fl H '1le!!!an JHWIMU'HWHWNW/l! J \ ll (DN-h ‘s‘HE PREDICTIELX‘E‘Y AN?) FAL'TGRED MIMIENSEGNS' (333 THE MJSCIALES F033;. ELEVE? TH (IRAQ—*3 PAREDCHEAE. SEIHSGL 3‘5 {.113in Thesis fix the fiegrw 9? Ph. D. MJCHWAN SEAT}? UNNERSETY Mark K. KEpi‘mua—Ner 39:63 0-169 NIVERSITY LIBRARIES IIIIIIII III III III‘IIII IIIIII I 31 1 This is to certify that the thesis entitled THE PREDICTIBILITY AND FACTORED DIMENSIONS OF THE M-SCALES FOR ELEVENTH GRADE PAROCHIAL SCHOOL STUDENTS presented by Mark K. Kipfmueller has been accepted towards fulfillment of the requirements for Ph D degree in Mn Major professor I '/ .’ wit'myl/ Henri” Date August 2, 1963 Michigan State THE PREDICTIBILITY [IND FACTORED DIMENSIONS OF THE l\'I--SC/IT.I".S FOR ELEVENTH GRADE PZIROCHl/‘IL SCHOOL STUDENTS Mark K. Kipfmueller [H ABSTRACT OF A THESIS Subnuttvd to trim Collogzé of Educatlon, NIlCIllC/Idzl Stat“ University ofilgrlcultt._1r0 and Applif‘d Sciwnch m partial fulfillment of frac- I*<‘qLIIrOIIIr‘;ItS for tho dvgrl‘wxé of DOCTOR OE Pl I ILOSOPHY 1963 AN ABSTRACT OF A THESIS MARK K. KIPFMUELLER This study was concerned with (l) the assessment of an experi— mental objective measure of motivation, the Michigan M-Scales, as a predictor of academic achievement for parochial school male and female eleventh grade students, and (2) a factor analysis of the male responses to the Generalized Situational Choice Inventory, a sub-scale of the M— Scales, to determine the underlying motivational structure of the paroch— ial school student. The results of the factor analysis were then com- pared logically with a previous analysis based on a public school eleventh grade sample. An aptitude test score (Differential Aptitude Test - Verbal Reason- ing), grade point average, and scores on the four sub—tests and total M— Scales were obtained for each GIFTVOnth grade student from two Class B Michigan parochial schools in the Lansing and Jackson area. The mean scores for the male. and female samples were compared with the mean scores from the public school sample studied by Farquhar. A regression analysis was conducted for both male and female sample. to estimate Jrade point average. when the M-Scales were added to the DAT-VR. Male re— sponses to 22 selected items and 4S valid items of the GSCI were factor analyzed by the principal axis solution; the normal varimax method of ro— tation enabled an interpretation of the factors. -11.. - iii — MARK K. KIPFMUELLER In this study, significant mean differences were found in GPA and DAT-VR test scores in favor of the parochial school students. The paroch— ial school male sample had higher mean scores in total score, GSCI and Word Rating List sub~tests of the M—Scales. Public school males had significantly higher mean scores on the Human Trait Inventory. The "t" test indicated significant differences in all but the Preferred Iob Character— istics Scale mean scores. The parochial school female had significantly higher mean scores in the WRL. No significant differences were found for the other scales. The regression analysis indicated that the precision of estimation of GPA was significantly improved when the GSCI, the WRL, and total M- Scales were added to the DAT—VR for both males and females. The HTI sub—scale also improved the precsion of estimation for females. The PICS did not improve the precision of estimation for either males or females. The factor analysis of the male responses to selected items of the GSCI indicated that although two factors were held in common by the pub— lic and parochial school samples, one factor was apparently unique to the parochial school student's motivational structure. The factors held in common were (1) EU.I£I‘;1C.-T\ZCF.S.U.S-9.0FITDQILQPSQQDLLSTJEQQL and (2) L’ZZLQLIEQEQ versits-rgljoiltjggmflypgement; The factor unique to the parochial school sample was QuL2L.v_Ci:_Su,S_iIli1€t-ditggiedimss. The public school male GSCI factor, competition with versus ease of meeting a standard, did not _. iv ._ appear as an identifiable factor for the parochial school students of this study. However, neither the unique nor the exclusion finding were given much weight because they most likely were traceable to the differences in rotational procedures . It was concluded that the M-Scales have predictive validity for estimated grade point average of the parochial students of the sample. TI—IE PREDICTIBILITY AND FACTORED DIMENSIONS OF THE M-SCALES FOR ELEVENTH GRADE PAROCHIAL SCHOOL STUDENTS Mark K. Kipfmueller A THESIS Submitted to the College of Education, Michigan State University of Agriculture and Applied Science in partial fulfillment of the, requirements for the degree of DOCTOR OF PHILOSOPHY 1963 «:25th 6%‘n'xlb‘f £19K. \' OWL}? DQEIEENLS To Dr. William W. Farquhar, chairman of my doctoral committee and thesis director, whose tireless efforts made the completion of this study possible. To the other members of my doctoral committee, Dr. Bill L. Kell, Dr. Walter F. Johnson, and Dr. John X. Jamricrr, for their interest and suggestions. To Dr. Van C. Johnson, Mr. Donald Kiel, Mr. John Gordon, Mr. William Darnell, and Mr. Arthur Ballas for technical and statistical as- sistamce. To my wife, Marilyn, for her patience and understanding. _ii- _'l_‘_A_BLE or commits ACKNOWLEDGEMENTS. LIST OI‘ TABLES. ..... CIIAPTER I. THE PROBLEM. . Purpose of the Study. . . ...... Importance of the Problem. . . ...... Theory. . . . . . . . . The Hypotheses . . . . . . ........ Organization of the Study. . . ..... 11. REVIEW OF LITERATURE ON APTITUDE, ACHIEVE- MENT AND MOTIVATION OF THE PAROCHIAL SCHOOL STUDENT. Aptitude Measures. Achievement Measures. Discussion of Studies. Summary. III. DESIGN AND METHODOLOGY. Sample Selection. Nature of the Data. Reliability. Analysis Procedures. Factor Analysis. Null Hypotheses. Summary. IV. MEAN DIFFERENCES AND REGRESSION ANALYSIS. . Mean Difference Between Parochial and Public School Student Scores. Correlational Analysis. . Regression and Multj p19 Correlation Analysis. Regression Analysis. . . . Multiple Correlation. Summary. Page ii CDNmC/OOU L0 10 11 15 17 18 18 19 21 22 23 24 24 26 3O .31 32 35 35 _iv._ _TABLE OF CONTENTS (Continued) ‘7‘- Page CHAPTER v. RESULTS or THE FACTOR ANALYSES. . . . . . . . 38 Results of the Twenty—Two Variable Factor Analysis Similar to the Farquhar Procedure. . . 38 Rotation of Factors with Altered Criterion - Twentv— Two Items. . . . . . . . . 41 Factoring of all Forty-Five VaIicl Items ofthe GSCI. . . . . . . . . . . . . . . . . 45 Comparison of Factors. . . . . . . . . . . . 47 Summary. . . . . . . . . . . . . . . . . . . 55 VI . SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS. . 57 Summary. . . . . . . . . . . . . . . . . . . 57 Conclusions. . . . . . . . . . . . . . . . . 60 Recommendations............... 63 APPENDICES............................ 64 BIBLIOGRAPHY........................... 67 TA BL E L.I_S_T_QF_ 113mg: SUMMARY OF A POLAR THEORY OF HIGH AND LOW ACADEMIC ACHIEVEMENT MOTIVATION. . SUMMARY OF HOYT'S ANALYSIS OF VARIANCE RELIABILITY ESTIMATES FOR THE M—SCALES FOR PAROCHIAL SCHOOL SAMPLE. MEANS, STANDARD DEVIATIONS AND t TESTS FOR PUBLIC AND PAROCHIAL SCHOOL MALES. . MEANS, STANDARD DEVIATIONS AND t TESTS FOR PUBLIC AND PAROCHIAL SCHOOL FEMALES. INTERCORRELATIONS AMONG ACHIVEMENT, APTITUDE AND M-SCALES SCORES FOR PAROCHIAL SCHOOL MALES. . INTERCORRELATIONS AMONG ACHIEVEMENT, APTITUDE, AND M—SCALES SCORES FOR PAROCHIAL SCHOOL FEMALES. . TESTS OF SIGNIFICANCE OF INCREASE IN PRE- CISION OF ADDING EACH OF THE SUB—TEST M—SCALES AND TOTAL M—SCALES TO THE DIF- FERENTIAL APTITUDE TEST—VERBAL REASONING TO ESTIMATE GRADE POINT AVERAGE — MALES ANDFEMALES................... ITEM LOADINGS FOR THE GSCI MALE RE— SPONSES - Factor I. ..... ITEM LOADINGS FOR THE GSCI MALE RE- SPONSES — Factor II. . . . ITEM LOADINGS FOR THE GSCI MALE RE— SPONSES - Factor I. ..... Page 21 28 3O 31 34 39 4O 42 TA BL I; \J .- 1V“ LIST OF TABLES (Continued) ITEM LOADINGS FOR THE GSCI MALE RE- SPONSES - Factor II. . ITEM LOADINGS FOR THE (LL—SCI MALE RE— SPONSES — Factor III. . ITEM LOADINGS FOR THE GSCI MALE RE- SPONSES — Facwr IV. ITEM LOADINGS FOR THE GSCI'MALE RE— SPONSES - Factor 1. ITEM LOADINGS FOR THE GSCI MALE RE- SPONSES — Factor 11. SUMMARY OF LABELS FOR THE FACTOR ANALYSES OF THE MALE RESPONSES TO THE GSCI FROM THE FARQUHAR AND THE PAROCHIAL SCHOOL STUDY. Page 43 43 44 46 46 48 £2 UAETEBLL mi; PROBLEM Researchers have made. great advances in the. assessment of achieve— ment and aptitude. There is agreement that academic motivation facilitates use of ability, but the. exact nature of facilitation is unclear. To more com— pletely understand academic achievement the need exists to understand aca— demic motivation. Attempts have been made by researchers to predict academic achieve- ment in terms of intellectual factors. The stueies by Scanneli, Garrett, Juola, Klugh and Bierly are Characteristic of this attempt. Recent emphasis has been an attempt to relate non-intellectual factors to academic achievement, such as studies by Chahbazi, S Scannell, Dale P. , "Prediction of College Success from Elemen— tary and Secondary Perforn'iance, " Journal of EducationalPsyghologv, 51, 1960, pp. 130—135. Garrett, Wiley S. , “Prediction of Academic Success in A School of Nursing," 13:15.92BCL3.Y}.C1Q}N_QQNCOIQLLILQQIl 38, 1960, pp. 500-503. Juola, Arvo E. , "Predictive Validity of Five College Level Aca- demic Aptitude Tests at One Institution, " _Pe_rsoniiti_ag§l_§giclance Journal, 38, 1960, pp. 637—641 . Klugh, Henry E. and Bierly, Robert, The School and College Abil— ity Test and High School Grades as Predictors of College Achievement," ECNISLQHQDQLQBQ-Eél/LQEQLQ92991 l\4e'iasurem<'7nts, 19, 1959, pp. 625—626. Chahbazi, Parzi, ”Use of Projective Tests in Predicting College ‘ College. Achievement, " Educational and Psyghjionical*Measurements, 20, 1960, pp. 839—842. _ 2 _ , 6 W, 7 . . . . Hac..,f_ Eir-f*.i:r3.t;i__G:1€l C 111.91% 1190. 1. Slam: - _C1__:1‘t_S_, Cooperative. Research Project No. 846 supported by the U. S. Office of Education, in cooperation with Michigan State University, Final Report, 1963. _ 3 _ There is a need to study parochial school students to determine whether or not the M-Scales is applicable to this population and evaluate factors that seem to be a part of their motivational pattern. Purpose of the Study: The purpose of the study is to investigate the differential predic- tions and factored dimensions of an objective measure of academic moti- vation, the Michigan M-Scales, when applied to a sample of parochial male and female high school students. Not only was the study designed to measure the predictive efficiency of the M-Scales when applied to pa— rochial school students, but also to study the underlying motivational structure by use of factor analysis. This seems particularly germane to this study because Farquhar and his associates have completed such an analysis of all the M-Scales for the public school sample. Importance of the Problem: According to Monsignor Frederick G. Hochwalt,10 Director of the National Catholic Welfare Conference, Washington, D. C. , the parochial school enrollment in 1961 consisted of 14 per cent of all children enrolled in elementary schools and nine and a half per cent of all high school stu— dents. In 2,376 secondary schools, there were 937,671 students; in 10,631 elementary schools, there were 4,445,288 students enrolled. 10 Hochwalt, Frederick G. , "Catholic Education U.S.A. - 1963, " The Catholic Schocijournal, 63, No. 1, January, 1963, pp. 12-14. _ 4 _ The image of the Catholic as a low achiever in intellectual activi— ties is apparently changing. Low social class membership, ethnicity, and recent immigration have characterized American Catholics in many . .. ll 7 y . . . . , , achievement studies. Certain of these SOCio—economic variables C10 not appear to be relevant to the emerging picture of third and fourth gener- ation Catholics in suburban America. The academic achievement of Catholic students and their general achievement orientation, reflected in 1 . . - . 1 12 . tieir High School Inventory responses of Denny s stud»: do not parallel the low achievement picture of Catholic students in research studies such 1 T ' l as tne hew Haven studies. . . . . l4 . 3. . The findings of Veroff, Curin, and Feld inciicate that for Catholics n-Ach scores are negatively related to income and positively related to ll Rosen, Bernard C., "Race, EthniCity, and the Achievement Angerigazjgsflggglggical Review, XXIV, (February, 1959), pp. S y n d r o m e , 4 7 - 6 U. 12 , . . . . . Denny, Terry, "Acnievements of Catholic Students in Public High Schools — II," The Cgtlgliquducational Review, LX, October, 1962, pp. 442—469. McClelland, D., Baldwin, A., Bronfenbreziner, U., and Strodtbeck, F., Talent‘and Society, Princeton, New Jersey: D. Van Nostrand Company, 19 58. 14 Veroff, Joseph; Gurin, Gerald; Feld, Sheila; "Achievement Motivation and Religious Background, " AmericanSociological Review, April, 1962, pp. 205—217. - 5 _. . . , . ,. . . ,. 15 family Size. Tnese findings were Opposite to wnat McCleiiand, et. al., 16 . . t . and Rosen have found in preVious researcn studies. This study was conducted to provide additional knowledge relative to the academic achievement, aptitude and motivation of the parochial school student. With the rapid growth in enrollment of the parochial school in the past ten years it becomes increasingly important to under— stand the achievement and motivation underlying achievement of the pa— rochial school student. T569111: 17 , , , .. . . McClelland theorized that tne need for acnievement motivation may be indicated by the subject's responses to certain TAT cards. Inter- pretation is made on the basis of the subject's concern for long—term in- volvement, unique accomplishment, and competition with a standard of excellence. l8 . .1 , ._ , . Parquhar polarized McCleiiand s theory by adding the constructs of short-term involvement, and common accomplishment. The third con— struct, competition with a standard of excellence, Parquhar adapted to in— clude maximal and minimal competition. These constructs were viewed 15 McClelland, op. ci_t_._ 16 , Rosen, op. Cit. McClelland, D., Atkinson, J., Clark, R. and Lowell, 13., The Achievement Motive, New York: Appleton—Century—Crofts, 1953, _.. - -- pp. 110—113 18Farquhar, op.cit. _ 6 _ as being related specifically to the academic setting. Earquhar's theory assumes that the first three factors theorized by McClelland characterize the individual with the high n-Achievement, while the three factors added by Farguhar characterize the student who possesses the low n-Achievement. The hypothesized constructs of academic motivation are summarized in Table 1 . 1. TABLE 1.1 SUMMARY OF A POLAR THEORY OF HIGH AND LOW ACA- DEMIC ACHIEVEMENT MOTIVATION. sm— A‘aags: if- m e 1”: —:_ -10111=7ia~+ ii:- ‘1? “Motivation . fi I “wMotivati-opnw 1. Need for Long—Term 1. Need for Short-Term __..._ Twelve—“meat...” - ._._ 13101:»! {went 2 . Need for Unique 2 . Need for Common _--__£\..999mp.1i.sihmerit‘ ._ A909 mpliSZ'UiiQnt..- --_ 3. Need to Compete with a 3. Need to Compete with a Maximal Standard of Minimal Standard of Excelleegeh Excellence From this theory, Parquhar and his associates developed the M- Scales, an objective measure of academic motivation which they hypoth— esized would increase the precision in predicting high school grade point average when added to an aptitude measure. The M-Scales is comprised of four sub—tests: (1) The Generalized Situational Choice Inventory which measures the student's need for academic achievement; (2) The Preferred Job Characteristics Scale which measures the occupational aspirations of the subject; (3) The Word Rating List which measures the student's self- concept in an academic setting; and (4) The Human Trait Inventory which _ 7 _ measures unique characteristics of high and low motivation students in an academic situation. An attempt is made. in this study not only to study the differential predictiveness of the M—Scales, but to examine the motivational structure of the parochial school student. Tie Hypotheses: The hypotheses to be tested in this study are: 1. There is a difference in the mean. scores on the four sub- tests of the M—Scales between parochial and public school males and females. 2. There, is a difference in mean scores on the total M~Scales between parochial and public school males and females. 3. There is a difference in mean grade point average between parochial and public school males and females. Research by Parquhar and his associates indicated that the M- Scales, when added to an aptitude measure, increases prediction of grade point average. The following hypothesis, therefore, will be tested in this study: 4. The M-Scales, when combined with an aptitude measure, will increase the. precision for predicting academic achieve— ment (GPA) for parochial school males and females. Previous research has indicated that the Generalized[Situational Choice Inventory is most closely aligned to the initial theory upon which -8- 1 the M-Scales are based. Also, the GSCI seems more applicabie to males than females. If sex is not taken as a soparate variable, the theory seems best reflected in male under— and over-achievement. Therefore, a factor analysis of male responses to the GSCI was used in the. study. The fol— lowing hypothesis will be tested: 5. The factor analysis of the male parochial school student's responses to the GSCI will yield an interpretable structure different than that found for the public school male. OYQEJEEE‘ILQH of the Study: In Chapter II the review of literature concerning achievement, ap- titude, and motivation of the parochial school student will be presented. The design and methodology used in the study are described in Chapter III. In Chapter IV the statistical analysis of the mean differences in scores and the Regression analysis will be presented. The results of the Factor anal- ysis is found in Chapter V. CHAPTER II REyJ§_W_Qf_*LIT[RAflIRFL on APTITUDE, ACHIEVEMENT AND MOTIVATION or THE PAROCHIAL SCHOOL STUDENT Few studies of the parochial school students' achievement in rela- tion to academic ability and motivation were found in reviewing the litera— ture. The majority of the studies tended to be descriptive rather than statistical in nature. Furthermore, many investigations did not separate parochial from private schools in their analysis procedures. In the re— view which follows attention is given to descriptive as well as actuarial reports. In addition, relevant studies focusing on private schools which include parochial populations will be reviewed as well as the obviously pertinent church-based reports. The Reverend C. Albert Koobl, Associate Secretary, Secondary School Department, The National Catholic Educational Association, in re- sponse to a request for information regarding studies of achievement, abil- ity and motivation conducted in the parochial schools replied that he did not have much to offer and that their office has not been able to do the kind of research that they would like to have done. The Carnegie Corporation has made available $350,000. for an ob- jective and factual study of Catholic elementary and secondary education. 1 Reverend C. Albert Koob, Personal Communication, National Catholic Educational Association, Washington, D. C. -9- L 10 _. Dr. William H. Conley, Notre Dame University, director of this study, outlined the following two purposes for the study: (1) to provide a pro- file of Catholic education, making available information about Catholic schools; and (2) to conduct investigations in depth in selected dioceses which will provide more detailed information to make judgments, decisions, and plans for the future of Catholic schools.2 The results of the. study, however, will not be available for at least another year. Aptitude Measures: 3 . . . Koos and Hill4 found parochial school students to be superior to public school students in scholastic aptitude as measured by the ACE Psy- chological Examination. Buegel5 found no significant difference in school ability as measured by SCAT scores in a comparison of high school juniors Conley, William H. , "The Study of Catholic Education, " Nae Koos, Leonard V. , Private and Public Secondary Education, Chicago: University of Chicago Press, 1931. ' ‘ Hill, Robert E., Jr. , "Scholastic Success of College Freshmen from Parochial and Public Secondary Schools, " School Review, 69, No. 1, Spring, 1961, pp. 60-66. Buegel, Hermann F. , "Comparison of SCAT Scores of High School Juniors in Parochial and Public Schools," Psychological Remrts, 7, 1960, pp. 497—498. _ 11 _ in. parochial and public scnools. A .JLlllCtln of the College Entrance Ex— amination Board states that young people who prepared to take the College Entrance Board Examinations through public high school instruction scored, on the whole, a little better than seniors from private college preparatory institutions. Dams and Frederiksen stated fiat on tne average, public school students did better academic work during the freshman year at Princeton in relation to ability as measured by the Scholastic Aptitude Test - Verbal Section - than did private school graduates. Achievement M ea sures: F) Lathrop and Kieffero studied 53 private school graduates and 51’) pub— lic school graduates enrolled at Iowa State College in 1952 to determine if the public and private Ellgi‘. schools in Iowa do equally well in preparing stu— dents for work at Iowa State College. The subjects were paired by sex, divisions of the college as well as previous high school course pattern, and as closely as possible on characteristics of aptitude as measured by total score on ACE Psychological Examination, high school grade point average, and high school size. They found that the difference in achievement -7.._'—.—.~--« _. —..._.-__ ' 6 Dyer, Henry S. and King, Richard C., "College Board Scores: Their Use and Interpretation. No. 2, " PriHCetun, New Jersey: College Entrance Examination Board, 1955, pp. 151-159. Davis, Iunius A. and Frederiksen, I‘~J:’nman, “Public and Private School Graduates in College, " Journal of Teacher Edugation, 6, March, 1955, pp. 18—22. , Lathrop, Irvin T. and Kieffer, Thomas 1., "College Achievement of Public Versus Private High School Graduates, " The Clearing House, Vol. 33, January, 1959, pp. 299-3-‘12. _ 12 _ between graduates of two types of high schools was not statistically sig— nificant at the end of the first or third quarter. However, the difference in achievement was significant at the time of leaving Iowa State College. It appears that the student who is a graduate of a public high school in Iowa enjoys a greater measure of academic success at Iowa State College than does a graduate of a private high school in Iowa. No evidence is available to explain why public high school graduates achieve higher grades. Conant9 said that the public high school graduate does better than the private school graduate in Harvard College though there may be consid- erable difficulty the first year. He did not support his statement with a presentation of empirical evidence. Seltzer10 compared the numbers of Harvard fresjimen on the Dean's List and the numbers of failing and passing students and found that the stu- dents admitted from public schools presented considerably better freshman- year academic records than the private school students in spite of the fact that there was little or no difference in intellectual potential or in academic background between the two groups. Conant, James Bryant, Education and Liberty, Cambridge, Massachusetts: Harvard University, 1953, p. 131. 10 Seltzer, Carl C. , "Academic Success in College of Public and Private School Students, " Journal of Psychology 25, 1948, pp. 419—431. A I _ 13 _ Wiggin11 analyzed the top 100 and bottom 100 men in the classes of 1954 and 1955 at Yale University and concluded that it is more likely that the individual public-school graduate will achieve a good academic record than a private-school graduate. The records at the Phi Beta Kappa Headquarters in Washington, D. C., indicated that in 1954 more public-school than private-school stu— dents were initiated into Phi Beta Kappa in three leading universities (Fiarvard, Yale University, and Colgate University). 13' 14 found that public school graduates earned significantly Shuey higher grades, on the average, than did the private school graduates when she compared the academic achievement during the freshman and sophomore college year of 189 pairs for the freshman year and 245 pairs of females for the sophomore year. The public and private high school graduates were 11 Wiggin, Lewis M., "The Yale Undergraduate: One Man's Findings," Yale Alumni Magazine, May, 1954, pp. 7—9. National Education Association Research Bulletin, Vol. XXXV, No. 4, December, 1957, p.148. 13 Shuey, Audrey M. , "Academic Success of Public and Private School Students at Randolph-Macon Woman's College. I. The Freshman Year,” Journal of Educational Research, 49, 1956, pp. 481—492. 14 Shuey, Audrey M. , "Academic Success of Public and Private School Students at Randolph—Macon Woman's College. 11. The Sophomore Year," Journal of Educationaeresearch, 52, 1958, pp. 35—38. _. 14 _ paired by age, intelligence, academic load, same geographical background, and same community size. She concludes that there is no evidence that private schools prepare pupils more effectively for college than do public schools. The most comprehensive study conducted on how public and private . . . . 1 school graduates compare in academic acnievement was conducted by Koos. His findings generally favored the graduates of public schools, a result re- markable since intelligence measures favored private schools. 16 . . lilll compared high school scholastic success of students from parochial and public elementary school backgrounds. Although his find— ings favored the public school, only the findings for Natural Science, Written Expression, and composite measures were significant at the .05 level. , , 17 . In another study, Hill compared the scholastic success of fresh- men at Ball State University from parochial and public schools. He found the parochial school sample to be statistically superior in scholastic apti— tude as measured by the ACE Psychological Examination. However, when 15 Leonard V. Koos, op. cit. 16 Hill, Robert E. , Jr., "An Investigation of the Educational De— velopment of Selected Iowa Secondary School Pupils from Varied Elemen— tary School Environment, " 14th Yearbook, National Council on Measure- ments Used in Education, New York: National Council on Measurements Used in Education, 1957. Hill, Robert E. , Jr. "Scholastic Success of College Freshmen from Parochial and Public Secondary Schools," School Review, 69, No. 1, Spring, 1961, pp. 60—66. _ 15 _ scholastic aptitude was controlled, the public school group at the end of the first and third term as well as for cumulative grade point average at the end of the third term was superior in scholastic achievement at the .05 level. I Denny studied the achievement of Catholic students in two IVIidwestern communities where there were no Catholic high schools. He found the general picture of academic achievement of Catholic students is one of equal achievement when compared with the non—Catholic contrast group's achievement. The academic achievement of Catholic students and their general achievement orientation, reflected in their High School Inven- tory (HSSI) responses, do not parallel the low achievement picture of Catholic students reported in research studies such as the New Haven Studies . 20 Denny suggests that any discussion of the achievement of Catholics should consider the socio-economic parameters of the Catholic and contrast group. Di scus_sion of Studies: The majority of the studies reviewed in this chapter were descrip— tive rather than statistical. Many of the studies were found lacking in 18 Denny, Terry, "Achievements of Catholic Students in Public High Schools - I. , " The Catholic Educational Review, LX, March, 1962, pp. 145-162. Denny, Terry, "Achievements of Catholic Students in Public High Schools - II. , " The Catholic Educational Review, LX, October, 1962, pp. 442-469. Strodtbeck, Fred L. , "Family Interaction, Values, and Achieve— ment," Talent and Socieiy, ed. , David C. McClelland, Princeton, New Jersey: D. Van Nostrand Company, 1958. hypothesis formulation, and control of external variables such as socio— economic factors. A number of the studies did not clearly specify the population from which the samples were drawn. Parochial schools were not separated from the private school classification in the discussion of findings related to private schools. In some of the studies no empirical evidence was presented upon which the conclusions were based; the conclusions appeared to be lOgi'— cal rather than empirical. Theoretical formulations a priori and a pos- teriori were lacking. Inadequate sampling would necessarily limit the population to which generalizations could be made. The evidence is not clear that the parochial school student achieves less than his public school counterpart. In fact, the findings are erratic, to say the least. This factor could probably be traced to such extraneous variables as the uncontrolled socio—economic variable where the selection effect in certain communities or the cost of education would tend to contam— inate the data. Furthermore, the studies which lump parochial and private schools together introduce inseparable effects because of the image of abil- ity and behavior problems encountered in the private school setting. Also, regional differences could make large differences in who does or can take advantage of a parochial school education. In the final analysis, it does not appear that there is any conclusive study regarding parochial school students' achievement, aptitude. and mo- tivation which accounts for the flaws in experimentation mentioned above. The need is paramount for such a study. _ 1'7 ._ Although there has been some research on Catholic adults' intellec— tual and cultural achievement, little is known about the achievement of Catholic youth until they leave the secondary school. The majority of the studies which concerned the parochial school student's achievement and ability tend to be inconclusive and contaminated with uncontrolled error. The Study of Catholic Education hopefully will provide an objec- tive and factual study of Catholic elementary and secondary education. In designing the study which follows particular attention was given to incorporating those features found lacking in the major share of the pre- vious studies completed. Particularly an attempt was made to (1) state clearly the population from which the sample was drawn and (2) use valid statistical procedures in testing stated hypotheses. CHAPTER III DESIGN AND METHODOLOGY The design of this study is described under three headings: (1) Sample Selection, (2) Nature of the Data, and (3) Analysis Procedures. Sample Selection: Two samples were involved in the analysis of this study. A pa- rochial school sample, comprised of 100 males and 100 females, was ran-— domly selected from a population of 280 eleventh grade students from two Michigan parochial high schools. These schools represent students from 15 cities and communities, and 20 parishes. The public school sample of 254 males and 261 females came from . 1 . an earlier study conducted by Farquhar. This sample was selected from 4200 eleventh grade students from nine Michigan high schools and ana— lyzed for the Farquhar project. The parochial school sample used for the factor analysis was the same sample that was used in the correlational analysis. The public school sample for this analysis was also a part of the Farquhar study.2 -..._._——- _—'.- M.-__.—... 1 William W. Farquhar, A_Compfirehensi_v‘e_ $9.9.qu thgjyfotiya: $391341- f,eetetellfisrirlzine Aehisy <2 me at. 9f . E 1-0515111: b .0 1: ed (2 fitshfi 911521 Students, Research Project No. 846 (8458) supported by the U. S. Office Of Education, in cooperation with Michigan State University, Final Report, 1963. William W. Farquhar, op. ciL ‘-18- -19_ Nature of the Data: Three types of data were gathered on each student: (1) an apti- tude estimate, (2) a grade point average, and (3) a motivational score. Aptitude Measure: The aptitude measure was the Verbal Reason— ing sub-test of the Differential Aptitude Tests (DAT-VR) . 3 Grade Point Average: The grade point average (GPA) was com- puted for each individual using grades received in ninth and tenth grade academic subjects, i.e. , subjects requiring homework. To obtain an estimate of the GPA criterion reliability a correlation was calculated be— tween ninth and tenth grade GPAs and was found to be .87 for males and .96 for girls. Motivational Score; The motivational score was obtained from the student's responses to the Michigan M—Scales. This scale. is com- prised of four sub—tests: (1) Generalized Situational Choice Inventory, (2) Preferred Job Characteristics Scale, (3) Word Rating List, and (4) Hu- man Traits Inventory. The Generalized Situational Choice Inven- tory was constructed to describe the aca- demic motivation situation. Students are required to make a forced choice between two types of situations, one which depicts a high and one which depicts a low academic A description of the DAT—VR may be found in the manual: Tilt?) Psychological Corporation, Megualjor tb£‘-_D_i._f_fi‘£§llllj§_1 Aptitude Tests, New York, 1959. -20- motivation situation. A high score on this scale indicates an individual who has a high need for academic achievement. A low score indicates an individual who chooses activities disassociated from the school's program. The Preferred Job Characteristics Scale was designed to differentiate between the job aspirations of high-low motivated students. Students are required to make a forced choice between two types of jobs, one which depicts high and one which depicts low job aspira— tion. Students who score high on the scale tend to want jobs where their individuality is recognized, where their talents and skills are used, and where opportunities exist for change and advancement. Students who score low on this scale tend to prefer jobs where entry is easy, requirements are low, time restrictions are absent, and where one is ”discovered" rather than worked into a position. he Word Rating List was designed to mea- sure the academic self-concept of the stu— dent. Students are asked to rate them- selves on a series of descriptive phrases and words describing high and low motivated and achieving students. A high score in— dicates an individual with academic self- concept oriented towards the school envir— onment. An individual who scores low on this scale has a fairly clear-cut picture that he is not academic. The Human Trait Inventory consists of items from past personality tests that have been found to differentiate between high and low achieving and motivated students. Stu— dents are asked to rate how they feel about these statements. Individuals who score high on this scale tend to have personali- ties more similar to highly motivated stu- dents and those scoring low tend to have -21- personalities similar to low motivated stu- dents. Begging A Hoyt's analysis of variance estimate for the total M-Scales was obtained for a sample of 100 parochial school males and a sample of 100 parochial school females. The reliability estimate for the male popula— tion was .62 and a reliability estimate of . 79 for the fen‘iale was obtained. The reliability estimates for all sub- and total scores are reproduced in Table 3 .1. TABLE 3.1 SUMMARY OF HOYT'S ANALYSIS OF VARIANCE RELIABILITY ESTIMATES FOR THE M—SCALES FOR PAROCHIAL SCHOOL SAMPLE. _ " Ih'm'“hq ..‘_‘_:___. ...._fi -- '"f‘i'T: 73—1-...2.’ ZZ:"—"'T"f:—‘"'_:__*""‘—_'_~ ‘ VarochiaI‘SchooI 7 Parochial School Males Reliability Females Reliability __S_ul3-T<2§t N Estimate Esti_mate GSCI1 100 83 h--_T_._,_...--._-76 -mw 2 , WRL ISO .49!) .90 ___fl_ Pics 3 100 _._9__6 w. 85 A. . , HE.:~_. 1'00 _.__‘-§_8_._______.r.. __-~ 1L..._.__“-_.+_—°-.80 m- M-Scales Total lOOrflfl . 62 ‘ * _;7.9-__w_ Based on 45 items for males and 30 items for females. Based on 48 items for males and 48 items for females. Based on 20 items for males and 33 items for females. Based on 26 items for males and 25 items for females. -22- AnaLx'sis Procedures: Three procedures were used in the analysis of the data: (I) mean test of significance, (2) multiple regression analysis, and (.3) factor anal- ysis. Mean Test of Sigiiifiggncg The Student "t" was used to test the difference in uncorrelated means betvtreen the responses of the parochial school student and the public school student to the Michigan M-Scales, both the four sub-tests and the total M—Scales, as well as the difference between the grade point average and the DAT-VRscores. Multiple Regr<3§§_ion Analysis: A multiple regression analysis was performed to predict the Grade Point Average (GPA) from the (1) Gen— eralized Situational Choice Inventory (GSCI), (2) Preferred Job Character- istics Scale (PICS), (3) Word Rating List (WRL), (Ll) Human Traits Inven- tory (HTI), and (5) Differential Aptitude Tests—Verbal Reasoning (DAT-VR). The multiple regression equations were solved using the general linear hypothesis and least squares estimates of the B's. This method is developed by Kempthorne in Design and Analysis of Experiments. The Michigan State University high speed digital computer, MISTIC, with a K—16M program, was used in the analysis. The assumption is made that a multivariate-normal population has been sampled. Oscar Kempthorne, Design and AmfllSEQLEEPETIiIDQQIEI John Wiley and Sons Book Company, New York, 1952. -23- Raptor Analysis: A factor analysis of the parochial school male responses to 22 se- lected items for the GSCI were conducted and compared with the factorial . 5 structure of the public school male responses to the same test. The entire scale (45 items) was also factor analyzed in order to more thoroughly understand the underlying structure of the GSCI. . 6 . . . The normal varimax method of factor rotation was used to Simplify the. factor loadings to interpret the factors. Rotation was continued until at least three items loaded on a factor. No item was interpreted with a factor loading less than .35 for the 22 item analysis. For the 45 item analy— sis no item was interpreted with a factor loading less than . 50. Prodedure 1. After building the two responses matrices (22 x100 and 45 X 100), the CCPA program was used to ob- tain the intercorrelation among the items as well as to factorize the matrix. William W. Farquhar, op. cit. Harman, Mgdern factor Analysis, University of Chicago Press, Chicago, Illinois, 1960‘, pp. 294-308. 7 The Correlation, Commonality, and Principal Axis program, de— signed by Charles F. Wrigley and Don Kiel, Department of Psychology and Department of Agricultural Engineering respectively, Michigan State University, was performed by the Michigan State University high speed digital computer, MLSILQ; - _. 21- 2. The factors Wt'rl’t‘ then rotated to sim— ple structure by the normal varimax method of factor rotation. The K—l7M program was used for this part of the analysis. 2 full W” rigothe- 9:: The following hypotheses are restatements of the hypothwses in null form: H01 There is no significant difference in mean scores of the four sub—tests of the M-Scales between parochial and public school males and females. H02 There is no Significant difference in the mean scores of the total M-Scales be ween parochial and public school males and females. IlO3 There is no Significant difference in mean grade point av~ erage between parochial and public school malts arid fe- males. H04 The M-Scales when combined with an aptitude measure will not increase the precision of prediction of academic achieve— ment (CPA) for parochial school males and females. The following hypothesis is stated in null form, but it should be under— stood that it is not tested against an exact probability («rror rmdr 1. Ho, The factor analysis of the. male paroc: ial school stirlr ctr-3' 3 responses to the GSCI Will yield an intx-rprr-fable structure not different from that found for t..r- public scnool male. A parochial school sample was drawn from two parrirxnal :3“(. 0"“ ' ' "'_‘ “ .I'W ‘ :5" f z ‘ ‘ " I- V ' ‘ l’ I i’ f' ’ r‘ f " v ‘ r. .7 r r- sc..ools where stud':r..s ion 1) citir s and corrimi....ti-~ and 2-) par...” "JI _ 25 _ were in attendance. To study the predictive efficiency of the M—Scales, the design was based on correlational analysis while the. factored dimen- sions of the M-Scales for males were. studied through factor analysis. The results of this analysis were compared with the results obtained by Farquhar in a study based on eleventh grade public school students. 4 CHAPTER IV MEQNODIFE 1131331 eggs .131ng “R.E_G_ RE SSION ANALYSIS In this chapter will be presented the analysis of (1) the mean tests of significance between parochial and public school male and female re— sponses to the four sub—tests and the total M—Scales, and (2) the correla— tional and multiple regression analysis to evaluate the relationship between the M-Scales, the DAT-VR, and GPA. Mean Difference Between Earochialgnd Public School Student Scores: Mean scores, standard deviations, and t—ratios for public and pa— rochial school male and female samples are presented in Tables 4.1 and 4.2. The null hypothesis tested for each of the sub—tests was: Hol There is no significant difference in mean scores of the four sub—tests of the M-Scales between parochial and public school males and females. The t-ratios of 1.11 and —.27 for males and females respectively indicate non-significant differences in GSCI mean scores. The null hy— pothesis was accepted. Responses to the PICS produced a significant "t" of 2.74 in favor of the parochial school males, while the t-ratio of 1. 36 for the females was not significant. The null hypothesis was rejected for the males with a higher mean being obtained by the parochial school male. The null hy- pothesis was accepted for the females. —26- -27- E3 H4O ocm mimomlfiz EBB .HHE :Hm>>. .mOE JED 5. meocgewfip 20055on mmé nulmo. .Q «. Ilill.ll....llll|llt - I 1 2.3.11?! loll 1|-HHlHHuHIHlllI! H-1H-WHHIHHHHHIII.HHHIHW ill: .IHI- lilalHHUHLHMHi. “HM H i to? 0 gm mm. m 3.3 mm... 3.2 . om ea .35 Ema SM 3.: 3&3! 3.8. glam m2 3&3 2 , :38. .em .5- Nmm f. .M: i 3.2 :1“ 3...: 111,311: 1111::me Hag Nmm 24w 3.? _ $6 3.8 I? we .33 file N3 ama . 3.: Nam, 8.2 l; zed! 11.1.3? i: 4 m2 2 .e . 2%: J 3 as 3.8 1.13.13} Heme W11 i -l v . J i i 1.... It '1! 1 - I. If Ill- lilo . o illllli 11:11. -1591- 11111111131 mmmp®>< em; N3 .3. H mm .m . S. VTN oo.m 26a 2030 is _ . e .e i -...m..m..- - Ll. ....... p auzlltl 401m. :1 -1..- clmczl ll. 80% l :3 -Enflaw -1 . . sohzfifihwmmmeww. Mewefl--lu.fieet -- befiwfiwfilnnlull .mmqfiz AOOIOm éHmOOm/E QZ< OHAmDm mOL mHmmH H 024 mZOHHmD Qm -8; 5.8 8.8 . 8.8 85 E88. - , ‘8 . moi 8.-.. u ta? 3% - 8.8 1 is; 3.8 . .8 w -9on .-..8:N ‘08 8. Sim _ 8. w 88, . 85 . newbie - . . ._ fizom oomemu a. .wc .mlwfi!,-.-_i.lltimmfl2wzél:10le ti. ESE .. eedmmilliilll,-.m..3.9-6.\,m11 . Momma 311.6% mueflumwmflwmbgmwlmbfio l1 E838 Reg - . . I. 3.228 AOOZOm A .195 at the .05 level .45* Intercorrelations between the sub-tests of the M-Scales range from .19 to .64 for the males and from .33 to . 73 for the females. The GSCI and the WRL sub—tests of the male sample correlated sig- nificantly from zero with GPA. The correlation of the PICS and the HTI was insignificant. For the female sample, only the PICS did not correlate significantly from zero with GPA. The total M—Scales correlatr-‘d with GPA significantly at .45 level for males and .48 level for females. -31- TABLE 4. 4 INTERCORRELATIONS AMONG ACHIEVEMENT, APTITUDE, AND M-SCALES SCORES FOR PAROCHIAL SCHOOL FEMALES (N = 100) GSCI PICS - WRL HTI DAT—VR ‘ GPA GSCH .62* .59* .66* .15 ' .28* PHSS .33* .48* .11 . .17 VVRL L .73* .43* 5 .59* liTI ‘ g .27* .40* 195?:VR__- . ~55* Total ‘ _,1.V1-:S.c.a.1e§L _ _ - - - . 4 8* - -_'._. ’g - - - --- ~. m- '- - —_- . _ o .— - - - - - ..-—. --_—- _ * p .05 for the test thatr=O, ifr>.195 at the .051evel The GSCI, the PICS, and the WRL for the males correlated signif- icantly from zero with the DAT-VR. The correlation of the HTI was not significant. The WRL and HTI for the female sample correlated significantly from zero with the DAT-VR, while the GSCI and the PICS correlations with DAT-VR were insignificant. - The DAT-VR produced higher correlations with GPA than any Of the sub—tests or the total M-Scales. The correlations for the males was . 66 and for the females . 65. Regression and Multiple Correlation Analysis: Both regression aid multiple correlation analysis were applied to the data of the study. _ 32 - Regressieri Mm The null hypothesis tested for each of the sub-tests and the total M—Scales was: Ho The M-Scales when combined with an aptitude measure will not increase the precision of predictiOn Of academic achievement for'parochial school males and females. There was need for a method of testing whether or not a variable (NI-Scales) when added to an aptitude measure (DAT—VR) sufficiently in- creased the precision of estimation of the criterion (GPA). Several approaches were considered before finally selecting the F test advocated by Guilford.1 The F tests the increase in multiple R rt.- sulting from an increase in the number of variables. An increase in mul- tiple R is equivalent to an increase in predictive value (precision). Another alternative which was given serious consideration was a t test of the regression coefficient Of the variables in the regression equa— tion. The t test would have indicated whether or not the correlation CO- efficient was significantly different from zero, i.e. , whether or not the variable is sufficiently weighted to estimate the criterion. The t test does not measure the increase in precision Obtained when a variable is added to the aptitude measure. The t test may contribute such useful information as indicating overlap or intercorrelations among variables. -—-— » -..__- ~ - - - ..- - - - ‘ . — c - - -. 1 -———~.—— __._-_._‘____ The formula used for making this F test is found in j. P. Guilford, Bandages 1116.1 .Stat is ti <2 §_ 1'11. P _8 2’91) 9 leery .i 1.1.1? excl 1521.09): .a ad 17613032191, M 0G raw - Hill Book Company, Inc., New York: 1956, p. 400. - 2 2 The formula used is: P(1, N-ml—l) = (R1 " R2) (N—ml - 1) 2 -. l (n 1-R _ 33 _ Consider, for example, the case where the t test is significant and the F is not significant. The t test would indicate that the variable is useful as a predictor of GPA, but it does not add to the precision of the estimate. The t test indicates that a significant intercorrelation between variables exists. Because a table of intercorrelations is included in the study, Table 4. 3 and 4.4, the t test was unnecessary. Therefore, only F tests on the multiple R were conducted because if a "t" is Significant, the var- iable contributing to the precision must be weighted sufficiently. If "t" is not significant, the variable could not contribute to the increase in pre— cision. The results of the tests of significance are found in Table 4. 5. Significant F's were found at the .01 level for males when the GSCI, the WRL, and the total M-Scales were added to the DAT-VR. The hypothesis was rejected for these scales. The F's for the PICS and HTI for the male sample were non-significant and the hypothesis was accepted. For the parochial school female sample, significant F's were found at the .01 level when the WRL, the HTI, and the total M-Scales were added to the DAT—VR; at the . 05 level when the GSCI was added. The hypoth- esis was rejected when these sCales were added to the DAT-VR. The ad- dition of the PICS scores to the DAT—VR did not increase the precision of prediction Of academic achievement for females; thus, the hypothesis was accepted for female PICS responses. ¥¥©mo.mfi _ ¥¥VH®.® _ *¥h©v.mm wuw.~ *amm.@ so w s new. . monEom _ mi 4' 'I‘!‘ til‘n‘l‘ll‘- IIII all! a l ¥¥Hmv.m mac. *«mow.HH moo. **ovfi.n I, so e 3 u we mofimaz Oyrlll‘tri'l‘l'-ll .5580 .5 $33 3. of we Emocflcgm m ,3.» .8889 .5 H96; mo. e5 pm 30055on L t.» meflmomrE EBB >eopco>5 33H. :mEDI v.3 95mm 90%. Boom mofimfi reopomemno ooh @88on 39:05: 005:0 Hmcoflmsfim eoNZmeecoU Baum “Y‘ —_.___ . it!" Olll. 'II‘.‘ ’I I' h m IN- C H NQW I 7W N N. memofogfim m ”om I‘ltl.t Il.'llal'll.|lllll" Ivl,0,0|l mug/22mm QZ< mud/dz I m0§m>< HZHOL mQ/JEU MHEZHHmm OH OZHZOmdm/m 4n9mmrw MQDHEAE A< “Eom epmeD ceegeoe cofigetoo eoeeo oeew eat 6qu . i (ll % S. E. 86%-2 EBB _ mo . . om . 59:53: :95. cmfimm ‘ E. E. ES 950m 303, mm. mm. Boom moSmCopomemLO QoH potomofi. l i . mo . , - , , mm . \CoEex/E @0550 Ecoflmsfim cemdmeodemw mchEmm Iii illilo- 1!. III. woflpvll rilwwmuflrrliltll Ill“ ‘1. ‘Oi'l-. -I.’ .- 5:28:00 29:3 mmqfizmm Q24 mméz I m0§m>< HZHOQ mflgmu LO ZOHHSZHHmm E>IHlhmme MODHHHm/w AQHHmemmmHQ Q24. mud/R512 12909 Q24. mméOmiZ MZH LO mHmMHImDm mOm ZOHH_§_-'.L<_3_”_§ ' .389 ,.._._._--—...._... »-_._. ..- —.—-——-Q _-“~—. —_—. -._._ 4- ‘.-- ~o Why 33. a. Work rapidly just skimming" along, or b. _WQrJ<_slowlywith-gr.<:§.t___tho,r§>tlghte S , -350 .—. —- ~--—~. .. — — . .- ‘ ~ *- c u - o - - ‘4 - - -- -“M_—. * Labeled unique versus common accomplishment. Facto_r_1_. Five items had the heaviest weighting and met the minimum criterion for interpretation. The underlying theme of the item content is concern with having immediate rewards or waiting for more en- during satisfactions. The factor was labeled longierm versus short— term involvement . _ 41 _. @Qri. Six items exceeded the . 35 criterion of factor load- ing on the second male GSCI factor. The content of the items appears to be a fusion of outer or inner directedness or unique or common accom- plishment. The factor was labeled unique versus common acflgpmplish- fight.“ BEEEUIQYLQI Factors with_Al’tere_d__Cyrite1;i_on - Twenty—Two Ite__m_s_: To yield a more interpretable factorial structure of the academic motivation of parochial school males as reflected by the responses to 22 items of the GSCI, the sum of squares criterion for including a factor for rotation was changed from 1. 00 to . 50. Six factors were found to meet the criterion and rotated. Four of the factors met the additional criter- ion of at least three items with factor loadings of .35 to be interpreted. The factors, their item content and their loadings are presented in Tables 5.3, 5.4, 5.5, and 5.6. Peeler}; Six items loaded on the first factor. Four met the minimum criterion of . 35. Item 13 was found in Factor I of the first anal— ysis. The content of the items is involved with academic risk-taking. The factor was labeled gli.anc_e;tal 0.3va .2 EeEoEo? , :5 E52 itogm mom 53> EEH laced .2 i 0:253 4 cm. :23 9:02 omit. omen: He 60 ILEE onCo> 20:20: ”Go .2 we mmooxo E mc3o> cemfim 5:5 whopomm US$952 Amara: m WV EcELmSQ IEooom coEEoo momeo> SEED pcoEchk/E Eccpufiorflm mom 15> E87952 mmecco upoefic FEE momeo> EEC mcgopueocogo oc momce> 9:. uxmp.-eococO .2 0:502 mm . 53> mEofi oofifi peer: be 6059: XmECe> 35.5: ”Om. 2.0 mmooxe c2 mo3m> coma :22, mcopomu copopoED EoELm2oEoo now :oEEoo m3m5> «53:2 . 2 EcE uo>2o>5 :23 ntorMm m3m€> EETGCOA .2 0.2253 mm . :23 9:82. era. “mom: pm “coficE onCm> 208.5: “00.2 20 mmeoxo c2 mo3m> comfi 53> meopomm 683052 23: N3 cEEem 2oogom BEoopm-m Emccopm o 9:59: we omwe momao> £33 :ofificoEoO . 22 :ofioo 123.05 ~53:ch womb; 8.025852 .2 EcE ugmzdficooom :oE quo momco> males. .2 9:22.602 mm . :23 miofi web: “meg E 6059: XmESumso ”00.2 20 mmooxo E me3m> coma 5:5 mecpomm 33.30ch TEE... NNV oEEmm umcsoumm 303012 coflmpoa LBS cam song tom pow :oCeaCO 2.9333832. . \wQDHm AOO20m A/EEOOMFE 22.2. 024. ~32wame ”22.2. 2202.2 HOmO M22. OH mmmZOmmmm 2632 M22. LO mumwh>HI>.._.Zm>>H mop X5532 ZOHedqmmmOOimuHZH DMQZDOK _ 64 _ -65- mm Hem omfl Hafi cm mvfl mmfi mum 0H A.Uepmo:2Hwomfl2CmF:o mmoflc: e>n2m0d OHN mmv ace: moo ma mom mHN mvfi omH mom [\ r—4 wmo oom QHN moo ova onfi vmoi QHH mnH OHHi Vmfi mmCi one mcsfim> “copfifico mum mdfltflocn: mmc cum mean who 9mm mvo mwo veal va Hem oma mmm m_m mmm HvH vwfi wow MM cmH .NN QMH .HN mam .QN Hma - .mH meI .mH 050 .NH wufi .©H coo .mH moo .VH mmo .mH IIII .NH NH .02 :eooH .coficfico we tozm Sow 2 L955 ce>c copier; meg .mcemzvm we Ezm £303 35 ecchmcoQ ow 03D m .323 %o emcee E cofifico ope mamczoeo ”cepmoupg emSCefio mmefic: oZfimOQ mum mego> H NNV. we. 3.. 3. $4 as 852% woficzm vm Han ma mo: vm: mm mm Annv .mm mm OOI we! NH! HO Wm Kw. Gov .HN we moi mm: mm: mHu mm: Ho Aomv .om on mm: we: Ho mo mm: mm Ammv .mH MW NH mm! NOI mo Wm HN Aomv .mH Hm can we can aw mm mm Ammv .vH mm we: ma mm mm mm ow Ammv .mH vm we ma om: 0H: oat mm Acmv .NH ma ma: no: em: moi em HA Ammv .HH ma OCI NNI VH WC NH MN ABNV .QH mm mo 3 am am. 5 mm 88 .m wH mo: HH Eu mm: om 0H Aomv .m mm moi Hm: mm em ma: mm Amfiv .n mo mo OH: mH mo: mHu vol Aoflv .9 mm mm me we: mm man Hm Awflv .m em 2- 2. 3' mm 8 mm 3: .w mm ma Ho mm- mo mm: on AHHV .m am No me: #5: mm: men we Amy .N mm no no mm co co mm on .H E - i. o m e m N H .22 :3: >93mc:::c0mv muopomu HDmU MIR LO mEmSL O>>Hi>HZm>>H mOm mmOHOxC OMH:.ipreh<:iisive Study of the 1‘\','\)11Vu11011(11 I‘actors Research I’roje<;.zt No. 846, supported by the U. S. Office of Edu- cation in cooperatioii with Michigan State University, Final Report, 1963. Garrett, Wiley, "1’1 (-xliut'imi of Rum» 3.1 a 3' LiILLT_Vl'Tc_J'.ll.;:‘ JLKEL’IWL V01. 38, 1960. r‘ ‘1‘ _ - II um: in ,; i'lfj, Eersonnel _. _' -- Guilford, James P., I“;:g'irlriimentalStatist'cs jn‘Psyc‘rielogy and Education, New York: McCraw—Hill Book Company, Inc., 1956. ~67- -68- Hackett, Herbert, ”Use of MMPI Items to Predict College Achievement, " 397599118153.1153.Q.U.idé1{l_§€‘_-1.9_ulflil_1I Vol- 39, 19%- Harman, Harry 11., Modern Factor‘An‘a‘ly‘sis, Chicago: University of Chicago Press, 1960. Hill, Robert E. Jr. , "An Investigation of the Educational Development of Selected Iowa Secondary Pupils from Varied Elementary School En- vironments, " 1__4jfea_r*b_o.ok_, National Council on Measurements Used in Education, New York: National Council on Measurements Used in Education, 1957. Hill, Robert E. Jr. , "Scholastic Success of College Freshmen from Paroch- ial and Public Secondary Schools," SchooLRevigy, 69, No. 1, Spring, 1961. ilochwalt, Frederick G., “Catholic Education, U.S.A. - 1.963," The Catholic_§c*h.o§_laJo.urna_l, 63, January, 1963. Juola, Arvo, "Predictive Validity of Five College Level Academic Aptitude Tests at One Institution, " Personnel .a_nd_G_uidan_c_e_J_o_u'r_na_l, Vol. 38, 1960. Kempthorne, Oscar, Design and Analysis_of_§_xp_e_rki’ment.s_, New York: John Wiley and Sons Book Company, 1952. Klugh, Henry E. , and Bierly, Robert, "The School and College Ability Test and High School Grades as Predictors of College. Achievement, “ Educatieaiel-aed_ESL/.'Qhel.o_gi9al _l\;1€‘§1-.9_‘:1§9111:?11Fr 19, 1959. Koos, Leonard V. , Private @139..EUb1iC-S.CllQQL551.295.11.931, Chicago: Uni- versity of Chicago Press, 1931. Lathrop, Irvin T. , and Kieffer, Thomas J. , "College Achievement of Public Versus Private High School Graduates, " The Clearing House, Vol. 33, January, 1959. McClelland, D., Studies.infllyl‘otivatipn, New York: Appleton-Century- Crofts,Inc.,19SS. McClelland, D., Atkinson, J., Clark, R. and Lowell, 13., T‘heAehieve: ESELMQQYQI New York: Appleton-Century—Crofts, Inc., 1953. McClelland, D., Baldwin, A. Bronfenbren'ier, U., and Strodtbeck, P., T_a-1i’_‘l__l_1._a-i_lfid_S(,')G_l.fjf;_, Princeton, New Jersey: D. Van Nostrand Com— Pany, 1958. -69- National Education Association Research .B_r_i‘l_lg_t_i_i, XXXV, No . 4, December, 1957 . Rosen, Bernard C. , ”Race, Ethnicity, and the. Achievement Syndrome,“ AmericanSociplogicalfiefiiewJ XXIV, February, 1959. Scannell, Dale, "Prediction of College Success from Elementary and Sec- ondary Performance, “ Journal of__Educ*ati_onalflPflsychglogy, Vol. 51, 1960. Seltzer, Carl C. , "Academic Success in College of Public and Private School Students, " Journalpjgsycj‘rglogy, 25, 1948. Shuey, Audrey, M. , "Academic Success of Public and Private School Students at Randolph-Macon Woman's College. I. he Freshman Year, " 12.91.1191..9§§.