IMPLICATIONS FOR EQUITY, PRACTICE & IMPLEMENTATION
  STATUS OF CLASS SIZE REDUCTION INITIATIVE In July 1996, California embarked on its largest ever education reform: a nearly $1 billion class size reduction effort to improve literacy in the primary grades. Now in its second year, the Class Size Reduction (CSR) initiative provides $800 (up from $650 the first year) per student to schools that reduce class size to 20 students or fewer in first grade, second grade and, then, third grade and/or kindergarten. 1 The impetus for CSR came from several factors. A revived state economy created a revenue windfall. Under Proposition 98, a minimum amount of this surplus must be allocated to education. The decision to funnel the additional money to CSR stemmed from the belief that smaller classes would help improve early literacy, an area of much concern in light of California’s scores on national tests. The National Assessment of Educational Progress (NAEP), for example, showed that in 1994 only 18 percent of California’s students were rated proficient or advanced in reading. 2 In 1994, California’s national ranking on NAEP was second from last in reading. 3 Whether this lackluster performance was due to the content of the state’s language arts curriculum frameworks, the increasing number of uncredentialed teachers, inadequate funding for schools or other reasons, is a matter of debate. 4 California’s class size, however, was undeniably among the highest in the nation, averaging approximately 28.6 students per K-3 classroom before the initiative was passed. 5 |
| ||
Hopes are high that CSR will significantly improve student performance. Surveys show tremendous enthusiasm among educators, and news stories report widespread public support. 6 But whether CSR will prove to be the crucial lever for enhancing California’s early literacy and overall academic achievement remains to be seen. As a massive experiment, California’s CSR program is being closely watched by other states. Since its enactment about half of the other states have passed or are considering class size legislation, including Nevada, New York, North Carolina, Utah, and Wisconsin. The idea is catching on internationally as well. In 1997, officials in England and Canada put forward proposals to reduce class size. WHAT THE RESEARCH SAYS Optimism surrounding CSR is not without reason. Research suggests that reducing class size is likely to improve academic achievement. 7 Yet there remains disagreement about the magnitude of that achievement gain and how small a class needs to be before a strong effect is observed. 8 Other research is more skeptical, concluding that class size reduction policies have little or no relationship to student performance. 9 Tennessee STAR Evaluation A recent and widely-cited study, which some observers say influenced California’s CSR legislation, is the evaluation of the Student Teacher Achievement Ratio (STAR) Project in Tennessee. Unlike most education studies, the STAR evaluation was scientifically controlled, including the random selection of schools, random assignment of students and tracking of student performance over four years. 10 Findings from that longitudinal study suggest: 1) small classes (17 pupils or fewer) are more effective academically than larger classes (22-26 students) in primary grades; 2) the advantage of small classes is the greatest for minority students in urban schools, and 3) gains are primarily attributed to students spending more time engaged in learning and teachers more time instructing and less time managing their classrooms. 11 The STAR program was able to control for certain conditions in ways California did not. Therefore, others caution against generalizing results from the Tennessee study to the California experience, especially given the speed, sheer size and complexity of its effort. 12 Factors cited as unique to California include: 1) the bigger scale of implementation (whole state versus a relatively small controlled study); 2) the pre-existing teacher shortage and the large numbers of teachers that had to be hired on an emergency credential basis (whereas all of Tennessee’s STAR teachers were experienced and fully certified); 3) the diversity of California’s racial and linguistic minority student population (as compared to Tennessee’s primarily African-American and white population), and 4) the larger average class size of California’s schools (compared to that of Tennessee’s). Other Influences on Student Performance Other research suggests that what actually happens in the classroom may influence learning as much or more than the number of students there. 13 The quality of teaching, the type of instructional strategies used, the curriculum content and student-teacher interactions, along with class size, also matter. 14 Finally, student motivation and family educational and economic background play a role in determining achievement outcomes. 15 Differences in district resources and implementation strategies may affect the results of CSR. As the Policy Analysis for California Education(PACE)-WestEd study shows, the playing field among California’s districts and schools was not level when CSR passed. Some districts had efforts already underway to reduce class sizes and were preparing to hire teachers and locate facilities. Resources for implementing CSR also vary. With the help of CSR state grants, some districts were able to lower class sizes in all three grades, while others struggled to find enough teachers and facilities for just one grade. Surveys of California's CSR Initiative Other surveys during the initiative’s first year indicate that districts face common challenges but vary widely in the degree to which these are obstacles to implementation:
Another set of unanswered questions relate to CSR’s effect on teaching and learning in the classroom. A handful of districts have begun classroom data collection. They report wide-spread teacher enthusiasm for CSR and say it is benefiting student learning. 20 Still unknown is how CSR has improved the conditions for, and learning opportunities of special populations, such as students with disabilities and limited-English proficiency. Are teachers rethinking or changing their instructional practice or are they doing what they have always done, perhaps better and quicker? To what extent do years of experience, credentials and professional development make a difference when looking at student performance in smaller classes? PURPOSE OF THIS STUDY This first-year implementation study aims to provide some contextual information as background for answering these and other important questions. Another purpose is to help refine some of the policy and research questions that a more extensive and comprehensive evaluation of CSR will then explore in greater depth. That larger evaluation is currently being undertaken by a consortium of research organizations who have joined forces to provide California’s policy makers and educators with formative and summative information about the CSR initiative (see box below). | |||
|
Because earlier studies have primarily relied on surveys of district officials about the facilities, finances, teacher supply and other logistical issues involved in implementing CSR, the PACE-WestEd study focused its data collection at the school level. We also purposefully chose to collect most of our data from urban districts and schools, since many of these districts reportedly faced a more complex set of implementation challenges. While some general information was sought about the implementation of CSR, many of these issues were already adequately covered through other state surveys. 21 Thus, we chose to explore the following issues:
| ||
STUDY DESIGN Most of our data was collected from large urban districts. Although we intended to compare challenges faced by urban and non-urban districts, we were unable to do so. Because urban districts have reportedly faced particular challenges in implementing CSR, our findings only pertain to other large urban districts—and are not representative of all districts and schools across the state. We therefore refer to our sample of schools and districts as purposive rather than random. Selection of Districts In conducting interviews and surveys of district level information, four large urban districts (two in Northern California, two in Southern California) were among a purposive sample of 12. Student enrollment in these urban districts range from 60,000 to 125,000 pupils. The remaining eight districts were chosen according to: 1) per-pupil expenditures, 2) enrollment, and 3) percentage of Limited-English-Proficient students. Districts were selected as being high (above 75th percentile) or low (below 25th percentile) in the range of the previous three variables. Districts that fell into eight cells, ranging from "low-expenditures, low-enrollment, low-percentage of LEP students" to high in all categories. One district was then selected from each cell, taking into consideration the geographic location and the proximity to research staff to accommodate in-person interviews when necessary or appropriate. Selection of Schools In selecting schools, we drew from the above-mentioned district sample and one additional urban district in Southern California, -which serves approximately 625,000 students. As such, schools were selected from 13 districts, including five urban districts. Using the California Basic Educational Data System (CBEDS, 1995-96), we generated a list of all schools containing grades K-3 (excluding charter schools) for each of the 13 districts. For each of the five urban districts, we ranked schools according to LEP percentage and then divided this sample into thirds. From each third, seven schools were randomly selected, so that a total of 21 schools from each of the five urban districts was included in our purposive sample. This stratified random sample assures that California’s diversity is adequately represented, in terms of both income and ethnicity. We also were interested in focusing this study on LEP student populations. From four of the remaining eight districts, we randomly selected five schools per district without stratifying by LEP population (because of the small number of schools located within these districts). Each of the remaining four districts contained only one school, which also was selected. In sum, 129 schools were selected to participate. | |||
| |||
Data Collection Methods Four interview protocols were used with district level officials. Depending on availability, a combination of phone and/or in-person interviews was conducted, followed by short questionnaires. The primary contact at each district was the person (or people) "most responsible for overseeing the class size reduction initiative." In some cases, this was a Superintendent or Assistant Superintendent, but could also have been a Director of Elementary Education or another program director. We also asked to speak with those most knowledgeable regarding the impact CSR was having on LEP programs, special education programs, and professional development services at the district. In a few cases, especially in smaller districts, there was one respondent for all four interviews. In larger districts, we often spoke with six or seven respondents, because many of the issues we were exploring spanned across several district offices and program areas. | |||
| |||
To capture the particular perceptions of each individual, four different staff-specific questionnaires were developed, consisting of both open-ended and discrete questions. The questionnaires vary, but topics often overlap. As figure 4b shows, topics include: implications for students, particularly LEP and special education student populations; qualifications of teachers and classroom practices; staff development issues; parental involvement; and general challenges posed by CSR implementation (see figure 4b). | |||
 | |||
