Arkansas Collegiate Performance

By Morgan Fincher

Acknowledgments

I would like to thank Dr. Seo-eun Choi for the opportunity to participate in this research and the National Science Foundation for the funding provided. I am grateful to the thesis committee for volunteering their time and talents to assist me in the completion of the project. A special thanks goes to my family for their constant support and motivation. Without these individuals, the completion of this thesis would not have been possible.

2 Table of Contents

ACKNOWLEDGEMENTS...... 2

TABLE OF CONTENTS...... 3

CHAPTERS

I. Prelude Abstract...... 4 Why this Topic? ...... 5 Research Population...... 6 What is Performance? ...... 10

II. Setting the Stage Defining Collegiate Performance...... 12 The Validity of this Definition...... 13 Research Summary...... 15

III. Collecting & Analyzing Data Collecting Quantitative Variables...... 17 Collecting Qualitative Variables...... 18 Factor and Cluster Analysis...... 20

IV. Conclusions Two Results...... 21 Final Conclusion...... 24 Beyond the Conclusion...... 27

REFERENCES...... 29

APPENDIX...... 30

3 CHAPTER I: PRELUDE

Abstract

Arkansas Collegiate Performance defined the meaning of “performance” as it

applies to higher education. A formal definition of collegiate performance was

established by identifying the variables possessed by a high performing school. The

definition was comprised of both quantitative and qualitative variables, and was used

throughout the research to compare all colleges and universities in the state of Arkansas.

All four-year institutions of higher education in Arkansas were included in this

research, both public and private. The quantitative data was collected from the schools

and the qualitative data from their students. A multistage clustering technique was used

to select a random sample of students from each school. All data received from these

schools was documented, and those schools not willing to participate were not included in the final results.

Factor Analysis was used in order to reduce the number of variables being compared from school to school. Factor Analysis placed highly correlated variables in clusters that had relatively small correlations to variables in other clusters. Once these

factors were justified to represent the original definition of performance accurately, they

became the new model for defining collegiate performance.

The data collected concerning these factors from each school was compared by

Cluster Analysis. This process grouped schools with similar performance. Results could

then be determined by analyzing schools within groups or from group to group.

Arkansas Collegiate Performance truly compares schools based on the factors

comprised of the most important variables defining high performance. The data from

4 each school was thoroughly analyzed in order to classify these schools in the most valid way possible.

Why this Topic?

There are several college and university rankings readily available to the public.

These include the Princeton Review, Forbes, U.S. News & World Report, Washington

Monthly, Kiplinger, Times Higher Education-QS World University Rankings, Academic

Ranking of World Universities, and many more. The art of comparing and ranking schools is not uncommon. However, rankings specific to individual states can be rare.

This research, Arkansas Collegiate Performance, was designed to be a valid ranking system for colleges and universities located in Arkansas.

The Arkansas lottery and scholarship program was approved in March of 2009.

This scholarship was available for prospective college students by the 2010-2011 school year. Lt. Governor Bill Halter stated at this time that,

"The Arkansas Scholarship Lottery is performing at a level that is

exceeding all expectations. Net lottery proceeds for college scholarships

are projected to reach $112 million in the first 12 months. Tens of millions

of Arkansas dollars previously spent playing lottery games in five

bordering states are now being invested in higher education here at home.

The Arkansas Department of Higher Education is just a few weeks away

from notifying applicants that they meet the eligibility standards for

scholarship awards that will pay a significant portion of their tuition upon

admission to a public or private college or university in Arkansas. Final

5 notification of more than 25,000 Arkansans that they are scholarship

winners will likely begin in May. (Terrebonne, 2010).”

Thanks to the lottery scholarship, money has been more abundant throughout the state. The lottery has allowed many students the opportunity to go to college and has led many prestigious students to remain in state to continue their education. Since more students are now entering these institutions of higher education in Arkansas, it is only appropriate that they are informed of all educational options available to them. Many schools in Arkansas are overlooked, and this research places all of them at an equal, unbiased level before comparing.

The results from Arkansas Collegiate Performance will be very beneficial to prospective students, because the schools will also be represented in a fair and accurate way. Schools can use these results to evaluate their performance in order to draw more students to their school. Arkansas schools should strive to appeal to future students, thus research results such as those from this study will be of high interest. This research is not only necessary, but proves to be beneficial to a large variety of individuals.

Research Population

According to Arkansas Department of Higher Education (ADHE), Arkansas has

ten public, undergraduate, four-year Universities and eleven private. All of these

were included in this research.

6 State Four Year Institutions:

• Arkansas State University (ASU) • (ATU) • Henderson State University (HSU) • Southern Arkansas University (SAU) • – Fayetteville (UofA) • University of Arkansas - Fort Smith (UAFS) • University of Arkansas at Little Rock (UALR) • University of Arkansas at Monticello (UAMont) • University of Arkansas at Pine Bluff (UAPB) • University of Central Arkansas (UCA)

Private Four Year Institutions:

(ABC) • Crowley’s Ridge Academy (CRC) • (CBC) • (Hendrix) • (JBU) • (Lyon) • Ouachita Baptist University (OBU) • (PSC) • (Harding) • University of the Ozarks (Ozarks) • Williams Baptist College (WBC)

Note: abbreviations following each school name will be used throughout this research

For the qualitative variables (the survey portion of this research), the sample

population was formed through a multi-stage, cluster sampling process. Two departments were selected from each school. A smaller department was selected from each of those departments, and finally, a student classification year was selected from each of the smaller departments yielding a final sample population of two per school.

The final sample populations selected were sent the student surveys.

7

Figure 1.1

SAMPLING METHOD

The samples found are illustrated in Table 1.1.

Table 1.1

SCHOOL SAMPLE GROUPS

University Divisions (Dept. or College) Sub-Divisions (1 from each Classification Division) Arkansas State University Business Accounting Senior Humanities History Sophomores Arkansas Tech University Business Business Education Senior Natural and Health Sciences Physical Sciences Senior Henderson State University Honors Whole College Junior Arts and Sciences Music Sophomore Southern Arkansas Business Organizational Management Freshman University Liberal and Performing Arts Behavioral and Social Sciences Freshman University of Arkansas, Education and Health Health Science, Kinesiology, and Senior Fayetteville Professions Recreation and Dance Engineering Chemical Junior University of Arkansas, Business Marketing Senior

8 Fort Smith Humanities and Social Behavioral Sciences Junior Sciences University of Arkansas, Business Accounting Sophomore Little Rock Education Teacher Education Senior University of Arkansas, Agriculture Plant and Soil Science Junior Monticello Computer Information Whole College Senior Systems University of Arkansas, Agriculture, Fisheries, and Agriculture Sophomore Pine Bluff Human Science Education Graduate Program Graduates University of Central Business Economics, Finance, and Insurance Sophomore Arkansas and Risk Management Health and Behavioral Health Sciences Freshmen Sciences Arkansas Baptist College Fine Arts Whole Division Junior General Studies Whole Division Junior Crowley’s Ridge Academy Freshman Junior Central Baptist College Behavioral Sciences Whole Division Freshman Fine Arts Whole Division Sophomore Hendrix College Humanities Religion Sophomore Pre-Professional Programs Public Health Freshman John Brown University Engineering and Construction Construction Management Sophomore Management General Studies General Studies Senior Lyon College Humanities Religion and Philosophy Junior Language and Literature English Sophomore Ouachita Baptist University Interdisciplinary Studies Whole Division Sophomore Fine Arts Applied Arts Sophomore Philander Smith College Education Sophomore Arts and Humanities Philosophy and Religion Junior Harding University Arts and Humanities Foreign Language and International Sophomore Studies Honors Whole Division Junior University of the Ozarks Education Whole Division Senior Science and Mathematics Whole Division Junior Williams Baptist College Art Studio Art Sophomore Education Mid-Level Math and Science Junior

Not all of the schools in Table 1.1 were included in the final survey results. Only those with accessible e-mail addresses and enough student response were included.

9 What is Performance?

Webster defines performance simply as “doing something successfully” while

Encarta Dictionary states that performance is “the manner in which something or

someone functions, operates, or behaves”. There are several definitions that attempt to

define performance as it would apply to a blanket of people, places, things, or actions.

Whatever the vague definition of performance is, it is apparent that the exact definition of

“performance” varies according to subject. High performance of an athlete differs greatly from what is considered high performance of a student. Thus it is only fair that

“performance” for this research be defined specifically for four-year colleges and universities in the state of Arkansas.

Because the definition of “performance” is so inconsistent, the first step in

Arkansas Collegiate Performance was to develop a more specific definition. This was a

long process throughout which many variables were explored. Appropriate variables

from many prestigious college rankings and other sources were considered. These

variables were grouped together with other similar variables to form five categories:

Academics, Atmosphere, Facilities, Selectivity, and Opportunities. Each variable was

easily placed in to one of these categories after which the groups were analyzed.

Redundant and inadequate variables were eliminated until a final list of 22 variables remained. These 22 variables became the definition for collegiate performance.

Each variable could be classified as either factual or opinionated information.

Information found in a school’s fact book should not be the only component considered when evaluating a school. The students that attend the college or university as well as their opinions of the school are also very important. Thus, Arkansas Collegiate

10 Performance’s definition of performance contains both factual and opinionated data. The

factual data was considered to be quantitative, since it can be measured, while the

opinionated data was considered to be qualitative, since it can only be rated. The 22

variable definition was comprised of 13 quantitative and nine qualitative variables.

The following sections will further examine this definition of performance and ensure its validity. The validity of Arkansas Collegiate Performance as a whole depends on this definition. If one does not agree that this is a competent definition of performance, the research results lose their legitimacy.

11 CHAPTER II: Setting the Stage

Defining Collegiate Performance

The final definition of Collegiate Performance was comprised of 22 variables that fell in to five categories. This definition proved to be suitable for the rest of the research.

Table 2.1

PERFORMANCE DEFINITION

Category Variables Variable Type ACADEMICS Number of Degrees Offered Quantitative Total Enrollment Quantitative Student-Teacher Ratio Quantitative Faculty Education Quantitative Average Hours Studied Qualitative Number of Returning Freshmen Quantitative

ATMOSPHERE Location Qualitative Community Involvement Qualitative Student Diversity Quantitative Campus Size (Transportation) Qualitative Friendliness (Professor / Qualitative Students)

FACILITIES Beauty of Campus Qualitative Condition of Buildings Qualitative Residence Comfort Qualitative Residence Cleanliness Qualitative

SELECTIVITY Acceptance Percentage Quantitative Entering Freshman GPA Quantitative Average ACT score Quantitative Percent of Out-of-State Students Quantitative

OPPORTUNITIES Greek Life Quantitative Clubs/Organizations Quantitative Intramurals Offered Quantitative The quantitative variables are factual information obtained directly from the college or

university while the qualitative variables are opinions obtained through student surveys.

12 The Validity of this Definition

The validity of this research relies heavily on this definition, so it was very

important to justify that this definition was appropriate for the intended purpose of

Arkansas Collegiate Performance. Because a number variables included in the definition came from other college ranking sources, it seemed appropriate to compare the definition to the various definitions from these mentioned sources.

According to an article from The Chronicle of Higher Education, there are 30

ways to rank a college: admission rates, standardized-test scores, class rank, percentage

of federal work-study grants focused on community service, peer assessment/ reputation

survey, student evaluation of faculty and institution, research spending, percentage of

alumni donating to college, faculty compensation, international attendance ratio,

international faculty ratio, Army/Navy ROTC size, alumni serving in the Peace Corps,

total cost, percentage of students receiving Pell Grants, average portion of financial need

met by student aid, student-borrower debt, student-loan-default rates, graduation/retention rates, faculty membership in the National Academies, prestigious awards/scholarships to students/faculty, Ph.D.’s awarded to students, faculty publications and citations, alumni salaries, professionally successful alumni, student-faculty ratio, percentage of full time faculty, class sizes, instructional spending per students, and instructor educational attainment (Richards, 2010). These variables are used by the

prestigious publications: U.S. News & World Report, Washington Monthly, Forbes,

Kiplinger, Times Higher Education- QS World University Rankings, and Academic

Ranking of World Universities.

13 Surprisingly, few variables are used by more than one ranking source, speaking to

the uniqueness of each of these rankings. The diverse definitions of what qualifies a

school as high performing show that there is no perfect and universally agreed upon way

to define collegiate performance. However, Arkansas Collegiate Performance compared favorably with these other ranking sources and the 30 ways in which they measured school quality. The definition used throughout this research contains nine of the 30

variables from the article. By comparison, the U.S. News & World Report contained

twelve, Washington Monthly - ten, Forbes - seven, Kiplinger - seven, Times - fiver, and

the Academic Ranking of World Universities - two. Thus, Arkansas Collegiate

Performance has a definition of performance that is above average in the number of

variables considered the most valid by various other school rankings.

As previously mentioned, most of the 30 variables were only used once among

the six ranking sources. However, variables that were used multiple times were

admission rates, standardized-testing, peer assessment/reputation survey, student-

borrower debt, graduation/retention rates, prestigious awards/scholarships to

students/faculty, faculty publication and citations, student-faculty ratio, and percentage of

faculty member who work full time. Arkansas Collegiate Performance contains five of

these nine most used variables, more than any of the other publication included in the

article except Kiplinger, which also contains five, and U.S. News & World Report, which

includes six.

14 Research Summary

Arkansas Collegiate Performance relied on three key steps: defining, collecting, and analyzing data. Up to this point, all four-year universities and colleges in the state of

Arkansas have been declared as subjects for the research, and the definition of

“performance” as it is to be applied to these subjects has been determined by a number of

variables.

Once the definition was formed, the variables were separated into either

quantitative and qualitative. The quantitative variables were gathered from existing

resources, such a school websites and fact books, while the qualitative variables were

provided by student survey results. Both of these processes were part of the second step

of research: collecting data. The data for each variable was collected from every school

and was collected from the most appropriate source: school fact books, faculty, staff, or

students. The data was then accurately recorded in files for each individual school. Once

complete data was collected across all schools, it could easily be compared. Though

collecting the data was the most time consuming step, the most complex and important

step was analyzing the data.

Due to the complexity of comparing such a large number of variables (22) across

such a large number of schools (21), it was obvious that Factor Analysis would be helpful

to this process. Factor Analysis is the process that removes the redundancy from the set

of correlated variables by representing them as a smaller set of derived variables known

as “factors” (Child, 1970). Each factor represents a group of underlying, highly-

correlated variables. Through this process, all the original variables are represented, but a

15 much smaller group (known as factors) can now be analyzed. These factors can be more easily be compared from school to school.

Cluster Analysis was the next method used to analyze the data. This method groups schools with similar performance, according to their factors, together in a group called a “cluster”(Aldenderfer, 1984). This method proved beneficial because comparing schools in a highly-correlated group is not only more efficient, but also more practical.

The next chapter more thoroughly explains the collecting and analyzing portions of this research.

16 CHAPTER III: Collecting & Analyzing Data

Collecting Quantitative Variables

The following information was collected from various school websites, fact

books, and appropriate collegiate staff. Missing data values were either not obtainable or

applicable for the corresponding school.

Table 3.1 QUANTITATIVE DATA

School ASU ATU UofA UALR JBU HARDING OZARKS HSU UAFS UAPB Degrees 7 6 5 5 3 4 1 4 4 5 Enrollment 12156 8814 19849 13132 2073 6484 625 3567 7335 3802 S:T 18 18 17 14 13 17 10 15 20 17 Faculty Ed 65.43 69 86.06 64.1 73 67 49.38 96 67.3 68.9 Retention 67.9 43.42 83.1 61 77 81 69 61 . . Diversity 30.91 15.09 21 35.38 19.02 15 30.88 30 24.69 5 Denial % 23.4 6.33 43.91 0 25 27 . 69 38.6 67.28 GPA 3.15 3.34 3.56 3.04 3.58 3.5 3.4 3.19 . . ACT 21.3 23.09 25.8 22 25.4 24.5 22 22.3 . . %Out of State 10.39 6.93 30.47 8 72.88 75 42.26 18 10 46 Greek 1 1 1 1 0 0 0 1 . . Clubs 114 102 356 118 43 100 . 69 . . Intramurals 1 1 1 1 1 1 . 1 . .

SAU UAMont UCA ABC CRC CBC Hendrix Lyon OBU PSC WBC Degrees 5 7 7 2 2 2 2 1 2 1 2 Enrollment 3226 3482 11781 626 201 625 1470 614 1503 668 619 S:T 15 17 18 15 10 . 13 13 12 9 16 Faculty Ed 50.3 59.6 67.5 22.2 80 83.3 94.2 97.8 92.9 79.6 92.3 Retention 56.36 . 56.1 . . . 88 . 81 . . Student Diversity 37 34 19.46 2.72 16 30.56 14 15 9.88 5 4 Denial% 26 . 42 0 0 20.22 18.7 32.82 33 30.75 35.82 GPA 3.13 34.49 3.3 . 3.09 . . . 3.57 . . ACT 20.9 23.6 . 19.3 . . . 24.5 . . %Out of State 33 10 10.49 . 32 . . 33 49 54 19 Greek . . 1 . . . 0 . 0 . . Clubs ...... 33 . . Intramurals . . 1 . . . 1 . . . .

17 Collecting Qualitative Variables

The following data was obtained from an online survey database. The randomly

selected students (Table 1.1) were sent an e-mail with the link to the survey. Anonymous

survey responses were taken from the website, then recorded as an Excel file. The

averages from the collected student responses were taken for each survey question, but

not all schools participated in this portion of the research.

Table 3.2

QUALITATIVE DATA

School ASU ATU UofA UALR JBU HARDING Ozarks HSU UAFS UAPB Qualitative 1 8.71 6.5 6.02 7.28 8.67 8.41 10.75 8.25 6.5 4.111 Qualitative 2 2.53 2.5 3.33 3.827 3 2.11 2.25 2 2.833 2.556 Qualitative 3 2.18 1.5 3.63 3.396 3 2.27 2.5 1.83 2.333 1.556 Qualitative 4 2.63 2 3.69 3.055 1.67 2.94 2.5 2 3 1.889 Qualitative 5 1.75 2 4.01 3.785 2 2.41 1.5 2.416 2.833 1.667 Qualitative 6 3 3 4.6 4.77 3.67 3.82 2 2.333 4.333 1.889 Qualitative 7 2.82 2.5 4.59 3.37 3 3.1 2.5 3.583 2 2.667 Qualitative 8 3.65 3 4.46 3.15 3 3.9 4 3.75 3 3.445 Qualitative 9 3.82 4 3.91 3.456 4 4.5 5 3.9167 4.333 3.667 Qualitative 10 4 3.5 4.13 3.97 4.33 4.5 4.75 4.083 4.167 4.111 Qualitative 11 4.33 4.25 4.37 3.797 4.33 4.68 5 3.83 4.167 2.333 Qualitative 12 4 3.25 3.85 3.97 4.67 4.22 4.5 3.4167 4 2.444 Qualitative 13 2.33 3 3.79 3.67 3.67 3.93 4 3.556 4.5 3 Qualitative 14 2.5 4 3.92 3.67 3.67 3.97 4 3.556 4.5 3 Qualitative 15 3.25 3 3.52 3.26 2.33 2.66 1.67 2.416 3.167 1.875 Qualitative 16 3.75 4 2.79 3.66 4 4.13 4.5 4.333 4 4.5 Qualitative 17 3 2.33 2.79 3.2 3 3.12 4.33 2.583 2.67 2.625 Qualitative 18 1.89 1 3.816 2.72 1 1.91 1 1.167 2 2.857

The 18 survey questions came from the survey in Appendix 1. Survey numbers 1, 2, 3,

16, and 17 were designed to provide background information on the students taking the

survey, and therefore are not included in the qualitative results above. The other 18

questions do correspond in order to those above. All questions (besides the number of

hours spent studying) are on a rating scale of 1-5. Those questions with only 3 possible

18 responses have responses that score either 1,3, or 5, making the rankings in Table 3.2 an

average rating from a 5 point system.

The 18 qualitative survey questions recorded correspond to the 9 original

qualitative variables used to define performance: Hours Spent Studying, Location,

Community Involvement, Campus Size, Friendliness, Beauty of Campus, Condition of

Buildings, Residence Comfort, and Residence Cleanliness.

These qualitative results, along with the quantitative values for the same ten

schools, make up the data to be analyzed by the statistics program SAS. The data placed

in to this program came from the values in Table 3.3.

Table 3.3

School ASU ATU UofA UALR JBU HARDING OZARKS HSU UAFS UAPB Survey 1 8.71 6.5 6.02 7.28 8.67 8.41 10.75 8.25 6.5 4.111 Survey 2 2.53 2.5 3.33 3.827 3 2.11 2.25 2 2.833 2.556 Survey 3 2.18 1.5 3.63 3.396 3 2.27 2.5 1.83 2.333 1.556 Survey 4 2.63 2 3.69 3.055 1.67 2.94 2.5 2 3 1.889 Survey 5 1.75 2 4.01 3.785 2 2.41 1.5 2.416 2.833 1.667 Survey 6 3 3 4.6 4.77 3.67 3.82 2 2.333 4.333 1.889 Survey 7 2.82 2.5 4.59 3.37 3 3.1 2.5 3.583 2 2.667 Survey 8 3.65 3 4.46 3.15 3 3.9 4 3.75 3 3.445 Survey 9 3.82 4 3.91 3.456 4 4.5 5 3.9167 4.333 3.667 Survey 10 4 3.5 4.13 3.97 4.33 4.5 4.75 4.083 4.167 4.111 Survey 11 4.33 4.25 4.37 3.797 4.33 4.68 5 3.83 4.167 2.333 Survey 12 4 3.25 3.85 3.97 4.67 4.22 4.5 3.4167 4 2.444 Survey 13 2.33 3 3.79 3.67 3.67 3.93 4 3.556 4.5 3 Survey 14 2.5 4 3.92 3.67 3.67 3.97 4 3.556 4.5 3 Survey 15 3.25 3 3.52 3.26 2.33 2.66 1.67 2.416 3.167 1.875 Survey 16 3.75 4 2.79 3.66 4 4.13 4.5 4.333 4 4.5 Survey 17 3 2.33 2.79 3.2 3 3.12 4.33 2.583 2.67 2.625 Survey 18 1.89 1 3.816 2.72 1 1.91 1 1.167 2 2.857 Data 1 7 6 5 5 3 4 1 4 4 5 Data 2 12156 8814 19849 13132 2073 6484 625 3567 7335 3802 Data 3 18 18 17 14 13 17 10 15 20 17 Data 4 65.43 69 86.06 64.1 73 67 49.38 96 67.3 68.9 Data 5 67.9 43.42 83.1 61 77 81 69 61 . . Data 6 30.91 15.09 21 35.38 19.02 15 30.88 30 24.69 5 Data 7 23.4 6.33 43.91 0 25 27 . 69 38.6 67.28 Data 8 3.15 3.34 3.56 3.04 3.58 3.5 3.4 3.19 . .

19 Data 9 21.3 23.09 25.8 22 25.4 24.5 22 22.3 . . Data 10 10.39 6.93 30.47 8 72.88 75 42.26 18 10 46 Data 11 1 1 1 1 0 0 0 1 . . Data 12 114 102 356 118 43 100 . 69 . . Data 13 1 1 1 1 1 1 . 1 . . Factor and Cluster Analysis

At this point in the research, 22 variables were to be compared across ten schools

(the schools that participated in the online survey). However, the data was still numerous

and complex, so to simplify this data, Factor Analysis was used.

Factor Analysis is a statistical approach that can be used to analyze relationships

between a large number of variables in order to explain them by a smaller set of

components. These components, known as factors, are a smaller set of variables derived

from the original variables. In other words, this ‘orderly simplification’ does what man

has engaged in throughout history, it makes order out of the chaos in the environment

(Child, 1970).

Using the Statistical Analysis System, SAS, the data from Table 3.3 was

simplified by Factor Analysis. The variables were reduced to a smaller number of factors

made up of original variables and given new and appropriate names as shown in Tables

4.1 and 4.2.

In order to compare all ten schools fairly according to these factors, Cluster

Analysis was used. Cluster Analysis is a multivariate statistical procedure that starts with

a data set containing information about a sample of entities and attempts to reorganize

them into relatively homogeneous groups (Aldenderfer, 1984). This method places the

ten very different schools into similar, homogeneous groups known as

“clusters”(Aldenderfer, 1984). Once grouped, schools can be compared more reasonably.

20 SAS was also used to run Cluster Analysis, and it placed the schools into the clusters shown in Figures 4.1 and 4.2.

CHAPTER IV: Conclusion

Two Results

Using the SAS program to run both Factor and Cluster Analysis on the data from all ten schools across all variables, five factors were found. Using the rotated factor pattern given, all the original variables with high correlation values were thought of as the main components of the new factors, and were therefore used in forming its new name. The factor names are illustrated in Table 4.1.

Table 4.1

FACTOR NAMES

Factor Components Name 1 Opportunity, Entertainment, Community Environment Enjoyment, Community, Campus Size, Walking, Transportation, Enrollment, Clubs 2 Friendliness, Beauty, Campus Environment Degrees, Retention, ACT, GPA, Out of State Percentage, Greek 3 Student Teacher Ratio, Opportunities Faculty Education, Town Size, Parking 4 Beauty, Degrees, Student Academics Teacher Ratio, Faculty Education 5 Hours of Study, Community Prestige Support, Campus Housing, Parking, Student Diversity, and Denial Percentage

21 Using these factors, the schools were also placed in clusters. The hierarchical cluster

analysis tree in Figure 4.1 allowed the formation of six groups (when evaluated around

1.05).

Figure 4.1

According to this procedure, Arkansas State University and University of Arkansas Little

Rock make up a cluster, Arkansas Tech University, John Brown University, and Harding

make up a second cluster, University of Arkansas and Henderson State University make up a third cluster, and the University of the Ozarks, University of Arkansas Fort Smith, and University of Arkansas Pine Bluff each form a cluster of their own.

These results, however, were unsatisfactory. The three schools in groups of their own had not been included in Cluster Analysis due to missing data values. In order to correct this error, the variables with missing data values were taken out of the data set.

These variables included seven quantitative variables: Retention Rates, Denial

22 Percentage, ACT, GPA, Greek Life, Clubs, and Intramurals. Therefore, the data had to

be re-analyzed leading to a second group of results.

Once again, using the SAS program to run both factor and cluster analysis, the

data was analyzed. This time all ten schools were compared across seven less variables

than the original analysis, but once again five factors were found. The new factor pattern

and components of these different factors are illustrated in Table 4.2.

Table 4.2

FACTOR NAMES II

Factor Components Name 1 Town Size, Opportunities, Opportunities Entertainment, Enjoyment, Transportation, Walking, Enrollment 2 Friendliness, Campus Size, Academics Degrees, Student Teacher Ratio, Out of State Percentage 3 Hours of Study, Beauty, Campus Environment Parking, Student Diversity 4 Friendliness, Campus Residence Life Housing 5 Community, Community Town Environment Support, Faculty Education

Using these factors, the schools were placed into their respective clusters. Using the

hierarchical cluster analysis tree in Figure 4.2, it was decided to divide the schools into 3

groups (cutting across around 1.25). This time no schools were clustered by themselves.

The results from this second run were much more reliable, and having three groups

proved more satisfactory than the six previous groups.

23

Figure 4.2

According to this procedure, Arkansas State University, University of Arkansas

Little Rock, John Brown University, and the University of Arkansas make up a cluster,

Arkansas Tech University, University of Arkansas Fort Smith, Harding University,

Henderson State University, and University of the Ozarks make up a second cluster, and

University of Arkansas Pine Bluff is in a cluster on its own. This second set of results

will lead to the final conclusion of Arkansas Collegiate Performance.

Final Conclusion

24 Schools placed in clusters together can be considered similar when it comes to

this definition of performance, and now it is possible to rank the groups as wholes. The

final groups are clearly demonstrated in Figure 4.3.

Figure 4.3

FINAL SCHOOL CLUSTERS

All 10 Schools

Arkansas State University Arkansas Tech University University of Arkansas Pine Bluff University of Arkansas University of Arkansas Little Rock Fort Smith

John Brown University Harding University

University of Arkansas Henderson State University

University of the Ozarks

In order to rank the groups, the results from factor analysis were once again

useful. The factor scores that each school received shows how they compare to other

25 schools according to that specific factor. Therefore, for each factor, the group with the highest average of scores for that factor was considered the highest performer in that specific area. Using Figure 4.4 (taken directly from the SAS program results), the average factor score for each group of schools was taken and placed in Table 4.3.

Figure 4.4

The SAS System

School Name Opportunities Academics Campus Residence Life Town

1 ASU -0.35881 1.19702 1.12349 -1.64756 0.39455 2 ATU -0.83092 1.53831 -0.27367 0.18011 -0.65333 3 UofA 1.69195 0.13109 -0.37738 0.28804 1.85965 4 UALR 1.83286 -0.09458 0.35597 -0.70542 -1.22973 5 JBU 0.20485 -1.10833 0.10806 -0.06872 -0.97939 6 HARDING -0.43486 -0.29614 0.00907 1.12939 0.97721 7 OZARKS -0.81232 -1.53663 1.57224 0.30252 0.05241 8 HSU -0.82389 0.15635 -0.12884 -0.35356 0.74727 9 UAFS 0.13915 0.86457 -0.18454 1.81469 -0.96433 10 UAPB -0.60800 -0.85166 -2.20440 -0.93949 -0.20430

Table 4.3

Cluster Opportunities Academics Campus Residence Town Environment Life Environment ASU, UofA, 0.84271 0.0313 0.302535 -0.53342 0.01127 UALR, JBU ATU, Harding, -0.55257 0.14529 0.19885 .61463 0.03185 Ozarks, HSU, UAFS UAPB -0.60800 -0.85166 -2.20440 -0.93949 -0.20430

Note: the highest score for each factor is highlighted

26 According to the results, the group containing Arkansas State University,

University of Arkansas, University of Arkansas at Little Rock, and John Brown

University, scored highest for Opportunities and Campus Environment. While Arkansas

Tech University, Harding, University of the Ozarks, Henderson State University, and

University of Arkansas at Fort Smith’s group scored highest for Academics, Residence

Life, and Town Environment.

Beyond the Conclusion

Arkansas Collegiate Performance was certainly limited by the participation of each school. Only the ten schools with a full set of quantitative data and high student response for qualitative data were included in the final results. However, there is still more that can come from this research. In the future, a larger group can be compared if more schools participate in the student survey. Also, qualitative data could be dropped all together so that the results depended only on collecting the factual school information.

In order to show how the results from quantitative variables only would compare

using Arkansas Collegiate Performance, cluster analysis was used to group together all

21 schools according to these variables. However, many variables had to be dropped due

to missing values. Therefore, the results in Figure 4.5 are those for these variables only:

Degrees, Enrollment, Student Diversity, and Faculty Education. Factor Analysis was not

necessary since there were so few variables.

Figure 4.5

27

ASU UALR UCA UofA ATU UAFS UAPB SAU UAMont HSU Harding CBC JBU CRC PSC Hendrix OBU UofO Lyon WBC ABC

References

Aldenderfer, Mark S. and Blashfield, Roger K. (1984) Cluster Analysis, Sage Publications, Inc.

Anderson, T.W. (1984) An Introduction to Multivariate Statistical Analysis, second edition, John Wiley & Sons.

Bailey, Daniel E. and Tryon, Robert C. (1970) Cluster Analysis, McGraw-Hill.

Bhattacharyya, G.K. and Johnson, R.A. (1977) Statistical Concepts and Methods, John Wiley & Sons.

Child, Dennis (1970) The Essentials of Factor Analysis, Holt, Rinehart & Winston Ltd.

Cochran, W.G. (1977) Sampling Techniques, third edition, John Wiley & Sons.

Johnson, R.A. and Wichern, D.W. (1992) Applied Multivariate Statistical Analysis, third edition, Prentice Hall.

Kachigan, S.K. (1991) Multivariate Statistics Analysis – A Conceptual Introduction, second edition, Radius Press.

Morrison, D.F. (1990) Multivariate Statistics Methods, third edition, McGraw-Hill.

Neter, J., Wasserman, W. and Kutner, M.H. (1990) Applied Linear Statistical Models, third edition, Irwin.

Richards, Alex and Coddington, Ron. (2010). 30 Ways to Rate a College. Retrieved from http://chronicle.com/article/30-Ways-to-Rate-a-College/124160 .

Scheaffer, R.L., Mendenhall, W. and Ott, L. (1986) Elementary Survey Sampling, third edition, Duxbury Press.

28

Sullivan, M. III (2007) Statistics – Informed Decision Using Data, third edition, Prentice Hall.

Tabachnick, B.G. and Fidell, L.S. (2007) Using Multivariate Statistics, fifth edition, Pearson.

Terrebonne, Amanda. (2010). Arkansas Lottery scholarships: Arkansas Legislature OKs $5k lotto scholarships. Retrieved from http://www.todaysthv.com/news/story.aspx?storyid=100049.

Appendix Online Student Survey

1. Which College/University do you attend?

< Select >

2. What is your classification?

< Select >

3. What is your area of study?

4. (Qualitative 1) Which best describes the number of hours you spend studying outside the classroom per week?

more zero 1 2 3 4 5 6 7 8 9 10 than 10 5. (Qualitative 2) Rate your personal opinion of the town's size: Very Small Small Neutral Large Very Large 6. (Qualitative 3) Rate the opportunities the town offers: Below Average Avergae Above Average 7. (Qualitative 4) How often do you spend weekends going out in your college's town?

29 Never Sometimes Often 8. (Qualitative 5) How often do concerts or other events get hosted in the town? Once a Once a Once a Never More year month week 9. (Qualitative 6) Do you enjoy spending time in your college's town? Yes No 10. (Qualitative 7) Rate the community attendance at collegiate events (such as sporting events): Poor Fair Neutral Good Excellent 11. (Qualitative 8) Rate the community's overall support of your college:

Poor Fair Neutral Good Excellent 12. (Qualitative 9) Rate the overall friendliness of students on your campus:

Very Very Unfriendly Neutral Friendly Unfriendly Friendly 13. (Qualitative 10) Rate the overall friendliness of the faculty on your campus:

Very Very Unfriendly Neutral Friendly Unfriendly Friendly 14. (Qualitative 11) Rate the overall beauty of your campus: Poor Fair Neutral Good Excellent 15. (Qualitative 12) Rate the condition of classrooms and learning facilities: Poor Fair Neutral Good Excellent 16. Do you live on campus?

30

Yes No 17. How would you describe your residence?

Campus Dorm Campus Apartment Greek Housing Other I do not live on campus 18. (Qualitative 13) Rate the comfort of your campus residence (Only if you reside or have resided on campus): Very Very Uncomfortable Neutral Comfortable Uncomfortable Comfortable 19. (Qualitative 14) Rate the cleanliness of your campus residence (Only if you reside or have resided on campus):

Very Very Unsanitary Neutral Sanitary Unsanitary Sanitary 20. (Qualitative 15) Rate the physical size of your campus: Very Small Small Neutral Large Very Large

21. (Qualitative 16) Rate the ease of walking across campus:

Very Hard Hard Neutral Easy Very Easy

23. (Qualitative 17) Rate the parking availability on your campus:

31 Little Enough Abundant No Parking Neutral Parking Parking Parking

24. (Qualitative 18) How would you describe Public Transportation options currently available on your campus? Little Effective Unnecessary No Public Public Neutral Public Public Transportation Transportation Transportation Transportation

32