Conclusions
Much of the literature suggests that grade retention can have a negative impact on students' academic performance. To test whether repeating a grade affects schooling in any way, we compared the retention rates of countries participating in PISA with their scores in the exam.
Retention at the primary level as a share of total enrollment does have an impact on PISA scores. Specifically, we found that math and science scores for the PISA exam were negatively affected by retention at the primary level but not by retention at the secondary level. Literacy was not impacted by grade repetition in any level of education.
Yet, two important problems become evident when considering the explanations for the connection between retention and test scores. The level of economic development and regional differences offer both complimentary and separate reasons for varying performance levels on PISA.
Economic development, controlled for in our tests with GDP per Capita, membership in the OECD, and the urban population as a share of the total population, can help explain why countries do better or worse on PISA. For example, many OECD countries with high levels of GDP per Capita perform better on PISA than countries with lower levels of GDP per Capita.
For instance, Japan does much better than Indonesia. Perhaps the economic, political, and infrastructure advantages that more developed countries enjoy are placing students in these countries ahead. Moreover, developed countries tend to be more urbanized, where infrastructure is likely to be better. These are modern advantages that a country like Indonesia may not have because of its lower level of development.
Conversely, when we look at Qatar we see a different picture. Qatar has a high level of GDP per Capita and is highly urbanized yet performs very poorly on PISA. Additionally, when we look at OECD countries like Chile and Mexico, they also perform poorly on PISA in comparison to other OECD nations in North America and Western Europe. Therefore, it is possible that rather than varying levels of economic development, it may be the case that regional differences can better explain performance on PISA.
Regional differences are important because they may account for varying institutional arrangements, different educational systems, etc. These factors can affect PISA scores in ways that economic development may not capture. In fact, we found distinctions among all PISA-taking countries according to region. In particular, we see that Latin American countries perform very poorly in comparison to Asian and European countries. Perhaps parental involvement in education, deficient institutional arrangements that impact education quality, and other factors specific to the various regions can better explain differences in PISA results.
Given these different possibilities, retention might reflect regional differences or varying levels of development. Yet, if we were to provide both a positive analysis and a normative claim concerning education and test scores we can point to the importance of retention at the primary level on PISA scores in math and science.
Contrary to arguments in favor of retention, we see that for Latin American countries high levels of retention and low PISA scores coincide. We show that there is a correlation between high levels of retention at the primary level and poor PISA results. We believe that retention does have an impact on test scores by negatively influencing student performance in school, as reflected by PISA results.
Retention at the primary level as a share of total enrollment does have an impact on PISA scores. Specifically, we found that math and science scores for the PISA exam were negatively affected by retention at the primary level but not by retention at the secondary level. Literacy was not impacted by grade repetition in any level of education.
Yet, two important problems become evident when considering the explanations for the connection between retention and test scores. The level of economic development and regional differences offer both complimentary and separate reasons for varying performance levels on PISA.
Economic development, controlled for in our tests with GDP per Capita, membership in the OECD, and the urban population as a share of the total population, can help explain why countries do better or worse on PISA. For example, many OECD countries with high levels of GDP per Capita perform better on PISA than countries with lower levels of GDP per Capita.
For instance, Japan does much better than Indonesia. Perhaps the economic, political, and infrastructure advantages that more developed countries enjoy are placing students in these countries ahead. Moreover, developed countries tend to be more urbanized, where infrastructure is likely to be better. These are modern advantages that a country like Indonesia may not have because of its lower level of development.
Conversely, when we look at Qatar we see a different picture. Qatar has a high level of GDP per Capita and is highly urbanized yet performs very poorly on PISA. Additionally, when we look at OECD countries like Chile and Mexico, they also perform poorly on PISA in comparison to other OECD nations in North America and Western Europe. Therefore, it is possible that rather than varying levels of economic development, it may be the case that regional differences can better explain performance on PISA.
Regional differences are important because they may account for varying institutional arrangements, different educational systems, etc. These factors can affect PISA scores in ways that economic development may not capture. In fact, we found distinctions among all PISA-taking countries according to region. In particular, we see that Latin American countries perform very poorly in comparison to Asian and European countries. Perhaps parental involvement in education, deficient institutional arrangements that impact education quality, and other factors specific to the various regions can better explain differences in PISA results.
Given these different possibilities, retention might reflect regional differences or varying levels of development. Yet, if we were to provide both a positive analysis and a normative claim concerning education and test scores we can point to the importance of retention at the primary level on PISA scores in math and science.
Contrary to arguments in favor of retention, we see that for Latin American countries high levels of retention and low PISA scores coincide. We show that there is a correlation between high levels of retention at the primary level and poor PISA results. We believe that retention does have an impact on test scores by negatively influencing student performance in school, as reflected by PISA results.