Sorting and Regression-Discontinuity Design in

Sorting and Regression-Discontinuity Design in a Public School Setting With Class-Size Caps The importance of class size in determining student achievement has grown even more contentious in recent years as class sizes are increasing due to decreases in state funding for education. Yet the new emphasis on standardized testing has made schools more determined than ever to ensure that 'no child is left behind' in the 'race to the top' for academic excellence.

Schools wishing to increase class sizes to reduce costs yet still ensure a high level of quality of education for their students will be heartened by the findings of the Second Year Paper of Olesya Baker, dated October 1, 2009, entitled "Sorting and regression-discontinuity design in a public school setting with class-size caps." Baker states that class size plays less of an emphasis on student achievement than previous educational studies have indicated.

Baker believes that other factors at least partially cause the apparent advantage of being a part of a class with low enrollment. Mandated limits on class size, she writes, automatically create a 'shortage' of spaces in the enrollment rosters of more desirable schools. More well-educated and knowledgeable parents tend to be more aware of prioritized registration dates, which "may explain why students in schools with class sizes of exactly 20 students have higher test scores and come from higher socio-economic backgrounds" (Baker 2009, p.3).

But while Baker's hypothesis is intriguing, her 'vigilance' hypothesis not proven: and it is difficult to see how it could be proven, without interviewing a cross-section of parents from top-tier and lower-tier schools. Baker states that the effect of reduced class size is overstated upon children's education, but even she herself admits this is not necessarily due to the lack of potential of individualized attention to make a difference in student's lives.

Rather, her findings indicate a lack of feasibility of class size reduction, given current economic constraints in the school districts that could benefit disadvantaged students the most: "Finally, this paper shows that class-size reduction policies can overlook those children who can benefit from smaller classes the most. Non-complying schools are composed of poorer students and previous research has shown that poor children benefit the most from class-size reductions.

Furthermore, the stacking of schools at enrollment multiples of 20 shows that even in complying schools, the financial incentives mechanism does not provide enough subsidies to promote creation of new classrooms and results in the overloading of excess students. Overloaded students are late registrants and are therefore more likely to come from a disadvantaged background. Such students could benefit the most from attending small classes in a highly demanded home school" (Baker 2009, p.15). This suggests that a lack of funding, not class size reduction policies themselves are the issue.

Schools that could not comply with class-size reduction plans and receive the financial incentive for doing so: "had difficulty acquiring the needed space to expand the number of classrooms because they were primarily located in densely populated, urban districts" (Baker 2009, p7). Baker's paper highlights the dangerous policy of 'overloading' -- where districts with class size limits over the minimum who still want the $800 financial bonus for limiting numbers of students can bus students to other areas, often inferior districts with no such class size restrictions (Baker 2009, p.5).

While such a policy is deplorable, it does not mean that class sizes have no effect upon student learning, rather that the subsidy policies may need to be rethought, and there may need to be restrictions placed upon how overburdened schools deal with 'excess' students.

The difficulty in measuring the benefits of class sizes is partially due to a host of factors difficult to control for statistically: students with mild learning disabilities, language difficulties, and special needs might benefit greatly from smaller class sizes even if they have resource room support. However, their measurable test gains may be less than students who do not have such challenges.

Quality of instruction is also an issue: perhaps good instructors and smaller class size is best, but what is 'better' -- an average teacher with a small class or a good teacher with a large class? Baker suggests the former scenario is superior, given that smaller class sizes can increase the need for teachers -- good or bad -- leaving many students in the hands of less competent teachers. However, at some threshold class management becomes impossible, even for a good teacher, and the line is difficult to draw.

Benefits may be different based upon student's gender, preparation level of the student, social class, and also the preparation level of the teacher. More experienced teachers may be better able to cope with larger classes, and also more adept at using smaller class size to individualize their curriculum while still meeting benchmarks. The benefits of small class sizes may be cumulative, even if the overall, measured year-to-year effects are small. That would explain such problems as the discrepancies between different grade levels in measured effects of class sizes.

But Baker sees such shifts as based on micro rather than macro concerns "the increase in test scores is offset by other test score fluctuations that occur between enrollments of 20 and 25. The positive ratio for third grade students (rather than negative as for second graders) is likely due to the fact that increases in test scores to the right of enrollment threshold are smaller for third than for second graders and, on the average, are offset by other fluctuations" (Baker 2009, p.13).

Even in light of her research hypothesis and skeptical conclusions Baker admits there are some indications that class size has a demonstrable effect upon student achievement. "Findings of school non-participation and creation of enrollment cutoffs [merely] demonstrate the unanticipated consequences that a class-size reduction program based on financial incentives can create, and should be taken into account when designing future class-size reduction policies" (Baker 2009, p. 5; 3). Baker's research highlights a number of questionable policies of California school districts, such as shuttling excess students to overcrowded districts.

It is arguable that given the population and financial challenges of California at this.