605 Giải pháp Marketing nhằm phát triển du lịch thành phố Đà Lạt - Lâm Đồng đến năm 2015

605 Giải pháp Marketing nhằm phát triển du lịch thành phố Đà Lạt - Lâm Đồng đến năm 2015

This manuscript has been reproduced from the microfilm master UMI films the text directly from the original or copy submitted Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer

The quality of this reproduction is dependent upon the quality of the copy submitted Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted Also, if unauthorized copyright material had to be removed, a note will indicate

the deletion

Oversize materials (e.g., maps, drawings, charts) are reproduced by

sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps Each original is also photographed in one exposure and is included in reduced form at the back of the book

Photographs included in the original manuscript have been reproduced xerographically in this copy Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge Contact UMI directly to

order

UMI

A Bell & Howell Information Company

300 North Zeeb Road, Ann Arbor MI 48106-1346 USA

313/761-4700 800/521-0600

APPLYING A CONSUMPTION VALUE MODEL

TO COLLEGE CHOICE

by

A Dissertation submitted to the Faculty of the Graduate School, Marquette University,

in Partial Fulfillment of the Requirements for the degree of Doctor of Philosophy

Milwaukee, Wisconsin

May 1997

UMI Microform 9811406 Copyright 1997, by UMI Company All rights reserved

This microform edition is protected against unauthorized copying under Title 17, United States Code

UMI

300 North Zeeb Road Ann Arbor, MI 48103

What drives a person’s decision of where to go to

college? This is a question I have always found

interesting With competition heating up in the area of higher education, college and university administrators are beginning to ask that same question There has been a lot

of research in the area of college choice but I felt that a more comprehensive model was needed With that, I began to read the material that was already available and look for additional models that could be examined What follows is

an article length synopsis of this research project

Marketing Higher Education:

Applying a Consumption Value Model to College Choice Introduction

Over the last 40 years there has been much published in the higher education literature dealing with the college selection process There are many reasons for this growing interest such as, the federal government's financial support

of higher education, a push to enhance the diversity of

college student bodies, a shift in demographics which has changed the pool of college students, as well as a reduced number of college applicants [Paulsen 1990] More recently, the increased competition among colleges for potential

students makes it crucial to understand the college

selection process in order for colleges and universities to develop better recruitment strategies [Kotler and Fox 1985] Colleges and universities have an obvious vested interest in understanding this decision process Those institutions that best understand the process will have the greatest

chance to attract the most qualified students, increase

their enrollment and decrease their transfer or drop out rate To this end, researchers have attempted to enlighten university administrators about the determinants of college selection

One way to organize this body of general research is in two main categories The first involves a distinction

between macro and micro research Another organizing

mechanism for college choice research is "by stage" in the college selection process The “stage” refers to the point

a student is at in the decision process

Macro vs Micro Studies

Macro research has to do with studies that utilize

aggregate student groups and focus on the environmental

effects associated with the total enrollment for all

institutions within a state, region, nation or other

designated area Universities often use environmental

monitoring in an attempt to stay on top of such macro issues

by tracking developments such as demographics, federal

funding policies, and average salaries of college graduates

Micro studies focus on individual students and how

environmental factors affect their enrollment behavior The

motivation for undertaking this research lies with

university administrators who are looking to determine the probability that a given student will attend a particular

institution

Three Stages of College Choice

The other most common way to organize the research on college choice is to break it up into three general stages

of the choice process This organizational approach is

based on the Hossler and Gallagher model (1987) Hossler and Gallagher (1987) identify three main steps:

predisposition, search, and choice These stages coincide with a long standing research approach utilized in marketing

to chart consumer decision making [Engel, Blackwell and

Miniard 1986]

In the context of college choice, the first stage,

predisposition, deals with the question of who chooses

higher education the so called "go-no go" stage In the second stage, search, researchers examine what factors

students use and what methods they employ to narrow down the colleges and universities considered These studies involve

issues such as, private vs public, large vs small, urban vs

rural, coed vs single gender, and possibly geographic

boundaries or cost constraints The final stage, choice, is the part of the time line during which the student enrolls

at one institution narrowed from the choice set developed

in stage two

Much has been published in the area of college choice

focusing on each of the stages Some of the common

variables found to be influential across stages are costs, financial aid, student background characteristics, parental income, educational level and occupation, aspirations and encouragement of others, admissions standards, institutional offerings, overall quality, academic programs, reputation and quality, distance from home, social activities, and

size Data on college choice has been collected in a

variety of ways from simple surveys and focus groups to more sophisticated methods like field experiments The methods

of analyzing choice data vary as well Multiple regression represents a relatively sophisticated form of analysis from the standpoint of linking the variables influencing college choice to the selection of a particular institution In other words, what are the correlates of enrollment at a

particular institution?

conducted, many questions remain A more complete,

comprehensive model of consumer decision making is needed in the area of college choice This research project centered

on a relatively new model that likely has application to the analysis of college choice

The SNG Model

‘The Sheth, Newman and Gross model of consumption values (1990), hereafter referred to as the SNG model, stems from the lifelong work of Jagdish Sheth in consumer behavior and decision making Over time, Sheth felt that some of the research on consumer decision making had gotten too muddled and that an overall, or "big picture" view was needed

Certain streams of research in consumer decision making were becoming lost in the detail Sheth took a step back to see

if there were any overriding factors influencing general consumer decision making He, along with fellow academics Newman and Gross, came up with a more universal model that incorporated five general values which they felt drove the entire consumer decision process

The SNG model is a general theory of consumption

behavior which is based on global consumption values that explain and predict consumer choices The authors label these values functional, social, emotional, epistemic, and

conditional According to the model, individual consumer choices will be influenced by one or more of these values

Why might one select the SNG model as an appropriate framework to examine college choice?

The SNG model evolved from the scrutiny

of various predictive models of consumer behavior It attempts to distill these

models to their essence

Central to the SNG model is the fact that the determinants of choice are global consumption values, not specific purchase criteria

The SNG model provides operational measures of these values that can be utilized in a variety of choice

Situations

The SNG model is conceptualized as a Managerial, pragmatic approach to consumer behavior

And finally, the SNG model represents a general theory of market choice

behavior.

All these points fit with and benefit the study of the

college choice decision process Let's briefly examine each

of the SNG values with regard to college choice

A product's ability to perform pragmatically,

physically or for its utilitarian purpose is the product's functional value The functional aspect of a product or service may be judged by any salient physical feature In regard to college choice, functional values may consist of the tuition charged, types of programs offered, or the job placement rate of recent graduates These variables can all

be measured and compared based on their practical utility to individual students

Social value consists of the influence consumers place

on the purchase decision based upon the reaction of parents, friends, their peer group, or the general public Consumers develop a social image, based upon the pressures exerted by Surrounding social groups These social groups influence various aspects of their behavior

Thus a potential buyer may first evaluate social

aspects of the purchase before exploring other values or features For instance, with regard to college choice, a potential student may first consider the perceptions their friends have of a particular college, or determine if any peers (e.g., boyfriend/girlfriend) are attending (or plan to

attend) a particular institution before engaging in further evaluation In many cases, such social perceptions may be more important to the potential student-consumer than the actual functional characteristics of the institution

Emotional value deals with the affective reaction of consumers to a specific purchase In other words, what

emotions are stirred from a particular purchase or the use

of a given product? In terms of college choice, this may include the feeling of safety and security around the

campus, the feeling of "home" felt in the residence halls,

or the community of common faith associated with attendance

at a college with religious heritage

Epistemic value is the feeling of curiosity that

surrounds a product purchase A consumer may purchase a product for its novelty or simply to satisfy one's

curiosity Epistemic value also encompasses the motivation

of knowledge-seeking, i.e., the desire to know more about a product or service because of inherent interest

A potential college student may allow factors such as regional weather or geography to influence their college selection For example, a Southern California resident may never have been exposed to snow and therefore becomes

curious about living in such conditions; thus, this student may choose to attend a mid-western or eastern university in

the snow belt Alternatively, a non-religious person may attend a sectarian institution to gain knowledge about a religious or value based education

Conditional value deals with market choice decisions made contingent upon particular factors In many cases, the purchase choices people have made may have been very

different had the conditions surrounding their decision been different For example, the decision to attend a particular institution may be contingent upon a conditional factor as basic as whether the institution accepts or rejects the

student Had a more prestigious university accepted a

particular student, they may have chosen to attend that

other college The world holds many Oxford, Harvard and Notre Dame wannabes Other conditional factors might be related to distance from home, financial aid, and

scholarships In today's economic environment it is not

uncommon to hear students opine that if they had received more financial aid from a particular institution, they would

have matriculated there In summary, the combination of these five independent values as influencers of market

choices make up the SNG model of consumer choice The model routinely covers a greater number of possible variable

categories than normally considered in typical college

This research study provides an exploratory test of the Sheth, Newman and Gross (1991) model of consumption values

as applied to the area of college choice The essential thesis of this research is that the SNG model will be able

to more accurately predict the attend/not attend decision of potential students than traditional approaches Therefore, this study is set in the choice stage of the college

selection process In order to conduct this preliminary test, one must select a sample of potential enrollees,

administer the SNG questionnaire, and analyze the data using the SNG and traditional approaches Included in the survey information must be the identification of group membership (1.e., enroll/not enroll) by the respondent

For this study, 1000 students were contacted 500 who chose to attend Marquette and 500 who had canceled their application The students were mailed a customized SNG

questionnaire with 88 specific items related to the SNG

values and approximately 20 basic questions about

demographics, financial aid, etc Item generation for the questionnaire comes from project customized focus groups The response rate was 47.1% (n=471) 58.4% for enrollees and 35.8% for cancels The data set that was gathered was used to test the five hypotheses which provide an

exploratory test of the SNG model in a college choice

setting

The above research procedure was specifically conducted

to test the following five hypotheses:

H,: Market choice is a function of multiple consumption

values

That market choice decisions are a function of multiple values is a fundamental premise of the SNG model The SNG model identifies five different values that influence choice

decisions Consumer choices are influenced to different

degrees by these various values depending upon the buying Situation The authors of the SNG model identify the five overriding values that may influence choice as functional, social, conditional, emotional, and epistemic In this

research, this hypothesis is supported if the items break into several distinct factors when factor analysis is

performed

H,: The SNG values (functional, social, conditional,

emotional and epistemic) exist and are independent of each other

The authors of the SNG model state that market choice

behavior is impacted by multiple values These multiple values are identified as functional value, social value, conaitional value, emotional value, and epistemic value

The values will influence different choice situations

differently Some choice decisions may be influenced by only one value while most are influenced by two or more

values Some choice situations are influenced by all five values [Sheth, Newman and Gross 1991] This hypothesis will test if these values do exist in college choice In this study, this hypothesis is supported if the components of the resulting factors seem to resemble the SNG values

Hạ: Each factor represents a differential contribution to choice determination

Given that market choices are a function of multiple values, the contribution of each value must be considered

It is hypothesized that the five values of the SNG model make differential contributions to specific choice

decisions In other words, in specific decisions, some

values contribute more than others Collectively, each value provides additional information about the choice decision

Admittedly, to propose that different factors influence buying decisions differently in particular buying situations

is somewhat trivial However, if one combines the theory of the SNG model with what is generally known about college choice situations, one should also be able to predict the relative contribution of each SNG model value to college

choice Therefore, for this specific research on the

college choice decision process, it is hypothesized that: H3,: The contribution of SNG functional value factors, to the explanation of variance in the college choice decision process, is greater than the contribution of the conditional value factors which in turn is greater than social value factors which is greater than the emotional value factors which is greater than the epistemic value factors (i.e., functional > conditional > social > emotional > epistemic)

Based upon a comprehensive literature review of the

college choice decision processes (see chapter one), it is hypothesized that functional and conditional values will contribute the most to the understanding of college choice decision behavior; social values will contribute the third greatest amount to the understanding of college choice

decisions, followed by emotional and finally epistemic

values This ordering will be determined by the ranking of both discriminant weights and discriminant loadings

H,: The SNG model will predict "correct" group membership (at higher than "chance") for a sample of accepted student applicants (In other words, the SNG model can be used to ascertain whether potential students are going to choose to attend or not attend a particular institution)

This hypothesis is perhaps the essential thesis of this Study The rationale for this hypothesis is obvious The purpose of model building in Marketing is explanation and prediction [Hunt 1991] The pragmatic test of the SNG model

is to see if it aids in the prediction of college choice decisions Discriminant analysis will be used on a holdout Sample of the data set to see if the SNG variables better Predict institutional choice than a "chance" categorization Hs: In terms of accurately predicting college choice, the SNG model will outperform a traditional model, using the

same data set

The SNG model is a novel organizing mechanism This model includes factors (such as emotional and epistemic

values) not extensively examined in college choice research Given this additional information and the unique SNG

Organizing method, it is hypothesized that the SNG model

will predict college choice as well as or better than more traditional methods of prediction, when using the same data set The prediction rates of a discriminant analysis using the SNG variables will be compared with a logistical

regression equation using the variables generally looked at

by college choice researchers

Examination and Discussion of Hypotheses

The first two hypotheses are closely related H, holds

that market choice is a function of multiple consumption

values H, states that all the values comprising the SNG model exist in this college choice setting These values (again) are labeled as functional, social, conditional,

emotional, and epistemic Factor analysis is conducted on all 88 variables from the questionnaire and the factors are examined Factors are simply a set of common underlying dimensions that summarize or represent multiple variables [Hair, Anderson, Tathan and Black 1995] Although 23

factors were derived, it was difficult to interpret the

results and label the factors To fully test this

hypothesis, confirmatory factor analysis should be

conducted However, to conduct factor analysis with this research, one would need an unrealistically large sample size of 19,580 Given the practical limitations of

conducting confirmatory factor analysis, the assumption that the values do exist was made, a fundamental premise of the SNG model Factor analysis was then conducted on each

individual section of the questionnaire, across 10 samples Twenty factors were derived and labeled After examining the factor analyses results one can conclude that there is limited support for H, and that H, is inconclusive

As detailed in chapter four, the SNG functional value,

as based on this data set and the individual factor analysis results, is made up of six factors: Quality, Friendly,

Activities, Cost, Facilities, and Crime The SNG social value for this Marquette sample is composed of six factors: Achievers, Rich, Religious, Party, Quiet, and City The conditional value is made up of three factors: Change in

cost, Change in family, and Change in expectations There

are just two factors which make up the emotional value: Joy and Fear And finally, the epistemic value is represented

by three factors: Curious, Interested, and Bored

With support for H limited and H, inconclusive, one may conclude that the values underlying the SNG model did not breakdown as cleanly as expected However, factor

analysis of the individual sections did produce factors that

were later useful in predicting college choice Thus with some qualification, this procedure gives college choice

researchers a new, comprehensive framework to apply at other institutions

Hypothesis three, H, states that each factor

contributes differentially to college choice More

specifically, H,, suggests the relative contribution of each factor based on the college choice literature It was

hypothesized that the order of contribution, in terms of

importance, would be the functional value followed by

conditional, social, emotional, and epistemic value,

respectively Again, the ordering of the values was based upon past research and the logical deduction of the

researcher

Discriminant analysis was now conducted The dependent variable was the decision to attend Marquette or to cancel the application The independent variables were the factors

derived in the previously discussed individual factor

analysis (20 factors total) Discriminant weights and

loadings were examined to determine the relative

contribution of each factor Discriminant analysis was

conducted on the entire sample (n=471) and on a series of 10 random samples in order to ensure the stability of the

results In other words, multiple holdout samples were

selected from the data set to repeatedly test the

consistency of the discriminant analysis prediction Based upon the discriminant analysis results across all samples,

H,, was not supported The actual ordering for this

institutional data showed that emotional and epistemic

values played a much more dominant role than predicted and

social and conditionai values contributed significantly less than hypothesized

The fourth hypothesis, H,, holds that the SNG model will be able to predict group membership accurately Again, the overriding goal of this research project was to

determine if the SNG model could more accurately predict the attend/not attend decision of potential college students than traditional approaches

The discriminant analysis mentioned above was conducted

to test H, The results of the discriminant analysis

support the fourth hypothesis (H,) The SNG model was able

to predict accurately the attend/not attend decision for Marquette University about 81% on average for the 10 sub- samples and approximately 70-83% of the time across all 10 holdout samples This is significantly better than simply assigning membership by chance which would result in a 62% accuracy of classification because 62% of the total sample actually chose to attend

Next is H, It states that the SNG model will

outperform the more traditional method of analysis The

traditional method uses regression and is so designated

because regression is the statistical method so often chosen

to look for associations between assorted variables and

ultimate attendance at a particular college or university

The traditional model used for comparison in this research

was a logistical regression analysis inputing variables

commonly used in past college choice research (i.e., quality

of an institution, cost of tuition, availability of

financial aid, etc.) The final logistical regression model included 16 variables The SNG model did in fact outperform the traditional model for this data set The logistical

regression model classified membership correctly 75% of the time on average (compared with 81% for the SNG model) This

1S significantly less (at 95% confidence level) than the SNG

model Hence H, was supported

Using the SNG Model to Help Understand College Choice

Based on this exploratory study, the SNG model is a more comprehensive framework for understanding and

predicting college choice than has been previously seen in the higher education choice research The SNG model

encompasses many of the traditionally gathered college

choice variables as well as adding new dimensions such as emotional and epistemic values The addition of these

values along with the unique SNG approach of using social and conditional values, provides a more complete and

comprehensive look at college choice decisions

In addition, the pre-specified structure of the

recommended SNG questionnaire allows researchers to gather a lot of information in a way that is fairly easy on the

respondent The conditional value, because of its

contingency format, allows the researcher to test particular changes in variables such as cost, financial aid, programs, etc For example, the respondent can quickly and easily indicate at what point a change in tuition would cause them

to revoke their decision to attend or not attend the

university The SNG format for conditional variables

provides one way to judge the impact of tuition changes

without actually having to change the tuition and suffer the possible negative effects This approach could also be

utilized when determining financial aid amounts or possible changes in programs that are offered, etc

The research approach suggested by the SNG model can be conducted at all stages of the college choice process That

is, a researcher could utilize this model at the "go-no go" stage to determine what values are most important in shaping whether students choose to go to college at all This model can be utilized at the choice set formation stage of the

interested in what values discriminate between groups of students who attend private versus public institutions

And, of course, the SNG model can be applied, as it was in this study, at the final decision stage (i.e., attend or not attend a particular college)

The SNG model can also be used within particular

institutions to shed light on other aspects of student

choice such as the selection of a college or area of study (i.e., nursing vs engineering vs business administration, etc.) In fact, the SNG model could be utilized within each

college to examine discipline choices, such as accounting vs

marketing, or electrical vs biomedical engineering And finally, the SNG model can be applied to graduate as well as undergraduate college choice decisions A particular

different factors, playing different roles for graduate

choice decisions versus undergraduate choice decisions [Webb 1993] Based upon the outcomes of this research, university marketers can adjust their strategies and better meet the needs of their market

Implications of this Research Study

Based on the results of this analysis of the college choice selection process of 471 students, one might

tentatively make the following observations

® The SNG model may be applicable to college choice The five SNG values were not found to independently exist, and further testing would have required an impractical

sample size However, the factors derived, when analysis of individual sections of the questionnaire was conducted (see

earlier discussion), do seem to be helpful to explaining college choice behavior

@ The SNG model suggests that at the choice stage of the college selection process, the emotional value and

epistemic value are more important than choice literature

suggests

@ The statistical analysis of this study suggests

that the SNG model may be a better predictor than regression type analysis using the commonly studied college choice

variables in classifying likelihood to enroll

What does this research tell Marquette University? Are the result practical in nature? In a time of shrinking

resources, both financial and human, universities must

become more efficient and effective in recruiting students Utilizing the SNG model, colleges like Marquette can begin

to reformulate its marketing strategies to better target and segment the marketplace by utilizing the evoked SNG values

to distinguish it from the other institutions in the

student's choice set Marquette can emphasize the values students are most likely to use as discriminating variables between institutions With the knowledge gained from using the SNG model with their own students, colleges and

universities can begin to segment the market and tailor

their educational products, along with their pricing,

distribution, and promotion

For example, given the lack of support for H,,, and the new information that it brings, the university can begin to utilize the prominence of emotional factors to adjust its

marketing strategies Specifically, Marquette could begin

to better leverage the emotional value in its communications with accepted students These communications could stress the joy or positive feelings surrounding the choice to

attend Marquette, while acknowledging and reducing the fear

of making the decision to commit to this institution

Marquette could also reinforce positive emotional feelings once the student accepts and enrolls, in order to minimize the chance of transfer to another university

Although hypothesized to play a minor role, the

epistemic value was also relatively important to the final choice decision concerning attendance at Marquette

University Recall, the epistemic value deals with the

"curiosity or novelty” that surrounds a product or service Based upon this information, Marquette University could

begin to adjust its marketing strategies to stress the

"mystique" of Jesuit education, or to target the "curious" student in order to get them more interested in unique

aspects of Marquette such as the faculty, programs, and

students

Additionally, Marquette needs to develop creative ways

to attract students who have been accepted to come to campus for a visit A campus visit gives prospective students the

"taste" of Marquette needed to enhance the epistemic value and helps them internalize the "fit and feel" dimensions of institutional choice

Marquette can begin to enhance its image with accepted students now Emotional and epistemic values may be easier

to change or adjust in the short term than say, the

Functional or social values To enhance the quality of an institution or to change the recreational facilities, both functional values, takes time and substantial resources Likewise, most students tend to choose institutions that have students similar to themselves [Kohn et al 1976;

Radner and Miller 1970a 1970b], so social change of this variety would be naturally gradual and slow

Because this research suggests functional and social variables are possibly not as important as originally

thought at the choice stage, Marquette may want to stress functional, conditional and social value aspects at the

early stages of recruiting when students are beginning to

applied and is accepted, the Office of Admissions at

Marquette, or any similar institution, could begin to

integrate and stress the emotional and epistemic aspects of the university

Also, because the SNG approach is shown to have an

encouraging (as compared to the logistical regression

approach) prediction rate, this allows Marquette to focus its efforts and resources on those students that are "most likely to attend" Marquette can drastically reduce

marketing efforts to those classified as "likely to cancel"

Not only can this model be used in academic decisions

as discussed, it can be applied to non-academic choice

Situations as well Some of the choices that may benefit from the use of the SNG model involve housing decisions

(i.e., on campus/off campus; dorms/apartments; selection of

resident hall coed/single gender, etc.), club

participation (i.e., academic/social; athletic/non-athletic; etc.), selection and scheduling of courses, volunteer

activities, use of student services (i.e., counseling

social activities (i.e., university sponsored or not;

musical events/sporting events, etc.) Again, during a

climate of limited resources, a university can benefit from the SNG model approach to help monitor choice perceptions

prior to investing time and money into less appreciated

activities The SNG model provides information on all

variables in the marketing mix product, price, promotion, and distribution With it, institutions of higher education can better strategically segment and target their markets effectively and efficiently

Limitations of this Research

Like all research studies, this one has limitations While it provides considerable support for applying the SNG model to the analysis of college choice, cautions are in order First, this is a single institution study It is not clear that these results are generalizable to other

colleges and universities However, there is reason to

think that, while the items composing the SNG factors will

be different across institutions, the SNG model itself is

not applicable and useful

Second, this study was conducted with a sample of

students, not 100% of the student applicants Again,

although the sample should provide a representative profile

of applicant opinions, it is not entirely certain that these results are generalizable to the entire Marquette student

body Further, with the fast pace of change in the urban and educational environment today, it is not clear that

these results reflect the feelings and attitudes of

subsequent incoming classes

Although factor analysis was conducted on all 88

variables and twenty three factors were derived, the results were extremely difficult to interpret Ideally confirmatory factor analysis would be conducted, however, the need for far fewer variables or an unrealistically large sample

(n=19,580), made it impractical for this study Instead, the fundamental premise that the factors do exist was

assumed and the analysis progressed

In terms of the data set, there were some missing data

in particular respondent files Although the amount of

missing data is quite small, it is not clear why the

respondents chose to skip particular questions One

wonders, if they had responded, would the results be

different? An additional concern with this research has to

do with low reliabilities of the factors that make up the values Low reliabilities indicate lack of internal

consistency and may suggest that the items be left as

individual items, not grouped into a factor

And finally, especially with the low response rate for specific financial aid received by respondents, one might conclude that a stronger regression model could have been constructed Perhaps additional variables, such as number

of institutions applied to or accepted at, or quality of athletic teams, could have been included Such variables

could possibly have strengthened the comparative (i.e.,

traditional) regression model One wonders, had this

stronger model been created, would the SNG model still

outperform (i.e., predict college choice better than) the traditional method?

Conclusion

The SNG model appears to be a strong research tool that could be utilized in higher education choice decisions The data required by the model can be gathered on an annual

basis to ensure changes in the external environment are

monitored Adjustments to the marketing strategy can then

be made periodically with shifts in choice determinants

monitored over time Also, individual units on campus, such

as the College of Business, ought to explore the use of the SNG model for input into their strategic plans For

example, the College of Business could utilize the model to determine what factors students use to select a major once enrolled at Marquette One may find that accounting majors tend to utilize functional value items to make their

decision of major, while human resource majors rely mainly

on the social value items In conclusion, based on the data

set from a single institution setting, the SNG model offers

considerable promise as a framework for understanding

college choice decisions

Acknowledgments

Thank you to committee members, Dr Thomas Bausch, Dr David Buckholdt, Fr Michael Kurimay, S.J and particularly

Dr Srinivas Durvasula, who assisted me from various

countries around the world A special and grateful thank you goes to my committee Chair and dissertation advisor, Dr Gene Laczniak who pushed me when I needed it and supported

me all the way

I'd also like to thank others in the Marquette

community who provided guidance and support along the way, such as Dan Gemoll, Dr Jamshid Hosseini and Dr Steve

Crane Thanks to Fr Thaddeus Burch, S.J for his openness

to the Interdisciplinary Program and his patience as I

progressed through

Special thanks to my mother, my family and friends Without such support I could not have been successful This work is dedicated to my father, Edward D Simmons who served

as a Marquette University faculty member and administrator

He always taught me what I needed to know and just a little bit more

TABLE OF CONTENTS List of Tables

Predisposition

Economic Models Sociological Models Search Stage

Final Choice Comprehensive Models of College Choice

Research Methods Applied to

College Choice Decision Making Limitations of the SNG model Utilizing the SNG model

Designing the Questionnaire Data Collection

Data Analysis Use of Multiple Regression and Logit Analysis

Multiple Regression Analysis Logit Modeling

Model Assumptions Limitations of Multinomial

Logit Modeling Conclusion

Response Rate

Data Entry

Testing of the Hypotheses

Data Analysis Specific to the SNG Model

Conclusion

Chapter Four

Introduction

Multiple Values Hypothesis

Do the SNG Values Exist in College Choice?

Is There a Differential Contribution

of the SNG Values?

Can the SNG Model Predict Group Membership

Accurately?

Does the SNG Model Provide a Better Prediction

of College Choice than the More Traditional Models?

Comparison of the SNG Model to

the Traditional Model Conclusion

Chapter Five

Introduction

Summary of the Research

Examination and Discussion of Hypothesis

What does the SNG Model’s Predictive Ability

Imply for Marquette University?

Why Utilize the SNG Model to Help Understand

the College Choice Decision Process?

Limitations of this Research

Suggestions for Use of the SNG Model at

Marquette University and Elsewhere

Conclusion

References

Appendix A Questionnaire

Appendix B Summary of Focus Groups Results

Appendix C Time Line of Returns

Variables Utilized in Previous

College Choice Research Table 1D

Integrative Models of College Choice

Table 4A

Factor Composition for

Functional Value Table 4B

Factor Composition for

Hypothesized SNG Values and

the Factors Within Each Table 4H

Table 4I

Standard Canonical Discriminant Weights for Entire

Sample & Discriminant Weights Across Samples

Is there a significant difference between the

prediction rates of discriminant functions

Introduction

Over the last 40 years much has been published in the higher education literature concerning the college selection process There are many reasons for this growing interest such as: the federal government's financial support of

higher education, a push to enhance the diversity of college student bodies, a shift in demographics which has changed the pool of college students, as well as a reduced number of college applicants [Paulsen 1990] Most recently, the

increased competition among colleges for potential students makes it crucial for colleges and universities to understand the selection process in order to develop better recruitment strategies [Kotler and Fox 1985] Such understanding will enhance the appropriate academic credentials of

matriculants, better match the student/institution fit and reduce the number of transfer/drop out students Therefore, colleges and universities have an obvious vested interest in understanding this decision process To this end, many

researchers have attempted to enlighten university

administrators about college choice Below, this body of research is examined, and a classification of the research

is prepared The other reason this information is to be reviewed is because it sets the stage for the introduction

Specifically, this review lays the groundwork for the

application of a consumer consumption model to the area of college selection

This general research review is organized in two main categories The first involves a distinction between macro and micro research Macro research deals with larger

environmental issues affecting college choice and micro

research is geared toward variables affecting individual institutions The second distinction or organizing

mechanism is "by stage" in the college choice process This

refers to the point a student is at in the decision process The stages are first, deciding whether or not to attend an institution of higher education; second, building a choice set of suitable institutions; and finally, the decision of which institution to attend

Macro VS Micro

The first distinction regarding the college choice

Macro level studies

Such research pertains to studies that utilize

aggregate student groups and focus on the environmental

effects associated with the total enrollment for all