Methodology

Introduction

The Campaign for College Opportunity took the following steps to ensure the accuracy of the data, including providing campuses with multiple opportunities to share data with us and confirm or correct the data we gathered.

Data for this report were collected from a variety of sources. For the purposes of this analysis, we defined “leadership” as:

CALIFORNIA COMMUNITY COLLEGES CALIFORNIA STATE UNIVERSITY UNIVERSITY OF CALIFORNIA
Campus Level
  • Faculty
  • Academic Senate Members
  • Campus Executives
  • Faculty
  • Academic Senate Members
  • Campus Executives
  • Faculty
  • Academic Senate Members
  • Campus Executives
District Level CCD Trustees DOES NOT APPLY DOES NOT APPLY
System Level
  • Academic Senate for California Community Colleges
  • CCCCO Leadership
  • Board of Governors
  • Academic Senate of the California State University
  • CSU Office of the Chancellor Leadership
  • Board of Trustees
  • University of California Academic Senate
  • UCOP Leadership
  • Board of Regents

Because leadership was defined as multiple roles at both the campus and UC, CSU, and California Community College systems, it was necessary to collect information from individual campuses as well as system and federal databases. Student data is also included as a measure of comparison, with the belief that if leaders are serving their student body, then the leadership bodies should match the structural diversity of the students. The following sections describe definitions of the data as well as the sources and the collection methods.

The populations used for this analysis include:

Term Definition
Students Included all enrollment status undergraduate students
Faculty

Included instruction staff categorized as “tenured,” “tenure track,” or “not on tenure track/no tenure system,” following the language provided by the Integrated Postsecondary Education Data System (IPEDS).

  • Tenured: Professor, Associate Professor
  • Tenure-track: Assistant Professors, Instructors, Lecturers, and other instructional staff who are on the academic ladder.
  • Not on tenure track/No tenure system: no academic rank
Campus Academic Senates Included voting members in the Academic Senate are only counted once, even if they serve on multiple committees
Campus Leadership

Included senior leadership from each campus

  • Community Colleges: President/Superintendent; Vice President and identified members of the President’s cabinet
  • California State Universities: President; Vice President; Provost and identified members of the President’s cabinet
  • University of California: Chancellor; Vice Chancellor and identified members of the Chancellor’s cabinet
Statewide Academic Senate Members who participate in the Academic Senates of the UC, CSU, and the California Community Colleges
System Leadership

Included the UC, CSU and California Community College system senior leadership

  • University of California: UC Office of the President (UCOP) - President; Vice President and members of the President’s cabinet
  • California State Universities: CSU Office of the Chancellor- Chancellor; Vice Chancellor and members of the Chancellor’s cabinet
  • Community Colleges: California Community College Chancellor’s Office (CCCCO) – Chancellor; Deputy Chancellor; Vice Chancellor and members of the chancellor’s cabinet and Consultation Cabinet
Governing Boards
  • Community Colleges: Board of Governors
  • California State Universities: Board of Trustees
  • University of California: Board of Regents

Data Collection

Student and faculty data were collected for the 2016-2017 academic year from the California Community Colleges Chancellor’s Office (CCCCO) Management Information Systems Data Mart, the California State University (CSU) Division of Analytic Studies, and the University of California Office of the President (UCOP) Division of Institutional Research and Academic Planning. Student and faculty data by race and gender are collected annually by the Integrated Postsecondary Education Data System (IPEDS), operated out of the National Center for Education Statistics (NCES), in the U.S. Department of Education.

While the collection of student and faculty data are routinely collected for submission to IPEDS and disaggregated by race and ethnicity and gender, no such data is reported to IPEDS for Academic Senate bodies and campus and UC, CSU, and CA community colleges system leaders. Therefore, academic senate rosters, college and UC, CSU, and CA community colleges system leadership information and organizational charts were collected primarily through publicly available information on institutional or organizational websites for the 2016-2017 academic year.  Even if leaders serve in multiple capacities they were only counted once and only voting members were included in the analysis.

Identification of Race/Ethnicity, Gender, International Status

The student and faculty data were collected from the UC Info Center, CSU Analytic Studies, and CCCCO DataMart and included demographic information such as racial/ethnic background, gender, and international status. Nonresident alien reported populations were excluded from the general analysis and a smaller, more focused analysis was conducted specifically for this population.

The classification we used to determine race/ethnicity was the U.S. Census Bureau’s standards on race and ethnicity :

  • Hispanic or Latino
  • White, Non-Hispanic
    • Middle East
    • North Africa
  • Black or African American
  • American Indian and Alaska Native
  • Asian American
    • Far East
    • Southeast Asia
    • Indian subcontinent
  • Native Hawaiian and Other Pacific Islander
  • Two or more races
  • Some other race or unknown

Due to small sample sizes within American Indian and Alaska Native (AIAN) and Two or More races, we combined these two categories into one group, which labeled “Other.” Similarly, we combined Asian American, Native Hawaiian and Pacific Islanders into one group, called “Asian American Native Hawaiian Pacific Islander (AANHPI).” While we acknowledge the importance of disaggregating the AANHPI data and exploring more in-depth the nuances for the AIAN and multiracial communities, given small sample sizes, we elected to create these combined categories.

We employed a process of racial identification using photos in a multiple-step process:

  1. Membership rosters, along with physical headshots of the corresponding individual were utilized to make a preliminary distinction of the race/ethnicity of the individual as well as their gender.
  2. If a physical photo was not available, the next phase in the pre-screening process was to cross reference the roster with the with professional organization websites (i.e. Academic Senate for California Community Colleges, ASCCC), professional websites (i.e. LinkedIn) as well as social media websites (i.e. Facebook, Instagram, etc.) to attempt to confirm a physical match as well as determine place of birth for international status determination.
  3. If still no physical confirmation could be matched with the roster, then a third phase was put into action to determine the leader’s surname with its country of origin from Ancestry.com to identify a racial/ethnic distinction for that individual. The guidelines on country of origin for racial/ethnic categorization followed the definitions given by the U.S. Census Bureau on current racial/ethnic categorization standards.
  4. Finally, all data collected led to a preliminary determination of racial/ethnic category, gender, and international status for each leader.
  5. Once a preliminary determination was completed, data were sent for verification to each institution or membership body for review (See description below). For instance, we sent academic senate demographic membership data to Academic Senate Presidents and Public Information Officers at each individual campus, requesting he/she verify or update the data we provided.
  6. Once data edits or confirmations were received from the institution, the preliminary determinations were re-evaluated and completed for final data analysis.

Subsequently, there were two formal requests for data verification from each leadership body and a varied number of interactions via email and phone throughout 2016-17. In January 2017, we emailed the data to the Academic Senate President, Public Information Officer and/or President of each campus to review and verify our findings on academic senate membership. In December 2017, we provided Academic Senate Presidents and Public Information Officers a final opportunity to verify the data before publication of the report. Furthermore, we also emailed our findings on campus leadership demographics to public information officers, chancellor and president’s offices, and executive support staff at every campus. With the community college district trustees, we contacted district public information officers and district executives to verify community college district trustee information as well. Finally, we emailed our UC, CSU, and CA community college system leadership findings to the California Community College Chancellor’s Office (CCCCO), the California State University Office of the Chancellor, and the University of California Office of the President (UCOP) for verification.

For a number of campuses, the available information on their websites was incomplete, outdated, or missing. For these colleges, we followed up with formal public information requests from the Public Information Officers. The campuses that were contacted by this additional method are included below:

  • Los Angeles Harbor College
  • Reedley College
  • Los Angeles City College
  • Las Positas College
  • Barstow College

While some campuses and districts responded positively to our request and agreed to verity or provide corrected data, many campuses objected to our verification requests and outright refused to provide any information. Other campuses claimed that they did not collect that type of data, therefore, would not be able to help us. Furthermore, some campuses expressed concern over the ways in which we collected data, yet did not provide any suggestions on how to improve the process or secure the information in the absence of having this data publicly available.

The number of institutions that responded to our request are shown in Table 1, along with the corresponding response rates.

TABLE 1. RESPONSES TO DATA VERIFICATION REQUESTS

Campus Academic Senate Campus LeadershipDistrict Trustees* System
Community Colleges
  • Number: 39
  • Percent: 34%
  • Number: 36
  • Percent: 32%
  • Number: 29
  • Percent: 40%
  • Number: 0
  • Percent: 0%
CSUs
  • Number: 9
  • Percent: 39%
  • Number: 6
  • Percent: 26%
  • Number: NA
  • Percent: NA
  • Number: 0
  • Percent: 0%
UCs
  • Number: 4
  • Percent:44%
  • Number: 5
  • Percent: 56%
  • Number: NA
  • Percent: NA
  • Number: 1
  • Percent: 100%

*Community Colleges Only

Community College n=114, CSU n=23, UC n=9, CCD=72

System Academic Senate, System Leadership, Governance n=1 (1 email sent to each system for all three leadership bodies)

The number of institutions that examined the data and confirmed the findings, or provided updated numbers that accurately reflect their leadership bodies is reflected in Table 2.

TABLE 2. DATA VERIFICATION OR CORRECTION

Campus Academic Senate Campus LeadershipDistrict Trustees* System
Community Colleges
  • Number: 23
  • Percent: 20%
  • Number: 33
  • Percent: 29%
  • Number: 27
  • Percent: 38%
  • Number: 0
  • Percent: 0%
CSUs
  • Number: 3
  • Percent: 13%
  • Number: 5
  • Percent: 22%
  • Number: NA
  • Percent: NA
  • Number: 0
  • Percent: 0%
UCs
  • Number: 1
  • Percent:11%
  • Number: 3
  • Percent: 33%
  • Number: NA
  • Percent: NA
  • Number: 0
  • Percent: 0%

*Community Colleges Only

Community College n=114, CSU n=23, UC n=9, CCD=72

System Academic Senate, System Leadership, Governance n=1 (1 email sent to each system for all three leadership bodies)

Finally, we conducted a data audit to measure and verify the accuracy of our estimates. We determined the accuracy by comparing our estimations to the verified or corrected data we received from responding campuses and the UC, CSU and CA community college systems. This is the average percent correct for each leadership body. For example, for the community college academic senates, we correctly estimated the racial identifications at an average of 81%. There were very few errors or corrections to the data from the responding leadership bodies, therefore, our accuracy across all estimations was 80% or higher. The average accuracy of our racial identification estimations is provided in Table 3.

TABLE 3. AVERAGE OF ACCURACY OF RACIAL IDENTIFICATION

Campus Academic Senate Campus LeadershipDistrict Trustees* System
Community Colleges 81% 81% 93% NA
CSUs 85% 88% NA NA
UCs NA 95% NA NA

*Community Colleges Only

Community College n=114, CSU n=23, UC n=9, CCD=72

System Academic Senate, System Leadership, Governance n=1 (1 email sent to each system for all three leadership bodies)

For a more detailed list of who responded and who did not, please refer to Appendix C in the full report where each college is summarized in detail.

Unlike student and faculty data, there are no publicly available data disaggregated by race or gender for campus academic senates, campus leadership, systemwide academic senates, system leadership, community college boards of trustees, nor governing bodies. Therefore, we had to employ a different method to identify race/ethnicity, gender, and international status for these particular groups. Using campus-supported biographies, professional organization membership summaries, social media, and professional websites, we employed a process of racial identification using this information, photos and all other available information.

We followed previous research across the social sciences and education that uses visual data (such as photographs and images) to classify people into racial categories, called photo elicitation and visual inspection . Specifically, social psychological research has explored the complexity of racial and gender identity not only from an individual estimation of self, but also from others’ perceptions of racial and gender identities . Through this body of research, social psychology scholars have learned about the importance of perceived categorization on racial and gender identity. Furthermore, one of the most highly regarded education journals recently published an article about racial and gender representation among higher education faculty and the authors used the same methods as we employed for this study. This study empirically confirmed the accuracy of their racial and gender designations, which validates our method as well. While externally applying race/ethnic and gender categories to someone can be problematic, enrollment for the Medicare database, for instance, shows that for African American and White enrollees, the accuracy of categorization is very high . Additionally, the research has shown that people of one race/ethnicity are more accurate at correctly identifying same-race participants . This means that through photo elicitation and visual inspection, Latinx individuals, for example, are much better at identifying other Latinx individuals. This research supports the accuracy of our estimations given the racial breakdown of the individuals who conducted the identification:

  • 2 White
  • 1 Asian American
  • 1 Latinx
  • 1 Multiracial

Data collection would have been strengthened if colleges collected and transparently reported disaggregated data by race on their campus leaders and academic senates and if had the colleges and universities been more cooperative in confirming or correcting our analysis as requested at least once to every single campus. Specifically, over the course of a year, we reached out at least once to campus academic senate presidents, public information officers, chief communication officers, directors of public relations, systemwide academic senates, and system central offices to provide them with the opportunity to correct or verify their data. Future analyses would greatly benefit from increased institutional verification and/or an integrated data system similar to the ones in Institutional Research offices that collect information on students.

Expert Feedback

We convened experts in the field of education to review our methodology and to discuss preliminary findings. The roundtable consisted of 10 faculty members, 1 campus senior leader, 1 system leader, 2 student representatives, and 3 public policy consultants from the UC, CSU, and CA community college systems to discuss the research and the framing of the analysis. We wanted to highlight the importance of diverse perspectives both in the classroom where there is more direct contact with students, and at leadership levels where critical decision-making about hiring, curricula, campus climate, and student support services are made that have impacts on students in concrete and intangible ways.

Preliminary findings were presented as well as a detailed explanation of the methods used to collect and analyze the data. One suggestion for improvement for the final report included the importance of displaying the findings for all racial/ethnic groups instead of aggregating all leaders of color into one category. Another suggestion for improvement was to expand the senior leadership data to not only include the most senior leaders on campus (i.e., President, Provost, etc.), but to also include the cabinet members as well. Based on their feedback, adjustments, additions, and edits were made to strengthen the validity and reliability of the data collection, thus bolstering the accuracy of the findings.

A draft of the report was sent to experts in the field to review and provide feedback and recommendations on how to strengthen the report. These reviewers included three tenured professors, one community college chancellor, two former college presidents, one non-profit policy director, and one higher education policy expert and entrepreneur. Reviewers were given two weeks to read, edit, and comment on the entire report. We incorporated their suggestions into the final draft.

Interviews

To inform the quantitative data collected on the composition of leadership bodies at the three public systems, interviews were conducted with students, faculty, and senior leaders from the community colleges and universities. Interviews lasting 45-60 minutes were conducted to provide insight into leadership, power, and practices on their campuses. The data from these interviews led to an in-depth understanding of leadership diversity among practitioners and those directly involved with campus diversity efforts. In addition, we conducted a brief online survey where respondents were asked a series of multiple-choice questions pertaining to leadership and an open-ended question about their perspectives about the importance (or not) of having diverse faculty and leaders in higher education.

Limitations

As with any research project, there are a number of limitations in the data and methods. One limitation was the way in which the race/ethnicity, gender, and international status was identified for campus and system leaders (with the exception of students and faculty). Given that there is no comprehensive database that includes this information, estimations were made based on publicly available data. Another limitation is the option of “Other” on some of the data sets, yet absent on others. This ambiguous racial category was difficult to operationalize across data sets and therefore, may not be consistent in the findings. Similarly, multiracial individuals were difficult to categorize given the variety of ways that multiracial individuals identify and are perceived. Therefore, some of the categorizations for multiracial individuals may be inconsistent.

This is the best possible approach to collecting the data necessary for this research project. We welcomed and provided numerous opportunities for colleges to confirm the data we collected. More importantly, we are confident with the findings that our campuses are not diverse enough, do not reflect the student bodies served, and that this hinders student success. Our colleges and universities need to be more aggressive about inclusivity at all levels of the campus and the UC, CSU, and CA community college systems.

Conclusion

A state as diverse as California and with a diverse student body in college deserves to have demographic data on academic senate and college leaders readily available and accessible. This would ensure that our public colleges and universities have the data to determine if they are adequately hiring and including diverse Californians in their leadership bodies. In doing so, colleges will be able to identify where they may be excluding and leaving critical members of our community out of these positions of power and influence. The state of California should require that these data be collected and made publicly available. If it were, all of California would be able to identify institutions that should be held accountable for their lack of diversity as well as recognize colleges that are more inclusive.