Reducing Disputes in Spartanburg, SC An Evaluation of a Problem Solving Partnership Edward R. Maguire George Mason University Shellie E. Solomon 21st Century Solutions, Inc. Craig D. Uchida 21st Century Solutions, Inc. Prepared for the Spartanburg Department of Public Safety and the Office of Community Oriented Policing Services March 2003 ACKNOWLEDGMENTS A thorough evaluation depends heavily on the time, commitment, energy, and passion of many stakeholders. The evaluation team benefited from contributions by a number of such people. First, we would like to thank Tony Fisher, Spartanburg’s Director of Public Safety, for granting us full access to his agency and for sharing his perspective with us on several occasions. We would also like to thank Joyce Lipscomb, Operations Analyst for the Spartanburg Public Safety Department, for helping to arrange access to people, places, and data. Other Spartanburg police officers also provided information and assistance to the evaluation team, granting interviews, taking us on ride-alongs, and providing other information on request. We would especially like to thank Vincent Johnson, School Resource Officer for Carver Junior High School, for granting us interviews and allowing our research assistant to “shadow” him in the normal course of his duties. Both the target and comparison schools fall within Spartanburg County School District 7. We received far more assistance than we could have reasonably expected from Dr. Arthur Brewton, Director of Alternative Programs, and Ms. Brenda Story, Student Services Coordinator. We are grateful to both of them, as well as their colleagues, for their assistance in this project. We would also like to thank the former Principal of Carver Junior High School, Dr. Robin Johnson, for granting interviews and providing access, assistance, and guidance in collecting data from different sources. We selected Whitlock Junior High School as a comparison school for this project. Former Principal Rusty Wright and School Resource Officer Roosevelt Young cooperated mightily in allowing us to administer surveys and conduct interviews and observations. We would like to thank George Appenzeller and Sara Meadows, proprietors of Social Work Solutions, the firm selected by the Public Safety Department to do the local evaluation of the School Based Partnerships program in Spartanburg. Through a memorandum of understanding that divided up the research tasks, Social Work Solutions conducted much of the qualitative analysis on this project. Andrea M. Philbrick and Mindy Davis of Social Work Solutions also helped us conduct student surveys in two schools. We would like to thank Kimberly Hassell, a doctoral student at the University of Nebraska at Omaha, for her research assistance. Kimberly conducted interviews and administered student surveys. In addition, her observations of the School Resource Officers helped us understand the daily lives and activities of these officers. We also thank Belinda Beale, a teacher in Prince William County, VA, for contributing various forms of research assistance during the preparation of the final report. Finally, we would like to thank the students at Carver and Whitlock for completing our surveys and sharing their thoughts with us about school safety. It was their world we sought to understand, and we are grateful to them for providing us a glimpse of it. -Edward R. Maguire -Shellie E. Solomon -Craig D. Uchida This report and project were supported by Grant No. 1999-CK-WX-K005 awarded by the Office of Community Oriented Policing Services, U.S. Department of Justice to 21st Century Solutions, Inc. Points of view in this document are those of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice or the law enforcement agency involved in the study. Copyright © by Craig D. Uchida and 21st Century Solutions, Inc. The U.S. Department of Justice reserves a royalty-free, nonexclusive, and irrevocable license to reproduce, publish or otherwise use, and to authorize others to use, this document for Federal Government purposes. No part of this document may be reproduced in any forms, by any means (including electronic, photocopying, recording or otherwise) without the prior written permission of the U.S. Department of Justice and the authors. Published by: 21st Century Solutions, Inc. P.O. Box 12279 Silver Spring, MD 20908 (301) 438-3132 www.e-21stCenturySolutions.com CDUchida@aol.co TABLE OF CONTENTS Exhibits 1 Executive Summary 2 I. Introduction 6 II. The Evaluation Process 14 III. Implementation Process 26 IV. Impact 35 V. Conclusion 40 References 41 Exhibits 43 EXHIBITS Exhibit 1: Selected Demographic, Social and Economic Characteristics Exhibit 2: Official Crime Statistics, 2000 Exhibit 3: Characteristics of the Spartanburg Public Safety Department Exhibit 4: Mission, Vision, and Values Statement, Spartanburg Public Safety Department Exhibit 5: Spartanburg’s Three Junior High Schools Exhibit 6: Student Survey Sample Sizes Exhibit 7: Nature of Disciplinary Infractions at Carver and Whitlock Exhibit 8: Total Disciplinary Infractions at Carver, and Whitlock, 2000-2001 Exhibit 9: Monthly Dispute Totals for Carver and Whitlock, 2000-2001 Exhibit 10: School Responses to Disciplinary Infractions at Carver and Whitlock Exhibit 11: In-School Suspensions, 1998-2002 Exhibit 12: Out-of-School Suspensions, 1994-2002 Exhibit 13: Expulsions, 1994-2002 Exhibit 14: Arrests by the SRO at Carver, August 2000 - May 2001 Exhibit 15: What Happens to Cases Resulting from an Arrest at Carver? EXECUTIVE SUMMARY In 1999, the City of Spartanburg, South Carolina received a School-Based Partnerships grant for $137,989 from the COPS Office. The grant enabled the Spartanburg Public Safety Department to implement problem-solving strategies in George Washington Carver Junior High School. Although Spartanburg as a whole suffers from poverty-related problems, the area served by Carver represents an even more concentrated level of economic disadvantage. The school represents an ideal area in which to focus community energies such as the problem-solving approach used in the School Based Partnership project. The focus of the problem solving efforts was disputes, including those between students, between ethnic groups, and between students and teachers. This report highlights the findings from a process and impact evaluation of the School Based Partnership project. A variety of qualitative and quantitative methods were used to conduct an intensive case study detailing the implementation of school-based problem-solving efforts, together with a multi-part quasi-experiment designed to determine whether the project resulted in a reduction in the number or severity of disputes. Spartanburg County has seven school districts; two of these districts are located within the City of Spartanburg. Carver Junior High School, the focal school for this project, is located in Spartanburg County School District 7. Two other junior high schools are also located within District 7. The first, McCracken, serves a primarily middle class area; the other, Whitlock, serves an economically disadvantaged population similar to Carver’s. For this reason, we selected Whitlock as a comparison school for this project. Evaluation Design The evaluation team relied on a variety of data sources for examining the School Based Partnership project in Spartanburg. We conducted two student surveys: one at Carver and one at Whitlock. We also analyzed disciplinary data from both schools from 1994 to 2002. Together with Social Work Solutions, the local evaluator selected by the Public Safety Department and our research partner on this project, we conducted dozens of interviews and focus groups in the two schools, in the school district headquarters, and in the police department. We “shadowed” the school resource officers in both schools, attended Neighborhood Watch meetings, and accompanied patrol officers on ride-alongs throughout the City of Spartanburg. We analyzed official disciplinary data from the schools, official arrest data from the police, and follow-up data from the courts. One of the basic principles of the evaluation was that a true understanding of program implementation and impact could only be achieved by triangulating from a variety of quantitative and qualitative data sources. We also considered a number of evaluation design options. Unfortunately, many of the more rigorous designs for evaluating programs, such as randomized experiments, were not available to us for a number of reasons. Principal among these reasons was the timing of the evaluation. The project team had already begun its work in Spartanburg in August 2000, five months before we arrived on site for the first time in late January 2001. Therefore it was difficult to collect “true” baseline data. Second, the School-Based Partnerships program was intended to be what evaluation researchers call a “full coverage” program; it was designed to benefit all students in the school. It is much easier to evaluate the success of partial coverage programs that are implemented among a subset of students, since other subsets can serve as an effective comparison or control group. Although we were unable to select a control group within Carver Junior High, we had the benefit of a similar school (Whitlock) located within the same school district. Because both schools are in the same school district, they keep similar disciplinary records that serve as useful data sources for this evaluation. On the other hand, different police departments serve the two schools and arrest data are only available for one of them (Carver). Using these various data sources, we studied both the implementation and the impact of the project. Program Implementation Evaluation research has long noted the inability of reformers to successfully implement promising programs. Researchers often use a medical metaphor to explain the need for evaluating program implementation. Suppose, for example, that cancer patients are assigned by a researcher to ingest two tablets of an experimental medication every morning and every night. The researchers conduct the study and learn that patient health did not improve; therefore they conclude that the medication did not work. If the patients complied with the experimental protocol, the researcher’s findings would be valid, but suppose the patients didn’t actually take the medication. Or more commonly, suppose they took the medication sometimes, but did not follow the researcher’s instructions and missed numerous doses. Can the researcher truly conclude that the medication does not work? In program evaluation, this is known as a “dosage” question. If a program is not completely implemented, or it is implemented in a weak fashion, how should the evaluator determine whether or not the program is effective? Because this question is crucial in determining whether a program is ineffective, as opposed to whether it was just implemented ineffectively, evaluating program implementation is vital. The implementation of the School-Based Partnerships project in Spartanburg involved four primary components: - External training and consultation in Crime Prevention through Environmental Design. - Peer surveys of children assigned to in-school suspension. - An after-school program designed to reduce the number of students being given repeat in-school suspensions. - Anti-violence and conflict resolution training taught by officers to students assigned to in- school or out-of-school suspension. In addition to these primary components, the school also experienced a number of other relevant changes, the most important of which was a new School Resource Officer (SRO) funded through a Cops in Schools grant. Program Impact Our ability to do a comprehensive impact evaluation was hampered by a number of factors, including the timing of the evaluation, the large number of programs being implemented, the move to a new school immediately after the project ended, questions about the comparability of the comparison school, and the lack of data. Nonetheless, we were able to draw some conclusions about impact. Although Whitlock has a much lower level of reported disciplinary infractions than Carver, the overall trajectories for each school during the 2000-2001 school year were strikingly similar. Focusing more specifically on disputes, we found that while both schools started off with monthly increases in disputes from August through November, Carver’s increase was much more pronounced than Whitlock’s. By November, Carver was experiencing nearly 12 disputes per day, compared with just over 6 for Whitlock. Yet, after November, Carver experienced a precipitous drop in disputes through January. Whitlock’s disputes also declined, though much less dramatically. Whitlock’s rate of disputes exceeded Carver’s rate slightly from December through May. However, absent Carver’s massive spike in disputes from August to November, and then its precipitous decline, the patterns for the two schools would be quite similar. Therefore, it is difficult to credit any program at Carver with having produced a decrease in disputes. Absent a longer data series, it is difficult to make any inferences from these trajectories. We were unable to gain access to monthly counts of in-school suspensions, out-of-school suspension, and expulsions over a long time period. Our only alternative was to examine yearly totals for each official response. The in-school suspension data show that Carver experienced about a 17% decrease in in-school suspensions, compared with a 52% increase in Whitlock. In the report, we downplay this finding due to a number of data problems. With regard to out-of- school suspensions, Carver experienced about a 10% increase in out-of-school suspensions, compared with about an 11% decrease in Whitlock. Since expulsions are frowned upon in the District, we were not able to discern any meaningful patterns from expulsion data. Police and court data were not useful for assessing impact because we could only obtain them from one school. Unlike the quantitative evidence presented above, “there was nearly universal agreement among those interviewed that this project had shown beneficial effects at Carver Junior High in several ways” (Social Work Solutions, 2001). Overall, the project team was successful in implementing all of the project’s elements. Not all of these elements appear to have played as central a role in the project as others. The School Resource Officer received virtually unanimous acclaim from students, teachers, school administrators, and peers in the Public Safety Department and we see a lot of benefits to continuing his role in the school. The anti-violence presentations by officers in the schools also appear to be targeting the appropriate audience and we see no reason why they should not continue. The peer surveys might have produced some information for the Guidance Counselor and the SRO, but we were unable to discern any specific benefits that would be likely to have a wholesale effect on disputes. If the peer surveys continue, we recommend investing energy in collecting and analyzing their findings more systematically, consistent with the precepts of problem-oriented policing. The after-school program attracted fewer students than it originally targeted. Furthermore, the students it did attract do not appear to have been responsible for participating in large numbers of disputes. Should it continue, we recommend increasing the number of students and making a careful effort to enroll those students participating in the most disputes. At the same time, we take to heart one comment from a student in the program who suggested that there are benefits to having a mixed group of participants, some of whom are more serious offenders than others. In addition to these substantive recommendations, we also have some recommendations for conducting periodic self-assessments. The key to becoming a learning organization capable of measuring, diagnosing, and transforming itself is to periodically examine evidence about the organization’s effectiveness. The School District’s disciplinary data have the potential to serve as a wonderful tracking device for detecting patterns in school-based disciplinary infractions and offending. Exhibits 8 and 9 illustrate some of the simple trend analyses that can be conducted regularly, with little expense, to determine whether offending patterns are changing. This kind of ongoing analysis will serve to enhance an already fruitful partnership between Spartanburg’s School District 7 and the Department of Public Safety. I. INTRODUCTION In 1999, the City of Spartanburg, South Carolina received a School-Based Partnership grant for $137,989 from the COPS Office. The grant enabled the Spartanburg Public Safety Department to implement problem-solving strategies in George Washington Carver Junior High School. Although Spartanburg as a whole suffers from poverty-related problems, the area served by Carver represents an even more concentrated level of economic disadvantage. The school represents an ideal area in which to focus community energies such as the problem-solving approach used in the School Based Partnership project. The focus of the problem solving efforts was disputes, including those between students, between ethnic groups, and between students and teachers. This report highlights the findings from a process and impact evaluation of the School Based Partnership project. A variety of qualitative and quantitative methods were used to conduct an intensive case study detailing the implementation of school-based problem-solving efforts, together with a multi-part quasi-experiment designed to determine whether the project resulted in a reduction in the number or severity of disputes. The report concludes with a series of practical recommendations for researchers, police departments, schools, and federal agencies providing grant programs for school safety. The City of Spartanburg Spartanburg, located in the northwestern corner of South Carolina, is the seventh largest city in the state with just under 40,000 residents. It is located in Spartanburg County, with approximately 250,000 residents, and within the larger Greenville-Spartanburg-Anderson, SC Metropolitan Statistical Area (MSA), which is home to nearly a million residents. Exhibit 1, containing data from the 1990 and 2000 Censuses, illustrates some of the demographic, social and economic characteristics of the City of Spartanburg in comparison with South Carolina and the United States. –SEE EXHIBIT 1– As Exhibit 1 demonstrates, compared with South Carolina and the United States as a whole, the City of Spartanburg has a substantially larger African-American population, a smaller white population, lower household and family income, a larger number of individuals and families living below poverty, fewer home owners and more home renters. Compared with the other four sites included in the national evaluation, the differences are even more pronounced. The proportion of African-Americans is at least five times greater in Spartanburg than in the other sites. The proportion of single mother households is much higher, median household and family income is substantially lower, the poverty and unemployment rates are much higher, and the proportion of people with high school diplomas or bachelor’s degrees is much lower. From 1990 to 2000, the City experienced a decrease in percent white and an increase in poverty and unemployment. Median household and family income rose, but at rates well below the increases for South Carolina and the United States as a whole. The proportion of owner occupied housing units increased, as did the proportion of people who graduated from high school or earned a bachelor’s degree. Although the population of Spartanburg County has been increasing steadily throughout the twentieth century, the population of the City of Spartanburg is decreasing. It has dropped steadily from more than 44,500 in 1970, to less than 40,000 in 2000. Exhibit 2 lists official crime rates in the year 2000 for the City of Spartanburg, the four other national evaluation sites, South Carolina, and the United States as a whole. Spartanburg has higher property, violent, and total crime rates than all of these other comparison groups. Its homicide rate is about eight times the average for South Carolina, and more than twelve times the average for the United States as a whole. Despite the sheer volume of crime that the Public Safety Department faces, its clearance rate is higher than the average rates for South Carolina and the United States. –SEE EXHIBIT 2– The Spartanburg Department of Public Safety The Spartanburg Public Safety Department employs 257 sworn and civilian employees. It has three primary divisions: a Fire Services Division with 69 firefighters, a Police Services Division with 146 Public Safety Officers, and a Technical and Support Services Division. As a public safety department, all police officers are cross-trained in firefighting; patrol officers carry firefighting equipment in the trunks of their vehicles. The Fire Services Division has primary responsibility for responding to fires, while the Public Safety Officers provide additional resources to the firefighters as needed. Exhibit 4 lists some basic characteristics of the Spartanburg Public Safety Department. –SEE EXHIBIT 3– Within the Police Services Division are three secondary divisions: A Field Operations division headed by a Patrol Captain, a Criminal Investigations division headed by a Captain of Investigations, and a Traffic and Special Events division headed by a Lieutenant. All three of these subdivision heads report to the Colonel in charge of the Police Services division, who in turn reports to the Director of Public Safety. Within the Field Operations division is the patrol team, a housing complex team, and a “downtown unit.” Among the sworn personnel assigned to patrol are 19 community-oriented policing officers, each one responsible for community policing activities within a geographic area in the city. The Department’s two school resource officers, as well as the Crime Prevention Unit, are housed within the Criminal Investigation division. The Public Safety Department has embraced an ongoing comprehensive effort to adopt community policing and problem solving throughout the department. Their community policing plans are included as an explicit component of their strategic plan, which is aptly named: Policing in Partnership...Preservation Through Prevention. Community policing is also woven into the very structure of the police department, with the community divided into 19 geographic areas based on census tracts, each one assigned to a community-oriented policing officer. There is also a formal “Community Organization/Problem Solving Award” that is designed to be awarded to an officer who distinguishes himself or herself in one or more of the following areas: solving problems, reducing crime, organizing neighborhoods, or resolving neighborhood issues. Furthermore, the Department has embraced the concepts of Crime Prevention Through Environmental Design. The Department routinely reviews new building plans prior to approval by the City to provide input on crime prevention before construction begins. These are just a handful of the community policing activities in which the Department engages; space does not permit a more thorough review of the Department’s community policing activities. The Department has also taken the lead on the racial profiling issue by becoming one of the first agencies in the nation to develop a bias-based policing policy. The policy encourages officers to use the following strategies during pedestrian and vehicle stops: - “Be courteous, polite, and professional.” - “Introduce yourself and explain to the citizen the reason for the contact as soon as practical, unless providing this information compromises safety. (During vehicle contacts, provide this information before asking for license and registrations).” - “Ensure that the length of the detention is no longer than necessary to take appropriate action for the known or suspected offense.” - “Answer any questions the citizen may have.” - “Provide citizens with court dates, including explaining options for criminal or traffic summons.” - “If your name and badge number is requested, provide it in writing or by furnishing them a business card.” - “Apologize and/or explain if it is determined that the reasonable suspicions were unfounded (e.g., after a BOLO contact).” As of October, 2000, the department employed 25% blacks and 73% whites. Nationally, blacks constitute only 11.7% of sworn officers in local police departments, thus the department is racially diverse in comparison with the national average. At the same time, approximately half the population of the City of Spartanburg is black, so the proportion of blacks in the department is substantially lower than in the community. Females constitute 28% of sworn employees and 50.8% of total employees. Nationally, females constitute only 10.6% of sworn officers in local police departments, therefore the department is diverse in this respect as well. The Department is led by an overall director, A. Tony Fisher. Chief Fisher is well-read and well- spoken. He is able to articulate very clearly his mission, vision, and values for the organization, and he is well connected with the larger police policy community. Chief Fisher began his police career in 1970 as an officer in the Montgomery County (MD) Police Department. While there, he served as chairman of the 20,000 member National Black Police Association. He was appointed at 33 years of age as Chief of Police in Takoma Park, Maryland, the first African- American to be appointed to that post. He served there from 1983 to 1993. While in Takoma Park, his work with community-oriented and problem-oriented policing was featured in a front page article in the Wall Street Journal. He left Takoma Park to serve as associate director of the National Organization of Black Law Enforcement Executives, where he worked for 18 months. He then became the first African-American Director of Public Safety of Spartanburg in August, 1995, an office he has now held for more than seven years. Exhibit 4 lists the mission, vision, and values for the Spartanburg Public Safety Department as outlined five years ago by Chief Fisher. –SEE EXHIBIT 4– Joyce Lipscomb, Operations Analyst for the Department of Public Safety, and our primary point-of-contact within the agency, appears to serve as the Department’s jack-of-all-trades in all matters related to grant funding, innovation, research, and evaluation. In 1999, she received the Director’s Award of Excellence for her work in helping the Department achieve national accreditation. The Public Safety Department received national accreditation from the Commission on Accreditation of Law Enforcement Agencies in March of 1999. The South Carolina Police Accreditation Coalition then accredited it at the state level in April of 2000. The Department is active in seeking out sources of external funding from a variety of federal, state, and private organizations. The Spartanburg School System Altogether, there are 88 school districts in South Carolina, all mandated by state law (the 1990 School Crime Report Act) to report data on school crime quarterly to the State Department of Education. Spartanburg County is home to seven school districts. Two of the County’s seven school districts are located in the City of Spartanburg: District 6 and District 7. The focal school for this project (Carver Junior High) is located in District 7. According to data from the 1999- 2000 school year, 38.4% of applicable school districts in South Carolina had higher school crime rates than Spartanburg District 7. The District averaged 2.04 school crime incidents per 100 students, compared with a statewide average of 1.70. The percentage of school districts with higher crime rates decreased to 33.7% in 2001, and then increased to 40% in 2002. In both 2001 and 2002, there were 2.14 school crime incidents per 100 students, compared with a state average of 1.92 (in 2001) and 1.86 (in 2002). During interviews conducted early in the evaluation process, various members of the Public Safety Department described their relationship with District 6 schools as troubled. They reported having good relationships with District 7 schools. While both districts were described as having been “in denial” about crime, gang, and drug issues, District 7 began to change before the School Based Partnership project period. It is in that district that the police department focused its efforts with the School-Based Partnerships grant. Within District 7 are three junior high schools: two are located in areas with low socioeconomic status, while the other is in a middle to upper middle class area. The Department chose one of the schools serving a disadvantaged area, Carver Junior High, as the place to focus its efforts in the School-Based Partnerships grant. Carver Junior High School was Spartanburg’s black high school before integration. In the 2000- 2001 school year, when the grant was implemented, it had 625 students and 46 teachers in 7th, 8th, and 9th grades. It is located in an impoverished and predominantly African-American neighborhood. More than 60% of students are on free or reduced price lunches. During the evaluation period, the Principal was Dr. Robin Johnson. Dr. Johnson described the mission of the school as follows: “Carver Junior High School is committed to the intellectual, emotional, and physical development of each student to the fullest extent possible in preparation for pursuit of a productive, fulfilling life by providing a challenging and diverse education in a safe learning environment.” Charles Redmond replaced Dr. Johnson as Principal prior to the 2002-2003 school year. During the school years ending in 2001 and 2002, the school received an absolute grade of “Below Average” and an improvement grade of “Unsatisfactory” on the State Department of Education’s Annual School Report Card. A rating of unsatisfactory means that “school performance fails to meet the standards for progress toward the 2010 South Carolina Performance goal.” A rating of below average means that the “school is in jeopardy of not meeting” the 2010 performance goal. Carver Junior High moved into a new building during the evaluation period, in August of 2001. The old building was in poor condition; the new building was designed with the concepts of crime prevention through environmental design in mind. This move had an important influence on our evaluation strategy because we were unable to rely on a data series that extended beyond the program implementation period. Examining a longitudinal data series that extended from the old school building to the new, both before and after the implementation of strategies to reduce the number of disputes, would mean possibly confounding the effects of the dispute reduction strategies and the new building. Quite simply, with the study period truncated in this way, it became even more important to rely on a control or comparison group. In this case, there is another school in the District, Whitlock Junior High, which serves students from similar socioeconomic backgrounds. It has a school resource officer from the Spartanburg County Sheriff’s Office. Because it is located in District 7, and because it is similar in many ways to Carver, we chose to use Whitlock as a comparison school. We will provide more detail about this decision, and about Whitlock, later in this report. The School-Based Partnerships Grant In 1999, the City of Spartanburg received a School Based Partnerships grant for $137,989 to implement problem-solving strategies in collaboration with Carver Junior High School. The focus of the problem solving efforts was disputes, including those between students, between ethnic groups, and between students and teachers. According to the grant application, Carver Junior High School has: “a large number of youth being referred to in-school and out-of-school suspension due to disputes that are occurring between students, and disputes that are occurring between students and teachers. The disputes are both verbal and physical. The result is that a number of youth are not in the regular classroom, which puts them behind in their school work. This effects [sic] the grades of those youth, but also effects [sic] the fear level at the school, and effects [sic] the parents when the youth are out of school.” Based on these concerns, the Public Safety Department, in concert with Spartanburg County School District 7 and Carver Junior High School, proposed a series of efforts to curb the problem with disputes. As we will describe in more detail later, the School-Based Partnerships grant was only one of many programs being implemented in Carver Junior High School at the time. According to the local evaluators, the grant was “used as a catalyst to integrate existing activities and ‘fill-in’ where necessary” (Social Work Solutions, 2001, p. 1). Because so much programming was occurring simultaneously, it was often difficult to isolate which activities were covered by the grant and which weren’t. To the best of our knowledge, the grant funded the following programs and activities: - Obtain external training and consultation in Crime Prevention through Environmental Design (CPTED) and Problem-Oriented Policing. - Conduct peer surveys of children assigned to in-school suspension. The surveys were conducted by other students and were designed to elicit information about why students were misbehaving and what could possibly be done about it. - Implement a Program Improvement Plan - this was an effort to reduce the number of students being given repeat in-school suspensions. It involved the development of an after-school program for students with two or more in-school suspensions, a point system with rewards for good behavior, and presentations by officers. We will describe in more detail later the exact elements of this program. - Assign an officer for 3-4 hours per day to work with students assigned to in-school or out-of-school suspension. Five officers appeared to play a central role in the grant funded activities, teaching violence prevention in classes and in the in-school suspension program. In addition to these programs, a number of others were also implemented in the school during the project period. Among the most important of these were: - A new School Resource Officer (SRO), funded through a Cops In Schools grant started in September 2000, replacing a Deputy from the Spartanburg County Sheriff’s Office who held the position earlier. Although the SRO was not formally part of the School Based Partnerships project, it is difficult to separate his influence from that of the grant. He came into the school at the same time the grant was being implemented, and he was personally responsible for carrying out much of its programming. As we will describe later, he received virtually unanimous acclaim from people we interviewed, including students, school staff and administration, school district authorities, fellow officers, and police administrators. - The Back in Control Program, which was originally included in the School Based Partnerships proposal but cut out by the COPS Office, was implemented using an alternative funding source. This program provides parent training to students with two or more in-school suspensions. Although COPS eliminated this program from the grant, the Public Safety Department clearly viewed it as an integral part of their response to the disputes problem. Once again, although this program was not formally part of the grant, it is difficult to isolate it from the grant activities for purpose of the evaluation. In addition to these programs, numerous others were either being experimented with or were ongoing during the project period. To outsiders like us, Spartanburg appeared to be a cauldron of creative energy, with a number of caring and compassionate people throughout the community trying to find solutions to the social ills affecting children, including poverty, despair, illiteracy, teen pregnancy, drug use, and delinquency. At the same time, as evaluators, we found the overwhelming number of programs being carried out simultaneously to be a challenging research problem which made it difficult to isolate the effect of the grant funded activities alone. According to the grant application, grant funds were used to pay for: - a Crime Analyst / Project Coordinator at the Public Safety Department; - overtime for officers to work in the school; - conference travel; - a GIS workstation with software, three laptop computers, and an LCD projector; - a color laser plotter and a color Laser Jet printer; - training and consultation in CPTED and problem-solving; - a local evaluation of the project; - Web site development and maintenance; - a variety of other miscellaneous supplies. Clearly, some of these line items were more integrally related to the core substance of the grant than others. The grant was accepted in September 1999 and originally scheduled to expire in July 2000. The Department received an extension until October 2001. The project was implemented at Carver from August 2000 to May 2001. The local evaluation was completed upon the grant’s expiration in October 2001. II. THE EVALUATION PROCESS This report describes a social scientific evaluation of the School-Based Partnerships program in Carver Junior High School. The evaluation has two major components. The first is a process or implementation evaluation designed to evaluate the extent to which the program was actually implemented. Research has shown that the process of implementing a program is least as important as the outcomes it produces (Pressman and Wildavsky, 1984; Winn, 1999). Therefore, we use a variety of methods, both quantitative and qualitative, to assess the extent to which the Public Safety Department and the school were able to implement successfully the program they described in their grant application to the COPS Office. The second component is an impact evaluation designed to assess the extent to which the program successfully reduced disputes in Carver Junior High School. From a technical perspective, this question is very difficult to answer for several reasons. We will describe these constraints shortly, together with the evaluation design we chose to determine whether the School Based Partnership project in Spartanburg produced any tangible or measurable benefits. This section of the report is divided into four subsections. We begin by examining the research questions of interest in the process or implementation evaluation. Next we discuss the principal research questions in the impact evaluation. We then describe the various data sources and research methods used to answer these questions. We conclude by discussing a variety of analytical and logistical issues inherent in the evaluation of school-based problem-solving initiatives. A. Research Questions: Implementation Process According to reformers, problem-oriented policing is a philosophy that must be tailored to the special needs and unique circumstances of the agency and the community (Goldstein, 1990). It is not a “program” that can be carelessly tacked onto the police agency as an appendage. Because the focus of POP is on designing customized solutions to unique local problems, there is some reluctance among reformers to endorse a “cookbook” approach to solving problems. On the other hand, for inexperienced agencies struggling to understand and implement POP, the availability of a concrete strategy can sometimes be useful for getting started. According to Goldstein (1990:66): “The principal danger in providing a detailed procedure is that it will be used to oversimplify the concept: that more effort will be invested in moving mechanically through the recommended steps than in the explorations and thinking that the steps are encouraged to stimulate.” The most well-known method for implementing POP is known as the SARA model, which stands for Scanning, Analysis, Response, and Assessment. The SARA model was first developed in Newport News, Virginia, by John Eck and William Spelman in 1987. A decade later, when the U.S. Department of Justice developed a new grant program to diffuse problem- oriented policing throughout the nation, all applicants were required to use the SARA model. Police officers throughout the United States can now be heard discussing the SARA approach to problem-solving. While it is a useful tool for many agencies, it is important to remember that it is only one method for implementing problem-oriented policing. We now introduce the four elements of the SARA process. Scanning involves a variety of methods for identifying the problems to be solved. Police and schools can do surveys, check calls-for-service records, or request information about problems from other agencies. Analysis means collecting and analyzing data about the problems selected during the scanning stage. During this stage, officers and others must open their minds to the various kinds of data that might be useful in solving a problem. Armed with useful data and insightful analysis, officers, students, teachers, and others involved in the project are often able to uncover information that is useful for developing creative response strategies. In the Response stage, officers and other problem-solvers use the information collected during the analysis stage in an attempt to solve (or reduce the scope of) the problem. In some instances, the response can constitute a single action (such as an arrest) by a single agency. In other cases, the response may be a complicated, multifaceted strategy that relies on the cooperation of police, school officials, parents, and students. In the Assessment stage, the agencies systematically examine the efficacy of their response strategy. In some cases, the assessment may be very simple. For instance, if the source of the problem is a single offender, and the offender has moved a far distance away or has been transferred to another school (or has been incarcerated), then the problem may have been effectively solved. In most POP cases, the assessment should be more involved and more systematic. The basic question to be answered during this stage is whether the response was effective. Good assessments move beyond this simplistic approach, examining the duration of effectiveness. For instance, if police attempt to alleviate a problem with speeding motorists by issuing citations over a short period, they should follow this up by determining how long it takes before the effects of the increased enforcement begin to diminish. Another important issue during assessment is displacement. A response cannot be judged completely effective for reducing crime in one area if it increases crime in another area. This stage of the SARA process is systematically ignored by many agencies professing to practice POP. Taking this stage seriously is one of the major challenges to implementing POP. Since implementing the SARA process is a fundamental component of all School-Based Partnerships grants, one important aspect of the evaluation is determining the extent to which it was successfully implemented in Carver Junior High School. Thus, the process or implementation evaluation will examine in detail each of these steps. The evidence used in this portion of the evaluation is entirely qualitative. It is drawn from field notes collected by members of the evaluation team during interviews and observations conducted within the school and the public safety department. Some of the research questions asked during this portion of the evaluation include the following: Scanning Phase - How were the problems identified? - What process was followed, and by whom? - What problems were decided upon? Analysis Phase - What kinds of analysis were done to examine the extent of the problem? - Who was involved? What types of data were used? - What were the sources of the information? - Did the analysis phase examine offenders, victims, and locations? - How long was this phase (months, days)? - What problems and obstacles did the project team encounter during this phase and how were they solved? Response Phase - What were the primary responses that were implemented? - What was the dosage or coverage of these responses? - How was the response implemented? - What resources did the implementation team draw upon to conduct the response phase? - How long was this phase (months, days)? - What problems and obstacles were encountered by the team during this phase and what solutions were decided on to deal with these problems? Assessment Phase - How was the assessment of the response conducted? - Who was involved? - What types of data were used in this phase? - What is the working relationship of the partners? B. Research Questions: Impact While it is important to know if programs or policies have been successfully implemented, all programs and policies are a means to one or more ends. Therefore, another fundamental element of a thorough evaluation is determining the impact of a program or policy on a set of outcomes. Impact evaluations are always normative in the sense that they require the evaluator to state a set of measurable outcomes and then determine whether the program or policy had any impact on these measures. It is important to recognize that the selection of outcome measures is an important step in this process. It would be simple to set the bar low so as to ensure a successful impact. Evaluators often do the opposite, setting the threshold for “success” too high, or defining it too narrowly, leading to the now-clichéd conclusion that “nothing works.” The solution is to adopt a middle-level stance, selecting a variety of potential outcome measures, neither guaranteeing that the evaluation will find a successful impact, nor setting the program or policy up for failure by setting the threshold for success too high. In this project, the primary yardstick for determining success is the incidence, prevalence, and perception of disputes. At the same time however, the project team also set at least two other ancillary goals. First was improving the relationship between police and the school. Second was improving overall feelings of safety within the school. Therefore, in our efforts to determine whether the School Based Partnerships program had an effect, we will consider these three dimensions: 1. The incidence and prevalence of disputes; 2. The relationships between the public safety department and the school, including students, teachers, staff, and administration; 3. The overall perceptions of safety within the school. Conducting an impact evaluation on these questions will require the use of numerous data sources. These are described in detail below. C. Data and Methods A good program evaluation chooses a set of appropriate research designs and then, within this framework, collects and analyzes quality data to learn more about the implementation and impact of the program. In this section, we outline the specific steps we followed to design the evaluation of the School Based Partnerships program in Spartanburg, and collect and gather the appropriate data. The evaluation team considered a number of evaluation design options. Unfortunately, many of the more rigorous designs for evaluating programs, such as randomized experiments, were not available to us for a number of reasons. Principal among these reasons was the timing of the evaluation. The project team had already begun its work in Spartanburg in September 2000, four months before we arrived on site for the first time in late January 2001. Therefore it was difficult to collect “true” baseline data. Second, the School-Based Partnerships program was intended to be what evaluation researchers call a “full coverage” program; it was designed to provide benefits to all students in the school. It is much easier to evaluate the success of partial coverage programs that are only designed to affect a subset of students, since other subsets can serve as an effective comparison or control group. Although we were unable to select a control group within Carver Junior High, Spartanburg County School District 7 has two other junior high schools that could serve as potential comparison schools. Whitlock Junior High School, which is located outside of the city, but serves students from several Section 8 housing developments within the city, is the most comparable to Carver. McCracken Junior High, also located within the city, serves a student body with a much higher proportion of whites and a higher socioeconomic status in an area that is largely middle-class. Exhibit 5 lists some characteristics of the three junior high schools. As a quick perusal of Exhibit 5 suggests, Carver and Whitlock are comparable in a number of ways, while McCracken appears to vary significantly from the others. For that reason, we chose Whitlock as the comparison school for purposes of this evaluation. –SEE EXHIBIT 5– While Carver and Whitlock are similar, they are also different in some important ways. First, different police departments serve the two schools. Carver is located within city limits, and is therefore policed by the Spartanburg Department of Public Safety. Whitlock is located in Spartanburg County, outside of city limits, and is therefore served by the County Sheriff’s office. Second, many students can walk to Carver from their neighborhoods; all students are driven or bused to Whitlock. As we have found in several other sites during the national evaluation, the commuting patterns by which students make their way to and from school have a significant impact on problems with crime and fear. Finally, though we were told early in the evaluation by the Sheriff’s Department that we would be able to obtain school arrest data from the County (for Whitlock), in the end we were only able to obtain such data for Carver. Thus, although we were able to compare the schools on a number of dimensions, having arrest data for only one school was difficult to overcome. The national evaluation of School Based Partnerships examined a number of local grant recipients, like the Spartanburg Department of Public Safety, that had already hired a local evaluator to examine their grant implementation efforts. In this case, the local evaluators were George Appenzeller and Sara Meadows, proprietors of Social Work Solutions, a Columbia, SC based firm focusing on research, training, and evaluation related mostly to social work, alcohol and drug abuse, and juvenile issues. Mr. Appenzeller spent some time as an intelligence officer for both the Army and the Atlanta Police Department, and both have had various experiences in working with justice agencies. Neither, however, claimed this as their specialty area. They described their evaluation experience as relying mostly on qualitative methods, and they recently completed an evaluation of the Department’s problem-solving efforts related to burglary. As of our arrival in January, 2001, there was $8,000 budgeted for the local evaluation. One important component of conducting a national evaluation in a site that already has a local evaluation underway is ensuring that the two sets of evaluators do not duplicate each other’s efforts. We were fortunate to work with Social Work Solutions for a number of reasons. First, they timed the local evaluation to begin at the same time as the national evaluation. To make things easier on our research subjects, we agreed to conduct all of our initial interviews in the Public Safety Department and the school district together. Second, since their evaluation experience focused mostly on qualitative research methods, we agreed to divide up the research tasks with Social Work Solutions conducting a number of interviews and focus groups, and 21st Century Solutions focusing on collecting various kinds of quantitative data. Since both firms needed to prepare and submit a comprehensive final report, the division of labor was not surgical: both firms collected and examined at least some qualitative and quantitative data. Social Work Solutions agreed to take on some additional evaluation duties that emerged as a result of adding a control school to the project. The national evaluation team provided some additional funds to pay for these extra duties. Together, the two firms signed a memorandum of understanding outlining the research tasks to be conducted by each. Social Work Solutions completed and submitted its final report in October, 2001, which was consistent with the schedule set by the COPS Office for completing local evaluations. 21st Century Solutions, as the national evaluators, needed to allow another school year to elapse before completing its evaluation. Therefore our final report contains one extra year of follow-up data. During the course of our research, the evaluation team relied on a variety of data sources, both quantitative and qualitative, for examining the School Based Partnership project in Spartanburg. We conducted two student surveys: one at Carver and one at Whitlock. We analyzed aggregate disciplinary data from both schools, including in-school suspensions, out-of-school suspensions, and expulsions, from 1994 to 2002. We also examined and coded detailed data on all disciplinary incidents at Carver and Whitlock during the 2000-2001 school year. We analyzed official arrest data and examined follow up data from the courts. Together with Social Work Solutions, the local evaluator selected by the Public Safety Department and our research partner on this project, we conducted dozens of interviews and focus groups in the two schools, in the school district headquarters, and in the police department. We “shadowed” the school resource officers in both schools, attended Neighborhood Watch meetings, and accompanied a number of patrol officers on ride-alongs throughout the City of Spartanburg. One of the founding principles of the evaluation was that a true understanding of program implementation and impact can only be achieved by triangulating from a variety of quantitative and qualitative data sources. In the following section, we describe each of those data sources in more detail. 1. Surveys As part of our effort to learn more about life within Carver Junior High School, we conducted student surveys in Carver and Whitlock on May 8th and 9th, 2001. The surveys were voluntary, with potential respondents instructed that no harm would come to them for failing or refusing to participate. Nonetheless, response rates were high. Given research demonstrating that seeking active consent from parents can lower response rates and contribute to sample selection bias, we chose to use a passive consent procedure in which parents could refuse to allow their child’s participation, but absent such a refusal from either parents or their children, the survey would be administered without actively securing parental approval (Andeman, et al., 1995; Esbensen, et al., 1996). The survey was 7 pages long and contained 35 questions, nearly all of which were closed-ended questions requiring the student to select from a set of responses. Many of the questions were drawn from the Youth Risk Behavior Surveillance System, a standardized instrument used in high schools and middle schools across the nation (Fetro, et al., 2001; Kann, et al., 1998). The survey instrument is contained in Appendix 1. In both schools, students were instructed that their answers would be anonymous and confidential. Each student placed his or her completed survey into an envelope, sealed it, handed it to a member of the evaluation team, and returned to their regularly scheduled class. Surveys were administered to a random sample of students at Carver Junior High School on May 8, 2001. The research team selected survey participants using a systematic random sampling method in which every third name was chosen from a list of enrolled students at each grade level. The resulting samples ranged from 64 - 68 students per grade level. Due to absences, assignments to in-school or out-of-school suspension, participation in external programs, or a number of other factors, the number of survey respondents is smaller than the samples selected for each grade level. Exhibit 6 lists the enrollment, sample size, and number of respondents for each grade level at Carver and Whitlock. The survey was administered in the cafeteria before and after lunch. Using the overhead speaker system, the principal called students by name and grade level from their respective classes at the beginning of the class period to participate in the survey. Each grade was surveyed separately during a different class period. Only one student actively chose not to participate. Others might have passively refused by not showing up; we were unable to check whether this occurred. –SEE EXHIBIT 6– On May 9, 2001 the student surveys were administered at James Whitlock Junior High School. The students selected to take the survey were again chosen using the same systematic random sampling technique in which every third name on the list of enrolled students was chosen to participate. The resulting samples ranged from 57 - 71 students per grade level. Again, due to a variety of attendance issues, the number of students participating in the survey was smaller than the selected sample sizes. As shown in Exhibit 6, the number of survey respondents at each grade level ranged from 44 - 54. Two classrooms were chosen for students to use during the administration of the survey. The Principal called the students by grade level at the beginning of the class period to report to the classrooms and participate in the survey. Two members of the survey team supervised one group of students and another member supervised the second group. Several problems occurred at Carver while the survey was being administered. While the eighth graders were taking the survey, another student wandered into the cafeteria. Some of the students taking the survey began to whisper about fighting with or attacking this student after the survey was completed. A member of the survey team had to retrieve the Principal to ensure that the fight did not happen. Some students told members of the research team that they increased their age level on the survey by one year if their birthday was coming up within the next few months. Several students tried to answer by committee, discussing their answers together before writing them down. The research team had to warn or separate several such students. At Whitlock, as in Carver, several of the eighth graders needed to be separated while the surveys were being administered to avoid behavior problems and prevent students from collaborating about their responses. Several of the students refused to place their surveys in the envelopes provided by the research team to assure anonymity of responses. In addition, several students failed to return to their classroom after leaving the survey room and chose to wander the halls. An announcement was made by the Principal and the Vice Principal directing them to their classes. During the survey, students asked a number of questions. Some of the common questions included: - Why are we asking about “carrying a weapon”? - What do we mean by “weapon”? - What is an “SRO”? Students wondered whether various kinds of implements should be considered as a weapon. One student, for instance, asked whether a razor is considered a weapon. We chose a liberal definition of the term, to include anything that the survey respondent viewed as a weapon. The survey also asked a question about the “School Resource Officer (SRO).” Subsequent questions then used only the abbreviation “SRO.” Students found this confusing and wondered what an SRO was. When the surveys were administered, students at both Carver and Whitlock had just finished taking the PACT test the previous week. The schools were scheduled to close for the summer in about two weeks. For those at Carver, it was the last year in the old school building, and many of the teachers, staff, and administrators were packing their belongings into boxes to prepare for the move. We note these issues to acknowledge that it was a stressful period for the students and employees of both schools. We are grateful for the overwhelming level of assistance, access, and support that both schools provided during the survey process. 2. Aggregate Disciplinary Data We obtained aggregate yearly disciplinary data from Spartanburg County School District 7. These included the total number of in-school suspensions, out-of-school suspensions, and expulsions in all three junior high schools for the years 1994-2002. 3. Incident-Level Disciplinary Data We obtained a data file containing incident level data on every recorded disciplinary incident that occurred in Carver and Whitlock during the 2000-2001 school year. This file included 5,218 incidents at Carver and 3,301 incidents at Whitlock. For each incident, a variety of information was available, including the date, a description of the offense, and a description of the school’s response. We used a variety of coding procedures to perform a systematic analysis of these incident level data. 4. Official Arrest / Court Data We originally hoped to obtain data on arrests at Carver and Whitlock for the 2000-2001 school year. These data would have been used to compare arrest patterns in both schools. We obtained arrest data on Carver from the Public Safety Department, but we were unable to obtain arrest data on Whitlock from the County Sheriff. In addition to the arrest data, we also obtained data from the Department of Juvenile Justice tracking the outcomes of arrests and summonses issued at Carver. We obtained these additional court data in July 2001. Some of the cases were still pending at that point, so we were not able to track the outcome of every case. Nonetheless, the police and court data together are useful for illustrating patterns in the processing of criminal cases initiated by an arrest or summons at Carver. 5. Interview and Observation Data In addition to the quantitative findings, much of the report is informed by our analysis of transcribed field notes assembled by the research team during interviews, focus groups, and observations. Together with Social Work Solutions, we conducted dozens of interviews in the two schools, in the school district headquarters, in the public safety department, and in the County Sheriff’s office. Our research assistant spent two-and-a-half days “shadowing” the School Resource Officers in both schools, observing their interactions with students, parents, teachers, administrators, and other officers. Her field notes from those 2.5 days alone filled 31 typed, single-spaced pages. We attended Neighborhood Watch meetings, accompanied a number of officers on ride-alongs throughout the City of Spartanburg, and toured nearly the entire city, including most of the housing projects. In addition, Social Work solutions conducted group interviews with Carver students in the peer counseling program and after school programs; focus groups with Carver students who took part in the after-school Program Improvement Plan and students who did not take part in the program; focus groups of Whitlock students who have and have not received in-school or out-of-school suspension; and focus groups with parents from both Carver and Whitlock. Together, data and findings from all of these sources of qualitative data will be used in concert with the quantitative data described above to evaluate the implementation and impact of the School Based Partnerships program. D. Evaluation Issues Our efforts to conduct a rigorous evaluation of the implementation and impact of the School Based Partnerships program in Spartanburg were hampered by a number of factors beyond our control. First, the timing of the evaluation relative to the implementation of the program limited the kinds of data we could collect and research designs from which we could choose. Any evaluation that begins several months after a program has already been in place is starting off on the wrong foot. Second, it was difficult to separate out the effects of the multiple interventions taking place in Carver Junior High School, only some of which were paid for through the School Based Partnerships grant. Thus, even if we found evidence that disputes or fear decreased, it would be nearly impossible to determine with any specificity which programs or policies were responsible for producing the decrease. Third, even though were fortunate to have a good comparison school in this case, data from both schools were not always available or strictly comparable. These three challenges –timing, multiple interventions, and the selection of an appropriate comparison group– represented significant hurdles in carrying out the evaluation and set limits on the strength of our findings. We now consider each one in greater detail. Timing Due to the timing of the national evaluation grant relative to the School Based Partnerships grant, Spartanburg was well underway in its program implementation efforts before the evaluation team began its work. One step that the COPS Office and other funding providers can implement to strengthen the scientific body of evidence on the impact of crime prevention programs, both in schools and in other venues, is to pay more careful attention to the timing of evaluation research. It is common for substantive programs (such as School-Based Partnerships) to be funded before evaluations of those programs are funded. At a minimum, evaluation grants should be funded concurrently with program grants. An even better solution, although perhaps logistically difficult, would be to fund evaluation grants before program grants. This will ensure that baseline data can be collected before students or citizens are contaminated by having been exposed to part of the treatment (such as the presence of a school resource officer). Furthermore, if evaluators arrive early enough, they can often make suggestions that assist program designers in formulating an implementation plan that is more amenable to a rigorous impact evaluation. The timing of an evaluation relative to the implementation of a program is always crucial. If the program has already been implemented, then the evaluator is limited to a relatively weak body of ex post evaluation designs. If the program has not yet been implemented, then the evaluator has more powerful ex ante design options available, including the most rigorous: the classic experimental design (Rossi and Freeman, 1993). This is a simple and powerful recommendation that funding agencies can implement with relative ease to enhance the scientific knowledge base about what works in school-based crime prevention. Multiple Interventions The grant funded the following programs and activities: training and consultation in Crime Prevention through Environmental Design (CPTED) and Problem-Oriented Policing (POP), peer surveys of children assigned to in-school suspension, an after-school program entitled the Program Improvement Plan designed for selected students with two or more in-school suspensions, and assigning an officer for 3-4 hours per day to work with all students assigned to in-school or out-of-school suspension. In addition to these activities funded under the School Based Partnerships grant, several other programs were implemented as well. These included a new School Resource Officer (SRO) funded through a Cops In Schools grant, and the Back in Control Program, which provides parent training to students with two or more in-school suspensions. In addition to these programs, numerous others were either being experimented with or were ongoing during the project period (Social Work Solutions, 2001). Although the proliferation of programs intended to help Carver’s students illustrates the commitment of the police and schools in Spartanburg, it makes it difficult from an evaluation perspective to isolate the effects of the grant-funded activities alone. Selecting a Comparison Group Several interventions were implemented in Spartanburg simultaneously. Some of these, like the after school program, could be considered “partial coverage” programs because they only focused on a small number of students. However, even those programs that focused on a sample or subgroup of students were meant to have school-wide effects. As noted in the grant application, the School Based Partnerships program was intended to reduce fear among faculty and students throughout the school, not just among individual program participants. Other programs, like the School Resource Officer, were clearly full coverage programs meant to influence every student in the school. Taken together, the various programs implemented under the School-Based Partnerships programs could really only be considered as full coverage programs meant to produce benefits for all students, not just a select group. Having a full coverage program within the school has important implications for selecting an appropriate evaluation design. For example, it is not possible to select an experimental and control group within the school. Such a design is only possible for partial coverage programs in which only a subset of the population (such as students or classes) is expected to receive access to the treatment or the program. This eliminates a number of possibilities for conducting a rigorous experiment or quasi-experiment, and leaves only a handful of other good options for potential evaluation designs. Given the impossibility of selecting an “internal” comparison group from within the target school, we chose the next best option – an external comparison group located outside the target school. As we explained earlier, we selected Whitlock Junior High School as the comparison school. While this was the best option available to us, it is not ideal for several reasons. First, although Carver and Whitlock are similar, there are still numerous differences between them. If we find that Carver has a lower rate of disputes, it will be difficult to attribute those differences to the School Based Partnerships program when they might just as easily be attributed to other differences between the schools. Second, arrest data, which are of vital importance in any evaluation study that focuses on offending levels, are only available from Carver. As stated earlier, we attempted to obtain them from the County Sheriff’s office for Whitlock, but we were unable to do so. Third, although we were able to obtain detailed data on officially recorded disciplinary incidents at Carver and Whitlock, we cannot be certain that differences in reported offending patterns reflect real differences in offending rather than simply differences in recording patterns. For instance, in 2000-2001, Carver recorded 5,218 disciplinary incidents, while Whitlock recorded only 3,301 incidents. Does this mean that Whitlock has 37% fewer disciplinary problems than Carver, or does it mean that the two schools use different criteria for recording such incidents? It is really difficult to know for sure. For all of these reasons, even though using an external comparison group is the best option available to us, it presents numerous challenges. For data that are only available from the target school, such as the arrest data, another potential set of design options are “reflexive control” designs. These compare an organization (or other unit) before and after the implementation of a program or policy. This can be a weak or strong design depending on the number of data points, both overall, and before and after program implementation. Having a large number of time points enables evaluators to use sophisticated “time-series” methods, which can be a good evaluation design. Having only two data points, known as a simple before-after or pre-post design, is a relatively weak design. We approached the evaluation with the intent of compiling a time series database to examine changes in offending levels and arrests over a large number of time periods. Unfortunately, data were not available over a large number of years, so it was not possible to develop a lengthy time series of yearly or semester-length data. Furthermore, even when such data are available, the systematic temporal fluctuations that occur in the normal rhythm of the school year make it very difficult to use a time series design with smaller units of time like months, weeks, or days. Summer vacations, the beginning of school, thanksgiving break, winter break, spring vacation, and the end of school are all periods that affect the volume of offending or misbehavior. It is difficult to control for these kinds of factors in a time series design. We could possibly minimize some of these concerns by creating a time series design using data aggregated at the monthly (rather than daily or weekly) level, but for some measures, this approach leaves us with only 9 months of data starting at the time the program was implemented. After considering the various challenges in selecting a good evaluation design, we were left with design options that are not very rigorous. We use some simple before-after analyses without a control group, some before-after analyses with a control group that is not strictly comparable, some qualitative impressions, and some minimal time series analyses where the data permit. We revisit these decisions, and the issue of program evaluation in schools more generally, in the conclusion to Part IV. III. IMPLEMENTATION “Implementation evaluations attempt to capture the essence of what programs do – the actions that make policy real.” -Winn (1999, p. 1) Implementing new programmatic innovations in complex organizations has proven to be a daunting challenge. This challenge arises in organizations of all types: public and private, profit and nonprofit, manufacturing and service (Pressman and Wildavsky, 1984). Implementation theory focuses on the process of implanting new programs, policies and practices in organizations. The fundamental assumption is that no matter how well an innovation works, it is of little utility if not properly implemented (Winn, 1999). The landscape of administrative practice, both public and private, is littered with promising programs whose champions were not able to implement them successfully. Research and theory on implementation has therefore become an important component of the larger study of innovation in organizations (Repenning, 2001). Understanding how implementation efforts succeed and fail is at least as important as studying the impact of programs and policies. This section discusses the implementation of problem-solving strategies in Carver Junior High School as a result of the School-Based Partnerships grant. Since the grant application asked all grantees to follow the SARA process, we initially sought to describe the implementation according to these four steps: Scanning, Analysis, Response, and Assessment. However, although the implementation process in Spartanburg contained elements of SARA, it was not as linear a process as SARA presupposes. Therefore we chose instead to describe the implementation process without forcing it into the SARA framework. As outlined previously in this report, multiple interventions that might affect the level of disputes were taking place within Carver Junior High School. Some of these were funded by the School- Based Partnerships grant, some were funded by other grants, and some were paid for through the normal budgetary process for Spartanburg County School District 7. Disentangling these different programs was sometimes difficult. For instance, although the SBP grant paid for officers to provide instruction during in-school suspension and the after school Program Improvement Plan, the School Resource Officer, who was not funded through this grant, participated frequently in this programming. Similarly, the Back in Control parent-training program, which was cut from the original SBP grant proposal, was implemented with an alternative funding source. It focused its efforts on the same population of children as the Program Improvement Plan: students with two or more in-school suspensions. We are careful to emphasize that we do not view these “crossover” efforts as inappropriate; we mention them only to highlight the difficulty of performing an evaluation under such circumstances. We begin this section by highlighting four programs or efforts that were most clearly funded by the grant: the CPTED training and consultation, the peer surveys of children attending in-school suspension, the after school Program Improvement Plan, and the assignment of officers to provide anti-violence training during In-School Suspension. When that is completed, we highlight some additional programs or efforts that were also introduced during the 2000-2001 school year. We chose only those that might affect the level of disputes, but which were not funded through the SBP grant. Grant Funded Programs (1) CPTED The Spartanburg Public Safety Department began the project by hiring Tim Crowe, a nationally known expert in Crime Prevention Through Environmental Design, to do some training and consultation on CPTED methods (see Crowe, 1990). This process began with Dr. Crowe providing training on CPTED methods to about 15 public safety officers. During training, the officers used Carver Junior High School as a class project to learn about CPTED and developed a CPTED diagram for the school. He also conducted a detailed CPTED analysis of the Tobias Hartwell housing development located across the street from Carver. His recommendations resulted in some changes in the design of housing units. Finally, four officers conducted a pseudo-experiment on the physical security environment at Carver. Four young-looking plain- clothed officers wearing jeans and t-shirts entered Carver Junior High School without permission, and without any visible signs that they were police officers. No one stopped or questioned these officers for more than an hour. This exercise demonstrated convincingly that there was a physical security problem at Carver. Although the Public Safety Department was not involved in the design of the new Carver building, the architects used CPTED concepts when designing the school. With regard to problem-oriented policing and the SARA model, the various CPTED processes that Spartanburg engaged in contained elements of scanning and analysis. (2) Peer Surveys Four students without disciplinary problems were selected by Carver to conduct peer interviews and surveys of students assigned to In-School Suspension. The purpose of these information- gathering activities was to learn more about why the latter students were getting into trouble. The interviewers earned points toward $5 gift certificates that could be redeemed at the mall. The interviewers also served as informants for the school administrators and the police officers working in the school. With regard to problem-oriented policing (POP) and the SARA model, the peer surveys constituted a form of scanning. The results of these surveys and interviews do not appear to have been analyzed in a systematic fashion and therefore do not constitute a form of analysis, at least from a POP perspective. (3) Program Improvement Plan (or Points for Partners) The Program Improvement Plan, or Points for Partners, was an after-school program focused on students assigned to In-School Suspension (ISS) two or more times. The focus of the program was broad-based: for participating students to work toward improving themselves in a number of important ways. Students were selected for participation for a number of reasons, not just serious disciplinary problems. As we demonstrate shortly, students are assigned to ISS for a variety of offense types or disciplinary infractions. The most frequent type of infraction resulting in being sent to ISS is being tardy. Therefore, students exhibiting poor effort or classroom performance were also eligible for the program, not just those engaging in disputes with teachers or other students. At the same time, many eligible students were not selected because the Guidance Counselor and the School Resource Officer felt they were “too far gone” and would not take seriously the notion of improving themselves. Thus, participants were those who met the following criteria: (1) they had two or more ISS placements, (2) they were selected by the Guidance Counselor and the School Resource Officer, (3) they agreed to participate, (4) their parents agreed that they could participate, and (5) they did not continue to get in trouble once they signed up for the program. The program was originally designed to host 20 kids. Initially, only ten met the criteria listed above. Three dropped out, leaving only seven participants. Other students requested permission to join, but were refused. According to our interview and focus group findings, some of those who elected not to participate, as well as those who dropped out, felt that the Guidance Counselor and SRO were too strict. The remaining participants represented a heterogeneous group. Two of the student in the after school program were selected due to lack of performance in the classroom. One of these was a child with a high IQ that was not achieving to her full potential academically. She was in the gifted and talented program and has never been a disciplinary problem. The others were selected based upon a variety of disciplinary issues. The group met each week on Monday and Thursday afternoons from 3:30 to 5:00 p.m. The program operates on a point system. Kids in the program have a daily log containing teachers' comments about their progress. Turning in the log earns them points. Behaving well earns them points. Each time they earn 50 points, they get a $5 gift certificate from the mall. One student earned $125; the average was about $30. The budget for the grant included $2,300 for these stipends. Kids who dropped out of the program got no money. If a kid was placed on academic probation, he/she lost all money earned. The kids also had to earn their way on field trips. Program participants kept journals, they talked about current events and various other issues, and they routinely read a magazine entitled Career World. Frequently the teacher assigned the students to read an article, write about it and present it to the class. The students learned about becoming leaders, about improving themselves, about the importance of having good self- esteem, and about adopting acceptable study skills. They took field trips to the zoo and museums. Participating students who continued to get assigned to ISS were asked to leave the program. The SRO’s role in the program included teaching lessons and reinforcing the Guidance Counselor’s lessons and instructions for the participants. In addition, he brought videos on crime prevention and the realities of detention in the criminal justice system. One particular video was called “Get Smart People.” It showed what prison life is like. From a problem-oriented policing perspective, this program constituted a major portion of the Department’s “response” phase. (4) Violence Prevention Instruction The grant also paid for officers to come into the school for 3-4 hours per day to work with students assigned to in-school or out-of-school suspension. Five officers appeared to play a central role in the grant funded activities, teaching violence prevention in classes and in the in- school suspension program. They made presentations on subjects such as “violence, gangs, decision-making, staying in school, domestic violence, and so on…” (Social Work Solutions, 2001, p. 4). Other Programs In addition to these three programs or efforts, three others were also ongoing during the implementation of the grant: a new School Resource Officer, the Back in Control Program, and the GREAT Program. There were numerous others as well that we do not describe here. (1) School Resource Officer (SRO) Whitlock Junior High School has had School Resource Officers from the County Sheriff’s Office since 1993. The 2000-2001 school year was the first year that the Spartanburg Public Safety Department assigned one of its officers to serve as an SRO in Carver Junior High School. In the years prior to the implementation of the School-Based Partnerships grant, the County Sheriff's Office gave an SRO to every school in the county in return for them paying the matching funds required under the Uniform Hiring Program grant from the COPS Office. According to several reports, the Deputy Sheriff assigned to Carver stayed in his office frequently and did not spend a lot of time in the classroom or engaging in conversation with the students. Most of the people we interviewed stated that there was a clear philosophical difference between the current and former SRO on the role of school resource officers. Our research assistant shadowed the current SRO at Carver for one-and-a-half days in May 2001. At that time, the SRO was 30 years old and married with an infant child. Prior to his assignment as a School Resource Officer, he had worked in various other assignments at the Department of Public Safety for 5 years. He had “never planned to get into law enforcement,” but he was laid off from an industrial job and applied for a job at the county jail. By the time he left the jail, he had learned to shoot a gun and was a Sergeant. A lot of the people who worked at the jail left to work at the Department of Public Safety. After his brother joined the DPS, he applied and was hired. Instrumental in his decision were rumors that “it is different now that Director Fisher is here.” Although this was his first year working as a School Resource Officer at Carver, he dealt with a lot of Carver’s students in his last assignment as an officer working in the housing complexes. In his current role as SRO, he is assigned to the Crime Prevention unit, which includes DARE, GREAT, SROs, and the Department’s “Safety Scene” cable television show. The SRO’s brother also works within the unit, teaching DARE and GREAT, and hosting the cable television show. When the SRO was a student in the ninth grade, there was a shooting at his school. He knows firsthand the fear that results from violence in schools. Furthermore, in 2000, his cousin was killed in a shooting on a street close to Carver. He said that some of the kids who live on that street go to Carver. The kids knew about the shooting; he came back and talked to them about it. He said, “I am honest and forthright with them and they understand.” He also recalled that when he lived in Michigan, there were only 4 blacks in his school. He first heard the word “nigger” there. The first day he was in Michigan, he got into a fight on the bus. He said he understands how students feel because he has been there. He tries to teach kids how to handle these types of situations without getting violent, relying instead on conflict resolution strategies. The SRO tries to teach kids that they have choices. Others have told the SRO, who is African American, that he is “working for the white man.” His reply is that he is just giving back to his community. He does not want his community to get overrun with drugs and other problems. At Carver, he is the assistant football coach and the head basketball coach. His fiancé (now his wife) attends all the basketball and football games. Most of the students know her. He believes that coaching has helped build his relationship with the kids. When he first started working at Carver, a group of girls attacked a younger and smaller girl. The victim had a black eye; one officer said, “They almost blinded the girl.” Two staff members said this was the most violent incident they had witnessed at Carver. The SRO charged the girls involved. He handcuffed them in front of everybody. In fact, when he arrests students, he always handcuffs them out in the open. He believes that “Advertising is the best tool in this school.” He tries to arrest and handcuff the bad kids right in front of their friends. When other kids see this, they know that he is serious. As we will discuss in more detail later, the SRO made 52 juvenile arrests at Carver during the 2000-2001 school year. Other officers made an additional 21 juvenile arrests at Carver for a total of 73 in one school year. When the SRO first came to Carver, he was excited but hesitant about dealing with the kids. He appears to believe in an approach that tempers cracking down firmly on misbehavior with forming enduring relationships with the kids. He started off with the idea that “you have to go in hard at first.” After that, the kids pass the word around; they know they will get arrested if they do something bad. At the same time, as we demonstrate shortly, he works very hard to establish relationships with them. His description of the balance he tries to achieve with the kids at Carver is reminiscent of an old metaphor that likens policing to forging an iron fist within a velvet glove (Platt and Takagi, 1982). When asked what his philosophy is, he replied, “A School Resource Officer is actually what it is: a resource for kids, a resource for students, and a resource for teachers. If a kid has a problem, I help. I am a resource for the kids.” He said that he never tries to “leave a kid or parent hanging. I try to find the answer. I am a resource. If you develop a relationship with these kids, the kids will tell you what you want to know.” He tries to give the kids encouragement; he tells them when he is proud of them; he enjoys “seeing the kids develop and change.” Many of the kids are from single parent homes. He stresses education a lot with these kids; he tells them they need their education because they can always fall back on it. He asks some kids what they want to do when they are out of school and some will say they want to play professional sports. He asks them what they will do if they get injured. He tries to teach them that if they get an education, they will always have skills to fall back on. He stresses education so much that the kids even show him their report cards. One girl stopped him in the hallway during our observation period to show him her GPA. He was proud of her and she was happy. He said, “You have to let kids develop and grow. You don’t want to stifle them too much.” Some of the following snippets from our field notes illustrate his efforts to achieve a balance between maintaining a “law and order” atmosphere while nurturing compassionate relationships with the students. - At lunch he plays ball in the gym with the kids. He walks the halls during class. He said that kids never really know where he is. They know he is around but they never know where. He said that after school, he goes out to the bus area. He stays out there until 4 p.m. - He said he enjoys interacting with the kids. He said, “First and foremost, you have to form a relationship with the kids.” He said the kids know everything. The only way to find out what is going on is to get it from the kids. - He frequently refers to his Carver kids as his “babies.” - After he arrested a young girl for possession of a knife, the SRO attempted to comfort her. He knows her family, knows what grades her sisters are in, and knows their names. He knew she was concerned and nervous about her mom coming to pick her up at the police station, and so he kept trying to make her relax and smile. She laughed with him while he teased her. She was frightened at the sight of her mom. SRO took her mom aside to speak with her. When SRO did this, she told me that the whole school likes the SRO. She said, “He’s nice.” She told me that he helps her and he laughs with the students. - While we were talking, a female student came into his office with some baklava. She made it in home economics and she brought in two pieces for him. - He said that most of the kids like him and the bad kids “tread lightly around him.” - While we were sitting in his office, he said hi to the kids as they passed by. He addressed some of them by name. - He went out to move his car; I stayed in the office. He walked back with some kids. He was talking and laughing with them. He seems to really know these kids. He makes these kids fell like they are really special. He seems to have a way of making each kid feel special. He knows a lot of the kids by name. He said he enjoys speaking with the kids on common ground. They love to tease him. - We saw a group of kids assembled in the gym. SRO asked the first kid what was going on. After SRO found out what was going on, and he knew the kids were where they were supposed to be, he started joking and talking with a lot of the kids. He arm-wrestled one of the kids. Everyone seemed to know the SRO and they seemed to respect him. - While we were outside, I saw another kid show SRO his report card. These kids seem to really love the attention they receive from SRO. - The SRO’s brother was outside Carver also. Both of them teased the students; they both know the students well. They addressed a lot of them by name. You can tell that these kids just love them. - While I was waiting outside, I saw one parent wave to SRO from her car. The kids appear to feel perfectly comfortable playfully hitting and slapping SRO. There appear to be no strong boundaries. - Upon our arrival, we again go to the principal’s office. After leaving the office, we went to get a kid out of class for hitting another kid. SRO approached the classroom, opened the door and called the boy out. He questioned the boy in the hallway, asking him why he hit the other kid. The SRO took both boys into a conference room in the principal’s office. He asked both boys, “What’s going on?” He made both boys apologize and shake hands with each other. He told them that talking and not hitting could have solved this misunderstanding. He said he is not going to charge them. He instructed them on how to handle this type of situation in the future. He also helped them see that assaulting someone is a criminal offense. From my observations so far, I can see that this SRO really tries to teach the kids that some of the things they are doing in school are actually violations of law and that there are legal consequences. He wants them to see the seriousness of the situation. Although he tries hard to build caring relationships with the students, it is clear that his role as a police officer is omnipresent. He keeps a “photo album” of arrestees. He keeps his police radio on at all times. When our observer noted during a visit to a student’s home that it was the first time she saw him wearing his police hat, he remarked that he does not wear it at school, but he does put it on while conducting his “law enforcement” duties. He obtains information from student informants about offenses committed by other students. He said that sometimes he leaves Carver to provide backup and respond to calls. If a call is close to Carver, then he goes if nobody else is available. “I’m still am officer,” he said. While he works hard to maintain the “velvet glove,” the “iron fist” is always there when it is needed. During the summer, the SRO said he tries to keep a “foundation with the kids.” He coaches the PAL leagues and stays in touch with the kids during the summer. When school is closed, he “works the street.” In the summer of 2001, just after our observation period, he was planning to attend a gang school in Mississippi for 3 days; planning to help out with the PAL league until the end of July; and planning to attend a three-day training session in Atlanta for police in schools. He said he will also help with DARE camp for a few days. One final snippet from our field notes serves to illustrate the complexity of the SRO’s role. “While we were waiting outside, a fight broke out. A student approached the SRO to tell him about the fight. The SRO immediately took control of the situation and broke up the fight. The other kids (the spectators) clearly enjoyed watching this fight. I kept trying to get in the front so I could watch the SRO break up the fight but the other kids kept pushing me out of the way. The spectators were obviously egging this fight. The SRO handcuffed both girls involved in the fight. This fight was between two black females. One stopped fighting when the SRO commanded her to do so. The other female did not want to stop; she kept trying to pull away from the SRO. He was VERY firm with her. After he had her handcuffed, she continued to battle him and he jerked her arms behind her body and yelled at her to stop. The SRO definitely changed his attitude. He went from the friendly School Resource Officer to a Spartanburg Police Officer. It is kind of like he put on a different hat. He was now very serious and firm. He did not tolerate disobedience from this girl. He arrested them and called their parents. He said he is releasing them to their parents with a custodial promise; the parents promise to bring the kids to Family Court. Both girls are being charged.” Although Carver’s School Resource Officer was not funded through the School-Based Partnerships grant, it is impossible to separate his influence from the other anti-drug and anti- crime efforts happening within the school. With regard to problem-oriented policing and the SARA model, he was involved in a variety of scanning and analysis activities, as well as forming the bulk of the response activities. (2) Back in Control Program The Back in Control program provides training for parents (Bodenhamer, 1983). District 7 authorities targeted approximately 60 parents of students with disciplinary problems, particularly those with two or more in-school suspensions. They hope to have 10 parents per workshop because they want the parents to bring their children. Students in the Program Improvement Plan are offered points for attending a session with their parents. Unlike the after school program, this program doesn't target a select sample of repeat ISS offenders - it targets all of them. The training was scheduled over an eight-week period and occurred every Monday at the District 7 headquarters. It started in April 2001. The COPS Office cut this program out of the grant, but the Department was able to get $3,000 in private funding to implement it. The Department still clearly viewed it as a fundamental part of the response phase. (3) GREAT Carver also has a Gang Resistance Education and Training (GREAT) Program, funded by the Bureau of Alcohol, Tobacco, and Firearms. GREAT is taught by the SRO's brother, who also serves as the anchorman on "Safety Scene," the Department's television cable program about the police. Conclusion Multiple interventions were taking place in Carver Junior High School during the 2000-2001 school year. Many more programs and efforts took place than we have described here. We have left out much of the programming that was implemented solely within Carver without participation from external agencies. The programs we have described in this section constitute a multi-faceted approach to problem solving in schools. However, from an evaluation perspective, distinguishing between the effects of these many programs is challenging. In the next section, we describe our attempts to assess the impact of these many programs. IV. IMPACT Introduction This project was much more useful for evaluating implementation than impact. Our ability to do a comprehensive impact evaluation was hampered by a number of factors, including the timing of the evaluation, the large number of programs being implemented, the move to a new school, and the lack of data. Nonetheless, we are able to draw some conclusions about impact, limited as these inferences may be. In this section, we trace various pieces of evidence about program impact, in each case highlighting the problems we encountered in reaching firmer conclusions. Hopefully, other agencies or evaluators can learn from our experiences. Reduction in Overall Disciplinary Infractions If the various programs implemented at Carver were successful, then we might expect to find evidence of a reduction in the total number of disciplinary infractions. Spartanburg County School District 7 maintains a standardized database that records disciplinary infractions in Carver and Whitlock. The District made the data for the 2000-2001 school year from this database available to the evaluation team. It contained the records for more than 8,500 separate disciplinary incidents. Carver’s data contained 75 separate infraction types, compared with 66 for Whitlock. The evaluation team collapsed some of these categories together for the sake of parsimony. Exhibit 7 lists the number of infractions by category for Carver and Whitlock. —SEE EXHIBIT 7— Exhibit 7 demonstrates clear differences between the schools. As with crime reporting in general, it is unclear whether these differences are real, or whether they are a product of differences in reporting patterns at each school. Differences in tardiness levels and bus violations, for instance, might be attributed to the large number of students bused to Whitlock compared with Carver. Even if there are differences in reporting practices between the two schools, these differences are likely to be temporally stable throughout the school year. Therefore, it may still be instructive to examine the monthly trends in the number of infractions, paying more attention to their relative trajectories than their overall levels. Exhibit 8 illustrates disciplinary infractions per month for Carver and Whitlock. —SEE EXHIBIT 8— Although Whitlock has a much lower level of reported infractions than Carver, the overall trajectories for each school are strikingly similar. Each appears to be somewhat of a bimodal distribution with peak infraction periods in October and March, and a lull during the holiday break in December and January. Unfortunately, the short time period covered by the data, when combined with possible differences in reporting or recording patterns in the two schools, make it difficult to glean much more than that from Exhibit 8. Reduction in Disputes The specific goal of the School Based Partnerships grant in Spartanburg was to reduce disputes. Using the same data set containing all reported disciplinary infractions at Carver and Whitlock, we coded each infraction type according to whether it constituted a dispute or not. We coded as disputes all incidents in which there was a disagreement between two or more people, in which one or more of the people was behaving disrespectfully, defiantly, disruptively, or violently. A number of infraction types were difficult to code and were based on fine shades of meaning. For instance, failure to remain in detention hall implies an interpersonal dispute since presumably the instructor is imploring the student to stay and the student refuses to comply. Yet, failure to show up for detention hall does not imply an active interpersonal dispute. It is a disciplinary infraction to be sure, but it does not appear to constitute a dispute in most cases. These kinds of decisions were necessary as we attempted to determine whether each infraction type constituted a dispute. Of the 75 unique disciplinary infraction types at Carver, we designated 22 as disputes. Of the 66 types at Whitlock, we designated 18 as disputes. Exhibit 9 illustrates the average number of disputes per day for each month during the 2000- 2001 school year. In computing the daily average for each month, we count only those days in which school was in session. —SEE EXHIBIT 9— The trajectories illustrated in Exhibit 9 are difficult to interpret. Unlike the previous graph, the two schools exhibit different offending patterns with regard to disputes. While both schools started off with monthly increases in disputes from August through November, Carver’s increase was much more pronounced than Whitlock’s. By November, Carver was experiencing nearly 12 disputes per day, compared with just over 6 for Whitlock. Yet, after November, Carver experienced a precipitous drop in disputes through January. Whitlock’s disputes also declined, though much less dramatically. Whitlock’s rate of disputes exceeded Carver’s rate slightly from December through May. However, absent Carver’s massive spike in disputes from August to November, and then its precipitous decline, the patterns for the two schools would be quite similar. Therefore, it is difficult to credit any program at Carver with having produced a decrease in disputes. Once again, absent a longer data series, it is difficult to make any inferences from these trajectories. Reduction in Official School Responses There are many official school responses used at Carver and Whitlock in response to disciplinary infractions. Based on the 2000-2001 disciplinary data provided by the school district, Exhibit 10 lists these responses together with the number of times they were used at each school. —SEE EXHIBIT 10— Once again, the differences between the schools are pronounced. Carver appears to rely far more heavily on in-school suspension and detention, while Whitlock seems to rely on out-of- school suspension and bus suspension. We do not know whether these differences represent actual differences in official responses to disciplinary infractions, or merely differences in recording and reporting practices. Nonetheless, Exhibit 10 illustrates the range of official responses used by both schools. Other than the 2000-2001 disciplinary data provided to us by the school district, we were unable to gain access to monthly counts of in-school suspensions, out-of-school suspension, and expulsions over a long time period. Our only alternative was to examine yearly totals for each official response. Exhibit 11 contains yearly totals for in-school suspensions from 1998-2002, Exhibit 12 contains yearly totals for out-of-school suspensions from 1994-2002, and Exhibit 13 contains yearly totals for expulsions from 1994-2002. —SEE EXHIBITS 11, 12, and 13— The in-school suspension data presented in Exhibit 11 suggest that Carver experienced a decrease from the 1999-2000 school year (prior to the project) to the 2000-2001 school year. At the same time, Whitlock experienced an increase. We are cautious in making too much of this pattern for three reasons: first, the number of in-school suspensions in Exhibit 11 does not match the total numbers listed in the detailed disciplinary data used to create Exhibit 10. Second, the numbers in Exhibit 11 also do not match the numbers contained in the local evaluation report (Social Work Solutions, 2001). Finally, there appear to be fairly large yearly fluctuations throughout the table, suggesting some temporal instability in suspension patterns. With these three concerns in mind, we find it difficult to draw any firm conclusions from Exhibit 11. Nonetheless, if we accept the data at face value, Carver experienced about a 17% decrease in in- school suspensions, compared with a 52% increase in Whitlock. With regard to out-of-school suspensions, Carver experienced an increase and Whitlock experienced a decrease from the 1999-2000 school year (prior to the project) to the 2000-2001 school year. Although the out-of-school suspension figures appear to be more temporally stable, we still have some concerns about their lack of correspondence with Exhibit 10. We also have some concerns about making inferences from yearly (rather than monthly) data. Nonetheless, if we accept the data at face value, Carver experienced about a 10% increase in out-of-school suspensions, compared with about an 11% decrease in Whitlock. Nearly everybody we interviewed told us that expulsions are frowned upon in the District. The data listed in Exhibits 10 and 13 are consistent with this interpretation. According to Exhibit 10, 17 students were “recommended for expulsion” during the 2000-2001 school year. Yet, according to Exhibit 13, none were expelled. Because there are so few expulsions in the District’s junior high schools, we are not able to detect any trends about program impact from expulsion data. Reduction in Official Police Responses From August 2000 to May 2001, there were 73 juvenile arrests made at Carver. The School Resource Officer made 53 of these arrests, with the remainder being made by a number of other officers. We hoped to discern some patterns in offending and official police responses to offending from arrest data for Carver and Whitlock. Unfortunately, arrest data were not available from the County Sheriff for Whitlock. Exhibit 14 lists the SRO’s monthly arrest data for the 2000-2001 school year for Carver. —SEE EXHIBIT 14— We also obtained family court data from the Department of Juvenile Justice to learn more about the results of arrests made at Carver. Once again, without similar data from Whitlock, these findings cannot address program impact. Nonetheless, we present them because they are interesting and informative. Exhibit 15 lists the outcomes of 80 separate charges filed against students at Carver. —SEE EXHIBIT 14— Nearly a third of the cases coming from Carver were dismissed. Only about 10% resulted in a sentence of probation. Consistent with the family court’s mission, many cases appear to be diverted to arbitration, truancy, or other programs. Qualitative Evidence In addition to the quantitative indicators of program success that we have just examined, we also examined qualitative evidence collected through interviews, observations, and focus groups. The local evaluators summed up the qualitative findings succinctly: “there was nearly universal agreement among those interviewed that this project had shown beneficial effects at Carver Junior High in several ways” (Social Work Solutions, 2001). According to the local evaluation report, the project had the following benefits: - “That Officer Johnson had established a presence of ‘friendly authority’ at the school. - “That Officer Johnson has become a sought-after resource to students and teachers for conflict resolution and early intervention.” - “That the specialized after-school groups have been helpful to the students who participated in them.” - “That the ISS conflict resolution classes have shown some positive effects.” - “That some students who do receive ISS referrals are now more likely to take responsibility for their actions and less likely to make excuses.” - “That the number of reported school incidents may have increased, but the severity of the incidents has declined.” Of course, not all of these successes can be attributed to the School Based Partnerships grant since the grant did not pay for the School Resource Officer. Yet, the qualitative evidence supports the assertion that other elements of the project that were grant funded also had some positive effects. For instance, participants in the Program Improvement Plan (or Partners for Points) program told the evaluation team that they realized a number of specific benefits: leadership skills, improved grades, and the opportunity to think about the important issues they face each day. Discussion and Conclusion In some ways, the evidence presented in this section is contradictory. We do not find stable and convincing evidence from the quantitative data we examined to suggest that the interventions implemented at Carver Junior High School produced a wholesale reduction in disputes or other disciplinary infractions. Perhaps this conclusion would have been different with more and better data. We are not concluding that the program had no effect - we simply are unable to discern such an effect with the data to which we had access. Findings from the qualitative portions of the evaluation are largely positive. The members of the project team, as well as students who participated in project-related programming gave the project high ratings. Some students, parents, and teachers reported that they did not know much about the project, leading the local evaluators to recommend that it be publicized more widely. Nonetheless, many people found that the project had achieved its objectives. The after-school program ended up having only a handful of participants, some of whom were not responsible for disputes. Therefore, although it produced benefits for participants, it is unlikely to have produced widespread dispute reduction benefits, particularly for students who did not participate in it. Some informants also suggested that the major benefit was a reduction in the severity of disputes rather than a reduction in the quantity of disputes. Since we did not attempt to measure severity in our quantitative data, this program benefit would not have showed up in our analyses. V. CONCLUSION The majority of Americans view community policing as a solution to problems of crime and disorder in schools. In a 1999 telephone survey of a nationally representative sample of more than 500 registered voters, 81% supported a proposal to establish community policing in schools “as a way of identifying early warning signals and preventing discipline problems or safety hazards from erupting into greater tragedies” (Democratic Leadership Council, 1999). This report provided the details of a process and impact evaluation of one such program implemented in Spartanburg, South Carolina. The project team invested a lot of energy in designing a multi-faceted project meant to reduce the incidence and prevalence of disputes in Carver Junior High School. Some of the project elements were not funded through the School-Based Partnerships grant, but it is clear that these unfunded or alternatively-funded elements were still considered a vital component of the overall project. Given the simultaneous implementation of these elements, it was not possible to isolate the effects of each one in conducting the evaluation. The project team was successful in implementing all of the project’s elements. Not all of them appear to have played as central a role in the project as others. The School Resource Officer received virtually unanimous acclaim from students, teachers, school administrators, and peers in the Public Safety Department and we see a lot of benefits to continuing his role in the school. The anti-violence presentations by officers in the schools also appear to be targeting the appropriate audience and we see no reason why they should not continue. The peer surveys might have produced some information for the Guidance Counselor and the SRO, but we were unable to discern any specific benefits that would be likely to have a wholesale effect on disputes. If the peer surveys continue, we recommend investing energy in collecting and analyzing their findings more systematically, consistent with the precepts of problem-oriented policing. The after-school program attracted fewer students than it originally targeted. Furthermore, the students it did attract do not appear to have been responsible for participating in large numbers of disputes. Should it continue, we recommend increasing the number of students and making a careful effort to enroll those students participating in the most disputes. At the same time, we take to heart one comment from a student in the program who suggested that there are benefits to having a mixed group of participants, some of whom are more serious offenders than others. In addition to these substantive recommendations, we also have some recommendations for conducting periodic self-assessments. The key to becoming a learning organization capable of measuring, diagnosing, and transforming itself is to periodically examine evidence about the organization’s effectiveness. The School District’s disciplinary data have the potential to serve as a wonderful tracking device for detecting patterns in school-based disciplinary infractions and offending. Exhibits 8 and 9 illustrate some of the simple trend analyses that can be conducted regularly, with little expense, to determine whether offending patterns are changing. This kind of ongoing analysis will serve to enhance an already fruitful partnership between Spartanburg’s School District 7 and the Department of Public Safety. REFERENCES American Civil Liberties Union (2000). “South Carolina Town Fights Racial Profiling with Data Collection.” WWW document (March 31). [http://archive.aclu.org/news/2000/w033100c.html]. Andeman, C., Cheadle, A., Curry, D. S., Diehr, P., Shultz, L., & Wagner, E. (1995). Selection bias related to parental consent in school-based survey research. Evaluation Review, 19, 663–74. Bodenhamer, Gregory (1983). Back In Control: How to Get Your Children to Behave. Simon and Schuster. Boyce, J.N. (1992, August 5). Softer style of policing takes hold in cities like Takoma Park, MD. Wall Street Journal, Section A, page 1, column 1. Platt, T., and P. Takagi (1982). The Iron Fist and the Velvet Glove: An Analysis of the U.S. Police. San Francisco: Crime and Social Justice Associates. Crowe, Timothy D. (1990). Designing Safer Schools. Los Angeles, CA: National School Safety Center. Democratic Leadership Council (1999). Survey of 502 registered voters conducted by Penn, Schoen and Berland Associates (June). Esbensen, F. A., Deschenes, E. P., Vogel, R. E., West, J., Arboit, K., & Harris, L. (1996). Active parental consent in school-based research: An examination of ethical and methodological issues. Evaluation Review, 20, 737–53. Federal Bureau of Investigation (2001). Crime in the United States, 2000. Washington, DC: U.S. Department of Justice. Fetro, Joyce V., Karin K. Coyle, and Phong Pham (2001). “Health Risk Behaviors Among Middle School Students in a Large Majority-Minority School District.” The Journal of School Health, 71(1): 30-37. Goldstein, Herman (1990). Problem-Oriented Policing. New York: McGraw-Hill. Hickman, Matthew J. and Reaves, Brian A. (2000). Law Enforcement Management and Administrative Statistics Local Police Departments, 2000. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics (January, NCJ 196002). Jeffrey, C. Ray (1971). Crime Prevention Through Environmental Design. Beverly Hills, CA: Sage Publications. Kann, L., Kinchen, S. A., Williams, B. I., Ross, J. G., Lowry, R., Hill, C. V., Grunbaum, J. A., Blumson, P. S., Collins, J. L., Kolbe, L. J., & State and Local YRBSS Coordinators. (1998). Youth risk behavior surveillance ––––United States, 1997. Journal of School Health, 68, 355-369. Kenney, Dennis J. and T. Steuart Watson (1998). Crime in Schools: Reducing Fear and Disorder with Student Problem Solving. Washington, DC: Police Executive Research Forum. Newman, Oscar (1972). Defensible Space: Crime Prevention through Urban Design. New York: Macmillan. Pressman, Jeffrey L. and Aaron B. Wildavsky (1984). Implementation, 3rd Edition. Berkeley, CA: University of California. Repenning, N.P. (2001). “A Simulation-Based Approach to Understanding the Dynamics of Innovation Implementation.” Organization Science. Rossi, P.H., & Freeman, H.E. (1993). Evaluation: A Systematic Approach, 5th Edition. Newbury Park, CA: Sage. Social Work Solutions (2001). Report on an Evaluation of a U.S. Department of Justice School Parternship Grant to the Public Safety Department of Spartanburg, South Carolina. Columbia, SC (October 16). Spartanburg Public Safety Department (1998). Policing in Partnership...Preservation through Prevention: Strategic Plan. Spartanburg, SC: January 1. Spartanburg Public Safety Department (2001). Law Enforcement Agency Annual Report. Submitted to the Commission on Accreditation for Law Enforcement Agencies (2/7/01). South Carolina Law Enforcement Division (2001). Crime in South Carolina, 2000. Columbia, SC. Takoma Park Police Chief Quits (1993, September 9). Washington Post, p. C5. Whitaker, J.D. (1982). Candidate Fisher, Sheriff Young Shoot it Out in Bethesda. Washington Post. Metro section, p. B5. Winn, Ellen (1999). “Understanding How Change Occurs: Implementation Research in the TANF Era.” The Forum: Research Forum on Children, Families, and the New Federalism, 2(3): 1-4. Exhibit 1: Selected Demographic, Social and Economic Characteristics % Black Spartanburg 1990 45.6% 2000 50.1% South Carolina 1990 29.8% 2000 29.5% United States 1990 12.1% 2000 12.3% % White Spartanburg 1990 53.1% 2000 47.8% South Carolina 1990 69.0% 2000 67.2% United States 1990 80.3% 2000 75.1% % Other Race Spartanburg 1990 1.3% 2000 2.1% South Carolina 1990 1.1% 2000 3.3% United States 1990 7.7% 2000 12.6% % Owner Occupied Housing Units Spartanburg 1990 48.2% 2000 49.8% South Carolina 1990 69.8% 2000 72.2% United States 1990 64.2% 2000 66.2% % Renter Occupied Housing Units Spartanburg 1990 51.8% 2000 50.2% South Carolina 1990 30.2% 2000 27.8% United States 1990 35.8% 2000 33.8% % Living in Same Home as 5 Years Ago Spartanburg 1990 52.4% 2000 51.5% South Carolina 1990 55.6% 2000 55.9% United States 1990 53.3% 2000 54.1% % Households with Single Mother Families Spartanburg 1990 14.5% 2000 14.3% South Carolina 1990 9.5% 2000 8.5% United States 1990 7.6% 2000 7.2% Median Household Income Spartanburg 1990 $22,423 2000 $28,735 South Carolina 1990 $26,256 2000 $37,082 United States 1990 $30,056 2000 $41,994 Median Family Income Spartanburg 1990 $29,304 2000 $36,108 South Carolina 1990 $30,797 2000 $44,227 United States 1990 $35,225 2000 $50,046 % Persons Below Poverty Level Spartanburg 1990 21.8% 2000 23.3% South Carolina 1990 15.4% 2000 14.1% United States 1990 13.1% 2000 12.4% % Families Below Poverty Level Spartanburg 1990 17.1% 2000 19.4% South Carolina 1990 11.9% 2000 10.7% United States 1990 10.0% 2000 9.2% % Unemployed Spartanburg 1990 4.2% 2000 5.4% South Carolina 1990 3.5% 2000 3.6% United States 1990 4.1% 2000 3.7% % HS Graduate or Above Spartanburg 1990 65.9% 2000 72.4% South Carolina 1990 68.3% 2000 76.3% United States 1990 75.2% 2000 80.4% % Bachelor’s Degree or Above Spartanburg 1990 23.4% 2000 26.0% South Carolina 1990 16.6% 2000 20.4% United States 1990 20.3% 2000 24.4% Exhibit 2: Official Crime Statistics 5 National Evaluation Sites Violent Crime per Capitaa South Carolina 12.0 United States 11.7 Colorado Springs, CO 19.3 Redlands, CA 65.4 Seattle, WA 1.5 Spartanburg, SC 8.2 Westwood, MA 5.1 Property Crime per Capitaa South Carolina 45.1 United States 30.6 Colorado Springs, CO 72.7 Redlands, CA 102.2 Seattle, WA 8.3 Spartanburg, SC 44.5 Westwood, MA 36.2 Total Index Crimes per Capitaa South Carolina 57.1 United States 42.3 Colorado Springs, CO 92.0 Redlands, CA 167.6 Seattle, WA 9.8 Spartanburg, SC 52.7 Westwood, MA 41.3 % of Offenses Cleared by Arrest South Carolina 45.2% United States 31.9% Colorado Springs, CO 22.6% Redlands, CA 31.2% Seattle, WA 31.2% Spartanburg, SC 21.2% Westwood, MA 20.5% a. Expressed per 1,000 people Exhibit 3: Characteristics of the Spartanburg Public Safety Department (Police Services Division) # Sworn Officers 146 # Civilian Employees 22 # Total Employees 168 Sworn Officers per Capita a 3.68 Percent Civilian 13% Percent Black 25% Percent Female 28% a. Expressed per 1,000 people Exhibit 4: Mission, Vision, and Values Statement MISSION We, the Public Safety Department of Spartanburg, are committed to providing high quality Police and Fire Services in partnership with the community through dedicated efforts and innovative programs. Our success is critical for improving the quality of life in our city through promoting safety while maintaining respect for human dignity and the individual rights of our citizens. VISION To assure the success of our mission as we move into the 21st century, we will develop partnership that are trusting and interactive, and we will invest in our employees to enhance their problem solving skills and leadership abilities. As a result, the City of Spartanburg will become a safer, more desirable and vibrant community. VALUES INTEGRITY We will continue to strive for total honesty and integrity in order to provide a strong foundation for building trust in our community and organization. RESPECT We are committed to maintaining the utmost respect for each other and the individual rights of the citizens we serve. PRIDE We are proud to serve in a growing and vibrant community. Our pride and confidence in what we do reflects how we feel about our department and the community we serve. LOYALTY We are committed to upholding the constitutions of our nation and state, the principles and values of the Spartanburg Public Safety Department, and the trust of the community we serve. ACCOUNTABILITY We recognize that in our chosen profession, we are held to a higher level of accountability. Therefore, we accept no less than the highest order of responsibility for our decisions and actions. PARTNERSHIPS We are committed to developing trusting and interactive relationships with the community in order to foster mutual respect and understanding. These partnerships will enhance the quality of life for all citizens. We realize that the expression of ideas is meaningless unless actively practiced. Therefore, we are dedicated to all of the above values in order to promote a professional work place and a safe environment for our employees and the citizens which we serve. Exhibit 5: Spartanburg’s Three Junior High Schools Carver Whitlock McCracken 2001 School Report Card Grade Unsatisfactory Unsatisfactory Below Avg. Teachers Satisfied with Learning Environment 32.6% 40% 88.2% Students Satisfied with Learning Environment 30.6% 53.1% 66.9% Students receiving free/reduced lunch 60% 84% 39% Students scoring basic or above on the PACT (English) 57.1% 37.3% 64.4% Students scoring basic or above on the PACT (Math) 52.1% 33.1% 50.1% Student-teacher ratio 20.8 to 1 20.1 to 1 10.4 to 1 Student retention rate 8.5% 10.3% .7% Dropout rate 1.1% .5% 0% Students that are older than usual for grade 8.6% 10.8% 1.7% Students with disabilities other than speech 13.4% 18% 11.7% Student attendance rate 93.9% 91.1% 95% Exhibit 6: Student Survey Sample Sizes School Grade Total Students Students in Sample # Survey Respondents % Survey Respondents* Carver 7 206 68 55 27% 8 202 67 49 24% 9 192 64 46 24% Whitlock 7 214 71 48 22% 8 173 57 44 25% 9 199 66 54 27% * This column lists the percentage of enrolled students at each grade level that responded to the survey. Exhibit 7: Nature of Disciplinary Infractions at Carver and Whitlock Nature of Infraction Carver Incidents Whitlock Incidents Excessive Tardies 1,758 337 Misconduct 1,533 752 Disrespect 496 311 Failure to Attend 466 614 Defiance 437 422 Profanity 154 145 Fighting 123 148 Leaving Without Permission 111 158 Bullying, Threats, Intimidation 68 136 Possession of Unauthorized Substance 21 34 Bus Violation 21 134 Destruction 10 4 Theft 9 14 Unauthorized Presence 6 13 Other 4 59 Without Pass or ID 1 20 Exhibit 8: Total Disciplinary Infractions, August 2000 - May 2001 Exhibit 9: Monthly Dispute Totals for Carver and Whitlock, 2000-2001 Exhibit 10: School Responses to Disciplinary Infractions at Carver and Whitlock Nature of Response Carver Responses Whitlock Responses In-School Suspension 2,369 1,374 Detention 2,102 710 Out-of-School Suspension 395 745 Conference 123 31 Removal from Activity/Class 97 63 Referral 52 27 Warning 29 45 Bus Suspension 18 132 Recommended Expulsion 17 18 Parent Contact 10 17 Community Service 1 4 Other 0 12 Exhibit 11: In-School Suspensions, 1998-2002 Year Carver Whitlock McCracken Total 1998-1999 993 * * * 1999-2000 2,499 1,419 753 4,671 2000-2001 2,077 2,160 1,080 5,317 2001-2002 2,569 1,061 987 4,617 Exhibit 12: Out-of-School Suspensions, 1994-2002 Year Carver Whitlock McCracken Total 1994-1995 489 522 341 1,352 1995-1996 200 226 181 607 1996-1997 218 244 219 681 1997-1998 246 265 159 670 1998-1999 167 190 113 470 1999-2000 187 325 171 683 2000-2001 203 288 196 687 2001-2002 189 358 181 728 Exhibit 13: Expulsions, 1994-2002 Year Carver Whitlock McCracken Total 1994-1995 3 7 7 17 1995-1996 8 19 9 36 1996-1997 3 2 1 6 1997-1998 1 6 2 9 1998-1999 1 2 0 3 1999-2000 1 0 0 1 2000-2001 0 1 2 3 2001-2002 3 2 0 5 Exhibit 14: Arrests by SRO at Carver, August 2000 - May 2001 Month Total Arrests 8 / 2000 9 / 2000 10 /2000 11 / 2000 12 / 2000 1 / 2001 2 / 2001 3 / 2001 4 / 2001 5 / 2001 Exhibit 15: What Happens to Cases Resulting from an Arrest at Carver? Category Cases % Dismissed 25 31.3% Unknown 14 17.5% Pending 11 13.8% Probation 8 10.0% Nolle Prosequi 6 7.5% “Choices” Program 5 6.3% Arbitration 5 6.3% Other 3 3.8% Determinate Sentence 2 2.5% Pre Trial 1 1.3% We use the term “project team” throughout this report to refer to the police and school officials who developed and implemented the School-Based Partnerships project. We use the terms “research team” and “evaluation team” interchangeably to refer to the individuals who conducted the local and national evaluations of the program. The population of Spartanburg is 39,673 according to the 2000 Census. The population of Spartanburg County is 253,791 according to the 2000 Census. The population of the Greenville-Spartanburg-Anderson, SC Metropolitan Statistical Area is 962,441 according to the 2000 Census. Spartanburg Public Safety Department, 1998. Spartanburg Public Safety Department, General Order 450.1. For more information on CPTED, see Jeffrey (1971), Newman (1972), or Crowe (1990). Spartanburg Public Safety Department, General Order 560.1. American Civil Liberties Union, 2000. Spartanburg Public Safety Department, General Order 100.5. Spartanburg Department of Public Safety, 2001. Hickman and Reaves, 2003. Hickman and Reaves, 2003. Whitaker, 1982. Boyce, 1992. “Takoma Park Police Chief Quits,” 1993. S.C. Code Ann., Sections 59-63-310-390 (2000). South Carolina School Crime Incident Reports, 2000, 2001, and 2002. The State of South Carolina, Annual School Report Card, 2001. Carver Junior High. The State of South Carolina, Annual School Report Card, 2001. Carver Junior High. Active consent means requiring participating students to submit written permission from a parent to participate in the survey. Passive consent means informing parents that a survey is being planned and instructing them to notify school officials if they do not want their child to participate. We were provided with only one year of disciplinary data and arrest/summons data for the year during which the program was implemented (2000-2001). A longer data series could have been useful for establishing offending trajectories prior to the program’s implementation. Unfortunately, arrest patterns from previous years would not have been useful because a different School Resource Officer with a different philosophy about his role was previously assigned to the school, therefore the previous year’s arrest data would not have been comparable with the current year’s data. We would also have liked to access data for the year following program implementation, but the school moved to a brand new building during that year. The old school was designed poorly, had graffiti on the walls, and offered numerous opportunities for offending in poorly guarded areas. Moving to a new school that was designed better and showed more evidence of being cared for might have independently affected offending levels. Therefore, had we collected offending or arrest data from the new school during the 2001-2002 school year, we would not have been able to attribute any changes to the effects of the School Based Partnership program. Any such changes might just as easily have resulted, directly or indirectly, from moving to a new building. For these reasons, some of our data series were only available for the 2000-2001 school year. Furthermore, these numbers do not match yearly totals for in-school and out-of-school suspensions listed in Exhibits 11 and 12. For instance, from 1998/1999 to 1999/2000, Carver experienced nearly a 152% increase in in-school suspensions. All of the items in the bulleted list are extracted verbatim from the local evaluation report (Social Work Solutions, 2001, p. 6). South Carolina Law Enforcement Division (2001). Federal Bureau of Investigation (2001). We use the term misconduct to refer to those rule violations that do not involve a “victim” or that do not bring pain or hurt to other people. Some of the behaviors classified within the misconduct category include locker violations, dress code violations, horseplay, being unprepared for class, disturbing the class, etc...