Identification and assessment of community college peer institution selection systems.

Introduction

Institutional effectiveness criteria, accountability measures, and performance standards drive the determination of methods for answering the question "How well are we doing in higher education in meeting the needs of the country, the state, private business, and the individual?" The President, the United States Congress, the governors of several states, state legislatures, and regional accrediting agencies, sometimes collectively and other times separately, champion a wide variety of plans and approaches for assessing and improving higher education. At the end of this domino chain rests the institution, and therein resides a large part of the responsibility to respond to the requirements of federal government regulations, enacted federal and state legislation, governor-directed initiatives, regional accreditation standards, and state higher education strategic and comprehensive plans. As part of higher education's attempt to provide cogent and meaningful answers, state higher education systems, institutional governance boards, institutional presidents, and recognized leaders in the field place heavy reliance on comparative analysis between, as well as among, higher education institutions. A very common basis for conducting institutional comparisons involves the use of institutional peer groups.

One of the first formal reports addressing the determination and use of peer states and peer institutions was done by Curry (1972) while at the National Center for Higher Education Management Systems (NCHEMS) in Boulder, Colorado. His focus was Washington state colleges and universities. Over the next 10 years, NCHEMS' membership, activities, and products grew, each becoming very popular and prestigious and having substantial lasting impact on higher education management, especially in the development and adaptation of quantitative-based, data-driven systems.

At the same time, the Association for Institutional Research (AIR) and its membership, primarily four-year institution directors of institutional research, developed into well-established and prominent vehicles for encouraging, promoting, collecting, publishing, presenting, and sharing quantitative research in higher education. In 1974, AIR sponsored a new journal, New Directions for Institutional Research, with each issue focusing on a distinctive topic.

Preceding New Directions by one year, the journal Research in Higher Education (RHE) began in 1973. Both journals, concomitant with the work of NCHEMS and AIR, directed attention and gave credence to applied quantitative research in higher education.

Peer institution selection systems became a higher education research interest and prompted Terenzini, Hartmark, Lorang, and Shirley's (1980) often cited work, A Conceptual and Methodological Approach to the Identification of Peer Institutions. It was published in RHE. Building on Terenzini's work, others, including Teeter (1983), Teeter (1987), Christal (1984), Christal (1987), Brinkman (1987), Berthold (1996), Szelest (1996), Zhao (1997), and Weeks (2000), published research studies that refined, expanded, validated, and replicated earlier studies and raised new issues with previous peer institution selection systems, methods, and applications in higher education. Most of these studies appeared in RHE, New Directions, or the proceedings of AIR national and regional forums, often made part of the Educational Resources and Information Center (ERIC). Although aspects of these published research works have meaning for the community college, they were, nevertheless, directed primarily toward four-year colleges and research universities.

Problem

A significant, yet preliminary, literature review of NDIR, RHE, ERIC, New Directions for Higher Education (NDHE), and other established databases suggested that either there has been very little formal effort in this area covering community colleges and community college systems, or that most of the work done remains unpublished. In instances where work has been done, it is shared only within the institution, state system, or regional environment from which specific needs for establishing peer groups arose.

Accordingly, the problem is three-fold. First, there is the need to identify peer institution selection methods for community colleges at both the institutional and state system levels. Second, there is a collateral requirement to assess--that is, describe, analyze, evaluate, and classify--those existing methods a) that have been published in academic journals, found in ERIC, contained in dissertation theses or made available through institution and state system official websites; b) that have a community college focus; and, c) that can be applied in community college settings. Third, if the research indicates that community college specific peer institution selection models are either not readily available, nor comprehensive or extensively applied, then it is imperative to direct research toward the development, construction, and analysis of institution peer group selection models in community college settings.

Importance

The primary uses of institutional peer groups seem to fall into three functional categories: finances, programs, and planning. Terenzini (1980, p. 348) identified planning, resource allocation, and performance measurement. Rawson (1983, p. 299) listed resource allocation and resource acquisition. Weeks (2000, p.1) spoke of funding needs and program quality as well as budget comparisons and performance indicators (2000, p. 2). Ingram (1995, p. 4) offered budget and program review. Zhao & Dean (1997, p. 6) referred to academic strength and financial health. Szelest (1996, p. 125) cited use in a system-wide planning process and also in a resource allocation review of campus funding formulae. Berthod (1996, p.3) noted requests for inter-institutional comparative data and management's need for such data. Terenzini (1980) stated the general uses "to classify higher educational institutions for descriptive, comparative, and analytical purposes" (p. 347).

Common terminology associated with the need for forming peer groups include performance-based funding, accountability, institutional performance, trend analysis, comprehensive and strategic planning, and institutional relative standing. Additionally, Teeter (1983), Teeter(1984), Terenzini (1980), and Weeks (2000) highlight the importance of melding political, academic, and analytical considerations and methods for developing peer groups and for reaching agreement on their final make-up.

With a strong foundation, dating to 1950, and considerable documented research and experience to draw upon, four-year colleges and research universities have a mature environment in which to approach issues surrounding the use, development, constituency, application, and acceptance of peer groups. Community colleges do not appear to have a comparable effort in the development and use of peer institution comparisons.

Expanded Literature Search

Conceived in four parts, the literature search, addressing the first two problems which are the focus of this paper, initially undertook searches of two well-known academic electronically accessible databases, ERIC (E*Subscribe for full-text availability) and Dissertation Abstracts (DA/ DAI). Then, attention turned to a search of published academic research using on-line electronic access to databases such as FirstSearch and WilsonSelect, especially for full-text on-line searches. After thoroughly searching the electronic databases for both published and unpublished research, substantial effort was expended on state higher education and community college boards and state system official websites, of which several provided a website search capability. Using the official websites of the Education Commission of the States (ECS), the State Higher Education Executive Officers (SHEEO), the American Association of Community Colleges (AACC), AIR, NCHEMS, and the National Center for Educational Statistics (NCES), searches were made primarily of comprehensive state plans, legislation, accountability reports, performance reports, meeting agendas and minutes, and data analysis systems as well as imbedded links often provided in the documents. Last, a general world-wide web search was performed to identify individual institution websites and other related websites indicating peer institutions and were not discovered in the first three approaches. Also, during the conduct of this research, the writer became aware of a session of the American Educational Research Association's Annual Meeting (Seattle, Washington, April 12, 2001) that addressed community college classification systems. Organized by the Carnegie Foundation, Dr. Alex McCormick, senior associate, led a symposium (Session 25.48) on community college classification systems. Follow-up communications with Dr. McCormick led to the identification of several resources cited in this paper and listed in the references.

Literature Review

The expanded literature search yielded information that consistently fell into three major characteristics of peer institution selection systems. Accordingly, the literature review is presented under three headings: purposes and requirements for institution peer groups' use in community colleges; identification of the existence of peer selection systems in community college settings; and, comparison groups (types, selection methods, and variables).

Peer institutions: purposes and requirements

The expanded literature search was very coincident with the initial search, providing most of the same reasons, purposes, and impetuses for developing peer institution selection systems. Budgeting, resource request and allocation procedures, and state financial support funding programs were consistently mentioned as the motives driving state systems and colleges to establish sets of peer institutions (Nebraska, 2000; Wyoming, 1999; Colorado, 2000: West Virginia, 2001; & New Jersey, 2000). Henry (2000) reports that at least eight states now use a comparator-based funding model. This approach requires the formation of comparators (peer institutions) to be used to determine a budget base request amount for each institution. Developed by NCHEMS in support of a project with the South Dakota Board of Regents (1999), the Wyoming Community College Commission subsequently contracted for NCHEMS' assistance to replace its funding formula with the comparator base budget approach.

Colorado's (2000) performance-based funding process uses 10 indicators with incentives, some that are for all postsecondary education institutions and others specifically for two-year colleges. For example, establishment and subsequent measurement of the benchmark for a 3-year graduation rate for each Colorado community college will be based upon national comparative institutional data taken from comparison institutional groups (Colorado, 2000). For each unique community college indicator, the benchmark will be derived from a national comparison group.

Accountability standards, performance indicators, and benchmarks are the other primary requirements placed upon community colleges, resulting in the need to develop a set of institutional peers (New Jersey, 2000; West Virginia, 2001; Ohio, 2000; Georgia, 2000; & Connecticut, 2000). The Ohio Board of Regents, responding to a request from the governor, identified performance goal areas, some that included benchmarking with peer groups (March 2, 2000) and others referred to as benchmarking indicators with peer institutions (June 8, 2000). Nebraska is required by a 1991 (Nebraska, 2000) legislative statute to develop peer institutions that will be used primarily for budget and program review, each requiring the establishment of standards.

Peer selection systems in community college settings

From the literature search, only one report, generated by a community college, emerged and related that institution's effort to determine a set of possible peer institutions. El Paso Community College (Soteriou, 1994) set out to establish a "pool" of possible peer institutions. El Paso in its attempt to create a set of peer institutions for comparative analysis found that their unique variables of Hispanic students and community could not be matched. Out of 22 institutions initially selected using eight variables, only one college proved to be a close match. All other instances of peer institutions in community colleges were found mainly in state comprehensive plans, budget and funding processes, accountability reports, performance indicator reports, legislation and special state system reports.

West Virginia (2001) indicated that its colleges and universities would be compared with their Southern Regional Education Board peers. Connecticut (1998) suggested that its institutions of higher education should establish formal liaisons among themselves and with all peer institutions across New England. New Jersey (2000) colleges and universities were encouraged to identify their strongest programs and make them competitive with peer institutions that are among the best in the region, the nation, or the world. Illinois (2000) annually performs a statewide salary survey wherein the data is compiled into seven peer groups structured in terms of enrollments, geography, and financial data. NACUBO has produced comparative financial statistics for two-year colleges since the late 1980s (Meeker, 1997). Each of these reports is presented based upon peer groups of national, multicollege districts, and single-college districts, classified into five enrollment subcategories. Both Georgia and Nebraska have developed, published, and implemented peer institutions for each of their community colleges. Nebraska's approach was extensive but based more on a descriptive comparative analysis given a predetermined variable list. Based upon sophisticated multivariate statistical procedures, including factor analysis, Georgia (2000) also established and published a set of peer institutions for each of its two-year colleges.

Arnold (1997) used NCES' Beginning Postsecondary Students Longitudinal Study (BPS: 90/94) to establish peer groups for California community colleges. Her primary intent was to create peer groups for use in comparing and evaluating "institutional-level student characteristics" (p. 4). El Paso was very interested in Hispanic students and community variables. Because California community college enrollments now have a majority enrollment of students of color, Arnold (1997) sought to determine national peer groups in consideration of levels of racial and ethnic diversity. Rather than developing peer groups for individual colleges, Arnold attempted to create several general peer groups of all two-year colleges to represent the range of California communtity colleges (Arnold, p. 4).

Comparison groups

The types of comparison groups, the methods for determining comparison groups, and the variables selected for use in establishing comparison groups are well established in the academic literature, especially for four-year colleges and major research universities. In the search for community colleges, the literature is representative of the broad range in general, but published research is very sparse.

Types

Previous research has identified several peer types that have been sufficiently operationally defined to constitute acceptance as standard definitions in peer group research. Brinkman and Krakower (1983) provided four peer group types: competitor, peer, aspiration, and jurisdictional. Later, Brinkman and Teeter (1987) folded jurisdictional into a broader category they named "predetermined." These four categories, for the most part, have remained constant across time. In 1997, the central office staff of the Oregon University System was asked to build a set of peer comparators (Weeks, 2000, p. 3). Providing caution for the potential of peer group proliferation at the institutional level, Whiteley and Stage (1987) described "unit-by-unit peers" and "issue-by-issue" peers.

Three distinct comparison groups were identified for community colleges in the current literature search: peer institutions and peer groups, comparator institutions, and benchmarking institutions. New Jersey, Connecticut, West Virginia, Nebraska, California, and El Paso Community College consistently refer to peer groups and institutional peer groups or peers. Ohio and Colorado use or refer to benchmarking institutions or peer institutions used to benchmark. Comparators and comparator institutions are used by South Dakota and Wyoming. Georgia refers to comparator groups and peer institutions.

It appears that these types are, to a degree, associated with the intended use. For example, the term "comparators" is used almost exclusively with budgeting, especially base budget level funding models. When performance budgeting and accountability are the primary indicated uses, benchmarking institutions appear most often. For general comparative analysis, peer groups and peer institutions are the most frequent choices.

Methods

Again, there is considerable research documenting the development, application, and implementation of several commonly used peer institution selection systems. One system aspect that received significant attention focuses upon selection methodology. Brinkman and Teeter (1987) captured the work initiated by Terenzini, et. al. (1980) and further developed by Teeter & Christal (1984) in a typology of procedures for developing peer groups (See Figure 1., p. 11). Subsequent research by Prather and Carlson (1991), Ingram (1995), Zhao and Dean (1997) and Weeks (2000) fit well within the typology. The Brinkman and Teeter classification provides a continuum (typology) from highly statistical, using multivariate statistical procedures of cluster analysis, factor analysis and discriminant function analysis to judgmental, using the panel review approach.

[FIGURE 1 OMITTED]

Panel

This is a true qualitative approach to peer institution selection. Its primary basis is non-statistical, without sophisticated algorithms. Relying on informed judgment, consensus results from the knowledge of the participants in the selection process. Hence, peer institution selection might derive solely from academic reputation whereas other peer institution sets could be directly attributable to membership in a conference or regional association. Simplicity of the panel approach constitutes its attractiveness for and facility of use in establishing peer institution groups.

Threshold methods

This procedure involves discriminating between institutions based upon nominal and interval level institutional characteristics. Decision variables usually include philosophy (open door and hands-on), mission (college parallel, technology, community service), control (public and private) and classification (Carnegie, NCES, NCHEMS and AAUP). An institution's possession or nonpossession of a specified attribute contributes to its inclusion in or exclusion from the peer group; hence, those are dichotomous variables without magnitude. On the other hand, ranges of values for selected variables can have interval level properties. However, since some of the intervals created for selection represent classifications such as small (<1000 students), medium (1000-5000 students), and large (> 5000 students), their metric is ordinal, especially since the intervals often are not equal.

In application, this method is objective dependent. First, to construct classifications of institutions, institutional assignment to particular classes proceeds according to the distribution of the combination of preselected variables. The objective could be to narrow the population of potential peer group institutions. Second, the purpose might be to identify institutions in relation to a target institution. The process centers around selecting those institutions with values of the criterion variables closest to those of the target institution.

The procedure begins with using the nominal level variables in an initial screening process to generate a potential peer institution set. Next, the target institution makes three decisions: the relevance of each measure; the importance (weight) of each measure; and, the acceptable ranges for each measure. Then, each institution receives a rank, based upon a weighting assignment, that considers the number of ranges missed. Each miss on an important variable moves the institution to a ranking farther from the target institution. From the list of institutions arranged in rank order, the target institution makes the final selection of an institution peer group.

There are three limitations associated with this approach. First, inclusion levels and ranges of values for the variables are arbitrary. Second, this method is affected by the number of variables selected and the extent of inclusion levels defined. As the number of each increases, so does the likelihood of experiencing very small cell frequencies in the data matrix. And third, there is the problem of selecting valid criteria for the variables and then attempting to reconcile the inclusion of an institution barely above a cutoff score and the exclusion of an institution slightly below the cutoff score.

Hybrid approach

While the threshold approach uses data and judgments, the hybrid approach combines data, statistics, and judgment (Zhao & Dean, 1997). The hybrid (Brinkman & Teeter, 1987) approach departs from the threshold method by using statistical algorithms to normalize raw data values into z-scores. Differences between potential peer institutions and the target institution z-scores form comparison scores which, once put into standard form, are weighted and summed to produce similarity scores (Brinkman & Teeter, 1987). This statistical process leads to the production of a rank-ordered list for final peer institution selection which necessarily involves judgment.

Cluster analysis

Statistically, cluster analysis is an umbrella for several sophisticated statistical algorithms used to identify groups possessing similar attributes. Specifically, the application of cluster analysis is objective in producing similar groups. It attempts to insure that institutions in a given cluster are more similar on chosen attributes than in any other cluster (Brinkman & Teeter, 1987). Cluster analysis is computer dependent and data driven. The statistical complexity of this approach is suggested by Della Mea (1989) when she describes the three common clustering algorithms used in the social sciences: hierarchical agglomerative, iterative partitioning, and factor analytic (p.30). Factor analysis and discriminant function analysis tend to accompany cluster analysis applications in higher education peer institution research.

Terenzini (1980) and Korb (1982) used factor analysis and discriminant function analysis in developing institution peer groups. In each study, factor analysis, used for data reduction, and discriminant function analysis, used for goodness of fit and likelihood of cluster membership, permitted a validity assessment of the institution peer clusters produced from the cluster analysis.

And again, the search for community college peer institution selection systems discovered methodologies consistent and coincident with the Brinkman and Teeter typology. Georgia used a sophisticated statistical approach, one that applied factor analysis, computation of distance scores, and ranking peers in terms of closeness to each Georgia community college. The NCHEMS model, a threshold approach (Brinkman & Teeter, 1987) based on significant comparative analysis, was used by both Nebraska and the Wyoming Community College Commission (1999). Similarly, the Colorado approach is primarily descriptive and comparative with a certain amount of judgment built into the system. El Paso Community College's approach also was comparative and descriptive using eight preselected variables. Both New Jersey and Illinois used preselected variables for establishing community college peer groups. While New Jersey's comparisons are to national peers, Illinois constrained its peer groups to in-state community colleges.

Variables

Once purpose has been decided, variables can be identified for determining peer institution selection. Choosing variables can proceed from using statistical analysis procedures to reliance on judgment (Teeter & Christal, 1984; Prather & Carlson, 1991; Zhao & Dean, 1997; Szelest, 1996; Berthold, 1996; Glover & Mills, 1989). Also, variable determination and selection can occur in stages such as initial selection for compatibility, statistical selection for comparability, and final selection for determining peer group constituency.

From the peer institution-related research, variables used can be placed into major categories. A suggested taxonomy for community college variables would include institution level, size, complexity, quality and finance. Additionally, and very important to the type of analysis desired, it is necessary to classify variables by type and level: qualitative (nominal and ordinal level) and quantitative (interval and ratio level).

For a taxonomy to have meaning and application, it must be comprised of aggregation levels and inclusion rules. The taxonomy proposed here (see Figure 1) consists of five levels: type, variable, measure, data element, and data specificity. It is also very important to note that Figure 1 is representative of what this researcher feels are key community college research variables. However, the taxonomy, as presented, is a general conceptualization and should not be taken as all inclusive. Additionally, Figure 1 details three of the taxonomy's five levels. The following examples are intended to provide illustrations that clarify application of the taxonomy.

For example, tuition resides in the taxonomy accordingly: type = finance; variable = revenue; measure = tuition; data element = cost/ credit hour; and, data specificity = $ xxx.xx (ratio level). Or, tuition might be more meaningful when the data element is annual tuition costs. The taxonomy readily accommodates both descriptions for tuition. Similarly, ethnicity occupies a taxonomy location as follows: type = complexity; variable = diversity; measure = ethnicity; data element = IPEDS racial reporting categories; data specificity = number or percent of students of color (ordinal level). For academic program mix, the taxonomical hierarchy proceeds in the following manner: type = complexity; variable = instruction; measure = academic program mix; data element = proportion of transfer and occupational or technical programs to total instructional programs; data specificity = number (ordinal) of instructional programs by Classification of Instructional Programs (CIP) Handbook, number of transfer programs and number of occupational or technical programs. Reference to a local public two-year community college is more specifically described in the taxonomy as: type = control; variable = P-20 system; measure = 2-yr public community college; data element = district board of trustees; and, data specificity = governor appointed (nominal). Or, the local board of trustees might be elected. Again, the taxonomy provides for that distinction. Given the large number of variables found in national, state, and institutional databases, this paper provides an application of the taxonomy to variables used in peer institution selection methods for community college peer groups identified in the current literature search (see Table 1). Scanning the table, one observes the wide diversity, hence subjectivity, in variables identified by several states for use in community college peer group selection systems.

Georgia, Nebraska, and Wyoming peer institution selection systems proved to be the most extensive in the variable selection process for community colleges. El Paso Community College, New Jersey, and Illinois used relatively small sets of preselected variables for peer institution determination, as did NACUBO. West Virginia, Colorado, and Ohio, given their stated intentions and directions, most likely will use comprehensive variable selection processes for determining community college peer institution groups. The California variable selection process was extensive, yet targeted toward racial and ethnic diversity of students.

Discussion

This paper provides partial answers to several questions posited about community college peer institution selection systems. First, what are the impetuses for conducting institutional comparative analysis? Budgeting, funding formulas, resource request and allocation procedures, accountability standards, performance indicators and benchmarks are the most commonly stated purposes. Second, who has developed community college institutional peer groups? From the research and sources utilized, several state systems (Georgia, California, Nebraska, New Jersey, West Virginia, Wyoming, Illinois, and Colorado) have formally determined various types of community college peer groups. El Paso Community College reported developing a set of institutional peers. No other single community college occurrences were found. Third, what constitutes a peer group? Are there different types? In addition to the four traditional types of peer groups commonly found in peer group research--aspiration, competitor, peer, and predetermined--this paper identified comparators and comparator groups, benchmarking institutions and institutions used for benchmarking. Fourth, how is peer group composition determined? What are the methods used and variables included? Community college peer institution selection systems are consistent with Brinkman and Teeter' s typology with methods ranging from complex multivariate statistical procedures to panel review. Variables were found to fall into seven suggested categories: mission, control, classification or membership, size, complexity, quality, and finance. Fifth, to what extent are peer groups present in community college settings? Response to the second question demonstrates that a limited body of research exists in the study of community college peer institution selection systems.

Implications

For states where community college peer institution selection systems exist or are being established, the process originates at the state level and is either performed, controlled, or coordinated at that level. With strong movements in budgeting, performance-based and base level, accountability standards, and performance indicators, it appears very likely that most states will become involved in developing and establishing community college institutional peer group selection systems.

Two primary concerns surface when contemplating future community college peer group research. First, total reliance on the long history of four-year college and university peer research would be inappropriate. The validity and reliability associated with key institutional variables in the four-year college and university levels have been established over a 50-year period and by using several different approaches. This is new research for community colleges.

Second, four-year colleges and universities have benefitted from classification systems that support national level data collection. For community colleges, these classification systems place all two-year institutions into one group. The absence of subcategories for different types of community colleges hinders historical and longitudinal research. Presently, characteristics that speak to community college uniqueness are not readily available in the national databases. Therefore, there is an urgency to determining the role of classification systems to further describe the diversity of community colleges.

Future Research

It appears that there are two very large efforts necessary to enhance and support community college peer group research. First, the concern with the perceived inadequacy of current institutional classifications systems must be resolved expeditiously. To this end, the Carnegie Foundation and the National Center for Educational Statistics have active projects directed toward an expansion of community college subcategories in the classifications systems. Also, the Carnegie Foundation organized and led a symposium on community college classification systems at the 2001 Annual Meeting of the American Educational Research Association in Seattle, Washington, April 12, 2001. Katsinas (1993,1996), a participant in the symposium, discusses a full range of issues about obstacles for community college research created by the various classification systems.

Second, there is a very strong need for objective research into the functioning of community college institutional peer selection systems. The focus should be directed toward appropriate and effective methodologies, hopefully superimposing or aligning methods with types of desired peer groups and types of community colleges. Similarly, extensive research should be conducted to identify those unique community college variables.

Table 1

Community college variables and measures identified and
used in establishing peer institution groups

                                         Georgia

Type        Variable

Sectoring/
Initial     Mission                      Mission
selection

            Control                      Public

            Classification               Carnegie two-
                                         year

Size

            Enrollments                  Headcount

                                         % PT

                                         FTE

            Staff                        Total staff
                                         FT total

                                           FT faculty

                                           FT
                                           non faculty

                                         PT total

                                           PT faculty

                                           PT
                                           nonfaculty

                                         % FT staff

                                         % FT faculty

Complexity  Academic programs            #6-digit CIP

                 Transfer

              Technical/
              occupational

              Certificate

              College
              preparatory

              General
              education

            Diversity

               Ethnicity

               Gender

               Internationality

               First-time first-
               year students

            Demography

               Geography/
               Location

                Urban

                Metropolitan

                Suburban

                Rural

                Region

              Population                 1-9 rating

                State

                Service
                area

                Institutional
                structure

Finance     E&G expenditures

            E&G revenues

            Budgeted $

            Restricted budget

            Expenditures

            Student costs

Quality     Selectivity

                 Applications

                 Admissions

            Degrees awarded              Total degrees

                                         # of < 1 yr

                                         1< # of <2

                                         # Associates

                                         # by 2-digit CIP

                                           Technology

                                           Health related

                                           Liberal arts

                                           Social Sciences

                                           Business

                                           Computer Sci

                                           Workforce Tm

                                         % Degrees awar.

                                         # of <1 yr

                                         1<# of <2

                                         # Associates

Quality     Degrees awarded              # by 2-digit CIP

                                           Technology

                                           Health related

                                           Liberal arts

                                           Social sci.

                                           Business

                                           Computer Sci

                                           Workforce Tm

            Completion/retention

            Placement/employment

            Average class size

            Student/faculty ratio

            Accreditation factors

             Faculty by disc. cred. hrs

             Faculty by terminal degrs

             Library holdings

              Books

              Periodicals

              Media

              Internet connections

            Workload

                                         California         Wyoming

Type        Variable

Sectoring/
Initial     Mission                      Mission            Mission
selection

            Control                      Public             Public

                                         Inst.
                                         sector

            Classification               IPEDS              NCHEMS
                                         two-year           two-year

Size

            Enrollments                                     Enrollment
                                                            size

                                                            FTE

                                         %FT

                                                            PT/FT
                                                            student mix

            Staff

Complexity  Academic programs                               Program
                                                            mix

                 Transfer

              Technical/
              occupational

              Certificate

              College
              preparatory

              General
              education

            Diversity

               Ethnicity                 % by
                                         gender

               Gender

               Internationality

               First-time first-         %
               year students

            Demography

               Geography/
               Location

                Urban

                Metropolitan

                Suburban

                Rural

                Region                   Region/state

              Population

                State

                Service
                area

                Institutional
                structure

Finance     E&G expenditures

            E&G revenues                                    E&G rev/fte
                                                            stud

            Budgeted $

            Restricted budget

            Expenditures

            Student costs

Quality     Selectivity

                 Applications

                 Admissions

            Degrees awarded              % of
                                         enrollment

Quality     Degrees awarded

                                         %
                                         race/gender

            Completion/retention

            Placement/employment

            Average class size

            Student/faculty ratio

            Accreditation factors

             Faculty by disc. cred. hrs

             Faculty by terminal degrs

             Library holdings

              Books

              Periodicals

              Media

              Internet connections

            Workload

                                         Nebraska           Colorado

Type        Variable

Sectoring/
Initial     Mission
selection

            Control                      Public

                                         Control

                                         Level

            Classification               NCHEMS             Two-year
                                         two-year

Size

            Enrollments                  Headcount

                                         FTE

            Staff

Complexity  Academic programs            #                  Deg. credit
                                                            required

                 Transfer                #

              Technical/
              occupational

              Certificate

              College
              preparatory

              General                                       # courses
              education                                     available

            Diversity

               Ethnicity                                    Retention
                                                            rate
               Gender

               Internationality

               First-time first-
               year students

            Demography

               Geography/
               Location

                Urban

                Metropolitan

                Suburban

                Rural

                Region

              Population

                State

                Service
                area

                Institutional
                structure

Finance     E&G expenditures             by Source

            E&G revenues                 by Source

            Budgeted $

            Restricted budget

            Expenditures                                    Inst. sup/
                                                            fte

            Student costs                Tuition &          # courses
                                         fees               available

Quality     Selectivity

                 Applications            %

                 Admissions              % admitted

            Degrees awarded

                                         Diplomas

                                         Certificates

                                         Associates

                                         by Areas

Quality     Degrees awarded

            Completion/retention

                                         Fresh ret rate     Three-year
                                                            grad rate

                                         %                  Retention
                                                            rate

            Placement/employment                            % Cert. &
                                                            AAS
            Average class size

            Student/faculty ratio

            Accreditation factors

             Faculty by disc. cred. hrs

             Faculty by terminal degrs

             Library holdings

              Books

              Periodicals

              Media

              Internet connections

            Workload                                        Faculty
                                                            teaching

                                         El Paso            NACUBO

Type        Variable

Sectoring/
Initial     Mission
selection

            Control

            Classification

Size

            Enrollments                  Range stud.
                                         enrollment

                                                            Credit FTE

             Staff

Complexity  Academic programs

                 Transfer

              Technical/
              occupational

              Certificate

              College
              preparatory

              General
              education

            Diversity

               Ethnicity                 % Hispanic

               Gender

               Internationality

               First-time first-
               year students

            Demography

               Geography/
               Location

                Urban

                Metropolitan

                Suburban

                Rural

                Region                                      National

              Population                 Range
                                         served

                State                    % Hispanic

                Service
                area

                Institutional
                structure

Finance     E&G expenditures

            E&G revenues

            Budgeted $                   Range

            Restricted budget            %

            Expenditures

            Student costs

Quality     Selectivity

                 Applications

                 Admissions

            Degrees awarded

Quality     Degrees awarded

            Completion/retention

            Placement/employment

            Average class size

            Student/faculty ratio        Students/
                                         faculty

            Accreditation factors

             Faculty by disc. cred. hrs

             Faculty by terminal degrs

             Library holdings

              Books

              Periodicals

              Media

              Internet connections

            Workload

References

Arnold, C.L. (1997). Using national data sets to create comparable national statistics for the student characteristics and outcomes in community colleges. (ERIC Document Reproduction Service No. ED421197).

Berthold, C. A. (1996). Aids to navigation: Using interinstitutional databases in the University of Maryland System. Paper presented at the Annual Forum of the Association for Institutional Research, Albuquerque, NM. (ERIC Document Reproduction Service No. ED397715)

Brinkman, P. T., & Teeter, D. J. (1987). Methods for selecting comparison groups. New Directions for Institutional Research, 53, 5-23.

Brinkman, P. T., & Krakower, J. (1983). Comparative data for administrators in higher education. Boulder, CO: National Center for Higher Education Management Systems.

Carnegie Foundation for the Advancement of Teaching. (2000). The Carnegie classification of institutions of higher education (2000 Edition). Available [on-line]: http://www.carnegiefoundation.org/Classification/CIHE2000/defNotes/ Definitions.htm.

Colorado Commission on Higher Education. (2001). Performance fudning process and quality indicator system for FY 2001-02. [On-line]:http://www.state.co.us/ ccheZ_dir/hecche.html.

Connecticut Department of Education. (1998). Transforming Connecticut higher education: An agenda for excellence into the 21st century, April 2, 1998, Report of the Public Agenda Advisory Council [On-line] http://www.ctdhe.org/dheweb/ toc.htm.

Curry, D.J. (1972). The seven comparison states: Their selection, use and applicability for higher education. Boulder, CO: National Center for Higher Education Management Systems. (ERIC Document Reproduction Service No. ED 095768)

Della Mea, C. L. (1989). A comparison of two procedures for peer group assignment of institutions of higher education. Unpublished doctoral dissertation, Virginia Polytechnic Institute & State University, Blacksburg, VA.

Glover, R. H., & Mills, M. R. (1989). Interinstitutional comparisons for decision-making. New Directions for Higher Education, 68, 79-92.

Illinois Community College Board. (1999). Fiscal year 1999 salary survey report for the Illinois Public Community Colleges. (ERIC Document Reproduction Service No. ED429638)

Ingram, J. A. (1995). Using IPEDS for selecting peer institutions. Lincoln, NE: Coordinating Commission for Postsecondary Education. (ERIC Document Reproduction Service No. ED387010)

Katsinas, S. G. (1996). Preparing leaders for diverse institutional settings. In J.C. Palmer & S.G. Katsinas (Eds.), Graduate and Continuing Education for Community College Leaders: What it Means Today (pp. 15-25). New Directions for Community Colleges, No. 95.

Katsinas, S.G. (1993). Toward a classification system for community colleges. Paper presented to the 1993 Annual Meeting of the Council of Universities and Colleges (Portland, OR.)

Korb, R. (1982). Clusters of colleges and universities: An empirically determined system. Washington, D.C.: National Center for Educational Statistics. (ERIC Document Reproduction Service No. ED227797)

Meeker, B. (1997). Comparative financial statistics for public two-year colleges: FY 1995 peer groups sample. Washington, D.C.: National Association of College and University Business Officers. (ERIC Document Reproduction Service No. ED409055)

Nebraska Coordinating Commission for Postsecondary Education (2000). Comprehensive Statewide Plan for Postsecondary Education [On-line]: http:// www.ccpe.state.ne.us/PublicDoc/CCPE/complantoc.asp.

New Jersey Commission on Higher Education. (2000). Accountability in higher education: The fourth annual systemwide report [On-line]: http:// www.state.nj.us/highereducation/ar04.htm.

Ohio Board of Regents. (2000). Performance Report Consultation [On-line]: http:/ /www.regents.stat.oh.us/perfrpt/.

Prather, J. E., & Carlson, C. E. (1991). Using institutional comparisons for administrative decision support. (ERIC Document Reproduction Service No. ED335794)

Rawson, T. M., Hoyt, D. P., & Teeter, D. J. (1983). Identifying "comparable" institutions. Research in Higher Education, 18, 3, 299-310.

Soteriou, N. E. (1994). Peer institution profile report: El Paso Community College. El Paso, TX: EPCC Office on Institutional Research and Evaluation. (ERIC Document Reproduction Service No. ED376874)

Szelest, B. P. (1996). In search of peer institutions: Two methods of exploring and determining peer institutions. Proceedings of the Annual Conference of the Northeast Association for Institutional Research (Princeton, NJ). (ERIC Document Reproduction Service No. ED432186)

Teeter, D. J. & Christal, M. E. (1987). Establishing peer groups: A comparison of methodologies. Planning for Higher Education, 15, 2, 8-17.

Teeter, D. J. (1983). The politics of comparing data with other institutions. New Directions for Institutional Research, 10, 2, 39-48.

Terenzini, P. T., Hartmark, L., Lorang, W. G. & Shirley, R. C. (1980). A conceptual and methodological approach to the identification of peer institutions. Research in Higher Education, 12, 4, 347-364.

University System of Georgia. (2000). Benchmarking and management review study [On-line]: http://www.usg.edu/admin/benchmark.

Weeks, S. F., Puckett, D., & Daron, R. (2000). Developing peer groups for the Oregon University System: From politics to analysis (and back). Research in Higher Education, 41, 1, 1-20.

West Virginia Higher Education Policy Commission. (2001). Summary of Senate Bill 703 [On-line]: http://www.hepc.wvnet.edu/.

Wyoming Community College Commission. (1999). WCCC hopes to change funding formula [On-line]: http://commission.wcc.edu/WCCC/Misc/news.htm.

Zhao, J., & Dean, D. C. (1997). Selecting peer institutions: A hybrid approach. Paper presented at the Annual Forum of the Association for Institutional Research (Orlando, FL.). (ERIC Document Reproduction Service No. ED 410877)

Rodney G. Hurley is a professor of business management and statistics at Hillsborough Community College and Ph.D. candidate in higher education administration at the University of South Florida in Tampa, Florida. [e-mail]