P! 109 239 THS THE EFFECTIVENESS OF A PUPIL ACTIVITY PROCEDURE IN THE TEACHING OF UNITED STATES HISTORY AT OTTAWA HILLS HIGH SCHOOL GRAND RAPIDS, MICHIGAN Thesis for the Degree of M. A. MICHIGAN STATE COLLEGE Ernest E. Ciddin gs 1940 Lg, _ .}~".$ 4J5 .m a A) 21; VI “(3 LI “rt THE EFFECTIVENESS OF A PUPIL ACTIVITY PROCEDURE IN THE TEACHING OF UNITED STATES HISTORY AT OTTAWA HILLS HIGH SCHOOL GRAND RAPIDS,MICHIGAN by Ernest E. Giddinge A THESIS Submitted to the Graduate School of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of: MASTER OF ARTS East Laneing,Michigan 1940 Ill-{551$ The writer wishes to acknowledge his indebtedness and to express his thanks to the following people who helped to make it possible to complete this study: Mr. Henry D. MacNaughton, Principal of Ottawa Hills High School,for permission to carry on this study; Dr. L. J. Luker, thesis adviser, for his valued super- vision and constructive criticism; Dr. G. P. Deyoe, and Dr. H. M. Byram for their help and suggestions; and Mrs.Giddings for typing the manuscript. 129324 111 CONTENTS Chapter Page I. The Problem - 1 A. Origin of Problem 1 B. ScOpe of the Experiment 8 C. Hypothesis 2 D. Definition of Terms 8 E. Purpose of the Investigation 4 F. Delimitation of Problem 5 II. Related Literature and Previous Studies 7 A. Related Literature 7 B. Previous Research 10 III. Procedures and Methods 18 A. Literature and Practice of Equating Groups 18 B. Equating the Groups 83 C. Procedure with Control Group 28 D. Procedure with Experimental Group 39 E. Initial and Final Tests 31 F. Establishment of Course Objectives 33 G. Major Factors Involved 34 IV. Findings and Interpretations 35 V. Summary,Conclusion,and Suggestions for Further Study 0 5 A. Summary 50 B. Conclusion. 51 0. Limitations of the Conclusion 51 D. Suggestions for Further Study 52 BibliOgraphy 53 Books and Bulletins - 53 Periodicals 55 Appendix 56 Columbia Research Bureau American History Test, Form A 57 Columbia Research Bureau American History Test, Form B 65 Otis Self-Administering Test of Mental Ability, Higher Examination: Form B 73 Number I. II. III. IV. VII. VIII. IX. XI. XII. XIII. TABLES, EQUALITY OF EXPERIMENTAL AND CONTROL GROUPS MEANS AND STANDARD DEVIATIONS OF CRITERIA USED TO EQUATE EXPERIMENTAL AND CONTROL SECTIONS MATCHED PAIRS EQUATED ON THE BASIS OF INTELLIGENCE QUOTIENTS,CHRONOLOGICAL AGE,AND HONOR POINT RATIO SCORES OF EXPERIMENTAL AND CONTROL GROUPS ON PRE- TEST AND FINAL TEST MEAN GAINS,STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF THE HIGH THIRDS OF THE EXPERIMENTAL AND CONTROL GROUPS ON THE BASIS OF INTELLIGENCE QUOTIENTS MEAN GAINS,STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF THE MIDDLE THIRDS OF THE EXPERIMENTAL AND CONTROL GROUPS ON THE BASIS OF INTELLIGENCE QUOTIENTS MEAN GAINS,STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF THE LOW THIRDS OF THE EX- PERIMENTAL AND CONTROL GROUPS ON THE BASIS OF INTELLIGENCE QUOTIENTS MEAN GAINS,STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF EACH THIRD OF EXPERIMENTAL AND CONTROL GROUPS, EQUATED ON THE BASIS OF INTELLIGENCE QUOTIENTS MEAN GAINS,STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF EACH THIRD OF EXPERIMENTAL AND CONTROL GROUPS, EQUATED ON THE BASIS OF AGE MEAN GAINS, STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF EACH THIRD OF THE EXPERI- MENTAL AND CONTROL GROUPS, EQUATED ON THE BASIS OF HONOR POINT RATIO MEAN GAINS,STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF EACH THIRD OF THE EXPERI- MENTAL AND CONTROL GROUPS, EQUATED ON THE BASIS OF PRE-TEST SCORES PRE-TEST AND FINAL SCORES, GAINS, AND DIFFERENCE IN GAINS OF THE TWENTY-EIGHT MATCHED PAIRS MEAN GAINS, STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF EACH THIRD OF EXPERIMENTAL AND CONTROL MATCHED PAIRS - iv Page 34 35 27 36 39 41 43 44 45 47 48 49 CHAPTER I THE PROBLEM A. Origin of Problem This problem has forced itself upon the author during the last five years as a result of casual ex- perimentation with various methods of classroom pro- cedure in the teaching of United States History. Debates, panel discussions, pupil forums and dramatizations were used at irregular intervals as means of enrichment of the course of study and vari- ation of the method of presentation. .The results from the standpoint of pupil interest, quantity of subject matter presented, and adaptability of the course of study to pupil participation were extremely satisfac- tory. The implication naturally followed that such devices could very well be a part of the established classroom procedure if the learning process proceeded either as well as or better than by the conventional classroom procedure. It then became the author's plan to conduct an experiment throughout one complete semester of the year 1938-1939 to determine the effec- tiveness of this method of teaching in comparison with the conventional classroom method. B. Scope of the Experiment The experiment was begun on January 27, 1939 and continued for one semester of nineteen weeks end- ing June 16, 1939. One hundred high school pupils participated, fifty as a_control group and fifty as an experimental group. The experimental group con- eisted of the nine-thirty section of twenty-seven pupils and the twelve-thirty section of twenty-five. The ten-thirty section and the one-thirty section of twenty-five pupils each constituted the control group. Ninety of the total group were in the second semester of the eleventh grade. Ten were seniors who had poet- poned taking the course because of some credit defici- ency. All were pupils at Ottawa Hills High School, Grand Rapids, Michigan. C. Hypothesis Experimentation with such pupil-classroom activ- ities as debates, panel discussions, group study, and dramatizations during the last four years have led the author to the hypothesis that in the teaching of United States History to eleventh grade pupils a procedure of pupil responsibility by means of these activities is measurably superior to the conventional teacher-dominated procedure. If this hypothesis is true the achievement or gains of the experimental group should be greater than those of the control group. If it is true there should be a higher degree of correlation between in- telligence and achievement in the group taught by that pupil presentation procedure than in the group taught by the conventional procedure. D. Definition og_Ierp§. For the purposes of this experiment the term pppr ventional classgggmprocedppg shall be used to designate a question and answer method in a teacher-dominated class. Expressions of Opinion, interpretations and comparisons were reasonably frequent by pupils but seldom originated in pupil initiative. Instead such participation was generally the result of a teacher originated question which itself probably often implied the answer. Pppil activity procedure shall be used to desig- nate a cooperative method of arriving at immediate ob- jeotives, of making assignments and of presenting the subject matter. It shall be understood to designate as much pupil participation as possible under existing conditions. Round table discussions directed by a pupil- ohairman constituted the pupil activity during a major part of the semester. A detailed account of this procedure appears in Chapter IV. Other activities were Open forum debates, formal two-man team debates, and whole class forums. _ Honor point ratig designates the point average of the individual's high school marks to the beginning of this experiment. values were assigned as follows:’ A, equals 4; B, equals 3; C, equals 2; D, eQuals l; E, equals 0. The Intelligence quotients used for equating groups were secured by the use of the Otis Self-administering test of Mental Ability, Higher Form A. E. gpppose of this Investigation; This experiment then was carried on in an effort to obtain some reliable and quantitative data on the question of the relative effect- iveness of teaching eleventh grade United States History by a student activity procedure as compared with teaching by the conventional question and answer procedure. Special phases of the purpose were: 1. To determine which method, if either is more effective with the high third, the middle third and the low third of the groups studied on the basis of mental ability. 2. To determine which method, if either, is more effective with the oldest third, the middle third and the youngest third of the groups studied. 3. To determine which method, if either,is more effective with that group of students who,regardless of age or intelligence have succeeded in obtaining teachers' marks which place them in the high third, middle third, and low third of the group studied. 4. To determine which method,if either, is more effective with the high third, middle third and low third of the groups on the basis of the pre-test scores. 5. To determine which method, if either, is more effective with the high third, middle third, and low third of the twenty-eight pairs equated on the three criteria: age, mental ability, and honor point ratio. 6. To obtain data to cOmpare the correlation between mental ability and achievement in the group taught by group activities procedure withthe correla- tion between mental ability and achievement in the group taught by the conventional procedure. F. Delimitation of Problem It is not alleged in the hypothesis upon which this investigation is based that results would support the hypothesis if the proposed activity procedure were carried on with pupils below the eleventh grade in high school. It is not contended that outcomes similar to the. anticipated outcomes in this experiment would be ob- tained in other situations unless reasonably equivalent influences existed. These rather immeasurable factors peculiar to the community include: a. A high degree of home ownership b. A high degree of professional occupations among parents c. An environment tending to make pupils well ad- justed socially d. A record of approximately seventy per cent of the school's graduates entering college e. A very negligible number of foreign born parents It is not alleged that the test used in this ex- periment measures more than one important quality. Although not exclusively a subject matter test, it is predominately that. The measurement of citizenship, attitude,and application of knowledge to current problems as examples of other out-comes of history study is not attempted objectively in this experiment. Such outcomes tending more to the subjective than to Ithe objective, fall in the class of philosophical judgments as pointed out by Murra1 in an analysis of research in the teaching of social studies. l. Wilbur F. Murra, ”Research in the Teaching of the Social Studies.u Harvard Educational Review, 7:393- 95, May, 1937. CHAPTER II RELATED LITERATURE AND PREVIOUS STUDIES A. Related Literature The concept of student activity is not new in the world of educational philosophy. There is evidence that the so-called progressive schools had their prototypes in Athenian and other classical educational systems. Aris- tctle advocated the principle of creative self-expression. He upheld a social system in which individual interests and group welfare were harmonized. In fact, much of the Greek concept of education was based upon freedom, cooper- ation, and intellectual creative activity. Mary E.Condon, in a study of the activity movement,3 concluded that it was “as old as civilization itself." She traces phases of its origin back to Quintillian, St.Augustine, and Alcuin. 3 L. C. Mossman summarized five important contributions on child activity as follows: lErasmus taught that knowledge of things is more important than knowledge of words. Bacon advocated the substitution of induc- tion for deduction. John Locke regarded the child as the center of the learning process. Rousseau contended that the child has his own way of seeing, thinking, and feeling. Herbert contributed the doctrine of interest.’’ 3. Mary L. Condon, "Classical Backgrounds of the Activity Movement.“ Unpublished Master's thesis, University of Washington, 1933. Pp. 39. 3. L. C. Mossman, The Activity Concept. New York: Macmillan Company. 1938. Pp. 168. Each of these philosophies would seem to support the hypothesis which the author proposes to test in this investigation. John Dewey confirms for us his belief in the valid- ity of experience when in a recent lecture he concluded:4 "In what I have said I have taken for granted the soundness of the principle that education, in order to accomplish its ends both for the individual learner and for society, must be based upon ex- perience -- which is always the actual life experience of some individual." He holds that this reconstruction and enrichment of experience constitutes the purpose and the end of living and education. Briggs in attempting to establish a broad concept of the function of education asserts these two rules:5 “The first duty of the school is to teach pupils to do better the desirable things that they will do anyway.“ "Another duty of the school is to reveal higher activities and to make them both desired and maximally possible.' Such a concept gives considerable support to the hypothesis of this study. Specifically applied it would help to justify a supervised cooperative group discussion procedure of attacking social, economic and political problems since these problems will be discussed anyway. 4. John Dewey, Education and Experience. New Eork: The Naomillan Company. 3 . p. 3- 14. 5. Thomas H. Briggs, Improving Instruction. New York: The Macmillan Company. 1938. Pp. 213-240. S. A. Courtis6 states the case for group activity, cooperation and democratic procedures in the April, 1938 Educational Method. He says cooperation is to be defined as simply 'working together". He says all must participate in the deliberations and all must work for the good of each. He warns that it demands self—control but promises that the procedure will yield rich dividends. Courtis7 has also pointed out the difference in practice between competitive discussion and cooperative discussion. He shows that our procedures of law, debate, parliamentary processes and diplomacy have in the past been competitive and implies that the result is a competetive philosophy. He proposes that a procedure of cooperative discussion must displace competitive discussion in order that our teaching procedures actually be contributing factors toward the end we seek. The National Education Association Research Bulletin, Modern Social and Educational Trends8 asserts that students are participating in the direction of their own learning through the use of debates, conferences and committees. 6. S.A.Courtis, “Techniques of Cooperation.’ Educational Method, 7:349-50, April, 1938. 7. National Education Association, Teachers and Cooperation. Yearbook on Cooperation, National Education Association, Department of Supervisors and Directors of Instruction, Committee on COOperation. Detroit, Michigan: November, 1937. Pp. 81. 8. National Education Association, Modern Socialnand Eduzp gational Trends. Research Bulletin of the National Education Association, Vol. XII, No. 5. Washington, D.C.: November, 1934, Pp. 284-387. lO Kilpatrick9 shows that since we live in continuous change it has become the responsibility of education to prepare the child not for any specific situation but to provide him with experience in adjustment to the inevitable new condition. He must learn by practice the procedure of interpreting the present by applying the lessons of the past. Kilpatrick asserts that education then must provide methods of attack rather than the solution itself. B. Previpus Research STUDY 110 1. Purpose. The aim of this experiment was to compare the effectiveness of teaching by the conventional textbook method with that of teaching by a procedure of pupil-activity. 8. Group Studied. Thirty pupils with an age range of from 13 to 16 years constituted the experimental group. This group showed a median chronological age of 13-6. They showed a reading grade-level range of from grade 6-8 to grade 9—A. Their English grade-level also ranged from 6-3 to grade 9-A, as did also their arithmetic grade- level. The control group of thirty-four also had a median chronological age of 13-6, but had a range of from 11-0 to 16-0 years. The control group ranged from 5-A to 9-A in 9. I.H.Kilpatrick, Education for a Changing Civilization. , New York: The Macmillan Company, 1938. Pp. 81. 10. Teresa Ryan, 'An Experiment to Study the Effect of Two Methods of Instructions in Attempting to give an Under- standing of Geometric Concepts in Two 8A Classes.‘ Unpublished Master's thesis, University of Detroit, 1933. Pp. 64. 11 reading grade-level, from 5-A to 9-B in English grade- level, and from 6-8 to 9-A in arithmetic grade level. 3. Specific Problem. To determine which method of instruction was more effective for each of the following levels: (a)above average; (b)average; (c)below average. 4. Means Used. Twenty-six items selected from the Butler Mathematics Test11 were used as the prertest and final test. The time allowed for the test was eight minutes. 5. Treatment of Data. Tables were prepared to show comparative gains made by experimental and control group of like mental rating, the two groups being devided into sub-groups having mental ratings listed as: A, B, C, and 0-. Separate tables were also prepared to show comparison of gains made by students of 9A, 9B, 8A, 8B, 7A, 7B, and 6B reading grade equivalent. A total of twenty-six similar tables were prepared. 6. Conclusions. In every mental group rating listed, the experimental group was superior to the control group in the average gains. This was also true in every group in which the similarity was reading ability. The experi- mental group was superior to the control group in all grade levels of arithmetic abilities except in GB grade. The author gives the following as possible reasons why 11. C. H. Butler, "Mastery of Certain Mathematical Concepts by pupils at the Junior Hi h School Level.'° The Mathemappcs Teacher, 35:13 , March, 1933. 13 pupils did better work under the experimental procedure: a. The work was more appealing b. Students did constructive work and did not have to be satisfied with theory only. c. Students had the satisfaction of seeing abstract formulas made concrete. d. The students had opportunities of self-expression. STUDY II 13 1. Purpose. This study attempted to discover the effectiveness of group cooperation. 3. Group Studied. The control consisted of 108 graduate students in education working individually. The experimental group consisted of these same 108 students working in various sized groups. 3. Factors Studied. The ability to make up words out of the letters of a given word. 4. Conclusion. Group thinking proved to be superior to the average of even the best members of the group work- ing alone. The group product in groups numbering up to ten, proved superior to individual production. 6. Evaluation. It would seem significant to the present study that Mr. Watson concluded that group coopera- tion showed superior results. It should support the hypo- thesis of the present study. 13. G. B. Watson, 'Do Groups Think More Efficiently Than Individuals?" Journal of Abnormal and Social Psychology. 39:336-353. May,1934. ) 13 STUDY III 13 1. Purpose. To determine the effectiveness of taking a Psychological Test as a group. 3. Group Studied. Twenty college upper classmen were paired. 3. Factors Studied. Group efficiency as compared with individual efficiency in answering questions of a PsycholOgy test. 4. Means Used. Form I, and Form II, used alternately by control and experimental groups. 5. Conclusion. Means for the group were considerably higher than for those working alone. 6. Contribution to present problem. This study seems to show that group attack on this type of problem is super- ior to an attack by individuals working alone. It would provide evidence to show the effectiveness of cosperative diagnosis, analysis, and solving of problems common to the group. STUDY IV 14 1. Purpose. To determine the relative effectiveness of three different methods of laboratory instruction in high school biology. 13. A. B. Weston, and H. B. English, "The Influence of the Group on Psychological Test Scores." American Journal 01 Psycholo , 37:600-603, March, 1936. 14. P. 0. Johnson, 'A Comparison of the Lecture-Demonstration, Group—Laboratory Experimentation and Individual-Laboratory Experimentation Methods of Teaching High School Biology.“ Jgurnal of Educational Research, 18:103-111, September, 1 38. l4 2. Groups Studied. Three classes of eleven pupils each, two of seventeen pupils each, and one of sixteen pupils, constituted the experimental groups. A rotation plan was used, to enable each class to perform eight lab- oratory sxercises under each of the following methods of instruction, the lecture-demonstration, the group-labora- tory and the individual-laboratory. 3. Means Used. Objective tests,containing from fifty to two hundred points, were used. 4. Conclusions. The test results were in favor of the lecture-demonstration method as compared to either of the other two methods. The difference, however, was not statistically significant. 5. Evaluation. Although Mr. Johnson did not claim significant difference in the three methods, he failed to admit two other factors which would seem to condition any possible conclusions from his experiment: (a) Only one factor, intelligence, was used as a basis for pairing the groups; (b) The total number of pupils was only eighty- four or an average of twenty-eight for each of the three methods of teaching. 15 STUDY V 1. Purpose. The object of this study was to deter- mine the effect of group performance upon college students. 15. P.R.Farnsworth, "Concerning So-Called Group Effects." Pedagogical Seminar, 35:387-94, September,l938. 15 3. Group Studied. The groups consisted of twenty and thirty-six respectively. 3. Means Used. Four different tests were used on separate occasions: a. Ohio State University, Form 10, part 5. b. Terman Group Test, Form A. c. Terman Group Test, Form B. d. Otis Self-administering Test. 4. Conclusions. No consistent group effect was ‘found. Students working alone were slightly superior on difficult items. STUDY v1 15 1. Purpose. This was an investigation to compare the results of teaching American History by the recita- tion method with the results of teaching by a combined contract and discussion method. 3. Group Studied. Twenty-seven eleventh grade pupils constituted the control group and another twenty- seven of the same grade , the experimental group. The data used, however, were those pertaining to twenty-one pairs of these pupils matched on mental ability as the only criterion. 16. J. T. Yurkewitch, "A Controlled Experiment on the Con- tract-Discussion Method of Teaching Vs. The Recitation Method in United States History." Unpublished Master's thesis, Pennsylvania State College, 1933. Pp. 33. 16 3. Means of measurement. The following tests were used: 1 Columbia Research Test in American History; 3 Pressey and Richardé Test in American History; 3 Wesley Test in Social Terms; 4 Wesley Test in Political Terms. In three of these one form of the test was used as an initial test, and another form as a final test. The same Pressey and Richards Test, however, was used at the beginning and end of the experiment. 4. Conclusions. Of six tests used, five indicated a definite superiority of achievement in the group taught by the contract-discussion method. Two tests show a superiority in the class taught by the regular recitation method. The author states that this superiority may be due to formal drill in the class taught by recitation method and implies that no formal drill was used in the class taught by the contract-discussion method. 5. Evaluation. The purpose of this study by Mr. Yurkewitch is almost identical to the purpose of the present study. The experimental factor "Arccntract-dis- cussion" method is shown by the author's definition to consist chiefly of pupil discussions from an outline con- tract. The chief limitation of the investigation is the small number of pupils, forty-two, and to the probability that pairing was in error since only one criterion, the intelligence quotient, was used for pairing individuals. l? The obtained results show a definite superiority in the contract-discussion method. Four of these studies found group activity to wield superior results to individuals working alche. One study found that individual study produced superior gains and in one study no difference in gains was dis- covered. Two of the four studies which found superior achievement resulting from group activity, dealt with high school students. The evidence, then, from these studies is inconclusive. 18 CHAPTER III PROCEDURES AND METHODS A. Litergturg and Practice of Equapgng Groups In attempting to equate the experimental and con- trol groups and to match pairs of individuals the author surveyed the devices and techniques suggested in the available literature on such procedures. A summary of certain of these techniques seems essential to justification of the methods used in the present study. Melby and Lien17 suggest using ordinary school classes without modification. Douglass18 in a study of two sequences in super- vised study equated groups on the basis of chronological age, Otis intelligence scores, and scores made on stand- ard achievement tests. He, however, transmuted the ob- tained intelligence and achievement test scores into regressed or estimated true scores in order to nullify to some extent the chance error of the raw scores. His next step was to convert the regressed subject matter scores into composite scores by the method of equal variables. 17. E. O. Melby, and Agnes Lien, "A Practicable Technique for Determining the Relative Effectiveness of Dif- ferent Methods of Teaching." Journal of Edu tional Rpsggrch, 19:355-364. April, 1 39. 18. Earl Roy Douglass, The Experimental Comparison of thg Relative Effectiveness of Two Sequences in Super- !ised Study. University of Oregon. 1937.Pp.l77-183. 19 Noll19 equated two groups of college students by matching individuals on percentile rank on American Council Psychological Test and on honor point ratio on previous work in college. Halvorsen80 considered six college classes closely enough equated by using only the scores made on the _ Nelson High School English Test. Courtis31 contends groups cannot reliably be matched without taking into consideration the "growth culture". In equating groups of second grade pupils averaging eighty-four months old he speaks of them as the "same age" when his data show them to vary by eight months. Monroe and Engelhartz3 say the following character- istics affect achievement in the field of experimentation; previous achievement in the field of investigation, study habits, and such personality traits as attitudes, ideals, and interests; health, sex, and race. 19. Victor H. Noll, "The effect of Written Tests Upon Achievement in College Classes: An Experiment and A Summary of Evidence." Journal of Educational Research, 33:345-58, January, 1939. 30. N. O. Halvorsen, Angxpepiment Comparing the Egpective- pose of Two Mgphods of Indicating Erropg In Students Themee. Research Report No.35. Cedar Falls, Iowa: Iowa State Teachers College. 1939, Pp. 1-3. 31. S. A. Courtis, "Criteria for Determinin the Equality of Groups" Schoolggnd Society, 35:874- , June 35,1933. 33. I. S. Monroe,and Max D. Engelhart, Experimental Research In Education. Bureau.of Educational Research,Bu11etin 48, Urbana, Illinois: University of Illinois, 1930. Ppe 19-300 20 Yurkewitch,83 in a controlled experiment conducted with two classes of twenty-seven pupils each, used only' the criterion of intelligence for forming matched pairs. Of these fifty-four pupils, he found that forty-two pupils could be matched with a difference of not more than three points between the individuals paired. Franzen24 conducted an experiment to test effective- ness of teaching with groups equated on the criteria of percentile rank on "intelligence examinations" and initial scores of the test used. Butterfielda5 used the criteria of age and intelligence quotients to equate two groups in an investigation of effec- tiveness of two methods of teaching history. 36 used the Nelson-Denny Reading Test as R. R. Fahrney a basis of pairing and considered it quite dependable for pairing purposes. 33. J. T. Yurkewitch, Op. cit. Pp. 3. 34. Carl G. F. Franzen, A CCOmparison of the Results made On pertain Standardizedi Tests by Pupils in the Bloomipgton High School Who Were Taught in Classes of the Same Grade pppUniversity Student Teachers and by Regular High School Teachers. Bulletin of the School of Education. Volume II, No. 4. Bloomington, Indiana: Indiana University. March, 1936. Pp. 4. 35. A.I.Butterfield, "A Study of the Relative Effectiveness of Two Methods of Teaching History." Unpublished Master's thesis, University of Michigan, 1935. Pp.9 36. R.R.Fahrney, AgCOmparison of the Effectggeness for Learn- inggof the Lecture-Quiz Technique with the Classrpom- discussion Technique in a Freshman S_phomore Course “in American History. Bureau of Research, Iowa State Teachers College. Cedar Falls, Iowa. Research Report, No. 36. March, 1939. Pp. 3. 31 In the ten studies reviewed here only two groups were equated on identically the same basis. Each of these used only reading test scores as a basis of equation. lone other of the ten experimenters used this factor even with some other. One equated on growth curve and none of the others attempted to make any use of it. Only one of the ten studies made use of pro—test scores and these were used in combination with intelligence quotients. One in- vestigation matched groups on honor point ratio while no other study used this basis of equating even with other factors. Only one study considered ordinary classes closely enough equated for experimental pur- poses and only one made use of a composite score. Four used intelligence quotients as a basis of match- ing but none of the four used the same combination of other factors. No agreement seems to exist in practice regarding the most scientific method of equating groups. However, a common objective does seem to exist. All authorities and experimenters seem to agree that the aim should be to have groups so nearly alike that either group could be used as the experimental group. 23 B. Eguating the Groups Two conditions would seem to Justify the method of equating used by the author in this study rather than the regressed score-composite technique used by Douglass. The present author is dealing with one hundred pupils while Douglass' experiment involved approximately five hundred. It has been possible be- cause of the small number of cases in the present ex- periment to conveniently re-test for mental ability in those cases where the teacher's judgment conflicted with the obtained score. A second difference is that the present author personally taught all of the classes involved while in Douglass' experiment the classes were taught by several different teachers. Monroe and Engelhartg7 suggest that since a com- posite score implies the weighing of criteria and since no uniformly acceptable plan of determining appropriate weights exists that therefore it seems best to equate on the basis of intelligence and a few other clearly definable factors. Two months previous to the beginning of this experi- ment, all the pupils of the school were given the Otis Self-administering Test of Mental Ability, Higher Exam- ination, Form B. The test was administered by the prin- cipal and two teachers with training and experience in 87. I. 8. Monroe, and M. D. Engelhart, op. cit. Pp. 29. 33 giving such tests. These few pupils who missed the test because of absence were given the test by the author under as normal conditions as possible during the second week of the semester. The honor point ratio obtained was based on all academic subjects studied by the pupil in high school. Although Monroe recommends that only those subjects in the field of experimentation be considered, it has seemed wise in this experiment to include achievement in all subjects except those in which the objective is the attainment of a skill such as typing. Grades in typing, woodworking, machine shop, printing, sewing, cooking and art were disregarded. At.the beginning of this experiment several individ- uals had been enrolled in only two previous semesters of history, and that in the first year of high school, it was therefore concluded that the most valid honor point ratio would be secured by the inclusion of all academic subjects studied and by the omission of marks in voca- tional subjects. Table I, Page 24, shows the intelligence quotient, chronological age, and honor point ratio of each individual in the investigation. TABLE I EQUALITY OF EXPERIMENTAL AND 006799; GROUPS 24 EXPERIMENTAL PQL; IOQO GOA. '1 00 mandamus-ma) ‘ Left school during semester. H. ... 133 16.41 3. 137 15.75 3.70 136 16.67 3.78 135 16.91 3.88 134 16.67 3.35 133 16.50 3.08 133 16.41 3.07 131 16.35 3.66 130 17.16 3.69 119 16.41 3.50 119 16.91 3.41 118 16.50 3.15 118 16.58 3.65 118 16.75 3.93 116 16.91 3.00 115 16.35 1.54 115 17.35 1.70 114 ' 16.33 3.53 113 16.75 3.83 113 16.67 3.33 111 16.35 1.91 111 16.75 3.58 110 16.16 3.08 110 16.41 3.15 110 17.16 3.00 108 19.00 3.00 108 16.35 3.33 108 17.33 3.69 108 17.83 3.00 107 16.75 3.31 105 16.41 1.83 105 16.83 3.36 104 16.41 3.50 103 ,16.67 3.34 103 16.35 3.91 103 17.41 1.77 103 16.67 3.57 101 17.35 1.41 99 17.77 3.37 99 16.75 3.00 97 16.33 3.57 97 17.16 1.41 95 16.83 1.50 95 18.33 1366 94 18.33 1.18 93 16.50 1.40 93 18.00 3.00 89 16.08 3.35 86 16.83 3.35 86 16.91 1.91 86 16.50 3.53 computations. CONTROL j._ 94911 1.0. 0.1. 9.3 a. 1 136 16.83 3.64 3 136 16.68 3.73 3 134 16.67 3.90 4 134 16.16 3.93 6 133 16.41 3.60 6 133 16.67 3.31 7 133 16.00 3.31 8 131 16.08 8.84' 9 130 16.83_ 3.66 10 2119 16.67 3.31 11 118 16.58 3.30 13 118 17.00 3.90 13 118 16.60 3.67 14 117 16.68 1.91 16 118 16.08 3.66 18 115 17.18 1.33 17 114 16.83 1.76 18 113 16.50 3.37 19 113 16.67 3.43 30 113 16.91 1.75 31 113 16.76 3.33 33 111 16.60 3.00 33 111 16.60 3.46 34 110 18.35 3.00 36 109 17.83 1.93 36 108 18.36 1.73 37 107 16.41 3.74 38 107 17.18 3.06 39 107 16.41 1.76 30 -106 16.68 3.41‘ 31 106 16.60 3.36 33 106 18.33 1:30 33 106 17.67 3.66 34 106 16.83 3.94 35 104 16.36 1.60 38 104 16.83 1.60 37 103 16.76 3.41 38 103 17.68 1.16 39 103 17.67 1.65 40 103 16.16 3.08 41 99 17.33 1.07 43 99 17.00 1.93 43 96 16.08 3.60 44 94 16.91 1.76 46 94 17.60 1.31 48 93 17.67 3.06 47 93 18.16 8.34 48 89 18.00 1.94 49 86 17.16 1.91 60 86 17.36 1.16 Not included in statistical TABLE II IEANS AND STANDARD DEVIATIONS OF CRITERIA USED TO EQUATE EXPERIMENTAL AND CONTROL SECTIONS t Group N. C.A. S.D. I.Q. S.D. H.P.R. Experi- mental 2nd Hr. Sec. 26 16.67 .31 110.2 9.2 2.45 5th Hr. Sec. 24 16.90 .70 107.8 12.4 2.27 Total Group 50 16.70 .46 108.7 11.6 2.32 Control 3rd Hr. Sec. 25 16.79 .72 108.8 12.0 2.23 6th Hr. Sec. 25 16.73 .56 108.2 8.9 2.19 Total Group 50 16.76 .64 108.7 10.8 2.19 25 S.D. .49 .61 .58 .61 .63 .60 Table II shows a difference of only .06 years between the mean chronolOgioal age of the control group and the experimental group. It shows a zero difference in mean I. Q. On this criterion then the total experi- mental group is more nearly equated to the total con- trol group than either of the experimental sections are equated to one another or to either of the control. 36 sections. For this reason and for the purpose of dealing with larger and therefore more homogenous groups it has been considered justifiable in this study to deal statistically with the total groups rather than with separate sections. Table III. Page 2?, shows twenty-eight matched pairs to be used as a check on the results obtained with the total experimental and control groups. Three criteria were used as a basis of matching or pairing individuals. No pairs were formed in which one individual differed from the other more than three points in intelligence quotient as determined by the Otis Self-administering Test of Mental Ability, Higher Examination, Form B. No pairs were formed in which there was more than nine months difference in chronological age and a nearness of twenty-five hundredths in honor point ratio was rigidly adhered to. On this basis it was possible to match fifty-six individuals into twenty-eight pairs. . ILDr r§. In... ‘5. TABLE III HATCHED PAIRS EQUATED ON THE BASIS OF INTELLIGENCE QUOTIENTS&CHRONOLOGICAL AGE AND HONOR POINT RATIO~ EXPERIMENTAL, CONTROL P’ I. .' C.A. H.P.R. Pupil 1.9. C.A. 33‘1” “"137 15.75 3.70 _ 4 134 16.18 ~ 5 134 16.67 3.35 6 133 16.67* 8 133 16.50 3.08 3 134 16.67 7 133 16.41 3.07 8 131 16.08 9 130 17.16 3.89 9 130 16.83 11 119 16.91 3.41 13 118 16.50 13 118 16.50 3.15 11 118 16.58 14 118 16.75 3.93 13 118 17.00 15 116 16.91 3.00 14 117 16.58 15 115 16.35 1.54 16 115 17.16 18 114 16.33 3.53 18 113 16.50 19 113 16.75 3.83 15 115 16.08 30 113 16.67 3.33 19 113 16.67 33 111 16.75 3.58 33 111 16.50 33 110 16.16 3.08 33 111 16.50 34 110 16.41 3.15 31 113 16.75 36 108 17.00 3.00 34 105 16.83 37 108 16.35 3.33 30 106 16.67 39 108 17.83 3.00 38 107 18.17 38 108 17.33 3.69 33 105 17.67 30 107 16.75 3.31 31 105 16.50 31 - 105 16.41 1.83 39 107 16.41 33 104 16.41 3.50 37 103 16.75 34 103 16.75 3.34 40 103 16.17 38 103 17.41 1.77 39 103 17.67 38 101 17.35 1.41 36 104 16.83 39 100 16.67 1.77 43 99 17.00 48 93 18.00 3.00 48 89 18.00 Kean 110.9 16.76 3.41 110.5 16.71 8'. Do 908 058 056 9.3 048 38 0. Procedure with Control Group. Samuel R. Powers38 warns against the inadequacy of merely calling a control group "old", "conventional“, 'traditional'. The following is an attempt to avoid this error and to provide a comprehensive and complete description. Assignments in the two control sections were by textbook chapters as units of work. There were as many units or chapter divisions as there were weeks in the semester, although the time devoted to units varied from three to seven days. In the control sections the assign- ments were arbitrary, no choice being allowed the pupils regarding the topic or topics each would study intensively. Assignments were made from three to seven days in advance. Preparation on the assignment depended chiefly upon the responsibility assured by each individual pupil. Thirty minutes, or half of the class period, was used for super- vised study twioe each week. No formal drill was used. In the recitation itself, the instructor asked for recitation by topic, and after the pupil had exhausted his information and opinions on the topic, the instructor questioned the pupil regarding material omitted or slighted. Thought and discussion questions were used as much as possible by the instructor. 38. Samuel R. Powers,"PsycholOgy and Methods in High School and College.'Review of Educational Research, 8:60-66, February, 1933. 39 Many such questions came from the students, but since pupil initiated discussion was the chief experimental factor, such questions were neither encouraged nor dis- couraged in the control sections. D. Prgceduge With Experimental Group. Student activity constituted the experimental factor in this study. Specifically then the following order of procedure was used: 1. Before starting a new unit of work, one-half hour was devoted to a preview of the unit about to be studied. 3. Major topics for special study and analysis were agreed upon in class. _ 3. Small groups of four or five students selected the topic of their choice for intensive study and re- search. 4. The following class period was usually devoted entirely to study and preparation by these study groups working in study rooms adjacent to the class room or adjacent to the library. The teacher seldom joined these groups except when requested. However, this was usually frequent enough to keep him occupied during the complete class period. The first act of the group meet- ing separately was to choose a chairman. 30 5. The discussion began on the third day following the choice of topics. It was understood to be the duty of the chairman to direct but not to arbitrarily dictate the course of the discussion. Participation by the other members of the class was always invited. During the first month of the experiment an attempt was made to formulate whatever principles seemed necessary to govern conflicts which arose or were anticipated. These principles thich gradually evolved as a result of pupil-teacher cooperation will help to clarify the experimental procedure: 1. Topics shall always be chosen at least two days in advance. 3. Panel shall choose a chairman. 3. Panel shall have at least one meeting in the study room before appearing before the class. 4. Hembers of the panel shall be free to interrupt the discussion at any time. 5. Other class members shall be free to contribute or question panel at any time. 6. The chairman shall be responsible for bringing out ideas from the panel members. 7. Differences of opinion are welcome. 8. Differences of opinion of authorities will be analysed and accounted for if possible. 9. Authorities quoted or mentioned should be identi- fied if not well known. 31 10. The chairman or someone previously designated by the chairman or group will summarize the findings. 11. Ihen assignments of topics for group discussions are given time will be allowed for individuals to suggest specific questions or problems they wish analyzed. 13. Group discussions are not to be considered debates. One enters a debate with rather a closed mind but a discussion.with an open mind. 13. One of the purposes of group discussions shall be to understand completely the meaning of the aur thority studied and of one another. 14. Permitting the discussion to proceed beyond a point not fully understood by any member of the class shall be considered an offense against him. 15. Remember that impuxwement in the procedure of group discussion may be as important as attain- ment of subject matter. E. Initia;_and Final Tests. 1. Selection of the test. Each of the United States History tests suggested 39 by Krey was surveyed to find the most suitable one to measure the initial status and final status of the individuals 39. T.L.Kelley, and A.C.Krey, Tests and Measurements in the Socialgggiences. Part IV Report of the Commission on the Social Studies. American Historical Association, New York; Scribner and Sons, 1934. gPp.610-616.. 38 to be tested in this experiment. The tests suggested in the National Education Association Research Bulletin, ”Improving Social Studies Instruction'so were also sur- veyed and considered. On the criteria of validity, re- liability, objectivity, completeness, variety of type of test, cost, method of scoring, and availability, the Columbia Research Bureau American History Test was superior to each of the others for the purpose of this experiment. The test was prepared by Harry J. Carman, Thomas N.Barrows and Ben D. Wood of Columbia University. The authors found a coefficient of reliability of .91 between the two forms A, and B, of the test. They also found that when carefully checked against the New York Regents' examinations and against the final high school ratings in American history in Dewitt Clinton High School in New York City, the Columbia Research Bureau tests correlated more highly with each of these criteria than they did with each other. The test showed a correlation of .883 with the final high school marks in American history, and .487 with the Regents' ex- amination marks.. 3. Use of the tests. Form A, of the test was given during the week preceding the first week of the semester. This was to 30. National Education Association, ImprovingASocial Studigg Instruction. Research Bulletin of the National Edur cation Association, Vol. XV, No. 5. Washington, D.C. November, 1937. Pp. 284-287. 33 determine the initial status of each individual. Form B, was used during the nineteenth week of the semester to determine the final status. Special care was taken by the two teachers giving the test to maintain the proper conditions during the giving of each test. Directions were followed exactly. Ninety minutes was the total time allowed for each test. Tests were corrected by the author and one paid student who was carefully supervised. F. Establishment of Course Objectives. During the second week of the semester one class period was devoted to the process of establishing and analyzing objectives for the course. The following ob- jectives were accepted. Pupils in all sections were kept aware of these throughout the semester: Objectives of the United States History Course. 1. To give pupils the most ggalistic knowledge that is possible of the conflicting groups and controversial issues in the past and present of our Nation. 3. To prepare pupils for promoting a wiser and more effective cooperation among individuals, states, groups and nations. 3. To develop character by giving pupils a desire to use knowledge for beneficentl social ends. 34 4. To develop the intellectual processes desirable in a democracy: a) Skill in exploring and stating both sides of controversial questions. (b; Skill in selecting and verifying information. c Skill in discussing social problems. 5. To give pupils knowledge of historical, sociallgeo- graphical, politicalJ and economic facts which may be of importance to them in later life. G. Major factors involved. 1. Constants. The following factors were alike for the control and experimental groups. (a) All pupils were from one school. (b) One teacher taught all four of the sections. (c) The same textbook was used in all sections. (d) Identical course objectives were adhered to with all sections. (e) Both groups were carried on during the same semester. (f) No pupil was told that an experiment was in progress. (g) One control and one experimental section met in the forenoon and one of each in the afternoon. (h) Motivation from tests was similar in all sections, since similar weekly and monthly tests were used. 3oJariah1L The only variable factor involved in so far as the author could control the experiment was the method of presenting the subject matter to the experimental group. CHAPTER IV FINDINGS AND INTERPRETATIONS In Table IV, Page 36, the individuals of the ex- perimental and control groups are in the order of the pre-test scores within the respective groups. Number 4 of the experimental group received the highest pre- test score and therefore appears first in the experi- mental list. The third column in each list shows the final score,that is,the score obtained by giving Form B, of the Columbia Research Bureau test which is de- scribed on page 38. In several instances irregular gains appeared. Certain of these are significant enough to be pointed out: Number 1, of the experimental group with the highest I.Q. of any pupil in the study, scored 84 on the pre-test and 98 on the final which was a gain of 14 or 1.3 points below the mean gain for the experi- mental group. This individual, in the teacher's opinion contributed more to the class than any other pupil during the semester. However, it is true that during the semester several other students in number l's section complained that he talked more than his share of the time. Number 53 of the experimental group with the 1cwest I.Q. scored 54 on the pre-test and lost 3 points to score 53 on the final. 35 1‘1 ‘17. (I)! . ”\vs‘r‘s. as“ . .. TABLE IV SCORES OF EXPERIMENTAL AND CONTROL GROUPS ON PRE-TEST AND FINAL TEST Experimental ‘_ Control Pupil Pro-test Final Pupil Pro—test Final 4 105 138 4 131 135 37 101 110 13 107 117 31 100 130 3 105 110 9 99 138 3 97 99 33 95 89 13 87 131 11 90 103 33 83 114 34 89 105 6 81 94 1 84 98 ll 79 75 5 83 7O 18 77 9O 37 78 83. 30 76 83 8 77 101 35 75 83 36 77 86 31 74 90 18 77 81 1 74 109 16 75 98 7 73 84 6 75 100 5 73 118 45 73 90 41 73 41 13 73 74 37 69 53 43 ' 71 81 48 69 83 3 69 91 9 68 71 36 68 103 34 68 96 3O 66 51 8 64 87 3 66 88 36 63 69 10 64 86 14 63 60 35 61 90 33 61 67 14 61 81 30 61 73 17 60 73 10 60 114 33 58 69 15 60 50 33 57 74 38 58 107 13 56 89 34 57 57 4O 54 73 37 56 119 53 54 53 33 55 60 39 54 73 44 55 58 34 53 75 16 53 37 15 53 55 31 53 63 33 50 54 49 51 45 47 48 36 36 50 84 41 48 ' 54 43 49 53 19 47 73 17 48 63 7 47 43 43 45 56 49 45 51 19 44 44 43 45 71 39 44 53 38 44 63 50 44 38 50 41 79 45 4O 53 38 39 56 35 39 76 31 '37 55 47 37 43 39 36 60 39 37 43 46 36 63 33 34 38 44 33 34 38 33 46 48 30 69 46 3O 33 30 35 49 4O 35 38 37 Pupil number 3, with the highest honor point ratio in the experimental group, 3.70, scored 69 on the pre-test and gained 33 points on the final test. Pupil number 46 who had the lowest honor point ratio scored 36 on the pro-test and 63 on the final, gaining 36 points. The greatest gain among the experimental pupils was 39 by pupil number 48 who had an I.Q. of 91. The greatest loss among the experimental group was 15 points by pupil number 30 with an I.Q. of 107. Pupil number 4 ranked first on both pre-test and final in the experimental group. lithin the control group, similar marked instances occurred: Pupil number 1, with highest I. Q. scored 74 on the pre—test and 109 on the final test, gaining 35 points. Pupil number 50 with the lowest I.Q. of 86 scored 44 on the pro-test and 38 on the final test, showing a loss of 16 points. The greatest gain was 63 points made by pupil number 37, with an I.Q. of 107. The greatest loss was 31 points, by pupil number 41 with an I.Q. of 99. Pupil number 4, with an I.Q. of 131 ranked first on both pro-test and final test. The oldest individual, number 36, made a gain of 34 points. The youngest, number 7 made a gain of only 11 points. Number 4, with the highest honor point ratio 38 in either the control or experimental group gained only 4 points between the pro-test and final but in spite of this small gain, ranked first in both pre-test and final. Table V, Page 39, shows that the high third of the experimental group made a mean gain of 16.5 in contrast to the mean gain of 13.3 of the control group. The range of gains among the experimental pupils in from -13 to 33 or 45 points while the .range in gains of the control pupils is from ~16 to 54 or 70 points. The wider range of gains of the control pupils would suggest that the experi- mental procedure was more effective since it induced more uniform gains. The standard deviation of a frequency distri- bution of the gains of the experimental pupils is 13.36 as compared to 18.4 in the control group. This dif- ference is an indication of greater consistency in gains among the experimental pupils. In the difference column, no algebraic sign be- fore the digit indicates a difference in gain in favor of the experimental individual. A negative sign indi- cates a difference in gain, in favor of the control in- dividual. This column shows eleven cases in which the experimental individuals showed superior gains and six cases in which the control individual made the superior gains. 39 TABLE V HEAN GAINS, STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF THE HIGH THIRDS OF THE EXPERIMENTAL AND CONTROL GROUPS ON THE BASIS OF INTELLIGENCE QUOTIENTS Experimental Control High Third High Third Pupil Gain Pupil Gain Diff. 1 14 1 35 —_ -31 3 33 3 5 17 3 33 3 3 30 4 33 4 4 39 5 -13 5 46 -58 6 35 6 13 13 7 -4 7 ll -15 8 34 8 33 1 19 39 9 3 36 10 33 10 54 —33 11 13 ll -4 9 13 33 13 10 33 13 3 13 34 -33 14 30 14 -3 18 15 3 15 -10 13 16 ' 33 16 -16 39 l7 13 17 19 -6 Mean 16.5 13.3 3.0 S. D. 13.36 18.4 40 Table VI, Page 41, shows the middle thirds of the experimental and control groups on the basis of intelli- gence quotients. The control group made a mean gain of 16.1 which was 3.1 greater mean gain than the 13.0 of the experimental group. The importance of this super- iority, however, is lessened by the fact that only seven control pupils made superior gains while eight experi- mental pupils made superior gains. In this group the range of gains for the control individuals was from 0 to 63 or 63 points. Among the experimental individuals the gains ranged from -15 to 30 or 45 points. It is obvious that pupils number 35, 36, and 37 of the control group differed enough from the mean as to account in a large degree for the apparent superiority of the control group. The standard deviation of 15.8 of the gains dis- tribution for the control pupils is an indication of inconsistency. It is an indication of extreme cases. The standard deviation of the gains of the experimental pupils is an index of somewhat greater consistency among them, than was found among the control pupils. 41 TABLE VI HEAR GAINS, STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF THE MIDDLE THIRDS OF THE EXPERIMENTAL AND CONTROL GROUPS ON THE BASIS OF INTELLIGENCE QUOTIENTS Experimental Control Middle third _7 Middle third Pupil Gain Pupil Gain Diff. 13 4 18 13 ~9 19 36 19 O 36 30 34 30 7 17 31 18 31 16 3 33 6 33 6 ~13 33 11 33 31 ~30 34 16 34 O 16 36 9 35 37 ~38 37 9 36 34 ~35 38 17 37 63 ~46 39 18 38 14 3 30 ~15 39 8 ~33 31 3O 30 ll 19 33 17 31 11 6 33 4 33 4 O 34 33 33 5 17 Mean 13.0 16.1 ~33.8 S. D. 10.8 15.8 43 Table VII, Page 43, shows the individual gains, mean gains, and the difference in gains between indi- viduals in the low thirds of the experimental and control groups, on the basis of intelligence quotents. The mean gain of the seventeen experimental pupils was 16.4. The range in gains was from ~13 to 39 or 51 points. Only two pupils show negative gains. The mean gain of the seventeen control pupils was 5. Their range in gains was from ~31 to 49 or 80 points as compared to range of 51 in the experi- mental group. Four control pupils made negative gains in achievement between the pre—test and final test. The greater range of gains in the seventeen con~ trol pupils gives so large a standard deviation that there are 1.4 chances in a hundred that there may not be a real difference. There are, however, 98.6 chances in a hundred that the student activity procedure would again be the superior method with the low third of any similar group. The standard deviation of 16.3 in the gains of the control pupils indicates an extremely wide dispersion of gains. The standard deviation of the gains of the experimental pupils shows more consistent gains. 43 TABLE VII MEAN GAINS, STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF THE LOW THIRDS OF THE EXPERIMENTAL AND CONTROL GROUPS ON THE BASIS OF INTELLIGENCE QUOTIENTS 3133’1123381 Lsanxagid Pupil Gain Pupil Gain Diff. 35 39 34 38 l 36 34 35 8 36 37 5 36 6 -l 38 18 37 ~17 35 39 34 38 49 -35 4O 19 39 5 14 41 6 40 13 ~7 43 36 41 ~31 57 43 10 43 6 4 44 1 43 4 -3 45 13 44 3 10 46 36 45 13 13 47 ~13 46 3 ~15 48 39 47 5 34 49 6 48 13 ~7 50 38 49 ~6 44 53 ~3 50 ~16 14 Mean 16.4 5.1 11.5 S. D. 14.0 16.3 TABLE VIII 44 MEAN GAINS, STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF EACH THIRD OF EXPERIMENTAL AND CONTROL GROUPS, EQUATED ON THE BASIS OF INTELLIGENCE QUOTIENTS Experimental Control 'Third N Mean S.D. Mean S.D. Diff. Chances Gain Gain in 100 High 17 16.5 13.3 13.3 18.4 3.3 73 Middle 16 13.0 10.8 16.1 15.8 -3.1' 73 Low 17 16.4 14.0 5.1 16.3 11.3 96.6 Total 50 15.3 13.7 11.3 18.0 4.0 88 Table VIII shows that there are 73 chances in 100 that the pupil activity procedure would again be the more effective teaching procedure under similar conditions with the high third. The conventional method would have 73 chances in a hundred of again being superior to the middle third. These do not constitute significant dif~ ferences. However, there are 98.6 chances in 100 that the pupil activity procedure would again produce superior gains in the low third. This does constitute significant‘ difference. l"Throughout this study, statistical significance was assumed if the probabilities were at least 95 out of 100 that the differences could not have occurred by chance. 45 TABLE IX MEAN GAINS, STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF EACH THIRD OF EXPERIMENTAL AND CONTROL GROUPS, EQUATED ON THE BASIS OF AGE ExperiHentalv_ Control Mean Mean Chances N Gain S.D. Gain S.DL Diff. ’ in 100 Oldest third 17 14.4 13.3 3.5 18.4 10.9 98 Middle third 16 11.6 13.8 10.9 16.4 0.7 50 Youngest third 17 15.5 11.3 16.5 16.9 ~l.0 58 Total 15 15.3 13.7 11.3 18.0 4.0 88 Table IX shows that the oldest third of the experimental group made a mean gain of 14.4 points and that the corree~ ponding third of the control group gained only 3.5 points. The difference is 10.9 points. There are 98 chances in 100 that the pupil activity procedure would again produce super- ior gains with similar groups and under similar conditions. The pupil activity procedure then, is significantly superior with the oldest third. There is practically no difference between the gains of the middle thirds. The youngest third of the control group made mean gains slightly superior to those of the experimental group. A comparison of the standard deviations of these means, however, shows that the experimental pupils made the most consistent gains. The difference is not significant. 46 TABLE I MEAN GAINS, STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF EACH THIRD OF THE EXPERIMENTAL AND CONTROL GROUPS, EQUATED ON THE BASIS OF HONOR POINT RATIO Experimental Control Mean Mean Chances ___ N Gain 8. D. ”Gain S.p.__ Diff. in 100 High third - 17 17.8 18.0 17.6 18.0 0.3 50 Middle third 16 10.1 14.4 9.5 13.6 0.6 56 Low _ - third 17 15.8 11.6 7.1 18.8 8.7 95 Total 50 15.3 13.7 11.3 18.0 4.0 88 Table I shows that the high third of the experimental group and the high third of the control group on the basis of honor point ratio made almost identical mean gains. There is, then, no superiority of one method over the other ‘ in its effect upon these pupils in the high third in achievement as indicated by teachers' marks. The data show that between the middle thirds on this basis there was no real difference. The low third of the experimental group made a mean gain of 15.8 while the low third of the control group gained only 7.1. The high standard deviation, 18.8, of the mean gain of the low third of the control group is indicative of a notable inconsistency in gains. The superiority of the pupil activity procedure with the 1 low third is significant. 47 TABLE XI MEAN GAINS, STANDARD DEVIATION OF THE GAINS, AND DIFFERENCE IN GAINS OF EACH THIRD OF THE EXPERIMENTAL AND CONTROL GROUPS, EQUATED ON THE BASIS OF PRES-TEST SCORES Experimental Control Mean 'Hgan Chances .. N Gagn S.D. Gain S.D. Diff. in 100 Hi h . th rd 17 14.6 13.3 14.4 17.6 0.3 50 Middle third 16 17.8 13.3 15.3 31.6 3.6 67 Low third 17 15.5 13.3 9.5 14.0 6.0 90 Total 50 15.3 13.7 11.3 18.0 4.0 88 Table XI shows the results of the experiment when the thirds of each group are matched on the basis of the pre-test scores. There is no difference in gains between “ the high thirds. Between the middle thirds the difference in mean gains, 3.6 points, favors the experimental group. The difference is not significant. The mean gain of the experimental low third exceeded that of the control low third by 6.0 points. This difference closely approaches. significance. Comparison of the means and standard deviations of the low thirds shows that the experimental pupils made the more consistent gains. . 48 TABLE XII PRE-TEST AND FINAL SCORES, GAINS, AND DIFFERENCE IN GAINS OF THE THENTY~EIGHTAMATCHED PAIRS Experimental Control _gc __ ngil Pre-testfiFinEI Gain Pupil Pre-tast Final Gain Diff, ‘3 105 138 33 4 131 135 4 39 5 83 ' 70 ~13 6 81 94 13 ~35 6 75 100 35 3 97 99 3 33 7 47 43 ~4 8 64 87 33 ~37 9 99 138 39 9 68 71 4 35 11 DO 103 13 13 87 131 34 - 31 13 56 89 33 11 79 75 4 37 14 61 81 30 13 107 117 10 10 15 53 55 3 14 63 60 ~3 5 16 75 98 33 16 53 37 ~16 39 18 77 81 4 18 77 90 13 ~9 19 47 73 39 15 60 50 ~10 39 30 35 49 34 19 44 44 O 34 33 95 89 ~6 33 83 114 -31 ~39 33 58 69 11 33 61 67 6 5 34 89 105 16 31 74 90 16 0 36 77 86 9 34 68 96 38 -19 37 45 71 36 30 61 73 ll 15 39 54 73 18 38 33 46 14 4 38 39 56 17 33 55 6O 5 13 30 66 51 ~5 31 53 63 11 ~16 31 100 130 30 g 39 44 53 8 34 33 50 54 4 37 69 53 ~17 31 34 53 75 33 4O 35 38 13 9 38 68 103 34 39 37 43 5 39 38 44 63 18 36 63 69 6 13 39 36 60 34 43 ‘ 45 56 9 15 48 30 69 39 48 69 83 13 36 Mean Gain 17.00 10.6 10.4 s, D. A 13.96 11.8 Table XII shows the pre—test score, final score, and gain of each individual in the twenty-eight matched pairs. Twenty-four experimental individuals show gains and four show losses. Their mean gain was 17. Twenty-two control individuals show gains and six show losses. Their mean gain was 10.6. Table III, Page 37, shows the data by which these pairs were matched. 49 TABLE XIII MEAN GAINS, STANDARD DEVIATION OF THE GAINS,AND DIFFERENCE IN GAINS OF EACH THIRD OF EXPERIMENTAL AND CONTROL MATCHED PAIRS Experimental Control Mean Mean Chances N Gain 8. D. Gain 8. D. Diff. in 100 High third 9 15.5 8.4 9.3 11.3 6.3 90 Middle third 10 15.4 8.4 9.3 14.0 6.1 88 Low third 9 30.3 13.0 5.8 9.3 14.5 99.7 Total 38 17.0 13.9 10.6 11.8 6.4 98 Table XIII shows the outcome of this investigation with the twenty-eight matched pairs. These results serve as a valuable check on the results obtained with the equated thirds of the total groups, since these twenty-eight pairs closely approached identity in the three criteria'on which they were matched. The pupil activity procedure produced " greater gains in each third of the matched pairs. In the high and middle thirds the differenge is enough to indicate a definite superiority in the pupil activity procedure. Between the low thirds there is a significant difference in< gains in favor of the experimental pupils. This outcome coincides with the results obtained with the low thirds of the total groups on each basis of comparison. * Table III, Page 37, shows the likeness of the two pupils in each matched pair. 50 CHAPTER V SUMMARY, CONCLUSION; AND SUGGESTIONS FOR FURTHER STUDY low The data presented in this study show that when United States history was taught to one group by a pupil activity procedure and to an equivalent group by the conventional procedure: 1. The high third of the experimental group,in mental ability who were taught by the pupil activity procedure, made sur~ perior but not significantly superior gains over the high third of the equivalent control group, taught by the con- ventional method. 3. The middle third of the experimental group,in mental ability;who were taught by the pupil activity procedure . made slightly inferior gains to those of the middle third of the equivalent control group who were taught by the conventional procedure. 3. The low third of the experimental group,in mental ability who were taught by the pupil activity procedure made significantly superior gains over the low third of the equivalent control group who were taught by the conven- tional procedure. 4. The oldest third of the experimental group made signif- icently superior gains over the oldest third of the con- trol group who were taught by the conventional procedure. 5. The middle thirds of the experimental and control groups chronologically made approximately equal gains. 6. The youngest third of the experimental group made slightly inferior gains to those of the youngest third of the con- trol group. 7. The high third of the experimental group,in previous academic achievement, made gains almost identical to the ' gains of the high third of the control group in previous academic achievement. 8. The middle third of the experimental group,in previous academic achievement,made slightly superior but not significant gains in comparison with the gains of the control group. 9. The low third of the experimental group in previous academic achievement, showed a significant superiority 51 in gains over the low third of the control group in previous academic achievement. 10. The high thirds of the experimental and control groups on the basis of pro-test scores made equal gains. 11. The middle third of the experimental group on the basis of pro-test scores made slightly superior gains over the control group on the basis of pre—test scores. 13. The low third of the experimental group on the basis of pre-test scores made definitely superior but not signif- icant gains in comparison with the gains of the low third of the control group, on the basis of pro—test scores. 13. Each third of the experimental matched pupils showed de- finitely superior gains over the equivalent third of the Control pupils. The gains of the low experimental third were statistically significant. B. Conclusion. Since the high and middle thirds of the experimental group were not consistently hindered by the pupil activity procedure,and since the low third made significantly superior gains, the implication follows that greater use of the pupil activity procedure in the teaching of United States history is justifiable. 0. Limitations of the Conclusion 1. It is impossible to determine the specific factor which produced the significant gains of the low third of the experimental group. Was it because each social problem was more thoroughly discussed, because greater individual re- sponsibility for participation had been established or because the procedure set up more favorable conditions for first hand experience in dealing with the problems analyzed? 3. The results prove only that the low third made superior 53 gains in the outcomes measured by this particular test. Since the high third of the experimental group contributed much more in the discussions, it is possible that they achieved more in some less tangible outcomes which this test does not measure. 3. The success of the pupil activity procedure does not prove that the low third of a similar group should be segregated and then taught by the pupil activity procedure. It is probable since the high third participated observably more than the low third, that this reaction among the more able pupils of the high third made a definite contribution to the achievement of the less able ones in the low third. D. Sggggstions for Further Study The following allied problems have arisen as a result of this investigation and are therefore suggested for further study: 1. Factors which account for occasional pupil loss in subject matter during a semester of enrollment in the subject. 3. The effect of using a pupil activity procedure in teach- ing United States history to the low third segregated from all others of the group. 3. The effect of using a pupil activity procedure in teach- ing United States history to the high third segregated from all others of the group. 4. The correlation between knowledge of the social studies and social conduct. 5. Means of measuring outcomes other than factual knowledge resulting from the social studies. 53 SELECTED BIBLIOGRAPHY (Books and Bulletins) Briggs, T. H. ImprovinggInstructipp. New York: The Macmillan Company. 1938. Pp. 587. Butterfield, A. W. “A Study of the Relative Effectiveness of Two Methods of Teaching History.‘l Unpublished Master's thesis, University of Michigan, 1935. Pp. 58. Condon, Mary L. "Classical Backgrounds of the Activity Movement.‘ Unpublished Master's thesis, University of Washington, 1933. Pp. 308. Dewey, John. Experience and Edupation. New York: The Mac- millan Company. 1938. Pp. 116. . Douglass, Harl Roy. The ExperimentalpComparison of the Re- lative Effectiveness of Two Sequence in Supervised Study. University of Oregon, 1937, Pp. 173-318. Fahrney, R.R.A ACompgrison of the Effectiveness foppLearnipg of the Lecture-Quiz Techhigpe with the Classroom-Discussion TQZhn:gpe in a Freshman Sophomore Course in American History. Bureau of Research, Iowa State Teachers College. Cedar Falls. Iowa. Research Report, No. 36. March, 1939. Pp. 7. Franzen, Carl G.F. A Comparison of the Results Made on Certain Standardized Tests by Pupils in the Bloomington High School Who Were Taught in Classes of the Same Grade by University Student Teachers and by Regular High School Teachers. Bulle- tin of the School of Education Volume II, No. 4. Bloomington, Indiana: Indiana University. March, 1936. Pp. 31. Garrett, Henry E. Statistics in Psychology and Education. New York: Longmans, Green and Company. 1937. Pp.493. Good, 0. 8., Barr, A. S., and Scates, D. E. The Methodology of Educational Researph. New York: D. Appleton-Century Company. 1936. Pp. 883. Halvorsen, N. 0. An Experiment Comparing_the Effectiveness of Two Methods of Indicating_Errors In Students Themes. Re- search Report No. 35. Cedar Falls, Iowa: Iowa State Teachers College. 1939, Pp. 35. , Kelley, T. L. and Krey, A. 0., Tests and Measurements in the Social Sciences, Part IV,Report of the Commission on the Social Studies. American Historical Association, New York: Charles Scribner's Sons. 1934. Pp. 635. Kilpatrick, W. H. Education for a chan nging Civilization. New York: The Macmillan Company. 1938. Pp. 143. 54 Monroe, I. 8., and Engelbert Max D. Egperimental Research In Education. Bureau of Educational Research, Bulletin 48, Urbana, Ill: University of Illinois. 1930. Pp. 105. Moesman, L. C. The Actiygty Concept. New York: The Macmillan Company. 1938. Pp.197. National Education Association, Improving‘Social Studies Instruction. Research Bulletin of the National Education Association, Vol. XV, No. 5. Washington, D.C: November, 1937. Pp. 187-258. National Education Association, Modern Socialzand Educational Trends. Research Bulletin of the National Education Association, Vol. XII, No. 5. Washington, D.C: November, 1934, Pp. 243-287. National Education Association, Teachers and Cooperation. Yearbook on Cooperation, National Education Association, Department of Supervisors and Directors of Instruction, Committee on COOperation. Detroit, Michigan: November, 1937. Pp. 81. - Ryan, Teresa, “An Experiment to Study the Effect of Two Methods of Instructions in Attempting to give and Under- standing of Geometric Concepts in Two 8A Classes." Un- published Master's thesis, University of Detroit, 1933. Pp. 64. lrightstone, J. Wayne. Appraisal g;_Experimenta1 High gphool Practices. New York: Bureau of Publications, Teachers College, Columbia University. 1936. Pp.194. Yurkewitch, J. T. "A Controlled Experiment on the Con- tract-Discuesion Method of Teaching Versus The Recita- tion Method in United States History." Unpublished gaster's thesis. Pennsylvania State College. 1933. p. 32. 55 SELECTED BIBLIOGRAPHY (Periodicals) Butler, C.H. “Mastery of Certain Mathematical Concepts by Pupils at the Junior High School Level." The ‘ Mathematics Teacher, 25:120, March, 1932. Courtis, S.A. "Criteria For Determining the Equality of ° Groups.“ School and Society, 35:874-8, June 25, 1932. Courtis, S. A. "Techniques of Cooperation." Educational Method, 7:349-50, April, 1938. Farnsworth, P.R. "Concerning So-Called Group Effects." Pedagogical Seminar, 35: 387-94, September, 1928. Johnson, P.O. "A Comparison of the Lecture-Demonstration, Group-Laboratory Experimentation and Individual-Laboratory Experimentation Methods of Teaching High School BiolOgy.' Jgurnal of Educational Research, 18:103-111, September, 1 28. a ”' Melby, E.O., and Lien, Agnes.,'A Practicable Technique for Determining the Relative Effectiveness of Different Methods of Teaching.’I Journal of Educational Research, 19:255-264, April, 1929. Murra, Wilbur F. "Research in the Teaching of the Social Studies", Harvard Educational Review, 7:293-95, May, 1937. Noll, Victor H. "The Effect of Written Tests Upon Achieve- ment in College Classes: An Experiment and A Summary of Evidence." Journal of EducayipnalyResearch, 32:345-58, January, 1939. Powers, Samuel R. "PsychOIOgy and Methods in High School and College." Review of_Educationa;_Research, 8:60-66, February, 1938. Watson, G.B. "Do Groups Think More Efficiently than Individ- uals?" Journgl.of Abnormal and Social Psychology, Weston, A. B., and English, 3.8. "The Influence of The Group on Psychological Test Scores.‘ American Journal of Ps1- cholOgy, 37:600-602, March, 1926. lrightstone, J. Wayne. “Appraising Newer Practices in Teaching Social Studies." School Review, 42:688-93, November, 1934. APPENDIX COLUMBIA RESEARCH BUREAU AMERICAN HISTORY TEST By HARRY J. CARMAN Associate Professor of History, Columbia University THOMAS N. BARROWS History and Civics The Lincoln School of Teachers College, Columbia University and BEN D. WOOD Associate Professor and Director Bureau of Collegiate Educational Research Columbia College, Columbia University TEST: FORM A For High Schools and Colleges Do not open this booklet, or turn it over, until you are told to do so. Fill these blanks, giving your name, age, etc. Write plainly. Name ....................................................................... (First name, initial, and last name) Age last birthday ............ years. Class ..................................... Date ................... 19 ......... School or college ............................................................. City ................................... Teacher .............................. How many school years have you studied American history in high school? (Indi- cate by drawing a circle around the right number.) .3, 1, l—lf, 2. GENERAL DIRECTIONS. This examination consists of four parts and requires ninety minutes of working time. The directions for each part are printed at the beginning of the part. Read them carefully and proceed at once to answer the questions. There is a time limit for each part. You are not expected to answer all the questions in any part before the time is up, but if you should, go back and make sure that your answers are right. You may then work further on any previous part or go ahead to the next part. If you have not finished a part when the time is up, stop work on that part and proceed at once to the next. No questions may be asked after the examination has begun. You need two sharpened lead pencils. Do not 0 pm the booklet yet. PART l SCORE I II III IV Total Rating Classification .— l’ublished by World Book Company, Yonkers-on—Hudson, New York, and Chicago, Illinois Copyright 1926 by World Book Company. Copyright in Great Britain All rights reserved. cnw : CRB-\HTI A-9 PD'YTFU IN ".9. k E” This test is copyrighted. The reproduction of any part of it by mimeograph, hectograph, or in any other way, whether the reproductions are sold or are furnished free for use, is a violation of the copyright law. 57 American history: A PART I DIRECTIONS. Read each of the following statements very carefully. If a state- ment is true, place a plus (+) in the parentheses following it; if it is false, place a zero (0) in the parentheses following it. If you are not sure whether a statement is true or false, leave the parentheses blank. Do not guess. Twenty minutes. SAMPLES. a. George Washington was the first President of the United States ........... ( + ) b. The panic of 1857 had no effect upon the economic life of the South. . . . . . .( 0 ) 1. All those who settled in Massachusetts previous to the Revolutionary War were required by law to become members of the Congregational Church .( ) 2. The English colonists in North America enjoyed more complete self-govern- ment than did the French colonists ................................. ( ) 3. Nothing resembling the feudal system of landholding existed in colonial America ......................................................... ( ) 4. Absolute liberty of conscience in religious matters was advocated by Cotton Mather .......................................................... ( ) 5. Class distinctions based on family and wealth existed in colonial America as well as in the Old World ........................................... ( ) 6. Throughout the colonial period the Americans depended largely on England for their tools and clothing ......................................... ( ) 7. The greatest barrier to the poor who wanted to migrate to America was the cost of the long sea voyage ......................................... ( ) 8. Peter Zenger was a New York publisher who was arrested for criticizing gov- ernment officials .................................................. ( ) 9. The Albany Plan of Union was a scheme for uniting the six Iroquois nations of central New York .............................................. ( ) 10. One of the principal industries of the Carolina colonists was cod fishing. . . .( ) 11. Some of the British colonies in North America never had colonial assemblies ( ) 12. In colonial Virginia most churchgoers belonged to the Church of England. . .( ) 13. The “Mayflower Compact” was the first written document to be adopted in America to provide a definite framework of government ............. ( ) 14. Indentured servitude was a system whereby ship owners and other persons of means furnished the passage money to immigrants in return for their promise to work for a term of years to repay the sum advanced ......... ( ) 15. The majority of the English colonies in America had compulsory education ' laws ............................................................ ( ) 16. By 1680 Philadelphia was the largest city in English colonial America ...... ( ) 17. Mrs. Harriet Beecher Stowe wrote a novel picturing the worst features of the slavery system ................................................ (_ ) 18. The Webster-Hayne debates set forth the opposing doctrines of nationalism and states’ rights ................................................. ( ) 19. The Crittenden Compromise was a proposal designed to prevent the P illman strike of 1894. ................................................... ( 20. The basic cause of the Mexican War was the discovery of gold in California. ( ) I 2] 5 7a ‘- . -. , -_» 4 - 4v- . .nv— .w‘