Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Cognitive style and children's performance on measures of elementary science competencies Sieben, Gerald Alexander 1971

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
UBC_1971_A8 S54.pdf [ 8.54MB ]
Metadata
JSON: 1.0055292.json
JSON-LD: 1.0055292+ld.json
RDF/XML (Pretty): 1.0055292.xml
RDF/JSON: 1.0055292+rdf.json
Turtle: 1.0055292+rdf-turtle.txt
N-Triples: 1.0055292+rdf-ntriples.txt
Original Record: 1.0055292 +original-record.json
Full Text
1.0055292.txt
Citation
1.0055292.ris

Full Text

COGNITIVE STYLE AND CHILDREN'S PERFORMANCE ON MEASURES OF ELEMENTARY SCIENCE COMPETENCIES by GERALD ALEXANDER SIEBEN B.Ed., University of British Columbia, 1966 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in the Department of EDUCATION We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA June, 1971 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Faculty of Education The University of British Columbia, Vancouver 8, B.C. June, 1971 ABSTRACT The purpose of this exploratory study was to determine the effect of Witkin's construct of cognitive style on children's performance on salient elementary science competencies. These competencies involved the ability to use science processes and the acquisition of specific attitudes. During the development of the study (see Appendix A), it was first necessary to determine the measurable objectives of the Elementary Science Study (E.S.S.). The Test of Science Processes was used to measure those E.S.S. objectives which pertained to science processes. In order to measure the attitudinal objective of the E.S.S. pro gramme, the author developed four attitude scales, utiliz ing proper attitude measuring techniques. The four scales measured children's attitudes towards the following beliefs: children will feel that "Messing about in Science" is fun (Fun Scale); children will follow-up phenomena encountered during E..S.S. experiences (Pursue Scale); children will impose a structure on their play to find out more (Structure i i Scale); children will themselves initiate their own inves tigations (Independent Investigation Scale). The develop ment of these scales is reported intact in Appendix B. Good reliability and factoral validity weve established for these scales. It was hoped that the four attitude scales would prove to be useful tools for elementary science educators. A natural experiment in a small city school district was utilized to determine the effect of years of E.S.S. experience, the effect of Witkin's construct of cognitive style, and the interaction effect of years of experience with cognitive style -- on children's per formance on measures of elementary science competencies. Utilizing a three by three factorial design, the test scores of 184 grade seven pupils were compared. The independent levelling variable used to determine cognitive style was based upon performance on the Children's Embedded Figures Test (C.E.F.T.). Years of E.S.S. instruction (one year, two years and three years) comprised the independent blocking variable. Groups were compared on fourteen dependent variables ( nine process variables and five attitudinal variables). Hotellings T2 statistic was com puted prior to analysis of variance in order to determine if the global group (C.E.F.T. score 0-15) would achieve i i i significantly lower scores than the analytical group (C.E.F.T. score 20-25) on the sets of elementary science competencies (processes and attitudes). The predicted inferior performance of the global group was confirmed on the set of attitudinal dependent variables and on the set of dependent variables concerning processes. The predicted effect of superior performance of students who had received more E.S.S. experience than other students was not generally supported by the statis tical tests. The predicted interaction effect was not generally significant either, although their appeared to be a trend which might indicate that the global group did less well when this group had more and more E.S.S. experience. Limitations of the cross-sectional design, however, made it difficult to come to any firm conclusions regarding the interaction effect and the effect of years of instruction. Analyses of variance confirmed the findings that the children with a more global cognitive style achieved significantly lower scores on elementary science competencies than children with more analytical cognitive styles. - Based on these findings, the implications of the construct of cognitive style on elementary science i v education was discussed in terms of methodological reform and curricular reform. Finally, a plan for further research was proposed. v TABLE OF CONTENTS Chapter Page I. INTRODUCTION 1 1.0 Importance of the Study 1 1.1 General Statement of the Problem ...... 3 a) A Definition of the Construct of Cognitive Style 3 b) The Nature of this Study 7 II. A REVIEW OF THE LITERATURE ON COGNITIVE STYLE. ... 9 2.0 The Classical Division between Cognitive Abilities and Perceptual Style is Unwarranted 9 2.1 Cognitive Style is Related to Dependence Upon Others 11 2.2 Studies Reveal the Ontogeny of 13 Differentiation 2.3 Discriminating Attributes of Cognitive Styles are Similar to the Attributes of E.S.S. Activities 17 III. PROCEDURES 22 3.0 A Natural Experiment Existed in a Small-City School District 22 3.1 Individual and Group Tests were Administered to the Subjects 24 vi Chapter Page 3.2 The Design of this Study can be Categorized as a Factorial Model 26 3.3 There was no Significant I.Q. Bias in Columns I, II, and III 28 3.4 Fourteen Dependent Variables were Tested in the Design 29 3.5 The Null Hypotheses 31 3.6 The Alternative Hypotheses 32 IV. THE STATISTICAL ANALYSES . 34 4.6 Raw Data was Analyzed by the Computing Facilities of the University of British Columbia .  34 4.1 Multivariate and Univariate Tests were Utilized for the Statistical Test of Hypothesis I 35 a) Statistical Tests and Their Results . 35 b) Summary of the Statistical Tests of Hn 40 UI 4.2 Multivariate and Univariate Tests were Utilized for the Statistical Tests of Hypothesis II 40 a) Statistical Tests 4b) Summary of Statistical Tests of UII 4.3 The Statistical Tests of Hypotheses III, IV, and V 45 a) Analyses of Variance Tests of Hypotheses III, IV, and V 45 b) Summary of Tests of Hypotheses III,IV, and V 59 vi i Chapter Page b b b V. CONCLUSIONS AND SUMMARY. . 67 5.1 This Study Developed from a Need which Arose in the Classroom 67 5.2 Limitations of the Study 68 5.3 Conclusions and Recommendations . 69 a) The Effect of Years of E.S.S. Instruction 6b) The Interaction of Cognitive Style and Years of E.S.S. Instruction 70 c) The Effect of Cognitive Style on Performance on the Test of Science Processes 71 d) The Effect of Cognitive Style on Performance on the Attitude Scales. . 71 e) General Conclusions and Recommendations 71 e-1) The Modification of E.S.S. Methodology 72 e-2) The Modification of the Way Curriculum is Used ... 75 5.4 Implications for Further Research 77 REFERENCES 79 v i i i -1) The Main Effect for Years of E.S.S. Instruction (H0 ) 60 u 111 -2) The Main Effect for Cog nitive Style (Hn ) 60 UIV -3) The Interaction Effect of Cognitive Style and Years of E.S.S. Experience (HQ ) . 61 Page APPENDIX A THE SELECTION OF PERFORMANCE CRITERIA 87 APPENDIX B THE DEVELOPMENT OF FOUR ATTITUDE SCALES TO MEASURE CHILDREN'S ATTITUDES TOWARDS THE AFFECTIVE OBJECTIVES OF THE ELEMENTARY SCIENCE STUDY 95 APPENDIX C ITEM AND TEST ANALYSES FOR C.E.F.T. AND THE TEST OF SCIENCE PROCESSES 152 APPENDIX D RAW DATA -- IDENTIFIED ACCORDING TO FORMAT 164 i x LIST OF TABLES Table Page 1. A comparison Showing the Similarity Between E.S.S. Objectives and the Descriminating Attributes of Global versus Analytical Functioning 18 2. t-Tests Comparing the Global versus the Analytical Group on Each of the Affective Variables 36 3. A Multivariate Comparison of Global versus Analytical Groups on the Affective Measures of the Elementary Science Study 38 4. A Comparison of Individual t-Tests of the Total Group versus the t-Tests of Each Sex Separately on the Affective Measures - 39 5. t-Tests Comparing the Global versus the Analytical Groups on Each of the Processes. 41 6. A Comparison of Individual t-Tests of the Total Group versus t-Tests of Each Sex Separately on the Cognitive Measures (Processes) 42 7. A Multivariate Comparison Between the Global Group and the Analytical Group on the Cognitive Elementary Science Competencies (Processes) 43 8a Total Groups Analyses of Variance Tables 47 8b Boys' Analyses of Variance Tables ..... 51 8c Girls' Analyses of Variance Tables 55 9. Summary of Rejected and Accepted Hypotheses (Hn , Hn » and H0 ) . 59 u111 UIV UV 1 Factor Loadings on the Four Attitude Scales 11.6 x Table Page B2 Alpha Coefficients, Means and Standard Deviations of the Four Scales 119 B3 Intercorrelation of the Four Attitude Domains - by Sex 121 CI Item Analysis of the Childrens ' Embedded Figures Test ....... 152 C2 Item Analyses for the Subtests of the Tests of Science Processes 154 xi LIST OF FIGURES Figure Page I. Sample Item from the Childrens' Embedded Figures Test 5 II. Stability of Cognitive Style over Time 14 III. The Design of the Study 27 IV. All Significant Main Effects for Years of E.S.S. Experience 62 V. Statistically Significant Interaction Effects 63 VI. Score Trends for Total Attitudes and Total Processes Based Upon Cell Means from Analyses of Variance 64 xi i ACKNOWLEDGEMENTS My thanks are extended to my supervisor, Dr. G.H. Cannon for his assistance and encouragement. Thanks are also due to Dr. L.L. Walters for his advice and support. I am grateful to Dr. Stephen Foster for his careful supervision of the preparation of the attitude scales. I also wish to express my gratitude to the teachers and students of School District #15 (Penticton) for their cooperation. x i i i CHAPTER I INTRODUCTION 1 .0 Importance of the Study This exploratory investigation focussed on an impor tant issue in curriculum development which was raised more than a decade ago: Shall educators search for the one best curriculum for all children, or shall educators seek to discover which curricula are best suited to children manifesting particular characteristics (Cronbach 1957, 1967; Cronbach and Snow, 1969). Statements such as the following still persist in major elementary science re visions and typify the unilateral approach to curriculum: Every child learns best when real things such as batteries, bulbs, bones and blocks are available for him to use, as tools for his inquiry. As a result, the pupils ex perience with real things will lead him to search for supporting vesources. . . .They [the pupils] need to have a fvee, unstruc tured period of time to feel, to smell, to listen, etc. . . .How the pupil learns things is more important than the things he learns {Elementary Science, Province of British Columbia, 1969, p. 12). This statement, though well-intentioned, is questionable and perhaps even contradictory. It is quite possible that when one considers how the pupil learns, it may be that 1 2 the individual pupil under consideration does not learn best within the structure of the methodology and materials that are prescribed for every child. For example, some children appear to lack direction and to experience great difficulty in the above mentioned elementary school science program which stresses individual, self-initiated experi mentati on. The writer does not wish to imply that he is con demning this particular curriculum either. Global con demnation is probably as inappropriate as complete endorsement for all children. Rather, he wishes to emphasize that a curriculum must not be regarded by those in positions of influence as some panacea—some magical elixir which is equally suitable for all children. Clearly, research is called for which attempts to uncover more about the individual differences of children who are learning within the particular framework of a curriculum. The implications of these individual differences and individual styles of dealing with the world should be matched with teaching strategies and curriculum materials. To this effect, this exploratory study attempted to analyze individual differences in terms of the cognitive styles of children who have been experiencing the learning stra tegies and materials of the Elementary Science Study. 3 During the development of this research, it was necessary to engage in a parallel study to determine and to develop criterion measures for Elementary Science Study experiences. This parallel research study is reported intact in Appendix A and Appendix B. It is hoped that the attitude scales which were developed by the author will be useful tools for elementary science teachers and for curriculum researchers. 1 .1 General Statement of the Problem a) A Definition of the Construct of Cognitive Style Witkin (1962) and his associates represent a school of psychology which is called "differential psychology." At the basis of this theory is a concept called field dependence which can be defined as the lack of ability to disembed or to decontextualize a stimulus figure from an irrelevant but organized stimulus background. There is a battery of tests which are used to determine this perceptual ability. For example, in the Rod and Frame Test, a simple square luminous frame provides a field which glows in a semi-darkened room. This frame is pivoted a.t its center so that a luminous rod may be tilted independently of the frame, clockwise or counterclockwise. The subject is asked to adjust the rod so that its position corresponds to the position of a hypothetical vertical 4 standard ("straight with a flagpole"). Meanwhile the experimenter adjusts the frame to various tilted positions. Some children are able to place the rod in a vertical position without being confused by the surrounding frame. They are able to perceive part of the field as discrete from the dominant part of the visual field. Others seem to rely on the misleading clues of the back ground to dominate part of the visual field and con sequently are referred to as field-dependent, while the former subjects are called field-independent. Subjects at either end of this continuum show a marked degree of consistency in other performance tests which also deal with disembedding one part of a field from the remainder. Figure 1 shows a pair of figures from the Children's Embedded Figures Test (Karp and Konstadt, 1969). The subject is asked to find the simple figure on the left within the complex figure on the right. Although this task does not involve perception of the upright, there is a basic commonality between the Rod and Frame Test and the Children's Embedded Figures Test. In the latter test, a figure is embedded within another and the experimenter attempts to determine how much the dominant whole of the visual field is inhibiting his subject's perception of a part that is embedded within it. Subjects Figure 1 Sample item from the Children's Embedded Figures Test. are trained to find two simple shapes within a more complex background. Then the simple shapes are removed from sight and they are not shown to the subject again unless he requests it; never, however, are both simple shapes and 6 complex shapes presented simultaneously (except in training sessions). As each of the complex shapes is presented, the subject is asked to identify it and name it. In this way, the experimenter is assured that the subject has "taken-in" the whole figure. The subject is then asked to find the simple figure which is embedded within the more complex,mis1eading figure. The number of correct first responses provides an index of field-independence. By means of perceptual indices such as these, Witkin and his associates have demonstrated that field dependent children differ from field independent children on a vast number of different criteria such as: dependence on others, the ability to structure ambiguous stimuli, the ability to see alternate uses for the familiar, the ability to resist persuasion by authority, the ability to be logical in the face of evidence that is contrary to the known attitudes of the subjects, and the ability to adopt analytical procedures when dealing with their environment.^ Witkin summarized these findings and used the term"cognitive style" as it will be used in this study. Evi dence Chapter II which sty1e. for this generalization is included reviews the literature of cognitive 7 To continue with, our main story, the studies cited and the numerous other ones as well, have made it quite clear that the style of functioning we first picked up in perception, where we were dealing with an immediately present stimulus con figuration, manifests itself as well in intellec tual activity, where we are dealing with symbolic functioning. As noted at the outset, we use the designation "cognitive style" to refer to this kind of characteristic, self-consistent way of functioning that an individual shows across per ceptual and intellectual (i.e., cognitive) activities. The particular cognitive style we have been discussing, of which field dependence of field independence is the perceptual component, may be described most broadly as follows: at one extreme there is a tendency for experience to be diffuse and global; the organization of a field as a whole dictates the way in which its parts are experienced. At the other extreme the tendency is for experience to be delineated and structured; parts of a field are experienced as discrete and the field as a whole as structured. To these opposite poles of the cognitive style we have applied the labels "global" and "articu lated. " It should be emphasized that there is no implication here that the world is populated by two kinds of human beings. Scores for any large group of people on tests of this cognitive style show a continuous distribution and depending on which sides of the mean or average a person 's score falls, we say his cognition is more articu lated or more global. It is clear from the evidence on hand that a tendency toward a more global or more articulated mode of functioning pervades a child's cognitive activity; and it may be added, on the basis of other evidence, that a given style of cognitive functioning is a stable characteristic of a child even over very long periods of time (Witkin, 1969, p. 206). b) The Nature of this Study Utilizing Witkins's construct of cognitive style, this study attempted to demonstrate that the children who 8 could be categorized as being at the extreme ends of the differentiation continuum would achieve significantly different results on elementary science competencies. It was hypothesized that the global group would achieve significantly lower scores than the analytical group on those measures which attempt to tap some of the essential objectives of the Elementary Science Study. Moreover, it was expected that the differences between the global (field dependent group) and the analytical (field inde pendent group) would increase as children had more and more experience within the E.S.S. programme. CHAPTER II A REVIEW OF THE LITERATURE ON COGNITIVE STYLE 2.0 The Classical Division between Cognitive Abilities  and Perceptual Styles is Unwarranted A great number of studies have shown that there is a relationship between field dependence and the kinds of attributes which the author believes to be very important in The Elementary Science Study (E.S.S.). There is con siderable evidence that the classical division between cognitive abilities and perceptual style is perhaps arbitrary and unwarranted. Differences in perception as measured by the Embedded Figures Test (adult version) have been found to be related to differences in cognitive functioning and in particular, to differences in analytical functi on ing. Factor analysis has revealed that field dependence, Guilford's construct of adaptive flexibility, Phillip's construct of spatial decontextualization , Dunkcer's notion of functional fixedness, Thurstone's "flexibility of closure," and the construct of "perceptional organization" 9 10 on the Weschler Intelligence Scale, all involve an ability to disembed (Goodenough and Karp, 1961; Witkin, 1962). Some of the research utilizing Einstellung tests revealed that on the extinction problem (not the critical problems), performance was related to field independence (Guetzkow, 1951, Goodman, 1960). Linton (1952) found field dependence significantly related to logical ability when syllogisms did not conform to the subjects' known views. Eccles (1966) and Pascual-Leone (1968, 1969) have shown that field dependent children are less able to integrate numbers of stimuli. Pascual-Leone believes that field dependence is an intervening variable which may restrict logical func tioning (in terms of Pascual-Leone1s Information Processing Model). He had demonstrated that many Piagetian situations involve a disembedding ability. Even humour appreciation was investigated and found related to perceptual style (Overlade, 1954, 1955). Evidence suggests also that field independent children tend to experience the world in a clear and structured fashion under everyday situa tions. Bieri, Bradburn and Galinsky (1958) found field independent children to score higher scores on measures of cognitive clarity.^ This dimension reflects the extent to which informa tion is discrete, structured and assimilated versus blurred, confused and unassimi 1 ated (Witkin et_. a]_. 1962 , pp. 103-114). 11 2.1 Cognitive Style is Related to  Dependence Upon Others Of special interest in terms of E.S.S. teaching strategies was the research of Gardner e_t. aj_. (1959) and Duram (1964) in which it was revealed that field dependent children learned less in terms of "non-human" incidental learning than did the field independent children, but when the incidental learning material was human faces, this trend was reversed. Gordon (1953) by means of a Thurstone scale, demonstrated that field dependent persons perceived themselves as socially dependent and were judged by others as more dependent as well. Pemberton (1952) in a factor analytic study found that the group which in her terminology corresponded to field dependent classifica tion, tended to be less ambitious, less perservering and less theoretical. Bell (1955) found field dependent sub jects more "other directed" than the field independent subjects. Similarly, Frenkel-Brunswick (1949) found that those people who relied on others for guidance tended to show more intolerance for ambiguity in perception. Numbers of studies have related high scores on the F--scale of authoritarianism to field dependence (Pollack e_t. aj_. , 1960 ; Jackson, 1 955; Linton, 1952). One of Fenchel's (1958) dimensions on his RAPH scale of social 12 rigidity was the belief in "rules for rules sake." This scale was correlated in the expected direction with field dependence. Bales and Couch (1956) showed that men who demonstrated a tendency to accept external authority were more field dependent. A mass of studies has indicated that field dependent people are more easily persuaded by group pressure or external authority (Linton, 1952; Asch, 1956, Crutchfield, 1957; Linton and Graham, 1959). Many studies indicated that global children would be prone to experience difficulty in unstructured learning situations. From case studies Witkin and his associates found children of "limited differentiation" (global or highly field dependent) to be characterized by the following attributes: . . . poverty of resources, lack of enterprise and initiative, underdeveloped interests, laok of well-structured controls and defenses and marked depen dence on others. Children showing this constella tion are perhaps the prototype of limitedly differ entiated children. A second subgroup showed as an outstanding characteristic severe problems of impulse control though also giving evidence of marked poverty of resources. Finally, the third and smallest subgroup consisted of children whose outstanding characteristic was a high level of verbal skills in the absence of developed unerlying structure. They presented a picture of uneven development^ (Witkin, 1962, pp. 268-269). These people could perhaps be called false accomo-dators in Piagetian sense. 13 The children described above would quite likely have more difficulty learning situations which call for the child to initiate and carry out his own investigations into phenomena that are revealed to him through play, than those children with many resources at their disposal. 2.2 Studies Reveal the Ontogeny of  Differentiation Some work has been carried out to determine the origins of individual differences in the levels of differentiation. The notion of environmental interaction upon a genetic base has long been a popular view. Vanderberg (cited by Witkin, 1962) studied Rod and Frame Test (R.F.T.) score variation among fraternal and identical twins; results were incon clusive. Perhaps a neurophysical approach to this topic will reveal mechanisms which will account for the dis-embedding process. J. G. Miller's The Individual as an Information Processing System (Fields and Abbot, eds., 1963) , provides a basis for interesting speculation about field dependence and differentiation, in terms of neural development and capacity. However, the "nature-nurture" question remains unresolved. Nevertheless, modes of perception have been studied in terms of their stability and in terms of the child-rearing practices of mothers. Witkin and his associates (1967) have demonstrated by means of both longitudinal and cross-sectional studies, that the perceptual style of an individual is surprisingly stable. During the growth years, children tend to become more differentiated, but it seems that relative to the group, children remain stable in terms of differen tiation. These trends can be seen in the graphs in Figure 24 21 18 15 12-9 • 6 • 0 W 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 AGE (YEARS) . Developmental curves for rod-and-frame. test based on cross-sectional data. 6-!3yeor mole* 0-----0 8'l5y*of («™O'M • '0'2Cy!O' tro'ti o OI0-24J.0* f-Tioiei 0 V> 8 9 10 II 12 IJ <* >i 'G l y .8 19 20 21 22 2J 24 ACC HEARS) Developmental curves for rod-and-frame test based on longitudinal data. ISO 150 120 90 60-30-8 9 '0 II 12 13 14 15 16 17 18 19 20 21 ACE(YEARS) Developmental curves for embedded-figures test based on cross-sectional data. UJ 0 8 9 10 II 12 13 14 15 16 17 18 19 20 21 AGE(YEARS) Fic. 1. Developmental curves for body-adjustment test based on cross-sectional data. Figure II: Stability of Cognitive Style over Time (Witkin, 1967), pp. 295-296) 15 Fliegel (1955), using college students, found that over a three-year period, test-retest correlations were extremely high on the entire test battery. Franks (1956) utilizing barbituates, amphetimenes, and placeboes, found that perceptual style was remarkably stable after treatment. Of note also is a study by Pollack, Karp and Fink (1960) in which it was found that convulsive therapy tended to reduce field dependence (as cited by Witkin, 1962). These findings suggest that a biophysical mechanism could be involved in the disembedding-analyzing process. Attempts to change perceptual style of adults via training schedules or via periods of sensory deprivation have not been successful in altering the individual's mode of perception (Witkin, 1948, 1967; Davis, McCourt, and Soloman, 1958; Gruen, 1955). In studies with young people, Witkin, Goodenough and Karp (1961) found that the tendency toward field independence increased generally to about age seventeen and then mode of perception stabilized. Relative to the group however, children remained remarkably stable in mode of perception. Witkin (1967) stated that these stability studies dealt with subjects who themselves  were reared in stable family and overall environmental  settings. Findings are not as of this time generalizable to unstable environmental child rearing conditions. Never theless, one may infer from this literature that if educators 16 wished to alter modes of functioning, the children involved would quite likely have to be quite young. By means of case studies and interviews, Witkin and his associates found that mothers of undifferentiated children complained often about their husbands and had difficulty themselves coping with everyday-life situations. Mothers of differentiated children tended to be more self-assured and to exhibit a sense of self-realization. In terms of mother-child interaction, mothers of undif ferentiated children differed greatly from mothers of differentiated child as a general rule. Undifferentiated children tended to fail to meet the mother's expectations--especially in intellectual achievement, appearance, behavior, and aggression (when agression was directed against the mother). Approval was given on a contingency that the child "be good" and not demanding of care. Mothers of differentiated children tended to be more approving with focus on such things as school achievement, creativity, and the child taking responsibility upon himself. These could be considered to be age-appropriate behaviors. Mothers of analytical children also stimulated curiosity and interest. The former mothers, however, tended to stress conformity, prevented the child's taking-up activities as described above, gave physical care which was not age adequate and often expected adult behavior from their child. 17 Attempts were made to control children by irrational means and by vacillation from indulgence to severe discipline and to coercion. Other studies showed that mothers of poorly differentiated or global children tended to be more poorly differentiated or global themselves (Witkin, 1962, pp. 286-367). 2.3 Discriminating Attributes of Cognitive Styles are  Similar to the Attributes of E.S.S.  Activities Many of the attributes of the kinds of activities that children do in the Elementary Science Study are similar to the attributes on which one can also discrimi nate between analytic and global children. This similarity is summarized in Table 1 on the following page. In general there seems to be strong suggestion that the construct of cognitive style is worthy of investiga tion in terms of science education. If children who are more field dependent do not experience as much success on elementary science competency measures (as defined in Appendix A) then that fact in itself would be worth knowing just from a theoretical point of view. Through the con struct of cognitive style teachers could be assisted in identifying these children who experience difficulty working 18 TABLE 1 A Comparison Showing the Similarity Between E.S.S. Objectives and the Discriminating Attributes of Global versus Analytical Functioning Attributes of E.S.S objecti ves Discriminating attributes of global versus analytical functioning Atti tudes: "Messing about" is fun. "Messing about" will lead children to pursue and follow-up phenomena. "Messing about" will lead children to impose a struc ture on their own. "Messi ng chi1dren on their about will lead to investigate own. variables Science Processes Observi ng CI as s i fy i ng Analyzi ng Control 1i ng Predi cti ng Handling data Experimenti ng Repli cati ng Posing problems Acquiring practical Creative Component: skills free wheeling speculation, creative problem solving, and intuitive, playful exploration. Humour appreciation (Overlade) (Pemberton) (Witkin) Ambi tion Perserverance Impulse control Cognitive clarity (Biera et. al.) Other directedness (Gordon ;Bel1 ) Authoritarianism (Pollack) Dependency (Bales and Cooch) Persuasabi1ity (Crutchfield) Integrating large numbers of stimuli (Pascual-Leone) Intolerance of ambiguity (Frenkel) Syl1og i sms (Linton) Insightful thinking (Guetzkow; Goodman) 19 TABLE 1 (Cont'd.) Manipulative and building  skills. Cognitive development: Specific concept develop ment. Incidental learnings Two Hand Coordination Test (Podell and Phillips) Incidental learnings (Gardener; Gordon; Duram) i 20 within the curricular framework of the Elementary Science Study. Teachers may be more understanding and may have more empathy with those children referred to above. Because the psychological construct of cognitive style has at its basis a rather wel1-articulated theory, this theory may be called upon to provide suggestions for specific teaching strategies for these children such as: providing more assistance in ambiguous situations and providing more direction, support, and encouragement for global children. Certainly the theory would suggest that coercive treatment by the teacher would only aggravate the learning situation for global children. In addition the theory would suggest that the initial laissez faire, "0" phase of E.S.S. method ology^ be modified for global children in order that they receive more supportive treatment from the teacher and more encouragement as well. Witkin himself has called for research in education, utilizing his construct of cognitive style. Using this particular cognitive style as a reference let us consider now the ways in which knowledge about the existence of cog nitive styles and their specific nature may be useful in dealing with some of the prob lems encountered in the educational system! In. particulars let us examine the implica tions of the cognitive style work for issues of evaluation, placement and teaching methods. . . . (Witkin, 1969, pp. 217-218). These phases of E.S.S. methodology are described in Appendix B. Witkin concluded his list of suggested implications for educational research by stating "It must be left to educa tors to assess the usefulness of these suggestions" (Witkin 1969, p. 226). In a personal communication with Witkin (January 26, 1970) the author informed him of the general nature of the proposal to use the construct of cognitive style in elementary science education. Witkin replied: I heartily agree with your concept that teaching methods need to be adapted to charac teristics of the individual student, rather than applied in the same way to all students. I have long had the feeling, too, that children who are located at different points of the differentiation dimension differ in the teaching approaches they need. I would include under 'teaching approaches' social as well as cogni tive aspects of teaching methodology. CHAPTER III PROCEDURES 3.0 A Natural Experiment Existed in a Small-City School District School District Number 15 (Penticton, B.C.) provided the site and subjects for this study. A natural experi ment existed there because the materials and teaching strategies of the Elementary Science Study were being phased-in over a number of years. Consequently, it was possible to locate fairly intact groups of grade seven pupils that had experienced the E.S.S. programme for varying numbers of years. It was a relatively easy task to find classes of children that had had experience with the E.S.S. pro gramme for time intervals of one year, two years, and three years . Because of pupil mobility, within each class there were a few individuals who had not been the same number of years on the E.S.S. programme as the rest of the group. These children were randomly assigned to one of the other classes which matched their years of E.S.S. experience. Data for pupils whose previous science experience was not 22 23 known were not included. During the administration of the criterion measures, five of the students moved away, one became seriously ill, and one student died. Students who were temporarily absent from school during the administration of the tests were tested individually upon their return. Before the testing could begin it was necessary to obtain the cooperation of the school officials and the teachers and pupils of the school district. With the help of the District Superintendent of Schools and the Supervisor of Instruction, principals and teachers were informed about the nature of this investigation.^ Conse quently it was possible to administer the individual and the multiple-sitting group tests in a predetermined schedule. Testing was carried out in four schools: Snowdon Elementary, O'Connell Elementary, Queen's Park Elementary, and Naramata Elementary. There were two classes with one year of E.S.S. experience; three classes with two years of E.S.S. experience; and two classes with three years of E.S.S. experience. The writer was introduced to each of these classes by the classroom teacher. Each class was The supervisor selected classrooms that would be comparable in terms of the teacher variable and he also drafted an affidavit testifying to this effect. 24 informed that a group of teachers at the university was trying to find out more about ". . . how kids learned science in school." Pupils were told that this group of university teachers was doing an experiment to learn more — just as the students themselves did experiments in science classes. Students were informed that they would be asked to take a number of tests, to respond to a number of quizzes and to solve a number of puzzles. It was emphasized that none of these tasks would be used for report cards and that all of the results would be confidential. Students were generally very interested in participating. During April, May and June of 1970 over two hundred children were tested. It was possible to utilize the data from 184 subjects who met the aforementioned experiential criteria. There were 92 boys and 88 girls in these groups. 3.1 Individual and Group Tests were Administered to the Subjects All subjects were first given the Children's Embedded Figures Test (C.E.F.T.), an untimed, individually admin istered test. This test provided the independent levelling variable of cognitive style. In general, the protocol utilized for the administration of the test followed the methodology outlined in the test manual. The disembedding 25 process was first introduced in a more casual manner, however. A friendly conversation was initiated which centered about a familiar cartoon figure (Snoopy). Children were then asked to find a simple rectangular figure within the more complex cartoon figure. The experimenter attempted to create an atmosphere in which the subjects were introduced to the task with as little anxiety as possible. Care was taken nevertheless, to have the subjects motivated to do as well as they could on the task. The time to administer the test varied from three to twenty minutes, depending on the subjects.^ Following the administration of C.E.F.T. in each of the schools, the attitude scales were administered to each class in two sittings. The four attitude scales attempted to measure the affective objectives of the Elementary Science 2 Study. The attitude scales were administered according to the methodology described in Appendix B. The attitude scales were administered before the rather lengthy Test of Science Processes in order to avoid the possibility of the attitude measures becoming contaminated by any hostility ^A more complete introduction to the children's Embedded Figures Test is given on pp. 4-6 above. Item and Test analyses are reported in Appendix C. 2 The selection of the performance criteria which were used as science competency measures is discussed in Appendix A. Complete details about the development, content, and method of administration of the four attitude scales can be found in Appendix B. 26 which may have resulted from the administration of the rather difficult process instrument. The four attitude scales provided measures of the first four dependent vari ables in this study. Finally, The Test of Science Processes was admin istered to each of the classes in four sittings. Each of the test items was read to the children and then read again, as they themselves read the items. It was the experimenter's intention to attempt to minimize the effect of reading difficulties confounding the process scores. As much time was allowed on each item as the subjects wished. Any reasonable questions about the wording of an item or the nature of the photographs were answered by the experimenter. As with the attitude scales, children responded on standard answer sheets which facilitated machine scoring and data processing. The eight processes measured by this 96 item process instrument provided the remaining dependent variables for this study. 3.2 The Design of this Study can be Categorized  as a Factorial Model The data were classified into a three by three, fixed effects, factorial design. The independent column variables were represented by the number of years the subjects had had 27 E.S.S. instruction: I, II, III. The independent row variables were represented by three levels of scores deter mined by the C.E.F.T. performance: 0-15 = global; 16-19 = average; and 20-25 = analytical. These three levels of performance were selected on the basis of the analysis by a computer programme entitled v.B.C. H-Gvoup. This program created groups by minimizing the within group variances and maximizing the between group variance. Figure III below represents this design « I II III 32 33 24 18 22 20 8 15 12 Figure III: The Design of the Study Columns I, II, and III represent the three levels of E.S.S. experience -- one year, two years and three years respectively. In addition to lending itself to analysis of variance tech niques, this model also provided categories for carrying out 28 pre-planned comparisons between levels A and C on the fourteen dependent variables. Rows A, B, and C represent the three levels of C.E.F.T. performance -- the analytic group, the average group and the global group respectively. The figures in each cell represent the number of subjects. 3.3 There was no Significant I.Q. Bias in  Columns I, II, and III It was the original intent of the experimenter to utilize the technique of analysis of covariance to arti ficially equate the column means of the dependent variables on the basis of I.Q. A more intensive review of the litera ture on cognitive style revealed however that cognitive style was related to intelligence because the disembedding process was intimately connected with spatial and analytic components of I.Q. tests. The experimenter became concerned about the danger of confounding the effect of cognitive style when covarying on I.Q. In order to determine if the populations represented by the columns in the above design were comparable on I.Q. {Otis Lennon Quick Score) a number of statistical tests were completed. A one-way analysis of variance produced an F ratio of 2.53 (probability of .08). These comparisons were the basis of hypothesis I and II. 29 Eta2 was then calculated from the sums of squares and this statistic revealed that there was no significant relation ship between years and I.Q. (eta2 = .028). Jaspens' multiserial correlation also revealed no significant cor relation between years and I.Q. (r = .05). Furthermore, the complete analyses of variance reported in Chapter IV were re-performed whilst covarying on I.Q. In general the main effect of cognitive style was diluted only slightly by covariance (as theory would suggest), nevertheless the magnitude of the F ratios were still far above the level required for statistical significance in most cases; more over the F ratios for the main effect of years were virtually unchanged when compared to results from analyses of variance. Hence, it was concluded that there was no significant I.Q. bias in the populations represented by the columns in Figure III. Therefore analysis of variance could be inter preted, minimizing likelihood of I.Q. bias. 3 .4 Fourteen Dependent Variables were  Tested in the Design The following dependent variables were found to be suitable criterion measures of Elementary Science Study 30 competencies: DI Attitudes concerning the belief that "messing about" is fun.2 D2 Attitudes concerning the belief that "messing about" will lead children to pursue (or follow-up) phenomena which are uncovered. D3 Attitudes concerning the belief that "messing about" will lead children to impose a struc ture on their play. D4 Attitudes towards the belief that "messing about" will lead children to investigate on their own. D5 Total attitudes score towards the belief state ments concerning the merits of "messing about" in science (E of Dx to Di*). D6 Observing - Test of Science Processes (T.O.S.P.). D7 Comparing - (T.O.S.P.). D8 Classifying - (T.O.S.P.). D9 Quantifying - (T.O.S.P.) D10 Measuring - (T.O.S.P.). Dll Experimenting - (T.O.S.P.). D12 Inferring - (T.O.S.P.). D13 Predicting - (T.O.S.P.). DI4 Total score on all science processes (T.O.S.P.). Justification for the selection of these dependent var iables as appropriate measures of E.S.S. competencies is given in Appendix A which is exclusively devoted to that task. The attitude's were measured by the author's four scales which are described in detail in Appendix B. The science processes were measured by The Test of Science Processes. The term "messing about" refers to an E.S.S. teaching strategy which is described in Appendix B. 31 These variables were tested with analysis of variance for the total group and then separately (post-hoc) for the boys and for the girls (Utilizing BMDX 64). t-Tests and Hotelling's T2 were also used to test the first two hypotheses noted below. 3 .5 The Null Hypotheses When the significance level is set at 5%: I. There will be no significant difference between the mean scores of the analytical group (level A) and the mean scores of the global group (level C) on the affective measures of ele mentary science competencies (Dl through D5). II. There will be no significant difference between the mean scores of the analytical group (level A) and the mean scores of the global group (level C) on the cognitive measures of ele mentary science competencies (D6 through D14). III. There will be no significant main effect for years of E.S.S. instruction on each of the fourteen dependent variables. IV. There will be no significant main effect for cognitive style for each of the fourteen dependent variables. 32 V. There will be no significant interaction effect between years and cognitive style on each of the fourteen dependent variables. 3. 6 The Alternative Hypotheses When the significance level is set at 5%: I. There will be a significant difference between the mean scores of the analytical group (level A) and the mean scores of the global group (level C) on the affective measures (DI through D5), as it is expected that the analytical group will achieve higher scores on these competencies than will the analytical group. II. There will be a significant difference between the mean scores of the analytical group (level A) and the mean scores of the global group (level C) on the cognitive measures (D6 through D14), as it is expected that the analytical group will achieve higher scores on these com petencies than will the global group. III. There will be a significant main effect for years of E.S.S. instruction on each of the fourteen dependent variables. 33 IV. There will be a significant main effect for cognitive style on each of the fourteen dependent variables. V. There will be a significant interaction between years and cognitive style as it is expected that the middle group (level B) and the analytical group (level C) will achieve higher scores with more E.S.S. instruction, but the global group is ex pected to make less progress or even to regress with more experience on this programme. CHAPTER IV THE STATISTICAL ANALYSES 4.0 Raw Data was Analyzed by the Computing  Facilities of the University  of British Columbia Group administered tests were scored by the IBM 1232 marking device and made ready for analysis. The results of the C.E.F.T. and The Test of Science Processes were then analyzed by utilizing a computer program developed especially for item analysis by the Department of Mathematics Education at U.B.C. This program is entitled ED46:TIA. All tests and subtests of these two dichotomously scored instruments were then marked and item analysis was carried out. Point biserial correlations with the test and sub test totals, as well as K.R.-20 internal reliability consistency coefficients for the process instrument and the C.E.F.T. are tabulated in Appendix C. Analyses of the four attitude measures was done by means of specially written computer programs listed at the conclusion of Appendix B. Item analysis was done during final validation 34 35 procedures, utilizing the output of U.B.C. Faeto. The results of these analyses of the attitude scales can all be found in Appendix B. After scoring all of the various tests, all of the data were entered on cards in a form acceptable to the analysis of variance program BMDX64, and to the U.B.C. Triangular Regression Package (TRIP). The final form of these data is listed in Appendix D. 4.1 Multivariate and Univariate Tests were  Utilized for the Statistical  Test of Hypothesis I a). Statistical tests and their Results: The statistical hypothesis to be tested was the null hypothesis H^: lJc = v^ versus the alternative hypothesis H,: yc < v^, where v represented the population row means of row A and where u represented the popula tion row means of row C for the set of four affective dependent variables (Dl through D4). Initially a one-tailed t-test was made to compare the means of the global group with the means of the analytical group on each of the five variables individually. Utilizing the "filter" statements and the "select" statements of the 36 computer program TRIP (U.B.C.'s Triangular Regression Package), the dependent variable scores of those individuals who scored between 0 and 15 on C.E.F.T. were compared to the dependent variable scores of those individuals who scored between 20 and 25 on C.E.F.T. The subroutine tj-Te_st was then employed and the following output was produced: Table 2 t-Tests Comparing the Global versus the Analytical Groups on Each of the the Affective Variables Name Name t-Value D.F. TPROB. Formulal Level C: Level A: DI Funatt vs . Funatt 2.152 122 0.016 (3) D2 Pursue vs. Pursue 3.276 122 0.001 (3) D3 Struct vs . Struct 2.282 122 0.0175 (3) D4 Indexp vs . Indexp 3.526 122 0.0005 (3) D5 Totatt vs . Totatt 3.427 122 0.0005 (3) These individual t-tests appeared to indicate that there was sufficient evidence to reject the null hypothesis for Hn . ul With this fairly large number of t-tests, however, there was considerable likelihood of misinterpreting the significance The formula utilized to obtain t (when population vari-iances were equal) for these comparison was t = Jl~ Iz — (ru-QS!2 + (n2-l)S22,l + 1 x ni + n2- 2 vnx n2 37 of individual t-values by being overly confident about the probabilities of each individual comparison. Hotelling's T2 statistic was computed, therefore, to determine if the global versus the analytical criterion would discrimi nate differences on the affective measures (Dl through D4)^ when these dependent variables were considered simultane ously. Table 3, on the following page, indicates the results of this multivariate comparison. Hotelling's T2 was sufficiently large to reject the null hypotheses for Hn . Further, because the confidence intervals for D2 and ul D4 did not contain zero, it was statistically justifiable to conclude that the variables D2 and D4 were each, indi vidually sufficient to cause rejection of the null hypothesis for Hn . Hence the probabilities for D2 and for D4 could ul be relied upon to be less than .05 when considered inde pendently. The probabilities for Dl and for D3 (reported in Table 2) quite likely could not be relied upon to be less than .05 as Table 2 would suggest. Nevertheless, the difference for Dl and D3 approached statistical significance. During the analysis that was done to test Hn it ul was noticed that there were interesting differences in D5 (the total attitude scale) had to be eliminated from this multivariate analysis because D5 is a linear combination of other four affective variables. It was felt that the individual t-test probability of D5 could be relied upon because of the magnitude of the t-value and because of the fact that D5 was a linear combination of all the attitude scales (Dl through D4). 38 Table 3 A Multivariate Comparison of .Global versus Analytical Groups on the Affective Measures of the Elementary Science Study. Confidence Intervals were Computed at the .05 level (one-tail) Level C Level A Di fferences Between Means Confidence Intervals for Differences Between Means Group 1 Mean Group 2 Mean Left Limit Right Limit Funatt Pursue Struct I ndexp 49.657 43.771 38.571 38.400 54.629 51 .719 44.640 47.449 -4.972 -7.948 -6 .069 -9.049 -11.573 -14.879 -13.667 -16.383 1 .629 -1 .016 1 .529 -1.716 Data give a Hotelling T-squared value of 15.623 and associated F-value 3.810 which is significant. The computed Probability of this statistic = .0061 compared to Pre-set probability of .05 (one-tai1) Degrees of freedom — 4 vs. 119 F-value used in determination — 1.991 t-values when the total population was divided according to sex and separate analysis done. It would appear that for girls especially, C.E.F.T. score groupings did not discriminate between performances on either Dl or D3 at probability levels that were' significant or were approaching significance. This comparison of t-values for sex groupings is summarized below in Table 4. 39 Table 4 A Comparison of Individual t-Tests of the Total Group versus the t-Tests of Each Sex Separately on the Affective Measures Variables Names Both Sexes (122 D.F.) Boys (67 D.F.) Girls (53 D.F.) Affective Variable: t-value T PR OB t-value T-PROB t-value T PROB DI Fun Attitude 2.152 .016 1 .823 .0345 1 .153 (-) .1265 D2 Pursue Att. D3 Structure Att. D4 Individ. Exper. Att. D5 Total Attitude 3.276 2.282 .001 .0175 2.104 2.037 .0185 .0265 2.607 1 .023 .0055 (-) .156 3.526 3.427 .0005 .0005 2.692 2.519 .003 .007 2.197 2.277 .0155 .0175 indicates that this difference not significant near .05 level. ( 40 b). Summary of the Statistical Test of HN : ul Hotelling's T2 was computed to test the null hypothesis that there would be no significant difference between the mean scores of the analytical group (level A) and the mean scores of the global group (level C) on the affective measures of elementary science competencies. The null hypothesis was rejected. It was found that the global group did significantly less well on the four attitudinal measures. Further, it was found that D2 and D4 were in themselves sufficient to cause rejection of the null hypothe sis and that the individual probabilities associated with these two variables could be relied upon to be less than .05. Sex influences were also noted, particularly for the attitudes measured by Dl and D3. 4.2 Multivariate and Univariate Tests were Utilized  for the Statistical Tests of  Hypothesis II a). Statistical Tests: The statistical hypothesis to be tested was the null hypothesis HQ : yc = versus the alternative hypothesis H^: yc < v^, where v represented the population row means of row A and where y represented the population row means 41 of row C for the set of eight cognitive (process) variables (D6 through D13). Initially a one-tailed t-test was utilized to compare the means of the global group (row C) with the means of the analytical group (row A) on each of the pro cesses separately. The results of these individual t-tests are reported below in Table 5: Table 5 t-tests Comparing the Global versus the Analytical Groups on Each of the Processes Level C Level A Variable vs . Variable t-value D.F. t.prob D6 Obs erv vs . Observ 5.885 122 0.000 (3) D7 Compar vs . Compar 2.447 122 0.0075 (3) D8 Class i vs . C1 a s s i 5.688 122 0.000 (1) D9 Q u a n t i vs . Q u a n t i 5.246 46 0.000 (3) D10 Meas ur vs . Measur 6.492 122 0.000 (3) Dll Experi vs . Experi 3.057 122 0.0015 (3) D12 I nferr vs. I nferr 5.704 122 0.000 (3) D13 Predi c vs . Predi c 4.313 93 0.000 (1) 1 DI 4 Totalp vs . Totalp 7.57: 122 0 .000 (3) (1) Note: t = I Si + s2 "NT (with unequal popula^ tion variances) (3) t = (nrl)s; + (n2-l)S2 (wi th equal popu-lations variances) 42 Because sex differences were noticed during the statistical test of Hn , the total population was cate-ul gorized according to sex and separate analyses were~ done. For the science processes (unlike the attitudes) there were no marked differences in C.E.F.T.'s discrimination attributable to sex. These statistics are summarized in Table 6 below. Table 6 A Comparison of Individual t-Tests of the Total Group versus t-Tests of Each Sex Separately on the Cognitive Measures (Processes) Variables Names Both Sexes (122 D.F.) Boys (67 D.F.) Girls (53 D.F.) Cognitive Variables: t 5.885 2.447 5.688 5.246 6.492 3.057 5.704 4.313 7.574 prob, .000 .0075 .000 .000 .000 .0015 .000 .000 .000 t 3.923 1 .486 4.602 4.170 5.569 2.087 4.231 1 .686 5.816 prob. .000 .069 .000 .000 .000 .0195 .000 .046 .000 t 4.562 1 .963 3.324 2.154 3.864 2.200 3.70 4.52 4.747 prob. .000 .026 .001 .022 .000 .0155 .0005 .000 .000 D6 Observing D7 Comparing D8 Classifying D9 Quantifying D10 Measuring Dll Experimenting D12 Inferring D13 Predicting D14 Total Processes Once again, individual t-test evidence indicated that each individual dependent variable revealed significant differences in the performance of the global group versus the performance of the analytical group. Again, however, it was not possible to rely upon the face value of the probabilities for this large number of t-tests. Hotelling's T2 was computed for this set of variables too, in order to determine if the global group achieved significantly lower 43 scores than the analytical group on the set of process variables when these dependent variables (D6 through D13) were considered simultaneously.^ This multivariate test was produced by utilizing the TRIP subroutine entitled HOTEL. Output is summarized below in Table 7. Table 7 A Multivariate Comparison Between the Global Group and the Analytical Group on the Cognitive Elementary Science Competencies (Processes) Confidence Intervals for the Differences Between the Means D6 Observ D7 Compar D8 Classi D9 Quanti D10 Measur Dll Expert D12 Inferr D13 Predic Group 1 Mean Level C 3.886 3.143 7.743 7.542 1 .400 4.400 4.857 3.486 Group 2 Mean Level A 6.067 3.629 9 .809 9.888 16.360 5.494 7 .427 4.685 Left Limit -3.600 -1.247 -3.456 -3.784 -7 .882 -2.464 -4.293 -2.462 Right Limit •0.764 0.274 •0.677 •0.906 •2.037 0.275 •0.846 0.063 Di fferences Between Means -2.182 -0.486 -2.066 -2.345 -4.960 -1 .094 -2.570 -1 .200 Data give a Hotelling T-squared value of 68.505 and associated F-value 8.072 which is significant. The computed probability of this statistic is 0.0000 compared with the criteria probability of .05 (one-tail). Degrees of Freedom - 8 vs. 115. F-value used in determination - 1.724. D14 (Total Processes) had to be eliminated from the computation of T2 because it was a linear combination of the other variables and hence was inadmissable to Hotel . The probability of D14 could be relied upon in any case, as it is a lineal combination of the eight processes. 44 Hotel 1 i rig's-2 was sufficiently large to reject the null hypothesis for Hn . Further, because the confidence UII intervals for D6, D8, D9, D10, and D12 did not contain zero, it was statistically justifiable to conclude that the variables D6, D8, D9, D10, and D12 were each in them selves sufficient to cause rejection of the null hypothesis for Hn . Hence the individual probabilities for these UII specific variables could be relied upon to be less than .05 when considered as separate entities. The probabilities for D7 and Dll could not be relied upon to be less than .05 as Table 6 might seem to suggest. Nevertheless, the differences for D7 and Dll were approaching significance at the .05 level. 45 b). Summary of the Statistical Test of Hn II Hotelling's T2 was computed to test the null hypothesis that there would be no significant difference between the mean scores of the analytical group (level A) and the mean scores of the global group (level C) on the set of cognitive measures of elementary science competencies (science processes). The null hypothesis was rejected. It was found that the global group did significantly less well on the eight science processes. Further, it was found that for the tests of Observation, Classification, Quantification, Measuring, and Inferring, these tests were each individually sufficient to cause rejection of the null hypothesis, and that the individual probabilities associated with these variables could be relied upon to be less than .05. No marked sex influences were uncovered for the eight science processes. 4.3 The Statistical Tests of Hypotheses  III, IV and V a) Analysis of Variance Tests of Hn , Hn , and Hn . U111 UIV UV During the tests of Hn and Hn , it was observed Uj un that the global group achieved significantly lower scores on elementary science competencies than did the analytical 46 group. Moreover, sex differences were found to play some part in determining the significance of the differences for the affective measures. It was considered prudent, therefore to test the three remaining hypotheses, not only for the total sample, but also for boys and for girls separately. The results of these analyses of variance are summarized below in Table 8. Findings are reported for the total group in Table 8a. Table 8b reports findings for the boys and Table 8c reports results for the girls. Because of the extensiveness of these analyses, the acceptance and rejection of particular hypotheses were included as part of the tables referred to. Following these tables a summary of accepted versus rejected hypotheses is also presented in tabular form. 47 Table 8a Total Groups' Analyses of Variance Tables Dl . FUN SCALE Hypo. Source Sum of Squares D.F. Mean Square F Ho Ill IV V Mean 439701.01119 1 439701.00000 3425.18384 accept* reject accept Years Cogs t cy 175.82117 920.97491 281.06846 2 2 4 87.91058 460.48730 70.26709 0.68481 3.58710 0.54737 Error -22465.27153 175 128.37296 D2. PURSUE SCALE III IV V Mean 368982.75732 1 368982.750000 2125.11768 accept reject accept Years Cogs t Cy 11.55316 1552.10600 418.24636 2 2 4 5.77658 776.05298 104.56158 0.03327 4.46960 0.60221 Error 30385.13441 175 173.62932 D3. STRUCTURE SCALE III IV V Mean 284245.60927 1 284245.46250 1549.29395 Years Cogs t cy 18.19544 970.19061 333.87566 2 2 4 9.09772 485.09521 83.46887 0.04959 2.64403 0.45495 accept aLc£ep_t( + ' accept Error 32106.86338 175 183.46779 Note: (+) = approaching significance at a = .05. 48 D4. INDIVIDUAL EXPERIMENTATION Hypo. Source Sum of Squares D.F. Mean Square F Ho Ill IV V Mean 295076.60211 1 295076.56250 1669.96802 accept rej ect accept Years Cogst Cy 186.68892 1723.10846 290.41381 2 2 4 93.34445 861.55420 72.60339 0.52828 4.87591 0.41089 Error 20921.79047 175 176.69592 D5. TOTAL ATTITUDES TOWARDS MESSING ABOUT IN SCIENCE III IV V Mean 5506305.52455 1 5506305.00000 3020.86548 accept rej ect accept Years Cogs t cy 378.31424 19480.75916 1994.77889 2 2 4 189.15710 9740.37891 498.69458 0.10378 5.34376 0.27359 Error 318982.60941 175 1822.75732 D6. OBSERVING III IV V Mean 3862.36532 1 3862.36523 1165.75171 reject rej ect accept Years Cogst cy 25.09983 101 .54618 11 .35123 2 2 4 12.54991 50.77309 2.83781 3.78786 15.32450 0.85652 Error 579.80960 175 3.31320 D7. COMPARING III IV V Mean 1772.53103 1 1772.53101 1685.45752 accept ac£e£t(+) accept Years Cogst Cy 1 .71745 6.28930 0.70462 2 2 4 0.85873 3.14465 0.17615 0.81655 2.99017 0.16750 Error 184.04091 175 1 .05166 49 D8. CLASSIFYING Hypo. Source Sum of Squares D.F. Mean Square F Ho Ill IV V Mean 12233.92816 1 12233.92578 4087.74048 accept reject reject Years Cogst cy 3.37249 95.06451 31.26261 2 2 4 1.68624 47.53224 7.81565 0.56343 15.88202 2.61145 Error 523.74596 175 2,99283 • D9. QUANTIFYING III IV V Mean 12129.96730 1 12129.96484 3769.07544 accept reject accept Years Cogs t Cy 8.25529 122.98666 20.71625 2 2 4 4.12764 61.49322 5.17906 1 .28256 19.10747 1 .60926 Error 563.20032 175 3.21829 D10. MEASURING III IV V Mean 29635.34867 1 29635.34766 1963.31738 Years Cogst cy 105.23352 634.96289 61.75043 2 2 4 53.61676 317.48120 15.43760 3.48582 21.03287 1.02273 reject reject accept Error 2641.54223 175 15.09453 Dll. EXPERIMENTING Mean 3638.34347 1 3638.34326 1101.12793 III IV V Years Cogst Cy 7.18945 37.41347 15.17475 2 2 4 3.59473 18.70677 3.79369 1.08793 5.66152 1.14814 accept reject accept Error 578.23441 175 3.30420 50 D12. INFERRING Hypo. Source Sum of Squares D.F. Mean Square F Ho Ill IV V Mean 5976.04107 1 5976.03906 1193.24414 reject reject accept Years Cogst cy 30.50133 158.01676 5.19325 2 2 4 15.25066 79.00838 1.29831 3.04512 15.77572 0.25924 Error 876.44015 175 5.00823 D13. PREDICTING III IV V Mean 2722.50407 1 2722.50391 1073.22925 accept reject accept Years Cogst Cy 0.55284 34.45086 3.72862 2 2 4 0.27642 17.22542 0.93215 0.10897 6.79037 0.36746 Error 443.92986 175 2.53674 D14. TOTAL SCIENCE PROCESSES III IV V Mean 471890.62186 1 471890.56250 3957.72778 Years Cogst Cy 512.62593 6769.76153 428.12938 2 2 4 256.31274 3384.87891 107 .03229 2.14968 28.38884 0.89768 ac£ep_t( + ) reject accept Error 20865.72399 175 119.23270 F values required for rejection of H : Df. a. 1-0 a .05 a.01 2 vs. 175 2. 30 3 .00 4.61 4 vs . 175 1 . 94 2 .37 3.32 51 Table 8b Boys' Analyses of Variance Tables DI. FUN SCALE Hypo. Source Sum of Squares D.F. Mean Square F Ho Ill IV V Mean 233838.13303 1 233838.12500 1777.18140 accept accept ( + accept Years Cogst Cy 94.19879 729.87535 434.34932 2 2 4 47.09940 364.94750 108.58728 0.35796 2.77354 0.82527 Error 11447.29659 87 131.57808 D2. PURSUE SCALE III IV V Mean 192378.15216 1 192378.12500 1087.43848 accept reject accept Years Cogst Cy 129.11036 1890.11850 808.34525 2 2 4 64.55518 945.05908 202.08630 0.36491 5.34205 1.14231 Error 15391.12160 87 176.90942 D3. STRUCTURE SCALE III IV V Mean 146585.84323 1 146585.81250 723.72192 accept reject accept Years Cogst cy 59.27946 1355.96059 132.54475 2 2 4 29 .63972 677.98022 33.13618 0.14634 3.34732 0.16360 Mean 1 7621 .3621 5 87 202.54436 BOYS (Cont'd) 52 D4. INDIVIDUAL EXPERIMENTATION SCALE Hypo. Source Sum of Squares D.F. Mean Square F Ho Ill IV V Mean 16294.81453 1 162694.81250 932.93555 accept reject accept Years Cogst cy 386.74678 1716.76755 684.53738 2 2 4 193.37329 858.38354 171.13434 1.10885 4.92220 0.98133 Error 15171.94799 87 174.39017 D5. TOTAL ATTITUDE TOWARD "MESSING ABOUT" III IV V Mean 2918621.07862 1 2918621.00000 1447.76831 accept reject accept Years Cogst Cy 1197.47363 21883.99120 175387.20713 2 2 4 598.73657 10941.99219 2015.94458 0.29700 5.42772 0.61057 Error 175387.20713 87 2015.94458 D6. OBSERVING III IV V Mean 2110.09156 1 2110.09155 586 .45752 reject reject accept Years Cogst Cy 27.75577 33.52035 27.20047 2 2 4 13.87788 16.76016 6.80012 3.85708 4.65815 1 .88996 Error 313.02880 87 3.59803 D7. COMPARING III IV V Mean 904.159 51 1 904.15942 736.23315 accept accept accept Years Cogst Cy 0.60633 1.89718 0.84639 2 2 4 0.30317 0.94859 0.21160 0.24686 0.77241 0.17230 Error 106.84375 87 1 .22809 BOYS (Cont'd) 53 D8. CLASSIFYING Hypo. Source Sum of Squares D.F. Mean Square F Ill IV V Mean 6062.16458 1 6062.16406 2068.88110 accept reject reject Years Cogst Cy 3.42785 50.11094 29.55982 2 2 4 1 .71393 25.05547 7.38995 0.58492 8.55087 2.52203 Error 254.92441 87 2.93017 D9. QUANTIFYING III IV V Mean 6205.42328 1 6205,42188 1989.42993 accept reject acce£t(+) Years Cogst Cy 6.65426 86.65572 29.68223 2 2 4 3.32713 43.32785 7.42056 1 .06666 13.89071 2.37900 Error 271.37018 87 3.11920 D10. MEASURING III IV V Mean 16816.69776 1 16816.69531 1193.97412 Years Cogst Cy 32.53012 306.90663 88.38174 2 2 4 16.26506 153.45325 22.09543 1.15481 10.89508 1.56876 accept reject accept Error 1225.36373 87 14.08464 Dll. EXPERIMENTING III IV V Mean 1772.25667 1 1772.25659 495.34985 Years Cogst Cy 2.35120 17.86458 8.14008 2 2 4 1.17560 8.93229 2.03502 0.32858 2.49660 0.56879 accept a_ccejDt( + ) accept Error 311.26777 87 3.57779 54 BOYS (Cont'd) D12. INFERRING Hypo. Source Sum of Squares D.F. Mean Square F Ho Ill IV V Mean 2908.82763 1 2908.82739 542.48350 accept reject accept Years Cogst Cy 7.48507 73.88997 12.34654 2 2 4 3.74253 36.94498 3.08663 0.69797 6.89007 0.5/564 Error 466.49914 87 5.36206 D13. PREDICTING III IV V Mean 1394.64486 1 1394.64478 554.35938 accept accept accept Years Cogs t cy 3.84833 6.28463 5.10169 2 2 4 1 .92416 3.14232 1 .27542 0.76484 1 .24904 0.50697 Error 218.87265 87 2.51578 D14. TOTAL PROCESSES III IV V Mean 245712.80379 1 245713.75000 2040.54761 accept reject accept Years Cogst cy 223.51510 2930.68879 723.62113 2 2 4 111 .75754 1465.34424 180.90527 0.92810 12.16906 1.50234 Error 10476.15713 87 120.41559 F values required for rejection of H-: Df. a .10 a.05 a .01 2 vs . 87 2 .37 3.15 4 .98 4 vs . 87 2 .05 2.53 3 .65 55 Table 8c Girls' Analyses of Variance Table DI. FUN SCALE Hypo. Source Sum of Squares D.F. Mean Square F Ho Ill IV V Mean 193423.48986 1 193423.43750 1530.63770 accept accept accept Years Cogst Cy 482.59342 328.65358 363.13841 2 2 4 241.29663 164.32678 90.78455 1 .90948 1 .30038 0.71841 Error 9983.06445 79 126.36787 D2. PURSUE SCALE III IV V Mean 165693.71054 1 165693.68750 1070.15430 accept reject accept Years Cogst Cy 333.06169 1081.66355 790.01929 2 2 4 166 .53076 540.83154 197.50482 1 .07556 3.49303 1 .27561 Error 12231.69907 79 154.83157 D3. STRUCTURE SCALE III IV V Mean 129091.89722 1 129091.87500 752.33862 accept accept accept Years Cogst Cy 5.82983 128.16886 713.60288 2 2 4 2.91492 64.08443 178.40070 0.01699 0.37348 1.03971 Error 13555.40699 79 171 .58742 GIRLS (Cont'd) 56 D4. INDIVIDUAL EXPERIMENTATION SCALE Hypo. Source Sum of Squares D.F. Mean Square F Ill IV V Mean 127259.35428 1 127259.31250 740.66821 accept a_c£e p_t (+) accept Years Cogst Cy 8.45721 981.82155 314.55069 2 2 4 4.22861 490.91064 78.63763 0.02461 2.85717 0.45768 Error 13573.53462 79 171.81685 D5. TOTAL ATTITUDE TOWARD "MESSING ABOUT" IN SCIENCE III IV V Mean 2440425.93034 1 2440425.00000 1552.68945 accept a_c£ej3t( + ) accept Years Cogst Cy 1990.16335 7408.92599 4297.74115 2 2 4 995.08154 3704.46289 1074.43457 0.63311 2.35692 0.68360 Error 124167.55420 79 1 571 .74048 D6. OBSERVING III IV V Mean 1594.21551 1 1594.21533 559.45532 Years Cogst Cy 17.56233 74.57972 13.35016 2 2 4 8,78116 37 .28955 3.33754 3.08156 13.08596 1.17124 reject reject accept Error 225.11713 79 2.84958 D7. COMPARING III IV V Mean 807.68098 1 807.68091 893.91797 Years Cogst Cy 1 .54633 4.80940 3.87694 2 2 4 0.77316 2.40470 0.96923 0.85572 2.66145 1 .07272 accept ac£ep_t(+' accept Error 71 .37879 79 0.90353 GIRLS (Cont'd) 57 D8. CLASSIFYING Hypo. Source Sum of Squares D.F. Mean Squares F Ho Ill IV V Mean 5770.62647 1 5770.62500 1872.84473 accept reject accept Years Cogst Cy 1 .0921 6 49.23180 15.19630 2 2 4 0.54608 24.61589 3.79907 0.17723 7.98904 1 .23298 Error 243.41550 79 3.08121 D9. QUANTIFYING t—11—i 1—4 Mean 5558.63" 155. 1 5558 .62891 1653.35815 t—11—i 1—4 Years Cogst Cy 2.54197 29.55254 4.33905 2 2 4 1 .27099 14.77627 1.08476 0.37804 4.39505 0.32265 accept reject accept t—11—i 1—4 Error 265.60000 79 3.36202 D10. MEASURING III IV V Mean 11900.29503 1 11900.29297 767.10522 Years Cogst Cy 87.58741 323.79244 53.71161 2 2 4 43.79370 161.89612 13.42790 2.82299 10.43599 0.86558 accept reject accept Error 1225.54662 79 15.51324 Dll. EXPERIMENTING III IV V Mean 1727.00083 1 1727.00073 529.33911 Years Cogst Cy 8.87023 20.68681 10.09581 2 2 4 4.43511 10.34340 2.52395 1 .35940 3.35940 0.77361 accept reject accept Error 257.74231 79 3.26256 GIRLS (Cont'd) 58 D12. INFERRING Hypo. Source Sum of Squares D.F. Mean Squares F Ill IV V Mean 2928.79588 1 2928.79565 620.17139 Years Cogst cy 36.20162 75.01964 11 .34160 2 2 4 18.10080 37.50981 2.83540 3.83284 7.94269 0.60039 reject reject accept Error 373.08217 79 4.72256 D13. PREDICTING III IV V Mean 1249.36653 1 1249.36646 501.94800 Years Cogst Cy 8.35602 38.78139 5.02702 2 2 4 4.17801 19.39069 1 .25674 1 .67857 7.79044 0.50492 accept reject accept Error 196.63380 79 2.48903 D14. TOTAL SCIENCE PROCESSES III IV V Mean 210661.38328 1 210661.37500 1758.59546 Years Cogst Cy 657.75766 3637.19460 491 .02467 2 2 4 328.87866 1818.59717 122.75616 2.74547 15.18160 1.02477 accept reject accept Error 9463.37879 79 119.78955 F values required for rejection of H : Df; a. 10 a.05 a.01 2 vs. 79 2.39 3.15 4.98 4 vs . 79 2.04 2.53 3.65 59 b). Summary of Tests of Hypotheses III, IV, and V. Table 9 below indicates those specific null hypotheses which were accepted versus those null hypotheses which were rej ected: Table 9 Summary of Rejected and Accepted Hypotheses (Hn , Hn and Hn ) u111 UIV UV FUN ATT. PURS ATT. STR ATT. INDIV. ATT. TOTAL ATT. OBSERVING COMPARING CLASSIFYING QUANTIFYING MEASURING EXPERIMENTING INFERRING PREDICTING TOTAL PROC. D 1 D 2 D 3 D 4 D 5 D 6 D 7 D 8 D 9 D 10 D 11 D 12 D 13 D 14 Main E.YEARS: TOTAL A A A A A R A A A R A R A A+ BOYS A A A A A R A A A A A A A A GIRLS A A A A A R A A A A ft R A A COG. ST. TOTAL R R A+ R R R A+ R R R R R R R BOYS A+ R R R R R A R R R A+ R A R GIRLS A R A A + A+ R A+ .R R R R R R R INTERACTION CY TOTAL A A A A A A A R A A A A A A BOYS A A A A A A A R A+ A A A A A GIRLS A A A A A A A A A A A A A A = approach significance at a = .05 D = significant beyond a.05 Total Number of Rejected H 's: Main E.Years: 6/42 0 COG. ST. : 31/42 INTERACTION : 2/42 60 b-1) The Main Effect for Years of E.S.S. Instruction (Hn ): u 111 In general, the main effect for years of instruction was only significant in six of forty-two tests. When data from these significant main effects were plotted (see Fig. IV), there were no clear-cut, obvious increases in scores attributable to years of E.S.S. experience. In general: the main effect of years of E.S.S. experience was not significant. b-2) The Main Effect for Cognitive Style (Hn ): UIV Table 9 indicates that there was a general rejection of the null hypothesis for Hn . The main effect of UIV cognitive style was significant in all but two variables for the total group. In these two exceptions, the F ratios were approaching significance at the .05 level. This general rejection of the null hypothesis for Hn was UIV consistent with the findings for Hn and Hn °I °II Sex differences were observed which were consistent with the findings for Hn and Hn as well. Worthy of UI UII emphasis was the finding that the main effect of cognitive style was not significant for the girls when the means were compared for Dl (the Fun Scale) and for D3 (the Structure Scale). 61 b-3) The Interaction Effect of Cognitive Style and Years of E.S.S. Experience (Hn ): UV Only two of forty-two interaction effects were statistically significant. These two significant inter actions can be seen visually in Figure V. Certainly two significant interactions out of a possible forty-two significant interaction effects, did not provide sufficient evidence to reject the null hypothesis for Hn . In general it was necessary to accept this null hypothesis. Key: ro Middle Analytical ^Global c £(i) Total Group Observing Global Middle Analytical CD u 8 CO Boys Group Observing • • Global o (iii) o CO Middle Analytical Girls Group Observing Middle Analytical SGlobal o m (iv) Total Group Measuring ^Global S(v) CO Middle ' Analytical Total Group Inferring 2 Global o (/j (vi) Girls Inferring Middle Analytical 63 1 yr 2 yr 3 yr 1 yr 2 yr 3 yr Years : Total Group Years : Boys Group Fig. V: Statistically Significant Interactions: Classifying Key -- = analytical = average group = global group 64 1 yr. 2 yr. 3 yr. 68 66 64 62 s 60 58 56 54 / 52 50 A f\ 48 46 O M /-\ o CO Years of Experi ence GROUP TOTAL PROCESSES 1 yr . 2 yr. 3 yr. 230 225 220 210 Years of Experience GROUP TOTAL ATTITUDES Fig. VI: Score Trends for Total Attitudes and Total Processes Based Upon Cell Means from Analyses of Variance (BMDX 64) Total Test Scores Key: Global (Score 0--5 C.E.F.T.) Middle (16-20) Analytical (27-25) Fig. Vl(Cont'd) Global Middle Analytical 66 Global Middle Analytical CHAPTER V CONCLUSIONS AND SUMMARY 5.1 This Study Developed from a Need which  Arose in the Classroom The reader should consider this exploratory study within its proper context. This investigation grew out of the author's classroom teaching experience in which it was observed that certain children appeared to experience much difficulty while operating within the rather un structured learning experiences of the Elementary Science Study. The theoretical framework of Witkin's Differential Psychology (1967) appeared to be of some assistance in explaining why the highly praised and relatively novel teaching materials and methodology of the Elementary Science Study did not appear to be successful with these children referred to above. It was the author's wish to initiate an investigation to determine if the apparently pervasive individual trait of cognitive style would be helpful in explaining differences in achievement in this rather unstructured curriculum. Achievement was measured in terms 67 68 of attitudinal and heuristic science competencies. Further, it was intended that by analyzing cross-sectional data, it would be possible to obtain some insight regarding the effect of cognitive style over time -- as children had more and more experience with the Elementary Science Study materials and methodology. 5.2 Limitations of the Study Because the analysis of the effect of years of E.S.S. experience was based upon cross-sectional data, interpretations about the effects of exposure to the E.S.S. program could only be described as a form of statistical speculation. Ideally a longitudinal study should have been done, but as this particular study was an exploratory investigation, it was reasoned that should Witkin's con struct of cognitive style bear significant relationship with performance on the type of elementary science com petencies appropriate to the E.S.S. curriculum, then the investment of time and labour on a longitudinal study would be warranted. The findings of this thesis appear to support the desirability for such a longitudinal study. 69 5 . 3 Conclusions and Recommendations a) The Effect of Years of E.S.S. Instruction: There was no statistical evidence to state that performance on the criterion measures improved as children had more and more exposure to the E.S.S. programme. This finding should not be viewed as a general indictment of the E.S.S. programme, but rather as an indication of the probable weakness of the cross-sectional research design employed by the author. There were likely too many inter vening variables unaccounted for, to come to any firm conclusions about the effect of years of E.S.S. instruc tion. A longitudinal study would eliminate this difficulty. b) The Interaction of Cognitive Style and Years of E.S.S. Instruction: Similarly interpretations of interactions between experience on the programme and cognitive style are extremely difficult, again because of the assumptions made by the design. Nevertheless, it seems apparent that the E.S.S. curriculum should not be viewed as some educa tional panacea. Graphs of score trends appear to indicate that the more global children may actually regress on competency measures as these children have more and more 70 experience with E.S.S. curriculum. A proper longitudinal study of intermediate-aged students may affirm this specu lative conclusion. The limitations imposed by the design on the con clusions regarding the effect of years of experience and interactions, do not apply for the main effect of cognitive style, because the parameters of cognitive style were clearly defined in terms of performance on the Children's Embedded Figures Test. c) The Effect of Cognitive Style on Performance on the Test of Science Processes: It was found that children who could be categorized as perceptually global, achieved significantly lower scores on both cognitive and affective measures of elemen tary science competencies. With regards to the cognitive objectives which were measured by the Test of Science Processes , it was found that the global group did signifi cantly less well on the set of eight science processes. Further, it was found that on the measures of Observation, Classification, Quantification, Measuring and Inferring --these tests were each in themselves sufficient to conclude that the global group achieved statistically lower scores than the analytical group on the set of science processes. 71 d) The Effect of Cognitive Style on Performance on the Attitude Scales: With regards to the affective objectives -- the attitudes concerning "messing about in science" -- there was a statistical basis to conclude that the global children achieved significantly lower scores than the analytical children on the set of attitudes scales. It was found that differences on the Pursue Scale (D2) and the Individual Investigation Scale (D4) were each sufficient in themselves to conclude that the global group achieved statistically lower scores than the analytical group on the set of attitude measures. Although there was some evidence that global children did less well on measures of the enjoyment attitude (DI) and on the attitude of imposing structure on play (D3), the evidence was perhaps confounded by a number of factors. For instance, cultural ethics regarding inculcated evaluations of the concepts of "work" (structure) and the concept of "fun" may have been involved. For example, sex differences were found with regards to DI and D3 which may have indicated the effect of role encul-turation, i.e. values which the culture inculcated in girls, such as being non-mechanical and "feminine." e) General Conclusions and Recommendations. Generally it should be viewed as an important finding that perceptionally global children appear to be less well 72 equipped attitudinally and heuristically to operate within the rather unstructured methodologies of the Elementary Science Study (and probably other similar "discovery oriented" curricula as well). This finding is significant both from a theoretical and from a practical viewpoint. Theoretically, it enhances the credibility of differential psychology to educators and suggests its broader applica tion for research. In terms of classroom practice, the author suggests the following recommendations: e-1 The Modification of E.S.S. Methodology: The teaching methodology defined by Hawkins (1965) was composed of three phases through which the learning environment was to develop.^ It is the first of these phases which requires modification. In this first stage, entitled the "0" phase, Hawkins defined a learning environ ment which he referred to as "glorious messing about" and "kindergarten revisited." During this phase, students of elementary school science were to be allowed to engage in a rather lengthy period of "free and unguided exploration"--this period sometimes was to take as long as two weeks. In his much cited article Hawkins stressed the need for a 'The "0", "A", and "•" phases of E.S.S. methodology are fully described in Appendix B. 73 re-emphasis upon this type of learning which typifies the style of learning exhibited by children before they come to school. Indeed, Hawkins even suggested the broader application of the "0" phase to the college level. In his zeal to de-emphasize the "•" phase, (abstract and theo retical teaching strategies such as lecturing), Hawkins committed an error of omission. Teachers have long been aware that some children appear to have a poverty of resources at their disposal. These pupils appear to be unable to engage in sustained play with materials. These children are not as capable of profitably utilizing extensive periods of unguided exploration in science classes. Differential psychology provides a theoretical framework which should be of great assistance in helping educators to provide for the type of individual differences referred to above. From the research of Witkin and his followers, it would appear that the degree of cognitive style (global versus analytical funtioning) is generally a pervasive trait after the age of eight years. It would seem to the author, therefore, that especially in the upper elementary school grades, the individual cognitive style of children is an important factor which educators should attempt to accommodate when designing and implementing curriculum. It would be naive to consider that the mere application of any one curriculum would suddenly alter the cognitive style 74 of older children (to drastically enhance the differentia tion process), because the child's particular mode of deal ing with the world has become well reinforced and entrenched by this time. Educators therefore must learn to provide for the individual differences in cognitive style. In the E.S.S. scheme for example, global children have a distinct disadvantage when compared to the more analytical children --while learning within a "fend for yourself" learning envir onment. The teacher should identify the children at the extreme ends of the differentiation continuum (utilizing such tests as C.E.F.T.) and he should modify the "0" phase for these children. Extremely global children require conscientious, systematic guidance and supportive assistance --while analytical children require very little interference by the teacher during this "0" phase. In the primary grades a more detailed methodology should be developed and tested; at this level it may be possible to enhance the differentiation process and by so doing, to actually alter the extremely global child's mode o.f deal ing with the world. This methodology for primary school instruction could be based upon the existing data and theory regarding the child-rearing practices of the parents of more analytical children. The actual development of special teaching strategies which would 75 enhance the differentiation process for young children is beyond the scope of this study. In short, differential psychology presents a theoretical and empirical basis from which to develop differentiated teaching methodologies. e-2 The Modification of the Way Curriculum is Used: Currently the "unit" approach to curriculum is a popular method of applying the materials of the many elementary science projects. For example, teachers may be planning to utilize Batteries and Bulbs (E.S.S.) followed by Pendulums (E.S.S.) -- or perhaps another unit from an entirely different curriculum project with a very different philosophy. Rather than approaching the study of topics or objects through the selection of "a unit" (unilaterally administered to the entire class), the author suggests the following modification in the way that curriculum materials are currently used. The study of topics or objects could be approached in such a way as to account for the styles of learning which are more appropriate to the cognitive styles of particular children. Several avenues or paths of learning should be available on any particular science topic. These paths would differ in the degree of structure and support which they afford the teachers and pupils who wish to investigate a topic. 76 The children and the teachers could decide which path is most suitable. The author does not necessarily suggest the design of entirely new curricula based upon the concept of differentiated routes to learning (although such a curriculum design may be feasible), he suggests however, that it may be presently possible to achieve the desired end by making much better use of curriculum materials which have already been developed. These many curriculum projects themselves have built-in features which reflect particular philosophies and methodologies. Consequently, the entire spectrum of structured versus relatively unstructured elementary science experiences is already available under the covers of these pre-existing curriculum projects. The author recommends that science educators extract the different approaches to given topics from these separate entities, and integrate the various approaches to the topics -- thus providing alternate routes to learning a topic. For example, A.A.A.S. lessons regarding seed germination could be compiled and offered as a more structured path to the topic than the more "free-wheeling" E.S.S. units such as Starting from Seeds. Students could then be afforded a choice as to which path they wished to follow -- structured versus relatively unstructured. 77 5.4 Implications for Further Research As suggested at the beginning of this chapter, there is sufficient evidence about the applicability of the psychological construct of cognitive style to justify development of a longitudinal study (similar in format to this one), in order to determine the interaction of cognitive style and years of E.S.S. experience on performance on elementary science competency measures. It is suggested that this study begin at the lower intermediate grade levels and should continue to the end of elementary schooling. In the primary grades, particularly in kindergarten and grade one, it would be an exciting study to determine the stability of cognitive-style when the teacher con scientiously adopts the type of supportive treatment which Witkin and his associates identified as the child-rearing style employed by parents of field-independent (more analytical) children. The author hypothesizes that such early school experiences would enhance the differen tiation process for exteremely global children.^ What is intended is that the environment of these children be manipulated conscientiously and intensively for several Cognitive style has been found to be stable and pervasive (relative to the group) in stable environments. 78 years in order to determine what changes could be effected upon cognitive style -- a construct which has been con sidered by researchers to be pervasive and relatively stable compared to the group norms. Cognitive-style would become the dependent variable of such a study. This investigation may indicate that it is truly possible to bring about a change in cognitive-style! It may be possible to foster the process of differentiation . . . in such a way that the child's own rich world of exploration becomes more disciplined3 more manageable, and more satisfying." (E.S.S., 1965, p. 4). REFERENCES . An Evaluation Model and its Application. 1 968. Science A Process Approach. American Association for the Advancement of Science. Miscellaneous Publication, 68-4. Asch, S.E., 1956. Studies of independence and conformity (Part 1), a minority of one against a unanimous majority. Psychological Monograph, 70 (Whole No. 416). Anastasi, A., 1968. Psychological Testing. New York: MacMi11 an. Bales, R.F. and Couch, A., 1956. A Factor Analysis of Values. Unpublished Study, Laboratory of Social Relations, Harvard University (cited by Witkin, 1962). Bell, Elaine, G., 1955. Inner-Pirectedness and Other-Pi rectedness. Unpublished doctoral dissertation, Yale University. Bieri, J., Bradburn, Wendy M., and Galinsky, D. , 1 958. Sex differences in perceptual behavior. Journal  of Personality, 26, 1-12. Bogardus, E.S., 1925. Measuring social distance. Journal  of Applied Social Psychology, IX: 299-308. Bruner, Jerome S., 1962. On Knowing, Essays for the Left  Hand. Harvard: Harvard University Press. Butts, Oavid P., 1963. The Pegree to which Children Conceptualize from Science Experiments. Paper read at the 36th annual meeting of The National Association for Research in Science Teaching, Washington, P.C. 79 80 Corey, Stephen M. (Chairman), 1960. Fifty-sixth Yearbook for the National Society for the Study of Education, Inservice Education of Teachers, Supervisors and  Admi ni sters. Chicago, Illinois. Cronbach, L. J. , 1960 (b). Essentials of Psychological  Testing. New York: Harper and Row. Cronbach, L.J., 1966. How can instruction be adapted to individual differences? In R.M. Gagnes (ed.). Learning and Individual Differences. Columbus, Ohio: Merrill Books. Crutchfield, R.S., 1957. Personal and Situational Factors  in Conformity to Group Pressure. Paper read at the 15th International Congress of Psychology, Brussels, Belgium. Davis, S.M., McCourt, W.F., and Solomon, P., 1958. Sensory  Depri vati on: (1) Effects of Social Contact, (2) Effects of Random Visual Stimulation. Paper read at the American Psychiatric Association meeting, Philadelphia. Dawson, S.L.M., 1967. Cultural and physiological influences on spatial-perceptual processes in West Africa, Part II. International Journal of Psychology, 2, 171-185~TBT^ Eccles, Elsie Marie, 1966. Field Dependence and a Neopiage-tian Model of Information Processing Capacity. Unpublished masters.thesis , University of Birtish Columbia. Edwards, A.L., 1957. Techniques of Attitude Scale Con-structi on , New York: Appl eton-Century-Crofts , Inc. Elementary Science, years 1-7, 1969. Province of British Columbia, Department of Education, Division of Curriculum, Victoria. Elliot, Arthur, 1970. Accountability. British Columbia  School Trustee, Vol. 26, No. 5. 81 Fenchel , G.H., 1958. Cognitive rigidity as a behavioural variables manifested in intellectual and perceptual tasks by an outpatient population. Unpublished doctor's dissertation, New York University. Fishbein, Martin, 1967. Attitude Theory and Measurement, New York: John Wiley and Sons, Inc. Frenkel-Brunswick, Else, 1949. Intolerance of ambiguity as an emotional and perceptual personality variable. Journal of Personality, 18, 108-143. Gagne, Robert M., 1963. A psychologists council on curri culum. Journal of Research in Science Teaching, Vol. 1 (27-32). ; Gagne, Robert M., 1963. The learning requirements for enquiry. Journal of Research in Science Teaching, Vol. 1 (144-1 53). Gardner, R.W., Holzman, P.S., Klein, G.S., Linton, Harriet B., and Spence, O.P., 1959. Cognitive control, a study of individual consistencies in cognitive behavior. Psychological Issues, 1 (Whole No. 4). Goodenough, D.R. and Karp, S.A., 1961. Field dependence and intellectual functioning. Journal of Social  Psychology, 63 (241-246). Goodman, Beverly, 1960. Field dependence and the closure factors. Unpublished study (cited by Witkin, 1962). Gordon, B., 1953. An experimental study of dependence --independence in a social and laboratory setting. Unpublished doctor's dissertation, University of Southern California. Gordon, Leonard, A., 1967. Validity of scoring methods for bipolar scales. Educational and Psychological  Measurement, 27 (1099-1 106) . Gottschaldt, K. , 1926. Uber den einfluss der erfahrung auf die wahrnehmung von figuren, I; uber den einfluss gehaufter von figuren auf ihre sichtbarkeit in umfassenden konfigurationen. Pschol . Forsch., 8, (261-317). 82 Gruen, A., 1955. The relation of dancing experience and personality to perception. Psychological Mono graphs, 69 (Whole No. 399). Guetzkow, H., 1951. An analysis of the operation of set in problem-solving behavior. Journal of General  Psychology, 45 (219-244). Guillford, J.P., 1966. Intelligence: 1965 model. American  Psychologist, 21 , 1 (20-26). Hawkins, David. E.S.S. Elementary science activities project. Science Education, Vol. 48, No. 4 (77-78). Hawkins, David. Messing about in science. Science and  Children (Feb. 1965), 5-9). . Introduction to the Elementary Science Study (1965 version). E.S.I. Inc. , Watertown, Mass. . Introduction to the Elementary Science Study (1 966 version). E.S.I. Inc. , Watertown , Mass. Jackson, D.N., 1955. Stability and resistance to field forces. Unpublished doctor's dissertation, Purdue University (cited by Witkin, 1962). Karp, Stephen A. and Konstadt, Norman, L., 1963. The  Children's Embedded Figures Test. Consulting Psychological Press, Paolo Alto, California. Konstadt, N., and Forman, E., 1965. Field dependence and external directedness. Journal of Personality and  Social Psychology, 1: 490-493. Krech, D. , Crutchfield, R.S., and Ballachey, E.L., 1 962. Individual in S o c i ty , New York: McGraw Hill. Linton, Harriet and Graham, Elaine, 1959. Personality correlates of persuasibi 1 ity. In Jam's, I.L., et. al. Personality and Persuasibi1ity. New Haven: Yale University Press. 83 Linton, Harriet B., 1952. Relations between mode of per ception and tendency to conform. Unpublished doctor's dissertation, Yale University. Linton, Harriet and Graham, Elain, 1959. Personality correlates of persuas i bi 1 i ty. In Jam's, I.L., et. al. Personality and Persuasibi1itv. New Haven: Yale University. Miller, James G., 1963. The individual as an information processing system in Information Storage and Neural  Control. Fields and Abbot (eds.) . Springfield, Illinois. Morrison, Phillip and Walcott, Charles. Enlightened oppor tunism. An informal account of the elementary science summer study of 1962. Journal of Research  In Science Teaching. Vol . 1 (48-53) . Morrison, Phillip., Tensions of purpose. E.S.I. Quarterly. Summer 1966 , (67). Nunnally, Jum C, 1967. Psychometric Theory. New York: McGraw Hill Book Company. Overlade, D.C., 1955. Humor: Its Relation to Abstraction. Paper read at Midwest. Psychological Association. Chicago. Pascual Leone, J., 1966. Piaget's Period of Concrete  Operations and Witkin's Field Dependence: A study on college students and children. Paper read at CP.A., Montreal, and U.B.C. (Mimeograph). Pascual Leone, J. 1969. A Mathematical Model for the Transition Rule in Piaget's Developmental Stages. Research report grant NCR APA 234 (Canada). York University. Pemberton, Carol, L., 1952. The closure factor related to temperament. Journal of Personality, 21, (159-1.75) . Podell, J.E., 1957. Personality and stimulus factors in adult cognition: a developmental analysis of decontextualization. Unpublished doctor's disser tation, Clark University. (cited by Witkin, 1962). » 84 Pollack, M. , Kahn, R.L., Karp, E., and Fink, M. , 1 960. Individual differences in perception of the upright in hospitalized psychiatric patients. Paper read at the Eastern Psychological Assocation, New York, (cited by Witkin, 1962) . Quarton, G.C., Evaluating new science materials. E.S.I.  Quarterly (spring-summer 1966), (77). , Report of the International Clearinghouse for Science and Mathematics Curricular Developments. 1968. Lockard, David, Director, (224-241). Scott, William A., 1968. Attitude Measurement. Handbook  of Social Psychology, Vol. 2, (204-273). Tannenbaum, R.S., 1969. Test of Science Processes. Published Servi ces. New York. d by R.S. Tannenbaum, Institute of Health Services. Hunter College. City University of Tannenbaum, R.S., 1969. The Development of the Test of  Science Processes. Paper read at the National Association for Research in Science Teaching. Pasadena, California. Winestine, M.C., 1969. Twinship and psychological differen tiation. Journal of the American Academy of Child  Psychiatry TIT No. 3. (436-455). Witkin, H.A., 1948. The Effect of Training and of Structural Aids on Performance in Three Tests of Space Orientation. Reprint no. 80, Div. Res., CAA. Washington, D.C. Witkin, H.A., and Asch, S.E., 1948a. Studies in space orientation: III. Perception of the upright in the absence of a visual field. Journal of Experi mental Psychology, 38, (603-614). Witkin, H.A. , and Asch, S.E., 1948b. Studies in space orientation: IV. Further experiments on perception df the upright with displaced visual fields. Journal of Experimental Psychology, 38, (762-782). 85 Witkin, H.A., 1949. Perception of body position and of the position of the visual field. Psychological  Monograph, 63, (whole no. 302). Witkin, H.A., Sex differences in perception. Trans. N.Y.  Academy of Science. 12, (22-26). Witkin, H.A., 1949. The nature and importance of individual differences in perception. Journal of Personality, 18, (145-170). Witkin, H.A., and Wapner, S., 1950. Visual factors in the maintenance of upright posutre. American Journal  of Psychology, 63, (31-50). Witkin, H.A., 1950. Perception of the upright when the direction of force acting on the body is changed. Journal of Experimental Psychology, 40, (93-106). Witkin, H.A., 1950. Individual differences in ease of perception of embedded figures. Journal of  Personality, 19, (1-15). Witkin, H.A., 1952. Further studies of perception of the upright when the direction of the force acting on the body is changed. Journal of Experimental  Psychology, 43, (9-20). Witkin, H.A., Wapner, S., and Leventhal , T., 1952. Sound localization with conflicting visual and auditory cues. Journal of Experimental Psychology, 43, (58-67). Witkin, H.A., Lewis, Helen, B., Hertzman, M., Machover, Karen, Meissner, Pearl, B., and Wapner, S., 1954. Personality through Perception. New York: Harper. Witkin, H.A., Karp, S.A., and Goodenough, D.R., 1959. Dependence in alcoholics. Quarterly Journal of  Studies in Alcoholism, 20, (493-504). Witkin, H.A., Dyk, Ruth B., Faterson, Hanna F., Goodenough, D.R., and Karp, S.A., 1 962. Psychological  Differentiation. New York: Wiley. Witkin, H.A., 1965. Some implications of research on cognitive style for problems of education. Arc hivio  di Psicologia, Neurologia e Psichiatria, 26 (1) , (Reprinted and modified in Professional School  Psychology, Vol. Ill, 1969, pp. 198-227). 86 Witkin, H.A., 1965. Psychological differentiation and forms of pathology. Journal of Abnormal Psychology, 70, No. 5, (317-336). Witkin, H.A., Faterson, H.F., Goodenough, D.R., and Birnbaum J., 1966. Cognitive patterning in mildly retarded boys. Child Development, 37, (301-316). Witkin, H.A., 1967. A cognitive-style approach to cross-cultural research. International Journal of  Psychology, 2, No. 4, (233-250). Witkin, H.A., Goodenough, D.R., and Karp, S.A., 1967. Stability fo cognitive style from childhood to young adulthood. Journal of Personality and  Social Psychology, Vol. 7, No. 3, November, Part I, (291-300). Witkin, H.A., Lewis, H.B., Weil, B.A., 1968. Affective reactions and patient-therapist interactions among more differentiated and less differentiated patients early in therapy. Journal of Nervous and Mental  Disease, 146, No. 3, (193-208). Witkin, H.A., 1968. Cognitive patterning in congenitally blind children. Child Development, (767-786). Wolcott, Charles, 1965. From the director. E.S.S. News letter, October. APPENDIX A THE SELECTION OF PERFORMANCE CRITERIA It is an enormous task to attempt to learn something about the nature of children who experience difficulty working within the framework of the Elementary Science Study. Of the many curricular projects which have been adopted by public schools, the author selected the Elementary Science Study, because it had been consistently adopted by many school districts, because it appeared to be one of the most popular programs, and because it was the one program with which the author had had a great deal of personal classroom experience. The selection of the Elementary Science Study as the treatment experience for the subjects in this study, how ever, presented some difficult problems which had to be overcome before the study of the general problem could continue. Determining the important measurable objectives was one such problem and the development of specific test instruments to measure these objectives was another. In the past, teachers of elementary science have evaluated their students by emphasizing in their tests 87 such things as factual.recall, recognition and other low level cognitive processes. With the advent of many new curriculum projects, a new emphasis has been placed on such objectives as the acquisition of scientific and humanistic attitudes and the fostering of creative talent Consequently, evaluation in the traditional sense has become insufficient and in some situations, inappropriate For example, teachers are now instructed to determine ". . . how each child is developing his own skills in scientific investigation" (Elementary Science, 1969, p. 26). Suggestions for carrying out these forms of evaluation include anecdotal reports, checklists, inter views, and paper and pencil tests. This trend away from simply attempting to measure how well children commit to memory facts from textbooks or facts from teacher presentations, is an obviously necessary one. In any case, since curricular objectives have been modified drastically, so must the evaluation techniques be redesigned to measure these new objectives. The literature in science education reveals that many articles have dealt with the problem of developing suit able evaluation techniques. For example, the curriculum which has been developed by the American Association for the Advancement of Science, possesses a specially develop 89 evaluation model which was tailored to meet the specific needs of this program (A.A.A.S., 1970). On the other hand, one of the most prolific and one of the most popular curriculum projects, the Elementary Science Study group, has stated that present evaluation techniques have con tinued to be so inadequate that such measurements might even be a dangerous and inhibiting influence on curriculum innovations. Consequently, this group has done relatively little in developing evaluation instruments and methods (Quarton, 1966, p. 7; Lockhardt, 1967, p. 240). Never theless, attempts must be made to develop more adequate evaluative techniques because school officials are required to justify innovations. This trend toward public account ability appears to be growing (Elliot, 1970). If appro priate evaluation methodologies are not developed, then inappropriate ones may find themselves in use in the absence of more suitable ones. The problem of developing evaluative strategies is further complicated because institutions which have developed such excellent materials as the internationally reknowned E.S.S. group, have been reluctant to state their objec tives in a complete, clear and orderly fashion. The following statement is the list of objectives that were reported to the International Clearinghouse for Science  and Mathematics Curricular Developments (Lockhard, 1967): 90 Behavioral objeotives identified: We have identified some. We feel that if the materials are well designed children will be deeply involved and highly motivated to continue with their owrk. We use such criteria as noise level, general order, attention to the work at hand and design of new experiments by the children. We also have as objectives an increase in  problem solving skills, an improvement in  the ability to predict what will happen  under certain experimental conditions with  ther materials involved. Res earch evidence of obj ectives achieved: Our evidence comes from anecdotal reports from teachers and from close and lengthy observations made by our own staff in classrooms (Lockhard, 1967, p. 241). It merits emphasizing, however, that this curriculum group's rather vague mode of dealing with the question of evaluation is epitomized in the introductory statement, "We have identified some." This vagueness made the task of determining the measurable objectives of this cur riculum very difficult. The writer felt that a necessary requisite for solving the problem of determining objectives should involve the following procedures: 1. Gaining experience with the program by teaching children of all grade levels, utilizing the E.S.S. materials, following the suggestions of guide books and employing E.S.S. teaching strategies. This experience should span a number of years. 91 2. Talking to other teachers and to children about their thoughts about the E.S.S. materials. 3. Studying and analyzing the literature that has been released by the E.S.S. group. Having undergone these procedures, the writer felt that he had the basis for an understanding of the objec tives of the program. His persistent belief was that there were many implicit objectives in addition to those that are listed in the official E.S.S. statement of objec tives.^ For example, after having experience with both E.S.S. material and A.A.A.S. materials, it was rather apparent that the children who were using the E.S.S;. materials were also acquiring practice in using the same types of processes of scientific investigation that the A.A.A.S. 2 group listed as primary objectives. In order to confirm these observations about other implicit objectives of the E.S.S. group, the writer then examined the published units and the following E.S.S. materials with the purpose of identifying the implicit objectives of E,S.S .- materials: This statement of objectives will be further analyzed in Appendix B. 2 In E.S.S. Rocks and Charts, the pupils classify. In Kitchen Physics, the pupils graph, use space-time rela tionships, predict, infer, etc. 92 1. The entire E.S.S. report submitted to the International Clearinghouse for Science  and Mathematics Curricular Development, (1967) . 2. The much cited article by David Hawkins, the former director of the Elementary Science Study. This article is entitled, "Messing About in Science." Hawkins (1965) outlined E.S.S. teaching strategies. 3. The general information bulletins which are entitled Introduction to the Elementary Science Study. Key words were underlined in each sentence. Key words were defined as words which told or described what children actually did in science classes. These words and phrases were then listed on cards, the cards were then shuffled and categorized. The final categorization revealed clusterings of key words which were labelled with the following headings: 1. Beliefs concerning the merits of "messing about" a. "Messing about" is fun. b. "Messing about" will lead the children to pursue and follow up phenomena which are uncovered. c. "Messing about" will lead children to impose a structure on their play. d. "Messing about" will lead children to investigate on their own. 2. Science processes: a. Observing b .- C1 a s s i f y i n g c. Analyzing d. Controlling variables e. Predicting f. Handling data 93 g. Experimenting h. Replicating i. Posing problems j. Acquiring practical skills 3. Creative component: free wheeling speculation, creative problem solving, and intuitive, playful explorati on. 4. Manipulative and building skills 5. Cognitive development a. Specific concept development b. Incidential learnings To avoid this study taking on unmanageable proportions, it was feasible to match only some of these objectives with suitable tests. The writer selected a recent test, The Test of Science Processes (Tannenbaum, 1969), because it appeared to provide a reasonable match to the second clustering of objectives mentioned above. It was reason ably easy to match the type of behaviors which are defined in the blueprint for the Test of Science Processes with those behaviors that occur in E.S.S. situations.^ Because For example, the blue-print defines five behaviors to be tested under the science process, Observing. The following illustrates that it is reasonably easy to match these behaviors described in the blueprint of The Test of Science Processes with E.S.S. activities, Behavior 1. Demonstrate an operational know ledge of the physical properties of objects. 2. Identify and describe objects which interact in a system (Tannenbaum, 1969, Appendix C) Questions EES Units 14,19 Attribute Games 14,19 Rocks and Charts Ki tchen Phys i cs ,~ ,Q Bones , Bulbs & Batteries , Gasses and Airs 94 the E.S.S. group stressed the need for children to acquire the attitudes described above (cluster 1), the author also chose to measure those objectives as well, and as a consequence four Likert-type attitude scales were developed. The techniques involved in developing these scales are fully described in Appendix B. APPENDIX B THE DEVELOPMENT OF FOUR ATTITUDE SCALES TO MEASURE CHILDREN'S ATTITUDES TOWARDS THE AFFECTIVE OBJECTIVES OF THE ELEMENTARY SCIENCE STUDY E.S.S. Literature Corroborates the Importance of the Affective  Objectives Established by the Procedures Listed in Appendix A In order to measure the attitudinal objectives dis cussed in Appendix A, it was necessary for the author to construct four attitude scales. In this appendix, the con clusions of Appendix A will be corroborated by statements from E.S.S. literature. It will be shown that if the aim of the E.S.S. program is to bring about attitude shifts in children, attitude testing is necessary if one is to assess the pupils, the teachers, or the learning environ ment. Following this argument are definitions of attitude and other relevant terms; reasons for the selection of a Likert-type scale; and the procedures employed in develop ing the scales. Finally, the scales in their final test form are included along with relevant statistical data about each scale. 95 96 Besides the usual objectives which emphasize the necessity for children to acquire the skills or processes of science, the E.S.S. group developed materials and strategies which they hoped would develop certain crucial attitudes in children. One could view these materials and strategies as a series of treatments which are supposed to bring about an attitude shift in students. Throughout the E.S.S. materials there are countless references to the effect that E.S.S. experience would bring about the attitude that science is fun to do. Perhaps the most obvious statement of this objective can be found in former E.S.S. Director David Hawkins1 article "Messing About in Science" (1965). Hawkins began his statement about E.S.S. teaching strategies by quoting from Kenneth Graham's poem The Wind in the Willows: 'Nice? It's the only thing, ' said the Water Rat solemnly, as he leant forward for his stroke. 'Believe me, my young friend, there is nothing—absolutely nothing—half so much worth doing as simply messing about in boats. Simply messing, ' he went on dreamily, 'messing—about—in—boats—messing— ' In this article, Hawkins defined three sequences of teaching strategies which he labels O , A , and • . In the o phase Hawkins demanded that there be a substantial amount of time for fun and play: 97 There is a time, much greater in amount than commonly allowed which should be devoted to free and unguided exploratory work (call it play if you wish; I call it work). Children are given materials and equipment--thing s--and allowed to construct, test, probe and experiment without superimposed questions or instructions. I call this 0 phase. 'Messing About' honoring the philosophy of the Water Rat, who absentmindedly ran his boat into the bank, picked himself up, and went on without interrupting the joyous train of thought (Hawkins, 1965, p. 6). It is clear that the E.S.S. group wished children to enjoy science. In their summary of units submitted to the International Clearinghouse for Science and Mathematics Curricular Development (1968), there were no fewer than thirty-one references to an enjoyment factor. One can also find statements in E.S.S. literature which emphasize the intuitive, the imaginative, the playful, and to use Bruner's term, "the left-handed" (Morrison and Walcott, 1962, p. 7). There is ample evidence that the E.S.S. group endeavored to produce materials and teaching strategies which would develop a positive attitude in children toward the belief that "messing about in science is fun." This statement was also the initial belief statement of the first scale. The second attitude scale was constructed around the belief statement that, "From Messing about in Science, 98 children will be lead to pursue phenomena that are revealed." There are many references in E.S.S. literature in which the importance of this attitude is stressed. It was felt that if the materials were appropriate then this attitude of pursuing phenomena would develop: We feel that if materials are well designed, children will be deeply involved and highly motivated to continue with their work, (Lockhard, 1967, p 241). In order for the child to make sense out of the phenomena revealed through "messing about" it is necessary that the quality of the play change and become more rig orous and more structured. Hawkins refers to this process in the learning sequences which he labels^ and Q . In these phases of "messing about" the teacher or a film loop may provide a situation in which an anomaly is made more recognizable. In order to make the anomaly behave, the child must impose a structure on his play. It is hoped that the child may take experiences with things and be able to ". . . analyze them, abstract from them, and perhaps even reach a generality which he can test in other situation" (E.S.S., 1965, p. 9). Clearly the E.S.S. group also wished to foster the attitude that "through messing about Children will impose a structure on their play:" . . . we have found basic agreement that a major aim of our project must be to en  courage children to examine, analyze and  understand the world around them and to  stimulate their desire to continue to do so (E.S.S., 1965, p. 7). This attitude of "imposing structure on play" became the basis for the third attitude scale. Finally, there was considerable evidence in the literary discourse that the E.S.S. group wished to foster the attitude that "through 'messing about' children would investigate independently." For example, even in the brief statement of objectives, there was a reference to the "... design of new experiments by the children" (Lockhard, 1967, p„ 241). It is hoped that children will be able to and inclined to investigate phenomena independently, without the teacher. This attitude that "messing about" will lead the children to be positive toward investigation of phenomena on their own, became the core of the fourth and final attitude scale developed by the author. Definitions of Attitude Indicated that Summated Ratings  Provided' a Workable Model for Attitude Measurement Because this curriculum project aimed to utilize elementary school experience as a vehicle which would 100 help the child become a scientifically curious person (E.S.S., 1965, p. 7), it can be concluded that this aim is primarily an attitudinal one. The author must agree with the emphasis that this curricular project places upon affective objectives. It matters little if teachers are capable of producing students proficient at scientific competencies, if the children do no carry these competencies out of the contrived environment of the classroom into the real world. The term "attitude" has been defined in many ways. In the psychological literature there are many attempts at categorical definitions while many writers prefer to give very limited definitions to the term. These definitions vary with convention and the author's purpose. For example, Krech, Crutchfield and Ballachey (1962, p. 152) define attitude as . . . an enduring organization of motivational3 emotional3 perceptual3 and cognitive process es with respect to some aspect of the individual's world. Newcomb, Turner, and Converse (1965), however, define attitude as "a state of readiness for motive arousal." One attribute of the construct of attitude seems to be important to this study in particular. The 'action tendency' 101 component of attitude seems to be especially important when considering the affective objectives of the Elementary Science Study: Attitudes are commonly distinguished from cognitions3 abilities, capacities, or intel ligence not only by the presence of an affec tive component but also by the conventional assumption that the mere presence of the relevant object is enough to trigger the pre pared response which does not require addi tional motivation. A person who 'knows how' to add will not necessarily do so in the presence of numbers3 but the person who likes to add may be expected to do so when given the opportunity (Scott, 1968, p. 207). Similarly in elementary school science, a person who is able to use the processes involved in scientific investi gation won't necessarily investigate in the presence of suitable materials, but a person who also likes to employ these processes may be expected to do so when given the opportunity. The following quotation from E.S.S. indicates that it is this "tendency for action" which is an essential part of their objectives: . . . we have found basic agreement that a major aim of our project must be to encour age children to examine, analyze and under stand the world around them and to stimulate  their desire to continue to do so. . . . [emphasis added] (E.S.S., 1965 , p. 7) . ". . . their desire to continue to do so" involves the notion of action tendency or motive arousal discussed above. 102 Other elementary science curriculum projects have not only emphasized the importance of asserting affective objectives, but have attempted to develop instruments which would measure attitudes. It is of interest to note that even one of the most structured curricula in ele mentary science, the programme developed by the A.A.A.S. (Science: A Process Approach) has recognized the need to foster the development of positive attitudes and has also initiated the research into the semantic differential technique as it applies to this particular curricular model. In order to evaluate the student's acquisition of the attitudes that have been shown to be objectives of the Elementary Science Study, the author decided to utilize the Likert method of summated ratings. Because the first part of each of the four belief statements involved what Hawkins called "Messing About in Science," summated ratings seemed to be.the most suitable technique to convey the meaning of the term "messing about" as Hawkins used it. It was felt that the very diversity of elementary science activities could be used to convey the notion of "messing about," and that these diverse situations would assist in providing many of the attributes of the construct of attitude that are referred to by Scott (1968, pp. 204-273). 103 Scott listed eleven properties of attitudes: direction, magnitude, intensity, ambivalence, salience, affective salience, cognitive complexity, overtness, embeddedness, flexibility, and consciousness. Scott emphasized that many properties of attitude have not been considered by scale makers: The critical point to be noted is that, if one is to 'measure attitudes ' as they are conceptualized in the literature3 one needs to find ways of operationalizing3 and con verting to numbers such properties as these. In actual practice3 most of them have not been operationalized satisfactorily3 let alone scaled. By far the greatest attention has been devoted to the measurement of magnitude (or intensity) so the ensuing description of measurement procedures will focus exclusively on this property. Com parable measuring procedures could3 in principle3 be applied to most of the other properties as well [direction, magnitude, intensity, ambivalence, salience, affective salience, cognitive complexity, overtness, embeddedness, flexibility, consciousness] (Scott, 1968, p. 208). Realizing that present day attitude scale techniques have many short-comings, the author felt that the Likert technique was probably more appropriate for this study than any of the other methods. Edwards (1959, p. 168) cited evidence that the correlation between Likert Scales and Thurstone Equal Appearing Interval Scales, indicated that there is nearly a perfect relationship between the scores of the 104 two different attitude scales. Edwards concluded that in his particular study: . . . we might predict that the relative ordering of the subjects on either an equal-appearing interval scale or a summated rating scale would be, for all practical purposes, essentially the same.(Edwards , 1 957 , p. 168) The Likert scaling technique was selected by the author after consideration of alternatives. The Likert scaling technique appeared to hold promise of guaranteeing a certain amount of salience. By including belief statements about dozens of actual elementary science activities (what the children really did in science classes) in the assertions about the attitude objects, the author felt that the scales would be pertinent and meaningful measures. By obtaining these statements by means of tapping the belief pool of elementary science students, there seemed also to be a guarantee that the test would provide measures of real attitudes towards real objects in the children's psychological world. Further, the Likert technique provided a situation in which assertions about the attitude object can be placed within the meaningful context of actual classroom behaviors. The author also felt that the Likert technique would best make use of his experience as a practicing elementary school teacher as the task of developing salient items in these 105 scales demands the tacit knowledge of a practicing teacher. Fishbein (1967, p. 395) cited Rosenberg's study in which it was found that . . . estimates of attitudes based on a consideration of an individual 's salient beliefs (i.e., those elicited by the sub ject) were considerably more accurate than estimates based on a consideration of beliefs selected on an a priori basis. Within each scale one can find items that involve many of the properties of attitudes mentioned by Scott. Many of the items reveal the diverse nature of the attitude domain that is being measured by each scale. Some of the items even resemble Bogardus' (1925) social distance measures.^ It has been this writer's persistent belief that when items are developed from the background of practical experience of the test maker and actual experiences of the testing population, the Likert technique provides a workable model for developing attitude measures. Further, the summated rating technique,when utilized as it has been in this study, does not appear to be inconsistent with more recent theories: In addition, it should be noted that according to the theory, [Fishbein's] every time an individual learns a new belief that associates the attitude object with some positively evaluated concept, his attitude will change in a positive direction. Similarly, if the new belief associates the For example, item 17 reads "People who like doing science experiments are creeps." 106 attitude object with a negatively evaluated concept, his attitude change, would be in a negative direction. That is, attitude change, as well as attitude per se, is viewed as a function of the total amount of affect asso ciated with an individual's beliefs about the attitude object. In contrast to this, most theories based on a notion of "consistency" would predict that attitudes and attitude change are functions of the mean amount of affect associated with an individual's beliefs. (Fishbein, 1967 , p. 398).. Summated ratings, as they have been developed for this study, attempted to measure attitude "as a function of the total amount of affect " Procedures Employed in Developing the Likert Scales The construction of the four Likert-type scales initially followed the procedures described by Edwards (1957), with other additional procedures employed as well. More elaborate methods were involved in constructing the final form of the test. This section will indicate how the preliminary version of the scale was drawn from the belief pool of the population to be tested; how this selection was validated by expert judges; and how item analysis of the preliminary scale was carried out. Following this presentation there will be a description of the procedures involved in producing the final scale, and the procedures employed in developing the audio-visual form of the test. 107 Tapping the Belief Pool of the Population Two classes of grade seven children who, in the opinion of the Supervisor of Instruction, had been follow ing correctly the program developed by the Elementary Science Study -- were selected to provide information about the belief pool of the population. A booklet was provided to each of the sixty pupils; on each page the students were asked to respond to one of the following statements: Page 1. What I feel about doing science in school. Page 2. "Messing about" in science is fun and interesting [Scale I]. Page 3. When "messing about" in science, I follow-up things that I notice [Scale II]. Page 4. Do you measure things that you notice? [Scale III] Page 5. By being allowed to "mess about" with things in science, I enjoy being able to figure things for myself [Scale IV]. Page 6. I wish I had more help from my teacher when I am doing science [Scale IV]. Students were told to react freely to these statements. They were informed that they need not comment unless they felt like doing so. Students were also advised that their comments were confidential and had nothing to do with 108 report cards and that their comments were to serve someone at U.B.C. who wanted to know how students felt about doing science in school. Students inquired whether "mess about" meant "horseplay--fooling around." The pupils were in formed that "messing about" meant "playing and using equipment and stuff, like you have been doing for several years in science classes." The children's responses pro vided a large pool of statements about the attitude objects of the four scales. Developing the Initial Set of Items Items were selected from the belief pool and some of these items were modified so that there would be the same number of positive assertions about the attitude object as there were negative assertions. Three outstanding elementary science teachers were called upon to validate whether each statement was indeed an assertion about the attitude object of each particular scale. The judges were provided with a list of the items and were also given the four key assertions about the attitude objects. Judges were asked to judge whether each item was a positive way of saying the assertion about the attitude object, a nega tive way of saying it, or whether the item was not a way of saying the assertion. Judges made a positive mark 109 beside each item, a negative mark beside the item or drew lines through irrelevant items. Judges also made sugges tions which improved, some items. Items for which there was complete agreement as a positive or a negative asser tion were kept intact. Items for which there was only partial agreement were rewritten and resubmitted to the judges. Items that were crossed-out were rejected. The items that survived with the complete agreement of the judges formed the preliminary version of the four attitude scales. Scale one consisted of 40 items, scale two 30 items, scale three 30 items and scale four 43 items. In order to reduce the reading ability factor in interpreting the results of the administration of these scales, a pre liminary taped version of this scale was also produced. Selecting the Final Set of Items The four scales were then administered to thirty-eight grade seven pupils at Dr. H. N. McCorkindale Elementary School. These pupils had also had several years of E.S.S. experience. Students were informed that their responses would be confidential and would not be shown to anyone at their school. Students responded anonymously on answer sheets designed to be mechanically scored. Students were told to show how they "really and honestly felt" about statementswhich described different situations in elementary school science. They were instructed to listen to the taped statement then to read it if they wished and then to respond to a five point agree or dis agree scale. Items were scored 4, 3, 2, 1, and 0. The scoring of negative items was reversed. Scores for each individual were summated. Following the method suggested by Edwards (1957, pp. 152-159), item analysis was then done. The top 33 per cent and the bottom 33 per cent of the subjects with the highest and the lowest scores were assumed to provide criterion groups in terms of which each individual statement could be evaluated. The following "t" ratio was used to evaluate each statement: t = XH " XL S7 S"2 4h h where = the mean score on a given statement for the top group X^ = the mean score on the same statement for the bottom group Ill the variance of the distribution of responses of the top group to the statement the variance of the distribution of responses of the bottom group on the statement the number of subjects in the high group the number of subjects in the bottom group Utilizing U.B.C.'s Triangular Regression Package (TRIP), "t" values for each item were calculated. To be signifi cant at the alpha level of .001, "t"'s needed to exceed 3.045; similarly "t"'s needed to exceed 2.750 for the .01 level and 2.041 for the .05 level of significance.^ With the exception of a few items which approached significance, items which failed to exceed the .05 level of significance were rejected. The two items which approached significance were rewritten and included in the final scale. Fifty-two of the items were significant at the .001 level, 9 at the .01 level, and eighteen at the .05 level. Utilizing a table of random numbers the items that survived were then randomized in order to reduce the possibility of one These critical values of t were determined at 38 degrees of freedom (one-tail). 112 type of response set manifesting itself in the responses of the subjects. The final forms of the four scales each contained twenty items half of which were positive and half of which were negative. The final form of the scales was composed of four separate test booklets. To accompany the visual form of the scale a final tape recording of each statement was also made. Mr. Mark Hartford a trained broadcaster from the audio-visual department in the Faculty of Education, U.B.C, read each item into the recorder in a clear and objective manner. Just enough time was allowed so that a person could listen to the statement, read it, and re spond on the mechanically marked answer sheet--the extent to which he agreed or disagreed. It was felt that this format would assist in making the scales a more pleasurable task to face and also that the taped version would minimize reading difficulties and reading errors. The instructions to the subjects were also given on the tape which helped to standardize the testing situation. The final scales were administered in two sittings to each of seven grade seven classes who have had experience with Elementary Science Study materials over a lengthy period of time. The responses of 184 subjects were then 113 used for further analysis. The response sheets were marked by the U.B.C. Computer Center's IBM 1232 Optical Scanner and responses were punched on cards. The analysis was done on U.B.C.'s IBM 360 facilities. The following special programmes were written by James Gaskill of the Mathematics Education Department at U.B.C: Program One adjusts the scores from the cards produced by the optical scanner and trans forms them to the correct mode and to the proper values of 0, 1, 2, 3, and 4. Program Two reverses the negative items, sums each of the four scales and sums an overall total score as well.^ Both scales are listed at the end of this appendix. Before the administration of the scales, students were told that their reactions to the statements were to be kept confidential. It was emphasized that the researcher wanted to find out how students really felt about the attitude statements. Students were asked to respond honestly rather than in a manner which considered how someone These programs are listed at the end of this appendix. 114 else like their teacher, would expect them to respond. These instructions were repeated in the tape recording.^ Statistical Analysis of the Attitude Tests Analysis was carried out utilizing the principal components factor analysis program which is entitled U.B.C. FACTO. The preliminary factor analysis revealed that there were thirty factors with Eigen values greater than one, and these thirty factors accounted for 70 per 2 cent of the variance. Because the first five factors accounted for much of this variance, principal components factor analysis with varimax rotation was carried out while restricting the number of factors to five. When the Eigen values of the factors were plotted the slope was negligible beyond five factors. The four attitude total scores and the grand total were also included as variables in this analysis in order to determine the factoral composition of the scales. It was hypothesized that the four scales would each be composed of mainly one unique factor. To some degree this hypothesis was upheld. ^Many of the students remarked that they found the taped instructions and scale items very helpful. o Nunnally (1 967 , p. 256) suggested that one can expect a large number of "factors" with 184 subjects and 80 items. i 115 An attempt was made to identify and label these factors by determining the common logical attributes of those items which possessed common factor loadings. Consequently, Factor A has been labelled "the fun factor"; Factor B "the insecurity factor"; Factor C "the imposing structure factor"; Factor D "the independence factor"; Factor E "the ego identification with science factor." Factor E appears to be the least important factor in all four of these scales. These factor loadings are summarized on the following page. It can be seen that Scale I is loaded primarily on "the fun factor." Scale II was judged to be logically unique, however it appears to be a com posite scale factorally. Scale II has been constructed to be logically different from the other three scales. It can be argued that the three major factors in the scale combine to produce a unique scale. Just as the color green can be shown to be composed of two primary colors, the color green, is still green -- neither yellow nor blue, but a unique color green! So too, Scale II is composed of certain combinations of Factors A, C, and D. Nevertheless the scale still stands as a logically unique entityJ Scale II has demonstrated that it is a ^Perhaps Scale II could be referred to as a secondary scale factorally, while the other scales could best be described primary scales, factorally. 116 Table Bl Factor Loadings on the Four Attitude Scales Scale A B C D E I .82106 -.33248 - .22851 .10787 .19605 II .69923 .19121 -^36221 .36948 .20400 III .21994 -.08511 -.89052 .30252 .16000 IV .24947 -.59813 .26475 .62628 .23662 Total .57830 -.35985 -.53364 .43256 .23850 117 sensitive instrument that discriminates differently from the other scales. Furthermore, Scale II has an extremely high internal consistency.^ The differences between Scale II and the other scales are subtle, but then too -- so is the attitude domain being tested. Scale III is factorally unique, and it is loaded primarily on Factor C. the appropriate factor. Scale IV is extremely interesting, factorally: as this scale appears to be composed of two opposing factors. The first component is Factor D (independence) which is positively correlated with this scale; the second factor is Factor B (insecurity) which is negatively correlated with this scale. In general, Scales I, II, and IV have been shown to be factorally unique. This finding enhances the validity of the scales. Item analysis was done by computing the Product Moment Correlation Coefficient between the scores of persons on each item versus the scores of persons on each appropriate total score. These correlations can be found in the left hand margin, beside each item in the scales which follow below. Although several items in each scale do not correlate highly with the total score, most items appear to correlate quite well. The internal reliability ^The Cronbach Alpha Coefficient for Scale II is .8239. 118 of each scale confirms the conclusion that it would not be worthwhile discarding these few items for the purposes of this study. Further refinements could be made at a later date however. The internal reliability coefficient selected to provide an index of homogeneity was Cronbach's Coefficient, Alpha (Cronbach, 1957, p. 161). The author computed these coefficients on a calculating machine, utilizing the standard deviations of individual items and total scale which was produced as output by U.B.C. FACTO. These standard devia tions were squared and the values were substituted into the following formula: Oj - EC2 K 01 = ITT Where K = the number of items in a scale I = the item T = the total or subscale total The following table summarizes the findings; the table indicates that the scales appear to be highly reliable i ns truments: 119 Table B2 Alpha Coefficients, Means and Standard Deviations of the Four Scales COEFFICIENT ALPHA Scale K al T a Mean S. D. I 20 35.15452 130.36433 .7687 54.26 11 .42 II 20 38.65241 177.95560 .8239 50.22 13.35 III 20 42.34158 185.03752 .8117 43.81 13.58 IV 20 40.41436 183.41962 .8207 45.12 13.54 TOTAL 80 156.56287 1881.10870 .9284 193.4 43.37 Finally, the four scales were correlated with each other, first for the total group and then for boys and then for girls. The resulting correlation matrices reveal that the four domains are more interconnected for the boys and more discrete entities for the girls. Computing the signifi cance of the differences between pairs of correlations in the first figure below indicates that five of the six pairs of intercorrelations for boys and for girls are significantly different. This finding would also lend support to the validity 120 of these attitude instruments. It is the writer's hope that these scales will be useful both to researchers and to practicing classroom teachers. Below are the intercorrelations among scales. This figure is followed by the actual test form. The item analysis and factor analysis of items preceeds the special computer programs referred to above. Table B3 Intercorrelations of The Attitude Domains by Sex GIRLS (88) Total Att. I II III IV Intercor relation I \l. 1 Xv .6759 .3812 .5249 .8005 A * Al C * Ll II .7445 \ 1 • iX .5085 .4847 .8405 III .5886 .6791 X 1 • 1 \ .4402 .7447 A * 2 * B2 bl IV .6391 ,7122 .6698 X 1 • i X .7803 C * L2 „ ** D2 ** t2 Total -8411 .9030 .8533 .8739 sig. different at .05. sig. different at .01. Table B3 (cont'd) Total Att. .8227 .8758 .8081 .8331 123 THE FOUR E.S.S. ATTITUDE SCALES I (1-20) "Messing about in science" is fun. (Fun Scale) II (21-40) "Messing about" leads the child to pursue (follow-up) phenomena that are noticed. (Pursue Scale) III (41-60) "Messing about" leads the child to impose a structure on his play. (Structure Scale) IV (61-80) "Messing about" leads the child to investigate on his own. (Independent Investigation Scale) 124 Please do not write on the blue pages. This is not a test for marks. There are no right or wrong answers. DIRECTIONS: Below are some statement about playing and experimenting in science. We would like to know how you really feel about them. Read each statement carefully, as I read it to you. Then decide whether you (1) agree a lot, (2) agree a little bit, (3) don't know how you feel about it, (4) disagree a little bit, or (5) disagree with it a lot. EXAMPLE: Grade six and sevens should get paid for coming to school. 1 . I agree a 1ot. 2. I agree a little 3. I don't know. 100 •••OD 4. I disagree a little bit. 12 3 4 5 5. I disagree a lot. Now look at the red answer sheet. Look at row 100 on this sheet. Fill in your answer by blackening one of the boxes. If you don't understand, raise your hand. Now do example 101: Grade six and sevens should not get paid for coming to school. ioi ••••• 1 2 3 4 5 Any questions? Now we are ready to being the statements. 125 FORM W 1. Doing science experiments is interesting 2. In science I have fun with stuff and it's interesting 3. Experimenting is fun. 4. Far too much time is wasted "just playing" with things in science. 5. I don't really like experimenting because I often don't know if I'm getting the right answer. 6. Doing science makes me notice that there are many beautiful things in the world. 7. Science makes me feel dumb. 8. By playing with batteries it helps me to get at all sorts of answers to question that bug me. 9. I'm glad when science periods are over Check your answer sheet to make sure that you are you are filling in the proper space. 126 FORM W (cont'd) . 10. I wish we could play around with things in all subjects as we can in science. 11. A person should want to do science experiments so that he can learn about things that he has wondered about. 12. In science experiments, I don't have to pretend -I can be myself. 13. The sooner that I can forget about science experiments the happier I am. 14. Experiments are a bore 15. Some things in science are beautiful and strange 16. A person really doesn't learn much by fooling around with things in science. 17. People who like doing science experiments are creeps 18. Science experiments are enjoyable 127 19. I sometimes brag a little at home about what I did in science. 20. Science experiments are never really fun. 128 FORM X 21. Fooling around with things makes me want to learn more about them. 22. I'm still experimenting and thinking about something that I noticed in science a long time ago. 23. I don't care about why things happen in a science experiment. 24. A person doesn't get many ideas for an exper iment from handling equipment. 25. If I noticed that a ball seemed to bounce about the same number of times if I dropped it from different heights, I'd go on and study something else. 26. It is nice to think about ways of discovering answers in experiments. 27. I really don't care why things happen the way they do. 129 FORM X (cont'd) 28. Anyone who goes to the library to get books about what he's noticed while playing around with things in science is a jerk. 29. If I can't find out why some strange things happen, it really bugs me. 30. If something unexpected happens in an experiment at another table, I don't think I'd bother going over there. 31. I want to discover more answers to things that bug me when we begin to experiment. 32. I'll work for hours on a science project if I think I've almost got an answer. 33. I wish we took a different unit every day. 34. I can think of a time when I did an experiment on my own because of something that I noticed. 130 FORM X (Cont'd) 35. Experiments are a challenge and I like to find out as many things as I can, before I go on to something else. 36. I don't think about science stuff unless I'm in class. 37. Although I know I should follow up more - from things that I notice in science, I usually don't bother. 38. If something is interesting I want to know what makes it tick even if it's hard work. 39. Once I've been introduced to an idea, I like to follow it up in an experiment. 40. Playing with things and messing around with things does not make me curious enough to  experiment with them. 131 FORM Y 41. I hate trying to figure out why things work; I'd sooner just play with them and then forget them. 42. I often make up my own names for things so that I can remember and compare. 43. If you measure a lot, you discover things that you never noticed before. 44. Blowing bubbles is O.K. until the teacher starts asking a lot of question. 45. I hate trying to discover rules about why things happen in a certain way. 46. The confusion when I begin to experiment soon goes away as I plan what I'm going to do. 47. I think that for me to plan an experiment is a waste of time. 48. Playing with things is O.K. but I like to plan ways to find out more of the detail. 132 FORM Y (cont'd) 49. After playing with ice melting in water, I'd like to measure the temperature and graph how the temperature changes as the ice melts. 50. People overdo all this "measuring stuff" in sci ence. 51. I'd rather think of things just as they are rather than by thinking about every part of them. 52. It's fun making up rules which might explain things you notice when you mess around with science stuff. 53. Having fun and measuring in science are two very different things. 54. If I experimented with pendulums, I would want to use a ruler and a timer. 55. If I was trying to find out how a mealworm explores a box I'd like to measure and record where he goes. 56. Experimenting can be fun, except I hate measuring and compari ng. 133 FORM Y (cont'd) 57. When I experiment, I like to keep some sort of record in my book so I can compare things. 58. Science is fun until you have to compare things exactly. 59. I like discovering a pattern in something which didn't seem to have one. 60. Science would be more fun wi thout rulers , graphs , and timers. 134 FORM Z 61. When I begin a new experiment I really get bugged when someone makes me stop. 62. It's neat to start right from the beginning of an experiment, doing everything for yourself. 63. I like subjects where the answers can be found easily in a book. 64. I'd like science a lot better if the teacher showed everyone how to do every experiment. 65. I wouldn't want to work with someone who usually told me the right answers. 66. I don't enjoy giving in and letting others do the work when we do an experiment. 67. I do extra experiments on my own. 68. I don't like the teacher to give away too many hints. 135 FORM Z (cont'd) 69. I wish the teacher would help me more so that I can do the right thing. 70. I'd sooner sit around and talk than play around with things in science. 71. I don't really like finding things out on my own 72. I really like to watch the teacher do an experiment, instead of me doing one. 73. It's more fun hearing about science than doing it, 74. I like it best when I'm told how to do the experiment exactly so that I know how to find the right answer. 75. If my friends thought that my ideas were crazy, I don't think I'd say them. 76. It would be great to have more time to work on experiments that you choose and figure out on your own. FORM Z (cont'd) I'd rather do my very own experiments instead of watching the teacher do one. Things get too confusing unless my teacher helps me. I don't like the teacher to give away many clue. I do a lot of experiments at home 137 Scales with item-per-scale correlations (on left margin). Factors with Eigen values are labelled A, B,C, D, and E, (these are located in columns). These factors seem to fit logically into the following classification: A. FUN -B. INSECURITY WHILE INVESTIGATING ON YOUR OWN C. IMPOSING A STRUCTURE ON INVESTI GATIONS -D. INVESTIGATION BY "DOING YOU OWN" THING" NOTE: Although Scale II does not appear too unique in factor analysis, it is logically different and has shown to discriminate differently from Scale I. E. EGO IDENTIFICATION WITH SCIENCE contributes little to the scales. Load + Load 00 CO FORM W FACTORS: (A) (B) (C) (D) (E) ITEM-TESl CORRELA TIONS QUESTION FUN 17.52 INSECURITY (i nvest) 3.60 IMPOSING STRUCUTRE ON PLAY 3.27 INDEPENDENCE (i nvest) 2.72 EGO 2.60 .54623 1 . Doing science experiments is i nteres ti ng. .60058 -.0555 -.06912 .06932 .09933 .53985 2. In science I have fun with stuff and it's interesting. .42337 -.18112 -.33764 -.03545 .17773 .51513 3. Experimenting is fun. .56009 -.16075 - .06260 -.00647 .12338 .19839 4. Far too much time is wasted "just playing with things in science. .14409 -.02038 - .02904 -.09922 - .24244 .49467 5. I don't really like experimenting because I often don't know if I'm getting the right answer. .38446 -.32006 -.20142 -.14893 .04735 .36343 6. Doing science makes me notice that there are many beautiful things in the world. .32437 .06632 -.21060 -.02250 .15910 .38284 7. Science makes me feel dumb. .22080 -.32647 -.08897 .1900*6 .01387 CO FORM W (Cont'd) I.T.C. QUESTION (A) (B) (C) (D) (E) .29686 E 8. By playing with batteries it helps me to get at all sorts of answers to questions that bug me. .16883 .07882 - .19906 -.11107 .30352 .53719 9. I'm glad when science periods are over. .38855 -.17182 -.13139 .03742 .18560 .23168 E 10. I wish we could pjay around with things in all subjects like we can in science. .14967 .08139 .19179 -.03607 .34571 .36284 11 . A person should want to do science experi ments so that he can learn about things that he has wonder about. .41204 .04046 - .05171 -.03529 -.14034 .33028 12. In science experiments, I don't have to pretend - I can be myself. .32510 .07855 -.00739 .01929 .10385 .58122 13. The sooner that I can forget about science experiments the happier I am. .39231 -.36395 -.23597 .17215 -.00844 .60703 14. Experiments are a bore. .55657 -.19212 -.18754 .22554 .01759 .30313 15. Some things in science are beautiful and strange. .33081 -.02382 .03720 .03796 -.01085 .36236 B 16. A person really doesn't learn much by fooling around with things in science. .22650 -.38355 -.00196 -.12109 -.01028 .53448 17. People who like doing science experiments are creeps. .43134 -.20635 -.70765 .37075 -.13102 I.T.C. QUESTION (A) (B) (C) (D) (E) .62053 18. Science experiments are enjoyable. .60991 -.07862 - .19459 .26091 .13742 .40271 E 19. I sometimes brag a little at home about what I did in science. .19193 -.20764 - .06118 .07874 .42429 .63294 20. Science experiments are never really fun. FORM X .58105 - .26814 -.11714 .18802 .01292 .36884 21 . Fooling around with things makes me want to learn more about them. .39345 .16761 -.06777 .12394 .11566 .41060 E 22. I'm still experimenting and thinking about something I noticed in science a long time ago. .13265 .03092 .01515 .22433 .44590 .59575 23. I don't care about why things happen in a science experiment. .34327 -.23911 -.26478 .26880 .03828 .33445 24. A person doesn't get many ideas for an experiment from handling equipment. .29694 -.09150 -.01605 - .01747 .06173 .24208 25. If I noticed that a ball seemed to bounce about the same number of times if I dropped it from different heights, I'd gon on and study something else. .1 0771 -.25867 -.31402 -.10458 - .1 2563 .57977 26. It is nice to think about ways of dis covering answers in experiments. .55977 .09358 -.14776 .22145 .22265 .60992 27 . I really don't care why things happen the way they do. .49632 -.29880 -.17366 .15079 .05034 FORM X (cont'd) I.T.C. QUESTION (A) (B) (C) (D) (E) .51067 28. Anyone who goes to the library to get books about what he's noticed while playing around with things in science is a jerk. .41633 - .07492 -.17052 .35282 - .26935 .39604 29. If I can't find out why some strange things happen, it really bugs me. .43375 ,12817 .01007 .14161 -.04007 .35485 30. If something unexpected happens in an experiment at another table, I don't think I'd bother going over there. .36753 -.07649 -.02421 .29746 - .39016 .59290 31 . I want to discover more answers to things that bug me when we being to experiment. .50157 .05581 -.18161 .22622 .22461 .51863 32. I'll work for hours on a science project if I think I've almost got an answer. .22594 -.09594 -.19170 .25493 .42867 .43051 C 33. I wish we took a different unit every day. .16425 -.19220 -.29848 .27414 -.27107 .38263 34. I can think of a time when I did an experiment on my own because of something that I noticed. .07367 -.12462 -.07236 .22517 .39285 .61475 35. Experiments are a challenge and I like to find out as many things as I can, before I go on to something else. .52054 .02980 -.19344 .26333 .26967 .58758 36. I don't think about science stuff unless I'm in science class. .36875 -.34019 -.19492 .14172 .08949 CM I.T.C. QUESTION (A) (B) (C) (D) (E) .47451 37. Although I know I should follow up more -from things that I notice in science, I usually don't bother. .31062 -.22379 -.3781 0 - .06808 .1 2690 .60654 38. If something is interesting I want to know what makes it tick even if it's hard work. .41722 -.01908 -.32780 .09542 .26979 .61892 39. Once I've been introduced to an idea, I like to follow it up in an experiment. .34686 - .00945 - .25664 .37986 .34743 .47744 40. Playing with things and messing around with things does not make me curious enough to experiment with them. FORM Y .40806 -.18300 -.20060 .10889 -.07108 .55789 41 . I hate trying to figure out why things work; I'd sooner just play with them and then forget them. .13501 -.26619 -.48816 .23840 .04915 .25789 42. I often make up my own names for things so that I can remember and compare. .01873 .03843 - .23849 -.09382 .18265 .27994 43. If you measure a lot, you discover things that you never noticed before. .11224 .27844 -.17150 .08342 .26516 .47828 44. Blowing bubbles is O.K. until the teacher starts asking a lot of questions. .01461 -.18070 -.51885 -.04214 -.03676 ro I.T.C. QUESTION (A) (B) (C) (D) (E) .48097 45. I hate trying to discover rules about why things happen in a certain way. .03647 -.20694 -.40948 .19730 .13960 .40528 46. The confusion when I begin to experiment soon goes away as I plan what I'm going to do. .13258 .19072 -.34189 .14161 .04407 .47630 47. I think that for me to plan an experiment is a waste of time. .27585 -.17883 -.35147 .26655 -.08953 .48205 48. Playing with things is O.K. but I like to plan ways to find out more of the detail. .17138 .18232 -.41290 .25148 .23467 .54235 49. After playing with ice melting in water, I'd like to measure the temperature and graph how the temperature changes as the ice melts. .15393 .21927 -.48391 .25680 .08562 .46992 50. People overdo all this "measuring stuff" in science. .4018 -.18584 -.42299 .17641 .0079 .57284 51 . I'd rather think of things just as they are rather than by thinking about every part of them. .10859 -.19702 -.53061 .1 0631 .04816 .42064 52. It's fun making up rules which might explain things you notice when you mess around with science stuff. .05087 .02204 -.33312 .23297 .21521 .43806 53. Having fun and measuring in science are two very different things. .10637 -.27228 -.34649 .03989 .1788 I.T.C. QUESTION (A) (B) (C) (D) (E) . 3751 6 54. If I experimented with pendulums, I would want to use a ruler and a timer. .18053 .16790 -.37070 .02185 - .05394 .52413 55. If I was trying to find out how a mealworm explores a box I'd like to measure and record where he goes. .17261 .16672 -.43277 .25422 .10650 .56373 56. Experimenting can be fun, except I hate measuring and comparing. .00965 -.07877 -.61715 .1325 -.07345 .66362 57. When I experiment, I like to keep some sort of record in my book so I can compare things. .1750] - .03897 -.62949 .22046 .05021 .43875 58. Science is fun until you have to compare things exactly. .10351 -.08544 -.45483 -.07462 .10951 .46137 59. I like discovering a pattern in something which didn't seem to have one. -.00308 -.00557 -.31490 -.39851 .19010 .62161 60. Science would be more fun without rulers, graphs, and timers. .12810 .26633 .57229 .24052 - .07036 LO FORM Z I.T.C. QUESTION (A) (B) (C) (D) (E) .49645 61 . When I begin a new experiment I really get bugged when someone makes me stop. .19069 - .01778 -.07447 .49868 .25444 .54753 62. It's neat to start right from the beginning of an experiment, doing everything for yourself. .18959 -.16980 - .14220 .49871 .15386 .51059 63. I like subjects where the answers can be found easily in a book. .07880 -.52900 -.30542 .06967 .10593 .42587 64. I'd like science a lot better if the teacher showed everyone how to do every experiment. -.02187 -.48862 -.04199 .20738 -.15674 .31766 65. I wouldn't want to work with someone who usually told me the right answers. -.02927 .04357 -.15332 .42387 -.10251 .48508 66. I don't enjoy giving in and letting others do the work when we do an experiment. .23687 .03218 -.07882 .61183 .02819 .45032 67. I do extra experiments on my own. .07386 -.31926 - .18378 .01629 .56325 .55538 68. I don't like the teacher to give away too many hints. -.01630 -.25859 -.02161 .65477 .04121 .38543 C 69. I wish the teacher would help me more so that I can do the right thing. .09760 .54643 .02235 .10558 .10211 I .T.C. QUESTION (A) (B) (C) (D) (E) .51362 C 70. I'd sooner sit around and talk than play around with things in science. .25329 -.22626 - .42503 .19834 .05435 .52831 71 . I don't really like finding things out on my own. .06190 - .39102 -.18319 .27233 .13643 .62820 72. I really like to watch the teacher do an experiment, instead of me doing one. .27863 -.36623 -.10943 .48832 -.02372 .50326 73. It's more fun hearing about science than doing it. .34716 -.30990 -.15833 .25156 .01276 .48872 74. I like it best when I'm told how to do the experiment exactly so that I know how to find the right answer. .10521 -.63646 -.06714 .01808 .09164 .42137 75. If my friends thought that my ideas were crazy, I don't think I'd say them. .14423 -.40356 .01292 .13940 .0092 .44772 76. It would be great to have more time to work on experiments that you choose and figure out on your own. .25968 -.05259 -.21380 .39406 .25460 .53559 77. I'd rather do my very own experiments instead of watching the teacher do one. .18036 -.18205 -.05547- .46840 .05167 .43482 78. Things get too confusing unless my teacher help me. -.04618 -.49689 -.1223 .15700 .04529 I.T.C. QUESTION (A) (B) (C) (D) (E) .49953 .37842 79. I don't like the teacher to give away many clues. 80. I do a lot of experiments at home. .10186 .11532 -.19182 -.25130 -.0666 -.14246 .57139 -.004 .05096 .6506 148 PROGRAMME ONE This programme converts the optical scanner output to proper-size mode. The blank items become scored as undecided. Data is then written according to the format (12X, 60A1/12X, 60A1). Logical unit "6" tells where this modified data is to be written (in this case in file "A"). $RUN *F0RTRAN DIMENSION 1(80) , M(l1 ) DATA M/,0,,,1I,,2',,3,,I4,>,5I,,6,>,7I,,8,JI9,>' '/ 5 READ (5, 10, END=80)I 10 FORMAT (12X,60A./12X,60A1) DO 20 J=l , 79 , 2 IF(I(J).EQ.M(11))I(J)+M(3) IF(I(J+1).EQ.M(11))I(J+1)=M(8) DO 20 K=6, 10 IF(I(J+1).EQ.M(K))I(J+1)=M(K=5) 20 CONTINUE WRITE (6,10)1 G0T05 80 STOP END $ENDFILE $RUN-L0AD# 6=A PROGRAMME TWO This programme reverses the negative items and then sums the four attitude scales independently and also gives a total score as well. Following the $Endfile card is the list of negative items (items for reverse scoring). The original responses are read from a file entitled -B. Cl 'Subjec' is a vector of items whose value is to be reversed. Cl 'Data' is a matrix of all the data with a row for each student. Cl 'Data' is a matrix of all the data with a row for each student and a column for each questi on. Cl "tot" is a vector of subtest scores. C2 Read in items to be reversed. C6 Read in data. C9 Reverse the required items. $RUN *F0RTRAN INTEGER SUBJEC(80) , DATA(184 ,80 ) , TOT(5) READ (5,5,END=10) (SUBJEC(I), 1 = 1 ,80) 5 F0RMAT(13) 10 NITEMS=I-1 DO 25 1 = 1 ,184 READ (6 ,1 5) (DATA(I,J) ,J = 1 ,80) 150 15 F0RMAT(12X,60Il/12Xa60n) (or any modification) DO 20 J = l , NITEMS DATA (I-SUBJEC(J)) = IABS(DATA(I ,SUBJEC(J))~4) 20 CONTINUE DO 220 J = l ,80 220 DATA(I,J)=IABS(DATA(I,J)-4) T0T(5)=0 M=l J = l K=20 21 T0T(M)=0 DO 22 L=J,K TOT(M)=TOT(M)+DATA(I ,L) 22 CONTINUE M=M+1 IF(M,GE.5) GO TO 23 J = K+1 K=K+20 GO TO 21 23 DO 24 M=l ,4 T0T(5)=T0T(5)+T0T(M) 24 CONTINUE WRITE(8,26) TOT 26 F0RMAT(5I7) 25 CONTINUE DO 40 1 = 1 ,184 WRITE(6,30) (DATA(I ,J),J = 1,80) 30 F0RMAT(12X,60I1/12X,60I1 ) (or any modification) 40 CONTINUE STOP END $RUN - LOAD# 6=-B 8=-H 4 5 7 9 13 14 16 These are the negative items: 17 one 13 format numeral on each 20 card. 23 24 25 27 28 (list of items for reversed scoring) 30 33 36 37 40 41 44 45 47 50 51 53 56 58 60 63 64 69 70 71 72 73 74 75 78 ENDFILE 152 APPENDIX C ITEM AND TEST ANALYSES FOR C.E.F.T. AND THE TEST OF SCIENCE PROCESSES Table CI Item Analysis of the Chi 1drens'  Embedded Figures Test Item No. Point Biserial P Vari ance 1 0.4388 0.7486 0.1882 2 0.3266 0.6667 0.2222 3 0.3960 0.6393 0.2306 4 0.1362 0.8251 0.1443 5 0.2796 0.8470 0.1296 6 0.4962 0.6339 0.2321 7 0.3922 0.8962 0.0930 8 0.1346 0.3770 0.2349 9 0.3310 0.8689 0.1139 10 0.4503 0.8197 0.1478 11 0.5087 0.5738 0.2446 12 0.4438 0.8525 0.1258 13 0.4414 0.4044 0.2409 14 0.4766 0.6831 0.2165 15 0.4612 0.8415 0.1334 16 0.4506 0.7978 0.1613 17 0.5295 0.6721 0.2204 18 0.4865 0.7705 0.1768 19 0.4952 0.5410 0.2483 20 0.3494 0.5355 0.2487 The Mean is 13.9945 The KR-20 is 0.7238 The standard deviation is 3.4665 Conti nued 153 * The first five items have been omitted from the analysis of items because of the method suggested in the manual for the administration of C.E.F.T. to this age group. Considering the total test of twenty-five i terns: Total Group: X = 18.86 a = 3.58 Boys Group: I = 18.77 a = 3.71 Girls Group: X = 18.961 a = 3.46 ^Note: there were no "clear-cut" sex differences on C.E.F.T. scores 154 Table C2 Item Analysis for Subtests of the Test of Science Processes D6: OBSERVING Item No. Point B i s e r i a 1 P Variance 13 0.4378 0.7717 0.1762 14 0.5340 0.7065 0.2073 15 0.489 3 0.5380 0.2486 16 0.4148 0.6957 0.2117 17 0.3075 0.7446 0.1902 18 0.5512 0.5815 0.2434 19 0.5269 0.4457 0.2470 20 0.5471 0.2935 0.2073 21 0.4636 0.5543 0.2470 The Mean is 5.3315 The KR-20 is 0.5736 The Standard Deviation is 2.0092 155 D7: COMPARING Item No. Point Biserial P Vari ance 3 0.5938 0.5761 0.2442 7 0.5249 0.7772 0.1732 22 0.5122 0.8967 0.0926 23 0.4426 0.3315 0.2216 24 0.5019 0.9076 0.0839 The Mean is 3.4891 The KR-20 is 0.2730 The Standard Deviation is 1.0214 D8: CLASSIFYING Item No. Point Biserial P Variance 1 0.2741 0.9511 0.0465 2 0.2202 0.8587 0.1213 4 0.3305 0.7880 0.1670 5 0.3470 0.7174 0.2027 6 0.3850 0.9348 0.0610 9 0.4857 0.2283 0.1762 10 0.4803 0.2989 0.2096 11 0.4417 0.4293 0.2450 25 0.2689 0.8859 0.1011 26 0.4088 0.8750 0.1094 27 0.4165 0.7772 0.1732 28 0.3203 0.7772 0.1732 29 0.4387 0.6630 0.2234 The Mean is 9.1848 The KR-20 is 0.4798 The Standard Deviation is 1.8993 156 D9: QUANTIFYING Item No. Point Biserial P Variance 30 0.3579 0.9674 0.0315 31 0.3326 0.8098 0.1540 32 0.5544 0.6957 0.2117 33 0.4156 0.6685 0.2216 34 0.2835 0.8641 0.1174 35 0.5774 0.6685 0.2216 36 0.5104 0.8478 0.1290 37 0.3578 0.3696 0.2330 38 0.4799 0.6739 0.2198 39 0.4446 0.8804 0.1053 40 0.4647 0.8207 0.1472 41 0.4209 0.9620 0.0366 The Mean is 9.2283 The KR-20 is 0.5911 The Standard Deviation is 1.9978 157 D10: MEASURING Item No. Point Biserial P Variance 42 0.4759 0.6359 0.2315 43 0.3269 0.9130 0.0794 44 0.4277 0.7174 0.2027 45 0.2692 0.7337 0.1954 46 0.3493 0.8207 0.1472 47 0.2359 0.6793 0.2178 48 0.4594 0.7065 0.2073 49 0.3921 0.4239 0.2442 50 0.3447 0.7174 0.2027 51 0.4251 0.6250 0.2344 52 0.2209 0.1957 0.1574 53 0.2711 0.1957 0.1574 54 0.4812 0.7228 0.2003 55 0.5031 0.7337 0.1954 56 0.4499 0.6087 0.2382 57 0.4716 0.5054 0.2500 58 0.3902 0.5707 0.2450 59 0.5072 0.5978 0.2404 60 0.1539 0.6087 0.2382 61 0.3486 0.7663 0.1791 62 0.3486 0.7663 0.1791 63 0.5401 0.5380 0.2486 64 0.4889 0.6359 0.2315 65 0.2117 0.2772 0.2003 66 0.3292 0.3261 0.2198 The Mean is 14.6848 The KR-20 is 0.7567 The Standard Deviation is 4.3637 158 Dll: EXPERIMENTING Item No. Point Biserial P Vari ance 67 0.3516 0.4674 0.2489 68 0.2911 0.5109 0.2499 69 0.1763 0.3152 0.2159 70 0.3873 0.3859 0.2370 71 0.5113 0.4511 0.2476 72 0.4610 0.4891 0.2499 74 0.3955 0.6032 0.2393 75 0.4536 0.6739 0.2198 76 0.4133 0.6033 0.2393 77 0.3610 0.5326 0.2489 The Mean is 5.0326 The KR-20 is 0.3453 The Standard Deviation is 1.8647 159 D12: INFERRING Item No. Point Biserial P Vari ance 12 0.3592 0.3641 0.2315 73 0.4162 0.2446 0.1848 78 0.2709 0.3967 0.2393 79 0.2531 0.4837 0.2497 80 0.4015 0.4924 0.2415 81 0.2245 0.2065 0.1639 82 0.4031 0.8641 0.1174 83 0.3504 0.5272 0.2493 85 0.4336 014891 0.2499 86 0.3335 0.3098 0.2138 92 0.4541 0.4978 0.2404 94 0.4773 0.4402 0.2464 95 0.2978 0.4163 0.2497 96 0.4921 0.6359 0.2315 The Mean is 6.6685 The KR-20 is 0.5090 The Standard Deviation is 2.4281 160 D13: PREDICTING Item No. Point Biserial P Vari ance 8 0.4170 0.7120 0.2051 84 0.4063 0.4457 0.2470 87 0.4212 0.6793 0.2178 88 0.4843 0.6957 0.2117 89 0.4570 0.3967 0.2393 90 0.4990 0.3587 0.2300 91 0.4730 0.6033 0.2393 93 0.2401 0.4946 0.2500 The Mean is 4.3859 The KR-20 is 0.3459 The Standard Deviation is 1.6245 161 Item Analysis of the Total Test of Science Processes D14: PROCESSES Item No. Point Biserial P Variance 1 0.1818 0.9511 0.0465 2 0.2005 0.8587 0.1213 3 0.1808 0.5761 0.2442 4 0.2514 0.7880 0.1670 5 0.2547 0.7174 0.2027 6 0.2792 0.9348 0.0610 7 0.3221 0.7772 0.1732 8 0.1334 0.7120 0.2051 9 0.1999 0.2283 0.1762 10 0.1804 0.2989 0.2096 11 0.2233 0.4293 0.2450 12 0.2870 0.3641 0.2315 13 0.3984 0.7717 0.1762 14 0.4307 0.7065 0.2073 15 0.4341 0.5380 0.2486 16 0.2652 0.6957 0.2117 17 0.2798 0.7446 0.1902 18 0.4387 0.5814 0.2434 19 0.3263 0.4457 0.2470 20 0.4273 0.2935 0.2073 21 0.2647 0.5543 0.2470 22 0.3583 0.8967 0.0926 23 0.2870 0.3315 0.2216 24 0.3363 0.9076 0.0839 25 0.2751 0.8859 0.1011 26 0.3948 0.8750 0.1094 27 0.2383 0.7772 0.1732 28 0.1814 0.7772 0.1732 29 0.3539 0.6630 0.2234 30 0.2959 0.9674 0.0315 31 0.2384 0.8098 0.1540 32 0.4440 0.6957 0.2117 33 0.3416 0.6685 0.2216 34 0.2076 0.8641 0.1174 162 Item No. Point Biserial P Vari ance 35 0.4688 0.6685 0.2216 36 0.4067 0.8478 0.1290 37 0.2968 0.3696 0.2330 38 0.3210 0.6739 0.2198 39 0.3440 0.8804 0.1053 40 0.3078 0.8207 0.1472 41 0.2928 0.9620 0.0366 42 0.4201 0.6359 0.2315 43 0.3609 0.9130 0.0794 44 0.3408 0.7174 0.2027 45 0.2439 0.7337 0.1954 46 0.3292 0.8207 0.1472 47 0.1193 0.6793 0.2178 48 0.4165 0.7065 0.2073 49 0.3849 0.4239 0.2442 50 0.2872 0.7174 0.2027 51 0.3509 0.6250 0.2344 52 0.1789 0.1957 0.1574 53 0.1811 0.1957 0.1574 54 0.4073 0.7228 0.2003 55 0.4855 0.7337 0.1954 56 0.4337 0.6087 0.2382 57 0.4777 0.5054 0.2500 58 0.3110 0.5707 0.2450 59 0.4404 0.5978 0.2404 60 0.1133 0.6087 0.2382 61 0.3035 0.7663 0.1791 62 0.3903 0.4293 0 .2450 63 0.4108 0.5380 0.2486 64 0.4139 0.6359 0.2315 65 0.1932 0.2772 0.2003 66 0.2616 0.3261 0.2198 67 0.1705 0.4674 0.2489 68 , 0.0883 0.5109 0.2499 69 -0.0151 0.3152 0.2159 70 • 0.2474 0.3859 0.2370 71 0.4731 0.4511 0.2476 72 0.3684 0.4891 0.2499 73 0.2954 0.2446 0.1848 74 0.1764 0.6033 0.2393 75 0.2612 0.6739 0.2198 163 Item No. Point Biserial P Vari ance 76 0.2213 0.6033 0.2393 77 0.2749 0.5326 0.2489 78 0.1264 0.3967 0.2393 79 0.1470 0.4837 0.2497 80 0.3606 0.5924 0.2415 81 0.1360 0.2065 0.1639 82 0.3584 0.8641 0.1174 83 0.2446 0.5272 0.2493 84 0.1079 0.4457 0.2470 85 0.3718 0.4891 0.2499 86 0.2428 0.3098 0.2138 87 0.3261 0.6793 0.2178 88 0.3878 0.6957 0.2117 89 0.3140 0.3967 0.2393 90 0.3894 0.3587 0.2300 91 0.3604 0.6033 0.2393 92 0.4194 0.5978 0.2404 93 0.0203 0.4946 0.2500 94 0.3727 0.4402 0.2464 95 0.2797 0.5163 0.2497 96 0.3969 0.6359 0.2315 The Mean is 58.0054 The KR-20 is 0.8887 The Standard Deviation is 12.6179 164 APPENDIX D RAW DATA—IDENTIFIED ACCORDING TO FORMAT 1 " 011 121 / 1 08421 316 6 3 09 G9 16 b Of 0 64 59 'JO i>5 4d 218"i 9 A l5 2 C2112153 10013 213 5 3 09 0 8 09 7 Oo 3 5 0 61 4 3 42 34 ld;> 2 L 17 OL 3 ^0311 1161 09(320 315 d 3 Od 09 12 7 10 6 63 YO r> 7 62 44 2<t3 is 30 lb 4 04111153 12020 315 64 08 11 18 7 Od 769 65 60 37 <+7 209 3d 39 Jl 5 C5111153 09923 318 3 2 Oi> 08 1 5 2 03 2 40 dl 62 62 49 234 l'+ 20 Ob 6 9.0^12162 11019 214 7 4 Ov 12 20 7 1 1 o 7<> 67 5 / 56 53 23 3_ 4H_0* 7 C7112~lod 0^/22 31 / 3 5 09 Od 12 0 Ob 7 58 39 51 27 48 165 22 34 12 8 C8H2167 08ol7 212 5 2 07 09 08 5 Oo 3 45 61 r>3 3<t 48 196 15 ^2 07 9 09111131 12317 212 7 5 09 10 16 2 07 5 61 54 58 64 t>0 236 3o 42 u£ 1C 101121O1 10323 318 5 t 09 09 18 5 07 5 62 54 56 51 50 211 26 33 05 11 11111155 12620 315 7 3 09 12 16 8 04 6 o5 58 51 33 46 186 J2 31 GO 12 , 12112163 12321 316 9 5 13 08 20 7 09 7 73 50 58 46 3 7 191 __3u 38 02 13 13112155 10514 109 4 4 il 07 11 u Oo 2 51 70 54 59 34 21/ la 31 05 14 14111152 11416 211 7 4 10 10 17 8 04 2 62 55 39 35 33 167 29 37 08 15 15112150 11419 214 7 5 09 10 18 5 07 3 64 65 41 33 HI 166 -31 37 C6 If. lolllldo 11419 214 f 3 09 09 15 5 08 5 61 62 66 54 63 243 15 24 09 17 1/112152 11222 317 6 5 09 11 16 8 06 5 68 68 53 49 197 26 32 06 13 18112160 10113 108 5 3 06 10 14 6 05 3 54 57 57 53 54 221 25 31 06 IS 19111155 13125 320 8 4 10 12 23 5 09 6 7/ 75 71 67 72 265 3/ 37 00 2C 20112170 10022 317 7 4 11 05 10 5 06 8 58 51 59 35 11 15t> 35 3o 01 21 21112156 09419 214 4 4 11 09 09 o 05 5 ^3 47 57 5/ 33 194 18 18 00 22 22112163 08615 110 0 2 08 06 07 3 01 2 29 39 34 34 28 135 10 11 01 23 23111157 09615 110 5 3 Od 09 11 5 04 3 43 61 63 4/ 48 219 2o 29 03 2^ 24112160 11125 320 7 5 09 10 13 5 Oo o 61 43 29 1 3 56 141 27 3j+_07 25 ~251121o7 10722 517 7 4 10 09 13 4 06 5 58 58 32 45 56 211 2o Jl 06 7t 26112150 1 1222 31 / 8 3 10 11 I 3 7 09 6 67 67 36 50 62 235 2<> 30 J*f 27 0112ilol 10413 108 6 3 08 05 13 6 05 4 50 41 24 34 35 134 17 20 03 28 C2121177 09424 319 9 3 i.0 09 15 7 09 5 67 76 71 /9 80 306 20 2<t 0^-2S C3121 150 10014 109 8 3 09 11 It 7 07 5 o4 54 3b 15 29 133 31 3o 05" _30 04122150 10 7^2 317 5 3 06 08 06 2 09 0 4 1 5 3 43 4 7 5 5 L9 8 2 0 2d 03 IDENTIFICATION OF DATA VIA "F" FORMAT - CONSECUTIVELY LABELLED: (1) F 2.0 = Class number (16) F 2.0 = Inferring. (2) F 1.0 = Year(first analysis of variance classification.) (3) F 1.0 = Division within the year. (17) F 1.0 = Predicting (4) F 1.0 = Sex (l=boy, 2=girl). (18) F 2.0 = Total Science Processes (5) F 3.0 = Age (19) F 2.0 = Fun Attitude (6) F 3.0 = I.Q: (20) F 2.0 = Pursue Attitude (7) F 1.0 = Actual CEFT Score. (21) F 2.0 = Structure Attitude (8) F 1.0 = Second analysis of variance classification accord i ng to CFJT soore . (9) F 2.0 = original CEFT score (Tot=20) - alternate scoring method. (10) F 1.0 = Observing (22) F 2.0 = Individual Investigation Att (11) F 1.0 = Comparing (23) F 3.0 = Total Attitude Score (12) F 2.0 = Classifying (24) F 2.0 = Sept. Reading (13) F 2.0 = Quantifying (25) F 2.0 = June Reading (14) F 2.0 = Measuring (26) F 2.0 = Reading Gain Score (15) F 1 .0 = Experimenting. I U vl 2 1 C 5 1 2 1 L35 1 0 0 1 9 214 2 3 1 1 0/ 1 3 3 0 ( 6 J2 4 b 3o 42 4o 169 2 4 34 I 3? 0ol21 L52 1 1 12 i 316 / 4 16 08 1 5 4 0 9 3 o2 5o o2 50 49 20 / 2 7 3 8 1 1 3 3 0 7 12 2 L5 I 10ol9 2 1 5 7 3 10 08 10 5 03 3 4 9 39 4 7 37 6 3 20o 2 / 38 1 I 34 C812 11 L 5o 10622 31 / 5 3 10 09 j. 8 6 10 6 o7 25 10 20 1 3 0/3 32 52 Ou 35 09121 L/5 09124 319 3 4 09 06 1 / 3 0 3 3 5 2 28 45 41 4 3 13 / 19 20 01 3c. 1G122 L5 1 1 192 0 315 o _3_ 11 1 1 1 7 i Oo 3 5 0 65 60 5 3 3 5 23 5 13 20 0 / 3 7 1112 1 L32 0992 1 'Hit' 3 3 Od" "u7~ 1 1 0 03 3 42 54 42 40 6 5 19 9 2 9 ~1 0 3 3 12121] L5 / 106 18 214 2 1 11 0 3 10 6 06 3 49 56 54 4 9 53 2t2 20 51 1 i 3S 13122] Lu5 094 1 3 108 4 2 0 9 0 8 06 3 04 4 40 35 44 45 3 \ 13b 30 31 6 i 4 0 14121 L6 1 103 14 109 7 4 08 11 13 2 06 6 o2 65 36 46 59 2/.o 10 2 0 lo 41 15 121 162 09621 316 b 4 06 10 21 / 0 7 5 6 7 52 4 0 3 2 40 1-4 l 9 2 9 lu 4? lol22 L 4-9[_ 11419 2i4 3 09 10 14 7 08 4 38 39 17 13 26 1O0 23 2d 0 3 4 3 1712 2] L53 09 9 I) 214 5~" 3 10 09 0 7 4 Ob 5 43 6 J 50 48 52 ~2To " 21 23 02 4 4 1«121 L50 11121 316 8 3 09 09 16 6 06 4 63 65 50 5 9 5 9 213 27 34 0 ( 4 * 1912 2] 169 096 1 7 212 5 3 09 08 10 3 04 3 45 47 35 3 7 3 8 15 / 19 27 o9 46 20121 15 5 10b 19 214 6 3 12 0 7 1 1 4 0 / 5 55 60 5 3 34 56 20 i 26 25 01 4 7 2 1121 16 7 09822 31 7 5 2 U6 08 13 4 05 5 50 40 2d 28 34 130 19 25 06 48 22122J L74 086 19 214 2 3 0 7 09 03 4 Ob 3 41 51 42 41 24 158 15 12 Ou 4 s 23 122 155 11023 316 6 3 11 I I 12 3 Od 4 58 53 t> 1 51 o4 234 3 5 35 00 5C 24121] 15 9 11422 317 6 4 09 09 16 4 06 1 59 50 61 16 35 162 3b 33 0 3 51 25122 L4 I 14820 313 9 4 lu> 12 19 7 1 i 6 81 48 56 24 4d 176 41 41 0 0 52 26121 L5o 10023 316 8 3 09 10 12 5 02 1 50 54 5 5 52 61 222 22 2/ oS «; -J 27121 L75 0 9619 214 5 4 10 07 11 5 05 4 51 5o 51 48 61 216 15 15 06 54 28 12220 1 09120 313 4 4 06 09 10 4 04 4 4 7 70 7 3 5 7 56 256 29 35 06 55 29121 L52 10020 315 o 4 09 1 1 15 4 07 4 60 53 5 7 o4 4 7 221 19 25 u* 56 30122 L4 7 11222 31 / 6 4 08 10 18 o 10 3 65 42 52 47 43 134 2 7 3d 09 57 31122 L5 9 10319 214 3 4 10 07 10 2 0 7 4 4 7 64 64 4 4 55 22 / 22 35 13 53 32121 L47 13622 31 / 9 4 11 12 19 4 07 5 71 06 48 40 34 188 40 37 oo 5 c C1212 158 09719 214 5 4 G8 10 12 2 05 5 51 57 56 50 46 211 15 24 6C 02211 Lu7 10725 320 4 3 10 06 15 4 06 3 53 33 32 4 3 ibt- 148 26 2 5 OP 6 1 C3 211 155 10320 315 8 4 7 09 1 7 t> Oo 4 61 56 63 58 50; 227 22 32 10 62 04211 L57 12321 31o 6 5 it 12 17 9 09 5 74 59 56 49 50i -214 52 36 04 6 3 05212J L49 12014 1 10 5 3 10 12 19 6 10 4 69 54 4 / 31 36 1 7 0 34 33 01 64 Co2l 1 L59 10616 213 6 5 10 11 23 4 05 3 6 7 62 50 5J 4 7 212 28 2 7 00 65 07211 L55 10520 315 4 1 0 7 08 12 2 07 2 43 54 44 36 39 17 3 50 36 06 6 6 C8211 154 13021 316 9 5 13 11 25 8 12 5 88 50 4 8 45 53 19 6 41 44 03 67 09211J L36 09516 21 1 5 4 09 08 13 2 06 6 55 5 8 65 37 42 202 20 30 10 68 10211 L50 11921 316 7 3 10 1 1 20 / 4 10 6 7i 63 61 51 54 229 3 9 36 00 6S 1121 U L5o 10ol3 106 4 4 08 09 16 4 06 6 57 43 26 33 14 l2o 31 20 0^ 70 1221 1 L50 11618 213 5 5 07 12 20 5 09 5 68 63 43 50 44 20b 26 23 0^ 71 13212J L55 11913 108 7 4 10 10 19 8 0 9 5 72 48 4 9 55 3d 190 30 3 3 o& 72 14212 L61 10014 109 3 5 09 09 11 6 07 4 54 50 45 35 42 172 2 1 29 o.8 73 15211] .o7 084 iO 105 2 3 05 06 08 2 04 1 31 47 52 56 46 201 13 13 Ou 74 1621 1 150 10o20 315 3 4 09 11 1 1 3 06 2 49 42 45 43 56 loo 24 22 75 16212 L 5 3 10918 215 4 4 12 1 I Id 4 09 5 65 5 9 53 35 44 19 1 26 29 o3 76 192111 L54 12517 213 9 4 11 12 20 4 09 5 74 54 40 34 51 1 79 5 / ii OP 77 20211 L55 092 8 104 3 4 Ob 08 14 4 05 3 44 42 36 34 26 138 27 22 OO 7 8 21212 L51 10013 106 5 4 08 07 1 1 3 OP 5 4 8 64 4 5 42 4 3 194 20 22 02 7S C1222] Lo3 09118 213 4 2 Oo 09 lb 6 0 7 4 53 62 00 66 64 252 23 2o o3 3C C2221] L 56 12221 3 16 8 5 10 09 16 5 Od 4 o5 44 37 12 1 9 112 3 / 35 OO 3 1 03221 L6 4 •lUOld 213 4 4 10 10 15 6 0 7 4 60 Gd 69 72 5 7 2 65 29 29 0 0 F.2 C4 2211 . 53 097 15 Ul 1 2 Oo 04 08 7 03 3 34 33 id 15 35 121 20 32 i 1. 8 3 05 221 . 54 11922 317 / 4 11 12 14 9 10 8 75 55 52 56 04 22/ 3/ 32 Ou 3 4 C622 1 LoO 10922 317 4 _4_ 0 9 11 14 _4_ Oo 1 3 3 6 7 o4_ 72 64 2o7 31 _3|L u -r 3 5" C 1221 L"51~ 13213 213 7 4 LV" 11 2 1 d 09 / 30 32 56 56 59 22 5 3 7 3 5 u C 36 C8221 L52 12 721 316 8 5 06 10 1 / 6 08 3 o7 69 64 60 60 25 3 3o 4l U'i 37 C9221 LoO 10313 10 8 4 3 09 09 14 6 Oo 3 54 o2 4 2 29 14 14 7 32 2 9 uO 33 1022 I L6 1 09320 J15 2 0 0 7 08 14 2 03 2 38 40 5 7 31 37 16b 53 21 0; J 8 9 1222 1 L63 0951 1 106 5 3 06 06 lb 4 03 1 45 56 32 2d 4 1 15 / 14 14 U U CC 1 3222 . DO 10216 2 1 1 6 3 0 9 09 L 5 0 03 5 5-5 VO 13 2 0 0 6 0 o 9 31 3 6 I DO 91 14222164 09910 lOo 4 2 0-j 05 07 5 04 2 34 34 5 3 3 4 31 17/ 15 a on Ou 92 15221172 09518 213 8 3 06 09 13 3 03 2 54 5 7 56 4 4 32 209 2b it S3 16222157 09717 212 4 3 09 0 7 09 3 04 3 42 33 56 4 5 i6 195 23 21 94 17222161 09320 315 5 3 08 07 09 5 0 7 6 50 45 50 31 13 144 28 28 00 S5 18221160 09616 211 4 4 08 10 15 4 03 7 60 49 62 3 9 50 200 2^ 27 0^ 96 19 221189 09219 214 4 4 10 08 17 5 0 7 3 58 bo 6 8 40 63 23 7 13 20 0* S7 20221162 09415 Ul 3 3 Ob 0 7 11 4 0 f 4 45 60 46 37 34 17 7 23 23 00 98 21222166 09*17 212 7 4 08 08 15 5 08 6 61 69 5 I 57 35 218 15 2 7 12 99 22222156 11819 214 8 4 09 08 15 8 10 5 67 51 60 50 66 227 35 37 02 ICC 01231176 09213 108 5 4 08 03 09 4 03 4 47 52 34 43 59 193 14 24 10 1C1 C2231157 11724 319 6 3 10 12 15 4 08 5 63 36 16 30 29 Ul 3b 39 03 IC2 03231153 13i22 31 7 a 4 10 12 19 6 07 5 71 43 4 3 2 9 36 151 42 45 03 1C3 C4231150 11819 214 7 2 C6 10 13 3 04 4 56 3o 30 34 37 15/ 31 38 0/ 104 05232182 08820 315 1 1 09 07 09 b 04 4 41 47 45 5 7 39 183 19 29 10 IC5 06232150 11621 316 8 3 11 11 14 5 09 7 68 58 57 65 40 220 34 37 03 106 07232177 09217 212 5 4 06 08 06 3 02 4 40 42 43 4/ 38 170 15 14 00 107 08231151 12521 316 7 4 13 12 19 3 08 4 75 64 49 30 37 180 J4 43 09 1C8 C9231149 12823 318 9 3 09 11 20 6 I 1 7 76 53 50 43 53 199 36 39 03 ICS 10232160 11122 317 6 4 09 11 17 6 10 4 67 o9 63 52 64 250 24 35 11 1 10 11231160 12320 315 9 5 12 12 25 6 13 6 88 67 62 60 70 259 41 43 02 111 13232155 12C20 315 6 3 11 12 15 5 09 6 67 63 65 53 56 242 30 36 06 1 12 14232149 14223 318 3 5 13 11 21 3 10 6 82 48 44 60 56 203 42 42 00 113 15232149 11824 319 8 4 08 11 lb 7 07 4 65 29 20 13 11 073 35 37 02 1 14 16231160 09917 213 4 3 05 08 17 6 08 5 56 72 66 68 57 263 It 24 10 115 17232164 09111 108 3 4 08 09 10 4 06 4 48 65 65 28 47 205 17 19 02 116 18232154 12123 318 6 3 12 10 20 6 09 5 71 41 40 38 50 169 37 39 02 117 19232152 12523 318 6 5 10 08 17 4 09 7 66 67 56 45 62 230 41 42 01 118 20231159 13123 318 8 4 13 10 21 3 09 2 7 5 63 52 41 49 20 5 41 41 00 US 21232153 12223 318 7 4 12 10 20 7 1 I 6 77 o7 59 32 47 205 25 36 11 120 22232156 12920 315 3 4 10 10 17 6 09 5 64 53 54 49 53 20 9 34 41 07 121 23231166 09718 215 3 2 06 10 10- 4 06 4 45 45 32 46 20 163 20 29 OS 122 24232153 12522 317 7 3 09 09 18 8 06 2 62 53 61 36 46 201 3b 38 02 123 25232153 11316 211 6 4 09 10 17 5 10 4 65 6 7 40 50 44 201 21 35 14 124 26231167 09622 317 6 5 11 10 17 5 07 3 64 55 46 51 51 20 3 42 2 7 oc 125 27232154 11217 212 7 4 08 10 18 7 11 7 72 b7 62 55 56 240 28 40 It 126 28231 167 09214 111 3 2 09 03 10 5 0 3 3 43 33 37 3 9 26 140 lo 31 1: 127 29231161 12822 317 5 4 10 12 22 8 04 6 71 74 77 o9 67 287 32 38 Ot 128 30232150 12123 318 7 4 10 11 23 6 10 6 77 47 53 39 41 180. 41 37 Ot 129 01312155 11419 214 3 4 10 07 16 6 05 2 53 55 33 17 46 151 25 30 Oc. 130 02311158 1022 5 320 4 3 10 10 15 6 07 3 53 58 49 35 50 192 27 29 Ot 131 03311151 09624 319 3 3 09 10 15 3 04 3 50 28 18 24 10 080 20 14 oc 132 04311172 10322 317 5 4 08 11 18 4 07 6 63 42 4 5 41 35 163 22 14 oc 133 C5311152 12917 212 8 3 10 12 23 7 11 5 81 71 72 58 75 2 76 31 33 0, I 34 06311149 11821 316 6 3 09 12 21 6 10 4 71 63 54 5b 53 231 2 7 33 Oc 135 C7311165 09521 316 9 2 09 11 21 7 08 5 72 64 58 55 54 231 19 25 Ot 136 08311169 09020 315 3 4 09 11 13 6 02 7 55 66 33 2b 37 16 7 18 19 01 137 0931U59 10820 315 5 3 li 06 20 5 06 6 62 46 33 4 5 43 172 30 34 0' 138 10311154 11914 109 4 3 08 06 09 1 05 4 40 33 3 7 20 29 124 3o 33 oc 139 11312170 09d22 317 7 3 10 11 21 8 09 8 77 65 53 34 39 191 17 24 0 14C 12312170 08420 315 3 2 06 06 08 7 05 2 39 47 o3 40 50 202 12 19 0 141 13312155 11625 320 4 4 13 08 12 4 08 4 5 7 53 60 3 7 29 179 23 30 0' 142 14312154 10217 212 5 4 11 11 14 8 08 6 67 50 57 31 44 182 23 33 K 143 15311159 12022 317 7 4 13 11 19 5 11 6 7b 72 7 1 64 4 7 254 38 41 0. 144 16312154 10918 214 6 3 10 10 16 7 06 7 65 57 48 2 7 27 159 26 41 l: 145 17311162 0y416 211 1 0 08 03 03 1 07 4 29 40 22 2 7 26 115 12 14 u 146 18312160 09012 107 0 2 03 02 08 4 02 3 24 44 38 43 24 154 20 21 0 147 19311169 09721 31o 6 5 10 11 17 5 06 3 63 63 67 44 50 224 19 24 0 148 20311174 07812 107 2 4 04 05 14 3 02 4 36 42 41 34 40 157 16 15 0 149 21312162 11216 212 6 4 08 10 18 7 1 1 6 70 o2 7b 75 51 264 36 39 0 150 22312152 0972 I 316 5 4 09 09 lo 4 05 4 5 6 61 5 7 46 55 2 19 20 23 0 I D / 1 5 1 2 3 3 12 15 > 12 9^3 did / 5 I * 1 1 2 1 / Oo 8 80 61 3 6 5 I :> 3 22 1 5o 40 0<f ! 243 I 2 136 10 2 31 J 5 4 10 1 1 1 7 7 0 / 3 5 4 •» 9 4 0 SO •> 5 1 od l / 22 0$ 1 5 3 25.) 1 1 lo / 0 7 o 12 10 J o 2 10 0 4 0 7 5 02 4 40 40 4 9 3 7 33 15 9 14 15 ot 1 5 4 26 3 12 15 2 1202 1 5 1.6 o 3 12 10 19 3 0 9 / 7 L 5 6 50 38 4 3 189 3 7 36 00 1 5 5 27312 164 0d<: 13 108 1 2 Oo 11 09 3 Oo 2 4 0 28 3 3 29 5 5 12 2 17 22 05 L '5 6 28312165 0 8 3 1 o 2 1 1 2 _L 0 7 0 5 I 1 _„.„ JL 03 2 3 8 4o 4 4 22 16 12a 12 16 04 157 29311159 130 lo 211 5 _> "l r 12 6 U9 6 7 1 67 "of "46 5 3 252 40' 40 00 153 CI 32 2 lo2 11424 319 6 3 12 10 1 7 2 0 5 4 5 9 6 9 71 o 3 61 2 o o 2 7 3 1 14 1 5 9 02321165 0 J 8 1 d 213 5 Oo 09 12 i 0 3 2 4 4 5 9 76 72 59 cob Id 20 02 If C C3322 15 7 11319 214 I 4 12 10 15 1 03 5 3 9 60 60 4 / 3d 20 3 3l 32 02 16 1 04322151 10719 214 3 4 08 03 11 2 0 5 4 4 5 51 60 55 4o 212 1 6 20 u2 16.? 05321155 10514 10 5 6 _06. 0 9 14 3 Od 3 51 55 3d 42 4 6 135 i 7 2 3 ...Q5 16 3 C6321136 0 9713 213 7 4 03 TT 10 3 05 4 52 64 5o 29 4 / 19o 20 1 / •JQ> 164 0/321loi 1 102 1 316 3 3 10 10 16 7 04 4 57 59 4 6 43 4 7 202 Jo 3 9 01 16?: C832 1137 09 716 211 3 4 07 08 1 1 5 04 5 4/ (0 6o 60 3d 2 54 12 14 6 2 166 ^)i^2l,7 10424 319 5 4 09 10 1 7 5 0 / 6 63 33 53 48 3 3 174 2 i 2J 03 167 10321 153 12424 319 o 4 10 08 lo 6 0 7 4 61 42 2 7 41 36 146 2d 3 9 1 i 16S 1 132217 1 095 16 2 1 1 6 3 1 1 1 1 13 4 05 5 53 39 34 60 30 133 __2o_ 24 00 169 12322 lo6 109 1 / 212 6 4 0 9 10 lo 4 0 7 2 5 7 6 / o3 5 3 4 / 230 33 2 3 0~0 170 13321156 10217 213 3 4 03 09 15 4 05 3 51 /4 71 47 50 242 18 23 03 17 1 143221613 09515 110 3 3 09 09 12 3 05 3 4 7 39 35 35 47 l5o 14 15 01 172 15321182 09318 213 5 4 10 09 1 7 5 08 6 64 64 6 3 63 44 2 54 24 26 02 173 1632 1164 09915 UO 3 3 09 03 17 5 05 5 5 5 44 44 3o 24 148 22 23 0i I 74 17321171 09113 106 3 5 09 07 12 3 04 3 4o 53 62 40 3 6 217 2 7 24 I 7 5 18321 163 10 3 15 no 5 1 04 02 0J 4 01 4 22 53 51 62 4 9 220 24 26 02 176 19322151 1 lol9 214 5 4 10 11 12 3 09 6 60 66 56 54 5 7 233 4 9 23 00 177 20322153 11219 214 6 4 08 12 20 7 06 6 71 34 30 28 1 / 109 31 31 00 1 7 3 21322158 10725 313 6 5 i3 11 14 7 05 6 6 7 62 59 67 59 24 / 2 9 39 10 17s 22322154 1 1220 515 4 4 10 09 1 7 5 05 3 5/ 64 62 53 50 234 2 9 33 04 IPC 255221o7 0 8 515 1 1 1 4 4 10 08 05 3 05 3 42 5o 34 33 5 7 ld3 14 20 06 13 1 24322153 "l 0521 316 5 2 11 09 06 4 09 4 52 33 44 54 43 1/9 19 22 03 132 25322 153 12316 211 5 1 09 03 15 0 7 5 32 34 06 5o 8o lv32 30 31 01 1 83 2632215 5 13824 319 8 5 09 12 2 1 9 1 1 7 32 37 43 52 5 5 £92 40 43 02 134 27321180 09215 110 5 5 10 09 13 7 06 3 56 48 51 41 46 I'd 3 lo 13 00. FNC 0 F FILE 

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
China 43 2
Canada 11 1
United States 9 3
Germany 5 2
United Kingdom 3 0
Colombia 2 0
Greece 1 0
Philippines 1 0
Myanmar [Burma] 1 0
Egypt 1 0
City Views Downloads
Hangzhou 18 0
Unknown 15 4
Shenzhen 5 2
Beijing 5 0
Nanjing 5 0
Ottawa 4 0
Ashburn 3 0
Pereira 2 0
Guangzhou 2 0
Vancouver 2 0
Mountain View 2 2
Chicago 2 0
Shanghai 2 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0055292/manifest

Comment

Related Items