AN INSTITUTIONAL STUDY OF CANCELLED ADMISSIONS Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY ELLIS STEBBINS HAMMOND 1972 I/ I T [Ii/j ,LLWEWMW W , \r Diggifiga: ‘ autism 94180 . Univusity -:—-———-- mm m. 51392 3 :ees- ABSTRACT AN INSTITUTIONAL STUDY OF CANCELLED ADMISSIONS By Ellis Stebbins Hammond When high school students apply to college, the pressures to be accepted generate applications to several different colleges. The resulting multiple applications create several problems for individual colleges, because the student will finally pick only one college to attend, however many may accept him. The other institutions receive a cancelled admission. To understand more about the reasons for cancellations and also to describe the type of student who may be most typical of those who cancel, data from students accepted as first-term freshmen at Michigan State University were analyzed with a multiple regression technique. Data derived from a content analysis of students' letters of cancellation were tabulated, as were the variables used in the multiple regression analysis, to generate a list of the frequencies of individual responses. Ellis Stebbins Hammond A post-card questionnaire provided responses used in a follow-up study . The results of these analyses showed no particularly sig- nificant differences between students who enrolled and those who cancelled when high school grade-point average, financial need, age, and other data available from the students' records were compared. The major findings are as follows: * A slightly larger percentage of the more academically able students tended to cancel their admission. Seventeen percent more of the B or better students cancelled than enrolled. More students receiving no offers of financial help from MSU cancelled their admission than enrolled (74. 0% vs. 25. 9%). More high-need students ($1000 to $2000) cancelled than enrolled. A sizable proportion (46. 3%) of those admitted cancelled their own enrollment. The University cancelled about half (53. 7%) of all who did not enroll. Many cancellations (74. 8%) came from state residents. Ellis Stebbins Hammond In retrospect, more detailed analyses of human behavior should be used to investigate the characteristics of the student who cancels. Although distance from home to college, costs, and preferences for a smaller rather than a large college were often cited by students as reasons for cancelling their admission to MSU, the same reasons were given by students in obviously different circumstances. "Distance" was used both by in-state residents living less than fifty miles from the MSU campus and by residents of California; "costs" were cited by students who never applied for financial’aid; and "size" was listed by students who apparently did not know about the three residential colleges within the University. In this study, at least, students' responses showed so little con- gruence with objective facts available that definitions of the reasons given must be considered relative rather than absolute measures. AN INSTITUTIONAL STUDY OF CANCELLED ADMISSIONS By Ellis Stebbins Hammond A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Administration and Higher Education 1972 For Cindy ii ACKNOWLEDGMENTS Like the iceburg, this dissertation shows only a small portion of its entirety. First of all, the MSU Admissions Office provided major support by making available the data on which this study is based, and I am grateful for this assistance. Many people contributed generously of their time and effort. Walter Johnson was perhaps the greatest single contributor in making this study assume its present form. His reactions are primarily responsible for the content and methods. James H. Stapleton, Chairman of the Department of Statistics and Probability at MSU, gave very valuable advice. Sheldon Lowry spent much time in patient review. Mike Gordon, a research consultant in the College of Education, helped by structuring questions that led to the use of the multiple regression analysis. Gene Rex offered counsel and advice that generated a more cohesive topic. Ed Blackman offered his usual thoughtfulness. Nancy Hammond gave editorial advice on the manuscript. And Zelma‘ Paine continually insisted that I finish my work. iii TABLE OF CONTENTS LIST OF TABLES CHAPTER I. COLLEGE ADMISSIONS AND MULTIPLE APPLICATIONS Background . . The College, the Student, and Multiple Applications . . . . The High School and the College The Student Who Cancels II. REVIEW OF THE LITERATURE . III. POPULATION, PROCEDURES, AND SAMPLES . Methods of Analysis Student Records Classification of Data . Coding Frequency Tabulations Questionnaire Summary IV. ANALYSIS AND RESULTS Analysis of Cancellation Letters . Multiple Regression Analysis Results Post- card Analysis iv Page vi 10 12 20 30 32 32 33 35 39 39 41 43 44 46 48 53 CHAPTER Page Frequency Tabulations . . . . . . . . . . . 53 Grade -point Average . . . . . . . . . . 58 Rank in Class . . . . . . . . . . . . . 58 Financial Aid . . . . . . . . . . . . . 59 Financial Need . . . . . . . . . . . . 60 Value . . . . . . . . . . . . . . . . . 61 Timing . . . . . . . . . . . . . . . . 62 Sex . . . . . . . . . . . . . . . . . . 64 Residence . . . . . . . . . . . . . . . 64 Age . . . . . . . . . . . . . . . . . 66 Distance . . . . . . . . . . . . Size of Town . . . . . . . . . . . . . . 68 Summary . . . . . . . . . . . . . . . . . 68 V. SUMMARY AND CONCLUSIONS . . . . . . . . . 70 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . 80 APPENDIX.......................88 Table 10. 11. 12. 13. 14. 15. LIST OF TABLES Classification of Students Admitted . Students' Reasons for Cancelling Admission In- state Students Who Applied for Aid In-state Students Who Did Not Apply for Aid Out -of-state Students Who Applied for Aid All Who Applied for Aid All Who Did Not Apply for Aid . Follow -up Reasons for Cancellation Primary Reasons for Cancellations Summary of Responses to Questionnaire Enrolled vs. Cancelled Students, by GPA . Enrolled vs. Cancelled Students, by High School Class Rank . . . . Enrolled vs .. Cancelled Students, by Type of Aid Offered Enrolled vs. Cancelled Students, by Analysis of Need . . . . . . . . Enrolled vs. Cancelled Students, by Amount of Need Met by MSU . . . . . . vi Page 43 46 49 50 5 1 52 52 54 55 56 58 59 60 61 62 Table 16. 17. 18. 19. 20. 21. Month of Cancellation, by Percent of Those Cancelling . Enrolled vs. Cancelled Students, Enrolled vs. Cancelled Students, Enrolled vs. Cancelled Students, Enrolled vs. Cancelled Students, from Home to Campus Enrolled vs. Cancelled Students, Home Town vii by Sex by Residence by Age by Distance by Siz e of Page 63 64 65 66 67 68 CHAPTER I COLLEGE ADMISSIONS AND MULTIPLE APPLICATIONS Today' s colleges and universities are receiving more applications for admission than ever before in the history of higher education. 1 The community college absorbs some, technical institutes enroll others, but the four -year college--primarily the large publicly- supported university-~admits the majority. Colleges with the most applicants are well known and receive applications from more students than they can accommodate. 1"Getting into College: Chances Now, " U. S. News and World Report, Vol. LXVIII, No. 12 (March 23, 1970), pp. 28-30; For Your Information (Washington, D. C. : Office of Institutional Research, National Association of State Universities and Land- grant Colleges, 1970). 2See, for example, John C. Hoy, Choosinga College (New York: Dell Publishing Company, 1967), pp. 213 -226; B. Alden Thresher, College Admission and the Public Interest (New York: College Entrance Examination Board, 1966), pp. 55 -58; Kate Hevner Mueller, "Admissions, " in Student Personnel Work in I-Iggther Education (Boston: Houghton Mifflin Company, 1961), pp. 149-157; and James R. Spence, "Whither Admissions Research, Not Whether, " College and University, Vol. 43, No. 3 (Spring, 1968), pp. 348-352. Theadmission staff must usually accept, reject, or postpone a decision on each applicant in keeping with the goals and facilities of each institution. The resulting need to select some students over others creates several kinds of problems, one of which is the problem of prediction: Of those applicants accepted, how many will actually enroll for the coming academic year? Several solutions have been proposed to help the student find a place in college and, at the same time, help colleges be more certain about the numbers of students who will finally enroll- Many of these suggestions--lotteries, "open" admissions, expanded state systems that assure some form of higher education for all resi- dents:3 --are still being debated and remain far from implementation. Until new methods are tested and verified, many colleges will continue to face the difficulty of too many applications for the open- ings available; and the related problems of forecasting enrollments will remain. The purpose of this study is to examine cancelled admis- sions at one institution and develop a predictive method for helping the admissions staff identify students likely to cancel their 3The California system, for example. See also Rixford K. Snyder, "Developing Nationwide Standards: Admissions, " in Logan Wilson (ed. ), Emerging Patterns in Higher Education (Washington, D. C. : American Council on Education, 1965), pp. 222 -224. admissions. So informed, the admissions office may be able to adapt, adjust, and respond more accurately to the publics it serves and the institution to which it reports. Background The process of admissions operates between the two worlds of its academic family and the public as a highly visible unit that must respond to the interests of each. To work at all, it must serve both; and to work well, it should serve both equally. In one domain, our industrial and highly technical society makes ever—expanding demands for the educated. At all levels, the citizen seeks educational opportunities for reasons of equality, progress, prosperity, and status. Today's student, who may seek entry, promotion, or advantage, is encouraged, if not actually required, to attain ever higher educational and technical competence. As the public seeks higher education, it often becomes involved with the admissions policies that govern entrance. Whether concerns about college admissions are locally, nationally, or privately centered, the demand has one common side-effect: Those who seek entry frequently compete for the limited places available. In the other realm, the college must confront its own set of problems. Because facilities and funds are limited, a college must restrict the number of applicants it accepts. It may wish to select students from among representative groups of, for example, non-residents, cultural minorities, or those with certain academic interests. It must also enroll enough students to assure the full utilization and efficient operation of the physical plant and staff. Given the responsibility for enrolling adequate numbers, the admis- sions office must evaluate and select the proper number of students from among all who apply. However, applications present a greater problem for the college than for the student, because the student usually has the single goal of entry, while the college must satisfy complex external demands and internal requirements. As a result, the college and student become shoppers in the two worlds of higher education. These acts of seeking and sorting generate the problems that are indicated by the admission figures reported by colleges nation wide. They show three distinct and diminishing sets of numbers for applications, acceptances, and enrollments. 4 For the college these figures represent the large number who apply, the smaller number 4College Admissions Handbook (Princeton, New Jersey: College Entrance Examination Board, 1969); see also Robert C. Anderson and Ellis S. Hammond, "Admissions at Michigan Colleges, 1967 -1968, ” Career Opportunity Guide III (East Lansing: Michigan State University, Institute for Community Development, 1968), pp. 26-43. who meet the criteria for admission, and the still fewer who finally enroll. For the student, the process is somewhat the same, because applications may be distributed to a number of colleges, two or three may accept him, and he enrolls at one. The problem is resolved for the student when at least one college accepts his appli- cation. The college, however, remains uncertain that the admitted student will actually appear on campus, and many colleges keep "paper" students in their files. To understand why students submit more than one application and why colleges may themselves encourage multiple applications, one must first examine the influences at work on both. The College, the Student, and ' Multiple Applications Multiple applications contribute directly to succeeding cancellations and the prediction problems for the admissions office. The problems come from several different sources. Parents often emphasize the necessity of attaining certain goals, and their children (who frequently absorb parental valuess) accept education as a necessary means. The business world often makes a college 5Robert C. Anderson, Russell G. Mawby, Joe N. Miller, and Andrew L. Olsen, "Parental Aspirations, " Adult Leadership (May, 1965). degree necessary for entry and promotion, and high school counselors advise students to go to college. This total emphasis on the need for higher education creates a desperate flood of applications . College admissions staffs also must assume some responsibility for encouraging multiple applications and subsequent cancellations. When talking to parents, students, and counselors, college representatives are sometimes vague about admissions criteria. 6 Some ambiguity may be unavoidable, because admis - sions staffs want to select from among the best applicants; yet they also know that standards for admissions may change at any time: A state legislature may change the level of appropriations to a public college and thus affect the number of students who can be accommodated. A specific policy (a quota on out-of-state students or a minimum high school grade-point average) may reduce the flow of applications and acceptances in a way that requires an immediate (and hence unanticipated) response. Cut- off dates or other general criteria for admission that were. established. for a given year may have to be modified to account for curricular changes, the influence of alumni, or a changing 6Robert L. Amsden, "Good and Bad Admissions Practices as Seen by High Schools, " Journal of the Association of College Admissions Counselors, Vol. 10 (Summer, 1965), pp. 3-6. drop -out rate. While many educators make honest pleas for more precise descriptions of admissions standards, 7 their reduction to "absolutes" is almost impossible as the two worlds of external forces and internal demands act upon the admissions office. That many colleges and universities are residential also influences admissions. Reasonable standards (variously interpreted) of academic achievement have to be maintained, but the full use of academic personnel and the physical plant are practical considera- tions that cannot be disregarded. 8 Experience shows that some accepted students never attend and that even the most careful elaboration of admissions criteria still produces applications from. some who cannot be accepted. As a result, in visits to high schools, participation in college -night programs, and in other prescribed encounters, admissions personnel are essentially seeking, encouraging, and soliciting applicants. This duty, among others, is precisely one for which they are hired. 7Richard W. Mecham, "Please Put an End to the Admis- sions Process Rat Race, " College Board Review, Vol. 74 (Winter, 1969-1970). pp. 14-15. 8"Off-campus Policy Reflects Dilemma of Residence Halls, " [Michigan] State News, March 11, 1970, p. 9. ( 9Hubert E. Mate, "Role of the Admissions Officer, " College and University, Vol. 44, No. 4 (Summer, 1969), pp. 705- 708; 5d David W. Peterson, "How High School Visits Influence College Admissions, " College Board Review, Vol. 68 (Summer, 1968), pp. 8-9, 18. Multiple applications are not a large problem for all colleges. Certain small and little -known colleges advertise for applicants, join applicant "pools, " register with agencies that help students get into college, and are listed in national publications as having freshman places available. 10 Forecasting enrollments is most difficult for the colleges of national reputation, who participate in multi -state recruiting efforts and use highly selective admis - sions criteria. In short, the problem of multiple applications is most common to popular and well -known institutions. 11 A college receives more applicants than it can accept for reasons that are as varied as the influences that lead a student to apply. The college may be known for its football team, academic reputation, research activities, or the "personalities" on campus. Often, t0p -ranking high school students receive what amounts to 10”Eighty New Colleges, " Changing Times, Vol. 17 (November, 1963), p. 13. The College Admissions Center, established in 1957 by the College Entrance Examination Board and cooperating colleges, has offices in New York City and Chicago. Reports, covering the Center' s activities and students processed, are issued annually and published in the College Board Review. See also, "The Clearinghouse Approach to Undergraduate Admis- sions, " Journal of the Association of College Admissions Counselors, Vol. 8, No. 4 (Spring, 1963), p. 12. 11Hoy, Choosing a Colleg_e_; Thresher, Collgge Admission; Mueller, "Admissions"; and Spence, "Whither Admissions Research?" direct-mail advertising from colleges who wish to recruit the academically able. Every college wants the "best" students, and, when the Merit Corporation publishes a list of test winners or state scholarship examination results are made known, the colleges and the students they seek enter the market place. Parents may influence their child to submit an application to their alma mater, or friends already in college may motivate a high school senior to apply there too. A visit to a campus may lead a student to apply before acquiring any real knowledge of admissions criteria. A wide variety of simple activities also may result in an application: leafing through college catalogues, attend- ing a college-night program sponsored by the school, or simply receiving a letter from a college that has recognized the student' 8 scholastic achievement. The reasons for filling out the application form are virtually endless. Basically, however, only two kinds of students create the largest proportion of multiple application problems. At one end of the academic scale may be the very able high school student who is searching for the "best deal. " These student "shoppers, " whether they be National Merit Scholars or High School All -Americans, may receive a number of offers because they are academically capable or they have a desired skill. They seek the best financial opportunity 10 to attend college and finally enroll at an institution that provides the most attractive offer. At the other extreme are the students who simply do not meet admissions criteria, and whose application is really a "shot in the dark. " They of course may be sincere in their wish for education; perhaps they even know that they do not meet the requirements. Nevertheless, the result of their often unrealistic aspirations is an additional burden of multiple applications for the admissions staffs in the colleges to which they apply. In between are the vast bulk of students who make an application, are accepted, and enroll at one college. These stu- dents do not contribute quantitatively to the multiple application problem because they usually are familiar with entrance require- ments, have visited the campus before submitting the application, are certain of their choice, and can be encouraged by both college and secondary school personnel. They usually apply to, and attend, the one college they have chosen. 12 The th School and the Colleg_e The relationship between college and secondary school also probably contributes to multiple applications. The college ¥2American Council on Education”, National Norms for Entering Colle_jge Freshmen--Fall, 1969, Research Report Vol. 4, No. 7 (Washington, D.C.: By the Council, 1969). 11 depends on the high school counseling office to recommend prospective students, while the high school needs the support of colleges through the acceptance and placement of its students. Both ideally must coordinate their judgments and agree that particular colleges are the best choices for particular students and then work together toward this goal. Few high school counselors, however, can be certain that the student guided to a particular college will be accepted, so applications to several colleges may be encouraged. Neither can the admissions officer be certain that the accepted student will actually enroll for classes, so that he may encourage more applications and accept more students than the college actually needs. Between guidance at the secondary level and decisions at the college level, varying degrees of uncertainty, poor communica- tion, misunderstood and misinterpreted facts, social pressure, and aspirations constantly operate. For reasons of social prestige, for example, parents may insist that their child apply to a certain college without regard for its entrance criteria and thus disrupt the most conscientious efforts of the school counselor. The counselor may be but partially informed about admissions standards, biased in favor of (or against) certain colleges, unaware that a college' 3 admissions standards have changed, or too busy with other 12 responsibilities, and hence he may misdirect students. Counselors also promote multiple applications when they urge a student towards a college whose program is too strenuous for his abilities or whose costs are beyond his parents' financial means. On the other hand, some college admissions staffs only sporadically inform their high school clientele about admissions requirements or changes. Col- leges send representatives into high schools to recruit candidates for admission knowing full well that they cannot possibly admit all whom their staff may see. The major participants in helping students from school to college thereby often "over -sell" their perceived responsibilities. The Student Who Cancels Some students change their minds after a visit to a campus. Having applied and been accepted, they may discover that the "real" place is something less (or more) than they had anticipated. Per- haps the original attraction of a good mathematics department becomes less significant in the light of poor dormitory accommoda- tions. Or, conversely, the potential competition from a student body of 20, 000 or more may be too threatening to a student from a small town. The campus may seem too big or the town in which it is located too small, or it is too far from home. Even the physical 13 setting (or lack of it!) can lead some students to cancel their admissions. As mentioned earlier, certain very bright high school students openly look for financial advantage to attend college. The college that offers only a job or a loan to this student may be the one that gets a cancellation letter in favor of the institution that has awarded an unrestricted scholarship. There are also students who simply hope for a financial "break. " They know at application time that they will need con- siderable assistance to attend an out-of-state or private college. If the college cannot meet their financial needs, these students have little choice but to cancel their admissionand attend a college whose costs are more financially realistic. The matter of "ranking" is also involved in applying to college. Thetsecond- and third -choice colleges will usually be rejected if the first-choice college accepts the student. The admis- sions office, of course, never knows what place it holds on the student' 3 preference list and therefore must treat each applicant as one who may enroll. There are perhaps as many other reasons for cancelling as there are reasons for applying. A parent may change jobs and residence, a girl -friend or boy -friend might transfer from the 14 college originally chosen, a late appointment to a military academy may change plans, family circumstances may require the student's presence at home, a newly-opened community college may seem more convenient, or the lure of a good job may alter previous plans to attend college. The admissions office has no way of anticipating the appli- cant' s changing aspirations, nor can it allow for changes in the student' 3 personal life. Each applicant must be considered a prOSpective student. Consequently, the admissions office must predict future enrollments on the basis of intangible and often unforeseeable factors. The result is often inaccurate and some- times unsatisfactory enrollment plans. In actual operation, the system of admissions is inefficient and imprecise. In any given term, colleges may find that they have more students than they were prepared to accommodate or less than they hoped for. International relations, domestic fiscal policy, and the public' 5 reaction to current events may all change the number of families interested in a particular college. A college department may terminate a program or make substantive changes that affect enrollments. Colleges can never be any more certain 13Simon J. Gormley, "How High Schools Can Help in Admissions, " College Board Review, Vol. 75 (Spring, 1970), p. 24. 15 of the numbers they will attract than can students be sure of their acceptance. The small college may suffer more drastically than the large university when conditions change enrollments, but the problem remains for all. The difficulty of multiple applications and its resulting ambiguities could be reduced if certain specific differences between students who enroll and those who do not were discernible. The present system typically uses previous enrollment and cancellation figures as a guide for the future, but these figures cannot account for changes and hence may produce misleading or inaccurate pre- dictions. Knowing more details about cancellations should make these predictions more accurate--the assumption being, of course, that distinguishable features can identify the student likely to cancel his admission. Furthermore, a method that predicts cancellations would allow the admissions staff to adjust acceptance rates. Too often, in order to adjust enrollments, admissions personnel have altered their methods in response to a ”crisis. " When selection criteria change, a student applying late may be admissible, whereas a classmate with comparable qualifications may have been refused admission at an earlier date. One mark of professionalism and a stable public image surely must be a relatively consistent pattern 16 of standards. While no admissions staff can predict what influence major policy changes may have, it can attempt to realign its review process at a pace that both assures desirable enrollments and also serves its public' 3 interests. The net effect of major and sudden changes in admissions criteria is mistrust by the secondary school and further confusion of the public. The student, parent, and high school counselor need the advantage of the best and most accurate admissions information consistent with institutional policy, goals, growth, diversity, and service. One element of this more truthful message can be supplied if the admissions staff has precise informa - tion about cancellations. Admissions officers then know what will happen to a given percentage of applicants and thus need not over- sell nor under-sell their institutional wares. Factors affecting the choice of a college have been docu- mented and studied, but probably no single device will serve as a final and perfect predictor of who will go where. 14 The present system of admissions, nevertheless, can be improved. Onesuch 14Arthur Dale, "Reported Determinants of Educational Choice, " Personnel and Guidance Journal, Vol. XLII, No. 6 (February, 1964), p. 564; Ralph F. Berdie and Albert B. Hood, "Personal Values and Attitudes as Determinants of Post -High School Plans, " Personnel and Guidance Journal, Vol. XLII, No. 8 (April, 1964), p. 754; and Lorraine S. Hansen, "Helping Students in Pre- College Decision Making, " Journal of the Association of College Admissions Counselors, Vol. 12, No. 3 (19677, p. 4. 17 improvement, which can best be accomplished within the admission operation itself, is the acquisition of additional data about the stu- dent who does not enroll. Better predictions about enrollments and cancellations might thus be obtained. If the admissions office had information about cancellations, what results could reasonably be expected? Three ramifications are possible: (1) The college could accept more students or it could reduce the number accepted as a higher or lower percentage of cancellations occurred. Faculty and staff could be appraised of normal or fluctuating conditions and adapt accordingly to meet immediate needs for housing, classroom space, faculty, or laboratory facilities. For long -term actions, the college might consider changing its over -all admissions policy--recruitment practices, for example, or participation in college-night programs-~to adjust to cancellations. A high rate of cancella- tions by out -of-state, urban, or high -ranking students might signify a need for greater recruiting efforts to attract those youth if the college deems them important. In other words, general admissions practices might be reviewed, curricula revised, and institutional policy changed. 18 (2) The characteristics of students who tend to cancel should be made known and discussed with secondary schools and other groups who supply the college with students. It should, in other words, become public information. Because colleges depend on the public, as we have seen, these consumers must have specific information about requirements and needs, about students who attend and those who do not. There should be no "secrets" in the admissions office. (3) One of the most important reasons for studying the student who cancels lies within the domain of the admissions office itself. Most admissions personnel spend considerable time making individual decisions on applications. When these efforts do not culminate in a student enrollment, they are tempted to wonder whether their energies are being dissipated. While any college might justify efforts to attract and retain the academically able student, we might also ask whether such expenditures are valid and what goals they serve. Does the admissions office really know its clientele? To what segment of its public is it most responsible? No responsibility can be perfectly executed, but the admis- sions office' 3 role is often measured by the number of "qualified" students it enrolls. The effectiveness of that role is open to 19 question if approximately half of the accepted applicants never enroll. 15 15College Admissions Handbook; and Anderson and Hammond, Care-e? Opportunity Guide 111. CHAPTER II REVIEW OF THE LITERATURE Literature about college guidance, admissions, and predicted college success is abundant, but the related problems of making a choice about college and changing student aspirations are sparsely represented in research and in studies of the transi- tion from school to college. For whatever reasons, those investigating the college -bound student have chosen to report the more obvious aspects of going to college. Perhaps, like the ice- berg, the most easily examined is that which is most visible. A search through more than ten years of professional journals, texts, dissertations, .and seminar reports uncovered only one clearly labeled study of admissions cancellations. Several writers talked about multiple applications as theyreported other data or mentioned confused admissions standards and related topics, but only Jones and Siegel1 of Miami University in Ohio made a 1R. L. Jones and L. Siegel, "Admissions Cancellation: An Institutional Study, " Collgge and University, Vol. 38 (Fall, 1962). PP. 89-95. 20 21 concerted effort to learn more about the student who is admitted and then cancels. Perhaps admissions staffs merely accept multiple applications and cancellations as part of their normal work routine. However, one is also tempted to conjecture that cancella— tions, as a particular aspect of admissions, carry such overtones of ego-threat and institutional rejection that they are just too embarrassing to confront with detailed study. The object of the Jones and Siegel study was to investigate three main problems: 1. What are the characteristics of the student who cancels? 2. What reasons for cancellation were given? 3. Where did those who cancelled go? To answer these questions, Jones and Siegel sent a ques- tionnaire to parents and examined the records of the 884 students in their sample. Parents' responses indicated that 60 percent of the applicants who cancelled favored another college as their first choice. Although the questionnaire asked parents for the reason for cancellation, the four options listed limited the response choices to financial reasons, admission to first-choice college, Miami per— ceived as too difficult, or a wish to be nearer home. Furthermore, "their comparisons of enrolled and non -enrolled students were restricted to male -female and Ohio -non -Ohio classifications. They 22 also used only general descriptive categories, such as "financial difficulty” and "geographic factors, " to classify the student who cancelled. Every subsequent investigator should profit from the flaws and errors of those who have gone on before, and while no one could dispute the need for examining the student who cancels, certainly Jones and Siegel have both contributed. to and at the same time left unanswered particular components of the over -all problem. Their approach has helped make this investigation more thorough. As noted in Chapter I, probably no more compelling reason leads a student to apply to more than one college than the general contemporary cultural "press" for higher education. All parties involved--parents, counselors, and admissions staff --contribute to the problems of multiple applications and the inevitable can- cellations to some colleges. In the reviews reported below, the investigator most often chose only a single element of the complex problem for research. Each one helps explain at least one part of the total, but, in combination, they should be viewed more as a synthesis of representative factors than as an exhaustive docu- mentation of every reason that may lead a student to apply, then cancel. 23 Ruch2 studied 100 high school students using the College Plans Questionnaire and concluded that short visits to the candidate' 3 chosen campus were less effective indicators of stability of choice than longer, more carefully planned tours. He also found some relationship between the activities of talking with a college repre- sentative and visiting the campus and the stability of college selection. Peterson' 5 study3 presents data showing no positive correlation between college representatives' visits to high schools and subse- quent matriculation at a specific institution. Hills4 applied a decision -theory technique at the secondary school level in order to helpstudents narrow their college choice. His method included assigning values, through multiple regression techniques, to test scores, high school grades, and college grade averages. By his own admission, however, the device proved "cumbersome" and time -consuming and could not be empirically validated. 2Charles P. Ruch, "Participation in Selected Guidance Activities and Stability of College Choice, " Journal of the Associa- tion of College Admissions Counselors, Vol. 12, No. 4 (September, 1967), pp. 2‘1-24. 3David W. Peterson, "How High School Visits Influence College Admissions, " College Board Review, Vol. 68 (Summer, 1968), pp. 8-9, 18. 4John R. Hills, "Decision Theory and College Choice, " Personnel agd Guidance Journal, Vol. XLIII, No. 1 (September, 1964). pp. 17-22. 24 Gladney5 contended that fees were a factor in determining one' 3 choice of colleges. Her presentation argued for higher initial application fees as a deterrent to those applicants who might other- wise make several applications to several colleges. In 1967, Mickelson6 investigated the hypothesis that parents' education and income affect a student' 3 choice of college. His questionnaire was distributed to 2, 580 freshmen at four colleges. He concluded that students select a college for more "intrinsic" reasons as parental income and education increase. Perhaps the most ambitious undertaking on the subject of college choice is a study done for the College Entrance Examination Board by the Center for Research and Development in Higher Edu- cation at Berkeley. 7 A 1968 report gave the results of questionnaires distributed over a seven-year span to 90, 000 students in four states. (More results will be published periodically as these students move from high school into various occupations and educational AA... A “4 A A A A v y—-— 1 v1 ' j 'T A A4. v-v—w'fiw v—r. 5Marilyn Bush Gladney, "The College Application Fee as a Factor in Percent of Accepted Applicants Who Do Not Enter, " Collige and University, Vol. 42, No. 1 (Fall, 1966), pp. 63-66. 6Howard 1. Mickelson, "Parental Factors Influencing College Selection, " Doctoral dissertation, Colorado State College, 1967. 7Dale Tillery, et a1. , "A Study of Student Decision Making and the Outcome, " in SCOPE (New York: College Entrance Examination Board, 1969). 25 institutions.) Item responses were tabulated for five aspiration levels, and high correlations resulted for aspiration, achievement, and family socio -economic levels. The report includes detailed responses to 69 different questions, about half of which are related to college choice. None, however interesting the study may be, sheds further light on why a student chooses one college rather than another. A more pertinent study of the perplexing problem of multiple applications was carried out at the University of Georgia by Donald Irvin. 8 His particular work involved a post-card and letter to 410 persons who cancelled their admission. Irvin con- cluded that the majority of the respondents would: (1) attend another college (rather than work or enter the Service), (2) for four years (rather than enroll for two years at a junior or com- munity college), (3) because of a "superior" or ”more suitable" program. Other tabulated entries show finances, social factors, and location as significant determinants of cancellation for a minority of students. The problem of multiple applications is clearly a major one because the term "college -choice" appears with great frequency 8Donald W. Irvin, "Ghost Applicants at the University of Georgia, " Collgge and University, Vol. 40, No. 1 (Fall, 1964), pp. 55-59. 26 in many works. For example, the College Board Review of Fall, 1969, lists 84 separate works published in 19 journals for the year 1968-1969 that deal with students going to college. 9 Nineteen items, more than any other single topic, relate to factors surrounding choice of college. Several noteworthy documents have focused attention. on the selection of a college but have not specifically mentioned the "ghost" problem for the admissions office. Thresher did so in his College Admissions and the Public Interest10 when he noted, "There is extensive literature on the theme of ' choosing a college, ' or 'how to get into college' " (p. 28). ". . . The student. . . becomes caught up in the process of choice, selection, comparison, and competitive differences among colleges . . ." (p. 36). He then noted seven college "interventions" that may influence the student; among them, ". . . Propinquity, is probably the single most powerful influence on college selection" (p. 37). In a related review of admissions problems for the 19603, the Educational Testing Service noted that "both school and college 9Stanley H. Cramer and Richard R. Stevic, "Research on the Transition from School to College. A.Review of the Literature, " College Board Review, Vol. 73 (Fall, 1969), pp. 22 -30. 10B. Alden Thresher, College Admission and the Public Interest (New York: College Entrance Examination Board, 19667. 27 are finding that, as the number of college applicants increases and as more students submit multiple applications, the preparation and utilization of information about school performance are becoming difficult. . . ."11 The article then goes on to propose a centralized records system as one means for dealing with the multiple appli- cations phenomenon. Another related and interesting study was carried out in the Fall of 1969 by the American Council on Education. 12 Among an array of items reported for 169, 190 freshmen at 270 institutions was one entry concerning the number of applications made by students attending Michigan State University. It showed that 50. 1 percent of MSU' s freshmen applied to no other'college, while 46. 2 percent nationally had not appliedto a college other than the one they were attending. Furthermore, 26. 8 percent applied to one additional institution, 15. 6 percent applied to two other colleges, and 6 percent had applied to three; in otherwords, almost half (48. 4%) applied to at least two colleges. No reasons for their final choice were listed. 11Frank Bowles and Richard Pearson. (eds. ), Admission to College. A Program for the 1960' s. (Fifty-eighth Report of the President.) (Princeton, New Jersey: College Entrance Examination Board, 1962). 2American Council on Education, National Norms for Entering Collegg Freshmen--Fall, 1969. Research Report, Vol. 4, No. 7 (Washington, D. C. : By the Council, 1969). 28 Possibly the most significant question uncovered in the literature review was asked at the end of a data presentation in a publication from the American College Testing Service. 13 Having shown tables and charts of freshmen characteristics, the author asked, "Are there differences between enrolled and non -enrolled students which can help to explain why the latter failed to enroll?" (p. 32). This question is most appropriate to this study. From these highlights of a decade of literature related to the subject of multiple applications, the influences that are most apt to cause them, and the three studies that focus on cancellations (Jones and Siegel, Gladney, and Irvin), certain dimensions of the problem become clear. We need more information about the dis- tinctions between enrolled and non ~matriculated students, and the reasons already uncovered for cancellation should be explored in depth. The following chapter, which outlines an analysis of the student who cancelled, may help the college, the student, the counselor, and the parents in their individual and collective roles as decision makers. It is the process of selection (and rejection) 13Donald P. Hoyt and Leo A. Munday, Your Colleg_e Freshman. Interpretive Guide to ACT Research Services for Higher Education (Iowa City, Iowa: Research and Development Division, American College Testing Service, 1968-69). 29 that concerns us here, not the final act of high school graduation or the introductory fact of matriculation at college. CHAPTER III POPULATION, PROCEDURES, AND SAMPLES The data for this study came from the Michigan State University admissions office, registrar! 3 reports, and cor-- respondence from students applying as first-term freshmen. Individual applications from students and tabulations of the number applying, admitted, enrolled, and cancelling were the primary sources of the information assembled. Records kept by the admissions office already contained some facts about cancellations. Reports indicated, for example, that about one -third of all applications came from out-of-state residents, that the University retains and enrolls about one- third of all out -of-state residents accepted for admission, and that the University cancels the enrollment of a large group of out -of-state residents in May because they had not paid the Advance Enrollment Deposit. 1 1As noted earlier, the Advance Enrollment Deposit is required of all newly -admitted students. Failure to pay this fee results in the automatic cancellation of the student' 3 enrollment. 30 31 The experience and observations of the admissions staff were also used in the design of the study. For example, some staff members speculated that most of the cancellations came from out -of-state residents who were simply "shopping" for the best financial opportunity and would therefore enroll at the college that offered them the highest monetary award. These speculations were used as guidelines for the research, however, rather than as hypotheses to be tested. Indeed, while experience can provide some insights, a study of this type calls for specific measures if it is to be at all helpful. Therefore, the first step-was to develop some systematic procedure for learning more about cancellations. Consultations with the admissions staff to determine what kind of information would be most helpful in forecasting enrollments and talks with the University' 3 Evaluation Services personnel and with the research personnel in the College of Education produced three general approaches: 1. A review of records, statistics kept by the admissions office, students' folders, and letters of cancellation to obtain data for analysis; 2. (a) A multiple regression analysis of the data obtained to determine whether students who enroll differmeasurably 32 from students who cancel, and (b) a simple frequency tabulation of the variables selected; 3. A post-card survey of students who cancelled their admission to check on the reasons for their action. Method 5 of Analysis As a first step, student records from 1969 were reviewed and studied with certain questions in mind. 2 These questions, at first based on experience and observation, became more specific as information and data were accumulated. As noted earlier, the first task was to isolate both descriptive and quantitative factors that would separate and identify students who cancelled from those who enrolled. Student Records The admissions records for all cancelled students were identified and the individual dossiers of application papers, school records, recommendations, correspondence, and work sheets were assembled for study. From this total, a ten percent sample was selected simply to learn more about the reasons for cancellation. 2Applicants from 1969 were selected for study because this group had one year to act on their decision to cancel. The post- card questionnaire asked not only why the student had cancelled his admission but what his current activities (attending another college, military service, etc.) were. 33 (Each sampling system used is explained at the beginning of the appropr.iate section below. ) Preliminary examinations of these records indicated that cancelled students could be classified into two groups:3 cancella— tions initiated by the student and cancellations initiated by the university. Almost half the cancellations were made at the request of the student, and their files contained enough correspondence to be useful in making generalizations about their reasons for cancel- ling. 4 Descriptive terms of explanation used by these students were classified by reason5 from every tenth folder in the sample. Classification of Data The second part of the analysis used selected data in a multiple regression technique. Items both discriminatory and descriptive were chosen so that the student who cancels might be 3See Gardner Lindzey, Handbook of Social Psycholfl, Vol. 1: Theory and Method (Reading, Mass.: Addison-Wesley Publishing Co., 1954), pp. 488-518; Carter V. Good, Introduction to Educational Research (New York: Appleton -Century -Crofts, 1959), Ch. 5, pp. 243, 250; and Claire Selltiz et al., Research Methods in Social Relations, rev. ed. (New York: Holt, Rinehart & Winston, 1963f, pp. 399—401. 4For sample letters, see Appendix, pp. 88-92. 5See Lindzey, Handbook of Social Psychology, Vol. I, pp. 269-271; William L. Hays, Statistics for Psychologists (New York: Holt, Rinehart& Winston, 1965), pp. 64-67, 215. 34 differentiated from one who enrolls. High school class rank, grade -point average, and other components, described below, were perceived as offering the most reliable information for making predictions about cancellations and enrollment. In essence the multiple regression analysis attempts to demonstrate that items _P: plus Bplus _C_ have a higher percentage of predictive validity than A alone or _A_ plus B. It tests the extent of item reliability singly and in combination. "In other words, there are often several variables with known values, each of which may contribute to the prediction about the value of the dependent variable. "6 A random sample method was used to select a matched group of 300 applicants who enrolled and 300 who cancelled. The 1969 Freshman class at MSU comprised 6, 909 new students. Every twenty—third record was selected for the sample of enrolled stu- dents; this sample served as the control. To obtain the sample of cancelled students, every fifteenth record was selected from the total of 4, 620. Their profiles were analyzed by computer, and the results appear in Chapter IV. One item commonly used in analyses of this type--a college admissions test score--is missing. It could not be included 6Hays, Statistics for Psychologists, p. 566. 35 because it was not consistently available for all students accepted in 1969. In the records used, the results of the A. C. T. examina- tion sometimes appeared, at other times the S. A. T. appeared (particularly for out-of-state students), and, in certain instances, only an I. Q. or other standardized secondary -level examination score was reported. This lack of consistency gave little hope that the test score could be used with confidence. The following data classifications for coding used in the multiple regression analysis are listed with the justification for their use. , 7 Coding Rank in High School Graduating Class Code Item 3 Code Range First Quartile 1 Second Quartile . 2 Third Quartile and below 3 7Code item #1 was used for a student identification number. Code item #2 (GPA) was used in the first multiple regression run, but was omitted from the analysis reported in Chapter IV. 36 Ranking is supplied by the high school and includes the student's rank and the total number in the class: e. g. , 14/238. This rank is converted to quartiles by the MSU admissions office. Only the percentages accounting for the largest number of accepted students were used, as the remaining quartile ranking accounts for less than one percent of the total. Financial Aid Code Item Code Range Applied for Aid 4 1 Did Not Apply for Aid Type 5 None Gift Gift + Obligation* Obligation .hCA‘DND-i Need** (in dollars) 6 No Need $1. 00- 500 $501 - 999 $1000-2000 $2001 & above 01$me Value*** 7 Zero (no award) Met Need Below Need 3 NH * Obligation = loan or job ** Classifications of need determined in consultation with MSU' 3 Financial Aid Office, Summer, 1970. *** Refers to the value of the award as related to ‘need. Any of the three entries may carry one of the five possible Need classifications; i. e. , Zero (no award offered) may occur for a student with a $1000—2000 Need. 37 The classifications of financial assistance offered are based on the kinds and amounts of aid extended to accepted applicants. They are also listed according to a "need factor, " as computed by the College Scholarship Service from the Parents' Confidential Statement. 8 If MSU did not meet an applicant' s need, perhaps the decision to attend another college may have been related to financial considerations. The _Ygl_u_e_ assignment reflects all kinds of assistance and includes either a single award or a combination of several types of aid, including a job (obligation), a loan (obligation), or a grant (gift). A large award ($1, 000 or more) is most frequently a combination of gift plus obligation, for example. _N_eg_d_ is usually related to thelaiug entry. For example, a $600 scholarship meets the need of a student in the $0 - 500 need category of awards, but a $600 scholarship and a $600 loan would meet the need of only some of those in category 4 ($1000—2000). One additional factor remains: the question of when the cancellation occurred. Awards from the Michigan Higher Education 8The College Scholarship Service is used by MSU to determine financial need. The Parents' Confidential Statement is the only evidence required to be considered for aid. A random sample of 300 applicants who were awarded aid was used for this portion of the analysis. 38 Assistance Authority9 are made in May, MSU sends out notifications of its awards in April, the Candidates' Reply Date is in May, 10 and MSU cancels the admission of students who have not paid the A. E. D. by the end of May. If money is an important consideration, as some students indicated when they cancelled their admission, a number of cancellations might occur after these various award letters have been sent. Therefore, a separate entry, labelled ”Timing, " shows the date of cancellation for those in our sample. Timing Code Item 8 Through July March April May June & After Code Range 1 2 3 4 5 9The Michigan Higher Education Assistance Authority, an agency under the Michigan State Department of Education, admin- isters four programs that provide financial assistance to college students: a student loan program directed through banks and other lending institutions; a tuition grant program, restricted to residents attending private colleges; a scholarship program, restricted to residents who take a competitive examination and who have a demon- strated financial need as determined by the C. S. S. Parents' Con- fidential Statement; and a War Veterans' Scholarship program. 10The Candidates' Reply Date is used by some colleges who subscribe to this C. E. E. B. service. See Annual Report (Princeton, N. J. : College Entrance Examination Board, 1969). 39 Frequency Tabulations A separate portion of this study included information to be tabulated. These factors include: sex of student, age, size of home town, and distance from home to college. While grade -point average, rank in Class, and financial needs might help to describe the student who will cancel, a further refinement for predictive purposes might be supplied by considering these additional character- istics. State universities serving residents but also seeking some geographic balance in their student bodies could profit from knowing about the students they attract and those they lose in the context of residence, for example; that is, do they receive more applications from in -state than out-of-state students? The classifications on the following page are based on the same coding system used in the multiple regression program. Items 4 through 10 in the classifications following were used in the multiple regression analysis 3E1. the tabulations. Items 11 through 16 were tabulated only. Questionnaire The third portion of this study used a post-card question- naire, sent to 600 students who had cancelled their admissions. 1For a sample card, see Appendix, p. 93. GPA RANK lst quartile 2nd quartile 3rd quartile & below Applied for aid Did not apply . Type None Gift Gift + obligation Obligation Need No need $1-500 $501 - 999 $1000-2000 $2000 8: above Value TIMING Zero Met need Below need March & before April May June July & after 40 SEX RESIDENCE AGE DISTANCE TOWN Small Moderate Large STATUS Male Female In- state Out- of- state 17 8; below 18 19 8: above 50 mi. or less 51- 100 mi. 101-500 mi. 501 - 1000 mi. Over 1000 mi. Population 4, 999 8: under 5, 000-49, 000 50, 000 8: over Enrolled Cancelled 41 Included in this sample were both those who had merely let their admission lapse by not paying the A. E. D. (so that their admission was cancelled by MSU) and those who had actually notified the university that they wished to cancel. Because some students originally stated their reason for cancelling, a follow -up seemed appropriate to test the validity of their statements. The question- naire, then, was an attempt to ascertain whether the students' actions following their cancellation were based on the reasons they gave in their letters to MSU. In other words, would we find, for example, that a majority of those who gave ”Distance" as a reason for cancelling actually were attending a college closer to home? Summary The methods and procedures used in this study were based on three questions about the student who cancels his admission: (1) Can we obtain information about the student who will cancel? (2) Can we predict who will cancel? (3) Are the reasons that applicants give for cancelling valid? With these questions in mind, three techniques of analysis were employed. A content analysis of letters of cancellation and 42 tabulations of data retrieved from students' folders and official records were used to obtain information that could be analyzed. A multiple regression analysis of these data was performed in an attempt to determine whether reliable differences between students who cancelled and students who enrolled could be ascertained. The attempt to develop a formula to predict cancellations would follow if the multiple regression analysis produced significant results. Finally, applicants who had cancelled their admission were sent a questionnaire that solicited information that could be compared with the data obtained from the content analysis. The results of these analyses appear in Chapter IV. CHAPTER IV ANALYSIS AND RESULTS As indicated in Chapter 111, data for this study came from records compiled by the MSU Office of Admissions and Scholarships and the Registrar' 3 Office. Included are all applicants admitted and classified as first-term freshmen for Fall term, 1969 (see Table 1). TABLE 1 CLASSIFICATION OF STUDENTS ADMITTED Number Percent Total Admitted 1 1, 342 Residents 8, 482 74. 8 Non -residents 2, 860 25.2 Total Enrolled 6, 722 59. 2 (of admitted) Residents 5, 787 86. 9 (of enrolled) Non -residents 935 ‘ 13. l (of enrolled) Total Cancelled 4, 620 40. 8 (of admitted) Cancelled by Student 2, 138 46. 3 (of cancelled) Cancelled by MSU (no A. E. D.) 2, 482 53. 7 (of cancelled) 43 44 Analysis of Cancellation Letters Although all students who cancel do not write complete explanations of their reasons or even correspond formally (some telephone), the records contain enough correspondence to justify a systematic study of letters from parents or from the students who cancelled. The first analytic method was similar to a content analysis and was applied to letters of cancellation. This analysis served only as a silnple device for discovering some clues about why students cancelled their admission and was a starting point that helped construct a later questionnaire; it also helped in selecting those items most pertinent for further study. The content analysis method for reviewing correspondence has been explained in several research texts and involves (1) a careful construction and definition of elements, words, units, or categories to be isolated and studied, and (2) tabulation of the frequency of such elements. 1 This review of letters helped identify the most common reasons given for cancellation. For example, money and distance from home were two frequently -cited items, and they were listed as two of the several possible classifications. 1Carter V. Good, Introduction to Educational Research (New York: Appleton-Century-Crofts, Inc. , 1959); and Claire Selltiz, et a1. , Research Methods in Social Relations, rev. ed. (New York: Holt, Rinehart & Winston, Inc. , 1963), pp. 335-342. 45 A closer examination of the correspondence indicated that certain key words and phrases were used often and could be grouped together. A few examples are listed below. DISTANCE too far closer to home more convenient great distance local college FINANCES money problems beyond financial means limited financial resources cannot afford home state scholarship money from another college PROGRAM cannot get program. at MSU program better at College X SIZE MSU too big going to a smaller college Some students who wrote indicated a preference for a local or community college and gave no explanation. Unless a specific reason, such as cost, was cited as the most significant reason, such letters were not included in the content analysis, because factors of size and distance and cost all could have contributed to their decision not to attend MSU. About half the students who cancelled their admission in writing offered some explanation to the admissions office. The most frequent reasons are shown in Table 2, below. 46 TABLE 2 STUDENTS' REASONS FOR CANCELLING ADMISSION* Financial Distance Size Program Both Resident and Non-resident 47’ 0% 26' 9% 22- 0% 4. 1% *N=213 These first investigations provided certain guidelines for the multiple regression analysis reported below. The most obvious reasons for cancellation cited by students, money, distance, and residency, were all accounted for in the selection of variables to be used in the multiple regression analysis. Additional variables were chosen to provide information that would result in a compre- hensive yet distinctive resumé of the student who cancels. Multiple Regression Analysis As noted elsewhere, numerous opinions were solicited in the initial phase of this study from those who had considerable experience with research design. The final selection of the variables 2Records of those communicating their notice of cancella- tion in some written form (2, 138) were separated from all other records. Every tenth letter was selected for study. CE] 47 used in the multiple regression analysis is the result of these several consultations. To develop a comprehensive profile of the cancelling students and to retrieve all data that might contribute to cancellations, twelve variables were originally chosen for analysis. These variables were selected with the following question in mind: What factors will help learn more about cancellations, and what data will distinguish the student who will cancel from the one who enrolls? The twelve independent variables were first analyzed according to the multiple regression method established for the MSU 3600 computer. 3 The results of this first analysis clearly showed that further refinement and the elimination of some of the variables were required. The first findings were not statistically justifiable and, in fact, contained elements that created a confused profile of the student who cancels. Using all twelve variables provided no way to distinguish those significant contributions to cancellations from those less important, nor did this procedure allow for the individual analysis of similar variables that alone might have provided significant results. Therefore, variables that 3William L. Ruble, Donald Kiel, and Mary E. Rafter, "STAT Series Description No. 7 LS. Calculation of Least Squares (Regression). Problems on the LS Routine. " (East Lansing: Michigan State University, Agricultural Experiment Station, November, 1969), mimeographed. Y“? (n m 86 g? 48 were similar for predictive purposes, such as rank and GPA, were combined or discarded. After further consultation and analysis, seven were selected for a second run of the multiple regression program. Those seven variables were additionally refined by running separate regressions for five different categories: 1. In -state students who applied for aid 2. In -state students who did not apply for aid 3. Out -of-state students who applied for aid 4. All who applied for aid 5. All who did not apply for aid This method simplified the interpretation of results and, at the same time, accounted for all of the variables originally selected, albeit in a different format. In short, whereas the first program tended to confuse any findings by running all elements together, the second was a simpler approach that clarified which variables or groups of variables accounted for the greatest amount of variance and significance . R esults The first analysis containing the seven selected variables was processed with 209 observations for in -state students who 49 applied for financial aid. For this group, Value (of financial aid) was the single most significant predictor of enrollment (Table 3). TAB LE 3 IN- STATE STUDENTS WHO APPLIED FOR AID 2 . Regression Beta . r vamable Coefficients Weights FB 31g Deletes Rank -0.053 -0.043 0.383 0.537 0. 113 Need -0.001 -0.003 0.002 0.966 0. 114 Value 0.212 . 0.329 19.70 0.0005 0.028 Sex -0.027 -0.028 0. 165 0.685 0. 114 Age 0.024 0.020 0.083 0.773 0. 114 Distance -0.036 -0.038 0.301 0.584 0. 113 Town -0.005 -0.008 0.013 0.910 0. 114 2 N = 209; r = 0. 114; F = 3.708; p= 0.001 At the end of this particular regression analysis, when all deletions had been made, Value (of aid) accounted for 10. 96 percent of the variance. In the remaining tables, only the first multiple regression is reported, because the variables deleted through the multiple regressions after the first did not alter the findings sig- nificantly. This first observation is perhaps the most interesting and _ does show a statistical level of significance. Although the amount of variance is not large, the calculations clearly show that the Value variable does account for the largest percentage of variance. 50 When the same kind of analysis was run on date for in-state students who did not apply for financial aid, no significant findings resulted. (See Table 4.) TABLE 4 IN-STATE STUDENTS WHO DID NOT APPLY FOR AID 2 . Regression Beta . r Varlable Coefficients Weights FB Slg Deletes Rank 0.044 0.044 0.420 0.518 0.021 Sex -0.044 -0.045 0.434 0.511 0.021 Age 0. 100 0.083 1.472 0.226 0.016 Distance 0.076 0.074 1. 195 0.275 0.017 Town 0.066 0. 104 2. 355 0. 126 0.012 2 N: 224; r = 0.023; F = 1.008; p= 0.414 The same variables were used in a regression analysis of data for out —of -state students. Again, for both those who applied for financial help and those who did not, none of the variables was significant. Nevertheless, the lack of statistical significance is interesting, particularly when one considers frequent speculation about non -resident students and those influences that may attract them to a certain campus. Given that Distance (from home to campus), Value of financial help, and Need factors do not produce statistically significant results (see Table 5), the college that 51 believes it can attract the more distant student through financial awards should be appraised that those in this particular sample showed no such inclination. In this analysis, money made no difference between enrollment and cancellation for the non -resident. Indeed, none of the variables used could account for any measurable variance (Table 5). TAB LE 5 OUT-OF - STATE STUDENTS WHO APPLIED FOR AID Variable Regression Beta FB 8‘ r2 Coefficients Weights 1g Deletes Need -0.073 -0.205 2.751 0. 101 0.072 Value 0.097 0.209 3.132 0.080 0.082 Sex 0.127 0. 153 2.138 0.147 0. 082 Age 0.120 0. 134 1. 546 0. 217 0.089 Distance -0. 069 -0. 127 1. 348 0.249 0. 091 Town -0. 068 -0. 124 1.269 0.263 0.092 N= 91; r2 = 0.105; F = 1.647; p= 0.144 In the last test of these variables, data on. all students who applied for aid were analyzed, and significant findings resulted. (See Table 6.) The two significant findings for this group of students had to do with the value of financial aid and residence (in-state or out -of-state). For all those applying for aid, the value of the money 52 awarded by MSU and their residence were significant predictors of enrollment. Given the previous finding that Value (Table 3) was a significant factor in the enrollment of in-state students who applied for aid, it is not surprising that these two factors corre- late significantly again in an analysis among all students who applied for aid . TABLE 6 ALL WHO APPLIED FOR AID 2 . Regression Beta , r Variable Coefficients Weights FB 81g Deletes Rank -0.068 -0.047 0. 806 0.374 0.224 Need -0.017 -0.037 0.407 0.532 0.225 Value 0. 166 0. 271 20. 902 <0. 0005 0. 171 Sex 0.009 0.009 0.029 0.842 0.226 Reside -0.272 -0.250 7.881 0.005 0.206 Distance -0.050 -0.096 1.287 0.256 0.223 2 N = 300; r = 0.226; F = 14.289; 3: 0.0005 For all students who did not apply for aid, the most significant finding was that the student' s residence accounted for the largest percentage of variation. (See Table 7.) 53 TABLE 7 ALL WHO DID NOT APPLY FOR AID 2 . Regression Beta , r Variable Coefficients Weights FB 81g Deletes Rank 0.078 0.073 1.610 0.203 0.051 Sex -0.047 -0.047 0.681 0.415 0.054 Reside -0.332 -0.289 0.100 0.005 0.030 Distance 0.049 0.094 0.865 0.356 0.053 N = 300; r2 = 0.056; p = 0.002 Although the over—all multiple'regression program for the given population of students produced little evidence that sufficient differences between enrolled and cancelling students existed to pro- duce a valid method for forecasting enrollments, certain findings war- rant additional exploration. The results reported above are discussed at greater length in Chapter V, where all results are analyzed. Post-card Analysis The first phase of the study was a content analysis of letters of cancellation received from students (or their parents) who cancelled their admission. In an attempt to verify this unsolicited information and to discover any discrepancies between what these students reported as their plans and what action they actually took, a questionnaire was sent to 600 students. The 54 questionnaire included the following items: present status, reason for cancellation, comments (see sample, Appendix). From this sample, 215 replies (35 percent) were received. The reasons for cancellation listed by these respondents are tabu- lated in Table 8. Some caution must be exercised in the interpreta- tion of these results because the return is relatively small. TABLE 8 FOLLOW-UP REASONS FOR CANCELLATION* Reason for Cancellation N % Financial 92 42. 7 Smaller College 8 l 37 . 6 Personal 60 27. 9 Distance 54 25. 1 Curriculum 35 16.2 TOTAL 322 * Cf Table 2, p. 46. ** Total = more than 215 because respon- dents checked more than one reason. Not included in the calculations are ten students (4. 6%) who reported that they were not in college; four who had married, four who were employed, and two who had entered the U. S. Armed Forces. Respondents were asked to rank -order their reasons for cancelling their admission by indicating the primary reason by "1, " the second-most important reason by "2, ” etc. Of the 215 55 respondents, 117 actually rank-ordered their reasons, and the tabulation of these responses is given in Table 9. Some students ranked as many as six reasons, but only the first three were used to calculate the percentages given below. TABLE 9 PRIMARY REASONS FOR CANCELLATIONS Reason Percent* Financial 47. 0 Distance 38. 4 Smaller College 37. 6 Personal 29. 0 Curriculum 15. 3 * Total exceeds 100% because responses were combined by category, not by rank-order, and some respondents checked more than one category. Table 10 gives a more detailed analysis of the responses from the questionnaire. These results clearly show that more defini- tive and discriminating categories for some items would have been desirable. Even though other choices were offered, "Personal" was obviously too broad and all— inclusive a category to delimit the reason for cancellation. The "Distance" classification is highly questionable when students living within 50 miles of East Lansing, as well as students from California, used it as their main reason 56 for cancelling their admission. Finally, no provision was made to allow respondents to combine categories, such as "Distance" and "Financial, " both considerations for the non- resident applicant who must pay out -of-state fees. TABLE 10 SUMMARY OF RESPONSES TO QUESTIONNAIRE (N = 215) Rank Reason Unranked Total Financial 30 18 7 0 1 0 36 92 Small College 22 17 5 1 l 1 33 81 Personal 18 6 10 3 0 0 23 60 Distance 11 27 7 1 0 0 8 54 Curriculum 11 6 1 1 1 1 0 35 Total Responses: 322 Total in College: 205* * Of this total, 82 were attending'Michigan colleges, as follows: 00 mwgmmdootb University of Michigan Wayne State University Western Michigan University Central Michigan University Community College Oakland University Eastern Michigan University Other H 57 Even with these limitations, the questionnaire was a useful device. Interestingly, the results of the content analysis and the answers from the post-card survey are somewhat similar. Both show large percentages of students (47% and 42.7%) who cite financial reasons, and about the same number who indicate distance (26. 9% and 25. 1%) as a reason for not attending. The differences between those who mentioned size as a factor and preferred a small college (content analysis = 22%; survey = 37.6%) may be explained by the possibility that the smaller campus became an important consideration later in the decision process; that is, students who finally decided in favor of a small college did so after having been admitted to MSU. Frequency Tabulations As noted in Chapter III, an analysis of the applicant's GPA, rank in class, and financial need may indeed contribute to our understanding of why he cancels his admission, but other factors, such as age, sex, size of home town, etc. , may also provide useful information. To refine the analysis, therefore, all variables were tabulated according to frequency, and enrolled students (300) were compared with those who cancelled (300). 4 4This cancelled student sample included both student initiated and MSU mass cancel students. In a sense, _a_ll who do not enroll are self -cancelled. 58 The frequency tabulations below show the results of this analysis, by variable. Grade - point Average As the data in Table 11 show, 12.5 percent more of the ”better" students (3. 0 GPA and above) cancelled their admission than enrolled. The better the .student' s high school record, the more opportunities he has for admission to college and, therefore, the greater his choice. By contrast, those in the middle range may wish to assure themselves of a place in a freshman class and hence enroll as soon as they have the opportunity. TABLE 1 l ENROLLED VS. CANCELLED STUDENTS, BY GPA Enrolled Students Cancelled Students - GPA N Percent N Percent N 3.0 and above 411 43.7 180 56.2 231 2.5-2.9 177 64.4 114 35.5 63 2.1-2.4 12 50.0 6 50.0 6 Rank in Class As expected, the results according to class rankings are similarto those by GPA in the first two categories. Higher -ranked 59 students cancel their admissions more frequently than mid -ranked students (see Table 12). The last category, which has only 9 cases, can perhaps be explained by the reality of competition. Although admitted, the bottom -of-the -class student may become convinced that competition would be too severe for survival. TABLE 12 ENROLLED VS. CANCELLED STUDENTS, BY HIGH SCHOOL CLASS RANK Enrolled Students Cancelled Students Rank - N Percent N Percent N lst Quartile 507 48. 1 244 51. 8 263 2nd Quartile 84 63. 0 53 36. 9 31 3rd Quartile 9 33. 3 3 66. 6 6 Financial Aid Of the 300 students who applied for financial aid in this sample (see Table 13), 150 were in-state (Michigan) students and 150 were out-of-state residents. Apparently, an offer of financial help, particularly gift monies, can affect enrollment. Jobs were less attractive than gifts as inducements to enrollment. The most striking segment here, of course, is the number‘who cancelled 60 their admission and received no award offer from MSU. (Both resident and non -resident applicants are in this category.) TABLE 13 ENROLLED VS. CANCELLED STUDENTS, BY TYPE OF AID OFFERED ' Enrolled Students Cancelled Students Type of Aid* N 3 Percent N Percent N None 108 25.9 28 74.0 80 Gift Only 37 72. 9 27 27.0 10 Gift + Obligation 123 72.3 89 27.6 34 Obligation Only 32 18. 7 6 8 1. 2 26 * None = No money awarded Gift = Scholarship, grant, etc. Obligation = Loan or job Financial Need When the "Need" factor (as interpreted by the College Scholarship Service) for enrolled and cancelled students was examined, the results (Table 14) showed that more "No need" students cancelled than enrolled. The exact opposite was true for those with modest needs ($500. 00 and under). 61 TABLE 14 ENROLLED VS. CANCELLED STUDENTS, BY ANALYSIS OF NEED Enrolled Students Cancelled Students Need (In Dollars)I N L Percent N Percent N No Need 38 36.8 14 63.1 24 1-500 48 60.4 29 39.5 19 501-999 84 57.1 48 42.8 36 1000-2000 115 44.3 51 55.6 64 2001 & above 15 53. 3 8 46. 6 7 V alue The value of financial aid awards also provides some insight. These comparisons show simply whether MSU offered financial help that was equal to the student' 5 evaluated need (Table 15). For this group, perhaps some award (even though need was not present) might have resulted in fewer cancellations. From these figures, one may speculate that financial aid is an important factorin the decision to enroll. Over-all, almost twice as many students offered some kind of assistance enrolled as cancelled their admission. However, some students, both residents and non- residents, still respond negatively even with financial help. 62 Perhaps their plans change or MSU does not offer enough financial support. TABLE 15 ENROLLED VS. CANCELLED STUDENTS, BY AMOUNT OF NEED MET BY MSU Enrolled Students Cancelled Students Award N Percent N Percent N None (Nkoard) 108 25. 9 28 74. 0 80 Need Met 100 60. 0 60 40. 0 40 Below Need 92 67. 3 62 32. 6 30 Timing The "Timing" Variable, that is, when the offer of admis- sion was cancelled, is shown below (Table 16). The large percentage of cancellations in May reflects the University' 8 policy of cancelling all admissions for which an Advanced Enrollment Deposit has not been paid by the end of May. But note also that a total of 40 percent of those cancelling their admission do so before the University takes any action. The increase between the March and earlier classifica- tion and April might be a reaction to MSU' 3 financial award notifications, which are sent out in April. Without financial help, 63 TABLE 16 MONTH OF CANCELLATION, BY PERCENT OF THOSE CANCELLING Month N Percent March and earlier 48 16. 0 April 72 24. 0 May 156 52 . 0 June 14 4. 7 July and after 10 3. 3 as noted earlier (Tables 14 and 15), the student may simply not be able to afford to attend MSU. The next figure (May), for which the cancellation percentages almost double, may reflect a number of factors: the University' s cancellation policy, the Candidates' Reply Date, and awards made by the Michigan Higher Education Assistance Authority. The Candidates' Reply Date is used by some colleges who subscribe to the agreement set forth by the Educational Testing Service. In essence, this agreement requires the participating colleges to withhold their letters of acceptance until May 1. A student receiving a notice of acceptance from a preferred college during the month of May may send a cancellation notice to other colleges that may have accepted him earlier but are, in his view, less desirable. For residents of Michigan, May is also the 64 announcement date for awards made through the MHEAA competitive examination. Those who made college plans based on the assump- tion of some financial help may change their plans if they do not receive this award. Sex Enrollments and cancellations do not differ appreciably according to sex, as we can see from Table 17. TABLE 17 ENROLLED VS. CANCELLED STUDENTS, BY SEX Enrolled Students Cancelled Students Sex - N Percent N Percent N Male 306 48. 3 148 51. 6 158 Female 294 51.7 152 48.2 142 Residence The comparison by residence category showed some surprising results, particularly for those in-state students who cancelled their admission (Table 18). Given that students cite factors such as "distance" and "cost" as reasons for attending or not attending a particular college, these responses alone might lead 65 to the prediction that more in—state than out-of-state students will enroll at MSU. These reasons for cancellation have consistently appeared over several years, and seem to reinforce Thresher' s "propinquity" hypothesis. 5 The figures for those who cancelled their admission, however, were hardly anticipated. If 40 percent of those cancelling live in Michigan, "distance" may have to be more strictly defined. TAB LE 1 8 ENROLLED VS. CANCELLED STUDENTS, BY RESIDENCE Enrolled Students Cancelled Students -Residence N . Percent N Percent N In‘s.tate 433 59.5 258 40. 4 175 ReSldents Out-of-state Residents 167 25. 1 42 74.8 125 (Not Foreign) Consider, for example, that the community college, as well as other factors, may have a sizable influence onattendance. If nearness does affect attendance, as reported, then perhaps "distance" should include the concept of "convenience. " A student 5Alden Thresher, Gouge Admission and the Public Interest (New York: College Entrance Examination Board, 1966). 66 may find attending a college in his own community easier than making the adaptive changes necessary for "going away to school, " even though that school may still be in his home state. A further analysis of this phenomenon appears in the next section of this chapter, which shows the responses of students who were asked to list their reasons for cancelling their admission to MSU. According to these calculations, "distance" turns out to be a relative term. Age The age of students both attending and cancelling might provide an interesting study in itself (Table 19). TABLE 1 9 ENROLLED VS. CANCELLED STUDENTS, BY AGE Enrolled Students Cancelled Students Age - N Percent N Percent N 17 and below 118 38.1 45 61. 8 73 18 472 52.7 249 47.2 223 19 and over 10 60. 0 6 40.0 4 The largest number of cancellations, among the youngest age group, may reflect the greater degree of parental influence over younger 67 students. Perhaps, too, the younger student is simply less sure of his plans than the older student. These figures indicate a need for further research, although a survey conducted by the American Council on Education at MSU showed a similar relationship between age and enrollment. Distance The largest proportion of enrolled students comes from a population that lives within 500 miles of the campus at East Lansing. (Table 20). TABLE 20 ENROLLED VS. CANCELLED STUDENTS, BY DISTANCE FROM HOME TO CAMPUS Enrolled Students Cancelled Students Distance ~ (In Miles) N . Percent N Percent N 50 or less 90 60.0 54 40. 0 63 51-100 317 58.3 185 41.6 132 101-500 113 37.1 47 62.8 71 501-1000 59 25.4 15 74.5 44 Over 1000 2 1 19.0 4 80. 9 7 6American Council on Education, National Norms for Entering Collegg Freshmen--Fall, 1969, Research Report, Vol. 4, No. 7 (Washington, D.C.: By the Council, 1969). 68 Again, Thresher' s "nearness” concept would seem to be a factor in college attendance . Siz e of Town Most enrolled and cancelled students come both from urban and suburban areas, a reflection of the dominant residence patterns nationally; and the difference between enrolled and cancelled stu- dents by residence is slight. (See Table 21.) TABLE 21 ENROLLED VS. CANCELLED STUDENTS, BY SIZE OF HOME TOWN » Enrolled Students Cancelled Students ' Size of Town N - Percent N Percent N Small (to 4,999) 132 53.0 70 46.9 62 Medium (5,000-49,ooo) 216 50-0 108 50.0 108 Large (50,000 a over) 252 48- 4 122 51.5 130 Summary While the multiple regression analysis by itself was not as valuable as anticipated, in combination with the tabulations and the post-card survey, it did provide considerable information about 69 cancellations. The entire study generated data that have not previously been reported; and these data could be used in a variety of ways in additional research. Some suggestions for further study are given in the next chapter. CHAPTER V SUMMARY AND CONCLUSIONS This study has produced some interesting and significant findings, a few disappointing results, and certain indications for future investigations into the problem of identifying students who will cancel their admission. Part of the original goal was to develop a predictive system that would help admissions staffs forecast enrollments more accurately. Given the results of the multiple regression analysis designed for this purpose, no such method could be established. Nevertheless, the finding that few differences exist between the students who will enroll and those who will cancel is important in itself. Also, the admissions office should note that the value of monies awarded can make some small difference between enroll- ment and cancellation. This particular finding might be studied in more detail to learn whether loans, grants, or job awards would have made any differences in enrollment. 7O 71 The tabulations provided considerable insight into the characteristics of the cancelling student and could perhaps be used with greater confidence for predictive purposes than the multiple regression analysis. That 17 percent more of the better students (3. 0 GPA and above) cancel than enroll may indicate one way for the admissions Office to insure more enrollments and fewer cancel- lations: admit more students with acceptable, but lower, GPAs. That 40. 4 percent of the cancelling students live within the state should also help plan future action. The post-card follow -up showed that 40 percent of the cancelling students actually attended some in-state college. Some students perhaps left the state to go to college, but obviously a large proportion remained in Michigan. The admissions office needs to know more about these residents' reasons for attending another college if it wants to attract and retain them. Certain discrepancies were uncovered when students' reasons for cancellation were checked against their individual records. Some students cited distance as the main reason for not enrolling at MSU, yet they lived within 100 miles of the campus-- a few as close as the Detroit area. Other students stated that financial reasons prevented their attendance, yet they never applied for financial aid. And of those out -of-state students who 72 finally chose to go to a small college instead of MSU, at least some perhaps were not aware of the three, small residential colleges on campus. These observations will remain speculations until they are tested by techniques that can reveal more precise reasons for cancellation. Perhaps some method could be applied to separate ”socially acceptable reasons" from actual ones. A probing tech- nique, such as the open -ended questionnaire or the personal inter- view, would be useful in determining the relativity of distance, money, and other reasons that seem to lack credence. For example, it is true that Ann Arbor is physically closer .to Detroit than East Lansing, but perhaps another factor, such as personal choice of college, is actually more decisive than distance. The very personal and individual view the applicant has of any one college and the ranking of institutional preferences would then be better indicators of the likelihood of cancellation than distance. This particular attitude system may have more to do with final choice than financial awards (or-lack of them) and other commonly -given reasons for cancellation, yet the college now has no means by which to assess attitudes. For example, applicants to MSU are never asked to rank -order their top three college choices. The school counselor may know, mother and dad may 73 know, but the institution directly involved seldom has the advantage of knowing exactly where it stands on the candidate's preference list. The problem has been discussed in journals and published reports that use student -supplied data, 1 but they invariably are after -the —fact studies and are of little help in forecasting enroll- ments for a given year. One rather noteworthy factor emerged from the multiple regression results, which showed thatXilgg (amount of money awarded) can make a difference in the decision to enroll for some students living within the state, but no difference at all. for non- residents. One might speculate that just the opposite would be true. As noted earlier, members of the admissions staff believed that most cancellations occurred among non -residents who wanted money from the university but did not get it. The evidence here seems to indicate that financial assistance may have little bearing on whether a non -resident student attends or not. He may indeed seek monetary help to pay non -resident fees, but not receiving it apparently is a small deterrent to actual enrollment. In other words, perhaps the out-of-state student originally applies with a 1American Council on Education, National Norms for Enteri_ng Collegg Freshmen--Fall, 1969, Research Report, Vol. 4 No. 7 (Washington, D.C.: By the Council, 1969). 74 strong determination to attend and is in actuality little influenced by the prospect of financial help. If this assumption is true, then gift monies, loans, and jobs may have to be assigned with both cancellation and enrollment facts on hand. For those students who do indeed cancel because they cannot afford to attend, the financial aids division may want to review its policy regarding awards to non -residents to determine how much and what kind of awards are needed. Perhaps loans and jobs would suffice for the non -resident who wants to enroll but needs financial assistance. On the other hand, the tabulation data indicated that about half the cancelling students (53.4%) received no offer of financial help from the university. This item alone would lead one to believe that money may indeed play an important part in enrollment. Another related result showed that the financial need of 42.6 percent of the cancelling students ranged between $1000 and $2000. Those with a proven need of considerable money perhaps had no choice but to attend some college that either came closer to meeting their need or was less expensive. Another finding from the tabulations showed that a large percentage of the cancelling students live within 500 miles of the campus. This fact alone may not be surprising given that 40. 4 percent *fi lrn 75 of the state residents cancelled their admission, but it does reveal that distance is a relative term. The MSU admissions office should recognize that some accepted students will attend another college closer to home, however accessible the East Lansing campus may ' seem. The post-card survey confirmed some of the observations first made through the content analysis of letters of cancellation. In fact, two of the three reasons most often cited by students as the major reason for cancellation differ by only a few percentage points when the figures compiled from the content analysis and the figures from the post-card follow -up are compared: Reason Content Analysis Post -card Survey Financial 47 . 0% 42 . 7% Distance 26. 9% 25. 1% Size 22.0% 37.6% The 15. 6 percent difference in the two figures for the Size variable can be explained only by some reasoned guesses. Recall that the content analysis began with letters of cancellation voluntarily written by the students. This group gave the admissions office the advantage of learning about their reasons for not attending (versus those who simply cancelled without comment). The data compiled from the survey reSulted from something of a forced 76 choice on the post-card and perhaps contained a smaller element of volition than the first. One is tempted to conclude that, in the final analysis, a large percentage of students is actually affected by the size of the college. In addition, certain students may be somewhat reluctant to admit openly that they will attend the local community college for convenience, money, or both. Few students who cancel can be expected to write down ”parental influence" as the major determi- nant of the decision to attend elsewhere. Neither can they be expected to state that a romantic interest swayed their final choice. In any study that utilizes large quantities of data, the temptation is to quote numbers, present statistics, and simply leave conclusions to the reader. However, some greater value may be realized from general statements based on the facts shown by the research. One such general conclusion from this study, for example, is that alarge percentage of cancellations came from in- state residents. This kind of concrete statement tends to synthesize what. otherwise might be tables of numbers and gives the study some focus for future consideration and perhaps action by the admissions office. The major findings of this study, there- fore, are summarized in the statements below. 77 * A slightly larger percentage of the more academically able students tended to cancel their admission. Seventeen percent more of the B or better students cancelled than enrolled. * More students receiving no offers of financial help from the university cancelled their admission than enrolled (74. 0% vs. 25.9%). More high-need students ($1000 to $2000) cancelled than enrolled. * A sizable proportion (46.3%) of those admitted cancelled their own enrollment. The University cancelled about half (53. 7%) of all who did not enroll. * Many cancellations (74.8%) came from state residents. In a study of this type, which purports to investigate one dimension of one problem, other views, possibilities, and studies can be suggested. For example, the problem of forecasting enrollments and the subsequent confusion when past patterns do not hold for a given year or class is itself rather an after -the -fact assessment. Perhaps more searching questions need to be asked about college choice and ranking, in -state or out -of-state preferences, or even the decisive factors that finally influence a 78 candidate' 3 choice of one institution. The limits of a forced -choice post-card survey are obvious. More likely, a combination of reasons applies and may not be ascertainable from a single ques- tionnaire. A questionnaire, after all, samples opinions at a particular point in time, whereas the final selection of a college may be the result of a number of decisions, made separately over time. True, something more has been uncovered about the can- celling student, but the matters 1121 investigated may uncover more revealing data about why a student will cancel or enroll. Whim probably cannot be measured, but an inquiry into why students actually chose to enroll at MSU would contribute an added dimen- sion. The point is, of course, to realize a balanced view of the process of college selection, so that "Why_<_i£ they come?" becomes as important a question as "Why don't they come?" Some studies now underway by the American College Testing Service and the American Council on Education and reports of the college-choice factors studied in recent years (see Chapter 2) indicate that we may be able to provide more answers soon. However, one must always be reminded that penalties (if such a label can be applied to cancelling students) will always accrue in a competitive, free enterprise system of education. As long as there are large and small, special and general, good and 79 less good colleges, there will be differences measured by each applicant. The church school, the city college, and the community college will offer some benefits and educational opportunities that will influence college choice for some students. BIBLIOGRAPHY BIBLIOGRAPHY Abbott, Charles F. "An Investigation of College Environmental Perceptions of Prospective College Freshmen and Their Relationship to the Choice of College or University. " Doctoral dissertation, Michigan State University, 1967. Allen, James E. , Jr. "Good Guidance and Enough of It. " The University of the State of New York at Albany, December 2, 1957 (mimeographed). American Council on Education. National Norms for Enteri College Freshmen--Fall, 1969. Research Report Vol. 4, No. 7. Washington, D.C.: By the Council, 1969. Amsden, Robert L. "Good and Bad Admissions Practices as Seen by High Schools. " Journal of the Association of College Admissions Counselors, Vol. 10 (Summer, 1965), pp. 3-6. Anderson, Robert C. , and Ellis S. Hammond. "Admissions at Michigan Colleges, 1967 - 1968. Career Opportunity Guide III. East Lansing: Michigan State University, Institute for Community Development, 1968. Anderson, Robert C., Russell G. Mawby, Joe N. Miller, and Andrew L. Olsen. "Parental Aspirations. " Adult Leader- ship (May, 1965). Astin, Alexander W. "Folklore of Selectivity. " Saturday Review, Vol. LII, No. 51 (December 20, 1969), pp. 57-58, 69-70. Beck, Joan. "There Is a College for You. " Journal of the Asso- ciation of College Admissions Counselors, Vol. 7 (Summer, 1961), p. 24. Benezet, Louis T. "College Admissions: The Hours Before the Dawn. " Journal of the Association of College Admissions Counselors, Vol. 8, No. 3 (Winter, 1963), p. 29. 80 81 Berdie, Ralph F. , and Albert B. Hood. "Personal Values and Attitudes as Determinants of Post-High School Plans. " Personnel and Guidance Journal, Vol. XLII, No. 8 (April, 19647, p. 754. Bloomgarden, Lawrence. "Our Changing Elite Colleges. " Journal of therAssociation of College Admissions Counselors, Vol. 7 (Summer, 19617, p. 12. Bowles, Frank, and Richard Pearson (eds. ). Admission to CollegegA Prggram of the 1960' s. (Fifty -eighth Report of the President.) Princeton, New Jersey: College Entrance Examination Board, 1962. "Changes in Today' 8 College Students." U. S. News and World Report, Vol. LVI, No. 7 (February 17, 1964), p. 66. Citizens' Committee on Higher Education. Remember the War Babies. Lansing: Michigan Coordinating Council for Public Higher Education, September, 1963. College Admissions Handbook. Princeton, New Jersey: College Entrance Examination Board, 1969. College Admissions Policies for the '70' 8. Princeton, New Jersey: College Entrance Examination Board, 1968. College Entrance Examination Board. Annual Report. Princeton, New Jersey: By the Board, 1969. Cramer, Stanley H. , and Richard R. Stevic. "Research on the Transition from School to College. A Review of the Literature. " College Board Review, Vol. 73 (Fall, 1969), pp. 22 -30.., Dale, Arthur. "Reported Determinants of Educational Choice. " Personnel and Guidance Journal, Vol. XLII, No. 6 (February, 1964), p. 564. Dyer, Henry S. "Ambiguity in Selective Admissions. " Journal of the Association of College Admissions Counselors, Vol. 9, No. 2 (Fall, 19637, p. 27. 82 "Eighteenth National Conference on Admission to College. " Journal of the Association of College Admissions Counselors, Vol. 8 (Winter, 1963), PP. 21-26. "Eighty New Colleges." Changing Times, Vol. 17 (November, 1963), p. 13. For Your Information. Washington, D. C. : Office of Institutional Research, National Association of State Universities and Land -grant Colleges, 1970. Gerritz, Ellsworth M. "How to Resolve the Hot Issue: Early Decision Plans for College Admission. " Bulletin of the National Association of Secondary School Principals, Vol. 44 (April, 1960), p. 102. "Getting into College: Chances Now. " U. S. News and World Report, Vol. LXVIII, No. 12 (March 23, 1970), pp. 28—30. Gladney, Marilyn Bush. ”The College Application Fee as a Factor in Percent of Accepted Applicants Who Do Not Enter. " College and University, Vol. 42, No. 1 (Fall, 1966), pp. 63 -66. Goldman, Leo. "Cooperative Plan for Guidance and Admission. " The Personnel and Guidance Journal, Vol. XLII (September, 1963). pp. 64-68. Good, Carter V. Introduction to Educational Research. New York: Appleton-Century-Crofts, Inc. , 1959. Goren, Arnold L. "The Dual Review Process. " Journal of the Association of College Admissions Counselors, Vol. 8 (Spring, 1963), p. 5. Gormley, Simon J. ”How High Schools Can Help in Admissions. " College Board Review, Vol. 75 (Spring, 1970), p. 24. Gross, Martin L. The Brain Watchers. New York: Random House, 1962. Hannah, John A. "Address to the Faculty Convocation. " East Lansing: Michigan State University, January 24, 1964 (m imeographed) . 83 Hansen, Lorraine S. "Helping Students in Pre -college Decision Making. " Journal of the Association of College Admis- sions Counselors, Vol. 12, No. 3 (1967), p. 4. Harris, Seymour E. "Future Demands for College Graduates, " in Plannigigollege Policy for the Critical Decade Ahead. New York: College Entrance Examination Board, 1958. Havemann, Ernest, and Patricia S. West. They Went to Colleggi The College Graduate in America Today. New York: Harcourt, Brace & Co. , 1952. Hays, William L. Statistics for Psychologists. New York: Holt, Rinehart & Winston, 1963. Healy, T. S. "Will Everyman Destroy the University?" Saturday Review, Vol. LII, No. 51 (December 20, 1969), pp. 54-56, 67 -69. Hechinger, Fred M. "College Search. " The New York Times, July 1, 1962. (Reprinted in Journal of the Association of College Admissions Counselors, Vol. 8 [Fall, 1962], pp. 13, 24.) 7 Heist, Paul, and Harold Webster. "A Research Orientation to Selection, Admission, and Differential Education, " in Research on College Students. Berkeley: University of California, Center-Lfor‘Higher Education, 1960. Hills, John R. "Decision Theory and College Choice. " Personnel and Guidance Journal, Vol. XLIII, No. 1 (September, 1964), pp. 17-22. Hoffman, Banesh. The Tyranny of Testing. New York: Crowell- Collier Press, 1962. Hollinshead, Byron S. Who Should Go to Colleg_e_? New York: Columbia University Press, 1952. Hoy, John C. Choos'fia Collegg. New York: Dell Publishing Company, 1967. 84 Hoyt, Donald P. , and Leo A. Munday. Your College Freshman. Interpretive Guide to ACT Research Services for Higher Education. Iowa City, Iowa: Research and Development Division, American College Testing Service, 1968-69. Iffert, Robert E. Retention and Withdrawal of College Students. Washington, D. C.: U. S. Department of Health, Education, and Welfare, Bulletin 1, 1958. Irvin, Donald W. ”Ghost Applicants at the University of Georgia. " Collegg and University, Vol. 40, No. 1 (Fall, 1964), pp. 55-59. J ambura, John W. "Analysis of Certain Relationships of Student Needs, College Environment, and College Choice. " Doctoral dissertation, Washington State University, 1967. Joint Committee on Testing. Testing, Testing, Testing. Washing- ton, D. C. : American Association of School Administrators, 1962. Jones, R. L., and L. Siegel. "Admissions Cancellation: An Institutional Study." College and University, Vol. 38 (Fall, 1962), pp. 89-95. Juola, Arvo E. "Selection, Classification, and Placement of Students, " in Paul L. Dressel & Associates (eds. ), Evaluation in Higher Education. Boston: Houghton, Mifflin Co., 1961, pp. 301-329. Kastner, Elwood C. "The Registrar and Director of Admission, " in Gerald P. Burns (ed. ), Administration in Higher Educa- tion. New York: Harper 8: Brothers, 1962, pp. 185-205. Kent, Raymond A.- (ed. ). Higher Education in America. Boston: Ginn and Company, 1930. Kerr, Clark. The Uses of the University. Cambridge, Massa- chusetts: Harvard University Press, 1963. Kerr, William D. "High School Counselors and College Informa- tion. " Journal of Collgge Student Personnel, Vol. V (October, 1963), p. 47. 85 Lindzey, Gardner. Handbook of Social Psychology. Vol. 1: Theory and Method. Reading, Mass. : Addison -Wesley Publishing Co. , 1954. Manual of Freshman Profiles. Princeton, New Jersey: College Entrance Examination Board, 1969. Mate, Hubert E. "Role of the Admissions Officer. " College and University, Vol. 44, No. 4 (Summer, 1969), pp. 705-708. Mecham, Richard W. "Please Put an End to the Admissions Process Rat Race. " College Board Review, Vol. 74 (Winter, 1969-70), pp. 14-15. Mikelson, Howard 1. "Parental Factors Influencing College Selection. " Doctoral dissertation, Colorado State College, 1967. Mueller, Kate Hevner. Student Personnel Work in Higher Education. Boston: Houghton Mifflin Company, 1961. National Association of State Universities and Land -grant Colleges. For Your Information. Washington, D. C. : Office of Institutional Research, the Association, 1970. Pearson, Richard. Annual Report of the College Board, 1965-66. (Report of the President.) New York: College Entrance Examination Board, 1966. Perkins, James A. "The Price of Choice. " Proceeding of the Annual Meeting, 1966. New York: College Entrance Examination Board, 1967. Perry, Margaret E. The Small School Talent Search. Chicago: The University of Chicago [n. d. ] . Peterson, David W. "How High School Visits Influence College Admissions." College Board Review, Vol. 68 (Summer, 1968), pp. 8-9, 18. Riesman, David. "College Subcultures and College Outcome, " in Selection and Educational Differentiation. Berkeley, California: Center for the Study of Higher Education, 1959. 86 Ruble, William L. , Donald Kiel, and Mary E. Rafter. "STAT Series Description No. 7 LS. Calculation of Least Squares (Regression). Problems on the LS Routine. " East Lansing: Michigan State University, Agricultural Experi- ment Station, November, 1969 (mimeographed). Ruch, Charles P. ”Participation in Selected Guidance Activities and the Stability of College Choice. " Journal of the Association of College Admissions Counselors, Vol. 12, NO. 4 (September, 7967), pp. 21-24. Rudolph, Frederick. The American College and University. New York: Alfred A. KnOpf, 1962. Schrag, Peter. "The Fetish of Growth." Change Magazine, Vol. 2, No. 2 (March-April, 1970), pp. 5-6. Selltiz, Claire, et a1. Research Methods in Social Relations, rev. ed. New York: Holt, Rinehart &. Winston, 1963. Snyder, Rixford K. "Developing Nationwide Standards: Admissions. " Pp. 222 -224 in Logan Wilson (ed. ), Emerging Patterns in HE her Education. Washington, D.C.: American Council on Education, 1965. Spence, James R. "Whither Admissions Research, Not Whether. ” College and University, Vol. 43, No. 3 (Spring, 1968), pp. 348 -352. The [Michigan] State News [East Lansing, Michigan], February 21, 1963; January 21, 1964; March 11, 1970. "Student Explanations of College Choice and Their Relationship to College Popularity, etc. " College and University, Vol. 33, No. 3 (Spring, 1958). PP. 313-320. "The Clearinghouse Approach to Undergraduate Admissions. " Journal of the Association of College Admissions Counselors, Vol. 8, No. 4 (Spring, 1963), p. 12. Thresher, B. Alden. College Admission and the Public Interest. New York: College Entrance Examination Board, 1966. 87 Tillery, Dale, et al. "A Study of Student Decision Making and Its Outcome, " in SCOPE. New York: College Entrance Examination Board, 1969. "Trends in College Admissions. " Changing Times, Vol. 18 (March, 1964). Wentworth, Russell. "A Descriptive Study of l, 464 Secondary School Students Who Were Denied Admission to Michigan State University. " Ph.D. dissertation, Michigan State University, 1970. ' Western Regional Membership, College Entrance Examination Board. "Statement of Principles on Admission, Honors, Advanced Pl'aéement, and Grading." Journal of the Association of College Admissions Counselors, Vol. 8 (Spring, 1963). Whitla, Dean K. "How Much Do You Really Know about Candidate Overlap?" College Board Review, Vol. 65 (Fall, 1967). "Why College Enrollments May Triple by 1970. " College Board Review (1960), p. 18. Wilson, Logan (ed.). Emerging Patterns in.Hi:_gi1er Education. Washington, D. C. : American Council on Education, 1965. APPENDIX LETTERS OF CANCELLATION* AND POST - CARD QUESTIONNAIRE * Street addresses and signatures have been omitted from these sample letters to preserve the privacy of the respondents. The descriptive words or phrases used in the content analysis appear in brackets. 88 Littleton, Colo. 80120 April 25, 1969 Gordon A. Sabine Michigan State University East Lansing, Michigan 48823 Dear Mr. Sabine: I regret to inform you that I will not be a member of the class of '73 at MSU. I would therefore like to have the $25 refund from my AED. I have discussed this with my father to great length, and we do not feel that we have enough assurances that the [money] would hold out for four years at MSU. Rather than take the risk of having to break off my college studies and go to work for a year in order to continue, I have elected to begin my college career at a Colorado school which I know I can attend for four straight years if I wish. My decision of course means I will not be accepting the $300 National Student Defense Loan. Again let me express my sincere regret that I will not be attending your fine school. I await your reply. Respectfully yours, 89 Casper, Wyoming April 26, 1969 Mr. Terrence J. Carey Director, Office of Admissions and Scholarships Michigan State University East Lansing, Michigan 48823 Mr. Carey, I have elected not to attend Michigan State University next fall. I was very pleased to be accepted but I have decided that you are a little [too far] from home, for a few years anyway. Please accept my apology for not writing sooner to tell you of my decision. Yours Sincerely, 90 April 28, 1969 Ft. Sill, Okla. Dear Mr. Arthur, I am so sorry to have to tell you I will not be able to attend Michigan State after all. Many factors enter in, the prime one being a reconsideration of factors involved in the [distance] from MSU to family, and our accessibility to each other. Please accept my sincere thanks for the time and effort you and your staff expended in order that I might attend State. I am most appreciative of the opportunity of participating in the ADS competition, and was so happy to be able to see the campus first hand, which makes this decision all the more difficult. Sincerely, 91 Glen Rock, New Jersey 07452 April 27, 1969 Michigan State University Office of Admissions and Scholarships East Lansing, Michigan 48823 Dear Sirs: I have decided not to attend Michigan State University in the Fall. I was impressed by Michigan State' 3 faculty and facilities but I am [not able to afford the cost] of attending. Therefore I have decided to attend College of Engineering, where I will be able to apply the scholarship they have granted me . Sincerely yours, 92 Fennville, Mich. November 18, 1969 Michigan State University University Business Office Box 311 East Lansing, Michigan Gentlemen; This letter is to inform you that I will not be enrolling at MSU as I had planned earlier. I have chosen a [smaller college] because of its size and more individual freedom. I regret that I won' t be attending. I am sending my certificate of admission, and the advance enrollment deposit card . Sincerely, 93 Dear Student: Our records show that you cancelled your admission to M.S.U. for the Fall Term of 1969. We are most anxious to learn of your precise reasons for cancel- lation. Your responses to the questions on the reverse side of this post card will help us better serve future applicants. Thank you for taking the time to respond. If we can help you in some way, please let us know. Cordially, Ellis Hammond Assistant Director Your present status: Student D Where Armed ForcesD EmployedD Other The basic reason you cancelled your admission was: (If more than one choice seems apprOpriate, please rank in order of the most to least influential, 1 — 2 — 3, etc.) Distance E] Financial D Armed Forces D Marriage [I Personal E) Smaller College B Job [I Curriculum Not Available at MSUD Please comment on any matters which affected your decision to cancel. 0-7577 "IIIIIIIIIIIIIIIIIIIII