Do smaller classes mean better educational achievements?
Author: Taub Center Staff Bulletin Articles

Parents want smaller classes in Israel’s education system, yet research shows that there is no notable relationship between class size in and of itself and educational achievements

The end of the school year, and preparation for the start of the next, provide a timely opportunity to evaluate large education policy trends. In fact, it was at this same time of year a couple years ago that parents and students in Israel took to the streets in what was deemed the “sardines protest” to demonstrate against overcrowded classrooms in Israel’s education system.

As part of the protest, parents from all over the country sent cans of sardines to the office of Minister of Education MK Naftali Bennett, to illustrate how students were packed into Israeli classrooms the way sardines are packed into a can.

The campaign drew legitimacy from international data on class size and student achievements. Of the 34 countries that participated in the relevant OECD study, Israel placed fifth in terms of large class size, with an average of 27 students per primary school class and 32 students per middle school class. At the same time, a look at Israeli student achievements on the OECD’s international PISA (Programme for International Student Assessment) exams, which test 15-year-olds in mathematics, reading and science, shows Israeli students to be in the lower half of the attainment ranking for all subjects tested.

It’s not a new phenomenon, nor one specific to Israel, that both parents and many teachers see smaller classes as a way to improve the achievements of students. Yet researchers Reut Shafrir, Yossi Shavit and Carmel Blank recently found that, when isolating class size from other factors that could influence student achievements – such as parental education levels and prior academic achievements – there is no notable relationship between class size and achievements in eighth grade.

The main challenge in trying to understand how class size affects student performance is that class placement is often not randomly determined, but rather reflects systemic educational considerations that themselves could potentially affect student achievements. For example, relatively low-achieving students are often placed in smaller classes from the start, in the hope that this will help improve their performance.

The study, which was published by the Taub Center, analyzes data on scores of students who took the Israeli Meitzav exams in language arts (Hebrew) in 2006 and 2009 as well as background data on these same students. In the initial analysis, the researchers found that there is actually a positive relationship between class size and achievements in Israel; the scores of students in large classes are seemingly higher than those of students in smaller classes.

However, when separating out parental education levels and prior achievements, the researchers found that the relationship between class size and achievements is not statistically significant in either direction. In this model, the greatest impact on a student’s Hebrew Meitzav exam score in Grade 8 is the score of the same student on the Grade 5 Hebrew Meitzav exam – that is to say, prior achievements. Higher levels of parental education were also found to be correlated with greater student achievements.

English box

Earlier studies in other countries indicate that ethnic minorities and those belonging to lower socioeconomic strata are likely to benefit more from smaller classes than others. The researchers evaluate the hypothesis that the impact of class size on achievements varies between students of different socioeconomic backgrounds and between stronger and weaker students in Israel. Yet the hypotheses did not hold up: no difference in the relationship between class size and achievements was found among the groups.

There are two important caveats to the researchers’ findings on class size. First, the result refers to eighth-grade test scores only, and does not guarantee that the same is true for, say, first graders. Second, while grades might not depend on class size, the classroom experience could certainly be affected, a factor that may be no less important than academic outcomes.

Similarly, small classes could facilitate the use of teaching methods that may help students achieve – for example, individualized or small-group instruction – but it is unclear whether teachers working in small classes do, in fact, take advantage of the possibilities that such classes present. If teachers use forms of instruction similar to those commonly employed in large classes, they may effectively neutralize the small-class advantage.


People who read this were also interested in: