Investigative Report Prepared on behalf of The Board of Trustees of Claremont McKenna College by O'MELVENY & MYERS LLP Apalla U. Chopra, Esq. Carolyn Kubota, Esq. April 17, 2012 Investigative Report of O'Melveny & Myers LLP TABLE OF CONTENTS Page I. Executive Summary ......................................................................................................... 1 A. Principal Findings .................................................................................................. 1 1. 2. B. II. A. Scope of Misreporting................................................................................ 1 Role of Leadership..................................................................................... 1 The College's Cooperation and Response ............................................................. 2 Scope of Investigation............................................................................................ 3 1. 2. 3. The College's Admission Data Processes and the VP's Related Conduct ...................................................................................................... 3 Accuracy of Material, Externally Reported Admission Statistics ............. 3 Role of Leadership..................................................................................... 4 Interviews................................................................................................... 4 Document Review...................................................................................... 4 Verifying the Accuracy of Selected Admission Statistics ......................... 4 Investigative Scope and Methodology............................................................................ 3 B. Investigative Methodology .................................................................................... 4 1. 2. 3. III. Investigative Report......................................................................................................... 6 A. Background ............................................................................................................ 6 1. 2. B. 1. 2. 3. 4. 5. 6. C. D. The Vice President..................................................................................... 6 Events Giving Rise to the Investigation..................................................... 6 CMC Uses a Holistic Admission Process.................................................. 7 SAT Statistics............................................................................................. 7 ACT Statistics ............................................................................................ 8 Top 10% Statistics ................................................................................... 10 Application Statistics ............................................................................... 11 Acceptance and Enrollment Statistics...................................................... 12 The Admission Process and Reporting of Admission Statistics at CMC .............. 7 Internal Reporting of Discrepancies Between Admission Statistics.................... 12 The VP Alone Knew of and Engaged in the Misreporting of Admission Statistics ............................................................................................................... 13 1. The Role of the Professional Staff in the Admission Office and IR........ 13 i Investigative Report of O'Melveny & Myers LLP TABLE OF CONTENTS (continued) Page 2. 3. 4. E. The Role of the Administrative Staff in the Admission Office ............... 13 The Role of Leadership............................................................................ 13 The VP's Conduct Remained Undiscovered for Multiple Reasons......... 14 The VP's Explanations for Falsely Reporting SAT Statistics, and the Role of Leadership ....................................................................................................... 14 1. 2. 3. 4. The VP's January 24, 2012 Statement..................................................... 14 The Pressure Described Was within the Norm ........................................ 15 The VP's Additional Explanation for His Conduct ................................. 17 Rankings Did Not Motivate the VP's Conduct ....................................... 17 IV. Conclusions..................................................................................................................... 17 APPENDIX................................................................................................................................. 18 SAT Statistics................................................................................................................... 19 ACT Statistics - Composite ............................................................................................. 22 Top 10% Statistics ........................................................................................................... 24 Application and Acceptance Rate Statistics .................................................................... 25 ii Investigative Report of O'Melveny & Myers LLP I. Executive Summary Claremont McKenna College ("CMC" or the "College") is a private liberal arts college located in Claremont, California. On January 9, 2012, a CMC employee reported to the College's General Counsel discrepancies between internally and externally reported SAT statistics for the College's 2011 incoming class. The College's president, Pamela Gann ("the President"), immediately initiated an internal investigation. On January 24, Richard Vos, the then-Vice President of Admission and Financial Aid (the "former VP" or "VP"), admitted to reporting false SAT statistics, starting in 2005. He resigned from his position on January 29. The next day, the President disclosed these events to the CMC community and various third parties. The College's Board of Trustees (the "Board"), shortly before the President's disclosure, retained O'Melveny & Myers LLP to perform an external investigation. The results of our investigation are described below. A. Principal Findings 1. Scope of Misreporting We found that the VP reported inaccurate SAT, ACT, class rank, and application statistics beginning as early as 2004.1 With respect to SAT statistics, we confirmed that the College's initial disclosures about the VP's conduct, and the College's corrected SAT statistics issued the week of January 30, 2012, were accurate. The degree of inaccuracy and the circumstances related to the misreporting of the other statistics varied.2 When engaging in this conduct, the VP did not alter individual student scores in the College's database or in students' files. We also found that the VP acted alone and no one at the College had knowledge of his conduct. Multiple factors enabled the VP to conceal his conduct from CMC's leadership and from employees within the Admission Office and Office of Institutional Research ("IR").3 The VP was a long-time and trusted executive who closely controlled and exercised ultimate authority over the reporting of admission statistics at the College. The absence of a process to independently verify the reports the VP generated, reinforced his control over the admission statistics reporting function. 2. Role of Leadership We found that no one in the College's leadership directed or encouraged the VP's actions; the VP does not contend otherwise. The VP did assert, however, that the President Tables containing comparisons of the misreported and accurate statistics, from 2004 through 2011, are set out in the Appendix. How and why the VP compiled and misreported each of these statistics is set out in this report. (See Sections III (B)(3-5)). 3 2 1 The Office of Institutional Research coordinates the collection and dissemination of information about the College for internal and external purposes. The Office also conducts research to support institutional decision-making and strategic planning. IR does not audit data or statistics provided to it from other parts of the College. 1 Investigative Report of O'Melveny & Myers LLP pressured him to maintain or increase SAT scores for the College's incoming classes. While we believe the VP sincerely felt pressure to achieve particular SAT goals, we do not believe that the pressure he described exceeded the norm for an executive-level employee. Instead, we concluded that the VP's disagreement with the President's admission priorities primarily motivated the VP's choice to falsify SAT statistics. In 2002, the College adopted a ten-year Strategic Plan. That Plan sets out the institution's overarching priorities, which included strengthening the College's alreadyaccomplished student body. While the College measured student quality across a range of attributes, SAT scores were among the important benchmarks considered. The VP, in consultation with the President, set annual admission goals derived from these priorities. One of his admission-related goals included maintaining or incrementally increasing the SAT statistics for the College's incoming classes. We believe that these institutional priorities and departmental goals were legitimate and reasonable. The manner in which the President held the VP accountable for reaching his admission goals also was reasonable. The VP did not describe, and we found no evidence of, verbal abuse, economic pressure, or any form of intimidation that might reasonably lead us to view the pressure exerted as coercive or improper. In addition, like all other employees within the College, the VP's compensation was never tied to meeting specific performance goals. Instead, his compensation was tied to his overall, qualitative job performance. During our investigation, the VP offered further explanation for his SAT-related conduct. He asserted that the President's admission priorities, while achievable, did not match his own. The VP stated that the College's applicant pool could readily support the goal of maintaining or incrementally increasing SAT scores. He also acknowledged that achieving the SAT goals required some adjustments to the priorities applied in the admission process. Rather than make those adjustments, however, the VP continued to apply his own admission priorities. When the VP's admission decisions failed to produce the desired SAT statistics, he chose to falsify them to conceal this outcome from the President.4 B. The College's Cooperation and Response Throughout our investigation, the College has been fully cooperative. We have had complete access to Board members, College employees, and documents (electronic or otherwise) during our process. The College's leadership promptly responded to the findings of their internal investigation and our report.5 CMC internally and externally issued corrected SAT statistics during the week of January 30, 2012, and issued corrected ACT, class rank, and Application statistics during the week of April 16. Interim processes also have been put into place around the The VP attributes no fault to leadership (or himself) for the misreporting of ACT, class rank, and Application statistics. The evidence shows no involvement or knowledge on the part of leadership. Written recommendations are beyond the scope of this report. In response to our report, however, the College's President has made a series of specific recommendations to the Board which they approved on April 12, 2012. 5 4 2 Investigative Report of O'Melveny & Myers LLP reporting of admission statistics. Going forward, CMC leadership is reviewing how statistics are internally and externally reported across the College and is creating appropriate processes to ensure the accurate reporting of the College's statistical information. II. Investigative Scope and Methodology A. Scope of Investigation Our investigation included three subjects: (i) the College's admission data processes and the VP's related conduct, (ii) the accuracy of material, externally reported admission statistics, and (iii) the role of leadership, if any, in the VP's conduct. 1. The College's Admission Data Processes and the VP's Related Conduct We examined how the College collects admission data, maintains that data, and compiles and reports admission statistics internally and externally. We also reviewed the role the VP and other CMC employees played in that process. We evaluated that information to determine how and why the VP compiled and misreported admission statistics, and also to determine whether any other employees participated in or had knowledge of the misreporting. 2. Accuracy of Material, Externally Reported Admission Statistics In light of the VP's admitted misconduct, the Board asked that we review the College's reported SAT statistics. The Board also asked that we review other material admission statistics to determine if they had been misreported. In response to the Board's requests, we selected the following externally reported admission statistics for review: Standardized exam statistics (i.e., SAT and ACT statistics); Class rank statistics (i.e., freshman who graduated in the top 10% of their high school class as a proportion of those students whose high school reported class rank ("Top 10%" or "class rank")); Application statistics (i.e., the total number of student applications the College received ("Application")); Acceptance statistics (i.e., the total number of students the College accepted through its admission process ("Acceptance")); and Enrollment statistics (i.e., the total number of freshman students who enroll at the College ("Enrollment")). Because the College considers either SAT or ACT exam scores when making admission decisions, we chose to review ACT, in addition to SAT, statistics. We selected the other metrics because, taken together, they are a reflection of the incoming class's academic excellence (Top 10%), the College's acceptance rate (Acceptance as a proportion of Application), and the College's enrollment yield (Enrollment as a proportion of Acceptance). 3 Investigative Report of O'Melveny & Myers LLP We chose and applied the same review period, 2004-2011, when determining the accuracy of all selected admission statistics. The start of the review period, 2004, was chosen principally to test the VP's assertion that he began to falsely report SAT statistics in 2005. 3. Role of Leadership We examined whether the College's leadership directed or encouraged the falsification of SAT or other admission statistics. As a part of that assessment, we reviewed how the College set and accomplished its institutional priorities and goals. While we reviewed the goal-setting process across the College, we focused most closely on the process behind the College's admission goals. Before and during our investigation, the VP offered some explanations for his SATrelated conduct, which involved the President. We assessed the validity of those explanations in the context of how leadership, if at all, directed or encouraged the VP's conduct. B. Investigative Methodology 1. Interviews We interviewed 24 people, including the VP. Most of the individuals we interviewed work in the College's Admission and IR offices. We also interviewed the President and each of the College's current Vice Presidents. Excluding the former VP, each of our interviews lasted between approximately 1.5 and 3.5 hours. We conducted three in-person and two telephonic interviews with the former VP for a total of approximately 15 hours. 2. Document Review In total, we reviewed more than 14,990 documents containing approximately 102,050 pages. We reviewed approximately 7,517 emails (including attachments) containing approximately 52,619 pages. We also reviewed 2,713 non-email electronic documents containing approximately 30,064 pages, including documents from the VP's imaged computer hard drive. We reviewed more than 5,300 hard-copy documents, containing approximately 19,367 pages. As part of our hard-document review, we examined all admission-related documents for each member of the 2011 matriculated class. We also reviewed available admission-related documents for each member of the 2010 matriculated class. 3. Verifying the Accuracy of Selected Admission Statistics We determined for multiple reasons that the data contained in the College's admission database was and is accurate. First, the data in CMC's database, most of which is uploaded directly from third-party sources such as the Common Application6 and the College Board,7 produces statistics different than those the VP compiled and reported. Second, the VP described 6 7 The Common Application is a form application that students may submit to colleges and universities. The College Board is the entity that proctors the SAT exam and reports student scores. 4 Investigative Report of O'Melveny & Myers LLP creating reports of inaccurate statistics based on printed reports of accurate data from the CMC database (see, e.g., Sections III (B)(2-3)). Finally, the VP stated he was not technologically proficient. He told us that he did not, and lacked the expertise to, alter data in the database. Instead, when he compiled inaccurate statistics, the VP used paper, a pencil, and a calculator. (See Section III (B)(2)). Because we concluded that the integrity of the College's database had not been compromised, we relied upon data contained in it to verify the admission statistics within the scope of our investigation. We used the following methods to do so: Method Used to Calculate SAT Statistics After collecting SAT data from CMC's database, we uploaded that data into an Excel spreadsheet. We then calculated the mean, median, range, and 25th and 75th percentiles for each freshman class from 2004 through 2011. In calculating SAT statistics, we considered only applicants' highest test scores used in the admission decision.8 That methodology is consistent with and derived from the College's admission policies. Method Used to Calculate ACT Statistics After collecting ACT data from CMC's database, we uploaded that data into an Excel spreadsheet and calculated the composite mean, range, and 25th and 75th percentiles for each freshman class from 2004 through 2011. In calculating ACT statistics, we considered only applicants' highest test scores.9 The College intends to use this methodology going forward and will be amending its admission policies to reflect this practice. Method Used to Calculate Top 10% Statistics Numeric class rank and class size data, if reported on matriculated students' high school transcripts or school reports,10 is contained in CMC's database. We calculated the Top 10% statistic based on that data only for the period 2004 through 2011. The College has preliminarily decided to use that same methodology going forward when calculating its Top 10% statistic. When an applicant submits multiple SAT exam scores, the College will use only the highest score achieved on each subpart of the exam (e.g., math or critical reading) in its admission decision. This practice is referred to as "superscoring." In addition, if an applicant submits both SAT exam scores and ACT exam scores, the College will use in its admission decision only the highest exam score achieved on either the SAT or ACT exam. We calculated the SAT statistics consistent with the College's admission policies. If an applicant submitted multiple ACT exam scores, we superscored the exams and used those figures in our ACT statistics calculations. In addition, if an applicant submitted both SAT exam scores and ACT exam scores, we included the ACT superscore in our calculation only if the ACT superscore was higher than the applicant's SAT superscore. The College, going forward, intends to follow this methodology. 10 9 8 A school report is a document provided by a student's high school counselor. It contains academic performance and other information about a particular student. School reports may or may not contain class rank information. In addition, the school report often is accompanied by counselor and teacher recommendations. 5 Investigative Report of O'Melveny & Myers LLP To determine the accuracy of the VP's Top 10% statistic for 2011, we also reviewed each complete Admission file for the 2011 freshman class independently and with the VP. (See Section III (B)(4)). Method Used to Calculate Application, Acceptance, and Enrollment Statistics We relied upon previously printed reports from the College's database to verify the Application, Acceptance, and Enrollment statistics. We obtained these reports for the full review period, 2004 through 2011. We compared these reports with data in the College's database for the period 2005 through 2011. Where there were inconsistencies between statistics we calculated from the database and the previously printed reports, those differences were explained through student movement between classes, e.g., student deferrals. For the year 2004, Application, Acceptance, and Enrollment data is not contained in the College's database. To verify these aggregate statistics for 2004, we compared the previously printed report with other internal and external reports of the same information from the same time period. III. Investigative Report A. Background Claremont McKenna College is a highly selective undergraduate liberal arts college with a curricular emphasis on economics, government, and public affairs. Established in 1946, CMC is a member of The Claremont Colleges, a consortium of seven independent institutions in Claremont, California. CMC is accredited by the Western Association of Schools and Colleges and has a current total enrollment of more than 1,200 degree-seeking undergraduate students. 1. The Vice President The former VP joined CMC in 1987 as Dean of Admission and Financial Aid. In 1989, he was promoted to Vice President of Admission and Financial Aid. He remained in that position until his resignation on January 29, 2012. Throughout his 24-year tenure, the VP was well-regarded at CMC, especially by the members of the Admission Office. He admitted students that the College collectively viewed as academically talented and well-rounded. The VP was considered a trustworthy member of the CMC community. 2. Events Giving Rise to the Investigation On January 9, 2012, a CMC employee reported to the College's General Counsel discrepancies between internally and externally reported SAT statistics for the 2011 incoming class. The General Counsel immediately notified the President. That same day, the President asked Jerry Garris, the Vice President, Dean Emeritus of CMC (the "Dean Emeritus"), to conduct an internal investigation. The Dean Emeritus began his work on January 9. During the investigation, the Dean Emeritus interviewed the VP. As the highest-ranking executive within the Admission Office, the VP was ultimately responsible for reporting admission statistics at the College, including SAT 6 Investigative Report of O'Melveny & Myers LLP statistics. On January 24, the VP admitted to the Dean Emeritus that he had falsely reported SAT statistics, starting in 2005. The next morning, on January 25, the President was notified of the VP's statements. The President immediately informed the Chair of the Board. The Chair notified the Board's Executive Committee on January 26 and the full Board on January 27. The VP tendered his resignation to the College on January 29. On January 30, the President distributed a letter to the CMC community, disclosing these events. That same week, the College posted on its website and distributed to third parties corrected SAT statistics. Before the President's announcement, the Board retained O'Melveny & Myers LLP to conduct an external investigation. B. The Admission Process and Reporting of Admission Statistics at CMC 1. CMC Uses a Holistic Admission Process CMC uses a holistic admission process that considers both qualitative and quantitative factors. Admission officers look at a range of attributes, including academic performance and promise, personal characteristics, achievements and goals, participation in activities and organizations, interests, talents, and leadership potential. SAT scores are only one of many admission factors considered. Each year, the Admission officers cast preliminary votes to admit about twice as many students as there are spaces in the class. Admission officers then meet to "shape" the class. CMC's admission priorities drive the decisions at these meetings. The VP maintained close control over this process. An important part of the admission function at CMC is the internal and external reporting of admission statistics about each incoming class. As the highest-ranking Admission officer at the College, the VP was ultimately responsible for that function. No process existed to verify independently the reporting of admission statistics. In the absence of such a process, the VP was able to exercise almost complete control over the reporting of admission statistics for many years. 2. SAT Statistics SAT statistics are one of the most commonly reported metrics about CMC's incoming classes each year. Beginning in at least 2004, the VP generated false SAT statistics for each incoming class. He compiled these false statistics about two to three weeks after the start of each fall semester. His process worked as follows: The VP first asked an Admission staff member to generate from the CMC database a report of SAT scores for the incoming class.11 The staff member printed a report containing 11 Obtaining data from the CMC database requires the use of a software reporting program called Cognos. To use Cognos, training is required. Few employees in the Admission Office received Cognos training, and most, even after the training, were unable to use Cognos proficiently. 7 Investigative Report of O'Melveny & Myers LLP accurate SAT data. Upon receiving it, the VP chose an SAT median he wished to publicly report. The VP then worked backward to determine the means, ranges, and 25th/75th percentile statistics that would be consistent with the false median. Although he manipulated these statistics, the VP did not change any individual student scores in the College's database or student files.12 The VP then asked a different Admission Office staff member, who had not seen the original accurate data, for a report template called the "Five-Year Report," on which the VP handwrote the false SAT statistics. That staff member typed and finalized the Five-Year Report, and the VP submitted it to IR. IR internally reported the VP's false statistics through a CMC publication called Factbook. Factbook is a compilation of institutional statistics assembled by IR. Since at least 2004, Factbook incorporated the VP's falsified SAT statistics. It was published on or about October 31st each year and circulated to certain members of the CMC faculty and staff. Once Factbook was published, the Admission Office relied upon it in responding to most external surveys throughout the year. One admission officer is designated as the "survey person." That person was responsible for responding to all admission-related external surveys -except U.S. News & World Report. Although the "survey person" has access to accurate admission data through the College's database, the admission staff's limited proficiency in generating data reports created a disincentive for doing so. Instead, the survey person typically asked the VP for the statistics or relied upon Factbook. The VP personally handled CMC's response to U.S. News & World Report rather than delegating that responsibility to the Admission Office "survey person." U.S. News & World Report issues its survey questions in about March or April each year. The survey addresses the prior year's entering class. IR compiles information from each CMC department that contributes to the College's response. The VP handwrote the Admission Office response, incorporating false SAT statistics he had prepared the previous fall for Factbook. He then submitted the Admission section to IR to be incorporated into the CMC response. Where the U.S. News & World Report survey requested mean SAT statistics, the VP instead submitted higher, false median statistics. We calculated the accurate SAT statistics reflecting the mean, median, range, and top 25th and 75th percentiles for each incoming class from 2004 through 2011. We compared the accurate statistics to the VP's falsely reported SAT statistics. Those comparisons are set out in the attached Appendix. Our SAT calculations, contained in the Appendix, are consistent with the corrected statistics the College reported during the week of January 30, 2012. 3. ACT Statistics In addition to SAT statistics, the College also externally reports its ACT statistics. During most years from 2004 to 2011, the VP reported inaccurate ACT statistics. 12 Admission decisions were based on students' individual applications, including their accurate standardized test scores. 8 Investigative Report of O'Melveny & Myers LLP The VP told us he appropriately reported ACT statistics and explained to us how he compiled and reported these figures.13 His explanations were unclear and internally inconsistent. He first said that he compiled his statistics from data reports an Admission Office staff member printed for him from the CMC database. The VP then added, however, that he often did not rely on information from the CMC database when reporting ACT statistics. On those occasions, he said he would either report the same ACT statistics he had reported the year before, or he would create statistics that appeared correct based on his "educated judgment" or "extrapolations." The VP also explained that he believed reporting ACT statistics was unimportant and that he devoted little time and attention to this task. Through our investigation, we found that each year from 2004 through 2010, the VP reported the identical composite mean (31) and composite 25th and 75th percentile statistics (29-33). The recurrence of identical scores for seven consecutive years appeared to be an unlikely outcome. The numerical variation reflected in the accurate statistics we calculated confirmed our assessment. In light of the VP's unclear explanation of how he calculated ACT statistics, his pattern reporting, and his statement that he viewed reporting ACT statistics as unimportant, the evidence suggests that the VP picked figures he believed were appropriate and reported them.14 While the VP provided his ACT statistics to IR, those figures were not reported internally within the College. The VP did, however, incorporate his mostly inaccurate ACT statistics into the Admission response to the U.S. News & World Report survey. Generally speaking, the Admission "survey person" also incorporated the VP's ACT statistics into the Common Data Set15 and IPEDS16 surveys. We calculated the most commonly reported ACT statistics -- composite mean, ranges, and 25th and 75th percentiles. We then compared the VP's reported figures to our calculations. The VP's reported ACT statistics were, at times, accurate. In a few circumstances, his reported statistics were lower than the true statistics. Generally, however, the VP's reported statistics were higher than the true statistics. The correct ACT statistics, compared to the VP's reported figures for 2004 through 2011, are provided in the Appendix. 13 Throughout our investigation, the VP asserted that he appropriately reported all admission statistics, with the exception of SAT statistics. The VP attributes no fault to himself or anyone in leadership for the misreporting of ACT statistics. The evidence demonstrates that leadership had no involvement in or knowledge of the misreporting. Numerous publishers collect information each year for their college guide and ranking publications through lengthy campus questionnaires. To reduce the amount of time and effort required to respond to duplicate questions on multiple surveys, publishers and the education community collaborated to produce a standard format -- the Common Data Set -- to capture most of the requested data. Each year from 2004 through 2011, CMC responded to the Common Data Set survey and posted that information on its website. IPEDS (Integrated PostSecondary Education Data System) is a data-collection system within the National Center for Educational Statistics, which is a part of the Department of Education. Each year from 2004 through 2011, CMC responded to an IPEDS survey. 14 15 16 9 Investigative Report of O'Melveny & Myers LLP 4. Top 10% Statistics We also reviewed an admission statistic commonly referred to as the Top 10%. The Top 10% is one of several metrics that external surveys use to assess the academic quality of colleges' and universities' incoming classes. According to these surveys, the Top 10% measures those freshman students who (i) graduated in the top 10% of their high school class, as a proportion of (ii) those students whose high school reported a class rank.17 Most students who apply to CMC attend high schools that do not release class rank information about their graduates. For example, only approximately 26% of the students in CMC's 2011 entering class graduated from high schools that reported numeric class rank information on student transcripts or school reports. If an applicant's high school provides numeric class size and rank information on a student's transcripts or school report, CMC's database extracts and stores that information. An applicant's file, however, may contain other information about class rank. For example, a college counselor's recommendation letter might refer to a student's class rank, or the school might provide a GPA distribution graph that, combined with the student's GPA, indicates where the student ranks within the class. This information is not collected in CMC's database. The VP told us that he appropriately reported class rank statistics and personally calculated those figures. He developed his own method of performing this calculation, which required a review of the complete admission file for each student in the incoming class. CMC maintains for one year, the complete admission files of its incoming classes. With respect to our efforts to verify the VP's Top 10% calculations, complete information was available for only the 2011 incoming class. For the class entering in 2011, we reviewed each admission file with and without the VP and tried to verify his Top 10% calculation. We were unable to do so. The VP also was unable to replicate the Top 10% statistic he reported for 2011. From this file-by-file review, we concluded that the VP's method was undocumented, not replicable, and ad hoc. The VP stated that the compilation of class rank statistics is an "art not a science." He stated that colleges receive little guidance as to what information should be considered when determining class rank, but that it was designed to reflect the excellence of the incoming class. Accordingly, when responding to survey questions seeking ranking information, the VP said he relied upon a variety of applicant information. Some of the information the VP relied upon to calculate his class rank statistic was appropriate, such as numeric class size and rank information reflected on a student's school transcript or school report. He also relied upon counselor letters which explicitly placed a student in the Top 10% of his or her class. 17 The Common Data Set survey defines the Top 10% as the "[p]ercent of all degree-seeking, first-time, first-year (freshman) students who had high school class rank within each of the following ranges (report information for those students from whom you collected high school rank information)." The first range described is "[p]ercent in the top tenth of high school graduating class...." The U.S. News & World Report survey uses substantially the same definition. 10 Investigative Report of O'Melveny & Myers LLP To "maximize" the Top 10% statistic, however, the VP relied upon methods that were unsupportable. For example, if a student attended what the VP considered to be an "elite" high school and was ranked below the top 10%, the VP disregarded the student for purposes of this calculation. As a student whose high school provided an explicit ranking, this student appropriately should have been included in the percentage calculation. The VP disregarded such a student even if the VP had included other students from the same school in the Top 10%. The VP used this same methodology when relying on GPA distribution graphs. The VP would disregard a student in his calculation if the student's GPA placed him or her outside the Top 10% on a GPA distribution chart, but would include a student in his calculation if the student's GPA placed him or her within the Top 10% on the same GPA distribution chart. By these and similar means, the VP inflated the Top 10%.18 According to the VP's calculations, 54% of the 2011 class reported a class rank; of them, he determined that 85.6% were ranked in the Top 10%. In contrast, using only numeric class size and class rank information from CMC's database, 26% of the 2011 class reported a class rank. Of those students, we determined that 70.9% were in the Top 10%. Internally, the Top 10% figure is not published in Factbook but is reported to CMC's leadership. Externally, the VP included his inflated Top 10% statistics in the Admission Office's response to the U.S. News & World Report survey. Generally speaking, the Admission "survey person" also incorporated the Top 10% statistic from the VP's Five-Year Report into the Common Data Set survey response. Going forward, CMC leadership has preliminarily decided to calculate the Top 10% based on numeric class size and rank information reported on applicants' final high school transcripts and school reports. CMC leadership recognizes that this calculation applies to a small portion of its incoming classes, but has concluded that it represents the most accurate and verifiable response to the questions posed by surveys requesting Top 10% statistics. The information also is stored in, and easily replicable from, CMC's database. CMC's Top 10% statistics for the years 2004 to 2011, applying the College's new methodology,19 are set out in the Appendix. 5. Application Statistics Each year the College reports the number of applications for admission it receives. Internally, CMC refers to this statistic as the number of "Applications Filed." Based on longstanding practice at CMC, the "Applications Filed" statistic includes a category called "Incomplete" applications. The "Incomplete" category includes files that do not meet all of the College's stated application requirements, such as the submission of a high school transcript. 18 The VP attributes no fault to himself or anyone in leadership for the misreporting of class rank statistics. The evidence demonstrates that leadership had no involvement in or knowledge of the misreporting. The numeric class rank and class size data in the College's database is currently extracted from both interim and final student transcripts and school reports. Going forward, however, the College will rely only on students' final transcripts and school reports when calculating its Top 10% statistic. 19 11 Investigative Report of O'Melveny & Myers LLP Internally, CMC accurately reported the number of "Applications Filed" from 2004 to 2011. In external reporting, however, the VP provided the number of "Applications Filed" even if the survey used a more restrictive definition of what constitutes an "application."20 The VP said he believed he appropriately reported Application statistics, and was never aware of any differences between the CMC term "Applications Filed" and the definition of "application" used by external surveys. As a result of the distinctions between these definitions, CMC likely over-reported the number of applications to certain surveys.21 When "Incomplete" applications are subtracted from the number of "Applications Filed," the resulting differences in CMC's acceptance rate (Acceptance as a proportion of Application) range from zero to two percent from 2004 through 2011. A chart reflecting the College's acceptance rate calculated with and without the "Incomplete" applications from 2004 through 2011 is set out in the Appendix. 6. Acceptance and Enrollment Statistics We address the statistics related to Acceptance and Enrollment together. "Acceptance" statistics refer to the number of freshman students the College accepted through its admission process. "Enrollment" statistics refer to the number of freshman students who enroll at the College. CMC has accurately reported both sets of statistics internally and externally from 2004 to 2011. C. Internal Reporting of Discrepancies Between Admission Statistics In 2011, the Admission Office and IR increased their data-analysis capacity. That change was an important factor in the discovery of the VP's actions. During fall 2011, IR requested that an Admission officer complete an external survey. The VP was away from the College at the time. The Admission officer responded to the survey by calculating the requested statistics, including SAT statistics, from accurate data in CMC's database. That response was provided to IR. Within days after the submission of that survey response, Factbook was published, which contained the inaccurate SAT statistics the VP had reported. Personnel in both the Admission Office and IR noticed discrepancies between the SAT statistics in Factbook and in the survey response. 20 For example, the Common Data Set defines an application as follows: "First-time, first-year, (freshmen) students: Provide the number of degree-seeking, first-time, first-year students who applied, were admitted, and enrolled (fullor part-time) in fall 2011. Include early decision, early action, and students who began studies during summer in this cohort. Applicants should include only those students who fulfilled the requirements for consideration for admission (i.e., who completed actionable applications) and who have been notified of one of the following actions: admission, nonadmission, placement on waiting list, or application withdrawn (by applicant or institution). Admitted applicants should include wait-listed students who were subsequently offered admission." The VP attributes no fault to himself or anyone in leadership for the misreporting of Application statistics. The evidence demonstrates that leadership had no involvement in or knowledge of the misreporting. 21 12 Investigative Report of O'Melveny & Myers LLP The discrepancies were reported to the VP, who provided demonstrably false explanations or did not respond. Concerns regarding the discrepancies were then raised with the College's General Counsel on January 9, 2012. D. The VP Alone Knew of and Engaged in the Misreporting of Admission Statistics The VP told us and the Dean Emeritus that no one at CMC knew he (the VP) was misreporting admission statistics. We investigated whether that assertion was true and concluded that it was. 1. The Role of the Professional Staff in the Admission Office and IR The Admission Office and IR professional staff members whose responsibilities include reporting admission statistics also identified the discrepancies that ultimately uncovered the VP's misconduct. Their roles strongly suggest no knowledge or involvement in the conduct at issue. The interviews we conducted and the documents we reviewed corroborated that inference. We concluded that the professional staff members did not know of or participate in the VP's actions. 2. The Role of the Administrative Staff in the Admission Office Certain members of the administrative staff in the Admission Office mechanically helped the VP prepare the reports containing false SAT and Top 10% statistics. Nevertheless, we concluded, for several reasons, that they did not know that the statistics contained in the reports were false. The Admission Office staff members do not recall seeing any discrepancies in the admission statistics the VP prepared or reported. Even if they had noticed any discrepancies, we believe they would have assumed that the differences existed for good reason. The staff members admired and respected the VP and never questioned his integrity. Indeed, on the day that the College disclosed the VP's conduct, administrative staff members sent the VP email messages expressing genuine sadness and surprise at what he had done. Those same staff members also provided the VP with reports from CMC's database that accurately stated admission data. Although this happened only occasionally, the staff members also provided accurate data reports to other Admission officers when asked. Moreover, unlike the VP (as discussed above), at least one of the administrative staff members in the Admission Office knew how to alter data in the CMC database, but all evidence indicates that no data was altered. 3. The Role of Leadership There is no evidence that any members of CMC's senior staff had any connection whatsoever to the misreporting of admission statistics. With respect to the President, the VP does not claim that she participated in, encouraged, or knew of the misreporting. Our investigation confirmed that she did not. 13 Investigative Report of O'Melveny & Myers LLP 4. The VP's Conduct Remained Undiscovered for Multiple Reasons Because the VP misreported admission statistics for over seven years, we investigated why personnel in leadership, the Admission Office and IR did not discover his activities before fall 2011. We concluded that multiple factors contributed to this outcome and timing. As the senior member of the Admission Office, the VP was ultimately responsible for the accurate reporting of admission statistics. In part because of their trust in the VP, the College had no processes in place to independently verify the VP's admission statistics. By personally calculating and reporting the College's admission statistics, the VP also was able to control that function for many years. The VP reported his inaccurate statistics to IR which incorporated those statistics into the U.S. News survey and Factbook. The Admission Office then relied upon Factbook when responding to other external surveys. The VP controlled both the internal and external reporting of admission statistics in this manner. He also took steps to ensure that his misreporting was not discovered. For example, he divided the process of creating the Five-Year Report containing false SAT statistics between two staff members. He also discarded his detailed calculations of false SAT statistics. The fact that few people in the Admission Office knew how to generate accurate reports from CMC's database facilitated the VP's control over the reporting function. The Admission "survey person" generally asked the VP for admission statistics or looked them up in Factbook because it was easier than generating a data report. Finally, for most of the years we reviewed, the VP reported the same false median SAT statistics he had reported the year before. For those years, CMC's true median SAT statistics were lower, and varied year to year, but the CMC community perceived that SAT statistics for the incoming classes were unchanged. Only in 2008 and again in 2011 did the VP increase the false SAT median statistics, in each case by 10 points (0.6%). By modestly inflating SAT statistics over time, the VP was able to conceal his conduct. E. The VP's Explanations for Falsely Reporting SAT Statistics, and the Role of Leadership 1. The VP's January 24, 2012 Statement As part of his internal investigation, the Dean Emeritus met with and spoke to the VP on several occasions. On January 24, the VP provided an explanation for his misconduct, which is summarized below: For many years, the President had pressured the VP to maintain or increase the SAT scores of the entering classes. In 2004, for the first time, the median SAT score for CMC's entering class reached 1400, which was an important milestone for an elite institution such as CMC. In 2005, however, the median SAT score dropped below 1400. The VP could not bring himself to tell the President about this development. Instead, to avoid disappointing her, the VP began to inflate SAT scores. 14 Investigative Report of O'Melveny & Myers LLP 2. The Pressure Described Was within the Norm Based on our collective experience and judgment, we believe the VP sincerely felt pressure to maintain or increase SAT statistics. We do not believe, however, that the pressure exceeded the norm for an executive-level employee at a high-performing institution, or that it best explains his actions. The College's Institutional Priorities and Departmental Goals Were Reasonable In March 2002, the College adopted a Strategic Plan (the "Plan").22 That Plan sets out the institution's overarching priorities. Continually strengthening the College's already accomplished student body is a key priority reflected in the Plan. The President set annual goals consistent with the College's priorities.23 Regularly included among the President's goals was ensuring the quality of students admitted to the College. Student quality is measured by a range of characteristics, one of which is strong academic skills. SAT scores, class ranking, grade point averages, and difficulty of course work are common benchmarks the College uses to assess a student's academic strength. Each of the College's Vice Presidents, in consultation with the President, also annually set goals derived from the College's priorities. Three times a year, the College's leadership would assess together their progress against the goals set for that year. That process ultimately would yield another set of job performance goals for the following year. Those goals would sometimes be the same, a more aggressive form of the same goal(s), or entirely new. This process equally applied to the former VP. From 2000 through 2011, the former VP set between 13 and 24 annual goals, including median SAT goals for the College's incoming class. His SAT goals included target ranges (e.g., 1380-1420) and specific median figures (e.g., 1420).24 He also identified goals related to the Top 10% statistic. In our experience and judgment, CMC's institutional priorities, goals, and measurements were legitimate and reasonable. That CMC hopes to attract and enroll the most talented students, and measures achievement of that goal against common benchmarks, is to be expected. The former VP's specific goals and the process used to set them, also were unremarkable. Executives are routinely required to set and meet reasonable performance goals. Those interviewed about this topic, including the VP, all stated that the College's applicant pool could readily support the VP's median SAT goals. 22 The Board adopted and approved the College's ten-year Plan in March 2002. The President submitted and the Board received an update to the Plan in April and May 2011, respectively. The Board also approves the President's annual goals. The VP set two median SAT goals during the review period. From 2004-2007, the VP's SAT goal was expressed as a range, i.e., 1380-1420. From 2008-2011, the VP set a specific median SAT goal, i.e., 1420. 23 24 15 Investigative Report of O'Melveny & Myers LLP The Manner in Which the President Held the VP Accountable Was Reasonable The President held each Vice President, including the former VP, accountable for meeting job performance goals. If a Vice President failed to meet a goal or goals, the President discussed that matter with the Vice President to assess why the goal was not achieved, and, at times, would express disappointment during those discussions. Critically absent from this process, however, was any evidence of verbal abuse, economic pressure, or any other form of intimidation that might reasonably be considered coercive or improper. The VP's statements that he enjoyed his job, despite the pressure, and felt secure in his position, also reflect the absence of undue pressure. Even when the VP admitted to falsifying SAT statistics, he said he did not believe he would lose his job. Nor was his compensation tied to meeting SAT goals. Like other CMC executives, the VP's compensation was tied to overall qualitative performance rather than meeting any specific goal or goals. While we believe that the VP sincerely feared disappointing the President if he did not meet his goals, we concluded that his subjective fear cannot reasonably be viewed as the product of undue pressure.25 Like the process of setting goals, the practice of holding executive employees accountable for performance goals is commonplace. The VP's Prior Conduct Contradicts His January 24 Explanation The VP's prior written reports to the President contradict his January 24 statement. In his statement, the VP explained that he began falsifying SAT scores in 2005, when the median SAT score fell below 1400, because he felt unable to disclose that information to the President. A similar drop in CMC's median SAT statistic, however, occurred between 2000 and 2001. In 2000, CMC reported26 a median SAT statistic of 1400. In a memorandum the VP authored the following year, he told the President that the SAT median had dropped below 1400. The VP experienced no discipline or reduction in pay as a result of reporting the decrease in the median statistic to the President. In his January 24 statement, the VP also stated that he began falsifying SAT scores in 2005. That assertion is incorrect. The VP explained that CMC reported an accurate median SAT statistic of 1400 for the first time in 2004. According to our calculations, however, CMC's actual median SAT statistic in 2004 was 1390. Although CMC publicly reported, in 2004, a median SAT statistic of 1400, that information was false.27 We concluded, therefore, that the VP began falsifying SAT statistics starting as early as 2004. In 2008, 2009, and 2010, the VP reported a false SAT median below his stated goal. The VP suffered no disciplinary or economic consequence for reporting median SAT statistics below his goal. 26 27 25 We did not verify SAT statistics prior to 2004. The VP told us that, at the time of his January 24 statement, he believed that CMC was unable to verify SAT data before 2005. The VP's understanding was incorrect. 16 Investigative Report of O'Melveny & Myers LLP 3. The VP's Additional Explanation for His Conduct During his interview, the VP provided further explanation for his decision to falsify SAT statistics. He explained that the President's admission priorities, while achievable, did not match his own. The VP stated that the College's applicant pool could support the goal of maintaining or incrementally increasing SAT scores. He also acknowledged that achieving the SAT goal required some adjustments to the priorities applied in the admission process. Rather than make those adjustments, however, the VP applied his own admission priorities. The VP stated that he knew what was in CMC's best interest when it came to admission decisions. And, although he embraced and achieved some of the President's admission goals, he felt she had too many goals and that some must give way. When the VP's admission decisions did not produce the targeted SAT statistics, he chose to falsify the SAT statistics to conceal this outcome from the President. 4. Rankings Did Not Motivate the VP's Conduct Even though the College sought to improve its standings within various college ranking surveys, the VP stated that improving the College's standings did not motivate his decision to report false SAT statistics. We believe the VP's assertion to be true. In particular, the evidence demonstrates that the VP's disagreement with the President's admission priorities -- not improving the College's rankings -- primarily motivated the VP's decision to report false SAT statistics. IV. Conclusions The conclusions contained in this report are based on the information we gathered in the course of our investigation, and are informed by our collective judgment and experience. We concluded that the VP, acting alone, compiled and reported inaccurate SAT, ACT, class rank and application statistics, starting as early as 2004. The College's leadership did not direct, encourage, or know about the VP's misconduct. The VP stated he falsified SAT statistics because he felt pressure from the President to maintain or increase them. We believe, however, that the pressure he described was within the norm for an executive at a high-performing institution. Finally, we concluded that the VP's disagreement with the President's admission priorities primarily motivated the VP's choice to falsify SAT statistics. 17 Report of LLP APPENDIX 1 8 Investigative Report of O'Melveny & Myers LLP SAT Statistics28 SAT Mean, Median, 25th-75th Percentiles Year 2011 Statistics Mean Median 25th-75th Percentiles Mean Median 25th-75th Percentiles Mean Median 25th-75th Percentiles Mean Median 25th-75th Percentiles Mean Median 25th-75th Percentiles Mean Median 25th-75th Percentiles Mean Median 25th-75th Percentiles Mean Median 25th-75th Percentiles Accurate 1386 1400 1300-1480 1385 1400 1310-1480 1377 1370 1290-1480 1380 1390 1290-1490 1342 1360 1250-1440 1354 1360 1250-1470 1359 1380 1260-1460 1358 1390 1270-1450 Externally Reported29 [1386]30 [1400]31 1300-1480 141032* 1410 1310-1510 1410* 1410 1310-1510 1410* 1410 1310-1510 1400* 1400 1310-1490 1400* 1400 1310-1490 1400* 1400 1310-1490 1400* 1400 1310-1490 2010 2009 2008 2007 2006 2005 2004 In calculating SAT statistics, we considered only applicants' highest test scores used in the admission decision, consistent with the College's admission policies. (See footnote 8). We also note that the corrected SAT statistics the College reported during the week of January 30, 2012, are consistent with our SAT statistics calculations. The column titled "Externally Reported" reflects the statistics the College submitted to U.S. News & World Report, except for 2011. For 2011, we obtained externally reported SAT statistics from the College's Common Data Set survey response where available. 30 29 28 Because the College has not externally reported the SAT mean statistic for 2011, we inserted the accurate mean statistic into that cell. Because the College has not externally reported the SAT median statistic for 2011, we inserted the accurate median statistic into that cell. * From 2004 through 2010, the VP reported his false SAT median statistics, in response to college survey questions seeking SAT mean statistics. For that reason, we inserted the VP's false median statistic into the cell reflecting the externally reported mean statistic for each year from 2004 through 2010. 31 32 19 Investigative Report of O'Melveny & Myers LLP Ranges for SAT Verbal/Critical Reading Year 2011 Ranges 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 Accurate 43.9% 46.1% 10.0% 0% 0% 50.2% 37.3% 12.4% 0% 0% 41.7% 49.0% 9.2% 0% 0% 48.2% 38.6% 13.1% 0% 0% 37.6% 48.2% 14.2% 0% 0% 45.1% 39.3% 14.4% 1.2% 0% 46.1% 39.5% 13.6% 0.8% 0% 46.3% 40.2% 12.2% 1.2% 0% Externally Reported33 55.2% 34.8% 10.0% 0% 0% 51.9% 37.0% 11.1% 0% 0% 52.9% 40.5% 6.4% 0% 0% 52.2% 39.5% 8.3% 0% 0% 51.5% 38.2% 10.3% 0% 0% 50.6% 37.8% 11.6% 0% 0% 2010 2009 2008 2007 2006 2005 2004 50.4% 37.4% 12.2% 0% 0% The column titled "Externally Reported" reflects the statistics the College submitted to U.S. News & World Report, except for 2011 and 2005. For 2011, we obtained reported SAT range statistics from the College's Common Data Set survey response. The College has no records reflecting the SAT range statistics it externally reported for 2005. 33 20 Investigative Report of O'Melveny & Myers LLP Ranges for SAT Math Year 2011 Ranges 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 700-800 600-699 500-599 400-499 -399 Accurate 61.7% 34.3% 3.5% 0.4% 0% 54.8% 39.4% 5.8% 0% 0% 49.5% 45.6% 4.4% 0.5% 0% 54.6% 38.2% 6.8% 0.4% 0% 41.3% 43.6% 14.2% 0.9% 0% 43.2% 45.5% 10.9% 0.4% 0% 52.7% 36.8% 10.5% 0% 0% 52.0% 37.8% 9.3% 0.8% 0% Externally Reported34 61.7% 34.3% 3.5% 0.4% 0% 54.8% 39.4% 5.8% 0% 0% 54.7% 37.2% 8.2% 0% 0% 55.0% 35.9% 9.1% 0% 0% 53.3% 36.9% 9.8% 0% 0% 53.0% 38.1% 8.9% 0% 0% 2010 2009 2008 2007 2006 2005 2004 53.3% 36.6% 10.1% 0% 0% The column titled "Externally Reported" reflects the statistics the College submitted to U.S. News & World Report, except for 2011 and 2005. For 2011, we obtained reported SAT range statistics from the College's Common Data Set survey response. The College has no records reflecting the SAT range statistics it externally reported for 2005. 34 21 Investigative Report of O'Melveny & Myers LLP ACT Statistics - Composite35 ACT Mean and 25th-75th Percentiles Year 2011 2010 2009 2008 2007 2006 2005 2004 th Statistics Mean 25 -75 Percentiles Mean 25th-75th Percentiles Mean 25th-75th Percentiles Mean th th 25 -75 Percentiles Mean th th 25 -75 Percentiles Mean th th 25 -75 Percentiles Mean 25th-75th Percentiles Mean 25th-75th Percentiles th Accurate 31 30-33 31 29-33 31 29-33 30 28-33 30 28-32 30 28-32 31 29-32 30 29-32 Externally Reported36 [31]37 29-32 31 29-33 31 29-33 31 29-33 31 29-33 31 29-33 31 29-33 31 29-33 35 36 In calculating ACT statistics, we considered only applicants' highest test scores. (See footnote 9). The column titled "Externally Reported" reflects the statistics the College submitted to U.S. News & World Report, except for 2011. For 2011, we obtained externally reported ACT statistics from the College's Common Data Set survey response. Because the College has not externally reported the ACT mean statistic for 2011, we inserted the accurate ACT mean statistic into that cell. 37 22 Investigative Report of O'Melveny & Myers LLP ACT Ranges Year 2011 Ranges 30-36 24-29 18-23 -17 30-36 24-29 18-23 -17 30-36 24-29 18-23 -17 30-36 24-29 18-23 -17 30-36 24-29 18-23 -17 30-36 24-29 18-23 -17 30-36 24-29 18-23 -17 30-36 24-29 18-23 -17 Accurate 76.0% 24.0% 0% 0% 74.0% 26.0% 0% 0% 68.4% 31.6% 0% 0% 66.7% 33.3% 0% 0% 62.0% 36.0% 2.0% 0% 55.6% 44.4% 0% 0% 64.3% 35.7% 0% 0% 71.0% 22.6% 6.5% 0% Externally Reported38 73.3% 26.7% 0% 0% 71.6% 27.0% 1.4% 0% 71.4% 28.0% 0.6% 0% 71.3% 28.1% 0.6% 0% 71.1% 28.4% 0.5% 0% 70.8% 28.6% 0.6% 0% 2010 2009 2008 2007 2006 2005 2004 70.5% 28.7% 0.8% 0% The column titled "Externally Reported" reflects the statistics the College submitted to U.S. News & World Report, except for 2011 and 2005. For 2011, we obtained reported ACT statistics from the College's Common Data Set survey response. The College has no records reflecting the ACT range statistics it externally reported for 2005. 38 23 Investigative Report of O'Melveny & Myers LLP Top 10% Statistics Year Statistics Percentage Based on College's New Methodology (Class Rank/Class Size Data from CMC Database)39 71% 26% 72% 38% 67% 34% 67% 36% 67% 38% 65% 34% 74% 37% 71% 36% Externally Reported40 2011 2010 2009 2008 2007 2006 2005 2004 Percent of Students Graduating in Top 10% Percent of Students who Submitted Class Rank Percent of Students Graduating in Top 10% Percent of Students who Submitted Class Rank Percent of Students Graduating in Top 10% Percent of Students who Submitted Class Rank Percent of Students Graduating in Top 10% Percent of Students who Submitted Class Rank Percent of Students Graduating in Top 10% Percent of Students who Submitted Class Rank Percent of Students Graduating in Top 10% Percent of Students who Submitted Class Rank Percent of Students Graduating in Top 10% Percent of Students who Submitted Class Rank Percent of Students Graduating in Top 10% Percent of Students who Submitted Class Rank 85% 26%41 85% 54% 85% 53% 85% 53% 84% 52% 84% 51% 84% 53% 83% 46% 39 The calculations in this column rely on numeric class rank and class size data, if reported on matriculated students' high school transcripts or school reports. That information is contained in CMC's database. The column titled "Externally Reported" reflects the statistics the College submitted to U.S. News & World Report, except for 2011. For 2011, we obtained the reported Top 10% statistic from the College's Common Data Set survey response. In 2011, the College reported the percent of students submitting class rank accurately (i.e., 26%), but misreported the percent of students graduating in the top 10% (i.e., 85%). We believe this reporting anomaly occurred because each statistic came from a different internal source -- an Admission officer provided the accurate statistic (i.e., 26%), and the VP provided the inaccurate statistic (i.e., 85%). 41 40 24 Investigative Report of O'Melveny & Myers LLP Application and Acceptance Rate Statistics Year Number of Applications Excluding Incomplete Applications42 4,412 (-69) 4,184 (-80) 4,196 (-80) 3,870 (-308) 4,055 (-85) 3,528 (-65) 3,588 (-146) 3,319 (-209) Externally Acceptance Rate Externally Reported Reported Excluding Incomplete Acceptance Rate44 Number of Applications Applications43 4,481 4,264 4,276 4,178 4,140 3,593 3,734 3,528 14% 17% 17% 21% 17% 23% 22% 23% 14% 17% 16% 19% 16% 22% 21% 22% 2011 2010 2009 2008 2007 2006 2005 2004 42 The number in parenthesis reflects those applications the College designated as incomplete for that particular year. The number of incomplete applications was found in previously printed reports of data from the College's database. The column titled "Externally Reported Number of Applications" reflects the statistics the College submitted to U.S. News & World Report, except for 2011. For 2011, we obtained the reported Applications statistic from the College's Common Data Set survey response. The column titled "Externally Reported Acceptance Rate" reflects the statistics the College submitted to U.S. News & World Report, except for 2011. For 2011, we obtained the reported Acceptance Rate statistic from the College's Common Data Set survey response. 43 44 25