As Ehrenburg et al (2001) said, in small-sized classes, there are less noise and less disruptive behavior, that gives the teacher more time to work individually with struggling students and gives the teacher the freedom of dividing them into groups to fulfil specific tasks (P.158). In addition, one hypothesis is that in small-sized classes, instructors are more prone to recognize low achievers and subsequently are more prone to give guideline intended to advantage these students in early grades. On the other hand, in small-size classes there is a higher probability for low achievers to communicate with educators and be more occupied with learning (Konstantopoulos & Chung, 2009, P.150). As a result, these students will engage more often in studies, which might result in a higher achievement level (Konstantopoulos & Chung, 2009, P.129). For instance, Angrist and Lavy (1999) found that CSR has increased the scores of fourth and fifth graders significantly in one the Israeli school based on nonexperimental data they used from Israeli elementary schools (as cited in Konstantopoulos & Chung, 2009, P.132). Therefore, the results show that CSR has significant positive impacts on achievement levels and engagement of low …show more content…
For example, Hoxby (2000) found that smaller classes had almost no effect on student achievement based on nonexperimental data from elementary schools in Connecticut (Konstantopoulos & Chung, 2009, P.132). In addition, Pong and Pallas found positive small-class effects on eighth-grade math achievement in the United States but he could not find in other countries. More recently, Milesi and Gamoran (2006) found no evidence of main class size effects on mathematics and reading achievement based on the data they used from the Early Childhood Longitudinal Study (as cited in Konstantopoulos & Chung, 2009, P.132). These results are good and important to be considered, but creating a small size class has a formula that some researchers may make mistakes in choosing the right formula. The formula a pupil-teacher ratio (PTR) is a division problem, but class size is an addition problem. Therefore, these two formulas are not the same, and thus PTR data cannot be used as a replacement for actual class-size data (Achilles, 2012, P.1). Although CSR did not work in some regions, but still we cannot ignore the findings of STAR and other experimental research; because, CSR worked for many of the countries and schools according to STAR and other