71 18,155 - ARNDT, John Richard, 1938A STUDY OF SELECTED FACTORS CHARACTERIZING FRESHMEN WHO ENTERED MICHIGAN STATE UNIVERSITY IN FALL 1968 IDENTIFIED AS UNDER-, OVER- OR NORMAL ACHIEVERS AFTER ONE TERM OF COLLEGE. Michigan State University, Ph.D., 1970 Education, higher U niversity M icrofilm s, A XEROX C om pany , A nn A rbor, M ichigan A STUDY OF SELECTED FACTORS CHARACTERIZING FUR S I M M UNO ENTERED MICHIGAN STATE UNIVERSITY IN FALL 1968 IDENTIFIED AS UNDER*, OVER- OR NORMAL ACHIEVERS AFTER ONE TEEM OF COLLEGE By J. Richard Arndt A THESIS Submitted to Michigan Stata University in partial fulfillment of tha requirements for the degree of DOCTOR OF PHILOSOPHY Department of Administration and Higher Education 1970 ABSTRACT A STUDY OF SELECTED FACTORS CHARACTERIZING FRESIMEN WHO ENTERED MICHIGAN STATE UNIVERSITY IN FALL 1968 IDENTIFIED AS UNDER-, OVER-, OR NORMAL ACHIEVERS AFTER ONE TERM OF COLLEGE »7 J. Richard Arndt Statement of tha Probl— It was tha purposa of this study to datarsiina vhsthar any differances axistad asxmg studants saparatad by sax and idantlflad as under-, ovar- or normal achievers aftar complatlng ona tarm of university study. Dlffarancas wars Investigated among salactad factors partalning to high school damographic charactaristics, high school acadaalc prepar­ ation, parsons1 academic behavior patterns, selected ability test score averages, first tarm collage grade point average and related variables, and tha awla:female ratio within achievement levels (whan studants ware placed in levels without regard for sax). Tha purpose of this research was not to determine how much the factors included for study contributed to predictive efficiency of standardised intell­ ective test scores. Procedures The study sample was composed of 520 swle and 573 female frestusen from a population of approximately 7500 new students who matriculated at Michigan State University in fall, 1968. The sample did not Include students who resided outside the United States of America, were designated as special part-thee students, carried less than six credit hours of class, dropped out of school during the first term of college or did not have complete test scores and other necessary data. During Freshmen J. Richard Arndt Orientation V M k , each atudent in tha aawple waa adainlaterad tha following inatruaanta: tha Collage Qualification Teata. Fora C, tha Michigan Stata Pnlvoraity Engliah Taat. tha Michigan Stata Pnlvoraity Raading Taat. and a apacially conatruetad inatruaant, tha Acadaaic Inventory. Fora C. A ragraaaion aodal aalaction tachnlqua raportad in tha litaratura waa uaad to diffarantiata tha thraa achlevemnt lovela. Flua- and- ■inua ona atandard arror of eatlmte wara uaad aa tha dividing pointa. Tha Chi-Square analyaia and a aultivariata analyaia of varianca wara tha wain atatiatical tachniquaa uaad to analysa data. Tha .05 laval of confidanca waa aatabliahad to dataraina atatiatical aignificanca for tha four hypothaaaa taatad. Major FindInga of tha Study Four hypothaaaa wara foraulatad and taatad in thia exploratory raaaarch atudy. Tha firat hypothaaia pertained to tha invaatigation of aalac tad non-intellective factora aaaoclated with atudanta at tha tine of m t r l culation at Michigan Stata Pnlveralty. (1) FindInga indicated tha following Few aignlflcant dlffereacee occurred aaong tha groupa re­ garding aalactad demgraphic charactariatlca about tha high achoal frow which each atudant graduated. (2) Mo waanlngful dlffarancaa wara dlacerned among tha groupa for either aax regarding tha kinda and nunber of t e r m of couraea atudanta took in high achool. (3) Mo dlffarancaa occurred aawng tha groupa for either aax regarding participation in aalactad high achool extra-curri­ cular actlvitlaa. J. Richard Arndt (4) Female overachievers had algniflcantly battar study habits than tha othar groups. (5) Mo diffarancas occurrad for males. Mo diffarancas wara notad for aithar sax ragarding sophisti­ cation in taking objactiva examinations. (6) Mo diffarancas wara discovered partalning to sajor reasons for attending collage. (7) Undarachiavars of both saxes had significantly lower selfexpectations than normal and overachievers of overall grades for their first year of collage. (8) Mo diffarancas occurred among achlavement levels ragarding studants of aithar sax who declared a major lsnedlately upon entering Michigan Stata University and those who made no lsaeedlate preferential choice. Tha second hypothesis was related to exploring diffarancas on average scores of selected intellective ability tests taken whan studants in tha study sample matriculated at Michigan Stata University. Results of data analysis (1) indicated tha following: Underachievers of both saxes scored lower than normal and overachievers on a test of reading comprehension proficiency. (2) Mo differences occurred among achievement levels for either sex on CQT-Total score. hypothesis Three pertained to the study of differences at the end of the first term of university study regarding average credit hours carried, average credit hours earned and grade point average. Findings were as follows: (1) No differences occurred awirag the groups of either sex for first term average credit hours carried; however, underachievers of both sexes had a significantly lower average for first term credit hours earned. J. Richard Arndt (2) Underachievers of both sexes had significantly lover first tens grade point averages than normal achievers who in turn had slgnl> flcantly lover grade point averages than overachievers. The fourth hypothesis regarded the male;female ratio vithin achievement levels vhen students were placed in the levels without regard to sex. Results of data analysis indicated that amiss were significantly overrepresented in the underachiever category. ACKNOWLEDGEMENTS I wish to express sincere appreciation to Dr. Walter Johnson for not only chairing tha guidance comittea and auggaating many changaa which lnprovad tha clarity of tha thasle, but alao for his helpfulness and encouragement throughout my entire doctoral progress and to Dr. Irvin Lehmann* Dr. William Sweetland and Dr. James McKee for their contributions and interest as mambers of the guidance committee. I am indebted to Dr. Charles Eberly for critically reading the swnuscript, offering numerous appropriate suggestions for its improve­ ment , and playing a supportive role during the study. 1 also wish to thank the following individuals for their signi­ ficant contributions: Dr. Willard Warrington, for willingly making the data available and defraying expenses of data processing, Dr. Arvo Juola, for patiently answering numerous questions, Mr. Robert Wilson, for advice regarding thesis design and data analysis, Mr. Ron Soltis, for help with data card processing, Mr. Keith Flogsterd and Mr. Howard Teltel Baum, for writing computer program control cards, and Mrs. Colleen Platt, for typing the manuecript. 1 am deeply grateful to my wife, Janette, who temporarily sacri­ ficed her goals of further schooling and limited her mobility by working to make attainment of the doctoral degree possible. I am also grateful to my children, Scott and Heidi, for always giving nw a lift at the day's end with their exuberant greetings. 11 Finally, X wish to thank God for His continuing fulfillment of Hla promise, Fear thou not; for X am with thee: be not dismayed; for X am thy God: X will strengthen thee; yea, X will help thee; yea, X will uphold thee with the right hand of my righteousness. (Isaiah 41:10) ill TABLE OP CONTENTS LIST OP T A B L E S ................................................... v CHAPTER ONE: THE P R O B L E M ......................................... 1 Introduction ................................................. N o o d ......................................................... ................................................. Purpose Hypotheses ......................................... Definitions • • • • ......................................... L i m i t a t i o n s ................................................. O v e r v i e w ..................................................... 1 1 3 4 5 6 7 CHAPTER TWO: REVIEW OP THE L I T E R A T U R E ........................... 8 Generel Overview ............................................. 8 Introduction .......... 8 Definitions ............................................. 9 Clessificetion S y s t e s m ........................ 10 Major Operetionei D e s i g n s ........................... 10 Alternative Approaches ............................. 11 Research Inadequacies • ................................. 13 Pocus: The Individual ................................... 14 Improvement Possibilities ............................... 15 B i b l i o g r a p h i e s ............................................18 Pec tors Unrelated To This Study ......................... 19 Personality Factors ..................................... 19 Biographical Factors ................................... 24 Factors Related To This S t u d y ................................. 25 Sex Differences............................................25 Self-Concept of A b i l i t y ................................... 27 Selected Extra-Curricular Activities ................... 31 Study H a b i t s .......................- .....................32 Test-Wiseness ........................................... 33 Choice of Major ........................................34 Reading Ability ..................................... 35 High School Curriculum ............................... 36 High School Graduating Class Sise ....................... 37 S ise of Community . . . . . . . . . . . . . . . . . . . . 37 Per-Pupll Expenditure of High School District ........... 37 Type of High School .............. 38 Summary ......................................... . . . . . . 38 ill lv CHAPTER THREE: THE DESIGM ....................................... 42 Introduction . . . . . . . . . ............................... 42 .................................................... 42 Population S a m p l e .......................................................... 43 Xnatrumentatlon . . . . . . . ............................... 43 43 Collage Qualification Taata . . . . . .................. Michigan Stata Unlveralty Reading Taat ................ 44 Michigan Stata University English Test ........ . . . . 44 45 Academic Inventory ..................................... Research D e s i g n .......... 46 Hypotheses ............ 48 48 Hypothesis I ......... 49 Hypothesis 1 1 . . . .................................... Hypothesis I I I ........................................... 49 Hypothesis I V . . .........................................49 Data Preparation................................................49 Statistical A n a l y s e s ............................................50 S u m m a r y ........................................................ 50 CHAPTER POUR: ANALYSIS OP D A T A ..................................... 52 Introduction .......... 52 Hypothesis I * . ........................ 53 Selected High School Demographic Characteristics . . . . 53 High School Curricula Taken . . . . . .................. 57 Participation In Selected Extra-Curricular Activities . . 61 Study H a b i t s .......... 62 Test-Vlesness .......... 65 65 Reasons For Attending College . . . . . . . . .......... Self-Concept of A b i l i t y ................................. 66 69 Choice of College M a j o r .......... Hypothesis I I .................................................. 69 Hypothesis I I I .................................................. 74 Hypothesis I V .............. 77 S u m m a r y ........................................................ 78 CHAPTER FIVE: SUMMARY, DISCUSSION AMD CONCLUSIONS .............. 82 Problem, Purpose and Procedures ............................. 82 Summary of Research Results ................................. 85 Discussion of Research Results . . . ......................... 89 Limitations of the S t u d y ........................................91 Suggestions for Future Research ............................. 92 B I B L I O G R A P H Y ........................................................ 94 APPENDICES LIST OF TABLES Table 4.1 4.2 4.3 Page N « b « r of Studants in Each AchievementLevel by Sex. . . . Percentage of Students by Sex Graduating From Public Parochial, or Private High Schools ....................... Percentage of Students by Achleveamnt Level and Sex AttendingHigh Schools from Various SlxedCommunities ... 52 55 55 4.4 Percentage of Students by Achievement Level and Sax Evaluating Their High School Preparation For College Accordingto Five C r i t e r i a ............................... 56 4.5 Percentaga of Students by Achievement Level and Sax Taking Varloua Numbers of Terms of Selected Courses In Grades......... 9 - 1 2 ............................... 58 4.6 Percentage of Students by Achievement Level and Sex Taking Various Muafcera of Terms of Selected Courses In Grades 9 - 1 2 ...................................................... 59 4.7 Percentage of Hales by Achleveamnt Level Indicating Their Participation In a High School Science C l u b ...........61 4.8 Percentage of Feawles by Achievement Level Indicating Their Participation In a High School Debating Club . . . . 4.9 62 Percentage of Students by Achievement Level and Sex Indicating Their General Study Habits Using a FivePolnt S c a l e .................................................. 63 4.10 Percentage of Students by Achleveamnt Level and Sex Indicating Their Reasons for Attending College .......... 66 4.11 Percentage of Students by Achievement Laval and Sax Indicating What Kinds of Grades They Expected In Their First Tear of C o l l e g e ........................................68 4.12 Percentage of Students by Achievement Laval and Sex Indicating Their Choice of M a j o r ........................... 69 4.13 Raw Score Means of Selected Ability Tests for Students By Achievement Level and S a x ....................................71 v vi Table 4.14 4.15 Pag* Hultlvariate Analyaia of Variance of Sex Differencee on Selected Ability Teat Scorea of Studenta . . . . . . . Multivariate Analyaia of Variance of Achievement Level Differencee on Selected Ability Teat Scorea of Studenta 72 . 73 4.16 Multivariate Analyaia of Variance of Interaction Between Sex and Achlevaamnt Level on Selected Ability Teat Scorea of S t u d e n t a .......................................... 74 4.17 Raw Score Meana of Selected Ability Teata for Studenta by Achleveamnt Level and S e x ................ 75 Multivariate Analyaia of Variance of Sex Differencee on Selected Ability Teat Scorea of Studenta ............ 75 Multivariate Analyaia of Variance of Achleveamnt Level Differencee on Selected Ability Teat Scorea ............ of Studenta 76 Multivariate Analyaia of Variance of Interaction Between Sex and Achievement Level on Selected Ability Teat Scorea of Studenta ............ . . . . . . 77 4.18 4.19 4.20 4.21 Percent of Malee and Femalea by Achievement Level Whan Levela Were Mot Choeen on the Baa la of S e x ........... 78 CHAPTER OHE THE PROBLEM Introduction Sine* the end of World War II, total attendance at America's institutions of higher education has increased so tremendously that a college or university education 1a no longer considered the privil­ ege of a select elite, but rather the right of every able cltisen. With the increase in numbers has coma a corresponding and necessary increase in research about the college student and his environment. One major area of concern has been the academic underachiever; i.e., that student mho does not perform up to the level of achievement predicted or expected of him. In this research study, the college freshman underachiever was the object of concern. One should resmnber, however, that many students do better than expected; i.e., they "overachleve." To understand over- and underachievesmnt, one must further realise that these are terms used in lieu of over- and underprediction. Since prediction of academic achievesmnt has been heavily researched and discussed from many aspects, cf. Thorndike (1963), Levin (1915), and Hilton A Myers (1967), another aspect of academic achievement was considered for this study. Heed It is unnecessary to docusmnt the fact that every year thousands of freshmen in the several types of higher educational institutions 1 2 u a d c r a c h i m acadtalcallj. N u M r o u a itudiaa covering aany Aspects of over- and undarAchievement have heen reported in the literature. Unfortunately, findings have been conflicting. Citing several conflict­ ing results, Peterson (1943) stated that "...the research has proved to be of ... little value.*' One reason is that sample slses of studies are too samll (Thorndike, 1963). Grading practices also differ easing institutions (Hood 6 Swanson, 1965); I.e., the implication is that a student classified as an underachiever in one college might possibly be a normal achiever In another, easier-grading college. Further, different operational procedures used to identify under- and over­ achievers may select different students within the same population (Pippert 6 Archer, 1963). Population characteristics also differ among and within Institution types. In fact, within a given institution over time, succeeding populations of students say change. Ratchlck (1953) studied 52 highly intelligent high school students and concluded that "...no simple element was found to be related to all cases of underachievement." Centl (1959) stated that An analysis of the published research which has attempted to determine the factors important to college success leads to the conclusion that the factors important to academic success are different from school to school. In view of this, it would seem Important for the college counselor to determine what factors Influence academic success or failure in the particular institution which he serves. The need for this study is based on the findings and conclusions presented in the preceding two paragraphs, especially on Centl's statement. It is necessary to ascertain what factors influence student academic underachievement in a given institution, especially in view of conflicting research findings obtained from studying different kinds of student populations. 3 It waa expected that findings fros this research would (1) contribute to a wore complete knowledge and understanding of freshsmn who entered Michigan State University In the fall of 1968, (2) gain knowledge which night be of value to a counselor or acadenic adviser in helping specific students find out why they are under­ achieving , (3) help University College faculty of Michigan State University determine what changes night be appropriate In existing curri­ cular offerings to better n e t needs of all Incoming students in general and potential underachieving students in particular, (4) offer evidence that some factors involved in underachievement extend beyond the bounds of a given type of institution, and (5) provide some guidelines or Ideas for future research by college and university faculty and administrators to identify potential underachieving freshmen early in their college experience. Purpose The purpose of this study was to determine what differences existed, if any, among three groups of students identified as under-, over- and normal achievers using a model reported in the literature. Differences were to be investigated smlnly among non-intellective variables; vis., biographical and dearegraphlcal factors associated with high school characteristics, high school academic preparation and personal academic behavior patterns. The purpose of this research project was not to determine how much non-Intellective biographical end demographlcal feetors Included for study contributed to the predictive efficiency of selected standardised Intellective test scores. Several researchers have already attempted this as Hilton and Myers (1967) reported. 4 Hypotheses As stated in the literature review in Chapter II, many results of published studies about over* and underachievesent are contradictory or inconclusive; therefore, a comprehensive, unifying theory about the causes of over* and underachlavement does not exist. Trends or patterns, however, can be discerned in various aspects of over* and underachleveamnt, although no single pattern or findings has universal application. The hypotheses, then, were not based upon a clearly delineated theoretical framework. in nature. This study was generally exploratory It was expected that results would be useful for drawing tentative conclusions and generating further hypotheses. The following questions, restated as four testable hypotheses in Chapter III, were formulated to give direction for analysing data about three groups of students separated by sex and identified as under-, over- or normal achievers after one term of university study: (1) Are there differences aawng the groups associated with selected desmgraphlc characteristics of hlqh schools from which they graduated? (2) Are there differences associated with high school preparation aaxmg the groups in regard to types end number of tersw of courses taken in high school? (3) Are there differences among groups in selected personal academic behavior patterns? (4) Are there differences sarong groups on selected intellective varlablas ? (5) Are there differences within groups pertaining to tha male: female ratio when groups are not Identified on the basis of sex? 5 D«fInitIona Thorndike (1963) and Lavin (1965) explained that tha concapta of over- and undarachlavanant involva dlacrapanclas between pradlctad and actual parfomanca and that tha t a n a ara nls leading If not correctly undaratood. Tha following definitions, therefore, wara offarad to clarify tha uaaga of t a n a In thla study. A fullar axplanatlon including nathoda for oparattonally idantlfylng tha extreaw groupa la includad In Chaptar II. Pndarachlavanant. That acadaailc parforaanca which falla balow a glvan dlatanca fron aoaa apaclflad laval of axpactad or pradlctad achlavanant.* Ovarachiaveawnt. That acadanic parforaanca which falla abova a glvan dlatanca fran aoaa apaclflad laval of axpactad or pradlctad achlovonant. Mornal achlavanant. That acadanic parfomanca which falla on or about a apaclflad laval of axpactad or pradlctad achlavanant. Lavln (1965) dlacuaaad tha llnltatlona of tha flrat two tarna (P. 25): Studlaa of ovar- and underach lavamnt ara found vary fraquantly In tha litaratura. Howavar, tha cholca of tarna aaana unfortunata. For ona reason, auch labala tand to ralaa Intalllganca and aptltuda taata to an alnoat aacroaanct laval. That la, aInca ovar- and undarachlava aw nt ara daflnad aa daparturaa fron what ability naaauraa would laad ua to expect, thara la a tandancy to think that thaaa daparturaa ara aoawhow nyatarloua and lnaxpllcabla--that aonahow tha Intalllganca or aptltuda taat juat cannot ba wrong. It would ba aora accurata to aay that for tha pradlction of acadanic parfomanca, ability la but ona kind of nacaaaary infomatlon. Fron thla point of viaw, what la laft aftar ability haa baan uaad aa a pradlctor la not ovar- and undarachlavaawnt, but unaxplalnad variation, nuch of which nay ba accounted for by othar pradlctlva factora. In abort, thaaa t a r n actually rafar to tha Inaccuracy Involved In predicting acadanic parforaanca fron ability n aaauraa alone. If thla la not recognised, wo nay fall to look for othar algnlficant claaaaa of pradlctora. *Xn thla atudy, tha "given distance" was chosen a priori to ba plus-and"ninua ona standard error of estlnate fron tha regression lino fornad by predicting first t a m collage GFA fron COT-Total score. 6 A second reason thet the choice of terns is unfortunate is that they have acquired negative connotations, arousing in sosm the idea that the overachiever is a "grind" who lacks such desirable qualities as sociability, "well­ roundedness, 11 and the like. On the other hand, the under­ achiever may be thought of as one who is lazy, undisciplined, and Innature. In this volume such connotations are not Intended. In spite of these shortcomings, the terms are currently used in the educational field. For this reason they are used here, with the hope that the reader will keep in mind Che limitations described above. Limitations Several limitations of this study should be noted. (1) There was no assurance that students comprising the study population were representative of succeeding student populations at Michigan State University or similar institutions, although it was assuswd that findings of the study based on this population may be useful for generating hypotheses about similar populations. (2) The first term college GPA used as the criterion in identifying each of the achievement groups was a heterogeneous criterion subject to the limitations pointed out by Thorndike (1963), p. 17: Whenever we combine data from different schools, different programs, or even different teachers, we are likely to intro­ duce heterogeneity Into the criterion...In any research of "over-" or 'underachievea»nt," we b u s t bewate lest our criterion measure of achievesmnt itself be heterogeneous, the same score or synbol representing different real levels of performance in different subgroups. GPA was used because it was the standard by which the University retained or dropped its students. (3) Factors found not to be significant for this population might be for another and vice versa. (4) The predictor, CQT-Total score, seemed to favor males. This limitation was based upon male/feamle comparative standings on CQT-Total score for all freshmen who entered Michigan State 7 University in fell, 1968, published in e report by the Office of Evaluetlon Services (1968). Feswle scores were about eleven points lower than wales at given percentile intervals. (5) The predictor also accounted for only about 19.4X of the variance in GPA for wales and 25X for fewales in the study sanple. How­ ever, this was not too different frow most slwilar cognitive predictors. (6) High ability students had relatively little roow to dewonstrate overachievewent because of the GPA "ceiling" and low ability students had relatively little roow to dewonstrate underachlevewent because of the GPA "floor" (cf. Duff and Siegel, 1960). Overview An introduction to the research problew, the need for and the purpose of the study were presented in this chapter. Questions used to construct hypotheses were posed, as well as definitions of essential tetws. Chapter II contains an extensive literature review in three parts: general overview, brief presentation of findings of factors not studied in this thesis, and results of previous studies of factors related to this thesis. Chapter III presents an account of the thesis design which Includes a description of the population and sawples, instrwentation, specific hypotheses and procedures utilised in studying the problew. Chapter IV contains the results of the analysis of data. A suwwary of purpose, procedures, findings, conclusions and lspllcatlons for future research are presented in the fifth and final chapter. The bibliography, Part I of Forw C of the Academic Inventory, and tables showing statistically non-significant results of data analysis of Inventory itews complete the contents of this thesis. cn rm zi U V U V OP THE LITOATQKE Canaral O v r v l w In troduc t Ion Slnco tha and of World War II, an axtraordinary anount of raaaarch haa boon conducted on varieua aapacta of atudent achlavanant In Anarlea's lavaral lavala of educational laatitufclona. For exanpla, Ra^k, Goldberg, and Passow (1966), concarnad only vitk "bright undar­ ach lavara ," countad 146 Investigations, 50 about eollaga atudanta only, reported during tha dacada 1953-1962, conyarad to only 37 In tha previous thirty yaara, 1923-1952, 23 about callaga atudanta only. Slnca 1962, tha nunbar of published atudlaa haa lncraaaad, aa an axanInation of PlaaartatIon Abatracta and tha Education Index denonstrates. Studanta ldantlflad variously aa having "high" or "superior" ability, balng "giftad," "bright," or "talantad," hava racalvad a graat daal of attantlon, aapaclally slnca tha advant of tha "apaca aga." Thla no at llkaly haa occurrad bacauaa tha fadaral govarnnant, prlnarlly via tha Wnltad Stataa Offlea of Education, haa fInanead nuanrous projects aland at idantlfylng and subsequently aaalating youngatara with auparlor ability to raallaa thair naxinun potantlal of acadaailc achlavanant, for raaaona of national dafanaa and Intarnal vitality. Tha govarnnant doaa not want to waata thla axtranaly valuabla and Irraplacabli raaourca, an undaratandabla concarn. 8 9 The p*rfon»oc« •£ atudtnta at all ability, aga- and grade-lavels, especially college students, haa haan and la continuously analysed. Researchers concerned with college atudaat achievement have atudlad what saena Ilka almost every conceivable variable and correlation of varlahlaa ralatad to tha subject: aga, aax, Major, Intalllganca, ability, h o M and collaga raaidanca, family background••Including socio-economic atatua, educational laval and nativity of parents--personality acalaa, high achool grade-polat avaraga, pravloua guartar *a grada-point avaraga, high achool and collaga extracurricular actlvltlea, and nunbar of houra working part-time, to nana aona factora aoat often analysed. Def la It Iona Tha underachieving youngatar la, according to Pataraon (1963), Ma atudent who haa tha ability to achieve a laval of acadanic auccaaa elgnlfleantly above that which ha actually attains," Raph, Goldberg and Paaaow (1966) defined tha "more able" undarachlavar aa "o m who, for whatever raaaon, falla to develop hla potential snximally.” (p. 2). Thaaa deflaitlona ara deceptive, however, bacausa whan put Into operational tarna, different atudanta can ba Identified aa ''underachlevera,M depending on tha nathod uaad, with little overlap. Plppert and Archer (1963) danonatratad thla phenenenon whan they uaad two najor methods-high ability/poor gradaa and high ablllty/poor achlavanant taat performance--la a ninth-grada clean. Tha former method Identified aa undarachlevera 14 boya and 7 glrla and tha latter nathod 7 boya and 12 glrla, with only two children Included In both groupa. load and Swanson (1965) further danonatratad thla when they looked at student achievesmnt In different types of Minnesota colleges. The neaa grade- point averages (GPA) for freshmen In various colleges ranged fron 1.9 to 2.6 (on a 4.0 scale) with little relationship found among tha colleges 10 ability laval of thair freshmen claaaaa and tha mean OFA aaraad by tha claaaaa. Tha implication la that a atudent idantiflad aa an undarachlavar at ona collaga might not ha aa daaignatad at another. Wallington and Walllogton (1961) alae warned aducatora and raaaarehara about tha lack of agraamant among aducatora and mathoda In accurataly ldantlfylng undarachlavara (p. 8). ClaaalfIcation Syatarna Malor Qparatlonal Daalana Thorndlka (1963), Farquhar and Payna (1964) and Jackaon (1968) hava daacrlbad and claaalflad tha mala methodological daaigna uaad to operationally define. Identify and subsequently study, under*, overand normal-achieving students and thalr performance. In his small but excellent volume, Thorndlka discussed two amjor designs: (a) tha classic dichotoaiy of axparlmaatal and control groups with experimental sumlpuletion and followup, and (b) examination of relationships among variables; i.e., "find out what variables correlate with achlavement, and haw they ara related to each other." (p. 34). three variations: Tha latter design had (a) prediction ovar time of tha affect of variables under study, (b) consideration of all variables at ona point In time (no antecedents), and (c) definition and comparison of two or three contrasting groups, usually "underachievers," "normal-achlevers," and "overachievers." Parquhar and Payne (1964) classified over- and underachieveemnt research into four descriptive categories. I. Central Tendency Splits. Under- and over-achlavement is determined by dichotomising a distribution of combined aptitude and achievement measures. . . (Cf. Dowd (1952), Fearlman (1952), Shaw and McCuen (I960)). II. Arbitrary Fartitions--Middle Croup Eliminated. Discrepancies are determined by contrasting extreme groups 11 In «chltvMMiiC-tptltu4« distribution*, and by eliminating a nlddla group...(Cf. Shaw and Brown (1957), Franks1 (I960)). III. Relative Diacrapancy Split*. Grada Point avaraga and aptltuda pradlctor* ara rankad indapandantly. Under- and ovar-ach lav— nt 1* datarminad by tha diacrapancy batwaan tha two ranks...(Cf. Dlanar (1960), Baynar and Pattaraon (1960), Duff and Siagal (I960)). IV. Raara**Ion Modal Salaction. A regression aquation is usad to pradlct achlevsamnt fron aptituda naasuras. Underand ovar-achlavanant Is than datamlnad on tha basis of tha dlscrapancy batwaan pradlctad and actual achlavanant. (Cf. Garbarlch (1941), Krug (1959)). Jackson (1968) raportad that "a ravlaw of tha litaratura suggasts at laast thraa major approaehas to tha identification of underachievers," which did not substantively differ fron tha above classification. Although not citing Farquhar and Fayae, his classification cowblned their first and second categories and duplicated tha third and fourth. Tha latter two classification systens did not include designs using notched pairs (cf. Wrenn and Hunber (1941), Shuay (1956), Fink (1962), Young (1967)) or notched groups (cf. Bruck and Bodwin (1962), Da Sana (1964), C.F. Conbs (1964)). Altarnative Approaches Othar researchers have offered alternative approaches to view ing aithar undarachlovers or undarachlavanant. Harrla (1940) *aw thraa factor* Involved In undarachlavanant: ability (Intalllganca), effort (notlvation) and circtmstances (non-Intelligence). Shaw (1961) pointed out that there was a great deal of differance batwaan tha "chronic* and "situational'* underachiever. Kewlts (1965) discerned thraa doninant approaehas to undarachlavanant in tha litaratura: (1) an illness Involving tha personality ("underachlavanant is, at bast, an inaccurate diagnosis of tha problen"), (2) a problen resulting fron Inadequate notlvation, and (3) a problen steaming fron poor educational adnlnlstratIon or organisation. Counselors and psychotherapists, such as 12 Meugeboran (1958), GoIdburgh and Ptnncy (1962), Roth and Meyersberg (1963), l i l p a n (1965) aad Bodnar and Walnbarg (1970), have studlad tha undarachlavar fron thalr particular perspective aad have offarad auggas tlons aa to how to daal with tha problen. Butebar (1967) dlatlagulahad batwaan intellective (GPA, achlavanant taat acoraa) aad noa-lntallactiva (personality, biographical and danograph leal data) factora. It la laportant to realise that raaaarchara have increasingly acrutlnisad non-intellective variables in tha paat dacada, 1961-1970, bacauaa thay ballavad that auch factora hold tha kay to variance unexplained by traditional iatallactiva variables uaad to pradlct acadanic parfomanca. Thaaa raaaarchara hypothaaIsed that non-intell­ ective variablae whan addad to Intallactiva variables would give a •ora accurata pradlction foraula. Fradarlkaaa and Melville (1954) found that tha Strong Vocational Iataraat Blank could ba nore pradlc tlva of acadanic auccaaa In an aaglnaarlng achool for noa-conpulaive atudanta than for cenpulslve. Thay concludad that Tha uaafulnaaa of a taat nay ba inprovad by discovering aubgroupa of paopla far which it la aapacially approprlata aa a pradlctor. Such a nathod any not only parnlt nora accurata pradlctlona for tha neafcera of tha aubgroupa, but for othar naabara of tha group any raduca arrera la pradlctlon which ara duo to tha uaa of a laaa valid pradlctor. Binder (1966), Lunaaborg aad Lunneborg (1966) and Staadrldga (1968) all indlcatad that non-lataHoctlva varlablaa could add algalfIcaatly to pradlctlon of acadanic parforaanca. ■lltaa and Mayers (1967) raachnd a dIffarant concluaion. Thay lavaatlgatad tha contribution of non*IntoHoctlva biographical guestionnalra data to acadanic pradlctlon fron oaves studies published batwaan 1950 and 1964 aad stated that each... of those studios haa reported significant correlation coefficients us lag a different biographical Inventory. Bona 13 of th«i has daaoMtrattf, howtvtr, that a cosprahtna ivc battery of ability aad achievement tasta would not ba wara highly ralatad to acadaalc parforaanca or that biographical data would contrlbuta anythlag unique bayond such a battary. It should ba aotad that tha above quotatloa 1s aa axcallaat example of tha two broad, major catagorias of rasaarch studios that this wrltar has dalinaatad: (1) Predictive studlas la which Intellective and/or non-lata1lactive varlablaa ara analysad to dataralaa which factors iadlvldually or la cowblaatlaa bast pradlct undarach lavaw a t , or which daslgas ara aost efficient^ aad (2) Pescrlatlvs studies, la which ovar- and undarachlavars ara idantlfiad aad than coaparad on various latallactlva aad/or aonlatallactlva varlablas. Sons of thasa studlas tast hypotheses, others ara exploratory, usaful oaly for hypothasls coastruetion. Rasaarch laadaquaclas Much confusion has occurrad and findings aulllfiad for gansral rafaranca bacausa rasaarchars hava not adoquataly and pracisaly dascrlbad their rasaarch daslgas, statistical analyses, populations, samples, hypotheses aad/or conclusions. (1961) and Shaw (1961)). (Cf. critiques by Aadarsoa In tha particular area of underachlavement, Patarsoa (1963) bluntly assarted that such rasaarch proved to ba of little value bacausa wost rasaarchars neglected the Individual, looking instead at tha phenomenon of undarachlevawant. Ha stated further that it waa difficult to compare results bacausa of different kinds of designs and analyses. Although his criticism has validity, not all rasaarchars hava arrad in thasa ways. Othar problems exist as wall. Harris (1960) stated that in many studies ha reviewed, aax and intalllganca wara not bald constant, statistical significance was not mentioned and heterogeneity of subjects 14 occurred. sosm These oversights still occurred in sore recent studies. In studies sample sises seemed too ssmll to warrant anything sore than the most tentative of conclusions or possible factors to investigate further. Thorndike (1963) stated that ..,the more of achievement that we are already accounting for by known predictors, the larger our experimental groups must become if we are to establish the influence of further, more subtle, influences. Correlational studies of factors related to gain in achievement that are based on 100 cases or less will generally be a waate of effort, (p. 39). Thorndike also stated that much research had little or no m eaning because of (1) errors in measuremsnt, (2) heterogeneity of criterion, (3) limited scope of predictors, and (4) Impact of unawasured inter­ vening variables upon the Individual. (pp. 4-5). ftaph, Goldberg and Passow (1966) stated the "burden of proof" lay with the researcher in "designating a student as an underachiever... He must have confidence in his predictors, in what is being predicted, and in the comparability of the samples he identifies to study." (p. 10). This statement is important because the terms "over-" and "underachievement" really should be "over-" and "underpredlctlon (cf. Chapter I , pp. 1 1 5). Focus■ The Individual Peterson (1963) was not alone in his attitude that the under­ achieving student's plight must he considered mainly as an individual problem. Shaw and Brown (1957) hypothesised that scholastic under- achievement on the part of bright college students was not an easily amdlflable phenomenon, but instead was related to the basic personality matrix of the individual. Passow and Goldberg (1958) found that uader- achlavement among gifted high school students appeared to be aymptoamtlc of a deeper, more basic personal-social problem. These consents are especially pertinent to "chronic" underachievers and are amplified In 15 the next M c t i o m concerning lnprovenent posslbilltloo in which rooonrch • Cudloo by counselors end psychothereplsts nro reviewed. FindIns* oooe reoeerchero ere epproprlece for both "sltuetlonel" end "chronic" underechlevero, Retchlek (1953) studied 52 highly Intelligent high school students end concluded thet "since no slnple elenenf wea found to be releted to ell ceses of underechleveasent, en Investigetlon slnulteneously Includes studies of the verlous pheses of the educetlonel process." Berrett (1957) intensively studied 32 gifted high school students end seid thet "only by e cereful end thorough study of eech Indlvlduel persenelity cen we find the reesons for underechievenent." Abe (1966) reinforced this position by steting thet results fron his study of non~lntellectlve indices of ecedenlc echlovonent Indlceted thet neny feetors were Involved, end thet no single neesure was edequete for ell. C. F. Cosfcs (1964) steted thet underechLevenent cennot be treeted in terns of eny one fecet of the problen, rether, underechlevenent suat be understood to be e coapletely persome1 end consistent edeptetion of the underechlever to his needs end cepecltles es he uniquely experiences then. Flnelly, Kowlts end Araaetrong (1961) concluded thet neny spec lei progrens in eleswntery end secondery schools devised to treet underechloving lndivlduels hed not found greet success beceuse they were not predlceted upon the feet thet underechlevenent Is en indlvlduel problen, verylng in ceuse(s) fron child to child. Znprovenent Foeslb111ties gnderechlevers cen be helped to Inprove their perfornence. The sltuetion le not es bleek es sane educetors would heve one believe. Appropriate counseling or psychotherepy cen be lnstrwMntel In helping the indlvlduel perforn better ecedenicelly end in other behevlor petterns 16 (Baymr and Patterson, 1960, blptrn, 1965). If a student la performing poorly in acado*lea, ha oftan la performing poorly in othar behaviors aa wall. If ha la halpad to do battar In ona area, ha aay concurrently do battar in othar araaa. that Roth and Mayarabarg (1963) alao concludad ..tha counaallng ralatlonahlp can aarva aa tha inpatua to changa tha achievement patterns." Thalr coneluaIon waa baaad on axtanalva clinical experience in tha Faychologlcal Services counaallng progran at Hanpton Inatltuta. Drasgew (1957) fornulatad thraa poatulataa aftar counaallng with at laaat twelve "gross" collaga undarachlavara who evidently vara racant high achool graduataa: 1. A rola of tha counaalor any ba to halp tha undarachlaving cliant divorce hlnaalf fron an allan curriculun and discover an appropriata ona, 2. Tha actual faallng of failure nay ba prerequisite to thla type of cliant*a beconlng ’’ready” for counaallng. 3. ...insight ia (not) nacaaaary for tha progreaa of counaallng (i.e., for counaallng thla kind of client). Motto (1959) recognised Draagov’a contribution, but offered conflicting coneluaIona baaad on a nuch different population of underachlevers--31 "gifted" veterans. Tha addaga of only generalising to tha study population ia again validated by thaaa studies. Meugeboren (1958), a psychiatric social worker, explored problens of 48 Tala Valverslty nan who entered tha achool between 1948>1954 and who ware seen at tha Division of Student Mental hygiene. Ha summarised that (1) naay passlbla explanations existed for under- and everachievement, (2) .similarities In pattarna of academic functioning occurred for students given tha saam diagnosis...." and (3) that tha severity of emotional disturbances cannot ba used aa tha sola criterion for predicting collaga success. Point two is especially significant as are two othar statements made in tha report: (1) problems of unadjusted 17 H B d t r « c h i « m a often result fron conflicts with psronts regard ins tha choice of major (ef. point one of Drasgow above), and (2) "Only students whose academic underachlsv— m at is seen as a symptom of inner problems would be expected to Improve their grades after therapy." Goldburgh and Penney (1962) developed and offered aa approach to help underachievers iaprove their perforumace. Their "primary focus" was on "•••speedy rehabilitation rather than long-tarn therapy." They called their technique "sector counseling" representing a form of "minimum change therapy" (Tyler, 1960). Bednar and Weinberg (1970) contributed a significant study to the literature whan they investigated the ingredients of successful treatment programs for underachieving students. They investigated 23 treatment programs for underachievers which used various counseling techniques. They ware seeking to answer the question, "What dimans ions of counseling treatment programs are associated with Improved acadearic performance?" rather than am rely asking, "Does counseling contribute to Improved academic performance?" They concluded that: The most potent variables that emerge from a survey of the research literature are duration and structure of the treatment method. Mot only are highly structured and lengthy prograsw the most effective in improving academic performance, aa measured by CPA, but the effects are lasting. Though structured programs are generally more viable than unstructured programs, consideration must be given to the population under study. For example, Independent students seem to profit most by an unstructured situation, which ia, however, lengthy. When the treatment consists of seam form of counseling, the higher the therapeutic conditions (empathy, warmth, genuineness), the mere effective the treatment. From the standpoint of economy as well as effectlvenass, group counseling appears to hold more promise as a treatment method than individual counseling methods or academic study courses. Mowsvsr, counseling, either individual or group, aimed at the dynamics of underachieveswnt and used in conjunction with an academic studies course seems the swat potent of all treatment methods. • • 9 s 3 *1 8 2 1 r tr X99. 3x s *1 0 e Hi 3 3 S ■o ft S ~ * e 9 a s ! * K K> M e :° 2 H* S £ Hi • n r x I e Hi !tr : s* is & I 1 i tr I X • IS a 9. 9 x » Vi •9 8 M * ft 9 ft D 1 : X 9 % ft ft IS A 8 8 I If 2 9 * 1 9 »1 ft ft 3 9 I m s ft 9 ft • H 3 ft ft • ~ VO m 'j H *1 9 ft it “ 0 B V o' v I! 9 ft 9 ft • 9. 9 X 9 9 9* 9 8 8 i 9 X 9 i 4 9 X 8 Hi 9 I t* B t IA ft ft 9 X 9 ft a 9 H. 0 Hi X X 9 t o e• H* V O w w X mKl t* 9 i 9 i 0 rr 0 9 ft 9 ft H * 3it 9 « 9 X 0 1 1 t V O 9 9> A at 1 r s 9 ■ j3*r ft. X ft 9 »^ X *"s V O O 9 •«vv v ? - 8 ft 9 9 9 ft 9 H> Hi i 9 9 t* 0 S 0 1 * i I 9 t* i 3 9 9 tr 0 3 «x• 1 & 89 \ 9 29* N* * r O' 8 9rr f 3 X f I9 ■ o H 0 I 2 x 1 X X 9 tr9 X X 9 9 B 9 9 IS S i 9 99 O3' 5 * 3 9 9 X r h* 1 f t X x 3 S 3 s 1 9 B 0 9 H O’ * 2 2 H- 3 9. 2 S 2 B 2 trA I 8 W • o X 9 1 0 X I 2 § r 9 * 9 I *X 3 J 9 a i M > 9 X X 9 I x o t * 3H> 2 8 8 ? 8 ft 9 ft 9 H> ft ^ 9 o A 9 9- 9 ft 9 ? n 9 9 M X E 9 ft ft H- 9 1 X 9 A r X I 9 O X 9 o 0 h *4 ft v 9 W h 10 Sf 2 Ml X 9 * 9 tr I t V H- ft 9 2 A • Hi H s b «Q x * m : x 3 X o. o* i1 A ft >0 a • 8 f£ t IS X r ft A. * IS I if m o tr • ►* 2it A l sr 0 9 * i T 8 0 <3 1 • S £ 8A 1 ? X £ 1 9 .• tr ir tr • 9 M *1 9 A t X O tr o i 9 X -■J I 3 A IS r • to 9 X 9 9 * : •S !9. f» *% is o x C h tr e 9 ft vo r 9 8 » ft pitAtfi 3 ft. 0 3 s. iiW(32«ttM * 1 1 s 9 l• 9 x m a • • A X 0 O' HI oxvq 0 9 ft O ftnt^Ats X « ft 9 tip I x tinitjtjn I!• tr pit 9. TO qrrq poifdaoa 3no^t qooq * porfpo (T96T) vooaopvy a Hi O M I 9 rt 9 9 ft ft ft 3 3 89 9 £ Jf • 9 £ 3 3 0 I * OB ft Vft 9 ft 9 i3s *! i N 8 I 8 * S 19 itudMtf. Miller (1961) edited a United States Office of Education tulle tiki on guidance for underach levers with superior ability which Included an excellent chapter by Shaw (1961) reviewing articles defining and Identifying underachievers of superior ability. Cowan (1961) cosq>iled an annotated bibliography on academically talented students. Lavln (1965) comprehensively reviewed both Intellective and nonintellective factors Influencing academic achievement. Goldberg (1965) compiled a lengthy bibliography while reviewing research on talented youngsters. Eaph, Goldberg and Passow (1966) also published a bibliography regarding bright underachievers. In his doctoral thesis, Butcher (1967) extensively reviewed theory and research about student self-concept and academic achievement. Factors Barelated To This Study Many researchers have Investigated personality, biographical and demographlcal variables, primarily in an attempt to Increase predictive efficiency of students' college achievement. Research findings concerning non-intellective factors not analysed in the present study are briefly reviewed in this section to give a broader view and understanding of under- and overachlavement. Seem studies reviewed were of students below college age. Personality Factors Many facets of personality structure have been analysed. I*avln (1965) Included an excellent review of personality factors in his book. Investigators have often used standardised inventories, such as the M innesota Multi-Phasic Inventory (HMPI), Strong Vocational Interest Blank (SV1B), Thematic Apperception Test (TAT), and Edwards Personal Preference Schedule (EPPS), to determine if significant differences 20 could bo dotoctod b t t m t n over- and underachievers. Perhaps tha original pub11ahad investigation using tha MMFI was that of Altus (1948). Ha equated two groups of 25 collaga studants studying eleamntary psychology and administered tha MMPI to thaw. Ha concludad that **... data appaar to Justify that adjustment ltesm can ba found which will ba associatad with academic achievement and hava no ralatlon whatever to intalligaaca as Is currantly measured." Morgan (1952) studlad University of Hlnnasota male sophomores and noticad that achievers and non-achiavars did not diffar significantly la variety of wall-dawnlopad interests, but did in typas of lntarasta. Morgan a^slaistered tha MMFI, TAT and S V H to his sample and from thalr raaponaaa concludad that several paraonallty variables appaarad to ralata positively to tha academic achievement of hlgh-ablllty collaga students: (1) smturlty and seriousness of interests, (2) aware* nass of and concern for othar parsons, (3) a sense of responsibility, (4) dominance, persuasiveness and self-confidence, and (5) motivation to achieve or need for achlavement. Burgess (1954) studlad a group of mala collaga freshsmn engineers and concludad that tha TAT had possibilities for usefulness for differ­ entiating groups of academic achievers. He did not find evidence that tha MfPX, Sill and Horschach would ba useful In this regard, although ha stated tha latter two teats ought to ba researched further, as wall as tha Borow Collean Inventory of Academic Adjustment. Burgess found that overachievers la his saagtle seeamd to (1) hava a greater need far achievement and lmproveanat of self or status, (2) ba more motivated for collaga study, (3) enjoy collaga study sore, (4) expect aw>ra from collaga study, (5) be mere efficient in planning and use of time, (4) ba battar adjusted and (7) hava more needs to ba aggressive. Underachievers seemed to ba (1) lasa into1lactually adaptive, (2) over-reactive to environmental cire urnstaacaa, (3) more dependent, (4) of weak academic motivation and (5) unabla to aaa tha value of a collaga aducatioa. Rjraa (1951) usad tha STI1 to atudy Tala ovar-, under- and normalachlevers and aotlcad aoaa differences, but ladleatad that tha laatruaant did not seem to hava awch uaafulaaaa in dlffaraatlatlng tha groups, However, in a later, similar atudy of Tala upperclassame, Rust aad Ryaa (1954) found aoaa significant diffarancaa. Thay aaaaad to ba favorabla toward using tha lastriaant for diagnoatlc purposas* ■uaaal aad 3printha11 (1965) atudiad a group of aala collaga praparatory atudanta and found raaulta slailar to Morgan and Burgaaa. Thay adala- iatarad tha SVZg aad thraa othar paraoaallty taata and fouad that ovarachlavara wara nora mature, thoughtful, planful, lndapandant aad purpoaaful than underachievers. Kish (1968) studlad sophomore nala undarachiavara la tha Collaga of Literature, Sclanea aad Arts at tha tfalvaraity of Michigan aad concludad that undarachiavara did not utilIsa thalr abilltlaa aa did ovarachlavara (cf. C.F. Combs, 1964). Rathar than lunplag than all together, Kish dlffarantlatad four sub-groups of undarachiavara: (1) ovarcenpeaaatlag for faallaga of aoclal Inadequacy by adopting an extroverted stance, (2) highly motlvatad but socially isolated, angry and alienated, (3) wall adjusted socially, but having weak acadealc lataraats, aad (4) authoritarian, conformist, non-Intellactual aad, therefere, poorly fitted for aajorlng la thla particular liberal arts program. Gabhart and Royt (1958) used tha EPPS to atudy paraoaallty aaads of over- and underachieving freehnea. Thay fouad that ovarachlavara scored slgalfleantly higher In Achievement, Order aad Iatraceptlen, and 22 Change. Proa t h w t f indln|<, they hypotkci ixed that ovtr«chi«vMMnt r«aulc«d fron a drive to (1) compete (Achievement), (2) organise aad plan (Order), aad (3) ba lata1lactually curioua (latraceptloa). Thay further hypothesised that underachieveasnt was assoc latad with (1) a aaad for verlety (Change) la which studlas M y appear boring or routine, aad (2) social motives (Affiliation aad Rurturance) la which frlaadshlp say ba placad above scholarship. In other words, thay demonstrated that several patterns or causes existed for under- aad evarachlaweaaat. Krug (1939) replicated tha Gabhart-Hoyt study In a collaga of engineering aad concluded that his results **clearly supported** their findings. Motivation haa bean extensively studlad because "one of tha basic assumptions la education Is that motivation Is a prime requisite for scholastic success** QtcBae and Duka (1940)--cf. Travers, 1949). Motivation la vary difficult to study because Its affect suist ba Inferred from observable (measurable) behavior; I.e., It is an Inter­ vening or **sK>deraterM variable (cf. Saunders, 1956, end Flaughter end Rock, 1969). KrumboIs (1957) defined achievement motivation as **...that Inter­ nal state of affairs which Impels aa individual to compete with some standard of excellence.** la hypotheeIxed that since Individuals differ la their level of achievement motivation, they achieve different levels of performance. Taylor and Farquhar (1966) Investigated the Interaction of person­ ality, achievement and motivation by using an original 94-item rosoarch scale, the Rumsn Traits Inventory (BTI). The scale contained person­ ality Items previous Investigators fouad positively related to academic 23 •eliltvMWBt. After croaa-validation, tha III vaa administered to high and lev motivated high achool atudonta, doflnod operationally aa overand undarachiavara. Significant dlffarancaa at tha .01 laval vara notad for 32 itana for nalee, 31 for fanalea, with 14 coamnn to both aexea. In tha araa of paraoaallty adjuatannt, Dowd (1952) and Paarljaan (1952) both notod no dlffaranco aamng “high capacity1* or MaupariorH ovar- and undorachiaving collaga atudanta. Conflicting raaulta wara raported by Barger and Sutker (1956) and Pierce (1962). Barger and Sutker atudied 199 nalaa and 154 fenelea In two different collagaa and concludad "...the atudy aaana to bear out tha general assumption that atudanta with high intellectual capacity and an adequate paraoaallty adjuatneat achieve higher academic performance. ** Pierce atatad that it appeared fron hia atudy that tha bright high achool low achiever waa laaa wall adjueted than hia high achieving pear. Goldnan (1961) aummnrixad hia review of tha literature by atatlng "Especially noteworthy la the fact that maladjustment can lead oithor to undarachiavaawnt or to ovorachievement. '* Hia c o n w a t lenda weight to Lavin'a (1965) criticlan that many raeearchere examine extreme groupe and ignore the middle group on the faulty assumption that only a linear relatlonahlp axlata between the extreme groupe. Lavin atatad that the middle group of achlevere alao nuat be etudled In caae a aituatlon would arlae in which a relatlonahlp waa not linear; i.e., the extranee aeon to be identical but different from tho central group, but if the middle group were deleted from the atudy, one would not know thla. It eeena a conaenaua that underachievera are more hoetlle than achlevere. Kirk (1952), who generalixed from intenalve, therapeutic 24 counseling c m m , ond cited one •• on example, definitely bo 1loved the undorochlovor woe occlng out his hostility toward o ■saber of his fsally who deaended success by performing poorly in school. Shaw (1941), citing ssveral articles besides his own research, stated that the under­ achiever generally showed hostility by an attltuds or feeling of distrust toward others. Smith (1965), studied achieving and non-achieving college freshsmn who scored in the top 51 of the Collose Qualification Tests and concluded that the latter were more negative aad hostile toward authority. Ralpern (1965) concluded that the student exhibits hostility, often subconsciously toward parents and other authority figures by passive resistance; i.e., the student just does not perform, (cf. Sutton, 1961). le stated that appropriate psychotherapy could effect positive changes in behavior in the areas of resolving hostility and improving performance in the classroom and elsewhere. Biographical Factors Biographical data have been intensively studied, usually with ths intent to add efficiency to intellective test scores in predicting college success or failure. Malloy (1954) developed a Life Bxnerience Inventory (LSI) for females at the University of Nebraska. The instrument sampled from school exporlencos, self-appraisal, family relationships and choice and type of friendships. Malloy and Ivanoff (1964) subjected the LSI to further validation for both sexes and reported the InstrusMut "...significantly increases the prediction of college marks over that of c o m e n l y used intellective measures and previous achlevsnsat in high school...It seems quite apparent that the LBI explores unique criterion variance over and above that presently accounted for by the sore traditional intellective tests and measures.'* 25 Duff *nd 81«|«1 (1960) davtLoptd a ten-area Biotrtphlea1 Inventory far collaga atudanta, and Anaataal (1960) alao davalopad a gloxraphical Invantory far tha Collaga Cntranca Examination Board. Milton and Myers (1967) reviewed aavaral auch Inventorlea and concludad that ...Each af theae atudlea haa reported algnlflcant correlation coefficieate uaing a different biographical Invantory. Mona of them haa demonstrated, however, that a comprehensive battery af ability and achieveomnt taata would not ba m o r e highly related to academic performance or that biographical data would contribute anything unique beyond auch a battery. Although numerous factors hava been found to be Involved In overachlavement, none was found to be a universal cause or Intervening variable. Probably the main reason was that different kinds of popu­ lations hava bean studied and/or different research models have been used. For example, Myers (1952) studied an eastern voean's college and found associated with academic success such factors as Jewish reli­ gion, urban living and foreign-born parents; however, these had no correlation with academic success at the University of Washington as reported by Lunneborg and Lunneborg (1966), who cited Myer's study specifically. Fee tors Belated To This Study In this portion of the chapter, research findings of factors investigated in this thesis are reviewed, including some studies concerning students below college level. Sex Differences Levin (1965) stated that the failure of many studies to analyse data separately far males and females hindered comparisons of findings, aad that sore research was needed in which sax differences were assessed and reasons for differences ware examined (p. 58). Farquhar and Fayae (1964) included separation by sex as a necessary criterion for effective 26 •tlMtion of ov«r- and undarachiavara (cf. Clark, 1953). Tka waight of raaaarch findings fron studios la which tha ssxas vara mot saparatad iadicataa that undarachiavara ara predominantly mala whlla tha majority of ovarachlavara taad to ba faawla. Dowd (1952)J atudlad high ability collaga atudanta and found that mora mslas undarachlavad and mora famalas achlavad up to expectation. For auparlor or gif tad atudanta balow tha collaga laval, Gowan (1955), Road (1955) and Shaw (1961) indicatad that undarachiavara wars pradomlnataly malsa. Tha flrat two reaaarchara gava a ratio of two malaa to ona famala for undoracklavamant and two famalas to ona amis in tha ovarachlavar catagary. Shaw aad McCuaa (1960) found that undo rack lavamant for malaa atartad la tha first grada in achool, bacama significantly diffaraat from achiavora at tha third grada aad incraaaad aach yaar through tha tanth grada. Tha p a t t a m waa diffarant for glrla. Faawle undarachiavara actually exceeded achlavars in gradaa ona to fiva, though not signi­ ficantly statistically. Thay droppad from grada six to grada alavaa, with tha dlffaranca bacaming significant in grada nlna. In tha alavaath grada, gradas of achlavars of both asxas droppad slightly. Bowman (1960) and Raph, Ooldbarg and Fassow (1966) atatad that sax dlffarancss wara more pronouncad balow tha uppar sanlor high gradaa (alavaath and twalfth) bacausa adult sax rolas bacama mora pronouncad starting at about tha alavaath grada. Thay thaorlsad that amny aldar high school girls did not want to camps to with aad ovarshadow boys who wara soon to bacaam family brsadvlaaars. raflactad by lowar achlavamaat. Tha girls* attituds was Tha rasaarchars hypothaaixad that whan malas antarad collaga, thay atartad raalislng thay must do wall if thay axpact to gradwata and to obtain adaguata employment. 27 Todd, Terrell and Frank (1962) danonatratW chat sex differences can occur whan several variables ara studlad. Thay studlad non-Incoll­ ective differences hatwaan collaga normal- and undarachiavara of suparlor ability and found significant dlffarancaa batwaan nala and fans la groups. Self-Concept af Ability Salf-coacapt of ability (SCA) sssns to hava baan studlad mast intanslvaly during tha dacada of tha 1960's. SCA was an Important factor la achlavamsat. Goldbarg (1965) statad A. W. Combs (1962) statad that "much of a parson's bahavlor Is tha rasult of his concaptlon of hlmaalf.** Sutton (1961) Infarrad from his study of 85 chlldran In gradaa 3-5 In an alamantary school In Athens, Gaorgla, that Achlavamant Implias a saIf-rafaranca and is not undarstandabla unions a cancapt of salf is adoptad. Tha achlaving aalf will display aa latagrativa organisation of parsonal traits... Educational achlavamant rasulta In parsonallty growth, Intagratlva bahavlor, and a mora haraumlous salf. Lucas (1968) daflaad SCA as "tha aggragata of knowladga, foaling, attltuda, ballaf and valua hold In ralation to ona's salf," and usad a scora on tha gills Indas of Adluatmsnt and Values to aparatlonally msasura It. With daflnltlons af six rasaarchars In mind, Patars (1968) formalatod tha following daflnltlon: "Salf-cancapt Is a psychological con­ struct usad to dascrlba a parson's pareaptlon of hlmaalf and...of his ralatloashlp to athnrs In tha anvlraamsnt." Sha lncludad thraa camponants (1) "parcaptual"— tha way la which ana saao hlawalf and tha ldaa ha has af tha lmprasslaa ha mskas an others; (2) "conceptual"--ana's ldaa of his "own peculiarly distinctive characteristics, abilities, limitations"; and (3) "attltudInal"--ana's sense af Identity within hia anvlraamsnt, his attltuda regarding tha present and future, and his degree of salfesteaau Sha administered tha Tennessee Salf-Concept Scale to 164 high 28 high achool s o b lor• aad found no differences between over- end under­ achievers In self-concept of ability; however, she stated that not controlling for sex and the swell sample alas limited the results of her study. Fink (1962) studied 20 pairs of wale and 24 pairs of female rural California high school students watched for sex and IQ. la found that underachieving boys definitely had a lower SCA, but not underachieving girls. Brookover (1942) studied ninth grade, urban, Michigan school children. Be reported that underachievers of both sexes with low SCA could Improve their achlevesmat by SCA enhancement on the part of parents acting as “significant others.** Counselors and outside discussion leaders from a nearby university did not effect any improvement. Borislow (1942) studied 184 freshmen students in the College of Arts and Science of a state university who completed a pre- and post­ semester Questionnaire designed to yield indices of the student's personal self-evaluation in general terms and specifically as a student, ■a formed four groups of students: 84 achievers and 21 underachievers oriented toward academic attainment, and 55 achievers and 26 under­ achievers not so oriented. Borislow concluded that: 1* Begardleas of an intention to strive for scholastic achieveamnt as a prime goal, students who underachieve scholastically cannot be distinguished from those who achieve scholastically on the basis of general self-evaluation prior to or subsequent to their first semester in college. 2. Students who underachieve scholastically have a peerer con­ ception of themsolves as students than do achievers subsequent to their scholastic performance, regardless of initial intention to strive for scholastic achievement as a goal. 3. Where students exhibit an intention to strive for scholastic achievemsnt as a prism goal, underachievers have a more pessimistic conception of themselves as students than do achievers prior to their actual scholastic performance. This does not hold true where scholastic achievemsnt Is not a prime goal. 4. Hlitra scholastic achievemsnt is a prist goal, Ohara tha student has a goad cancapt af himself as a student, and where he does achieve scholastically, his general self-evaluation becomes ware favorable from pro- to post-semester assessaeats. This does not hold true where scholastic achleveasnt is not s prlas goal. lorialaw apparently established that poor achleveasnt was ante­ cedent to poor SGA and goad performance enhanced one's SCA. Conversely, ■alpera (1965), a psychotherapist, proposed that in early childhood eoraal development of a separate Identity went wrong for the underachieve ■e hypothesised the phenomenon was caused by an Inadequate parent-child relatlonahlp and the lack of normal development of a separate Identity manifested Itself In peor performance. Srookover (1962) showed that Improving SCA can, in turn, improve achlevesmnt. Shaw and Alves (1963) fouad that a negative self attitude was associated with lower academic achievement of bright, underachieving smle high achoel students compared to normally achieving male students. They were unable to determine a cause-effect relationship, however, and stated the subject needed further research. In short, the long-standing "Which causes which?** question does not appear to be solved. An Important fact, however, is that once the "low SCA— low achlevesmnt" syndrosm is started, focussing on improving one's SGA can help to counteract it. A few researchers have added a variation to studies of SCA by investigating atudanta* self-prediction of future academic performance. V. C. Yeung (1954) studied a sample ef 100 students who made self­ predictions of future academic performance and self-estimates of academic ability after the first sin weeks of college to the same counselor under the same conditions. Young concluded that there was a significant positive correlation between eelf-predlctien of college scholastic achievement and actual achievement. 30 Galer (1941) M k a d 132 undergraduate* te e e t l M t e their final grade in an Adeleeceat Psychology class, aad their reasons. Be found that high-ranking students appeared to be nost capable in estimating their final course grade. Middle-ranking students seened to be least accurate in their predictions. Todd, Terrell and Frank (1942) invest­ igated college swle and fasttle nornal and underachievers of superior ability on certain non-lntellactlve factors. They found that both aale and fenale achievers had higher expectancies for success in acadenic pursuits than underachievers. Delays and Eesaglla (1943) asked a sanple of 183 freshnen at Southern Illinois University te predict their GPA for their first two quarters of college. They analysed the students* self-estimated GPA, actual CPA and School and College Aptitude Test (SCAT) scores. They concluded (1) the self-estlsmted GPA, though significant, corre­ lated less with actual OTA than did the SCAT scores, .41 to .43, (2) the self-estinate did net account for criterion variance when conhlned with SCAT scores in Multiple correlation, (3) students in the sanple over-as time ted their actual OTA, 3.45 (estimate) to 3.06 (actual), and (4) higher ability students tended to under- or accurately estimate future grades; whereas, lower ability students tended to over-estimate future grades. Keefer (1945) studied the entire student body of 195 students of Bryan College, Tennessee. Bis main finding appeared to be that self- confident students of superior intellect were mere accurate in selfpredlctlen of future grades. It seems that this aspect of student achlevesmnt has been neglected and ceuld be a fruitful area of investigation. 31 Salactad Extracurricular Actlvltiaa Holland and Elcharda (1966) and Baird (1968) of tha Anerlcan Collaga Taating Progran (ACT) danonatratad that intaraat waa a battar pradictor of non-acadaaiic (axtra-curricular) achlavaanat than aptituda or high achool gradaa, and that acadanlc and non-acadanie achiavananta wara largoly indapandant of aach othar. Holland and Bicharda aalactad a 31 aanpla fron 612,000 high achoal aaniora who took tha ACT pragran taat battary during a oua-yaar parlod andlng Octabar, 1963. Bapllaa to nonacadaaiic achlavanaat acalaa wara axanlnad to laarn what bar atudanta had won prisaa or achiavanant awarda in laadarahlp, nua1c, drawn and apaach, art, writing or acianca. Ho ralatlon waa found batwaan racalpt of awarda in thaaa araaa to ACT aptituda acoraa or high achool gradaa la Engliah, nathanatlca, aoclal atudlaa or natural ac lanca. In hia artlcla, Baird rapartad two atudlaa of nonacadanlc achlavananta of "bright” and **avaragaM collaga atudanta. In ona atudy, non­ acadaaiic ach iav anan ta of 3700 atudanta attandlng 33 dlvaraa collagaa who acorad in tha avaraga ranga on tha ACT (naan acora, 20) wara conparad with tha achiavananta of 323 national MarIt Finallata. alight dlffarancaa wara found batwaan tha grnupa: Only "bright” atudanta achlavad nora fraquaatly la litarary araaa and "avaraga" atudanta in artlatic araaa. In tha aacoad atudy, Baird conparad high achoal achiavananta af about 14,400 Michigan high achool atudanta who had ACT acaraa in tha acholarahlp ranga--22 or ahava--wlth about 10,700 atudanta acaring undar 22. Analyalng achiavananta in acianca, art, laadarahlp, nwalc, writing and draaw, ha found alight, but atatiatically algnlficant diffarancaa favoring bright atudanta la laadarahlp, aclanea and writing, and avaraga atudanta in art. 32 In light •£ these results, lalrl, Bolland and Richards statad that callage admissions affleers aught ta ha nora cencerned with, and taka lata accaunt, non-academic achiovoswat whan selecting students. Bolland, Richards aad Baird alsa teak inta accaunt research findings which shewed that success In life after collage was not necessarily related te acadanlc achievement ia callage, and that eminent people began achieving early in life. In other words, they advocated that college should place more emphasis upon non-academic achlevesmnt by positively rewarding such behavior, especially in the admissions precess. Pavek (1968) studied the relationship of several non-academic high school variables to college achievemsnt and participation In selected extra-curricular activities for male freshmen at the Balverslty of Borth Dakota. Be concluded that college GPA and ACT scores correlated peerly with non-academic achievement in areas such as science, art and writing. Study Babits Vrenn and Bwnber (1941) stated that Wrenn's Study BabIts Inventory (SBI) had possibilities for differentiating between smle and female underachievers. Dowd (1952) administered the SBI to first semester freshsmn at the Balverslty of Bow Baapehire. Be distinguished high capacity achievers and underachievers en several items In fever of achievers, but found no sex differences. The Brown-Bel trnan Survey of Study Bab Its and Attitudes (SSBA) developed in the early 1950's, has been frequently used In research studies. Beltsman aad Brown (1953) stated that the SSBA could contri­ bute te prediction of academic success and guidance of college students. Studying college aale sophomores, Juniors and seniors at the Balverslty 33 of Arkansas, Disssr (1960) found that overachievers had ksttar study habits on tha basis of SSBA Items. Las (1960) studlad under-, normal and overachlevlng collaga famalas at tha Oniverslty of Ismail and datsctad no dlffaranca In professed study habits. experimental farm of tha SSIA. Sha usad an Da Sana (1964) attamptad to ldantlfy non-lntallactual characteristics of consistent over-, under- and normal achlavars enrolled in science curricula at tha Pennsylvania State Vnlverslty. Ha concludad that non-Intellective factors could ba useful in predicting academic performance, and statad that tha most affective discriminating and predlctiva lnstruawnt ha usad was tha SSIA. Smith (1963) analysed 134 male, University of Kentucky freshman on several variables, la found that achlavars had battar study habits than undar- achlevers. Gallant (1966) found that high school staff ratings of students' study habits correlated positively with their academic achievement In collaga. Tast-Wisenass Boer ana and Hahlstrom (1968) statad that "tast-wiaeaaes"--the knowledge of techniques concerning how to taka objective tests and hew te take advantage ef various "cues" within them to manlmlse ana's test scare--could be taught and would halp improve one's achievement. In their paper, they cited Kbel's (1963) pertinent warning that H... more error In measurement Is likely to originate from students who have too little, rather than too much, skill In taking tests." In other words, a student with poor skills In taking objective tests might receive a lower score ("inaccurate measure") than he "ought" to receive. Later, If he received higher grades based en other criteria, ha might possibly be falsely identified as an overachiever. Or, In a prediction study, he might be classified as a potential low achiever but turn out 34 to bo Identified aa a u o n a l or high ochlovor. In abort, on Inaccurate ■aaivraMat on tbo lov a Ida could anally canaa a prediction arror whlcb In tarn nlgbt falaely Identify a atudant aa an overacblever aftar a parlod of tine alapaaa. Juola (1969) analysed raaponaaa to Iteam about objective taattaklng practicaa includad on tha Acadanlc Inventory. a apaclally conatruetad lnatrunant administered to about 85% of all freshmen Matriculating at Michigan Stata University In fall, 1968. Ha dlacovered that raaponaaa Indicating test-wlseness wara generally favored wore often by high achieving and high ability groupe. ravaraa true far some procedurea. Ha aleo found tha Juola hypotheelaed that tha latter finding night hava bean due to low achlavara taking renedlal couraaa In high achool. Choice of Major McQuary (1954) atudled flrat aanaatar freshmen nalaa at tha University af WlaconaIn and found that elgnlfleantly nara undarachiavara than everachlavare wara Hvary uncertain** about thalr vocational chalcaa. Stoner (1956) atudled 19 antched paIra of high ability high achool atudanta to determine factora related to underachleveneat. Ona finding waa that undarachiavara wara not aa certain of thalr future educational plana aa wara achlavara. Todd, Tarrall and Trank (1962) atudled non*Intellactlva dlffarancaa batwaan noraal and underachieving collaga atudanta af auperler ability. Thay analysed data for each sex aaparataly and found that normal achieving nalaa had decided an apacific vocational goals amre often than undarachiavara. Mo dlffarancaa existed batwaan tha female groups. Taylor (1964) reviewed 39 studies concerning personality traits and discrepant achlavamant published batwaan 1933*1963. Me dlaearned 33 •mral factors which had baan found to positively rolata to achiavonant lavaI. Ona factor was tha realism of a student's goals. Taylor cited seven studies which indicated the underachiever either had no stated goals or had unrealistic goals. Favek (1949) studied first semester male freshswn at the Vnlverslty of North Dakota and found that "undecided" majors graduated frow high school classes of less than 25 students, had the lowest GPA and particlpatad in less selected extra-curricular activities than other students. Baird (1949) reviewed several studies which cewpared a student's currlcultm choice and his acadewlc achlavewsnt. He found that evidence, though sparse, suggested undecided students differed little frow decided students. Baird attewpted to study this area as a priwary concern in two ways hy studying (1) college students near the end of their first year, and (2) college-bound high school students. He concluded frew study one that no real difference existed between a student who had decided upon a vocation (near the end of his first year in college) and the student who has not. In study two, Baird found that undecided college-bound high school students sore often emphasised the goal of developing their wind than decided students, and chose the goal of vocational or professional training less fre­ quently. Baird further stated that no evidence existed to shew that west undecided students are maladjusted or abnormal, as many people seem to think. Beading Ability Wedenoyer (1943) studied 18 male and 5 feswle college students who scored In the top 2% of an IQ measure, and found the achievers had a better reading level than non-achievers. Stoner (1957) administered 36 a diagnostic raadlng c««t to 19 natchad pairs of high ability high achool atudanta. Ba fouad that tha undarachlaving group had a lowar naan acora than tha achlavara. Ha furthar atatad tha undarachiavara had nora raadlng handicaps than tha achlavara and conprohandod lass wall tha typa of raadlng notarial Maasurad by tha taat. Cowan and Schalbal (1960) daaianatratad that a high posltiva corralatian axlatad batwaan raadlng ability and achlavaannt ansng callaga atudanta of aqulvalant into1lactual ability. Stabana (1968) avaluatad tha affacts of a raadlng iaproviaant progran on low achlavlng collaga atudanta and cancludad that a raadlng skills class could halp sana collaga atudanta lnprova thalr acadanlc achlavanant. High School Currlculun Staton (1962) studlad now fraahnan fron Oklahona high schools who anrollad In tha Onlvaralty of Oklaham. Ona of tha variablas ha aalactad for atudy waa atudanta* high achoal currlculun. Ha cancludad that tha currlculun takan la high achool did not lnfluanca callaga gradaa. H. V. Young (1967) analysad tha high achoal currlculun pattarna of claaaly natchad pairs af collaga atudanta. Ha found no significant dlffaraaca in collaga achlavanant batwaan atudanta who took 7.9 buainaas and industrial couraaa in high achool and atudanta who took 0.9 auch coursas. Ashcraft (1969) invastlgatad tha affact af tha high achoal currlculun upon callaga achlavanant. Ha cancludad that for callaga atudanta of bath saxas, high achool currlculun aaanod to ba no at Influ* antla1 on achlavanant In tha flrat yaar af collaga. Ha found no significant dlffaranca batwaan atudanta who took 731 or aora collaga praparatary couraaa csnparad to thosa who took 59*1 or lass collaga praparatory couraaa whan ability was aqua1land. 37 ll|k School Graduating Class Slsa Several rasaarchars have attempted to determine If tha slsa of a studaat*s high school graduating class had any Influsnca on his collaga performance. Hoyt <1959) statad that rasults of 20 previous studios wara conflicting duo to Inadequate methods of analysis. In his study, ha found no significant dlffarancas whan ha compared flva subgroups of studants according to slsa of graduating class and sax using thraa a sasuras--first yaar collaga GPA, Aaarlcan Council on Education Psycho­ logical Ixaalnatlon tast scorn and high school rank (USE). Dowd (1952) studlad 19 achlaylng and 16 non-achlaylng collaga frashaan at tha Valvaralty of Haw Hampshire and found tha slsa of ona'a graduating class did not dlffarantlata among thaa. Slsa of Coaaunlty Salth (1965) studlad 156 Valvaralty of Kentucky aala frashaan to dataralno dlffarancas batwaan hlgh-ablllty achlaving and nonachlaving studants. Studants la his saapla scorad la tha uppar fifth parcantlla on tha Collaga Qualification Tests. Ea found that achlavars (• avaraga and hlghar) c a m am inly froai comuualtles of 50,000 - 100,000 population aad aoa-achlavars (C avaraga and lowar) caan from cItlas with 600,000 or mora Inhabitants. Ha tantatlvaly concludad that atudanta who cam s from largar metropolitan araaa possassad a oat of valuas and attitudes concerning education which seamed to make thorn awre prone to underachleveneat. Par-Pupil Expenditure of U g h School District Gallant (1966) found no pattern of relationship evident batwaan collaga achlavamant on par-pup11 axpaadltura of high school districts whan ha studlad salactad aspects of 663 students' backgrounds fron Ashland Collaga and Kant State University. 38 Typ« of High School Shuey (1936) analysed tbs acadeaic a u c c u a of 189 aatchad palra of frashaan who attended public aad private schools, and who antarad Bandelph-Macon Woaan*s Collaga. Shuoy fouad that public school studaats sarnad sigaificantly highar gradas than studaats fron private schools. Studaats wara aatched for age, intelligence, acadaaic load, section of country aad hosM-town sixe. Shuay also cited several studies, the results of which favored public school graduates over parochial or private school graduates. ■ill (1961) iavestigated the scholastic success of 103 satchad pairs of collaga frashasa at Ball State Teachers College who attended public aad parochial secondary schools. He fouad students frea public schools earned superior grades when scholastic aptitude was controlled. The review of the literature was presented la three phases: general overview of the subject of over- aad underachleveneat, factors unrelated to this study, aad factors related to this study. Aa underachiever was defined as *'a student who has the ability to achlave a level of acadaaic success significantly above that which he actually attaiaaM (Petersen, 1963); however, it was shown that this definition ia deceptive since different operational procedures often Identify different students as over- and undarachievers. It was further pointed out that under- and overachlavsasnt actually night he wore accurately called under- and everpredlctlon. Major operational designs were reviewed, after which other approaches frea the literature were cited and research inadequacies briefly discussed. Attention was then focussed on the individual and possible techniques and procedures which night help hia escape his problea of poor achleveaent. Several 39 bibllographla* and r c r l m of the literature regarding o»tr- and under­ achievement vara cltad to and tha flrat aaction of the chapter. The aacond portion of tha chaptar wta diractad toward a briaf review of factora not analysed In thla thesis, hut deemed necessary to provlda a hattar understanding of tha amnlfold facata of under- and overachlavmmt. Personality and biographical variables conatltutad dir la Iona of thla aaction. Although no alngla paraonallty factor universally appliaa to all typaa of atudant populations, In ganaral It aaana that achievers and/or ovarachiavara have tha following charactariatlca coaiparad to underachievers: (1) hattar adjustment, (2) hlghar motivation, (3) morn maturity, (A) hattar organisation, (5) mora efficiency, and <6) laaa hoatlllty toward paraata and othar authority flguraa. Biographical Invaatorlaa wara found to ha gaaarally uaad In an attaa^t to find non-Intellective factora which would algnlfleantly add to prediction of criterion variance above and beyond that accounted for by Intellective teat maaauraa. Illton and Myera (1967) damonatrated that auch afforta aearned to hava bean In vain. Individual factora aaaamd to vary greatly from population to population. Tha third and final aaction of thla chaptar dealt with eleven of tha non-lntallactiva factora and one intellective factor (reading ability) laraatlgatad in tha atudy. Findlaga are ainmarliad aa fallowa: (1) Parguhar and Fayne (1964) and Lavln (1965) atated that tha sexes ahauld be atudiad aaparataly. Several atudiee ravlawad seamed to damanatrata that dlffarancaa actually do occur among varloue factora analysed. When a sample of atudanta not aaparatad by aax are identified aa under- and ovarachiavara, more boya uaually are in tha underachlever group and mora glrle in tha everachlever bracket. 40 (2) It seems that research results indicate that underachievers have a lover self-concept of ability than normal and overachlevers; however, many underachievers can be helped to improve their self-concept and hence their poor academic performance through appropriate counseling focussed on their poor self-concept. Further, underachievers seem to under-predict their future grades aoreso than normal and overachlevers. (3) It seems that no significant correlation exists between academic and nonacademic (extra-curricular) achievement. (4) in general, underachievers have poorer study habits than normal and overachievers. (5) Tentatively, higher achievers seem to have more test-vlseness than lower achievers. (6) Evidence seems to conflict concerning whether or not early college underachievers significantly differ from normal and overachievers with regard to certainty of choice of academic major. (7) It seems normal and overachievers have a higher level of reading ability or perforswnce than underachievers. (8) - (10) Wo pattern of relationship seems to exist between overand underachievement and the influence of high school curriculum taken by a student, size of his high school graduating class or the perpupll expenditure of his high school district. (11) Very tentatively, students who coma from metropolitan areas (600,000 or sore Inhabitants) soy underachieve more often than students from ssmll population concentrations. (12) It seems more underachievers cost from parochial high schools than from public high schools. Finally, It must be remembered that no variable has been found to have universal application to all under-, norsal- or overachievers in all types of populations or In any single population. 41 la the M k t chaptar, tha raaaarch daalga la daacrlhad, a daacrlptlan af tha atudy papulation and aaapla, hypathaaaa and data analyala. Including laatruawntatloa, CMAFTUL m i l TB DBSICM Introduction Thla choptor conntoto of dascriptlons of tho population and sanplas, instrusMntatioa, rasasrch daaiga, hypothasas, nathod of data collactlon and proparatlon and statistical procaduros usad to aaalysa data. Population Tha papulation for this study consists of all frashnan who antarad Michigan Stata Va Ivors Ity O M V > in tha fall t o m of 1968. Tha Ragistrar's Offlea racordad 7474 now studants who ragistarad for cradlt coursas at that tint, not counting transfar studants. Maw studants axclwdad fron tha study population not ana or nora of tha following crltarla: 1. Rasidad outsIda tha Unitad Statas of Anarlea, 2. Vara daslgnatad as apaclal part-tina studants, 3. Carrlad lass than six cradlt hours of classas, 4. Droppad out of MSU hafora conplatlng fall tarn, 5. Mad Incosgilata taat scorns, 6. Mad unlatalllglhly cadad tost scorns, 7. Had lnconplata data cards. 42 43 Sampla During O r l M U t i o a Waak nil aatarlag atudanta warn inn true end tn go to n cortnln location to taka tha Acadaailc Invantory (Al). At aach location, ain forma of tha AI, A through F, vara laid out on armchalr-daaka in alphabatical ordar. of tha inatrumaata. Abaut 4355 fraahawn complatad ona Sampla a isa for aach form vaa approximataly 1100, axcapt for Form F which auabarad about 650. Only Sampla C, conaiating of 1093 atudanta, waa choaan for thla atudy. Ina tr uamn ta tion Inatrumaata admlniatarad to all or moat now fraahmaa wara Form C of tha Collaxa QuaIlficatloa Taata, MSP handlax Taat. and MSP Knxliah Taat. Cach atudaat la tha aamplaa complatad ona of tha aix forma of tha Acadawlc Inventory. Collaxa Qualification Taata Tha Collaxa Qualification Taata (CQT) (Baanatt at al, 1957) conaiat of thraa ability taata: and Informational (75 itama). Farbal (75 Itama), Pumarleal (50 itama) Half tha Information taat itama daal with aclanca; tha othar half with aoclal atudiaa. ba darivad for aach half. Saparata acoraa may Scoraa for aach taat ara glvan aaparataly and than combinad into a total acora. All aix CQT acoraa wara uaad in thia atudy. Tha CQT-Total acora aaama ta hava battar pradlctlva powar for aarly collogo achlavamant than do individual taat acoraa whan uaad aaparataly. Applaton (1945, p. 41) lndicatad that carralatlona from .50 to .70 ,(aaam to ba tha uaual flndlaga** whan ralating total-acora to aarly collaga parformaara. In ana longitudinal atudy, Juola (1943) datarmlnad that tha CQT-Total acora waa aapaclally uaaful far pradicting a atudant'a firat quartar gmda-point avaraga (GPA). Flrat quartar GFA waa tha 44 criterion •£ « c k U v « M n t uatd la tills atudy, Correlation b a t m a n the CQT-Total acora and first quartar GFA for tha study sampla was .44 for males and .50 for females. In tha CQT manual (Bennett, at al, 1957, p. 2 7) total-score reliability coefficients of .97 and .94 were reported for groups of freshmen men end women, respectively, from two state universities. These coefficients were obtained by using the split-half method, In which differences between scores of odd and even test Items are compared. Individual test score reliability coefficients for the groups ranged from .81 to .75 for men and .78 to .94 for women. Science and Social Science scores had the levest coefficients. Mlchlaaa State University BeadInn Test The MSB BeadIns Test mss developed by the Office of Evaluation Services. The test was designed to measure a student's ability to comprehend Ideas eapressed In paragraphs representative of these found In textual materials of various academic areas at MSB. The test con­ sists of 50 Items and Is used on a supplementary basis for selecting students for the Preparatory English Program as well as for selection Into honors programs. Bellablllty of the test has been estimated on several occasions by the Office of Evaluation Services to be approximately .80. Corre­ lation between the reading test and first quarter CPA for the study sample was ,45 for males and .49 for fsarnies. Michlaaa State Bnlvorslty Enallah Tost The MSB Enallsh Tost was developed by the Office of Evaluation Services. The tost mss designed to measure a student's proficiency in grammar and expression. It consists of 38 objective Items repre­ senting several aspects of English usage and 1s primarily used to select students requiring assistance in the Preparatory English Program. 45 Reliability of tha test haa been eatlaw ted on aeveral occasions by the Office of Evaluation Services to be approximately .80. Corre­ lation between the test and first quarter GFA for the study sample was .39 for males and .44 for females. Academic Inventory1 The Academic Inventory (AX) was developed by the Office of Evaluation Services to assess high school background, preparation and academic skills of incoming freshmen and transfer students. It was expected that Vnlversity College faculty could determine if the college mas adequately and appropriately meeting the needs of entering students by evaluating student responses to items in the inventory. Each of the six AI forms consisted of two sections. Depending upon the form, the first section contained from 80 to 89 items of a non-Intellective nature. Items 1 to 23 were identical on all six forms. These items concerned such things as also of the student *s high school graduating class, slxe of the community in which his high school was located, type of high school he attended--publie, parochial, prlvate-and its administrative arrangeiMat; i.e., three-year senior high with a three-year Junior high school, and information about courses he took in grades 9 through 12. Reswialag items in part one were grouped according to content, although not all subjects were the same on each form. Each student was asked questions about the following subjects: 1) Books he mao rsqulred to read for class, 2) Innovations in his high school's curricula and instructional mathods , 3) Bis personal study and objective test-taking habits, 4) Extra-curricular activities avallabls in his high school, a p p e n dix A. 44 5) His ouc>o£-cl«ss accomplishments, 4) A prediction of his first yt«r's academic performance la college. Tha second part of Forms A to E Items of a cognitive nature. c o b s Istod of 84 four-alternative Each Item was designed as an entity er "task" In Itself to determine degrees of proficiency of skills In various academic areas of students at the time of their matriculation Into MSI. Major areas were smthematlcs, physical, natural and social sciences, and the fine arts. Perms A to E each contained a dlfferant set of Items, although their content was similar. As an added measure, each student waa asked to Indicate the level of certainty or confidence with which he answered each Item according to a five-point scale: 1) very certain, 95X sure; 2) almost certain, 75X sure; 3) educated guess, SOX sure; 4) remote chance, 3SX sure; 5) pure guess, 25X sure. Form F differed from this procedure hy duplicating Itesw from Form A hut offering a fifth alternative, HI Don't Know**, and eliminating the indication of confidence level. Mo total scoro on either part was expected or computed. Data from part two were not analysed in this study. Kesearch Design The primary objective ef this study was te determine non-intellective factors characteristic ef each of three groups of students, fer each sex, Identified as under-, ever- or normal achievers. The precedure chosen to Identify these groups for this study was the "ftegresslea Medel Selection,** described hy Farquhar and Fayme <1944), In which "a regression equation Is used to predict achievement from aptitude measures. Vader- aad over-achievement Is then determined on the hasIs of the discrepancy between predicted and actual achievement.** 47 Thorndike (1943) also 41acutaa4 tha ragraaaloa aodial la a chaptar entitled "laait* II-C: Caacurraat Comparison of Contrasting Groups." Ha stated (pp. 59-41) that tha advantage of this method is that it ...caa provide a more sensitive taat of tha existence of a re­ lationship par caaa completeIt tested thaa doas a corralatloaal analysis of a csmpleta, latact group. By taking casas at tha extremes (assuring a llaaar relationship), we gat casas la which aay lafluaaca will haws tha M x i s u a opportuaity to show itself, wa ara, la effect, putting a Magnifying glass upon tha relation­ ship that wa ara trying to discover... If wa caa safely think of ''degree of achievement in relation to expected achievement" as a single continuous variable, differing in degree hut not in kind, the use of the "overachlever" group may be expected to provide the msrimum amount of information for tha amount of data gathered. Tha sharp difference in achievement between the contrasting groups will make them arnre sensitive, case for case, to any genuine differences in related variables. Thus, this becomes an efficient experimental design. However, for the results from such a contrast of extresm groups to be interprotable, we must assise that the "everachiever" differs only quantltatlvely--aot qualitatively--from the "under­ achievers". .. Insofar as "over-" and "underachievement" are qualitatively different phenomena, with different causes and cerrelatas, the camperiaon of extreme greupa may be ambiguous and confusing. The comparison of "underachievers" with a group of average or normal achievers may be less efficient in bringing out differences between the two groups, but the differences that are established will be more clearly associated with "vaderachlavement" per so. Insofar as our interest focuses upon the "underachiever," the strategy of using a group of average achievers as the contrasting group will be the safer one and tha one leading to mora clear-cut interpretations. Lavln (1945) has offered cautions to be used in Interpreting results obtained in studying extreme groups. He stated that tha middle group ought not to be dropped out because more information would be obtained from studying all three groups. Further, he questioned the assumption that a linear relationship exists between variables of extrema groups. Therefore, with the cautions ef Thorndike and Levin in mind, it was decided to study all three groups and not Just tha extremes. It was expected that more information would be provided about the sample; 48 I.e., * aer* aecurit* M t l a i M of a alapla linear ralatloaahip laaag tha groups for a givan varlahla or evidence of the proaone* of a nonllaear relationship (cf. Feldwaa and Ravcoab, 1989, pp. 285*6). Tho rogroooloa lino required by tho rogrooaloa aodol technique was determined by prodictlag flrot tona collogo GFA from CQT“Total acora. Flus-aad*mlaus oao ataadard orror of estimate^ waa aoloctod a priori aa tho boat level to difforontlato tho throo groupa. A program writtoa by gublo aad Faftor (1969) waa uaad to obtain tho prodlctod GFA, tho dlfforoaco botwoon tho actual GFA and prodlctod GFA, aad tho atandard orror of estimate. Studonta ono standard orror of estimate bolow tho rogroasloa llao woro doalgaatod underachievers, atudoata oao standard orror of estimate abovo tho rogroasion llao woro doalgaatod overachievers, aad tho remaining atudoata within oao standard orror of ostiaato of tho rogroasloa llao woro doalgaatod aa noraal achievers. waaaor: Throo sots of throo groups woro ldontiflod la this (1) aad (2) aaloa aad fesmles aoparatoly--to tost Hypothoaoa I, II aad III, aad (3) not on tho basis of aox-*to tost Hypothoala IT. Hypothoaoa Tho following null hypothoaoa woro forwulatod frow previously stated purposes aad Investigative questions. Tho groupa referred to la tho hypothoaoa woro undorachlovora, aorwal achiovors aad overachievera. Hypothesis X Ho differences existed at tho tiwe of university watriculatlea awoag tho groups on tho following non-Intelloctivo characteristics: (1) Selected high school daw ographic characterlaties. (2) High school curricula taken. Htale group ■ 0.677; foaalo group * 0.619; group not chosen by sox ■ 0.650. 49 (3) Participation ia aalaccad oxtra curricular activities la hlfk school, (4) Study hahIts, (3) OhJactIt s test-taking habits ("test-wlseaess"), (6) Reasons for attaadlag collaga, (7) Self-concept of ability as ladlcatad by salf-expectation (prediction) of academic achievement In tha first yaar of collaga, (8) Daclarad/uadaclarad amjor. Hypothesis II ■o dlffaraacas existed at tha time of university matriculation among tha groupa oa salactad taats of acadsale aptltuda. Hypothesis III Mo dlffaraacas axlstad at tha and of tha first term of ualvarslty study among tha groups on: (1) First tern average cradlt hours carrlad, (2) First term average cradlt hours aaraad, (3) First tara grada point avaraga. Hvpothasls IF there was no dlffaraaca la tha ratio of aalas ta faaalas within aach group (whan tha groups wara not chosaa on tha basis of sax). Data Fraparatlon Data far aach atwdaat wara put oa thraa data procaaslag cards. Two cards coatalaad all raspoasas to ltaws on tha first half of tha Acadswlc Iavaatarr aad ona card coatalaad all tost acora data aad othar aacassary Information. Data wada avallabla farthis study fraw tha Registrar's Offlea wara aach student's cradlt hourscarrlad, 50 credit hours urnad, and GFA for tha 1968 fall tan. Remaining data wara obtalnad fro* tha Offlea of Evaluation Sarvlcas. Statistical Analysis All computations of data analysis wara parforaad using althar tha COC 3600 or CDC 6500 computer. ^ Tha Calculation of Least Squaras prograsi (Rub la and Fafter, 1969) was usad to compute tha statistics nacassary to identify tha thraa groupa studiad in this rasaarch project. This procedure was discussed in greater detail In tha previous section concerning rasaarch design. Chi-Square analysis, using tha ACT program (Leegold, Zarby and Foster, 1969X was performed to analyse data for tasting Hypo­ theses 1 and IV. A multivariate analysis of variance program (Finn, 1968) was used to analyse data for tasting Hypotheses II and III. The population of the study consisted of all new freshmen who entered Michigan State University in fall, 1968. Studants wara deleted from this population who wara classified as special studants, who resided in a foreign country, who dropped out during the quarter, who carried less than six credit hours of classes, or who had Incomplete or unintelligible test score data. Each student in the study population completed Form C of the College QualiflcatIon Tests. the Michigan State University Reading Test and the Michigan State University English Test. From this population, six random samples of students completed one of six forms, A through F, of a specially constructed instrument, the Academic Inventory. The instrument was deslgnad to ascertain the levels *Use of the Michigan State University computing facilities was made possible through support, in part, from the National Science 51 of acadeaic proficlancy n w studants wara bringing with thea to Michigan Stats University as wall as salactad charactarIstlcs of thalr high schools and personal acadsalc habit patterns and experlances. Tha saapla which coaplatad Fora C was chosen for this study. Qypotheses wara developed to test differences aaong three groups Identified as under-, over- and noran1 achievers. Differences were based on Intellective test scores, GPA, and responses to non-intell­ ective iteas of a deaographlcal and biographical nature on the first half of Fora C of the Acadeaic Inventory. Student data were put on data processing cards and analysed on either the CDC 3600 or 6500 coaputer. Various statistical analyses, Including linear regression, a sniltlvarlate analysis of variance, and Chi-Square analysis were perforaed using prograas available for Michigan State University researchers. The following chapter is concerned with results of dace analysis. CHAPTER FOUR AHALYS1S OP DATA Introduction The purpose of this chapter is to present the analysis of data as described in Chapter Three. ing upon their nature. Data were analysed in several ways depend­ Tables of statistically significant results are included in the body of the chapter. Tables of statistically non-slgnlfleant results nay be found in Appendix 1. Tha data analysis is presented in four parts: (1) Hypothesis I, (2) Hypothesis II, (3) Hypothesis III, and (4) Hypothesis IP. Groups nentionsd are underachievers, normal achievers, and overachievers. The nunber of students in each achievement level, when the levels were identified separately by sex are given below (Table 4.1). T A B U 4.1 HIMBER OP STUDDVTS IH EACH ACHUVBfEITr LEVEL BT SIX Level of Achievement Sex Male Under Hormal Over 78 356 86 520 Total 52 Pernsle 81 414 78 573 53 Hypothesis 1 Ho differences existed et the time of M t r leu let ion asong the groups on the following non-Intellective characteristics: (1) Selected high school demographic characteristics, (2) High school curricula taken, (3) Participation in selected extra-curricular activities in high school, (4) Study habits, (5) Objective test-taking habits ("test-wlseness"), (6) Reasons for attending college, (7) Self-concept of ability as indicated by self-expectation (prediction) of acadesilc achievement in the first year of college, (8) Declared/un­ declared major. The Chi-Square analysis technique was used to test Hypothesis 1. Results of the analysis are Included in the following subsections. Statements of rejection or non-rejection of each part of the hypo­ thesis are Included at the end of the discussion in each sub-section. The .05 level of confidence was established as the critical level In testing all hypotheses. In sosm cases, cells had to be collapsed because not enough expected frequencies occurred in two or more cells to provide an accurate analysis (Walker and Lev, 1953). Selected Hlah School Demographic Characteristics The first eleven questions in Part I of the Academic Inventory* were concerned with various demographic characteristics of and personal opinions held by each student about the high school from which he graduated. Questions dealt specifically with number of students In one's high school graduating class, type and administrative arrangement of the high school, percentage of fellow graduates definitely planning to go on to college Immediately, description of the community in which one's high school was located, one's mode of transportation to school, and age of the school building. Kach student was also asked his personal opinion regarding his high school's financial support, instructional procedures and success in preparing its graduates for college work. ^Appendix A 54 It was hypothesized froa the literature review that no dlffaraacas aaong sizes of high school graduating classas or tha financial support of tha high school would ba discerned aaong tha groups. No dlffaraacas wara found to occur upon analysis of itans 1 and 2 (Appendices g . 1 and B .2). In Itaa 4, aach student was asked to identify tha type of high school froa which ha graduated--publie, parochial (church related) or private. It was anticipated froa tha literature review that a greater percentage of public high school graduates would be norwal and over­ achievers than parochial school graduates. Contrary to expectation, no significant differences occurred aaong achieveasnt levels aaxmg the three types of graduates for feswles, and a larger percentage of parochial than public high school graduates were noraal and overachievers for aales (Table 4.2). Further, one aight expect private school graduates to perfora better than the public or parochial school grad­ uates. Results of data analysis Indicated that no significant differences occurred for feaales aaong the types of graduates aaong achieveasnt levels. For aales, a auch larger percentage of private school graduates than public and parochial school graduates were class­ ified as underachievers. Results of analyses of Itea 7, pertaining to the else of the coasunlty In which each student's high school was located (Table 4.3), were significant only for feaales, but were difficult to interpret aeanlngfully. Noraal achievers caae aore often than the other groupa froa high schools in suburbs of large cities (200,000+) and saall city or rural schools than frea otber-alzed coaninltles. Underachievers caae aoat often froa saall city or rural schools and about the saae percen­ tage of under- aad overachievers caae froa high schools In large cities. 55 TABLE A. 2 PERCENTAGE OF STUDENTS BY ACHIEVEMENT LEVEL AMD SEX GRADUATING FROM PUBLIC, PAROCHIAL, OR PRIVATE HIGH SCHOOLS Level of Achlevesmnt Under Normal Over Total Public Parochial 14* (15) 68 (72) 18 (13) 100 (100) 9 (7) 82 (72) 9 (22) 100 (100) X2 Private A0 (9) A8 (78) 12 (13) 100 (100) 17.681** (5.233) *First set of figures for males : figures in parentheses for feswles. **Slgnifleant at the .05 level or beyond. TABLE A.3 PERCENTAGE OF STUDENTS BY ACHIEVEMENT LEVEL AMD SEX ATTENDING HIGH SCHOOLS FROM VARIOUS SIZED COMMUNITIES Level of Achievement Under Normal Over 200,000+ Suburb of 1 18* (20) 16 (10) 12 (18) 19 (20) 30 (35) 35 (2A) 50,000199,000 1A (1A) 13 (13) 16 (21) Suburb of 3 Small City, Town, Rural 6 (9) 8 (9) 3 (12) A2 (38) 33 (3A) 3A (26) X 9.178 (19.067! *First ««t of figures for a i l u ; figures in parentheses for f m l t s . **Slgnlfleant at tha .05 laval or beyond. Mo directions wara found In tha literature reviewed which would have given tentative expectations for results of analysing Items pertaining to Instructional procedures In one's high school (Item 3), approximate percentage of fellowgraduates definitely planning on ismedlately going to college (item 5), one's evaluation of high school preparation for college (Item 6), degree of Industrialisation of the high school 56 community (item 8), the administrative arrtag«Mnt of tho high school (item 9), one's aod* of transportation to high school in his san lor year (ltas 10), and approximate age of the high school physical plant (item 11). Results of data analysis for the above items indicated no signifi­ cant differences occurred among achievement levels for either sex, except for faMles on item 6, pertaining to the student's evaluation of his high school preparation for college (Table 4.4). Almost three- quarters of female students who expressed a "very poor" evaluation of high school preparation for college work were underachievers; whereas, the vast majority of students expressing a "very adequate" evaluation of high school preparation were normal and overachlevers. TABLE 4.4 PERCENTAGE OP STUDENTS BY ACHIEVEMENT LEVEL AND SEX EVALUATING THEIR HIGH SCHOOL PREPARATION FOR COLLEGE ACCORDING TO FIVE CRITERIA Level of Achlevesmnt Under Normal Over Total Vary Adeq 14* (9) 74 (74) 13 (17) 100 (100) Above Ave 15 (15) 66 (75) 19 (11) 100 (100) Below Ave Ave 16 (12) 68 (74) 16 (14) 100 (100) 17 (18) 68 (83) 15 (18) 100 (100) Very Poor 0 (73) 88 (18) 13 (9) 100 (100) X2 4.319 (38.345)** *Flret aet of figures for males; **Slgnlfleant at the .05 level or beyond. In conclusion, this section of Hypothesis I, which pertained to selected demographic characteristics of the high schools from which students in the study sample graduated, was rejected for both sexos. Differences were found to occur on some of the characteristics investigated. 57 Hifih School Curricula Taken It waa hypothaalead from tha lltaratura raviaw that faw if any aignificant dlffarancaa would ba found among tha achiavamant lavala ragarding tha kinda and nuafcar of tarma of couraaa thay took in high achool. For malaa, only thraa Itama in twanty-aavan indieatad atatiatically aignificant dlffarancaa--partainlng to numbar of art (itam 12), phyaical aciamca (itam 15) and political aclanca (itam 25) couraaa takan (Tablaa 4.5 and 4.6). Undarachlavara took laaa t a n a of art couraaa and mora of phyaical and political aclanca couraaa than normal and ovarachiavara. Anothar quaation (itam 28) partalnlng to art waa not aignificant for malaa. Only thraa itama wara aignificant for famalaa--nunbar of couraaa takan in phyaical aclanca (itam 15), rnathaamtica (itam 22) and vocational agrlcultura (itam 30) (Tablaa 4.5 and 4.4). Undarachlavara took mora t a n a of mathaamtica and phyaical aclanca than tha othar groupa. Xtam 30 waa not hald to ba maaningful, howavar, bacauaa of tha aubjact and tha nuabara of famalaa not avan taking ona couraa. Thia aaction of Hypothaala I waa not rajactad for aithar aax. TABLE A.5 PERCENTAGE OF STUDENTS BY ACHIEVEMENT LEVEL AMD SEX TAXIMG VARIOUS NtMBERS OF TEMS OF COIRSES DV SELECTED AREAS IM GRADES 9 - 12 Itea Number A Course Area 12. Art 13. 14. IS. 16. 17. 18. 19. 20. 21. 22. 23. 1 Underachievers 2 3 4 $8* (68) 68 (41) 4 (5) 1 (20) 37 (21) 4 (5) 1 (4) 13 (7) 1 (0) 5 (0) 5 1 Normal Achievers 2 3 4 5 Overachievers 2 1 3 4 5 X2 • * 12 0 16 9 20 75 72 8 9.764**+ (25) (4) (U) (3) (63) (23) (8) (4) (2) (66) (19) (8) (1) (6) (8.703) Music 9 6 5 12 17 7 6 12 57 60 13 20 3 3 9.611 (25) (6) (li)(17) (37) (25) (10) (6) (22) (40) (23) (18) (5) (14) (12.150) 69 Biological 23 6 71 21 2 3 1 1 64 1 29 1 3.972 5 (61) (28) (6) (-> Sciences (12.210)++ (3) (73) (20) (3) (-) (4) (83) (12) (1) (-> 21 8 6 Physical 18 53 4 28 48 13 1 18.020** 21 60 16 1 Sciences (36) (31) (14) (-) (20) (46) (28) (5) (“) (13.498)**++ (21) (55) (19) (5) (-) Cowercial 53 3 0 54 8 37 40 7 1 1 53 6 0 1 3.369 (11.222) Arts (54) (17) (2) (5) (59) (16) (6) (2) (17) (61) (12) (3) (6) (18) English 21 13 22 13 16 17 10 49 6 6 46 17 14 47 4.991 (13.291) (8) (13) (19)(10) (50) (4) (10) (28) (6) (51) (ID (10) (15)(59) Literature 10 2 11 32 16 18 41 30 23 40 29 1 6 39 4.551 (20) (15)(46) (12) (4.448) (23)(18) (48) (23)(15) (49) (15) (2) (9) (1) Foreign 22 34 21 9 49 15 14 13 42 15 7 11.541 13 16 16 (10.440) Language (6) (46) (19)(22) (5) (8) (33)(17) (37) (3) (8) (36)(24) (29) 26 50 2 History 11.064 13 10 16 44 27 1 14 11 15 43 27 (20) (49) (17)(14) (7.296) (0) (20) (45)(26) (8) (0) (26) (41)(27) (6) Social 50 10 4 6 31 45 38 8 51 30 6 4.806 3 8 5 Sciences (54) (36) (10) (-) (6.845)++ (5) (50) (37)(10) (-) (8) (54) (31) (8) (-) Mathe­ 9 17 74 13 21 6 16 7.101+++ 78 65 matics (18.918)** (0) (6) (25)(37) (32) (0) (12) (35)(21) (33) (1) (6) (20) (27)(45) 72 0 6 Vocational 18 4 20 3 22 7.082 67 6 3 67 1 6 3 Education (53) (30) (10) (32) (3.670) (58) (29) (81 (41 (1) (54) (9) (3) (3) Key: 1. Did not taka courses in tils area 3. Took three or four terms in this srea 2, Took one or tiro tons in thisarea 4. Took five or six tone in this tree 5. Took seven or tore terms In this tree ♦First sot of figure* for males, figures la parentheses for feaales; ♦♦Significantat the ,05 level or bevond: -tCaanutad with 4 D.F. calls V k sad I l « u t i tn > * TAILS 4.6 PERCENTAGE OF STUDENTS IT ACHIEVEMENT LEVEL AND SEX TAKING VARIOUS NW1ERS OF TERMS OF SELECTED COURSES IN GRADES 9 - 12 Iten Nwber 6 Course Area 24. Statistics Undarachievers 1 2 3 4 85* (86) 44 25. Political (44) Sciaact 26. Anerican 3 Litarat. (4) 27. Enflish 5 Litarat. (7) 28. Art 91 (70) 29. Music 73 (40) 30. Vocational 94 Agricul. (99) 100 31. H o b s Ecoa. (53) 32. Industrial 74 Arts (55) 33. Europaaa 55 History (44) 34. World 29 History (15) 73 35. Geography (45) 14 (14) 31 (30) 21 (21) 29 (22) 3 (12) 1 (11) 1 (1) 0 (11) 6 (2) 26 (21) 14 (25) 12 (23) Horns1 Achievers 1 2 3 4 5 5 1 (0) 19 (19) 54 (47) 53 (54) 6 0 0 (0) (0) 3 3 (4) (4) 18 5 (9) (20) 2 14 (2) (14) (9) (2) (6) 18 1 (2) (31) 0 5 (0) (-) 0 0 (4) (ID 1 3 (1) (0) 1 1 (5) (1) 1 1 (0) (4) 0 1 (1) (0) 6 (1«) 0 (0) 0 (21) 15 (1) 17 (28) 54 (57) 14 (18) m m 87 (97) 46 (49) 3 (3) 7 (7) 78 (64) 60 (39) 95 (100) 96 (59) 67 (98) 44 (49) 17 (20) 62 (67) 10 (7) 39 (33) 22 (17) 24 (24) 10 (10) 9 (13) 1 (0) 4 (14) 10 (2) 24 (24) 23 (19) 22 (19) 2 (1) 12 (14) 52 (54) 50 (50) 1 (13) 8 (14) 1 (0) 0 (17) 15 (1) 26 (24) 54 (57) 14 (13) Overachlevers 2 1 3 4 0 0 (0) (1) 2 2 (2) (2) 8 15 (9) (17) 8 11 (8) (U) - - (2) (11) 2 21 (4) (30) 0 3 (0) (-) 0 0 (1) (8) 1 6 (0) (0) 3 2 (1) (1) 3 3 (2) (2) 1 1 (0) (1) 84 (96) 45 (44) 1 (1) 8 (10) 80 (68) 63 (44) 97 (97) 92 (55) 65 (95) 44 (55) 21 (25) 66 (74) 12 (4) 29 (36) 16 (18) 17 (18) 9 (9) 9 (12) 1 (0) 8 (18) 9 (3) 27 (18) 14 (12) 23 (15) 5 0 0 3 (0) (0) (0) 0 24 1 (18) (0) (3) 6 17 59 (58) (12) (12) 17 1 56 (55) (9) (8) 10 (8) (3) (13) 0 3 24 (10) (1) (33) 0 2 0 (3) (0) (-) 0 0 0 (15) (1) (10) 17 2 6 (0) (1) (1) 2 24 2 (22) (3) (3) 2 3 59 (60) (0) (4) 1 0 9 (10) (0) (0) X2 4.154 (7.766) 13,671**+ (6.387) 4.542 (5.037) 11.835 (7.068) 7.059++ (5.024) 11.486 (2.928) 2.760 (15.335)** 8.361 (6.021) 6.333 (12.929) 5.270 (10.184) 11.687 (9.920) 8.032 (9.871) TABLE 4.6--Continued Itew Muaber A Course Ares Vnderachlevers 1 2 3 4 1 34. Aeronau­ 99 tics (100) (0) 74 37. Philosophy 19 or Logic (73) (20) 3B. General or 74 8 ConsuMr (67) (7) Math. Key: 5 0 0 0 (0) (-) (-) 3 1 (4) (0) (4) 12 4 0 (1) (12) (73) 1. lever studied this subject 2. One tern or seaester Normal Achievers 1 2 3 4 98 1 (99) (1) 81 17 (71) (16) 77 8 (7) (8) 3. 4. 5 0 0 0 (0) (-) (-) 0 2 (5) (1) (1) 1 7 4 (2) (10) (69) Overachlevers 1 2 3 4 1 98 (97) (3) 12 84 (71) (19) 10 45 (9) (6) Two terM or aoMSters Three tens or s u m s ters *Pirst set of figure* for Ml***, figure* in parentheses for feMle*. **Signifleant *t the .05 level of confidence or beyond. -tCoaputed with 6 D.P. becsuse cells 4 end 5 were collapsed. -HConputed with 4 D.P. because cells 3, 4 and 5 were collapsed. 5 1 0 0 (0) (-) (-) 0 2 (8) (0) (3) 1 8 15 (1) (14) (14) 5. X2 2.252 (2.999) 7.471+ (9.762) 10.431 (5.483) Pour or nore tens or seaestera 61 Participation in Salactad Extra-Curricular Activities In alavan itens in tha Academic Inventory (56 through 66), aach student waa asked whether selected extra-curricular organisations and activities wara available in tha student's high school, and, if so, tha degree to which ha actively participated in then. In nine items, 67 through 75, each student was requested to indicate the nuad»er of Mout-of-class" experiences or accomplishments he had in the area of social science in high school (Appendix B.9). Based on the literature review, it was hypothesised that no differences would be noted asong the twenty items devoted to this subject. (Science Club) was significant (Table 4.7). For Bales, only item 62 For females, only item 64 (Debating Club) was significant (Table 4.8). In both cases, a greater percentage of underachievers indicated the club did not exist in their high school as far as they knew. No significant differences were found aaong the groups for either sex regarding accomplishments or experiences of a social science nature (Appendix B.10). TABLE 4.7 PERCENTAGE OF MALES BT ACHIEVEMENT LEVEL INDICATING THEIR PARTICIPATION IN A HIGH SCHOOL SCIENCE CLOB Level of Achievement Under Noraal Over Not Available Actively Partlcipated 49 32 37 17 13 17 Did Not Participate *Slgnlfleant at the .05 level of confidence or beyond. 35 55 45 X2 12.339* 62 TABLE 4.8 PERCENTAGE OP FEMALES BY ACHIEVEMENT LEVEL INDICATING THEIR PARTICIPATION IN A HIGH SCHOOL DEBATING CLUB Not Available Level of Achieveaant Under Noraal Over 37 16 17 Actively Partlcipated Did Not Participate 13 19 8 50 65 76 X2 27.047* *Slgnlfleant at the .05 level of confidence or beyond. Since analysis of data for only ona item for aach sax showed significant differences aaong the groups, it seeawd results occurred due to chance. Therefore, this section of Hypothesis I was not rejected for either sex. Study Habits In nine iteas, 42 through 50, each student was requested to indi­ cate his general study habits. No differences were noted for sales, but significant differences occurred in two iteas for feaales--ltesw 43 and 44 (Table 4.9). It waa expected froa the literature reviewed for this study that soae trends or differences in study habits aad attitudes alght be discovered indicating that underachievers woula have poorer study habits and attitudes than the other groups. In the first ltea, each student was ashed If he tried to ask hiaself questions and to answer thea as he studied. In the second ltea, each student was asked if he regularly reviewed his class notes froa lectures and assigned readings. Overachievers had better study habits than the others. TABU 4.9 PERCENTAGE OP STUDENTS BY ACHIEVEMENT LEVEL AMD SEX INDICATING THEIR GENERAL STUDY HABITS USING A FIVE-POINT SCAU 42. In taking notM on required library reading*, do you try to record the exact sentence* as written rather than rewriting the thoughts in your own words? 43. Do you try to ask yourself questions aad to answer then as you study? 44. Do you regularly review your class notes fron lectures and assigned readings? 45. Is the final copy of your lecture notes organised in som kind of an outline forn? 46. Ilhen you listen to a lecture, do you try to copy down as nuch as you can, rather than relying on key words to help you recall general ideas? 47. When you study history, do you try to look for the relationships aaong various events? 48. In first reading an assignaent, do you try to read each paragraph thoroughly? 49. Do you pause during your study to think about the aaterlal? 50. After you learn a general principle or rule, do you try to think of exaaples which illustrate it? ltea 42. 43. 44. 45. 46. 47. 48. Hnderachievers 1 2 3 4 1* (4) 6 (12) 9 (20) 12 (12) 5 (6) 22 (17) 21 (20) 12 (13) 21 (23) 22 (19) 13 (23) 5 (12) 24 (28) 22 (22) IS (15) 32 (30) a (36) 18 (15) 13 (17) 26 (21) 22 (14) 36 (41) 31 (27) 37 (21) 31 (28) 27 (30) 22 (26) 18 (23) 5 36 (28) 9 (7) 9 (5) 27 (21) 50 (35) 4 (9) 18 (21) Noraal Achiever* 1 2 3 4 5 7 (5) 15 (14) 12 (16) 13 (19) 5 (10) 22 (19) 22 (23) 13 (15) 21 (24) 18 (26) 19 (23) 12 (13) 26 (26) 26 (26) 16 (17) 23 (22) 25 (25) 15 (20) 14 (16) 26 (22) 15 (16) 31 (35) 28 (28) 30 (25) 24 (21) 25 (28) 16 (25) 21 (21) 33 (28) 14 (11) 15 (8) 28 (17) 44 (33) 10 (7) 16 (14) Overachievers 1 2 3 4 5 (5) 17 (32) 22 (32) 14 (21) 7 (15) 23 (23) 26 (26) 17 (13) 22 (28) 23 (22) 27 (29) 13 (12) 28 (35) 24 (31) 10 (5) 27 (17) 24 (21) 13 (18) 14 (18) 27 (23) 23 (9) 34 (37) 24 (17) 23 (17) 24 (15) 30 (21) 20 (15) 20 (21) 5 34 (29) 9 (16) 7 (9) 22 (17) 36 (35) 2 (4) 7 (14) X2 7.151 (1.550) 8.720 (21.753)** 15.354 (18.830)** 7.229 (7.717) 6.153 (5.302) 6.643 (6.802) 8.815 (6.190) TABLE 4.9--Continued Itea 49. 50. Bey: Underachievers 1 2 3 4 13 33 35 15 (16) (40) (35) (4) 15 27 28 25 (17) (31) (26) (21) 5 4 (6) 4 (5) Horae1 Achievers 1 2 3 4 5 16 29 30 19 (17) (31) (33) (16) 17 30 22 23 (17) (27) (26) (24) 6 (3) 8 (6) Overachievers 1 2 3 4 23 28 28 20 (26) (29) (28) (15) 19 20 29 27 (12) (38) (19) (31) 5 I (1) 6 (0) I. Alaoet always (ovar 901 of the tlae) 3. Ofcan (261 - 741 of Che tlae) 2. Usually (751 - 901 of tha tins) 4. Soaetlnss (101 - 251 of tha tlae) 5. Barely (lass chan 101 of Che else) ♦First sat of figures for asles; figures in parencheses for feMles S i gn i f i c an t at the *05 level of confidence or beyond. X* 7,949 (15.190) 7.542 (12.383) 65 Since Che enalyels for etlei showed no statlsclcslly significant differences and since female overachievers had better study habits than normal and underachievers, this section of Hypothesis I was not rejected for males, but was rejected for fesmles. Test-wiseness In nine ltesm, 76 through 84, each student was asked about the procedures he used in answering questions on objective examinations (Appendix B.11). No significant differences occurred among the fesmles achievement levels. For M i e s , significant differences occurred only in item 79, in which each student was asked if he tended to choose one of two very similar possible answers on a swltiple choice question. The meaning was unclear, however, and the results were considered of doubtful value. This section of Hypothesis I was not rejected for either sex. Reasons for Attending College In itesm 39 through 41, each student was requested to select his first, second and least important reasons for attending college from a list of five suggestions (Table 4.10). Only item 41, concerning the least important reason, indicated significant differences asmng male achievement levels; however, the results were not possible to Interpret meaningfully because the items regarding the two most Important reasons showed no significant differences. No significant differences were noted in the items among the female achlevesmnt levels. This section of Hypothesis I was not rejected for either sex. TABLE 4.10 PERCENTAGE OP STUDENTS BT ACBIEVDtENT LEVEL AND SEX INDICATING THEIR REASONS FOR ATTENDING COLLEGE 39. 40. 41. Itea No. Which of these reasons seeas to be nos t laportant to you personally? Which appears to be next In laportance? Which of these reasons seesa to be the least laportant to you personally? Level of Achieveswnt 39. Under Noras 1 Over 40. Under Nontal Over 41. (20) 17 (18) 5 (16) Noraal Over 1. 2. 3. 4. 5. 29* (26) 15 (1») 20 (13) 33 (23) 38 (20) 30 (12) 6 Under K«y: 1 2 3 (7) 5 (2) 4 (ID 13 (15) 18 (21) 20 (23) 28 (17) 25 (25) 33 (27) 3 51 (32) 62 (25) 60 (32) 28 (35) 20 (28) 27 (35) 9 (U) 3 (13) 1 (3) 4 0 (1) 2 (2) 1 (6) 4 (7) 6 (7) 4 (9) 31 (41) 36 (39) 49 (45) 5 17 (33) 16 (45) 14 (40) 22 (20) 18 (23) 19 (20) 26 (11) 19 (5) 12 (9) X2 12.203 (9.215) 7.494 (7.007) 27.880** (13.453) For the prestige of a collogo degree It is oxpoctod in our faally To atka aore aonay To bo with school frloads Enjoy going to school *First sot of flguros for aeles; flguros in psronthosos for feaeles. **Signifleant ot tho .05 level of confldonco or boyond. Self-Concept of Ability In iteas 51 through 54, ooch studont was asked to indie*to his oxpoctod grodos for his first yoor of collogo in (1) *11 coursos taken, (2) aejor subjects, (3) required general education coursos and (4) social science coursos should ho enroll in any (Table 4.11). In a fifth question, itea 55, each studont was asked what kinds of grades ho received in high school social science courses in the last year any It ms hypothesis ad froa ths litsrsturs rtritvtl that under­ achievers would indicate lover sxpoctstlons when prsdlctlng thoir first year's acadeaic performance. In itea 51, asIs and faasla underachievers had significantly lower expectations than aoraal and overachievers for earning overall high grades in their first year of college. Further, in itea 52, feaale underachievers expected significantly lower grades in their aejor courses than the other groups. In itea 54, aale over* achievers expected significantly higher grades in social science courses for their first year in college than the other achlevesant levels. Finally, in itea 55, fesale underachievers received signifi­ cantly lower grades in high school social science grades than both n o n a l and overachievers. Hypothesis II ms Because of these results, this section of rejected for both sales and feaales. TABLE 4 . U PERCENTAGE OF STUDENTS BT ACHIEVEMENT LEVEL AND SEX INDICATING WHAT KINDS OF GRADES THEY EXPECTED IN THEIR FIRST TEAR OF COLLEGE 51. 52. 53. 54. 55. Whatkinds of grades do you expect to receive this coning year at MSU? Whatkinds of grades do you oxpoct to receive in tho subjects of your najor? Whatkinds of grades do you expect to receive in required general education courses? Whatkinds of grades would you expect to receive should you enroll in social science courses at MSU? What kinds of grades did you receive in high school social science courses in the last year you took courses of this kind? Iten 51. 52. 53. 54. 55. Key: Underachievers 2 4 3 1 30 <1> 12 (4) 4 (1) 8 (9) 40 (40) 23 (15) 52 (52) 32 (20) 44 (23) 39 (43) 1. Mostly A's; 42 13 (04) (20) 3 34 (41) (4) 17 44 (48) (22) 22 24 (38) (28) 7 14 (12) (4) 5 0 (0) 0 (-) 1 (2) 1 (1) 0 (1) 2. Mostly B's; Nornal Achievers 2 1 4 3 5 9 (3) 16 (12) 12 (0) 13 (7) 43 (49) 31 (27) 51 (59) 30 (28) 35 (37) 40 (43) 42 16 (54) (14) 7 25 (24) (0) 20 34 (47) (18) 19 31 (35) (19) 13 4 (6) (2) I (0) 1 (“> 3 (2) 2 (1) 0 (0) 3. Mostly B's & C'o; Overachievers 2 1 4 3 14 (4) 25 (17) 14 (5) 8 (8) 59 (04) 4. 43 (46) 50 (02) 40 (33) 51 (38) 27 (31) 10 33 (41) (8) 22 3 (19) (3) 16 28 (49) (12) 22 15 (38) (13) 12 2 (4) (1) Mostly C's; ♦First set of figures for neles; figures in parentheses for fenales. ♦♦Significant at the .05 level of confidence or beyond. -tCooputed with 6 D.F. because cells 4 and 5 were collapsed. 5 0 (1) 0 (-) 0 (1) 3 (3) 0 (0) a X2 23.25800 (25.886V* 11.176 (17.157)00+ 12.167 (6.906) 17.450 (9.679) 10.519 (19.142)oo 5. Mostly C's & D's. 69 Choice of College Major A freshmen say or aay not declare a major upon antarlng Michigan Stata Univarsity. Data regarding cholca of sMjor vara analysed by using a simple dichotomy--preference/no prafaranca. This analysis ravaalad no significant diffarancas among tha groups for smles or famalas (Tabla 4.13); thsrsfora, this section of Hypothesis I was not rajactad for either sax. TABLE 4.12 PERCENTAGE OF STUDENTS BY ACHIEVEMENT LEVEL AND SEX INDICATING THEIR CHOICE OF MAJOR Achievement Level Under Normal Over Mo Preference 19* (27) 27 (26) 21 (30) X2 81 (73) 73 (74) 79 (70) 2.876 (0.395) ♦First sat of figures for males; figures in parentheses for females. Hypothesis II Mo diffarancas existed at tha tism of university matriculation among tha groups on selected tests of academic aptitude. A multivariate analysis of variance (Finn, 1968) was used to carry out the test of the second hypothesis. The .05 level of confidence was chosen to test for statistical significance. academic aptitude were: Selected tests of (1) Michigan State University English Test. (2) Michigan State University Reading Test, and (3) the College Qualification Tests (CQT), Form C. The CQT consisted of three tests from which the following six scores were obtained: (a) Information- Social Science, (b) Information-Natural 8clance. (c) InformationTotal . (d) Verbal. (3) Numerical and (f) CQT-Total. 70 Tha analysis of data ravaalad tha following about tha study sampla: (1) Sax Diffarancas: Zt was axpactsd that sax diffarancas and slaillarltlas would ba found for tha study sampla ragardlng avaraga scoras on salactad tasts of acadamic aptituda which would raflact charac taris tics notad within tha study population (Offlea of Evaluation Sarvicaa, 1968). In tha study population, malas scorad hlghar than famalas on fiva of tha six CQT tasts--about four points sora on tha Munarleal. six points hlghar on tha Information Tasts (thraa points hlghar on aach information subtast) and about alavan points aora than famalas on tha total scora. Both saxas had about tha saam avaragas on tha CQT-Varbal tast and tha MSP Eaadlna Tast in tha study population, and famalas scorad about two points mora than malas on tha MSP Baalish Tast. Basults of analysis warn in accordanca with axpactatlons. Mo significant diffarancas axlstad on tha MSP Eaadlna Tast and CQT-Parbal tast scora avaragas. Significant sax diffarancas did occur on tha ramainlng tast scora avaragas, famalas parforming battar on tha MSP English Tast and am las scoring battar on tha othar fiva CQT scoras (Tablas 4.13 and 4.14). TABLE 4.13 KAV SCORE MEANS OP SELECTED ABILITY POR STUDENTS BT ACNUnHENT LEVEL AND SEE Acbiavatat Laval Padar Noraal Ovar Bag Rdg Stlactad Ability Taata** ISS IScl IT Var Nua CQT-T 24.17* 30.68 26.90 51.59 54.24 142.68 24.69 36.85 (24.72) (30.40) (20.58) (23.88) (44.46) (53.09) (30.35) (127.89) 24.73 53.12 26.56 139.26 23.84 51.29 32.15 34.85 (25.50) (31.96) (21.62) (23.14) (44.75) (54.03) (30.15) (128.91) 38.52 33.31 24.80 24.66 52.15 51.58 142.26 27.49 (27.33) (34.18) (22.10) (22.17) (44.87) (54.33) (29.91) (129.12) *?lrat M t of figuraa for mitt: figuraa In paranthaaaa for faaalaa. **Collaga Qualification Taata (CQT): Infoiaatloa-Soclal Sc lanea (ISS) Inforaatlou-Eatural Sclanca (IScl) Iafomatloa-Total (IT) Varbal (Var) B u t rleal (Nub) COT-Total (CQT-T) MSP Knallah Taat (Ent) MSP Eaadlna Taat (Mg) Scoraa aro rav acoras 72 TABLE 4.14 MULTIVARIATE ANALYSIS OP VARIANCE OP SEX DIPPERENCES ON SELECTED ABILITY TEST SCORES OP STUDENTS P-Ratlo - 58.61* D.P. for hypothasis ■ 8 D. P. for arror - 1080 Varlablas** Eng Rdg ISS IScl IT Var Nun CQT-T Batwaan Maan Squaras 825.98 2.04 2753.01 3481.00 12451.08 225.69 8577.00 35884.76 Unlvarlata P 23.75* 0.04 98.58* 129.04* 148.01* 1.41 107.48* 54.72* D.P. for hypothasls - 1 D.P. for arror* 1087 *Slgnlfleant at tha .05 laval of confldanca or bayond. **Saa Tabla 4,13 for abbraviatIona of varlablaa. Thla aspact of Rypothasls II vaa not rajactad for aithar tha CQT-Varbal taat or tha HSU Raadlna Taat. but vaa rajactad for tha MSU English Taat and rasnlnlng fiva CQT taat scora avaragas according to tha diffarancaa or aiaiilaritlaa notad. (2) Group Diffarancaa: it was axpactad that undarachiavara would aarn significantly lowar taat scora avaragas than tha othar two achlavanant lavala on aalactad taata of acadanlc ability. Rasults of data analysis wara only partially congruant with this hypothasla (Tablaa 4.13 and 4.15). Undarachlaving nalas and fanalas scorad significantly lowar on tha MSU Raadlna Tast and undarachlaving fanalas scorad significantly lowar on tha MSU English Taat than tha othar groups. Ovarachlaving nalas did significantly battar on tha CQT-Munarical taat than tha othar groups; howavar, on tha sans tast undarachlaving nalas unexpectedly scored better then norms1 echlevers. Mo slgnlficsnt differences among echlevement levels were found on the rcsalnlng five CQT test score evereges. TABLE 4.15 MULTIVARIATE ANALYSIS OF VARIANCE OF ACHIEVEMENT LEVEL DIFFERENCES ON SELECTED ABILITY TEST SCORES OF STUDENTS F-Ratio - 3.94* D.F. for hypothesis - 16 D.F. for error ■ 2160 Varlablea** Between Mean Squares Univariate F Eng Rdg ISS IScl IT Ver Mum CQT-T 193.12 411.65 25.59 22.07 17.26 32.08 346.48 221.88 5.55* 7. 72* 0. 92 0.82 0. 21 0.20 3.09* 0.34 D.F. for hypothesis - 2 D.F. for error - 1087 *Slgnlfleant et the 0.5 level of confidence or beyond. **See Teble 4.13 for ebbrevletlons of varlebles. In light of these results, this pert of Hypothesis II wes rejected for females for the MSP English and Reading Tests. and for swles for the MSP Reading Test end the CQT-Numerical test. Hypothesis II was not rejected for either sex for the resmlnlng five CQT test score averages. (3) Interaction Effects (Sex x Achievement Level): were held concerning this area. No expectations Results of the data analysis indicated that no significant differences regarding sex x achievement level interaction effect occurred among the variables (Table 4.16). section of Hypothesis II was not rejected. This 74 TABLE 4.16 MULTIVARIATE ANALYSIS OP VARIANCE OF INTERACTION BETWEEN SEX AND ACHXEVBfEMT LEVEL ON SELECTED ABILITY TEST SCORES OP STUDENTS F-Ratio - 1.47 D.F. for hypothesis - 16 D.F. for error - 2160 Varlablas** Between Mean Squares Eng Rdg ISS IScl IT Ver Hum CQT-T 13.96 20.33 25.52 35.02 12.52 154.24 279.60 396.27 Univariate F 0.40 0.38 0.91 1.30 0.15 0.96 3.50* 0.60 D.F. for hypothesis - 2 D.F. for error - 1087 *Slgnlfleant at .05 l«v«l of confidence or beyond, **See Table 4.13 for abbreviations for variables. Hypothesis III Mo diffarancas existed at tha and of tha first tans of university study asosg tha groups on: (1) First tans average credit hours carried, (2) First tens average credit hours earned, and (3) First term grade point average. A multivariate analysis of variance program (Flan, 1968) was used to carry out the test of the third hypothesis. The .05 level of confidence was chosen to test for statistical significance. The analysis revealed the following about the study sample: (1) Sex Differences: Mo expectations were held regarding sex differences among the three variables analysed In this study. Results of data analysis Indicated that no significant sex differences occurred (Tables 4.17 and 4.18). Therefore, this aspect of Hypothesis III was 75 TABLE 4.17 RAW SCORE MEANS OP SELECTED ABILITY TESTS FOR STUDENTS BY ACHIEVEMENT LEVEL AND SEX Lovol of Achlovonsnt CHC 13.60** (12.98) 13.55 (13.50) 13.36 (13.49) Undor Noras 1 Ovor Varlablas* CHE 10.40 (9.83) 13.39 (13.41) 13.36 (13.47) GPA 1.43 (1.42) 2.55 (2.56) 3.47 (3.38) *Crodlt hours carrtod (CHC); Crodlt hours oarnod (CHE); Grads point avorago (GPA). **First sot of flguros for silos; flguros In paronthosos for fonaloa. TABLE 4.18 MULTIVARIATE ANALYSIS OP VARIANCE OP SEX DIFFERENCES ON SELECTED ABILITY TEST SCORES OP STUDENTS P-Ratio - 0.93 D.P. for hypothasis - 1 D.P. for orror - 1087 Varlablas* CHC CHE GPA Botwoon Moan Squaros Univarlata F 2.98 0.13 1563.64 0.57 0.02 0.65 D.P. for hypothasis ■ 1 D.P. for orror - 1087 *Crodlt hours carrlod (CHC); Crodlt hours oarnod (CHE); Grado point avorago (GPA). 76 (2) Group Differences: It was oxpoctod that significant difforoncoo among achieveottnt lovols would bo found by naturo of tho dooign. Mo hypothooio was made c o n c o m i n g difforoncoo among tho groups for tho avorago crodit hours carriod and oarnod. Rosulta of data analysis showod that underachievers of both soxos did not havo a significantly lower crodlt hours carriod avorago than normal and overachievers, but that undorachiovors did havo a significantly lower crodlt hours oarnod avorago. Undorachiovors of both soxos had a significantly lower grade point avorago than normal achievers w h o , in turn, oarnod a significantly lowor grade point avorago than ovorachiovors (Tables 4.17 and 4.19). Because of those results, this section of Hypothesis III was not rejected for either sox regarding avorago crodit hours carriod, but was roJoetod for both soxos for avorago crodit hours oarnod and grade point avorago. TAILS 4.19 MULTIVARIATE ANALYSIS OF VARIANCE OF ACHIEVEMENT LEVEL DIFFERENCES ON SELECTED ABILITY TEST SCORES OF STUDENTS F-Ratio - 287.35* D.F. for hypothesis - 6 D.F. for orror - 2170 Var lab loo** CMC CHE GFA Between Moan Squares 4.17 739.11 1626371.60 Univariate F 0.79 94.62* 678.33* D.F. for hypothesis - 2 D.F. for orror - 1087 *Signlfleant at tho .05 level of confidence or beyond. **Credit hours carriod (CHC); Crodit hours oarnod (CHE); Grade point avorago (GFA). 77 (3) Interaction Effect (Sex x Achievement Level): was made concerning this factor. Mo hypothesis Mo significant sex x achievement level interaction effect was noted among the variables analysed (Table 4.20). Therefore, this section of Hypothesis was not rejected for either sex. TABLE 4.20 MULTIVARIATE ANALYSIS OP VARIANCE OF INTERACTION BETWEEN SEX AMD ACHIEVEMENT LEVEL ON SELECTED ABILITY TEST SCORES OF STUDENTS F-Ratio - 1.03 D.F. for hypothesis - 6 D.F. for error ■ 2170 Variables* CHC CHE GPA Univariate F Between Mean Squares 6.77 6.43 1636.93 1.29 0.82 0.68 D.F. for hypothesis - 2 D.F. for error ■ 1087 ♦Credit hours carried (CHC); Credit hours earned (CHE); Grade point average (GPA). Hypothesis IV There was no difference in the ratio of males to females within each group (when the groups were not chosen on the basis of sex). Hypothesis IV was tested by the Chi-Square analysis technique. The .05 level of confidence was established to test for statistical s ignlflcance. It was hypothesised from the literature reviewed that more males than fesMles would be identified as underachievers when the groups were not chosen with regard to sex. The results of data analysis were in line with this expectation--*lgnifleant differences did occur. Males 78 w«r« overrepresented in the underechlevers category (Table 4,21). Hypothesis XT was therefore rejected. TABLE 4.21 PKBCEMT OP MALES AMD FINALES BY ACHIEVEMENT LEVEL WHEN LEVELS HEBE NOT CHOSEN ON THE BASIS OF SEX Achievement Level Males Under Normal Over 60 45 46 Females X2 40 55 54 11.228* *Signlfleant at the .05 level of confidence or beyond. This chapter has presented results of statistical analyses of data collected for this study. Results indicated the following for under-, over- and normal achievers: (1) High School Demographic Variables: A few significant dlffar- ences occurred aswng the achievement levels regarding selected demo­ graphic variables and opinions students held about their high schools. Unexpectedly, a larger percentage of male parochial schools were noraal and overachievers than graduates of public schools. As expected, a greater percentage of graduates of private schools were underachievers. AlaK»st three-quarters of female students who regarded their high school preparation for college as "very poor" were classified as underachievers; whereas, the vast smjorlty of femala students who evaluated their high school preparation as "very adequate" were normal and overachievers. Differences also existed among the feswle groups pertaining to the else of the community in which one's high scheol was located, but were difficult to Interpret. town/rural schools. Underachievers came most often from small Normal achievers caam most often from high schools 79 la suburbs of largo cltlos (200,000+ population). This soction of Hypothesis 1 was rajactad for both saxas. (2) High School Curricula Takan: Only thraa ltens of twenty- seven partaining to hinds and nuabar of tarns of coursas takan in high school wars found to ba statistically significant for aach sax. neanlngful pattarn was discerned. No This saction of Hypothasis 1 was not rajactad for aithar sax. (3) Participation in Salactad Extra-Curricular Activities: No significant diffarancas occurred aamag tha groups according to parti­ cipation in various kinds of social science "out-of-class" experiences, and only one Itan for aach sax regarding participation in various kinds of high school clubs. This saction of Hypothesis I was not rajactad for aithar sax. (4) Study Habits: No significant diffarancas occurred for mala students; however, two of nine itana for feswles ware significantly different in favor of overachievers. This saction of Hypothesis I was not rajactad for nalas and was rajactad for fancies. (5) Test-wiseness: No significant diffarancas occurred anong faswla achlevenent levels regarding procedures used in answering quest­ ions on objective exawinatlons. significant for nalas. One Itan of doubtful sManlng was This saction of Hypothesis I was not rajactad for aithar sax. (6) Reasons for Attending Collage: No diffarancas occurred anong aithar faamla or nale groups concerning tha nost or second-sx»st inportent reasons for attending collage. This saction of Hypothesis I was not rajactad for aithar sax. (7) 8aIf-Concept of Ability: Underachievers of both saxas had lowar expectations than nornal and overachievers for earning overall 80 high grades In their first y u r of collsgo. Further, f«ul« underachievers expected significantly lowar gradas in thair major coursas and swle ovarachlavars axpactad significantly highar gradas in social scianca coursas for thair first yaar In collage than tha othar groups. This saction of Hypothesis 1 was rajactad for both saxas. (8) Choice of Collage Major: No significant diffarancas for aithar sax ware found among tha achievement levels whan students who declared a major immediately upon matriculation ware compared to students who did not declare a major at that time. This saction of Hypothesis I was not rajactad for aithar sax. (9) Tasts of Academic Aptitude: (a) Sax Dlffarancas: No significant sax diffarancaa existed on tha Michigan State Universlty Heading Tast and CQT-Verbal tast scoras. Famalas did significantly battar on tha Michigan State University English Tast. Malas scorad high on tha remaining CQT tasts. This aspect of Hypothesis II was rajactad for tha MSN English Tast and tha Collage Qualification Tasts except tha Verbal Tast. and not rajactad for tha MSU Heading Tast. (b) Group Diffarancas: Undarachlaving malas and faawles scorad significantly lowar on tha MSU Heading Tast and undarachlaving famalas scorad significantly lower on tha MSU *"glieh Tast than tha othar groups. Over-achieving males did significantly battar than tha othar levels o n tha CQT-Humarical tast. No significant differences asK>ng achlavement levels ware found on tha remaining CQT tast scora avaragas. This aspect of Hypothesis II was rajactad for feswles for tha MSU English and Heading Teats. and was rajactad for malas for tha MSU Heading Tast and CQT-Humorical tast. Hypothesis 11 was not rajactad for tha MSU English Taat for malas, nor was it rajactad for aithar sax for tha remaining five CQT score avaragas. 81 (c) Interaction Effect (Sex x Achievement Level): No significant differences regarding eex x achievement level interaction effect occurred among the variables analysed; therefore, this section of Hypothesis II vas not rejected. (10) Grade Point Average/Credit Hours: (a) Sex Differences: No significant sex differences occurred asK>ng the variables analysed; therefore, this aspect of Hypothesis II was not rejected. (b) Groups Differences: Underachievers of both sexes earned a significantly lower grade point average than normal achievers who in turn earned a significantly lower grade point average than overachlevers. Further, underachievers of both sexes had a significantly lower average for credit hours earned, although no differences occurred among the groups for average credit hours carried. This part of Hypothesis III was not rejected for either sex regarding average credit hours carried but was rejected for both sexes concerning average credit hours earned and grade point average. (c) Interaction Effect (Sex x Achievement Level): No sex x achievement level Interaction effect was noted among variables analysed; therefore, this section of Hypothasis III was not rejected. (11) When achlevesMnt levels were not identified on the basis of sex, males were significantly overrepresented among underachievers. Hypothesis IV was rejected. CHAPTER FIVE SUMMARY, DISCUSSION AND CONCLUSIONS Problem, Purpose end Procedures It was the purpose of this study to determine if eny differences existed among three groups of students for each sex identified es under-, over- end normal achievers using e model reported in the literature. Differences were to be looked for using mainly non-intell­ ective items; vis., biographical and demographical factors associated with high school characteristics and academic preparation and personal academic behavior patterns. The purpose of this research was not to determine how much the factors included for study contributed to predictive efficiency of standardised intellective test scores. The following questions, restated as four testable hypotheses, were formulated to give direction for analysing data about three groups of students separated by sex and identified as under-, over- or normal achievers efter one term of university study. (1) Are there differences among the groups associated with selected demographic characteristics of high schools from which they graduated? (2) Are there differences associated with high school prepar­ ation among the groups with regard to types and number of terms of courses taken in high school? (3) Are there differences among groups in selected personal acadesiic behavior patterns? (4) Are there differences among groups on selected intellective variables? 82 83 (5) Are there dlfftrtncaa within groups partaining to tha male: few*la ratio whan groups ara not idantifiad on tha basis of sax? Tha review of litaratura was prasantad in thraa parts: (1) A ganaral overview of tha subject of ovar- and underachievement, including definitions, major oparational designs, raaaarch lnadaquaclas, improvement posslbilltlas and bibliographies, (2) A briaf raviaw of factors--msinly parsonality and biographical--not lavastigatad In this study, but which gava a broadar viaw and undarstanding of rasaarch on ovar- and underachievement, and (3) A raviaw of litaratura appropriata to tha varlablas undar study in this rasaarch. In ganaral tarns, undarachlavars ara studants who do not p a r f o m up to axpactatlon, normal achiavars parform acadamlcally as axpactad and ovarachlavars parform battar than axpactad. In othar words, under- and ovarachiavamant ara actually undar- and ovarpradictlon. Thasa tarns can ba oparationally daflnad in savaral ways; howavar, studants idantifiad at ona achlavamant laval by ona method ara not nacassarily so idantifiad by another method. No cosqpreheasiva theory of ovar- and underachlevement was found to exist; howavar, soma directions wars discerned from tha litaratura. (1) Underachievers had a lowar self-concept of ability than normal and ovarachlavars; howavar, away undarachlavars have bean helped to Improve thair self-concept and hence thair academic performance through appropriata counseling focussed on thair self-concept. Further, undarachlavars seamed to under-predlct thair future gradas moreso than normal and ovarachinvars. (2) It seamed that na significant correlation existed between academic and nonacademic (extra-curricular) achievement. 84 (3) In general, underachievers had poorer study habits than normal and ovarachlavars. (4) Tentatively, underachievers seemed to have less definite plans for an academic smjor than normal and ovarachlevers. (3) It seemed normal and ovarachlavars had a higher level of reading ability or performance than underachievers. (6) - (7) Mo pattern of relationship seesmd to exist between over- and underachlavement and the Influence of high school curriculum taken by a student or the slxe of his high school graduating class. (8) It seemed more underachievers cast from parochial high schools than from public high schools. (9) No variables had been found to have universal application to all achievement levels in all types of populations or In any single population. The study population consisted of approximately 7300 new freshmen who swtrlculated In Michigan State University In the fall term, 1968. The study sample comprised 1093 students: 520 suiles and 573 females. The sample did not Include students who met the following criteria: (1) Resided outside the United States of America, (2) Were designated as special part-time students, (3) Carried less than six credit hours of class, (4) Dropped out of Michigan State University during the-fall term, 1968, and (5) Did not have complete test scores and other necessary data. These students completed Form C of the Academic Inventory. an Instrusmnt especially constructed by the Office of evaluation Services of Michigan State University to learn what academic preparation and proficiencies students brought with them to college and to Investigate selected high school deswigraphic character Is tics. 85 Student scores were eveliable for the Michigan State University EnalIsh and Raadlna Tests. and the College Qualification Tests, Form C. First term average credit hours carried, average credit hours earned and grade point average were also nsde available. A regression model selection technique reported in the literature (Farquhar and Fayne, 1964) was used to differentiate the three achieve­ ment levels. Each student was identified as either an under-, normal or overachiever by predicting his first term GPA from his CQT-Total score and then ascertaining whether or not the difference between his actual and predicted GPA was greater than minus one standard error of estiswte (underachiever), greater than plus one standard error of estimate (overachiever) or within plus-and-minus one standard error of estimate (normal achiever) from the least squares regression line. The Chi-Square analysis and a multivariate analysis of variance (Finn, 1968) were the main statistical techniques used to analyse data. The .05 level of confidence was established to determine statistical significance for all hypotheses. Summary of Research Results Four null hypotheses were tested using data collected for this study to determine if significant differences existed asmng under-, over- and normal achievers. In Hypothesis I, selected high school demographic characteristics were investigated. of academic aptitude were analysed. In Hypothesis II, selected tests In Hypothesis III, average credit hours carried, average credit hour s earned and grade point average were analysed. In Hypothesis IF, the male:female ratio within each achieve­ ment level (when levels were not differentiated with regard to sex) was inves t iga ted. 86 A l u a u r y of raiulta of this research are as follows: (1) High School Demographic Variables: A few significant differ* ences occurred asmng the achievement levels regarding selected desmgraphlc variables and opinions students held about their high schools. Unexpectedly, a larger percentage of male parochial high school graduates were normal and overachievers than graduates of public high schools. As expected, a greater percentage of graduates of private schools were underachievers. Almost three-quarters of female students who regarded their high school preparation for college as "very poor*' were classified as underachievers; whereas, the vast majority of faautle students who evaluated their high school preparation as "very adequate" were normal and ovarachlevers. Differences also existed among the female groups pertaining to the else of the community in which one's high school was located, but were difficult to Interpret. Underachievers came most often from small town/rural schools. Normal achievers came most often from high schools in suburbs of large cities (200,000+ population). This section of Hypothesis 1 was rejected for both sexes. (2) High School Curricula Taken" Only three items of twenty- seven pertaining to kinds and number of terms of courses taken in high school were found to be statistically significant for each sex. meaningful pattarn was discerned. Mo This section of Hypothesis I was not rejected for either sex. (3) Participation in Selected Extra-Curricular activities: Mo significant differences occurred asmng the groups according to parti­ cipation in various kinds of social science "out-of-class" experiences, and only one item for each sex regarding participation in various kinds of high school clubs. far alrhav This section of Hypothesis I was not rejected 87 (4) Study bblti: No significant diffarancas occur rad for M i a studants; howavar, two of nlna items for females wara significantly dlffarant In favor of ovarachlavars. This saction of Hypothesis I was not rajactad for wales and was rajactad for females. (5) Test-wlseness: No significant diffarancas occurred asmng fawala achlavawant levels regarding procedures used in answering quest­ ions on objective examinations. significant for smles. Ona ltaw of doubtful smanlng was This saction of Hypothesis I was not rajactad for aithar sax. (6) Reasons for Attending Collage: No diffarancas occurred among aithar fesmle or suila groups concerning tha aK>st or second-most important raasons for attending collage. This saction of Hypothesis I was not rajactad for aithar sax. (7) Saif-Concept of Ability: Undarachlavars of both saxas had lowar expectations than normal and ovarachlavars for earning overall high gradas in thair first year of collega. Further, fasuila under­ achievers axpactad significantly lowar gradas in thair major coursas and mala ovarachlavars axpactad significantly hlghar gradas in social science coursas for thair first year in collage than tha othar groups. This saction of Hypothesis 1 was rajactad for both saxas. (8) Choice of Collage Major: No significant diffarancas for aithar sax wara found among tha achlevesMnt levels whan studants who declared a major Ismmdlately upon matriculation wara compared to studants who did not declare a major at that time. Hypothesis I was not rajactad (9) This saction of for aithar sax. Tasts of Academic Aptitude: (a) Sax Diffarancas: No significant sax diffarancas existed on tha Michigan State University Raadlna Taat and CQT-Verbal tast scoras. Famalas did significantly 88 better on the Michigan State University English Teat. high on the remainlog CQT teats. Melee scored This aspect of Hypothesis II was rejected for the MSU English Test end the College Qualification Tests except the Verbal Test, and not rejected for the MSU Reading Test. (b) Group Differences: Underachieving males and females scored significantly lower on the MSU Reading Test and underachieving feawlas s*ored significantly lower on the MSU English Test than the other groups. (Her-achieving males did slgnlficar.tly better than the other levels O' the CQT-numerical test. No significant differences among achieveamnt levels were found on the remaining CQT test score averages. This aspect of Hypothesis II was rejected for females for the MSU English and Reading Tests. and was rejected for males for the MSU Reading Test and CQT-Nunsrice1 teat. Hypothesis II was not rejected for the MSU Ergliah Test for males, nor was it rejected for either sex for the remaining five COT score averages. (c) Interaction Effect (Sex x Achievement Level): No significant differences regarding sex x achieveamnt level interaction effect occurred among the variables analysed; therefore, this section of Hypothesis II was not rejected. (10) Grade Folnt Average/Credit Hours: (a) Sex Differences: No significant sex differences occurred asmng the variables analysed; therefore, this aspect of Hypothesis II was not rejected. (b) Groups Differences: Underachievers of both sexes earned a algnifleantly lower grade point average than normal achievers who in turn earned a significantly lower grade point average than ovarachlevers. Further, underachievers of both sexes hod a significantly lower average for credit hours earned, although no differences occurred among the groups for average credit hours carried. This part of Hypothesis III was not rejected for either sex regarding average credit hours carried 89 but was rtjscttd for both saxas concerning avaraga cradlt hours aarnad n d grade point average. (c) Intaractlon Effect (Sax x Achievement Lawal): Mo aax x achiawaaant laval Intaractlon affact was notad aaong varlablaa analyaad; therefore, this aaction of Rypothaala III was not rejected. (11) Whan achleveaant lavals wara not idantiflad on tha basis of sax, males wara significantly ovarraprasantad aaong undarachlavars. Hypothesis IV was rajactad. Discussion of Rasaarch Rasults Tha purposa of this saction is to discuss tha aaanlng of salfont findings of data analysis parforasd for this study, educators concerned about helping each underachieving student perfora up to his ability laval should be Interested in evidence gained froa this study which are congruent with tha literature regarding possible reasons causing or factors influencing undarachlavaaant. Counselors, especially, should be aware of "factors Important to acadealc success" in their respective institutions bocause such factors vary froa collage to collage (Cantl, 1959). Further, reasons for a given individual's underachieveaent varies froa student to student (Ratchick, 1953; Barrett, 1957; and Aba, 1966). It saosa appropriate froa tha rasults of this study that one concerned about helping underachievers at Michigan State University improve their academic performance should focus primarily on the following factors congruent with the literature which seeaad to influence student achievement for both sexes in this study sample: (1) Self-concept of ability, Implied by one's expected or pre­ dicted future performance: Underachievers of both sexes had significantly 90 lower expectations than ovtrachitvert and normal achlavara of overall grades for their first year of college. (2) Reading comprehension level: Underachievers of both sexes scored significantly lower than nonsal and overachievers in this area. (3) Declaration of major: No statistically significant difference occurred anong achievement levels regarding students of either sex who declared a major immediately upon entering college and those who made no Immediate preference. Secondarily, one could then focus his attention upon the under­ achieving student's proficiency in English g r o m a r and expression, his study skills and habits, his sophistication in taking objective exam­ inations, his evaluation of high school preparation for college, type of high school (public, parochial or private) from which he graduated, and else of the community in which his high school was located. The reason for evaluating these factors secondarily was that results of the data analysis were either not identical for both sexes, difficult to Inter­ pret or seemingly incongruent with the literature reviewed for this study. At this point, it does not seem that one would gain much Insight into an underachieving student's problem by attempting to determine what kinds of courses he took in high school, his amount of participation in high school extracurricular activities, his reasons for attending college or the kinds of demographic characteristics about the student’s high school found not to be statistically significant in this study. This conclusion is congruent with the literature reviewed for this study. The first three hypotheses considered students differentiated into three achievement levels after separation by sex to determine if sex differences would occur. Since the two groups were similar in else at 91 the start, ths achlevsamnt groups were also about tha same slaa (Tabla A.l). However, to tast Hypothesis IT, students wara not saparatad by sax whan placad into achlavaacnt lavals. Rasults of tasting tha hypo- thasis indicated that amles wara significantly overrepresented in tha underachiever category by a ratio of three smles to two females (Table A.21). One lnportant Implication of this finding is that in whole numbers, counselors may be working with more males than females regard­ ing academic underachlevesmnt. Finally it should be realised that no statistically significant differences occurred among achlevesmnt levels within each sex on the average CQT-Total score. This indicated that underachieving students in the study population did not perform poorly because they did not have the basic ability to succeed in college at tha atart. Stated another way, tests of acadsmlc ability indicated the underachieving students considered in this study had the ability to be academically successful at Michigan State University. They were unsuccessful for a complexity of reasons soma of which were researched in this study. Limitations of the Study Several limitations of this study were noted. (1) There was no assurance that students comprising the study population were representative of succeeding etudent populations at Michigan State University or similar institutions, although it was assumed that findings of tha study based on this population sty be useful for generating hypotheses about similar populations. (2) The first term college GFA used as the criterion in identifying each of the achievesmnt groups was a heterogeneous criterion. The measure was used because it was the standard by which the university retained or dropped its students. 92 (3) Factors found not to bo significant for this study aanplo night bo for onothor and vlco versa. (4) Tho prodlctor, CQT-Total scoro, soonod to favor smles. This 1initation was basod upon nalo/fonalo conparatlvo standings on CQTTotal scoro for all froshnon who ontorod Michigan Stato University in fall, 1968, pub1ishod in a report by tho Office of Evaluation Services (1968). Fosuile scores wore about eleven points lower than smles at given percentile intervals. (5) The predictor also accounted for only about 19.4X of the variance in GFA for nales and 25X for females in the study sample. However, this was not too different from nose similar cognitive predictors. (6) Hlgh-abillty students had relatively little room to demonstrate overachievement because of the GFA "ceiling" and low-ability students had relatively little room to demonstrate underachievement because of the GFA "floor" (Cf. Duff and Siegel, 1960). Suggestions for Future Research Firstly, it seems appropriate for this research to be extended to analysis of Identical and similar items contained in Fart I of the other five forme of the 1968 Academic Inventory to determine if results would be congruent with findings of this study and/or to enable one to better understand results of this study difficult to interpret. Secondly, it is strongly recnwniled that this research also be extended to analysis of two Instrwants constructed from the six forms of the Academic Inventory--the Exam and Study Skill Survey and the Academic Questionnaire. Both instruments were administered to all students entering Michigan State University in the fall term, 1969. 93 The study i M p l t list would be larger than that of this study which would allow a graatsr nunbsr of studant subject to be analysed. More rssponsas to items ltesm per subject existed on of a given the 1969 lnstruswnts which would allow more aspects of the subject to be explored. It is possible that some variables not found significantly different in this study night be upon analysis of items in the 1969 forme; a.e., test-wlseness and reasons for attending college. Thirdly, in light of theliterature reviewed for and the findings of this study, it is suggested that further research of non-intellective predictors of academic success be performed concerning the entering college student's own expectation or prediction of future academic performance. Fourthly, it is urged that research be done into the kinds of successful and unsuccessful programs colleges and universities have used or are using to immediately identify and help potential under­ achievers enrolled la their institutions. Presentation of a swdel program of such endeavors should be a logical result of such research (Cf. Boxak, 1969; Bodnar and Weinberg, 1970). BIBLZOGKAm BIBLI0GBAFHT Aha, C. , MA Factor Analytic Study of S m n Non-Intellectual Indices of Academic A c k i t r a m a t , H Journal of Educational Meaner amant. 3:31-44, 1946. Ahmana, j. S., Smith, W. L. and Clock, M. D. , "Predicting Academic Succeaa In Callage hy Moana of a Study Hahlta and Attltudo Inventory," Educational and Favchologleal Maaauroamnt. 18: 853-7, 1958. Altua, W. D., "A Col logo Ach Invar and Non-achiever Sea la for tha Mlnnaaota Multlphaalc Faraonallty Invantory," Journal of Aoollad Farehoioar. 32: 385-97, 1948. Anaataal, A., Maada, M. J. and Schnaldara, A. A., **Tha Validation of a Biographical Invaatory aa a Pradlctor af Collaga Succaaa,** CEEB Baaaarch Managraph No. 1, Now Tork; CEEB, 1980. Andaraon, K . , ad., Baaaarch on tha Academically Talantad Student. Waahlngton, D. C . : National Education Aaaoclatlon, 1981, 922 pp. Applaton, J. , "A Caaq>arlaon of Freehmea Caanutar Studant a on Salactad Charactarlatlca, Experlencea, and Chaagea," Uapuhllahad Ph. D. thaala, Michigan Stata tFuiverelty, 1985, pp. 40-1. Aahcraft, M. 6., MAn Analyala of tha Effact of tha Nigh School Curriculum Upon Collaga Achlavamant.M Vnpuhllahad Ed. D. thaala, Nam Haxlco Stata tfnlvaralty, 1988. Baird, L. L . , "The Achlavamanta of Bright and Avaraga Studanta,** Educational and FavchoIonleal Maaaaramant. 28:891-99, 1988. Baird, L. L . , **Blg School, Small School: A Critical Examination of tha Hypothaala,** Journal of Educational Favcholonv. 80:253-80, 1989. Baird, L. L . , "Prediction of Accempllahmaat in Collaga: A Study of Achlavamant,** Journal of Caunaallnn Paychalonv. 18:248-33, 1989. Baird, L. I,, **Tha Vndacldad Studant--How Dlffarant la Ha?** and Guidance Journal. 47:429-34, 1989. Faraonnal Barratt, H. O . , **An Zntanalva Study of 32 Glftad Children,** Faraonnal and Guidance Journal. 38:192-4, 1957. 94 95 Baymur, F. B. and Patterson, C. 1,, "A Comparison of Three Methods of Assisting Voderachieving llgh School Students,” Journal of COMaaallaa Psychology. 7:83*90, 1980. Bodnar, R. L. aad Weinberg, S. L., "Ingradionta of Succoaaful Treatment Program for Vnderachlevers," Journal of Coanaaliaa Psychology. 17: 1-7, 1970. Bonaott, G., Sanaa tt, M. G., Wallace, W. L. aad V t a a u , A. G. , Co liana Qualification Taata Manual. Row Tork: Tho Psychological Corporation, 1957. Bargar, I. L. aad Sutkor, A. R., "Tha Ralationahlp of Emotional Adjustment aad Into1lactual Capacity of Academic Achievement of Collaga Students," Manta 1 Hynlaaa. 15:85-77, 1958. Blgga, D. A. aad Tlaalay, D. J . , "Studaat-Mada Academic Predictions," Journal of Educational Baaaarch, 83:195-7, 1970. Bladar, B. M . , "Relationships Aaong Saif-Expectations, Salf-Ceucapt, and Acadaalc Achievement," Vnpubllshod Ph. D. Thesis, Valuerslty of Wiscoasla, 1985. BoarsiM, P. J., aad Wahlstrem, M . , "Tha Influence of Teat-tflseaass Vpoa Achievement," paper proseatad at 1988 American Educational Research Association annual convention. Borislaw, g . , "Self-Evaluation and Academic Achievement," Journal of Counsalian Psvcholonv. 9:248-58, 1982. Bouman, P. I., "Personality and Scholastic Bader-Achlavement," in Free inn Capacity to L o a m . Washington, D. C. : Association for Supervision and Curriculum Development, 1980, pp. 40-55. Bosak, I. M . , "Suesser Project for Oaderachleviag Freshmen," Improving College and University Teaching. 17:208-11, 1989. Bristow, W. H. , Low Achievement: A Memorandum and Bibliography. Hew Tork: Board of Education of City of Hew Tork, 1959. Brookevar, W. V. and Associates, "Self-Concept of Ability and School Achievement," East Lansing, Michigan, Office of Research and Publications of Michigan State Valuersity, Final Report of Cooperative Research Project Ho. 845, 1982. Burgess, S., "Personality Factors of Over- and Vnder-Achievers in Engineering," Journal of Educational Psychology. 47:89-99, 1958. Butcher, D. 8., "A Study of the Relationship of Student Self-Ceucept to Academic Achievement in Six High Achieving Elementary Schools," Vnpublished Ed. V. thesis, Michigan State Vniverslty, 1987. Butsow, J. W. and Williams, C. M . , "College Freshmen Achievement of Parochial and Public Secondary School Graduates," Journal of Educational Research. 80:215-7, 1987. 96 . Caapbell, 9 T. «nd SUil«]r, J. C. , I x p i r i i n U l and Q m u 1-E k p w l n a U 1 Boa inns for M > — rch. ChicAgo: land McRal ly 6 Co., 1969 , 84 pp. Carey, H. B . , "The Bright Bad«rAchi*v«r la Reading," in Robinson, 1. A., od,, Tho Underselllover in B o a ding. ChicAgo: University of Chicago Press, Supplemental BducAtloaAl Monograph 92, December 1962, pp. 70-86 Conti, P., "highest and Lowest Ranking Students At tho School of IducAtion At a Largo, Urban University," Poroonnol and Culdanco journal, 37:457-9, 1959. Clark, J., "Partial CorrolatIona," Technical Report Ro. 16, Computer Inatitute for Social Science Reeearch, Michigan State University, March 31, 1968. Clark, J. and Gllljaore, G . , "Data Plot aad Statlatica for Variables and PaIra of Variables: DAP1, !*F2 and RAP22," Technical Report 15, Ceaiputer Inatitute for Social Science Reaearch, Michigan State University, February 7, 1968, 15 pp. Clark, J. I . , "Grade Achievement of Female College Studenta in Relation to Ron-Intellective Factora: M1P1 Xtema," Journal of Social Psychology. 37:275-81, 1953. Cenba, A. W . , "A Perceptual View of the Adequate Peraeaality," in Forceivlay. Behavina. Becoming. Waahlngton, B. C. : Aaaociatlon for Supervlalon and Curriculum Bevelopmeat, Rational Bducation Aaaociatlon, 1962 Yearbook, pp. 50-64. Conbs, C. P., "Perception of Self and Schelaetlc Vnderachlavement in the Academically Capable," Peraennel and Guidance Journal. 43:47-51, 1964. Coemba, R. R. and Ravlea, P., "SoIf-Conception and the Relatlonehip Between U g h School and College Scholaatlc Achleveme nt," Sociology and Social Reaearch, 50:460-9, 1966. Bn Sena, P. A., "The Rffectlveneae of Two Study Rablt Inventorlea in Predicting Conalatent Over-, Bnder-, and Rormel Achievement," Journal of Counaelina Pavcholoav. 11:388-95, 1964. Be Sena, P. A., "Problema of Conalatent Over-, Vnder-, and Hormal Achieving College Studenta aa Identified by the Mooney Problem Check Liat," Journal of Rducatlonal Reaearch. 59:351-5, 1966. Biener, C. L., "Similarity and Blfferencea Between Overachieving and Underachieving Students," Personnel and Guidance Journal. 38: 396-400, 1960. Beleys, I. J. and Rasaglla, G. A., "Accuracy of Student Prediction of College Grades," Personae 1 and Guidance Journal. 41:528-30, 1968. Baw d, R. J,, "Vaderachlevlag Students of High Capacity," Journal of ■Inher Bducation. 23:327-38, 1952. 97 Dr«a|ow, J. , ^ W a r a e h & t v l a i / 1 Journal of Counseling Psychology. 4:210-1, 1957. On lala, P. and lackact, Eralnntion of Qyor aad O W C No. Hoar 814(14). Haehington University, Conference on Research E. P., «41cora, Tho Haaauraoant aud gnJarachlavanant. Technical Report Ho. 8, St. Louie: Department of Pajckology, April 1943, 111 pp. (Proceedings of a Methodology in Training, 14 April 1941.) Duff, O. L. and Siegel, S., "Biographical Factors Associated With Academic Oyer and Vnderach lavement," Journal of Educe t loam 1 Psychology. 51:43-4, 1940. Dulles, R. J . , "The Myth of Oaderachieveemat," Education Digest. 27(4): 18-19, 1942. Ebel, R. L . , Measuring Educational Achievement. ■all, 1945. Mew Jersey: Prentice- Edwards, A. L . , Experimental Design in Psychological Research. Tork: Rinehart 4 C o . , 1948, Third edition. Mew Engle, K. B . , Davis, D. A. and Baser, 6. E., "Interpersonal Effects on Vnderachlevers," Journal of Educational Research, 41:208-10, 1948. Eseklel, M. and Fox, K. A., Methods of Correlation and Regress ion Analysis. Hew Tork: Wiley 4 Seas, 1959. Parquhar, W. W. and Payne, D. A., "A Classification and Comparison of Techniques Used in Selecting Vnder- and Over-Achievers," Personnel and Guidance Journal. 42:874-84, 1944. Parquhar, W. W. and Payne, D. A . , Chuhh, W. and Thorpe, M. D . , Motivation and Academic Performance," Selected Bibliography to 1945, Michigan State Vniverslty College of Education mimeograph. Feldman, K. A. and Mewconb, T. M . , The Impact of Colleges on Students. San Francisco: Jessey-Bass, Inc., 1949, p. 285-4. Feldt, L. S ., 'The Vse of Extreme Croups to Test for the Preaence of a Relationship," Psychemstrika, 24:307-14, 1941. Fessender, B. I., "A Case Study of Factors Contributing to the Academic Deficiency of Selected Michigan State College Students on Probation," Vnpubliahed Ph. D. thesis, Michigan State College, 1953. Fink, M. B . , "Self-Concept as it Relates to Academic VaderachlevesMat," California Journal of Educational Research. 13:57-42, 1942. Finn, J. D., "Valvarlate and Multivariate Analysis of Variance and Covariance: A FVRTRAM IV Program," Version 4, Department of Educational Psychology, State University of Mew Tork at Buffalo, June, 1948. 98 rithtr, J. L. , "FacCcrt A£f«ctla| Academic S a c c M t , " Vnpub 1 lshed Ph. D. t h M i a , l o r t h M f t t n Viivtrtity, 1964. F l t u i h M r , ft. L. tad Rock, D. A., "A Multiple Moderator Approach to tho Identification of Over- and Vnderachlovara," Journal af Kdncatlonal Measurement. 4:223-8, 1949. Flieglar, L. A., **Vndara tending tha Vnderach lav lag Gifted Child,** Psychologies1 ftaporter. 3:533-4, 1957. Foreman, F. S., "Study of Self-Reinforcement and Study Skllla Programs with Bright Collaga V n d e r a c h l a v a r a V n p u b l l a h a d Ed. V. thaala, Vnivaraity of Nebraaka, 1949. Frankal, E ., "A Cooperative Study of Achieving and Vnderachlaving High School Boya of High Intellectual Ability," Journal of Educational ftaaaarch. 53:172-180, 1940. Fradarlckaan, V. and H a l v 11la, S. 0., "Differential Predictability In tha Vaa af Teat Scoraa," Educational and Faychological Measurement. 14: 447-54, 1954. Galar, E. L . , "Studant Self-Estimate of Final Courae Grades," Journal of Genetic Favcholoav. 98:43-7, 1941. Gallagher, J. J. and Rogge, W . , "The Gifted," Review of Educational Reaearch, 34:37-55, 1944. Gallant, T. P . , "Academic Achlavamant of Collaga Freohmen and Ita Relationship to Selected Aapacta of the Student's Background," Vnpubllahad Ed. D. thesis, Western Reserve Vnlverslty, 1945. Garrett, H. F . , "A Review end Interpretation of Investigations of Factors Related to Scholastic Succeaa In Colleges of Arts and Science and Teachers Colleges," Journal of Enperlmental Education, 18:91-138, 1949. Gebhart, C. G. aad Hoyt, D. P., "Personality Heeds of Vnder- and OverAchieving FreshaMu," Journal of Applied Psychology. 42:125-8, 1958. Gerberich, J. R. , "Factors Related to the College Achlavamant of HlghAptltude Students Who Fall," Journal of Educational Psychology. 32:253-45, 1941. Goldherg, M. L . , Research on Talented. Hew Tork: Bureau of Publications, Teachers College, Celuubla Vnlverslty, 1945, 72 pp. Oeldberg, M. I#., "Studies In Vndorachiovanant Among Academically Gifted," In Freeing Capacity te Learn. Washington, D. C . : Association for Supervision and Curriculum Development, 1940, pp. 54-73. 99 Geldburgh, 3. J. u 4 Pttiay, J. T., "A Hoc* on C o n M O l l a f Vo4«rachlovin| Collot* Scodooto," Jaurnol »f C o w o IIm P o T c h o l o p . 9:133*8, 1942, Ooldaoo, L . , fotoi Tooto la Counseling. Crafts, 1941, pp. 321-42. Haw Turk: Appletea-Century- Oaaaa, J. C. , A n Annotated Bibliography aa tka Acadoalcally Talented. Washington, D. C . : Rational Education Association, 1941. Oaaaa, J. C. , " D y a u l c o af tka Haderacklaveaaat af Gifted Students," Exceptional Children. 24:98-101, 1957. Oaaaa, J. C. , "Factors af Achleyeasat la ll|h Sckaol and Collaga,** Jouraal of Counseling Psychology. 7:91-5, 1940. Oaaaa, J. C. aad Sckalkal, t. W . , "Tka Xnproveasat af Reading la Giftad Cklldraa," Educational Adaiaiatratioa aad Saparvlalaa. January 1940, pp. 35-40. Graff, 0. H. aad loasaa, J. C . , "Relationship of Q4XS Scorao to Collaga Achleyeasat aad Adjustaaat," Jaaraal of Callaaa Studaat Faraoaaal. 11:129-34, 1970. Guilford, J. P . , Fundanaato1 Statlotlca la Psychology aad Bducatloa. Raw Tork: McGraw-Hill, 1945, 405 pp. Hackatt, g. T . , "A Bibliography of Oyar- and Oadarachlavaaaat la tka Classrooa," la DuHola, P. H. and Hackatt, E. ▼., adltora, Tka M a — ura— at aad gyaluatloa of Qyar aad Undorachioyoaaat. St. Loula: Washington University, April 1941. Halpara, H. M . , "Can Psychotherapy Halp tka Underachiever?" In Halparn, H. M. , A Parent'jb Guida to Cfcild Psychotherapy. Raw Tork: Barnaa, 1943, pp. 44-81. Harris, B . , "Factors Affactlag Collaga Gradas: A Baylaw of tha Lltaratura: 1930-37." Psychological Bullatla. 37:125-44, 1940. Hill, H. E . , Jr., "Scholastic Succass of Collaga Freshaea froa Parochial aad Pukllc Secondary Schools," School Raylaw. 49:40-4, 1941. Hlltoa, T. L. aad Myars, A. E . , "Parsoaal Background, Exparlaaca and School Achleveaeat: A a Xayastlgatloa of tha Contribution of Quastioaaalra Rata to Acadealc Prediction," Jouraal of Educational Maaaurasaat. 4:49-80, 1947. Holland, J. L. and Rickards, J. M . , Jr., "Acadealc aad Hoaacadanlc Acceaplishaent la a Raprasaatatiya Saaple Takaa froa a Population of 412,000," lawa City: Aaerlcaa Collaga Tasting Rasaarck Reports Ho. 12, May, 1944, 17 pp. Roltwasn, V, H. aad Brawn, V. F . , "Study Habits aad Attitudes la tka Prediction of Acadsale Success," Asaricaa Psychologict. 8:349-72, 1953. 100 Hood, A. B. and Svanaoa, E. O. , "A Look at Student Achlavaaaat In Dlfftraat Typo* of Collagaa,H Faraoaaal and Guidance Jo u r n o 1 . 44:282-5, 1945. Hoyt, D. P., "Sis* of High School ond Collogo Grade*," Por*onnol ond Guidance J o u r n a l . 37:569-73, 1959. Hoyt, D. P., and Munday, L. A., Tour College Freshmen. Iona City: American College Testing Prograaa, Inc. , 1969, 249 pp. lu*eel, R. and Sprinthall, H. , "Uuderachlevement Related to Interest*, Attitudes and Value*.*' Per*onnel and Guidance Journal. 44:388-95, 1965. Jackson, R. M . , "In Support of the Concept of Vnderachievenent," Personnel and Guidance Journal, 47:56-62, 1968. Jones, J. G. and Grieneeks, L . , "Measures of Self-Perception as Predictors of Scholastic AchieYeaant," Journal of Educational Re s e a r c h . 63:201-3, 1970. Juola, A. E . , "Freshman-Level Ability Tests and Long-Range Prediction,** Paper presented to Rational Council on Measurement in Education annual aaeeting, February 1963. Juola, A. E . , "Prediction of Successive Terms Performance in College From Tests and Grades," Amarlean Educational Research Jou r n a l . 3:191-7, 1966. Juola, A. E . , "The Predictive Validities of the Rev Orientation Test Battery for Diverse Academic A r e a s ,** Michigan State Vnlverslty Office of Evaluation Services mimeograph, June 1959. Juola, A. E . , "Variations in Exam Sophistication Among College Freshaaen," Paper presented to American Educational Research Association annual aaeeting, February 1969. Keefer, K. E . , "Self-Fredletion of Academic Achievement by College Students," Unpublished Ed. D. thesis. University of Tennessee, 1965. Khan, S. B . , "Affective Correlate* of Academic Achievement," Journal of Educational Psychology. 60:216-21, 1969. Kirk, B . , "Test Versus Academic Performance in Malfunctioning Students," Journal of Consulting Psychology. 16:213-6, 1952. Kisch, J. M . , "A Comparative Study of Patterns of Underachieving Male College Students," Unpublished Ph. D. thesis, University of Michigan, 1967. Klein, S. P., Rock, D. A. and Evans, F. R . , "The Use of Multiple Moderators in Academic Prediction," Journal of Educational M e a s u r e m ent. 5:151-60, 1968. Kornrlch, M . , e d . , Undorachievement. 1965. Springfield, Illinois: Tlismss, 101 Kovlts, 6. T. , MAn Aul j i i a of Underact*ltvnaat" la Kornrlch, M . , ed., Uaderachievomeat. Springfield, Illinois: Th o m s , 1945. Kovlts, 0. T. sod Armstrong, C. M . , "Vnderach lavement: Concept or Artifact?** School aad Society. 89:347-9, 1941. Krathvehl, V. C., “Specificity of Over aad Under Achievement in College Courses,** Journal of Aoallod Psychology. 34:103-4, 1952. Krug, R. K., “Oyer- and Underachievement aad the Edwards Personal Preference Schedule," Journal of Applied Psychology. 43:133-4, 1959. Kruabols, J. D., 'Measuring Achleyesmnt Motivation, A Key lew,** Jouraal of Counsellag Psychology. 4:191-8, 1957. Layla, D. E . , The Prediction of Acadewlc Performance: A Theoretical Analysis and Keylew of the Literature. New Tork: Russell Sage Foundation, 1945, 182 pp. Lesgold, A. N . , Zerhy, R. and Fester, G . , "Analysis of Contingency Tables (ACT II),H Cowputer Institute for Social Science Research, Technical Report No. 14.1, Michigan State Vniyerslty, July, 1949. Lucas, L. A., "Relationships Between Self-Concept, Selected Scholastic Variables, and the GPA*s of Nigh School Students,** Unpublished Ed. D. thesis, Vniyerslty of Southern California, 1948. Lise, M. K . , "A Comparison of Under- and Oyer -Ach ley lag Femele College Students," Journal of Educational Psychology. 51:109-14, 1940. Luaaaberg, P. W. and Luaneborg, C. E., "The Differential Prediction of College Grades from Biographic Information," Educational and Psychological Measurement. 24:917-25, 1944. Malloy, J. P., "An Inyestlgatloa of Scholastic Oyer- and UnderAch lavement As m g e Female College Preshswn," Journal of Counseling Psychology. 1:240-3, 1954. Malloy, J. P., "The Prediction of College Achievement With the Life Experience Inventory," Educational and Psychological Measurssmnt. 15:170-80, 1955. Malloy, J. P., and Ivaaoff, J. M . , "Further Use of the Life Experience Inventory la Predicting College Achievement," Journal of Educational Resaarch. 57:522-5, 1944. May hew, L. B., "Non-Test Predictors of Academic Achievement," Educational and Fovchological Measurement. 25:39-44, 1945. McQuary, J. P., "Seme Differences Between Over- and Vnder-Achievers in College," Educational Administration and Sunaryision. 40:117-20, 1954. McQuary, J. P., and Truax, W. E., "An Under-Achievement Scale," Journal of Educational Research. 48:393-409, 1955. 102 Merrill, R. M. Murphy, D. T . , "Parson*11ty Factors and Acadealc Achleveaent in College," Journal of Csunasling Psychology. 5:20711, 1959. Mlddloton, 6* and Guthorlo, 0. M . , "Porsonallty Syndroaes and Acadoaic Vnderachleveaent," Journal of Educational Psychology. 50:66-9, 1959. Millar, L. M . , ad., Ouidanca for tho Indorsehloyor with Superior Ability, Washington, D. C . , Vnlted Statos Office of Education Builotin OE25921, Mo. 25, 1961. Morgan, H. H., "A Poychoaatrlc Coaparison of Achieving and Monachlaylng Collaga Students of High Ability,** Journal of Consulting Psychology. 16:292-8, 1952. Motto, J. J., **A Kopiy to Drasgow on Vnderachlevers," Journal of Counao1Ing Psychology. 6:265-7, 1959. Myors, R. C., "Blographlcal Factors and Acadoaic Achioyoaont: An Experiawatal Inyostigatlon," Educational and Psychological Moaauroaant. 12:615-26, 1952. Myors, R. C. and Scbults, D. G . , "Predicting Acadoaic Achlawsnaat With a Mow Attltude-Intaraat Quastlonnalre," Educational and Psycho­ logical Measureaant. 10:656-63, 1950. Maugoboron, B., "Clinical Study of Acadoaic Vndorachlayers," in Wedge, B. M . , ad., Payshasocial Probloas of Collego Mon. Mow Rayon: Tala Vniyerslty Prass, 1958, pp. 63-96. Offlea of Evaluation Servlcos, "Conparatlva Standings of tho Various Collage end Curriculiai Groups on tho Orlontation Wook Exaalnatlons for tho Fall of 1968," alaoographod roport, Michigan State Vnlvarslty, 1968, p. 20. Ovens, W. A. and Johnson, W. C., "Sena Maaeurad Personality Traits of Cellaglata Vnderachlayeaant," Journal of Educational Psychology. 60:61-6, 1969. Passev, A. R. and Goldberg, M. L., "Study of Vnderachleving Gifted," Educational Leadership. 16:121-5, 1958. Pavek, F. L . , "The Relationship of Mea-Acadaalc High School Variables to College Achlsvsawnt and Participation in Selected ExtraCurricular Activities," Vnpubllahad Ed. B. thesis, Vnlverslty of Rorth Dakota, 1968. Pearlaan, S., "An Investigation of the PrObleas of Acadealc VnderAchlevaaent Aaong Intellectually Superior College Students," Vapubllehed Ph. V. thesis, Row Tork Vnlverslty, 1952. Pederson, J. 0., "An Exploratory Study of Ronintolloctive Variables and Related Deviation froa Predicted Acadealc Perforaance," Republished Ph. D. thesis, Purdue Vnlverslty, 1968. 103 P«t«r«, D. M . , "The Self-Concept m a Factor in Over- end Underachieveeent," Unpublished Ed. D. thesis, Indiana University, 1968. Peterson, J . , "The Researcher and the Underachiever: Never the Twain Shall Meet," Phi Delta Happen. 44:379-81, 1963. Phillips, E. L . , "A Note on the Use of the Term Over-Achleveaent in Guidance and Personnel Work," Journal of Educational Psychology. 34:300-6, 1943. Pierce, J. V., "The Bright Achiever and Underachiever: A Comparison," in Bereday, C. Z. #. and Lauwerys, J. A., eds., The Gifted Child. New York: The Yearbook of Education, Harcourt, Brace and World, 1962, pp. 143-54. Plppert, R. and Archer, N. S., "A Comparison of Two Methods of Classifying Underachievers with Respect to Selected Criteria," Personnel and Guidance Journal. 41:788-91, 1963. Powell, W. J. and Jourand, S. M . , "Some Objective Evidence of iMsaturity in Underachieving College Students," Journal of Counseling Psychology. 10:276-82, 1963. Raph, J. B . , Goldberg, M. L. and Passow, A. H . , Bright Underachievers: Studies of Scholastic Underachievers Among Intellectually Superlor High School Students. New York: Teachers College Press, 1966, 289 pp Ratchick, I., "Achievement and Capacity: A Comparative Study of Pupils with Low Achleveswnt and High I.Q. 's with Pupils of High Achievement and High I.Q.*e," Unpublished Ed. D. thesis, New York University, 1953. Reed, C. E ., "A Study of Three Groups of College Preparatory Studenta Who Differ in Relative Achievement," Unpublished Ed. D. thesis, University of Houston, 1955. Richardson, W. T . , "Nonacademic Influences on Academic Achievement," annotated bibllogrephy prepared for College Student Personnel Institute, April 1965. Roth, R. M. and Meyersberg, H. A., "The Non-achievement Syndrome," Personnel and Guidance Journal. 41:535-40, 1963. Ruble, W. L. and Fafter, M. E . , "Calculation of Least Squares," Agricultural Experlamnt Station, STAT Series No. 7, Michigan State University, November 1969. Rust, R. M. and Ryan, F. J . , "The Strong Vocational Interest Blank and College Achievement," Journal of A p p 1led Psychology. 36:341-5, 1954. Ryan, F. J . , "Personality Differences Between Under and Overachievers in College," Unpublished Ph. D. thesis, Columbia University, 1951. Saunders, D. S., 'Moderator Variables in Prediction," Educational and Psychological Measuresmnt. 16:209-22, 1956. 104 Shuey, A. M. , "Academic Success of Public end Private School Students in Randolph-Macon Women's College: 1. The Freshman Year," Journal of Educational Research. 49:481-92, 1956. Schultz, D. 6., and Green, B. F . , "Predicting Academic Achievement with a Hew Attltude-Xnterest Questionnaire 11," Educational and Psychological Measurement. 13:54-64, 1953. Shaw, M. C., "Definition and Identification of Academic Underachievers," in Miller, L. M . , ed., Guidance for the Underachlever with Superior Ability, Washington, D. C . : United States Office of Education Bulletin OE25021, No. 25, 1961. Shaw, M. C., "Underachlavement Is Predominantly a Male Problem," Personnel and Guidance Journal. 44:1091-2, 1966. Shaw, M. C. and Alves, G. J . , "The Self-Concept of Bright Academic Underachievers: Continued," Personnel and Guidance Journal. 42:401-3, 1963. Shaw, M. C. and Brown, D. J . , "Scholastic Underachievement of Bright College Students," Personnel and Guidance Journal. 36:195-9, 1957. Shaw, M. C. and Grubb, J . , "Hostility and Able High School Underachievers, Journal of Counseling Psychology. 5:263-6, 1958. Shaw, M. C. and McCuen, J. T . , "The Onset of Academic Underachievement in Bright Children," Journal of Educational Psychology. 51:103-8, 1960. Siegel, L . , "A Biographical Inventory for Students: Validation of the Instrument," Journal of Applied Psychology. 40:122-6, 1956. Smith, L . , "Significant Differences Between Hlgh-Ablllty Achieving and Monachleving College Freshmen as Revealed by Interview Data," Journal of Educational Research. 59:10-12, 1965. Standrldge, C. G . , "The Predictive Value of Nonintellectual Factors and Their Influence on Academic Achievement," Unpublished Ed. D. thesis, University of Arkansas, 1968. Staton, J. T . , "The Relationship of Selected Factors to Academic Success for Beginning Freshmen," Unpublished Ed. D. thesis, University of Oklahoam, 1962. Stebens, L. D., "A Study of the Relationship Between Reading Skills and Academic Achievement in Specific Subject Matter Areas," Unpublished Ed. D. thesis, Oklahosw State University, 1967. Stern, G. "Environsmnts for Learning," In Sanford, N . , e d . , The American College. New York: Wiley and Sons, 1962, pp. 690-2. Stoner, W. G . , "Factors Related to the Under-achievement of High School Students," Unpublished Ed. D. thesis, Stanford University, 1956. 105 Stroup, A. L., "Tho ProdletIon of Academic Porforotnco Froo Personality and Aptitude Variables," Journal of ExperAjsontal Education, 38:83-6, 1970. Sutton, R. S., "An Analysis of Factors Rolatod to Education Achievesmnt," Journal of Gonotic Psychology. 98:193-201, 1961. Taylor, R. 6., "Personality Traits and Discrepant Achievesmnt: A Review," Journal of Counseling Psychology. 11:76-82, 1964. Taylor, R. 6., "Tutorial Services and Acadesilc Success," Journal of Educational Research. 62:195-7, 1969. Taylor, R. G. and Parquhar, W. U . , "The Validity and Reliability of the Hunan Trait Inventory Designed to Measure Under- and OverAchievesmnt," Journal of Educational Research. 5:227-230, 1966. Thiel, L. and Patrick, L., "PERCOUMT," Technical Report Mo. 18, Cosiputer Institute for Social Science Research, Michigan State University, May 6, 1968. Thorndike, R. L . , The Concepts of Over- and Underachlevesmnt. York: Teachers College Press, 1963, 79 pp. Mew Todd, F. J . , Terrell, G. and Frank, C. E . , "Differences Between Normal and Underachievers of Superior Ability," Journal of Applied Psychology. 46:183-90, 1962. Travers, R. M . , "Significant Research on the Prediction of Academic Success," in Donahue, V . , ed., The Measurement of Student A d ­ justment and Achievement. Ann Arbor: University of Michigan Press, 1949, pp. 147-90. Tyler, L., 'Minimum Change Therapy," Personnel and Guidance Journal. 38:475-9, 1960. Walker, H. M. and Lev, J., Statistical Inference. Holt and Company, 1953, pp. 106-8. Mew York: Henry Walsh, A. M . , Self-Concepts of Bright Boys with Learning Difficultles. Mew York: Teachers College Press, 1956. Wedaamyer, C. A., "Gifted Achievers and Mon-Achievers," Journal of Higher Education. 24:25-30, 1953. Welts, H. and Wilkinson, H. J., "The Relationship Between Certain MonIntellective Factors and Academic Success In College," Journal of Counseling Psychology. 4:54-60, 1957. Walllngham, W. W . , "The Interpretation of Relative Achievement," American Educational Research Journal. 1:101-12, 1964. Wellington, C. B. and Wellington, J., The Underachlever: Challenges and Guidelines, Chicago: Rand McMally and Company, cl963, 122 pp. 106 Willie— , R. L. and Cola, S., "Self-Concept and School AdjustMnt/' Personnel and Guidanca Journal. 46:478-81, 1968. Willie— on, S. G., How to Counaal Studanta. Book Co., 1939, pp. 361-86. Raw Tork: McGraw-Hill Wolf, S. J., "Hlatorlc Background of tha Study of Peraonallty aa it Ralataa to Succaaa or Pallura In Academic Achieve— nt," Journal of Canatic Psychology. 19:417-36, 1938. Wrann, C. G. and Husri>er, W. J., "Study Habita Aaaociatad with High and Low ScholarahIp," Journal of Educational Psychology. 32:611-6, 1941. Wylie, R. c . , Tha Salf-Concapt--A Critical Survay of Partinant Raaaarch Lltaratura. Lincoln: University of Nebraska Press, 1961. Young, P. C., "Collaga Frashwan Judge Their Own Scholastic Pronise," Personnel and Guidanca Journal. 32:399-403, 1954. Young, R. W., "The Relationship of Business and Industrial Courses in High School with Acad— ic Achievansnt in tha Firest Tear of Collaga," Unpublished Ed. D. thesis, University of Michigan, 1966. APPENDIX A PART I OP ACAPBilC INVENTORY. FORM C c MICHIGAN STATE UNIVERSITY Academic Inventory Stati' University iu n t in u a I Iy at tempts tn tleve ]o p eduiat Iona I programs t h a t will b e s t .. . the needs of Its students. In order to do this el fee lively It is necessary to know as much as passible about the academic b a c kgrounds ol our enter i n g students. This Inventory Is designed to produce some information in this area. Mu-liin an The inventory consists of two parts. Part have 1 is conce r n e d with your high school experiences and Part II is c o n c e r n e d with what you learned as a c o n sequence of these and other educational experiences. Do not 1. begin working until you have read, On gagji of the answer sheets, print the d a t e . and yonr student n u m b e r . understood and carried out in the the directions below. appropriate places your name (LAST NAME FIRST), 2 . lPn less your S TUDENT NUMBER IS C O R R E C T L Y M A H K E D in pencil in the six rows of spaces under the words STU D E N T NUMBER, your answer sheet C A N N O T BE PROCESSED. First, W R ITE your student number in the vertical c olumn at blank boxes under the heavy arrow. Then MARK ONE SPACE in EACH of the SIX ROWS of ten spaces that c o r responds to each number of your student number. Be sure that (a) you have ONE MAR K in each ol the six rows, . Approxlaately what percentage of students in yiiur graduating class definitely plan in go to college this coning year? I \a ^ than 25'f J . 2&*; Id 4i*» to 744; i, 4. 7 W to 89% 5, 90% or ovtr - 2- 11. Approximately whan