland’s Leaving Certificate, England’s T-Levels, Poland’s Egzamin zawodowy and Lithuania’s Vocational upper secondary certificate (4th-5th EQF). These certificates do not include internal assessment, and so are categorised as Model 1.
In France, the Contrôle terminal includes the grand oral, the oral part of the French exam. For scientific subjects it includes the assessment of experimental skills. For students taking the Technological Baccalauréat, it includes practical examinations according to the chosen specialisation.
Source: Transitions in Upper Secondary Education Certificate and Assessment mapping exercise, data collected across 2024 and 2025.
Key characteristics of non-exam external assessment include giving students a set amount of class time to engage with a centrally-set task e.g. to craft writing based on a set prompt, to design a product for a specified audience or undertake an investigation or practical activity (see Annex 3.B – examples for Ireland and New Zealand). Students may not be able to take work home and may have to attest to the authenticity of their own work (e.g. to confirm if AI is used, that it is referenced appropriately). Students are typically asked to submit their final output at the end of the specified timeframe. In the build up to these assessment events, students might have some class time preparing for the assessment event e.g. doing a mock exercise, learning what to highlight in their work to explicitly showcase their alignment with the assessment criteria.
Countries might use non-exam external assessment in upper secondary certification to support assessment of different and a wider range of skills than what would be typically assessed in ‘exams’ (Chapter 3 provides an extended discussion of the skills, knowledge and understanding assessed by different assessment tasks). For example, in New Zealand, non-exam external assessment has been a feature of the system since the introduction of upper secondary certification, the National Certificate of Educational Achievement (NCEA) in the early 2000s. In particular, the submission of portfolios or formal reports for external marking is used for subjects like Visual Arts and Technology, where a timed, closed conditions exam may not be the most appropriate way to assess student skills (see Chapter 5).
Other types of assessment formats might also be felt to help to reduce the stress and pressure that some students might feel from exams. In Ireland, for the Leaving Certificate, the shift to include non-exam assessment in every subject – representing 40% of the final grade and delivered at scheduled times prior to exams – responds to concerns about the stress reportedly experienced by students associated with this time (National Council for Curriculum and Assessment, 2021[8]).
The design of certificates and the assessments that contribute to them is a much-debated topic in countries by educators, students, their parents and society. External exams form a large and important part of high-stakes assessment in many countries. Most upper secondary certificates (50 out of 71) analysed for this report include external exams. Exams can be an effective way to assess a wide range of knowledge and skills, including some higher order competencies like applying knowledge and skills to interpret, analyse and solve problems (Kellaghan and Greaney, 2020[9]). Exams can also be seen as important for motivating students by providing clear goals for students to work towards with important stakes for their futures (Bishop, 2006[10]). Yet, there is extensive discussion about the perceived negative effects of exams on students’ well-being and stress (Woods et al., 2019[11]). There is also extensive debate in educational contexts about the contribution of exams for learning, and the skills and knowledge they assess. Frequently, this is articulated as concerns about the constrained nature of exams as time-pressured single events to assess the diversity of young people’s strengths and capabilities and to assess the skills that they will need for life after school (Campbell, 2023[12]; UN, 2023[13]; Independent Assessment Commission, 2022[14]).
This section contributes to these discussions by presenting original analysis about assessment design for young people’s outcomes. This report, and the mapping that underpins it, provides a unique perspective on the composition of upper secondary certificates internationally. In many cases, it provides information about the range of certificates that exist in systems – across general and vocational education – and which certificates include which assessment types. This analysis goes beyond assessment tasks to consider also the broad continuum of internal and assessment responsibilities and the possible variations for scheduling and defining assessment conditions. Notably, this report identifies that several systems require students in upper secondary general education to sit external exams, but this may not be a requirement for students in upper secondary vocational programmes. It also finds that not all students in general upper secondary education may take upper secondary exams, for example, because the subjects or option they choose are not assessed by exams. This more granular understanding about which students are assessed by exams and other assessment tasks for high-stakes purposes provides the foundations to better understand the relationships between assessment design and student outcomes.
Having external exams does not mean all students engage with those exams
In many cases where systems report having external exams, not all students in that system are necessarily required to engage with that exam. There are several reasons why this occurs:
Some vocational certificates do not include external exams.
In Austria for example, only around half (48%) of vocational upper secondary students take external exams (Annex Table 6.B.1). This is because some students in vocational education are completing apprenticeships which do not include external exams. This is also the case in several other systems with multiple vocational programmes, such as Croatia, Czechia and Estonia, where certification in one programme does not require students to take an external examination.
Flexible programme and certificate design.
In some systems, students can choose whether they sit exams, either by virtue of the combination of subjects and courses they take or by opting into or out of exam-based assessments. For example, in New Zealand, under the National Certificate of Educational Achievement, it is technically possible to achieve upper secondary certification based entirely on internally-assessed units.
External exams sit outside upper secondary education.
In several systems where upper secondary certificates are based entirely – or almost entirely – on internal assessments at the school level, there are separate external exams or tests for entry to higher education. These exams, when based on the upper secondary curriculum, appear similar to a high-stakes exit exam, although they are not requirements for upper secondary certification. This is the case in Greece, Japan, Korea, Spain and Türkiye. In these systems, the share of young people engaging in entrance exams or tests is less than the full cohort since not everyone will want to attend higher education and/or take higher education entrance tests. For example, in Korea, around 70% of students completing upper secondary take this test (Korean Education Statistics Service, n.d.[15]).
Exploring the relationship between exams and students’ performance
Using data about the shares of students who are exposed to curriculum-based external exams in upper secondary systems, this report explores the relationship between the existence, and prevalence, of external exams and student outcomes in PISA. Annex 6.B. provides an overview of the shares of students taking such examinations across OECD countries. While similar analysis has been undertaken in the past, an important contribution of this new analysis is that it is based on a better understanding of the share of students who take external exams. Previous analysis has been based on whether a country has external exams or not. Understanding the share of students who take external exams provides an important distinction because, as discussed in the preceding section, external exams may exist at the upper secondary level, but not all students are exposed to them.
In school systems where everyone takes external exams, students perform better in PISA for maths, reading and science, but the relationship is not significant
Previous analysis has found positive relationships between external exams and student learning outcomes (Wößmann et al., 2007[16]; Bishop, 1998[17]). PISA 2009 found that students in school systems that use external examinations perform, on average across OECD countries, 16 points higher in reading, than students in school systems that do not use these examinations (OECD, 2010[18]). Analysis undertaken for this report identifies that countries where a higher share of students take external exams tend to have higher mean scores in mathematics, reading and science in PISA 2022 (see Annex 6.C). Table 6.4 shows that students in school systems where everyone takes external exams perform 43.6 points higher mathematics, 30.8 points higher in reading and 37.4 points higher in science than students in countries that do not have exams, before accounting for students’ and schools’ socio-economic background. These results mean that a 10 percentage points increase in the share of students taking exams is associated with an average increase of 4.36 points in mathematics.
However, the relationship between the share of students taking exams and performance is not statistically significant when accounting for students’ and schools’ socio-economic background, suggesting that the result is mainly driven by other factors. Similar analysis in PISA 2006 also found the positive association between exams and student performance in science was not statistically significant when accounting for demographic and socio-economic background factors (OECD, 2007[19]).
The share of students taking external exams and performance in mathematics, reading and science in PISA 2022
Note: Statistically significant results are shown in bold.
Socio-economic status is measured by the PISA index of economic, social and cultural status.
The results were calculated using the pooled sample of students in PISA 2022 from OECD countries, where each national sample was equally weighted.
Source: Author original analysis using OECD (2023[20]), PISA 2022 Database, and data collected from countries (see Annex 6.B).
Exams could help support student performance, but further analysis is needed
The findings from this report suggest no conclusive role for exams in improving students’ performance. Many systemic factors beyond the presence of exams such as young people’s prior learning, teacher qualifications, continuous professional development and resources, may influence student performance. Limitations of the analysis presented in this report include that external exams may not cover the same content that is assessed by PISA, in particular in relation to reading where most upper secondary exams in national language tend to focus on written expression and literacy analysis. Also, in many countries, students are not yet in upper secondary at 15 when PISA is taken and may not yet be exposed to a curriculum or academic culture that is oriented towards final exams. Consequently, the possible positive impacts of exams may not yet be visible at age 15. The OECD will explore further analysis to address some of these challenges, such as by analysing the relationship between exams and young adults’ skills in the OECD Survey of Adult Skills (PIAAC) in the future.
Exploring the relationship between exams and students’ self-beliefs
While academic outcomes are important and a key measure of success for education systems, there are other factors, such as students’ attitudes and predispositions to learning, that inform policymakers about how the education system is preparing students to be effective learners. For example, PISA collects information on students’ self-beliefs about how well they can perform specific tasks or subjects. These self-beliefs affect students’ attitudes towards learning, the development of their skills and their decisions about educational pathways and careers (OECD, 2024[21]). The analysis in this report explores how students’ mathematics anxiety and growth mindset relate to the existence of external exams in upper secondary education.
Analysis suggests that there is no significant relationship between the share of students taking exams and mathematics anxiety
In many systems, the perceived negative consequences of exams on student well-being and stress are a prominent concern for parents and policy makers and may even be a driver of policy change. In the literature, there is significant discussion about the negative effects of exams for student well-being (Banks and Smyth, 2015[22]; Kouzma and Kennedy, 2004[23]; Roome and Soan, 2019[24]). PISA 2022 does not provide an overall measure of student stress or anxiety, however it does capture anxiety linked to maths specifically. PISA 2022 asked students to report whether they agreed with a series of statements about experiencing anxiety about mathematics, such as “I worry that I will get poor marks in mathematics”, “I get very tense when I have to do mathematics homework” and “I feel anxious about failing in mathematics”. These responses were then combined to construct the index of mathematics anxiety (OECD, 2024[25]). Mathematics anxiety is a challenge for students' well-being but also for their learning, as it is negatively correlated with performance (OECD, 2023[26]).
In countries where a higher share of students take external exams, levels of mathematics anxiety tend to be lower (Figure 6.4). However, regression analysis does not find a positive association between student’s maths anxiety and the prevalence of external exams, after accounting for students’ and schools’ socio-economic background and students’ mathematics performance (Table 6.5). These findings are interesting for our understanding of student stress, anxiety and exams. Notably, the limited relationship between maths anxiety and presence of external exams contrasts with the discussion in the literature suggesting that exams are detrimental for student well-being. There may be several reasons for this. Firstly, in the absence of external exams, students are likely to still experience high-stakes assessment, but it may be conducted in schools, repeatedly throughout the year and using a wider range of formats. Secondly, student stress and anxiety may not be fully captured by the measure of maths anxiety as reflected in PISA 2022. Finally, stress related to high-stakes assessment is likely closely linked to how assessment results are used. Exam-related stress may be more significant depending on how important those results are for entry to higher education, the diversity of post-school pathways that are perceived to be ‘successful’ and the influence of the broader socio-economic context on the stakes – both real and perceived – of exams at the end of school.
Note: Underlying data on the share of students taking external exams in each system is set out in Annex Table 6.B.1
Source: PISA 2022 Database, and data collected from countries.
A higher share of students taking exams is associated with lower levels of growth mindset
The concept of a growth mindset is the belief that abilities and intelligence can be developed over time. Students who report having a growth mindset score higher in PISA, value school more, set more ambitious learning goals, report higher levels of self-efficacy and display higher levels of motivation and lower levels of fear of failure (OECD, 2021[27]). PISA 2022 asked students to report whether they agreed with a series of statements about their mindset, such as “Your intelligence is something about you that you cannot change very much” and “Some people are just not good at mathematics, no matter how hard they study”. These responses were combined to construct the index of growth mindset (OECD, 2024[25]). Figure 6.5 shows that countries where a higher share of students take external exams tend to have lower levels of growth mindset.
Note: Underlying data on the share of students taking external exams in each system is set out in Annex Table 6.B.1
Source: PISA 2022 Database, and data collected from countries.
On average across OECD countries, having all students taking external exams in the education system is associated with a 0.3-point decrease in the growth mindset index, after accounting for students’ and schools’ socio-economic background and students’ mathematics performance (Table 6.5). These results mean that a 10-percentage point increase in the share of students taking exams in a country is associated with a 0.03 standard deviation decrease in students’ growth mindset. Unlike for performance and mathematics anxiety, the association between all students taking external exams in the education system and growth mindset is significant, even after accounting for students’ and schools’ socio-economic background and students’ mathematics performance. There is a strong relationship between having a growth mindset and certain learning behaviours and dispositions, like asking questions in class, considering multiple valid perspectives and enjoying challenging schoolwork (OECD, 2024[21]). One possibility is that there may be less time and space for students to develop and practice these behaviours in exam-oriented systems, and students may naturally find challenging schoolwork less enjoyable when it is underpinned by high-stakes exams. More broadly however, the finding underscores the importance of diversifying assessment types in order to draw on the positive contributions of different types of assessments for student performance, well-being and growth mindset.
The share of students taking external exams and index of students’ mathematics anxiety and index of growth mindset
Note: Statistically significant results are shown in bold.
Socio-economic status is measured by the PISA index of economic, social and cultural status.
The results were calculated using the pooled sample of students in PISA 2022 from OECD countries, where each national sample was equally weighted.
Source: Author original analysis using OECD (2023[20]), PISA 2022 Database, and data collected from countries (see Annex 6.B).
Few upper secondary certificates fully and perfectly achieve the four principles identified in this report (relevance, credibility, fairness and manageability). Yet when certificates become significantly unbalanced, with their relevancy perceived to be very low, or they are not seen to be credible or fair, or they are simply unmanageable for stakeholders – those certificates are unlikely to endure.
Table 6.6 summarises the different kinds of certificates analysed in this chapter – entirely exam-based, entirely internally-assessed and mixed certificates. It gives an overview of the different certificates in relation to the principles of credibility, relevance, fairness and manageability.
Source: Transitions in Upper Secondary Education Certificate and Assessment mapping exercise, data collected across 2024 and 2025.
Balanced certificates combine internal and external assessment and different types of tasks
External assessment is essential for credibility and facilitating selection for higher education
External assessment plays a fundamental role in upper secondary certification, particularly for facilitating selection into higher education. Most certificates for general education include external exams (36 out of 45), which serve as trusted signals of achievement and facilitate selection of students from different schools and regions. When upper secondary certification lacks external components, systems typically implement separate higher education entrance exams to fulfil this function, providing the credibility higher education institutions seek (i.e. Korea, Poland, Spain, Türkiye and the United States). In Poland, students who achieve the internally-assessed school leaving certificate must also sit the Matura oral and written exams for higher education entrance. Similarly, in the United States, high school graduation diplomas are rarely sufficient for higher education entry and many students take standardised tests like the SAT and the ACT.
Relevant, credible and fair vocational certificates depend on assessments that recognise the unique value and skills of vocational programmes
In many systems, low esteem of vocational upper secondary education is a challenge and contributes to shortages of young adults with technical skills for the economy. Upper secondary certificates can influence perceptions of vocational education programmes through their recognition of vocational skills and how they provide access to post-school pathways. In designing upper secondary vocational certificates and considering how they influence perceptions and pathways, there is a central trade-off for systems: a) authentically assessing practical skills with the involvement of employers, and b) providing pathways to higher education – typically by designing vocational certificates that are more aligned to general certificates (i.e. including a greater focus on external exams).
Country experiences illustrate the challenges for upper secondary vocational certificates when balancing alignment with general certificates and relevant assessment of vocational skills. In Slovenia, there are separate upper secondary certificates in general and vocational education – the general matura and vocational matura – which both provide access to ISCED 5 and 6 higher education programmes. The two certificates have different designs – the general matura has far greater external setting and marking than the vocational matura – leading to perceived differences in standards across the two certificates. The low completion rate of students from vocational upper secondary at bachelor level, ISCED 6 (41%, compared with 68% for students from general upper secondary) reinforces concerns that the vocational matura is not adequately ensuring that students have the necessary skills and knowledge to succeed in the higher education programmes they enter (OECD, 2023[28]). Slovenia has recently published a new law to reform the upper secondary vocational matura to address some of these concerns (Slovenian Ministry of Education, 2025[29]). In contrast in Lithuania, students from vocational and general education must achieve the same upper secondary certificate to access higher education. Given the different focus of the vocational programme with less time focused on general skills, few vocational students gain access to higher education. In 2022, only 1.7% of VET graduates progressed to higher education compared to 57.8% of all general graduates (OECD, 2023[30]).
Ultimately, “parity” across vocational and general upper secondary education certificates is not about assessing the same skills and providing the same pathways. Rather, vocational upper secondary certificates must recognise the unique achievement and skills of students completing both programmes. In Poland, for example, recent changes to the Matura enable students who have achieved professional qualifications to replace additional extended subject with results from their professional qualification when accessing ISCED 5 (Republic of Poland, 2025[31]). This approach maintains credibility through standardised external assessment while recognising the specific achievement of vocational students and their occupational skills.
The involvement of a wider range of stakeholders and assessment formats is also an important aspect of credibility and relevance for vocational certificates. Assessing students’ occupational skills through practical exams and workplace assessments and the involvement of professionals and employers (e.g. through local juries or assessment boards) can help to ensure that vocational certificates represent the skills and knowledge needed for vocational further education and employment.
Internal assessments can promote perceived relevance
It is normally left up to schools and teachers to determine when and how students are assessed for the internally-assessed parts of certificates, often labelled as school-based assessment or coursework grades. In practice, this means that a wider range of assessment formats can be used, supporting relevance by validly assessing complex skills that are important to society. Complex skills like evaluating and creating, social and emotional skills like teamwork and occupationally oriented skills can be integrated into and fostered through programmes of learning that are assessed at the school level (and in partnership with workplace and employers) through projects or using naturally occurring evidence.
Yet the relevance of internal assessment needs to be weighed against the challenges of ensuring common standards and approaches across internal assessment which can undermine certificate credibility. The need to balance the relative merits of internal assessments and external exams is likely a central reason why upper secondary certificates most commonly combine these two types of assessment.
Manageability is often the forgotten principle
Throughout the development of this report and in frequent conversations with policymakers, manageability emerged as the ‘forgotten principle’ and yet as the factor that makes or breaks certification reform. When countries are reforming their upper secondary certificates, matters of credibility, relevance and fairness tend to dominate policy debates and stakeholder consultation because individuals naturally focus on the overriding objectives they are seeking to achieve, or change, through reform. For example, in Scotland, a June 2025 announcement signals a reduced focus on exams to better align with the curriculum and support the recognition of all students’ achievements (emphasising principles of relevance and fairness) (Scottish Government, 2025[32]). While, in New Zealand, an August 2025 announcement proposes replacing the existing system of certification with new upper secondary certificates that would require all students to sit external exams (emphasising credibility) (New Zealand Ministry of Education, 2025[33]). In both of these and other cases, manageability, is less visible in the language of reform.
However, manageability is important when undertaking assessment reform because exams, other types of external assessment and internal assessment all present unique challenges for student and teacher workload. Additionally, when implementing reforms, when a system needs to adapt from an old to a new way of doing things, manageability (or a perceived lack of it) emerges as the factor that may derail well-intentioned reforms. In some systems, teachers may not be able to take on additional responsibilities for the setting and marking of high-stakes assessment where this has not historically been scoped into their roles. In other systems, central assessment agencies may not have the capacity to deliver and quality assure a wider range of assessment formats beyond traditional, standardised assessments. As discussed in Chapter 2, unmanageable certificates are unlikely to remain unchanged for long. Systems currently reforming upper secondary certification should ensure that manageability is a prominent and visible part of the dialogue and stakeholder consultation, alongside principles of credibility, relevance and fairness.
Building on the new analysis developed by this report, there are several areas that stand out as being important for future work to further enhance international understanding of upper secondary certification.
Upper secondary certification as a passport to higher education
A potential avenue for future work is examining the relationship between external exams (either as a part of upper secondary certification or separate higher education entrance exams), the supply of higher education places and the success of students in higher education. Work on this topic would aim to better understand how certification and entrance to higher education can be most effectively organised to support learning in upper secondary education and equitable student success in higher education.
Marking, moderation and grading systems
Credibility of upper secondary certificates relies on trust in student grades and overall results. While externally-assessed components tend to be marked centrally and anonymously, internally-assessed components are frequently more variable and, consequently, seen to be less reliable. While some countries have developed sophisticated procedures to promote consistent standard-setting and high-quality task development, the merits and contributions of the approaches used across different systems have not been discussed in comparative assessment research so far.
The Transitions in Upper Secondary Education project plans to look in greater detail at marking, moderation and quality assurance, with a focus on the role of social moderation structures and local-national feedback loops for supporting confidence in teacher-given grades. Planned work would also look at the potential for AI-based marking systems to complement human-led marking and what this means for workload and trust in results. Understanding the grading systems used across certificates will be important for this work as the grading system employed (i.e. qualitative judgements of competence, number/letter grades, percentages) and how grades are ultimately used by students, employers and higher education providers (i.e. for granular differentiation of students, for signalling overall competence) may necessitate particular approaches to moderation and quality assurance of grades. This work would seek to build awareness of available moderation practices, analyse what works for different systems and help countries identify approaches that would improve their practices. Future work on marking, moderation and grading would also provide any opportunity to look further at the relative weighting of assessment components and how different assessment approaches may be used for different subjects.
Exploring innovations in high-stakes assessment: cross-curricular assessment and self-assessment
The matrix of upper secondary assessment developed for this report and used for mapping upper secondary certificates has advanced understanding of the assessment components that are currently used within upper secondary certificates. However, many countries are actively exploring innovative assessments to reflect changing curricula and emerging skills needs. Building on OECD work to advance international assessments of social and emotional skills (OECD, 2024[34]) and previous work on assessing, documenting and recognising social and emotional skills within high-stakes assessment contexts (OECD, 2023[35]), the OECD could explore ways of assessing the skills, knowledge and understanding that sit outside or across subject boundaries, as well as further enquiry into the assessment of social and emotional skills, for example, through self-assessment and peer assessment.
Further exploring associations between exams and outcomes: subject-specific analysis and the skills of young adults
Future work will build on the initial analysis about the shares of students exposed to exams and student outcomes undertaken for this report. Future analysis will focus on exploring relationships between young people (i.e. 16-24-year-olds) who take exams at the end of upper secondary education and their outcomes as captured in the OECD Survey of Adult Skills (PIAAC). Outcomes of interest include literacy and numeracy skills, and potentially some social and emotional skills. Future analysis will also explore links between the content assessed in exams and student outcomes. For example, do students who are required to take upper secondary exams in maths have higher numeracy skills as young adults? If these analyses yield valuable results, the feasibility of similar analysis focused on other types of assessment tasks like projects and performance assessment could also be explored.
References
[22] Banks, J. and E. Smyth (2015), “‘Your whole life depends on it’: academic stress and high-stakes testing in Ireland”, Journal of Youth Studies, Vol. 18/5, https://doi.org/10.1080/13676261.2014.992317.
[10] Bishop, J. (2006), Chapter 15 Drinking from the Fountain of Knowledge: Student Incentive to Study and Learn - Externalities, Information Problems and Peer Pressure, https://doi.org/10.1016/S1574-0692(06)02015-0.
[17] Bishop, J. (1998), “The Effect of Curriculum-Based External Exit Exam Systems on Student Achievement”, The Journal of Economic Education, Vol. 29/2, pp. 171-182, https://doi.org/10.1080/00220489809597951.
[12] Campbell, C. (2023), All Learners in Scotland Matter: The National Discussion on Education - Final Report, Scottish Government, https://www.gov.scot/binaries/content/documents/govscot/publications/independent-report/2023/05/learners-scotland-matter-national-discussion-education-final-report/documents/learners-scotland-matter-national-discussion-education-final-report/learners-scot (accessed on 19 January 2024).
[14] Independent Assessment Commission (2022), Qualifications for a New Era: Equitable, Reliable Assessment - Final Report, A New Era, https://neu-era.s3.eu-west-2.amazonaws.com/downloadable/BJsJE4HobiNYRd7QUSNbyiraxtbFSCWu6eahtU7Q.pdf.
[9] Kellaghan, T. and V. Greaney (2020), Public Examinations Examined, World Bank, https://doi.org/10.1596/978-1-4648-1418-1.
[15] Korean Education Statistics Service (n.d.), Korean Education Statistics Service, https://kess.kedi.re.kr/eng/index;jsessionid=4KaosifPLciugHj0gMM3wJm3KdhaU79Z2uJzJtfa4t0SeuQQvr1JvqofvyF17KjV (accessed on 10 November 2025).
[6] Korean Ministry of Education (2022), The National Framework for the Elementary and Secondary Curriculum.
[23] Kouzma, N. and G. Kennedy (2004), “Self-reported sources of stress in senior high school students”, Psychological Reports, Vol. 94/1, https://doi.org/10.2466/pr0.94.1.314-316.
[7] Kwon, S., M. Lee and D. Shin (2017), “Educational assessment in the Republic of Korea: lights and shadows of high-stake exam-based education system”, Assessment in Education: Principles, Policy and Practice, Vol. 24/1, https://doi.org/10.1080/0969594X.2015.1074540.
[8] National Council for Curriculum and Assessment (2021), Senior Cycle Review Advisory Report, https://ncca.ie/en/resources/senior-cycle-review-advisory-report/ (accessed on 9 May 2025).
[33] New Zealand Ministry of Education (2025), Proposal to replace NCEA with new national qualifications, https://web-assets.education.govt.nz/s3fs-public/2025-08/NCEA%20Discussion%20Document%202025_web%206Aug.pdf?VersionId=s3FX.AjGhhVp5riTS4iUi_ZIvI_Ul7xP (accessed on 4 August 2025).
[2] OECD (2025), Education at a Glance 2025: OECD Indicators, OECD Publishing, Paris, https://doi.org/10.1787/1c0d9c79-en.
[21] OECD (2024), PISA 2022 Results (Volume V): Learning Strategies and Attitudes for Life, PISA, OECD Publishing, Paris, https://doi.org/10.1787/c2e44201-en.
[25] OECD (2024), PISA 2022 Technical Report, OECD Publishing, https://doi.org/10.1787/01820d6d-en.
[34] OECD (2024), Social and Emotional Skills for Better Lives: Findings from the OECD Survey on Social and Emotional Skills 2023, OECD Publishing, https://doi.org/10.1787/35ca7b7c-en.
[35] OECD (2023), “Assessing, documenting, and recognising social and emotional skills in upper secondary education: An overview of practices, approaches, models, and strategies from OECD countries”, OECD Education Policy Perspectives, No. 84, OECD Publishing, Paris, https://doi.org/10.1787/69c7abe6-en.
[28] OECD (2023), Education at a Glance 2023: OECD Indicators, OECD Publishing, Paris, https://doi.org/10.1787/e13bef63-en.
[20] OECD (2023), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html.
[26] OECD (2023), PISA 2022 Results (Volume I): The State of Learning and Equity in Education, PISA, OECD Publishing, Paris, https://doi.org/10.1787/53f23881-en.
[30] OECD (2023), Strengthening Upper Secondary Education in Lithuania, OECD Publishing, Paris,, OECD Publishing, https://doi.org/10.1787/a69409d7-en.
[4] OECD (2022), OECD Economic Surveys: Korea 2022, OECD Publishing, Paris, https://doi.org/10.1787/20bf3d6e-en.
[27] OECD (2021), Sky’s the Limit: Growth mindset, students, and schools in PISA, https://www.oecd.org/content/dam/oecd/en/about/programmes/edu/pisa/publications/national-reports/pisa-2018/brochures/Sky-s-the-limit-pisa-growth-mindset.pdf (accessed on 23 July 2025).
[3] OECD (2018), Education for a Bright Future in Greece, Reviews of National Policies for Education, OECD Publishing, Paris, https://doi.org/10.1787/9789264298750-en.
[5] OECD (2018), Education Policy in Japan: Building Bridges towards 2030, Reviews of National Policies for Education, OECD Publishing, Paris, https://doi.org/10.1787/9789264302402-en.
[18] OECD (2010), PISA 2009 Results: What Makes a School Successful?: Resources, Policies and Practices (Volume IV), OECD Publishing, https://doi.org/10.1787/9789264091559-en.
[19] OECD (2007), PISA 2006: Science Competencies for Tomorrow’s World: Volume 1: Analysis, OECD Publishing, https://doi.org/10.1787/9789264040014-en.
[31] Republic of Poland (2025), Item 881, https://isap.sejm.gov.pl/isap.nsf/download.xsp/WDU20250000881/O/D20250881.pdf.
[24] Roome, T. and C. Soan (2019), “GCSE exam stress: student perceptions of the effects on wellbeing and performance”, Pastoral Care in Education, Vol. 37/4, https://doi.org/10.1080/02643944.2019.1665091.
[32] Scottish Government (2025), Curriculum, Qualifications and Assessment Reform: progress to date and next steps, https://www.gov.scot/publications/curriculum-qualifications-assessment-reform-progress-date-next-steps/pages/2/ (accessed on 23 June 2025).
[29] Slovenian Ministry of Education (2025), Act on Amendments to the Matura Act, https://e-uprava.gov.si/si/drzava-in-druzba/e-demokracija/predlogi-predpisov/predlog-predpisa.html?id=17640&lang=si (accessed on 8 September 2025).
[13] UN (2023), Report on the 2022 Transforming Education Summit, https://www.un.org/sites/un2.un.org/files/report_on_the_2022_transforming_education_summit.pdf (accessed on 16 November 2025).
[1] UNESCO Institute for Statistics (2012), International standard classification of education: ISCED 2011, Comparative Social Research, http://uis.unesco.org/en/topic/international-standard-classification-education-isced (accessed on 4 December 2023).
[11] Woods, K. et al. (2019), “Linking rights, needs, and fairness in high-stakes assessments and examinations”, Oxford Review of Education, Vol. 45/1, https://doi.org/10.1080/03054985.2018.1494555.
[16] Wößmann, L. et al. (2007), School Accountability, Autonomy, Choice, and the Level of Student Achievement: International Evidence from PISA 2003, OECD Publishing, https://doi.org/10.1787/246402531617.
Note: ¹ This column indicates the extent to which standardised external examinations exist in the education system. Where there is a value between 0 and 1, standardised external examinations exist in some parts of the system concerned, but not throughout the system (e.g. regional variation or variation between different types of education programmes). The data collection for the existence of external standardised examinations values is based on the John Bishop’s definition of “external examination system”, which has the following characteristics: it produces signals of student accomplishments that have real consequences for the student, and it defines achievement relative to an external standard, not relative to other students in the classroom or the school. To enable fair comparisons of achievement across schools and across students at different schools, it is organised by discipline and keyed to the content of specific course sequences, which focuses the responsibility for preparing the student for particular exams on one or a small group of teachers; it signals multiple levels of achievement in the subject and not only a pass-fail signal, and it covers almost all secondary school students (Bishop 1998, 2001).
² New Zealand: In 2024, there were 45,045 students participating in NCEA Level 1 (defined as entered for 80 credits at Level 1 or above) and 42,190 students engaging in one or more exams at Level 1. From this, it can be interpreted that approximately 94% of students taking NCEA Level 1 take external exams. At NCEA Level 2, the figure would be 80% and at NCEA Level 3 the figure would be 77%. In a single year, approximately 83% of students taking any level of NCEA take external exams. However, as students typically complete NCEA Level 1 before progressing to higher levels, it can be understood that approximately 94% of students take external exams at least once during upper secondary education while working towards NCEA qualifications.
³ In England, Northern Ireland and Wales (United Kingdom), most students sit GCSEs. However, there are some students who will not take GCSEs. Data is not available on the share of students who do not sit GCSEs.
Source: Transitions in Upper Secondary Education Certificate and Assessment mapping exercise, data collected across 2024 and 2025.
Source: OECD (2023[20]), PISA 2022 Database, and data collected from countries (see Annex 6.B).
Source: OECD (2023[20]), PISA 2022 Database, and data collected from countries (see Annex 6.B).
Source: OECD (2023[20]), PISA 2022 Database, and data collected from countries (see Annex 6.B).
← 1. In Austria, since 2024, the extended essay is no longer a compulsory part of the standardized competency-oriented school leaving examination at AHS (general education schools). Approximately, 50% of students in general education schools still opt to do the extended essay. For students in vocational education schools, the diploma thesis remains mandatory.