Perspectives in Education- Part 14: Does Classroom Size Matter?
Numbers and Educational Outcomes
Shifting the subject matter somewhat from vouchers and funding, one important aspect of educational reform is the subject of classroom size. Specifically, what is the ideal class size for optimal learning? Unfortunately, the results of numerous studies when implicated at the policy level results in a one-size-fits-all attitude that often flagrantly ignores other equally important factors. Such is the mindset of the NEA and other liberal groups. Intuitively, one would think that the smaller the class size, the more optimal the learning experience. As the NEA and teacher unions assert, with smaller class size comes more individualized learning, greater opportunity for teachers to identify and work with slower learners, better discipline, and a host of other teacher Utopian ideals. Today, the national teacher to student ratio is 15.5:1 overall which is damn well near the 15:1 ratio the NEA officially has settled upon. However, their argument is that in the lower grades, this ideal is not being met. The national average in kindergarten- a child’s first exposure to formal public education- is 20.6:1 and in elementary schools (K-8) it is 20:1. The overall 15.5:1 is achieved by factoring in high school ratios which average 11.9:1. In other words, the NEA is not satisfied until the lower grade ratios are lowered and/or the secondary school levels are raised.
For every child added or subtracted from these ratios on a national basis amounts to $12 billion annually. So, if we wanted to add a child to that ratio, it would save the country $12 billion and to reduce class size would cost the country an additional $12 billion annually. This is the cost required to add or subtract teachers from the workforce. Classroom size reduction (CSR) is not a new phenomena by any means. It is, in fact, one of the primary causes of the increase in educational costs over the past twenty years. One of the biggest payoffs in the Obama Porkulus was to the teacher unions and educational spending which sought to decrease class sizes. Whenever Obama opens his mouth in this area, usually to justify more educational dollars, the goal is more teachers teaching fewer children. If we lowered that national elementary school ratio from the current 20.1:1 to the NEA/Obama-preferred 15:1, it would cost $60 billion annually.
Left out of that equation, however, is the need for more physical classrooms which would clearly inflate these costs considerably above the $60 billion. But, the NEA contends, we are concentrating on upfront costs. Instead, we should be considering the longer term benefits like a better-educated elementary or middle school student, higher graduation rates, more kids college-ready, lower incarceration rates, and a whole host of other societal benefits. Where have we heard these “promises” before?
Any discussion of classroom size and its effect on academic performance must begin with the STAR study in the 1980s involving Tennessee schools. Without getting into the details, the results showed that a 30% difference in class size- 22 versus 15 students- resulted in academic achievement improvement equivalent to three extra months of schooling per year. The smaller the class size, the greater the academic achievement results. In fact, the NEA to this day relies solely on the STAR study- a study out of only one state- as justification for their 15:1 ideal. The STAR study was replicated in Texas and in Israel. In both those cases, decreased class size showed an improvement in academic achievement, but not to the degree illustrated in Tennessee. Results from California showed a mixed bag of results- improvement in some metrics in some grades, but not in others. Conversely, replicated studies in Connecticut and Florida indicated absolutely no correlation between classroom size and academic performance or improvement.
What concerns the liberal education establishment is this fact: with every additional student in a class it necessitates 7% less teachers. With average costs nationally to educate a single child around $11,000 per year, about $3,600 of that is dedicated to teacher salary. But, lets go the other way. Lets assume the national student to teacher ratio is 16:1. To hire enough teachers to lower that ratio to 15:1 would cost $12 billion alone. But, it would also necessitate the construction of another 225,000 classrooms nationwide. Besides the payroll concerns, there is the capital improvements to be made to the schools themselves. This cost is neglected by the NEA. Instead, they insist on some “magic number” to create educational reform.
The Heritage Foundation did an analysis on class size and found that there were several factors which probably affected student performance to a greater degree than class size. One key factor was the amount of reading material in the home and the educational level of the parents. Also, teacher quality figured into the equation.
Choosing ten states from among a list of high black population states, high Hispanic population states and high white population states, I chose the highest test score state, lowest test score state and average test score from each grouping. Then, looking at the teacher to student ratio in each state, I looked to see if there were any differences on the average of the 4th and 8th grade reading and math NEAP scores. For high schools, using the same states and secondary school ratios, I used graduation rates (since the NEA claims smaller class size means higher graduation rates) and SAT scores which are an indication of college readiness. The states chosen were: California, Illinois, Indiana, Louisiana, Massachusetts, Minnesota, New Jersey, Tennessee, Virginia, and West Virginia.
For elementary school levels, three states- California, Louisiana, and Virginia- had teacher/student ratios above the national average for this grade level. These were the high class size states. The average states in terms of class size were Indiana, Minnesota, New Jersey and West Virginia. The small class size states- those with ratios below the national average- were Louisiana, Massachusetts and Tennessee. The large class size states averaged 256.7 on the reading test and 238 on math. The average class size states averaged 257.4 in reading and 243.4 in math- better performance than the large class size states. This would seem to partially confirm the beliefs of the NEA. Except, when we look at the low class size states, the reading scores averaged 254.7 (the worst of the three groupings) and 240.5 in math. It is difficult to discern why states with smaller than the national average class size perform worse than states with average (i.e., larger) class sizes. The most amazing thing is that the difference between the large and average class size states was three students- 23 to 20. But the low class size states averaged 13 students. The only thing that can be gleaned from this is that low class size does not guarantee increased student performance in elementary school.
Looking at the high school figures, the higher than normal class size states numbered four and averaged 16 students. Their graduation rate was the lowest of the three groups at 72.5 and SAT scores averaged 1609.5. The two average states class size was 11.4 and they had the highest graduation rate at 79.2 (or about 10% better than the large class size states), and their SAT scores were the highest at 1633. Meanwhile the small class size states had a graduation rate of 76.5% and SAT score average of 1578.5. The lower class size does not nessarily translate into enhanced college preparedness. Nor does it increase graduation rates. One would expect that as class size decreases, one should see corollary increases in SAT scores and graduation rates. We see that expected trend in neither instance. The reason for the lowest graduation rate in the large class size states is likely attributable to that age group. An few disruptive 16-year-old students in a class of 17 is more deleterious to the educational environment of a few disruptive 11-year-old students in a class of 22 students.
Obviously, something more than a magical number is at play in determining student performance. The NEA will still claim that the student to teacher ratio is unacceptable in the most formative years- kindergarten through third grade. What the NEA neglects to mention is that in respect to kindergarten, most classes today have an aid in that classroom. If we count them as perhaps 1/4 of a teacher, the national average in kindergarten drops from 20.6 students per teacher to 16.5 students. So right off the bat, the NEA is misstating some facts.
As the statistics prove, we are at the ideal number of students per class if we accept the status quo. Of course, the goal is improvement. But simply decreasing class size which necessitates the hiring of more teachers and the building of more classrooms is not the panacea to cure academic stagnation.
In private schools, the average class size is 13.5 students per teacher. Here, the NEA must be in a pickle. While they can use this as an argument for smaller class sizes, they would have to also acknowledge two key things: (1) private schools outperform public schools and (2) teacher salary makes no difference in student outcomes. As has been mentioned in previous articles. there are likely other factors which determine the better performance by private schools. With charter schools, which are a relatively new phenomena, class size also is lower than that in public schools. But as they grow in popularity and as a choice versus public schools, class size will likely increase. Regarding their student performance metrics, they are generally mixed with some doing well and others being failures.
This may eventually boil down to the old quality versus quantity conundrum. The NEA goal is simply numbers, although they give lip service to quality factors like teacher credentials, pay scales, etc. Ironically, as I tried to illustrate in a previous article, providing vouchers for private or parochial schools and increasing the number of charter schools, public school class size will necessarily decrease as private, parochial and charter school class size increases in a natural sorting of school demographics as choices become available to students and parents.
Having taught elementary level classes, do I prefer a class of 15 versus 25? Absolutely since the laws of average dictate that with more students, there is a greater chance of disruptive students. But, can a class of 25 students be taught adequately? The answer again is “yes.” To illustrate, a lot depends on the subject matter, student interest and how the class is taught. In one particular case, I had to teach seventh grade science- three different classes. One class had 14 students, another 15 and a third 23 students. The subject was the human reproductive system. Although all three classes were involved and full of questions (they did not want to go to language arts), the class with 23 students, despite having four very disruptive students, learned more and asked the more demanding questions. Compare this to a special ed class of four seventh grade students. Learning writing and math in a setting outside the classroom one day, those four students were the most unruly and likely learned nothing. The moral of the story is: class size makes little difference. Or to paraphrase the situation: “Its the motion of the ocean, not the size of the vessel that counts.”