Impact of Class Size 1 Running Head: Impact of Class Size AN ANALYSIS OF THE IMPACT OF CLASS SIZE ON STUDENT ACHIEVEMENT By Roe, Stephen Submitted to The Educational Leadership Faculty Northwest Missouri State University Department of Educational Leadership College of Education and Human Services Maryville, MO 64468 Submitted in Fulfillment for the Requirement for 61-683 Research Paper Summer 2012 July 30, 2013
Impact of Class Size 2 ABSTRACT The purpose of the study was to analyze whether or not class size has an effect on student achievement based on a series of common formative assessments taken throughout the year. The research includes findings that answer the question, Is there a difference in student achievement based on common formative assessment data between smaller and larger classes? The research was conducted using the same students in a 6 th grade classroom throughout the 2012-13 school year. The findings were analyzed through Microsoft Excel and A Statistical Program (ASP) software utilizing ANOVA analysis. Findings indicate that there is no difference in student achievement according to class size. Further data analysis of class size and its affect on student achievement through other forms of assessment, such as state standardized testing, is recommended. The school district may also want to look at other schools within the district that are seeing higher levels of incoming student population and less outgoing student population.
Impact of Class Size 3 INTRODUCTION Background, Issues, and Concerns The question of whether or not classroom size affects the education of students has been around for decades in the United States. This study s concern is within a single grade level of middle school students in a Midwestern School District. In order to compensate for higher numbers of students moving into the district from surrounding areas and an ever-decreasing budget, the district has opted to increase classroom size. It can be argued that this is the other choice in the situation whereas any other choice will result in cuts across the board, but does it affect student learning? The district stands by the policy that the teacher is the single tool most responsible for student learning regardless of what they have to work with. The school s motto is 850 students + 100 staff = 1 goal. Many teachers in the building believe that their classes with 30 or more students do not learn as much or as well as classes with only 22 students. Practice under Investigation achievement. The practice under investigation is whether class sizes have an effect on student School Policy to be Informed by Study In order to avoid budget cuts across the board, schools opt for increased class sizes. School teachers and officials can view student achievement in the same grade with the same teacher and analyze the differences in student achievement based on larger and smaller class
Impact of Class Size 4 sizes. Other school officials will be able to determine if larger class sizes would positively or negatively impact their own school system Conceptual Underpinning Class size is a factor in student achievement. Students achieve more in a smaller classroom setting. These smaller classes for students result in higher scores on common formative assessments given to all classes. In theory, it is the teacher who matters most to education, but there are variables that impact the teacher. One of the most important such variables is class size. With a large class size, there are more discipline problems, more needs, more concerns, and more to grade. All of these create more stress on both the teacher and the student. With a smaller class, the teacher is able to provide more personal assistance to each student individually. Following this theory, the classes that have the lowest number of students will be highest achieving and the classes that have the larger number of students will be lowest achieving based upon CFA data. Statement of the Problem There is a lack of evidence in the school district to show whether there is a relationship between class size and student achievement. Larger class sizes have been appearing particularly within the past five years. Larger class sizes have been implemented as a result of large numbers of students moving into the district and a decreasing budget. These larger class sizes have resulted in increased disruptions during class, more needs from students, and more for teachers to grade and thus resulted in lower achieving scores.
Impact of Class Size 5 Purpose of the Study The purpose of this study is to ascertain if a relationship exists between class size and student achievement at the middle school level. The information gained will help administrators and school officials know if smaller class sizes have a direct benefit on students and school environment in the future. Research Question RQ: Is there a difference in student achievement based on common formative assessment data between smaller and larger classes? Null Hypothesis Ho: There is significant difference in student achievement based on common formative assessment data between smaller and larger classes. Anticipated Benefits of the Study The result of this study will inform school officials and administrators about the impact of class sizes upon student achievement. It will help convince school officials to hire new teachers and decrease class sizes. It will help inform other schools and districts about the benefits of smaller class sizes. Definition of Terms DESE: Missouri Department of Elementary and Secondary Education The Department of Elementary and Secondary Education is the administrative arm of the Missouri State Board of Education. STAR: Student Teacher Achievement Ratio
Impact of Class Size 6 STAR is a research program in Tennessee that specifically targeted class size and its impact on student achievement at various levels. NEA: National Education Association NEA is the largest professional educator organization in the United States. Summary The school district has had to increase class size as a result of students moving into the district and a decreasing budget. This research investigates the use of smaller and larger class sizes within a single middle school at the 6 th grade level to determine the impact of class size on student achievement. The research studies various arguments by published authors.
Impact of Class Size 7 REVIEW OF LITERATURE Whether or not larger and smaller class sizes have an impact on student achievement has been an argument for decades among teachers and school officials. It has primarily been an argument in public schools, due to the fact that private schools can limit the number of students in each classroom by controlling who is and is not accepted. Public schools must offer education to all who enroll. One of the many misconceptions when discussing class sizes is the student/teacher ratio. When DESE calculates that the school district has a student/teacher ratio of 14 and the state of Missouri has a student/teacher ratio of 13, that is including all teaching professionals in the building (DESE, 2012). This includes principals, counselors, Special Education teachers, and Encore teachers. This includes teachers who have 30 students per class and teachers who only have 2 students per class. This includes Instructional Assistants who do not usually have specifically defined classes. This number is not accurate when discussing the actual school classroom with the number of students in the room and one teacher. The earliest studies on classroom size and the affect it has on student achievement were done by Ross and McKenna in 1955 and the NEA in 1968. Both of these studies favored smaller class sizes as being more beneficial to students. In 1978, the Educational Research Service conducted a similar study and found that class sizes only affected a small number of students in the primary grade levels. It argued that students above primary grade levels were not affected by being in larger classes. In 1986, Slavin introduced a new study that showed that when class sizes were decreased from 27 to 16 (about 40%), students achieved more, although it was a very slight
Impact of Class Size 8 increase. No matter how slight the increase, Slavin still counted it as higher achieving through smaller class sizes. (Robinson, 1990) The most recent peer-reviewed study was the STAR Experiment. The STAR Experiment focused on 25 years of data from Tennessee schools. STAR found that smaller class sizes in grade K-3 resulted in short and long-term benefits throughout students educational career. Some of the short-term benefits listed include improved test scores, improved school engagement, reduced grade retention, and greater benefits for poor, minority, and male students. Some of the long-term benefits from STAR include more students taking the SAT and ACT exams, higher graduation rate, and taking more advanced courses in high school. This data helps show that primary grades are where class sizes need to be effectively lowered to increase student achievement in the short-term and long-term. (Achilles, 2012) A similar study called the Wisconsin Student Achievement Guarantee in Education led to similar conclusions. Students in classrooms of 12 to 15 students achieved higher school than students in classrooms of 21 to 25 students. Teachers were also reported as having more knowledge about students, more daily instructional time, and fewer discipline and behavior issues. Teachers were better able to work individually with students to help them understand difficult concepts. (Robinson, 1990) The United States federal government has even been a large part of both sides of the argument over class size. In 1999, the federal Class Size Reduction Act was authorized. The CSR initially allowed for 1.2 billion dollars to be spent on hiring and training new teachers. The funding rose in 2000 to 1.3 billion and again in 2001 to 1.623 billion dollars. This led to over 37,000 new teachers being hired with federal money that helped to decrease class sizes (USGE, 2004). On the same page, NCLB, introduced in 2001, repealed the CSR program. NCLB does
Impact of Class Size 9 not continue the direct link that CSR had of lowering class sizes by hiring more teachers (NEA 2013). NCLB follows guidelines that many districts have adopted during the recent economic recession. It believes that the teacher is the most important tool in the classroom and that a highly qualified teacher may overcome any obstacles within the classroom including the number of students in the room. NCLB provides funding for grants to train teachers, but no longer to hire them (Chen, 2008). The minimum class size standard for Missouri for grades 5-6 is 30 with the desirable standard being 25. This number can raise by five students if a half-time aide is placed in the classroom and by ten students if a full-time aide is placed in the room. With a full-time aide, Missouri approves up to 40 students in a single classroom. (DESE 2006) New York has faced similar issues with large numbers of students enrolling with a decreasing budget. As with many other cities and states, they have not developed a permanent solution. An article from the NY Times in 2003 displayed a scenario that has been played out in classrooms across the country with the opening of the teacher, Brent Wyso, on the first day of school being: Boys and girls, I apologize for not having enough chairs We will be getting more chairs (Winerip 2003, p.1 ). With thirty-seven students in the class, some students sat on a bench. Other classes did not have a bench. They used the floor with clipboards to do their work. Other classes used the teacher s desk. Students operating under these conditions are expected to achieve similar results to students who have a classroom size of twenty and more one-on-one time with teachers. (Winerip, 2003)
Impact of Class Size 10 RESEARCH METHODS Research Design A quantitative statistical analysis of CFA data served as the research design. The alpha level was set at 0.25 for all tests with this research. The independent variables is class size. The dependent variable are the assessment scores. An ANOVA test will be used. Study Group Description The middle school under study is an accredited school district within the school district. The study group is 103 students. Of those students, 76 (73.8%) are White, 10 (9.7%) are Hispanic, 8 (7.8%) are Black, 7 (6.8%) are Asian, and 2 (1.9%) are Other. 72.8% percent of students are eligible for free or reduced lunch. The average class size is 26:1. Class 1 has 29 students, Class 2 has 27 students, Class 3 has 26 students, and Class 4 has 21 students. The Classes will be arranged in sub groups of high, medium, and low class sizes with Class 1 being high (1), Classes 2 and 3 being medium (2), and Class 4 being low (3). The student make-up of the classes have seventy-six white, ten Hispanic, eight black, seven Asian, and two other minorities. Data Collection and Instrumentation For this research, collected data was used from common formative assessments given throughout the year. This data was taken from the 2012-2013 school year.
Impact of Class Size 11 Statistical Analysis Methods A Statistical Package (ASP) software was used to complete the statistical calculations in this study. An ANOVA test was calculated.
Impact of Class Size 12 FINDINGS Data for ANOVA Analysis of Class Size and Student Achievement Table 1 Summary of Descriptive Statistics for CFA 1 Size Groups N Mean SD 1 29 10.7931 0.57373 2 53 10.5849 0.42439 3 21 10.9524 0.67421 One hundred three students were selected for a study to determine if there is a correlation between class size and students achievement based on Common Formative Assessment scores taken throughout the year. These one hundred three students were divided into three groups based on class size. Group 1 contains 29 students from the largest class. Group 2 contains 53 students combining medium size classes 2 and 3. Group 3 contains 21 students from the smallest class. The null hypothesis states that there is not a difference between student achievement based on class size. These groups were analyzed and broken down using the ANOVA test to identify if there was a significant difference to propose that class size directly impacted student achievement based on CFA scores. The mean of Group 1 is 10.7931 and the standard deviation is 0.57373. The mean of Group 2 is 10.5849 and the standard deviation is 0.42439. The mean of Group 3 is 10.9524 and the standard deviation is 0.67421.
Impact of Class Size 13 Table 2 Summary of ANOVA Test of Significance Results for CFA 1 Source SS df MS F p-value Class size 2.25603.57373 1.12801 CFA 1 954.579 100 9.54579 0.118169 0.88867 Note: Significance = < 0.25 After dividing the results of CFA 1 into three categories based on class size, an ANOVA test was completed to test the null hypothesis. The null hypothesis stated there is no difference in students achievement based on class size. The data collected for CFA 1 illustrates the SS, sum of squares, was 2.25603; the df, standard error of freedom, was 1) 0.57373, 2) 0.42439, 3) 0.67421; and the MS, mean squared, was 1.12801. The F, Fisher ratio, was found to be 0.118169. The test reported a p-value of 0.88867, which is higher than the alpha level of 0.25; therefore, the null hypothesis is not rejected. This suggests that student achievement is not affected by class size. Table 3 Summary of Descriptive Statistics for CFA 2 Size Groups N Mean SD 1 29 11.1724 0.46183 2 53 10.8113 0.34162 3 21 11.2857 0.54271
Impact of Class Size 14 One hundred three students were selected for a study to determine if there is a correlation between class size and students achievement based on Common Formative Assessment scores taken throughout the year. These one hundred three students were divided into three groups based on class size. Group 1 contains 29 students from the largest class. Group 2 contains 53 students combining medium size classes 2 and 3. Group 3 contains 21 students from the smallest class. The null hypothesis states that there is not a difference between student achievement based on class size. These groups were analyzed and broken down using the ANOVA test to identify if there was a significant difference to propose that class size directly impacted student achievement based on CFA scores. The mean of Group 1 is 11.1724 and the standard deviation is 0.46183. The mean of Group 2 is 10.8113 and the standard deviation is 0.34162. The mean of Group 3 is 11.2857 and the standard deviation is 0.54271. Table 4 Summary of ANOVA Test of Significance Results for CFA 2 Source SS df MS F p-value Class size 4.45344.46183 2.22672 CFA 2 618.537 100 6.18537 0.359998 0.698578 Note: Significance = < 0.25 After dividing the results of CFA 2 into three categories based on class size, an ANOVA test was completed to test the null hypothesis. The null hypothesis stated there is no difference in students achievement based on class size. The data collected for CFA 2 illustrates the SS, sum of squares, was 4.45344; the df, standard error of freedom, was 1) 0.57373, 2) 0.34162, 3)
Impact of Class Size 15 0.54272; and the MS, mean squared, was 2.22672. The F, Fisher ratio, was found to be 0.359998. The test reported a p-value of 0.698578, which is higher than the alpha level of 0.25; therefore, the null hypothesis is not rejected. This suggests that student achievement is not affected by class size. Table 5 Summary of Descriptive Statistics for CFA 3 Size Groups N Mean SD 1 29 11.5862 0.38379 2 53 11.2642 0.28389 3 21 11.2381 0.54271 One hundred three students were selected for a study to determine if there is a correlation between class size and students achievement based on Common Formative Assessment scores taken throughout the year. These one hundred three students were divided into three groups based on class size. Group 1 contains 29 students from the largest class. Group 2 contains 53 students combining medium size classes 2 and 3. Group 3 contains 21 students from the smallest class. The null hypothesis states that there is not a difference between student achievement based on class size. These groups were analyzed and broken down using the ANOVA test to identify if there was a significant difference to propose that class size directly impacted student achievement based on CFA scores. The mean of Group 1 is 11.5862 and the standard deviation is 0.38379. The mean of Group 2 is 11.2642 and the standard deviation is 0.28389. The mean of Group 3 is 11.2381 and the standard deviation is 0.54271.
Impact of Class Size 16 Table 6 Summary of ANOVA Test of Significance Results for CFA 3 Source SS df MS F p-value Class size 2.27158.0.38379 1.13579 CFA 3 427.146 100 4.27146 0.265902 0.767054 Note: Significance = < 0.25 After dividing the results of CFA 3 into three categories based on class size, an ANOVA test was completed to test the null hypothesis. The null hypothesis stated there is no difference in students achievement based on class size. The data collected for CFA 3 illustrates the SS, sum of squares, was 2.27158; the df, standard error of freedom, was 1) 0.38379, 2) 0.28389, 3) 0.54100; and the MS, mean squared, was 1.13579. The F, Fisher ratio, was found to be 0.265902. The test reported a p-value of 0.767054, which is higher than the alpha level of 0.25; therefore, the null hypothesis is not rejected. This suggests that student achievement is not affected by class size. Table 7 Summary of Descriptive Statistics for CFA 4 Size Groups N Mean SD 1 29 11.7931 0.36800 2 53 11.9245 0.27222 3 21 11.7143 0.43246
Impact of Class Size 17 One hundred three students were selected for a study to determine if there is a correlation between class size and students achievement based on Common Formative Assessment scores taken throughout the year. These one hundred three students were divided into three groups based on class size. Group 1 contains 29 students from the largest class. Group 2 contains 53 students combining medium size classes 2 and 3. Group 3 contains 21 students from the smallest class. The null hypothesis states that there is not a difference between student achievement based on class size. These groups were analyzed and broken down using the ANOVA test to identify if there was a significant difference to propose that class size directly impacted student achievement based on CFA scores. The mean of Group 1 is 11.7931 and the standard deviation is 0.36800. The mean of Group 2 is 11.9245 and the standard deviation is 0.27222. The mean of Group 3 is 11.7143 and the standard deviation is 0.43246. Table 8 Summary of ANOVA Test of Significance Results for CFA 4 Source SS df MS F p-value Class size 0.772115 0.36800 0.386057 CFA 4 392.742 100 3.92742 0.0982979 0.906466 Note: Significance = < 0.25 After dividing the results of CFA 4 into three categories based on class size, an ANOVA test was completed to test the null hypothesis. The null hypothesis stated there is no difference in students achievement based on class size. The data collected for CFA 4 illustrates the SS, sum of squares, was 0.772115; the df, standard error of freedom, was 1) 0.36800, 2) 0.27222, 3)
Impact of Class Size 18 0.43246; and the MS, mean squared, was 0.386057. The F, Fisher ratio, was found to be 0.0982979. The test reported a p-value of 0.906466, which is higher than the alpha level of 0.25; therefore, the null hypothesis is not rejected. This suggests that student achievement is not affected by class size. Table 9 Summary of Descriptive Statistics for CFA 5 Size Groups N Mean SD 1 29 12.0000 0.32772 2 53 12.1509 0.24242 3 21 12.6667 0.38512 One hundred three students were selected for a study to determine if there is a correlation between class size and students achievement based on Common Formative Assessment scores taken throughout the year. These one hundred three students were divided into three groups based on class size. Group 1 contains 29 students from the largest class. Group 2 contains 53 students combining medium size classes 2 and 3. Group 3 contains 21 students from the smallest class. The null hypothesis states that there is not a difference between student achievement based on class size. These groups were analyzed and broken down using the ANOVA test to identify if there was a significant difference to propose that class size directly impacted student achievement based on CFA scores. The mean of Group 1 is 12.000 and the standard deviation is 0.32772. The mean of Group 2 is 12.1509 and the standard deviation is 0.24242. The mean of Group 3 is 12.6667 and the standard deviation is 0.38512.
Impact of Class Size 19 Table 10 Summary of ANOVA Test of Significance Results for CFA 5 Source SS df MS F p-value Class size 5.84185 0.32772 2.92093 CFA 5 311.459 100 3.11459 0.93782 0.394896 Note: Significance = < 0.25 After dividing the results of CFA 5 into three categories based on class size, an ANOVA test was completed to test the null hypothesis. The null hypothesis stated there is no difference in students achievement based on class size. The data collected for CFA 5 illustrates the SS, sum of squares, was 0.5.84185; the df, standard error of freedom, was 1) 0.32772, 2) 0.24242, 3) 0.38512; and the MS, mean squared, was 2.92093. The F, Fisher ratio, was found to be 0.93782. The test reported a p-value of 0.394896, which is higher than the alpha level of 0.25; therefore, the null hypothesis is not rejected. This suggests that student achievement is not affected by class size.
Impact of Class Size 20 CONCLUSIONS AND RECOMMENDATIONS The null hypothesis stated that there is no difference in student achievement based on class size. The results of this study indicate that there is no difference in students achievement based on class size. The scores of students on the various common formative assessments through the 2012-13 school year improved as a whole from beginning to end, but did not improve at a significantly higher rate in the small class compared to the medium or large class. Group 3 scored higher than Groups 1 and 2 on only three assessments; the scores were not significantly higher on these assessments. There are some causes that might explain why class size did not affect student achievement. Only one hundred three students were used because only one hundred three students completed all five assessments. There were another thirty-two students who moved in and out of the classes without completing all five assessments. Only two of these thirty-two students were enrolled in Group 3, the small class. Another cause may have been the scheduling of the classes. The only class with a built-in disruption in the schedule is in Group 3. Group 3 s schedule goes to lunch directly in the middle of class. While the schedule reflects the same number of teaching minutes as compared to other classes, the realistic teaching minutes is five to ten minutes less depending upon multiple facts such as disruption from lunch, bathroom and drink breaks following lunch, and returning late from eating with administration or other staff in the building. Although class size did not affect student achievement in this case, it may affect student achievement when compared to other assessment indicators such as state standardized testing. It has also not been significantly proven that class size has no affect on other aspects of school. One such factor that class size may still affect is disciplinary actions taken by teachers and
Impact of Class Size 21 administrators in large class vs. small classes. Class size may impact behavior due to the lower ratio between student and staff. Another possible factor of class size is the lack of materials. Not all classrooms are fit to handle large class sizes based on space, books, and various other materials. Districts should continue to consider the consequences of increasing class sizes in their buildings. Future studies in this area may include various testing. The current test could be altered to include more data from student testing such as state standardized testing and other classroom assessments. Learning is not only shown from a single set of common formative assessments. Another possibility for future studies may include more classrooms with more varying class sizes. The comparison of classrooms of fifteen and thirty students should prove interesting. At what point, if any, does class size impact more greatly student achievement. The same one-way ANOVA test can be used to calculate these possible future studies.
Impact of Class Size 22 REFERENCES Achilles, Charles M. (2012). Class-size policy: the STAR Experiment and related class-size studies. Retrieved from http://www.eric.ed.gov/pdfs/ed540485.pdf Chen, Grace. (2008). Class size: federal funding of class size reduction. Retrieved from http://www.publicschoolreview.com/articles/17 Department of Elementary and Secondary Education. (2006). Standards and indicators manual. Retrieved from http://dese.mo.gov/divimprove/sia/msip/fourth%20cycle%20standards%20and%20indic ators.pdf. Department of Elementary and Secondary Education. (2012). Missouri Comprehensive Data System. Retrieved from http://mcds.dese.mo.gov. National Education Association. (2013). Class size: priorities for changing no child left behind. Retrieved from http://www.nea.org/home/13120.htm Robinson, Glen E. (1990). Synthesis of Research on the effects of class size. Retrieved from http://lcsd.k12.ny.us/cms/lib/ny01001015/centricity/domain/366/synthesis%20of%20re search.pdf United States Government of Education. (2004). A descriptive evaluation of the federal classsize reduction program. Retrieved from http://www2.ed.gov/rschstat/eval/other/classsize/report.pdf. Winerip, Michael. (2003). On education; no child left behind law leaves no room for some. The New York Times. Retrieved from http://www.nytimes.com/2003/09/10/nyregion/on-education-no-child-left-behind-lawleaves-no-