Does class size matter?
An update...

Why are we doing so poorly?

One reason commonly heard, is that classes are too large. Teachers don't have the time to provide individualized attention when they've got kids stacked up to the rafters in their classrooms.

But does class size really matter? SuperKids looked at this question 10 years ago, and the answer then was "yes". Using data from a standard 4th grade reading test given in all 50 states, we looked to see if there was a correlation between test scores and class size. Our analysis then showed that states with more small classes generally had higher scores than states with fewer small classes. Smaller classes presumably allowed teachers to provide more attention to their student's individual needs, resulting in more learning, and better performance on the tests.

Ten years later, we wondered if the relationship still held true, and if it extended to other subjects, like math and science?

There are large differences in student/teacher ratios among the states. In 2004, public school student/teacher ratios varied by a factor of two, or 100%, from a low of 11.3 in Vermont, to a high of 22.6 in Utah. That sounded like a big difference to us, until we looked at the difference in test score performance among the states.

Using the percent of test-takers that scored 'proficient' or higher for their grade level on their tests, we discovered a vast gulf between the high and low-scoring states.

• 4th (and 8th) grade reading proficiency ranged from a low of 18% in Mississippi, to a high of 44% in Massachusetts.
  
  
• 4th grade math proficiency levels were virtually the same, from a low of 19% in Misissippi, to 49% in Massachusetts. 8th grade math scores were slightly lower, ranging from 14% in Mississippi (and New Mexico) to 43% in Massachusetts and Minnesota.
  
  
• 4th grade science proficiency levels were even lower, from just 12% in Mississippi, to 40% in Vrginia. 8th grade performance was slightly higher, ranging from 14% in Mississippi to 43% in North Dakota.

Bottom-line: we're not sure if we're more disturbed by the three-fold difference in proficiency between some states on these tests, or the fact that less than half of the students on any test, in any state, received scores of proficient or better.

Methodology
Using the same methodology used in our previous study - a standard linear regression model - SuperKids analyzed the relationship between student/teacher ratios and scores on NAEP tests given in 2005 to 4th and 8th graders in Reading, Math, and Science.

Findings
Although we found once again that there was a general correlation between class size and test scores - with smaller classes producing better results - this time we discovered that the influence of class size was stronger in some subjects than in others:

• Reading. Class size explained almost 25% of the difference in test scores in both 4th and 8th grade. How big an impact do class sizes have? For example, based only on this variable, reducing California's average student/teacher ratio (21.1 students) to the national average (15.6 students) would be expected to produce a 30% increase in the percentage of students scoring proficient or better.




• Science. Class size explained 28% of the difference in 4th grade test scores, and 16% of the difference in 8th grade test scores.




• Math. Here was the surprise of the study; class size and test scores showed an extremely weak correlation in Math, both in 4th and 8th grades, explaining only 8% and 5% of the difference in scores, respectively. Why might this subject be so different? One possibility is that many schools utilize 'tracking' in math, wherein students are placed in classes based on their ability in the subject. This could result in smaller classes for students of either advanced ability, or for those needing remedial help - or both. Either way, the impact of tracking on the data would hide any other correlation between class size and test scores.

Want to see more information on how your state compares? Check out the 2005 NAEP test results.

The Data
For our performance data we elected to continue using scores from the National Assessment of Educational Progress (NAEP) tests, a congressionally mandated survey that has been administered by the National Center for Education Statistics (NCES) in the U.S. Department of Education since 1969. Unlike many other tests now given as part of the No Child Left Behind Program, these tests are uniform across the country, do not have entire curriculums built around their questions, and don't drive teacher, school and district performance evaluations. They are a relatively pure assessment test designed to evaluate the quality of our education system, and the results are submited to Congress as part of "The Nation's Report Card."

For class size data, we turned to data for public elementary and secondary schools from the NCES for the same time period. (Note: In an ideal analysis, we would have matched class size data for specific grades with their test scores. Lacking that specific data, we used average student/teacher ratios for all grades after determining that there was a strong correlation between elementary and secondary school class sizes within each state.)

---

Questions or comments regarding this site? webmaster@superkids.com
Copyright © 1998-2024 Knowledge Share LLC. All rights reserved. Privacy Policy