Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effect of the interaction between test expectancy and question placement on incidental learning Koe, George Gerald 1974

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1974_A8 K63_9.pdf [ 7.07MB ]
Metadata
JSON: 831-1.0054407.json
JSON-LD: 831-1.0054407-ld.json
RDF/XML (Pretty): 831-1.0054407-rdf.xml
RDF/JSON: 831-1.0054407-rdf.json
Turtle: 831-1.0054407-turtle.txt
N-Triples: 831-1.0054407-rdf-ntriples.txt
Original Record: 831-1.0054407-source.json
Full Text
831-1.0054407-fulltext.txt
Citation
831-1.0054407.ris

Full Text

THE EFFECT OF THE INTERACTION BETWEEN TEST EXPECTANCY AND QUESTION PLACEMENT ON INCIDENTAL LEARNING by GEORGE GERALD KOE B.Ed., University of British Columbia, 1970 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in the Faculty of Education (School Psychology) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA MARCH, 1974 In presenting this thesis i n pa r t i a l fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make i t 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 i s understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Education The University of British Columbia Vancouver 8, Canada. Date February 27, 1974 ABSTRACT Koe, G.G. The effect of the interaction between test expectancy and question placement on incidental learning. Unpublished master's thesis, University of British Columbia, 1973. Problem The purpose of this study was to investigate the effect of and the interaction between test expectancy and question placement on incidental learning from prose materials. It was expected that post-adjunct questions would lead to higher scores on a test of incidental learning than pre-adjunct questions. It was also expected that a test expectancy condition would lead to higher scores on a test of incidental learning than questions without test expectancy. Methods of Investigation The study sample was limited to Grade Ten and Eleven students enrolled in the Mission Senior Secondary School (Mission, B.C.) during the 1972-73 session. The study used a two by two by two factorial design. The factors were: a) question placement (pre-adjunct or post-adjunct), b) test expectancy (test expectancy present or test expectancy absent), and c) race (Indian or non-Indian) . The major hypotheses of this study were based on a two by two by two analysis of variance of scores on a test of incidental questions. This analysis tested the factors described i n the design. General Conclusions The major conclusion of t h i s study was that neither question placement nor t e s t expectancy supported expectations which were based on the t h e o r e t i c a l contentions of Rothkopf (1966, 1970). Moreover, fin d i n g s with regard to question placement d i d not support expectations which were based on the findings of previous i n v e s t i g a t o r s (Frase, 1967; Rothkopf, 1966; Rothkopf & B i s b i c o s , 1967) . The only s t a t i s t i c a l l y s i g n i f i c a n t findings of t h i s study were that Indian subjects achieved higher scores than non-Indian subjects on the t e s t of the i n c i d e n t a l questions when t e s t expectancy was absent and that Indian subjects achieved lower scores on the t e s t of i n c i d e n t a l questions than non-Indian subjects when t e s t expectancy was present. However, the low r e l i a b i l i t i e s f o r the Indian subjects' scores, the wide variance i n the non-Indian group, and the small sample s i z e of the Indian group placed considerable constraint on the i n t e r p r e t a t i o n of t h i s f i n d i n g . i v TABLE OF CONTENTS Page ABSTRACT i i TABLE OF CONTENTS i v LIST OF TABLES V LIST OF FIGURES v i i ACKNOWLEDGMENT v i i i CHAPTER I. THE PROBLEM 1 Background of the Problem 1 The Problem 5 D e f i n i t i o n of Terms 6 Hypotheses and Research Questions 9 Assumptions of the Study 11 Delim i t a t i o n of the Study 12 J u s t i f i c a t i o n of the Study 12 CHAPTER I I . REVIEW OF THE LITERATURE 14 CHAPTER I I I . METHODOLOGY 24 Design 24 Measuring Instruments 27 Experimental Subjects 31 Experimental Procedures 33 S t a t i s t i c a l Procedures 34 Limitations of the Study 34 CHAPTER IV. PRESENTATION AND INTERPRETATION OF DATA 36 Relevant Questions 36 Incidental Tests 43 CHAPTER V. SUMMARY, FINDINGS, AND CONCLUSIONS 54 Statement of the Problem 54 Theoretical Framework and Model 54 Procedures 54 Findings 55 Conclusions 58 Implications of the Study 60 Areas f o r Further Study 61 V LIST OF TABLES Table Page 1. A Comparison of the S t a t i s t i c a l Results of Studies regarding the e f f e c t s of Pre-Adjunct and Post-Adjunct Questions on Incidental A s s o c i a t i v e Learning 15 2. A Comparison of the S t a t i s t i c a l Results of Studies regarding the Interaction between Pacing and Question Placement i n Studies of Incidental Learning 17 3. A Comparison of the S t a t i s t i c a l Results of the Studies i n v e s t i g a t i n g Time as a confounding 21 f a c t o r i n Incidental Learning 4. Stages of the Study 25 5. Factors i n the Analysis of Variance 35 6. Analysis of Variance of the Independent Variables: Question Placement and Race, on Relevant Question Set One 38 7. Analysis of Variance of the Independent Variables: Question Placement, Race and Test Expectancy on Relevant Question Set Two 42 8. Analysis of Variance of the Independent V a r i a b l e s : Question Placement and Race on Incidental Test One 45 9. Analysis of Variance of the Independent Variables: Question Placement, Race, and Test Expectancy on Incidental Test Two 50 A. C h a r a c t e r i s t i c s of the Pre-Study Population 103 B. Baseline Data for Tests 104 C. Table of S p e c i f i c a t i o n s for Experimental Questions 105 D. Comparison of S t i m u l i and Response Items i n Questions 106 v i Table Page E. Distance i n Words between Questions i n Relevant Question Set One and Incidental Test One 107 F. Distance i n Words between Questions i n Relevant Question Set Two and Incidental Test Two 108 G. Number of St i m u l i and Response Term Replications i n Questions 109 H. C h a r a c t e r i s t i c s of Parent and Sample Populations 110 I. Number of Subjects by Grade selected f o r each Condition of the Experimental Design, A t t r i t i o n , and Number of Subjects cast out of each Condition of the Design by Grade I l l J. R e l i a b i l i t y of Relevant Question Set One, Relevant Question Set Two, The Mathematical Computation Test, Incidental Test One, and Incidental Test Two 112 K. Mean Scores on Relevant Question Set One, Relevant Question Set Two, The Mathematical Computation Test, Incidental Test One, and Incidental Test Two 113 L. Comparison of Variance on Relevant Question Set One, Relevant Question Set Two, The Mathematical Computational Test, Incidental Test One, and Incidental Test Two 114 M. Analysis of Variance of the Independent Variables Question Placement and Test Expectancy on Incidental Test Two f o r Non-Indian Subjects 115 N. Analysis of Variance of the Independent Variables Question Placement and Test Expectancy on Incidental Test Two f o r Indian Subjects 116 0. Debriefing Questionnaire: Frequency of Answers 117 P. Debriefing Questionnaire: Verbal Answers 121 v i i LIST OF FIGURES Figure Page Interaction of Race by Test Expectancy for Incidental Test Two 51 v i i i ACKNOWLEDGMENT The w r i t e r wishes to express h i s appreciation to the members of h i s committee: Dr. S.F. Foster Dr. T.D.M. McKie Dr. D. Thomas The w r i t e r also wishes to express h i s appreciation f o r the help received from h i s wife: C a r r y l D. Koe, B.Sc. CHAPTER I THE PROBLEM Background of the Problem During the last decade, considerable research attention has been directed towards investigating the learning of information from highly factual prose materials (Frase, 1967, 1968a, 1968b, 1971; Frase, Patrick & Schumer, 1970; Morasky & Willcox, 1970; Natkin & Stahler, 1969; Rothkopf, 1966; Rothkopf & Bisbicos, 1967; Rothkopf & Bloom, 1970). The emphasis of this research has been upon the relationship between question placement and 1 associative learning. Studies have involved manipulating the location of questions before and after the particular portion of the prose selection to which they pertain; 2 observing relevant learning by retestmg the subjects on the same questions 3 used i n the experimental condition; and observing incidental learning by testing the subjects on the information content of the prose selection not previously questioned in the experimental treatment. Studies by Rothkopf (1966), Rothkopf & Bisbicos (1967) and Frase (1967) have pointed out that significantly more incidental learning occurred in studying prose when post-adjunct questions were used than when pre-adjunct The term 'associative learning" r e f e r s to assumed stimulus-response connections made by the subjects between words or groups of words i n the prose s e l e c t i o n to answer hi g h l y f a c t u a l questions. The term 'relevant' i s used only i n the sense that the subject has previously encountered the question while reading the prose. The term ' i n c i d e n t a l ' i s used only i n the sense that the subject has not previously encountered the question while reading the prose. 2 questions were used. In these cases, post-adjunct questions were questions that immediately followed the page of the prose to which they pertained, while pre-adjunct questions were questions that immediately preceded the page of the prose to which they pertained. This f i n d i n g i s of p a r t i c u l a r importance as i t suggests that a s s o c i a t i v e learning of f a c t u a l information i n prose may be shaped and modified through question placement. In attempting to develop a t h e o r e t i c a l construct f o r shaping and modi-fy i n g learning from prose, Rothkopf (1963, 1970) proposed that content learning i n reading depends upon the strength of associations between s t i m u l i . The strengths of associations between s t i m u l i seem r e l a t e d to the organization of the questions and prose and the inspection behaviours of the subject. The inspection behaviours of the subject r e f e r to the manner i n which the sub-j e c t ' s eyes move while reading. Inspection behaviour, according to Rothkopf (1963, 1970) i s thought to be a function of the organization of the material (Tinker, 1936) , the subject's learned reading behaviours (Hoffman, 1946), and the subject's motivation to read the material (Carmichael & Dearborn, 1947; Schroeder & Holland, 1968). I t can be hypothesized from Rothkopfs (1963, 1970) t h e o r e t i c a l model, that the reason f o r the apparent d i f f e r e n c e s i n i n c i d e n t a l learning i n the pre-adjunct and post-adjunct question conditions l i e s , at l e a s t p a r t i a l l y , i n the motivational aspects of question l o c a t i o n (Frase, 1967; Rothkopf, 1966; Rothkopf & B i s b i c o s , 1967). S p e c i f i c a l l y , i f the subject has to guess at the answer to a question before reading the passage he would be expected to read the passage mainly to determine the correctness of h i s guess. His hypothe-size d i n s p e c t i o n behaviours would equal or exceed the average number of regressions per l i n e , f i x a t i o n s per l i n e , and duration of f i x a t i o n pauses per l i n e which might be considered normal f o r h i s p a r t i c u l a r age group reading 3 that s e l e c t i o n e i t h e r with or without questions. Consequently, the subject would be expected to make the same number of, i f not more, associations between words r e l a t e d to answering relevant questions and the same number of, i f not fewer, associations between words r e l a t e d to answering i n c i d e n t a l questions than would be expected under the condition of reading without questions or reading with post-adjunct questions. In the post-adjunct question condition, the subject knows that questions w i l l be asked, but does not encounter those questions u n t i l a f t e r he has read the pertinent m a t e r i a l . As such, he would be expected to read the prose i n such a manner as to e i t h e r glean as much information from i t as p o s s i b l e , or to obtain such information as he thinks l i k e l y to be questioned. The subject's inspection behaviours would be hypothesized to include fewer regressions per l i n e , f i x a t i o n s per l i n e , and duration of f i x a t i o n pauses per l i n e than might be considered normal f o r h i s p a r t i c u l a r age group reading that s e l e c t i o n with pre-adjunct questions or without questions. Consequently, the subject would be expected to make the same number of, or fewer, associations between words r e l a t e d to answering relevant questions and the same number of, i f not more, associations between words r e l a t e d to answering i n c i d e n t a l questions than would be expected under the condition of reading without questions or with pre-adjunct questions. In t h i s context, i t i s apparent that subjects who do not know the exact nature of the questions to be asked, p r i o r to reading the prose, have a higher p r o b a b i l i t y of learning information not d i r e c t l y r e l a t e d to answering those questions than do subjects who know the exact nature of the questions to be asked before reading the prose s e l e c t i o n . I t i s also apparent that subjects who know the exact nature of the questions to be asked before reading the prose s e l e c t i o n have a higher p r o b a b i l i t y of learning the information 4 d i r e c t l y r e l a t e d to answering those questions than do subjects who do not know the exact nature of the questions to be asked before reading the prose s e l e c t i o n . However, Rothkopf's (1963, 1970) t h e o r e t i c a l p o s i t i o n does not adequately explain the f a i l u r e to f i n d s i g n i f i c a n t question placement e f f e c t s f or r e l e -vant questions. This f a i l u r e i s observed i n the c o n f l i c t i n g r e s u l t s obtained by Rothkopf (1966) and Frase (1967, 1968a) regarding relevant learning. Rothkopf (1966) claimed that pre-adjunct questions f a c i l i t a t e d relevant l e a r n -ing i n accord with h i s t h e o r e t i c a l p o s i t i o n . Frase (1967, 1968a) claimed that post-adjunct questions f a c i l i t a t e d relevant learning more than pre-adjunct questions which contradicted Rothkopf (1966). The majority of the evidence (Frase, 1968b; Frase, P a t r i c k & Schumer, 1970; Morasky & Willcox, 1970) points to the assumption that question placement does not n e c e s s a r i l y influence relevant learning. While Rothkopfs (1963, 1970) t h e o r e t i c a l p o s i t i o n does account f o r the data obtained by Frase (1967), Rothkopf (1966) , and Rothkopf & Bisbicos (1967), the r e s u l t s of these studies may have been confounded by the experimental d i r e c t i o n s . These experimenters e i t h e r d i r e c t l y informed t h e i r subjects that they were to be tested on the prose material being read or informed t h e i r subjects that they were involved i n a study concerned with how people learn from reading prose. The assumption that they were going to be tested i n order to quantify the r e s u l t s follows 'a p r i o r i ' . I t i s reasonable to assume that t h i s knowledge may have acted as an a d d i t i o n a l i n c e n t i v e to the subjects and may have been motivational by nature. Test expectancy may be postulated to act i n a motivational capacity s i m i l a r to that hypothesized f o r post-adjunct questions. Within Rothkopfs 5 t h e o r e t i c a l framework (1963, 1970), t h i s contention i s supported by a compar-ison of the findings of Hoffman (1946) and Carmichael & Dearborn (1947). Hoffman (1946) investigated eye movements i n Harvard subjects over prolonged periods of reading. Carmichael & Dearborn (1947) conducted a s i m i l a r study with the addition of t e s t s at one hour i n t e r v a l s . Carmichael & Dearborn (1947) reported s i g n i f i c a n t differences between t h e i r study and Hoffman's (1946) study i n terms of the number of b l i n k s , f i x a t i o n s , l i n e s read, and regressions. These r e s u l t s i n d i c a t e d that Hoffman's (1946) subjects had engaged i n scanning reading behaviours a f t e r one hour of reading. Carmichael & Dearborn's (1947) study suggests that the eye movements associated with the reading of a p a r t i c u l a r passage may be maintained by t e s t expectancy. The eye movements associated with t e s t expectancy are complemen-tary to those postulated f o r post-adjunct questions. However, the eye movements associated with t e s t expectancy i n t e r f e r e with the scanning behaviours postulated for pre-adjunct questions. Consequently i n the studies by Frase (1967), Rothkopf (1966) and Rothkopf Si Bisbicos (1967) , the incentive induced by t e s t expectancy may have disproportionately a l t e r e d the variance a t t r i b u t e d to question p o s i t i o n f or i n c i d e n t a l learning. The Problem The purpose of t h i s study was to inv e s t i g a t e the i n t e r a c t i o n between t e s t expectancy and question placement on i n c i d e n t a l learning. The study u t i l i z e d Grade Ten and Eleven, Indian and non-Indian students from the Mission Senior Secondary School, Mission, B.C. and s i m i l a r materials to those used by Rothkopf (1966). 6 D e f i n i t i o n of Terms Subjects One hundred twenty subjects were randomly selected from the four hundred seventy-eight students which comprised the Grade Ten and Eleven student population of the Mission Senior Secondary School, Mission, B.C. Prose S e l e c t i o n The s i x thousand three hundred f o r t y - f o u r word s e l e c t i o n from Rachel Carson's book, The Sea Around Us, e n t i t l e d "The Sunless Sea" was divided i n t o twenty pages of approximately t h i r t y l i n e s each and presented i n two sections. (See Appendix A). F i r s t s e c t i o n . The f i r s t s e c tion contained the f i r s t ten pages (the f i r s t three thousand one hundred f i f t y - s i x words of the prose s e l e c t i o n ) . Second s e c t i o n . The second section contained the l a s t ten pages (the f i n a l three thousand one hundred eighty-eight words of the prose s e l e c t i o n ) . Relevant Questions Those questions that subjects encountered while reading the prose 4 s e l e c t i o n were termed relevant questions. The fourteen questions used by Rothkopf (1966) as relevant questions were assigned to relevant question set one or relevant question set two by the i n v e s t i g a t o r . The i n v e s t i g a t o r con-structed s i x a d d i t i o n a l relevant questions so that a t o t a l of ten relevant questions pertained to section one of the prose s e l e c t i o n and ten pertained to section two of the prose s e l e c t i o n . (See Appendix B). 4 These questions appeared on a separate page e i t h e r immediately preceding or following the l o c a t i o n of t h e i r answer i n the prose. 7 Relevant question set one. Relevant Question Set One consisted of the ten relevant questions which pertained to the f i r s t section of the prose s e l e c t i o n . Relevant question set two. Relevant Question Set Two consisted of the ten relevant questions which pertained to the second section of the prose s e l e c t i o n . Relevant Learning^ Relevant learning has been defined i n other studies as the subject's raw score on a t e s t c o n s i s t i n g of the relevant questions administered a f t e r the prose s e l e c t i o n had been read. Pre-adjunct Question Condition The pre-adjunct question condition was defined as a question condition i n which relevant questions were placed immediately before the page i n the prose s e l e c t i o n containing the information necessary to answering them. The subjects were required to answer each question before reading the page of the prose containing the answer to the question. Pre-question Condition The term 'pre-question condition' has been used i n other studies to r e f e r to a question condition i n which a l l relevant questions were presented as a group p r i o r to reading the prose s e l e c t i o n . This t e s t was not administered i n t h i s study as i t was not a p p l i c a b l e to the hypotheses being tested. However, other i n v e s t i g a t o r s of i n c i H e n t a l learning (Frase, 1967, 1968a, 1968b; Frase, P a t r i c k & Schumer, 1970; Morasky & Willcox, 1970; Rothkopf, 1966) d i d administer such a t e s t immediately a f t e r the prose s e l e c t i o n had been read and the relevant questions answered. This condition was not used i n t h i s study as i t was not applicable to the hypotheses being tested. However, i t has been r e f e r r e d to by other i n v e s t i g a t o r s of i n c i d e n t a l learning. (Rothkopf, 1966) . 8 Post-adjunct Question Condition The post-adjunct question condition was defined as a question condition i n which relevant questions were placed immediately a f t e r the page i n the prose s e l e c t i o n containing the information necessary to answering them. 7 Post-question Condition The post-question condition has been re f e r r e d to i n other studies as a question condition i n which a l l relevant questions were presented as a group a f t e r the reading of the prose s e l e c t i o n had been completed. Incidental Questions Those questions that the subject encountered only on a p o s t - t e s t were termed i n c i d e n t a l . The twenty-five questions used by Rothkopf C1966) as i n c i d e n t a l questions were assigned by the i n v e s t i g a t o r according to whether they pertained to section one or section two of the prose s e l e c t i o n . The in v e s t i g a t o r constructed twenty-five a d d i t i o n a l i n c i d e n t a l questions so that a t o t a l of twenty-five i n c i d e n t a l questions pertained to section one of the prose s e l e c t i o n and twenty-five i n c i d e n t a l questions pertained to section two of the prose s e l e c t i o n . Incidental t e s t one. Incidental Test One consisted of the twenty-five i n c i d e n t a l questions which pertained to the f i r s t section of the prose s e l e c t i o n . (See Appendix C). Incidental t e s t two. Incidental Test Two consisted of the twenty-five i n c i d e n t a l questions which pertained to the second section of the prose s e l e c t i o n . (See Appendix D). This condition was not used i n t h i s study as i t was not applicable to the hypotheses being tested. However, i t has been r e f e r r e d to by other i n v e s t i g a t o r s of i n c i d e n t a l l earning. (Rothkopf, 1966; Rothkopf & B i s b i c o s , 1967). 9 Incidental Learning Incidental learning was defined as the subject's raw score on Incidental Test Two. Mathematical Computation Test The Mathematical Computation Test consisted of a set of twenty-five elementary algebra questions drawn from Modern Elementary Algebra, by Nichols, C o l l i n s & MacPherson (1963). (See Appendix E ) . Test Expectancy Test expectancy r e f e r r e d to whether the subject was able to i n f e r from the experimental d i r e c t i o n s that he would be tested on the section of the prose s e l e c t i o n being read. No Test Expectancy No t e s t expectancy r e f e r r e d to whether the subject was able to i n f e r from the experimental d i r e c t i o n s that he would not be tested on the section o f the prose being read. Hypotheses and Research Questions Several i n v e s t i g a t o r s claim that i n c i d e n t a l learning i s greater when post-adjunct questions are used than when pre-adjunct questions are used. (Frase, 1967; Rothkopf, 1966; Rothkopf & B i s b i c o s , 1967). I t was therefore hypothesized that: 1. Subjects i n a post-adjunct question condition would score higher on a t e s t of i n c i d e n t a l learning than subjects i n a pre-adjunct question condition. Carmichael & Dearborn (1947) p o i n t out that the number of eye move-ments may be maintained by t e s t expectancy. As such, t e s t expectancy should be associated with eye movements s i m i l a r to those postulated f o r post-adjunct questions. A lack of t e s t expectancy should be associated with eye movements s i m i l a r to those postulated for pre-adjunct questions. I t was therefore hypothesized that: 2. Subjects i n a t e s t expectancy condition would score higher on a t e s t of i n c i d e n t a l learning than subjects i n a no t e s t expectancy condition. Rothkopf's (1963, 1 9 7 0 ) t h e o r e t i c a l p o s i t i o n and Carmichael & Dearborn's (1947) study suggest that post-adjunct questions would lead to higher scores on a t e s t of i n c i d e n t a l l earning, as would questions under a t e s t expectancy condition, than would pre-adjunct questions or questions without t e s t expectancy. I t was therefore hypothesized that: 3. A s i g n i f i c a n t o r d i n a l i n t e r a c t i o n would be found between question placement and t e s t expectancy on * a t e s t of i n c i d e n t a l l earning. There was a lack of information i n the l i t e r a t u r e on i n c i d e n t a l learning regarding Indian subjects as opposed to non-Indian subjects. The following research questions were asked: 4. Would a race d i f f e r e n c e be found? 5. Would a s i g n i f i c a n t i n t e r a c t i o n between question placement and race be found? 6. Would a s i g n i f i c a n t i n t e r a c t i o n between t e s t expectancy and race be found? 7. Would a s i g n i f i c a n t three way i n t e r a c t i o n between t e s t expectancy, question placement and race be found? The s i g n i f i c a n c e l e v e l f o r t e s t i n g these hypotheses and research, questions was set at the .05 l e v e l . Where hypotheses or research questions are d i r e c t i o n a l the c r i t i c a l value of F used was F.10 rather than F.05. 11 Assumptions of the Study There were several assumptions inherent i n the hypotheses of t h i s study. 1. One of the basic assumptions o f t h i s study was that Rothkopf's (1963, 1970) t h e o r e t i c a l p o s i t i o n d i d not adequately explain e f f e c t s r e l a t e d to relevant l e a r n i n g . This f a i l u r e was r e f l e c t e d i n the c o n f l i c t i n g r e s u l t s obtained by Rothkopf (1966) and Frase (1967, 1968a) and the lack of s i g n i f i -cant fin d i n g s reported by Frase (1968b), Frase, P a t r i c k & Schumer (1970) , and Morasky & Willcox (1970). The t h e o r e t i c a l p o s i t i o n upon which t h i s study was based does not adequately explain the experimental r e s u l t s associated with relevant learning at t h i s time. Consequently, relevant learning was omitted as a dependent v a r i a b l e i n t h i s study. 2. Another basic assumption of t h i s study was that the t e s t s used would provide a measure o f the extent of the a s s o c i a t i v e process o f learning f a c t s while reading. 3. Another assumption of t h i s study was that question placement and t e s t expectancy would a l t e r eye movements which, i n turn, are assumed to provide at l e a s t a p a r t i a l p h y s i c a l manifestation of the a s s o c i a t i v e process of learning facts while reading. This contention was supported by the t h e o r e t i c a l construct of Rothkopf (1963, 1970) with regard to a s s o c i a t i v e learning from prose and by the comparison of Hoffman's (1946) findings with Carmichael & Dearborn's (1947) fin d i n g s . 4. S t i l l another assumption inherent i n the hypotheses of t h i s study was that question placement e f f e c t s could be found i n Canadian subjects. Previous studies of i n c i d e n t a l learning d i d not i n v e s t i g a t e the motivational aspects of c u l t u r e . Consequently, the authors of these studies provided neither experimental nor t h e o r e t i c a l evidence upon which to base an 1 2 expectation of possible cultural bias. 5. A f i n a l assumption of this study was that question placement and test expectancy effects would occur in Grade Ten and Eleven subjects. Previous studies on incidental learning were conducted on subjects ranging from Grade Ten to adult telephone workers. The studies mentioned on visual fatigue were conducted on university undergraduate subjects. Neither the visual fatigue studies nor the incidental learning studies suggest that age may be a factor in the results. Delimitation of the Study The subjects of this study were Grade Ten and Eleven students at the Mission Senior Secondary School, Mission, B.C. The materials of this study were the specific incidental question content of the chapter, "The Sunless Sea", in Rachael Carson's book, The Sea Around Us. Thus, the findings of this study may be considered to be applicable to Grade Ten and Eleven students at the Mission Senior Secondary School, Mission, B.C. under conditions similar to those used in this study. Justification of the Study This study was limited in terms of i t s generalizability. However, i t did attempt to provide a methodology for determining whether question place-ment alone was sufficient to modify incidental learning, and whether test expectancy interacted with question placement. This investigation of test expectancy and question placement was of importance in that i t attempted to c l a r i f y the theoretical position upon which a number of studies have based an examination of incidental learning. It also attempted to provide a replication of the findings of Rothkopf (1966). 13 Moreover, t h i s was the f i r s t study of question placement e f f e c t s on inc i d e n -t a l learning from prose which has discussed content v a l i d i t y and r e l i a b i l i t y of the dependent measure. This study was of p r a c t i c a l importance when the a p p l i c a t i o n of question placement i s considered as a means f o r modifying learning from prose. In both the i n s t r u c t i o n a l and the remedial s e t t i n g s , i t i s frequently the case that the information a subject i s required to learn f a r exceeds the constraints of time required to d i r e c t l y question t h i s information. As such, i t i s e s s e n t i a l to investigate whether questions and tes t s may be used to f a c i l i t a t e i n c i d e n t a l learning. CHAPTER II 14 REVIEW OF THE LITERATURE This review i s limited to those studies which deal with question placement effects on incidental associative learning. Studies were selected primarily on the basis that they used pre-adjunct and post-adjunct question conditions and that they differentiated between relevant and incidental associative learning. The discussion in this chapter i s limited to those experimental manipulations which apply directly to incidental associative learning and irrelevant variables which may have confounded those experimen-t a l manipulations. A description and additional criticism of each study discussed in this review is presented in Appendix I. Theoretically, more incidental associative learning should occur when post-adjunct questions are used than when pre-adjunct questions are used. Studies by Frase (1967, 1968a, 1968b), Frase, Patrick, & Schumer (1970), Rothkopf (1966), and Rothkopf & Bisbicos (1967) point out that greater incidental learning occurred from prose when post-adjunct questions were used than when pre-adjunct questions were used. The s t a t i s t i c a l results of these studies with respect to pre-adjunct and post-adjunct question effects on incidental associative learning are presented in Table I. With the exception of one contrast made by Rothkopf & Bisbicos (1967) the mean for the post-adjunct group was consistently reported as significantly higher than the mean for the pre-adjunct group. Unfortunately, the results reported by Rothkopf (1966) and Rothkopf & Bisbicos (1967) reflect the use of multiple tests. Hopkins & Chadbourn (1967) point out that the use of multiple tests exploits chance differences and introduces inextricable patterns of dependency and non-orthogonality in the partitioning. This leads to an inaccurate probability model upon which to base Ho. 15 TABLE I . A COMPARISON OF THE STATISTICAL RESULTS OF STUDIES REGARDING THE EFFECTS OF PRE-ADJUNCT AND POST-ADJUNCT QUESTIONS ON INCIDENTAL ASSOCIATIVE LEARNING Study S t a t i s t i c a l Test Degrees of Freedom Prob. Post-adjunct Mean 8 Pre-adjunct Mean Frase (1967) F = 16.10 1/60 .001 14.28 10.67 Frase (1968a) F = 10.93 1/* .005 * * Frase (1968b) F = 10.43 1/96 .001 13.02 : 1 1 . 4 9 Frase, P a t r i c k , & Schumer (1970) F = 4.61 2/261 .01 29.76 27.74 Rothkopf (1966) t = 4.54 1/159 + .01 41.82 32.38 Rothkopf & t = 2.03 l/252 + .05 * * Bisbicos (1967) t = .65 l/252 + .50 * * t = 3.03 l/252 + .01 * * + 8 Not reported by the experimenter. Calculated from the data reported by the experimenter. Mean scores represent mean number of questions answered c o r r e c t l y . 16 According to Ladas (1973), Hopkins & Chadbourn (1967) re-analyzed Rothkopfs (1966) data using the Newman-Keuls technique of multiple compari-sons. This method seemed to be a more sensitive approach given Rothkopfs (1966) design. Hopkins & Chadbourn (1967) found no significant differences even allowing one-tailed tests. The use of one-tailed tests by Rothkopf (1966) was also questionable. A one-tailed test infers a directional hypothesis. Rothkopfs (1966) study did not state any specific hypotheses although he did refer to Rothkopf (1963, 1965) in which the f a c i l i t a t i n g effects of test-like events were hypothesized. However, i t i s not clear how these hypotheses apply to Rothkopfs (1966) study. Rothkopf (1966) should have used two-tailed tests. The criticisms regarding s t a t i s t i c a l procedures used by Rothkopf (1966) also apply to the study by Rothkopf & Bisbicos (1967). Both studies used non random assignment of subjects to treatments which challenges the assumption of independence. Both studies used paid volunteers which severely limits the generalizability of results. No information on the r e l i a b i l i t y or validity of the dependent measures was reported in the studies by Frase (1967;,1968a, 1968b), Frase, Patrick, & Schumer (1970), Rothkopf (1966), or Rothkopf & Bisbicos (1967). While the questions were factual in each study, the lack of r e l i a b i l i t y information suggests that significant findings may have been par t i a l l y a function of measurement error. However, the trend of the investigations of the effect of question placement on incidental associative learning suggests that greater inciden-t a l learning does occur when highly factual questions accompanying prose are presented in the post-adjunct format as compared to the pre-adjunct format for subjects ranging in educational levels of Grade Ten to university. 17 TABLE tl A COMPARISON OF THE STATISTICAL RESULTS OF STUDIES REGARDING THE INTERACTION BETWEEN PACING AND QUESTION PLACEMENT IN STUDIES OF INCIDENTAL LEARNING Study Test Statistic Degrees of Freedom Prob. Pacing Mode Frase (1968a) Frase (1968b) F = 4.00 1/* .05 10 vs 20 sentences F = 3.24 3/96 .05 10, 20, 40, 50 sentences Frase, Patrick & Schumer (1970) F = 4.2 2/168 .025 1 vs 5 pages Not reported by experimenter 18 Frase (1968a, 1968b) and Frase, Patrick, & Schumer (1970) not only-indicated that post-adjunct questions resulted in greater incidental learning than pre-adjunct questions. They also reported that the farther away pre-adjunct questions were from the portion of the text to which they applied, the greater the incidental learning: the closer post-adjunct questions were to the portion of the text to which they applied, the greater the incidental learning. The comparison of the s t a t i s t i c a l results regarding the interaction between question pacing and question placement are presented in Table II. Pacing refers to the number of lines separating the question from i t s answer in the prose. However, in the case of these studies, pacing is confounded by differential placement of questions in the prose. This confusion between distance from the answer and the number of questions occurring together con-founds an interpretation of the effect of pacing. Moreover, the s t a t i s t i c a l result presented by Frase, Patrick, & Schumer (1970) portrays a four-way interaction among question placement, pacing, incentive, and item type. Natkin & Stahler (1969) suggested that the effects for post-adjunct questions for short term rec a l l noted in the studies by Frase (1967, 1968a, 1968b), Frase, Patrick, & Schumer (1970), Rothkopf (1966), and Rothkopf & Bisbicos (1967) may be reversed in long term recall (F = 22.09, D.F. = 1, p = .01). Natkin & Stahler (1969) claim that under conditions of high pre-exposure to questions, there was a decline from immediate to delayed testing. When questions were introduced with no pre-exposure, they resulted in a marked increase in delayed performance. However, this study incorporated a possible confounding factor in equating incidental questions with relevant questions. While the post-adjunct questions used in the study were relevant, the questions on the f i r s t 19 administration of the test were incidental. However, the incidental questions on the f i r s t administration of the test became relevant by definition on the delayed criterion test. This test measured relevant rather than incidental learning. This confusion resulted in the comparison between short and long term retention being a comparison between relevant and incidental questions. Natkin & Stahler (1969) did attempt to provide some information on the validity of their dependent measure. They reported a personal communication from Rothkopf indicating that the procedure for differentiating between incidental and relevant questions was about 95% effective for ensuring d i f -ferent content. However, Natkin & Stahler (1969) did not provide any r e l i a b i l i t y information on the dependent measure. While Natkin & Stahler's (1969) study was not directly comparable to other studies in this review, i t was included as i t suggested that short term memory was a possible restriction on the interpretation of the findings of Frase (1967, 1968a, 1968b), Frase, Patrick, & Schumer (1970), Rothkopf (1966), and Rothkopf & Bisbicos (1967). While the trend of the effects of question placement on incidental learning seem relatively unambiguous, several studies have indicated possible confounding effects associated with the time taken to read the prose selec-tion. Morasky & Willcox (1970) in two studies reported that when time was a dependent measure the post-adjunct question group took longer to read the prose than did the pre-adjunct question group. This finding was supported by Frase, Patrick, & Schumer (1970) and Rothkopf (1966). However, Rothkopf & Bisbicos (1967) did not support this contention and reported no significant differences in time. The s t a t i s t i c a l results of the studies investigating time as a dependent variable are presented in Table III. 20 I t was i n t e r e s t i n g to note that Morasky & Willcox (1970) calculated 9 the rate of comprehension and found i t to be greater f o r the pre-adjunct question group on relevant t e s t scores and t o t a l t e s t scores. On the basis of t h i s f i n d i n g , they concluded that i f the rate of comprehension were con-sidered important to reading then pre-adjunct questions were most advantageous to the reader. Morasky & Willcox's (1970) use of multiple t comparisons e x p l o i t s chance d i f f e r e n c e s and questions the orthogonality of p a r t i t i o n i n g . Rothkopf (1966) d i d not apply the means to a s t a t i s t i c a l t e s t . However, the trend of these i n v e s t i g a t i o n s suggests that prose with post-adjunct questions takes longer to read than prose with pre-adjunct questions. This trend i s consis-tent with t h e o r e t i c a l p r e d i c t i o n . Morasky & Willcox (1970) also d i d an analysis of the scores f o r each paragraph. Differences between the pre-adjunct and the post-adjunct question groups were not evident u n t i l the 12th, 15th, 17th, 18th, 19th, 20th, and 21st paragraphs. These r e s u l t s were interpreted to i n d i c a t e that a shaping procedure was taking place during the reading rather than a constant d i f f e r -ence e x i s t i n g from the f i r s t to the l a s t paragraph. Another f a c t o r which may have confounded the i n t e r p r e t a t i o n of studies of i n c i d e n t a l learning from prose was the nature of the experimental d i r e c -t i o n s . Rothkopf (1966) pointed out that d i r e c t i o n s to read c a r e f u l l y and slowly re s u l t e d i n greater i n c i d e n t a l learning than when no d i r e c t i o n s were given (t = 2.09, d.f. = 1/139, p = .05). However, the use of multiple t comparisons makes t h i s f i n d i n g questionable. Rate of comprehension was equal to the percent of the t e s t items corr e c t m u l t i p l i e d by the completion time. 21 TABLE III A COMPARISON OF THE STATISTICAL RESULTS OF THE STUDIES INVESTIGATING TIME AS A CONFOUNDING FACTOR IN INCIDENTAL LEARNING Study St a t i s t i c a l Test d.f. Prob. Post-adjunct Mean Pre-adjunct Mean Mean Unit Frase, Patrick & Schumer (1970) F = 195.28 2/261 .01 33.88 31.92 min. text Morasky & Willcox (1970) t = 4.92 1/41+ .001 954.6 754.6 sec. text @ T = 196.5 l/32 + .01 932.4 898.8 sec. text Rothkopf (1966) * * 79.9 69.0 sec. page Rothkopf & Bisbicos (1967) F = .80 6/245 * * * Not reported by the experimenter. Wilcoxon Rank Sum Test Calculated from the data reported by the experimenter 22 Rothkopf & Bloom (.1970) indicated that teacher presentation of questions was superior to a written presentation of questions and no questions (F = 6.87, d.f. = 2/59, p = . 0 1 1 . However, the results of this study may have been confounded by a differential time sequence for groups. The oral question group took, longer than the written question group which in turn took, longer than the no-question group. Rothkopf & Bloom (19701 also indicated that differences i n experimental directions for the three groups may have con-founded the results. Subjects may have communicated during the two minute break after the 66th slide and after the experiment with other subjects. Another possible confounding factor i n studies of incidental learning is the motivational level of the subjects. Frase, Patrick & Schumer (1970) reported and interaction between incentive and time (F = 43.37, d.f. = 2/261, p = .001). The ten cent condition took longer than the three cent condition which in turn took longer than the zero cent condition. Frase, Patrick & Schumer (1970) also noted a four-way interaction among question position, frequency, incentive, and item type (F = 4.2, d.f. = 2/168, p = .025). This interaction revealed that the expected ordering of incentive factors only occurred when questions were infrequent. Frase (1971) also investigated the effect of incentive variables and type of adjunct questions upon text learning. Frase (1971) did not find a significant main effect for incentive noted by Frase, Patrick & Schumer (1970). Frase (1971) attempted to explain this result i n terms of the high incentive of paid volunteers used in his study as compared to the captive subjects used by Frase, Patrick & Schumer (1970). Associated with the motivational level of the subjects is the question of provision of answers to subjects. Rothkopf (1966) claimed that providing subjects with knowledge of results for relevant questions did not e f f e c t i n c i d e n t a l l e a r n i n g . Th is conc lus ion was supported by Frase (1971). 24 CHAPTER I I I METHODOLOGY Design In order to t e s t the e f f e c t s of question p o s i t i o n , t e s t expectancy, and c u l t u r a l group on i n c i d e n t a l learning, a two by two by two f a c t o r i a l design was used. The factors were: a) relevant question condition (pre-adjunct or post-adjunct), b) t e s t expectancy (knowledge of i n c i d e n t a l t e s t i n g or absence of knowledge of i n c i d e n t a l t e s t i n g ) , anddc) c u l t u r a l group (Indian or non-Indian) . This design r e s u l t e d i n the four conditions depicted i n Table IV, within each r a c i a l group. The study was conducted i n the four stages also depicted i n Table IV. The materials f o r each stage of the study were presented to the subjects i n a sealed envelope. Stage 1. Stage 1 of the study f u l f i l l e d two functions. I t was necessary for introducing the t e s t expectancy condition i n Stage 2 and i t allowed for the shaping procedure taking place during reading described by Morasky S Willcox (1970). In Stage 1 a random h a l f of the subjects received pre-adjunct relevant questions embedded i n prose s e l e c t i o n one. The other h a l f of the subjects received post-adjunct relevant questions embedded i n prose s e l e c t i o n one. Ten relevant questions were used i n t h i s stage of the experiment as Morasky & Willcox had indic a t e d that s i g n i f i c a n t d i f f e r e n c e s between pre-adjunct and post-adjunct questions d i d not occur u n t i l t h e i r 12th question. Stage 2. Stage 2 of the study provided the means f o r introducing t e s t expect-ancy to the subjects. A random h a l f of the subjects i n the pre-adjunct and TABLE IV STAGES OF THE STUDY Condition Stage One Stage Two Stage Three Stage Four 11 21 12 22 Prose Selection One with Pre- adjunct Relevant Questions Prose Selection One with Post- adjunct Relevant Questions Prose Selection One with Pre- adjunct Relevant Questions Prose Selection One with Post- adjunct Relevant Questions Mathematical Computation Test Mathematical Computation Test Incidental Test One Incidental Test One Prose S e l e c t i o n Incidental Two with Pre- Test Two adjunct Relevant No Test Questions Expectancy Prose S e l e c t i o n Incidental Two with Post- Test Two Adjunct Relevant No Test Questions Expectancy Prose Se l e c t i o n Incidental Two with Pre- Test Two adjunct Relevant Test Questions Expectancy Prose Se l e c t i o n Incidental Two with Post- Test Two adjunct Relevant Test Questions Expectancy 2 6 post-adjunct conditions received the Mathematical Computation Test while the remaining subjects received Incidental Test One. As subjects had been i n -formed that the second t e s t they received would be s i m i l a r to the f i r s t t e s t , subjects who received the Mathematical Computation Test were expected to be l i e v e that they would receive another Mathematical Computation Test a f t e r reading Prose Se l e c t i o n Two. These subjects would be expected to have no t e s t expectancy that might modulate t h e i r reading behaviours. Those subjects that received I n c i d e n t a l Test One were expected to b e l i e v e that they would receive another Incidental Test a f t e r reading Prose Se l e c t i o n Two. These subjects would be expected to have s u f f i c i e n t t e s t expectancy to modify t h e i r reading behaviours. The Mathematical Computation Test was chosen so that there would be as l i t t l e confusion as p o s s i b l e on the part of the subjects that the information required by the t e s t was not i n the prose s e l e c t i o n . Confusion on the p a r t of the subjects could lead to changes i n reading behaviours i n an attempt to f i n d t h i s information i n Prose Se l e c t i o n Two. Stage 3. Stage 3 of the study was i d e n t i c a l to Stage 1 except that the second section of the prose s e l e c t i o n was used and the ten questions relevant to section two of the prose were used. Stage 4 . Stage 4 consisted of the administration of the dependent measure, Incidental Test Two, to a l l subjects. 27 Measuring Instruments One of the problems with previous i n v e s t i g a t i o n s of the e f f e c t s of question placement on i n c i d e n t a l learning was that i n v e s t i g a t o r s (Frase, 1967, 1968a, 1968b, 1971; Frase, P a t r i c k , & Schumer, 1970; Morasky & Willcox, 1970; Natkin & Stahler, 1969; Rothkopf, 1966; Rothkopf & B i s b i c o s , 1967; Rothkopf & Bloom, 1970) had f a i l e d to provide data comparing the d i f f i c u l t y of t h e i r t e s t s and relevant questions, and data on the r e l i a b i l i t y and v a l i d i t y of t h e i r t e s t s - In an e f f o r t to avoid t h i s shortcoming of e a r l i e r studies a pre-study was conducted at the G a r i b a l d i Junior-Senior Secondary School under conditions s i m i l a r to the main study. Pre-study. On the basis of a f l i p of a coin, f o r t y randomly drawn subjects from Grade Ten and Eleven were di v i d e d i n t o two groups. The Baseline Data group contained twenty-one subjects and the R e l i a b i l i t y group contained nineteen subjects. A t t r i t i o n due to absenteeism r e s u l t e d i n the loss of four subjects from the Baseline Data Group and three subjects from the R e l i a b i l i t y Group. C h a r a c t e r i s t i c s of the samples and a t t r i t i o n f o r the Baseline Data Group and the R e l i a b i l i t y Group are presented i n Table A of Appendix H. Baseline Data Group. The purpose of the baseline data group was to provide an i n d i c a t i o n of how guessing would influence scores on the experi-mental instruments. This group also helped to determine i f the t e s t s were s u f f i c i e n t l y d i f f i c u l t to permit the scales to r e f l e c t d i f f e r e n c e s a t t r i b u t -able to the experimental manipulations. Each subject i n the Baseline Data Group was provided with a sealed envelope containing the Experimental D i r e c t i o n s , Relevant Question Set One, Relevant Question Set Two, the Mathematical Computation Test, Incidental Test 28 One, and Inci d e n t a l Test Two. The order of the t e s t s i n the envelope was randomized using a table of random numbers. The subjects were requested to work through the tes t s i n the order stapled. The mean scores and standard deviations are presented i n Table B of Appendix H. The mean scores and standard deviations f o r Relevant Question Set One, Relevant Question Set Two, Incidental Test One, and Incidental Test Two were s u f f i c i e n t l y low to permit the scale to r e f l e c t d i f f e r e n c e s that might occur as a r e s u l t of the experimental manipulations. R e l i a b i l i t y Group. The purpose of the r e l i a b i l i t y group was to provide an i n d i c a t i o n of the p r e c i s i o n of Incidental Test Two as a measuring instrument. Each subject i n the R e l i a b i l i t y Group was provided with a sealed envelope containing the Experimental D i r e c t i o n s , Prose S e l e c t i o n One, and Prose S e l e c t i o n Two. The order of the contents of each envelope was the same. The subjects were requested to read the two prose s e l e c t i o n s , and to return the contents to the envelope when they had f i n i s h e d . The subjects then brought the envelope to a desk at the front of the room and exchanged i t f o r I ncidental Test Two. The sixteen subjects' scores on Incidental Test Two i n the R e l i a b i l i t y Group had a mean of 9.50, standard deviation of 4.99, and K-R20 of .82. This r e l i a b i l i t y estimate gave some assurance that t h i s t e s t was measuring s u f f i c i e n t l y accurately to give reasonably dependable scores i n a short term study. V a l i d i t y . One of the major problems with studies on i n c i d e n t a l learning has been the lack of a systematic basis upon which to compare the format and l o c a t i o n of the i n c i d e n t a l and relevant questions in, and between studies. To a i d i n 29 such comparisons, a Table of Specifications was prepared. This Table of Specifications described the question and answer location, in the prose selection, as well as the stimulus and response location within questions and answers, for Incidental Test Two, Incidental Test One, Relevant Question Set One, and Relevant Question Set Two. Mean scores were provided for both word location and line location Cwhere applicable). These data are presented in Table C of Appendix H. An inspection of the data i n Table C of Appendix H indicated that variations in question and answer location i n the prose selection and variations in stimulus and response location within questions and answers were not extreme across tests. Questions in each test were in stimulus-response format. An example of this format i s : Scientists who study the 'fr i l l s h a r k 1 are called (ichthyologists) ? In this question the noun "scientists" (as qualified) was termed the stimulus. The noun 'ichthyologists' was termed the response. The stimulus term pre-ceded the response term in the above question. The format for this type of question was termed 'stimulus-response'. The answers to each test question were also in stimulus response format in the prose sentences. In order to provide additional data on the similarity of Incidental Test Two, Incidental Test One, Relevant Question Set One, and Relevant Question Set Two, a comparison was also made on the parts of speech and the format of the stimulus and response terms. This comparison was made as Rothkopf & Bisbicos (1967) had investigated the differences between using measures and proper names as answers to questions as opposed to using common words and technical phrases. Rothkopf & Bisbicos (1967) did not report significant results for this comparison. However, their use of multiple tests and their lack of control for item d i f f i c u l t y confounded the inter-30 p r e t a t i o n of t h i s f i n d i n g . Data on the comparison between the parts of speech and the format of the stimulus and response terms f o r Incidental Test Two, Incidental Test One, Relevant Question Set One, and Relevant Question Set Two are presented i n Table D of Appendix H. Incidental Test Two d i f f e r e d from Incidental Test One by using considerably fewer proper nouns and considerably more common nouns as s t i m u l i . I ncidental Test One used more numerical items as responses and more multi p l e word items as s t i m u l i than Incidental Test Two. Otherwise, the v a r i a t i o n s between t e s t s were not extreme. The other major problem with studies on i n c i d e n t a l learning has been the lack of a systematic basis upon which to contrast the format and l o c a t i o n of the relevant and i n c i d e n t a l questions i n and between studies. Rothkopf (1966) and Rothkopf & Bisbicos (1967) r e l i e d on the use of a "transfer evaluation" group to ensure that there was no t r a n s f e r of s p e c i f i c t r a i n i n g between the relevant and i n c i d e n t a l t e s t s . This group wrote the i n c i d e n t a l t e s t , d i d not read the m a t e r i a l , learned the relevant t e s t item answers to a c r i t e r i o n of one hundred per cent, regardless of order, and then were tested on a comparative form of the i n c i d e n t a l t e s t . Other i n v e s t i g a t o r s (Frase, 1967; Natkin & Stahler, 1969) have r e l i e d on s e l e c t i n g i n c i d e n t a l and relevant questions from d i f f e r e n t portions of the prose s e l e c t i o n . In support of t h i s procedure, Natkin & Stahler (1969) reported a personal communication from Rothkopf i n d i c a t i n g that t h e i r procedure for d i f f e r e n t i a t i n g between in c i d e n -t a l and relevant questions was about n i n e t y - f i v e per cent e f f e c t i v e f o r ensuring d i f f e r e n t content. To a i d i n s p e c i f y i n g the distance between relevant and i n c i d e n t a l questions i n t h i s study, the word distances between the answers to questions i n Relevant Question Set One and Incidental Test One, Relevant Question Set 31 Two and Incidental Test Two were ca l c u l a t e d . These data are presented i n Tables E and F of Appendix H. The shortest distance between the answer to a question i n Relevant Question Set One and Incidental Test One was twenty-six words. The shortest distance between the answer to a question i n Relevant Question Set Two and Incidental Test Two was eleven words. In order to provide a d d i t i o n a l evidence on the d i f f e r e n c e s between Relevant Question Set One and Incidental Test One, Relevant Question Set Two and In c i d e n t a l Test Two, the number of r e p l i c a t e d s t i m u l i and response terms was analyzed. These data are presented i n Table G of Appendix H. The number of s t i m u l i and response term r e p l i c a t i o n s was quite small considering the number of questions that were compared. S t i m u l i term r e p l i c a t i o n s tended to occur more frequently within t e s t s than between t e s t s with the exception of Relevant Question Set Two. S t i m u l i terms from Relevant Question Set Two were r e p l i c a t e d twice i n Incidental Test Two. Incidental Test Two was the only t e s t to have a response term r e p l i c a t i o n s within i t . Responses from Relevant Question Set Two were r e p l i c a t e d twice i n Incidental Test Two. The number of s t i m u l i term r e p l i c a t i o n s i n responses never exceeded one. Incidental Test One was the only t e s t to contain an i d e n t i c a l response and stimulus term. Experimental Subjects One hundred twenty subjects f o r the study were randomly drawn from the Grade Ten and Eleven student population of the Mission Senior Secondary School, Mission, B.C. The student population i n Grades Ten and Eleven, at the time of the study, was four hundred seventy-eight students, i n eighteen home-room clas s e s . The Mission Senior Secondary School was somewhat a t y p i c a l of second-ary schools i n B r i t i s h Columbia, due to a rather large native Indian popu-l a t i o n . Of the four hundred seventy-eight students r e g i s t e r e d i n Grades Ten 32 and Eleven, almost s i x per cent were Indian. Ninety-two per cent of these Indian students boarded at Saint Mary's Re s i d e n t i a l School. An analysis of the composition of the Indian and non-Indian students by grade and sex i s presented i n Table E of Appendix H. One hundred twenty subjects were chosen using a t a b l e of random numbers. The breakdown of the experimental subjects by race, grade, and sex i s a lso presented i n Table H of Appendix H. The one hundred twenty experimental subjects were divided by the experimenter i n t o two groups: Indian and non-Indian. This d i v i s i o n was based on information provided by a teacher of these s-tudents, as to t h e i r r a c i a l background. The Indian subjects were assigned the numbers one to twelve. The non-Indian subjects were assigned the numbers one to one hundred and eight. Using a procedure s i m i l a r to that described f o r the s e l e c t i o n of experimental subjects, the Indian and non-Indian subjects were assigned to the four conditions of the design. A t t r i t i o n resulted i n the loss of t h i r t y -one non-Indian subjects. An a d d i t i o n a l nine subjects were cast out of the design using a table of random numbers i n order to equalize c e l l s i z e within race groupings, and to avoid the unc e r t a i n t i e s of non-orthogonal analyses. The c h a r a c t e r i s t i c s of the experimental subjects, a t t r i t i o n , subjects cast out, and subjects used i n the study, are presented i n Table I of Appendix H. 33 Experimental Procedures The study was conducted between 10:20 and 12:20 a.m. on Thursday, June 7, 1973, i n the c a f e t e r i a of the Mission Senior Secondary School. The subjects had previously been informed that they would be engaged i n a study, and parental consent had been obtained The c a f e t e r i a seats were arranged i n rows with s i x tables i n each row. Twe rty a l t e r n a t e tables were selected. The packages of experimental materials were d i s t r i b u t e d i n a non-systematic manner to each t a b l e . Boxes were placed at each table to provide a receptacle f o r the experimental materials. As the subjects entered the school c a f e t e r i a , they were requested to s i t where they found a package bearing t h e i r name. The subjects were requested to not open t h e i r packages u n t i l t o l d to do so. A f t e r a l l the subjects were seated, they were d i r e c t e d to read the d i r e c t i o n s c a r e f u l l y , to open package A, and to proceed through the materials at t h e i r own r a t e . The subjects were monitored by two adult volunteers and the experimenter. The monitors ensured that the subjects followed d i r e c t i o n s and d i d not communicate with one another. Upon completion of the materials, the subjects were allowed to leave the c a f e t e r i a and return to t h e i r c l a s s e s . The atmosphere i n the room was 'test l i k e ' . 34 S t a t i s t i c a l Procedures The data were processed at the U n i v e r s i t y of B r i t i s h Columbia Computing Centre. Two computer programs were used. The TIA (Test and Item Analysis) program was used to c a l c u l a t e i n t e r n a l r e l i a b i l i t y c o e f f i c i e n t s (K"tR20's) f o r a l l t e s t s . The BMDX64 pro-gram was used to compute the analysis of variance. The factors i n the analysis of variance are presented i n Table V. Limitations of the Study The study sample was l i m i t e d to Grade Ten and Eleven students e n r o l l e d i n the Mission Senior Secondary School, Mission, B.C. during the 1972-73 session. The materials used i n the study were l i m i t e d to those s p e c i f i e d p r e v i o u s l y . The s i z e of the Indian sample was an a d d i t i o n a l l i m i t a t i o n of t h i s study. There were only three Indian subjects i n each of the four experimental conditions. TABLE V FACTORS IN THE ANALYSIS OF VARIANCE Factors Degrees of Freedom Hypotheses and Research Questions M 1 Q 1 1 T 1 2 R 1 4 QT 1 3 OR 1 5 TR 1 6 QTR 1 7 S (QTR) 72 M = Grand Mean Q = Question Condition T = Test Expectancy R = Race S(QTR) = Subjects (nested within QTR c e l l s ) 36 CHAPTER IV PRESENTATION AND INTERPRETATION OF DATA Relevant Questions One of the major concerns of t h i s study was that the manipulations of the independent v a r i a b l e s , question placement and t e s t expectancy, would r e s u l t i n observable changes i n reading behaviour. The question immediately arose as to whether these manipulations of the independent variables were, i n f a c t , v a l i d manipulations. Analyses of variance were conducted on the answers to Relevant Question Set One and Relevant Question Set Two i n an attempt to answer t h i s question. Relevant Question Set One Question p o s i t i o n and race were the only independent variables consid-ered i n analyzing Relevant Question One. Test expectancy was not introduced u n t i l a f t e r the eighty subjects had encountered e i t h e r the Mathematical Computation Test or Incidental Test One. No hypothesis could be stated f o r race with respect to performance on Relevant Question Set One. However, i f question p o s i t i o n were a v a l i d manipulation, one would expect that subjects i n the pre-adjunct question condition would be required to guess the c o r r e c t answers to relevant questions. Subjects i n the post-adjunct question condition would have to remember the c o r r e c t answers to the relevant questions. Consequently subjects i n the post-adjunct question condition would be expected to obtain higher scores on Relevant Question Set One than would subjects i n the pre-adjunct question condition. How question placement and race would i n t e r a c t was uncertain. 37 The research question was stated as follows: Would a s i g n i f i c a n t main e f f e c t f o r question placement be found, on Relevant Question Set One with the post-adjunct question condition scoring s i g n i f i c a n t l y higher than the pre-adjunct question condition? The acceptance l e v e l f o r the s t a t i s t i c a l t e s t of t h i s research question was set at p = .05. C o e f f i c i e n t s of the i n t e r n a l consistency r e l i a b i l i t y of each group's scores on Relevant Question Set One are presented i n Table H of Appendix H. The K-R20 c o e f f i c i e n t s were at acceptable l e v e l s i n a l l conditions except Condition 21. Condition 21 was a condition i n which subjects received post-adjunct questions and d i d not have t e s t expectancy. As t e s t expectancy was not introduced u n t i l Incidental Test One or the Mathematical Computation Test was administered, Condition 21 was i d e n t i c a l to Condition 22. I t would therefore appear that the low r e l i a b i l i t y encountered i n Condition 21 may have occurred by chance alone. The r e s u l t s of the analysis of variance are presented i n Table VI. A s i g n i f i c a n t main e f f e c t was found f o r question placement OF = 12.23, d.f. = 1/76, p = .00). The post-adjunct question group scored s i g n i f i c a n t l y higher (X = 5.80) than the pre-adjunct group (X = 2.85). A comparison of the variances on Relevant Question Set One i s presented i n Table L of Appendix H. This comparison suggested that the assumption of homogeneity of variance may have been v i o l a t e d . This v i o l a t i o n would make the true l e v e l of s i g n i f i c a n c e d i f f e r e n t from the nominal alpha of .05. However, p r o b a b i l i t i e s of l e s s than .005 are l i k e l y to i n d i c a t e s i g n i f i c a n c e at a true l e v e l of .05. Therefore t h i s f i n d i n g would not l i k e l y a l t e r the s i g n i f i c a n c e of the factors presented i n Table VI. An a n a l y s i s of variance for the non-Indian subjects' scores on 38 TABLE VI ANALYSIS OF VARIANCE OF THE INDEPENDENT VARIABLES: QUESTION PLACEMENT AND RACE, ON RELEVANT QUESTION SET ONE Source Sum of Squares D.F. Mean Square F Prob Mean 759.22 1 759.22 115.77 0.00 Q 80.19 1 80.19 12.23 0.00 * R 0.016 1 0.02 0.00 0.91 PR 0.19 1 0.19 0.03 0.84 Er r o r 498.39 76 6.56 N = 80 *p-C .05 Q = Question placement R = Race 39 Relevant Question Set One supported the findings noted previously. In t h i s supplementary analysis the assumption of homogeneity of variance d i d not appear to be v i o l a t e d . In conclusion, the analysis of variance of scores on Relevant Question Set One i n d i c a t e d that, as hypothesized, a s i g n i f i c a n t main e f f e c t e x i s t e d such that the post-adjunct question group scored higher than the pre-adjunct question group. This f i n d i n g tended to confirm that the manipulation of question placement f o r relevant questions i n the text was e f f e c t i v e i n a l t e r -ing performance on those questions. However, t h i s i n t e r p r e t a t i o n should be taken with caution as the scores of the post-adjunct group, i n which the subjects d i d not receive t e s t expectancy, were below an acceptable l e v e l of r e l i a b i l i t y (K-R20 = .56). Relevant Question Set Two The independent v a r i a b l e s , question p o s i t i o n and race, were again analyzed with respect to the dependent v a r i a b l e "Score on Relevant Question Set Two". Based on the r a t i o n a l e stated previously, the following research question was asked: Would a s i g n i f i c a n t main e f f e c t f or question placement be found on Relevant Question Set Two such that the post-adjunct question condition would score s i g n i f i c a n t l y higher than the pre-adjunct question condition? However, Relevant Question Set Two d i f f e r e d from Relevant Question Set One i n that an a d d i t i o n a l independent v a r i a b l e , t e s t expectancy, was contained i n t h i s a n a l y s i s . Test expectancy was i n f e r r e d from the experimental d i r e c t i o n s which stated that the second t e s t would be s i m i l a r to the f i r s t t e s t . I f t e s t expectancy were a v a l i d manipulation, one would expect that 40 subjects who received Incidental Test One would read more c a r e f u l l y and con-sequently obtain more corre c t answers to the relevant questions than subjects who received the Mathematical CGoniputation Test. Based on t h i s r a t i o n a l e , the following research question was asked: Would a s i g n i f i c a n t main e f f e c t f o r t e s t expectancy be found on Relevant Question Set Two such that the t e s t expectancy condition would score s i g n i f i c a n t l y higher than the no-t e s t expectancy condition? The acceptance l e v e l f o r the s t a t i s t i c a l t e s t s of these research questions was set at p = .05. C o e f f i c i e n t s of the i n t e r n a l consistency r e l i a b i l i t y of each group's scores on Relevant Question Set Two are presented i n Table J of Appendix H. The K-R20 c o e f f i c i e n t s were low f o r the Indian subjects' scores i n Conditions 11, 21, and 22 and the non-Indian subjects' scores i n Condition 11. The low r e l i a b i l i t i e s f o r the Indian subjects' scores are based on such a small sample s i z e that they may be sampling a r t i f a c t s . However, the low r e l i a b i l i t y f o r the non-Indian subjects i n Condition 11 places constraint on the i n t e r p r e t a t i o n of r e s u l t s . The r e s u l t s of the analysis of variance are presented i n Table VII. A s i g n i f i c a n t main e f f e c t was found f o r question placement (F = 20.91, d.f. = 1/72, p = .00). The post-adjunct question group (X = 6.17) scored s i g n i f i c a n t l y higher than the pre-adjunct question group (X = 3.00). A s i g n i f i c a n t main e f f e c t was found f o r t e s t expectancy (F = 7.49, d.f. = 1/72, p = 0.01). The t e s t expectancy condition (X = 5.33) scored higher than the no-test expectancy condition (X = 3.85) The comparison of the variances on Relevant Question Set Two i s presented i n Table L of Appendix H. The discrepancies between variances suggests that the assumption of homogeneity may have been v i o l a t e d . However, 41 i t seemed highly u n l i k e l y that t h i s v i o l a t i o n would a l t e r the i n t e r p r e t a t i o n of the s i g n i f i c a n c e of e i t h e r of the main e f f e c t s reported. An analysis of variance f o r the non-Indian subjects' scores on Relevant Question Set Two supported the findings noted previously. In t h i s supplementary analysis the assumption of homogeneity of variance d i d not appear to be v i o l a t e d . In conclusion, the a n a l y s i s of variance of scores on Relevant Question Set Two confirmed the analysis of variance of scores on Relevant Question Set One with respect to the main e f f e c t f o r question placement. The post-adjunct question group scored s i g n i f i c a n t l y higher than the pre-adjunct question group. This f i n d i n g tended to confirm that the manipulation of question placement for relevant questions i n the text was e f f e c t i v e i n a l t e r i n g per-formance on those questions. However, the low r e l i a b i l i t i e s reported f o r the Indian subjects i n Conditions 11, 21 and 22 and the non-Indian subjects i n Condition 11 places a constraint on the i n t e r p r e t a t i o n of t h i s r e s u l t . The a n a l y s i s of variance of scores on Relevant Question Set Two also indicated a s i g n i f i c a n t main e f f e c t f o r t e s t expectancy. The t e s t expectancy group scored higher than the no-test expectancy group. This f i n d i n g tended to confirm the manipulation of t e s t expectancy as e f f e c t i v e i n a l t e r i n g perform-ance on the relevant questions. 42 TABLE VII ANALYSIS OF VARIANCE OF THE INDEPENDENT VARIABLES: QUESTION PLACEMENT, RACE, AND TEST EXPECTANCY, ON RELEVANT QUESTION SET TWO Source Sum of Squares D.F. Mean Square Prob Mean 795.00 1 795.00 131.95 0.0 Q 126.00 1 126.00 20.91 0.00 * R 2.50 1 2.50 0.41 0.53 T 45.11 1 45.11 7.49 0.01 * QT 1.04 1 1.04 0.17 0.68 QR 2.40 1 2.40 0.40 0.54 TR 8.21 1 8.21 1.36 0.25 QRT 1.24 1 1.24 0.21 0.66 E r r o r 433.80 72 6.03 N = 8 0 * P <e . 0 5 Q = T = R = Question placement Test expectancy Race 43 Incidental Tests The major t h e o r e t i c a l concern o f t h i s t h e s i s was the e f f e c t of the i n t e r a c t i o n between question placement and t e s t expectancy on i n c i d e n t a l l e a r n i n g . Two t e s t s were used to measure i n c i d e n t a l l earning: Incidental Test One and Incidental Test Two. Race and question placement were the inde-pendent v a r i a b l e s considered i n the analysis of scores on Inc i d e n t a l Test One. Race, question placement, and t e s t expectancy were the independent v a r i a b l e s considered i n the analysis of scores on Incidental Test Two. The major research hypothesis of t h i s study was based on the analysis of scores on Incidental Test Two. Incidental Test One Incidental Test One was used i n t h i s study as a means of introducing t e s t expectancy to the subjects. However, t h i s t e s t also provided a p a r t i a l r e p l i c a t i o n of Rothkopf's (1966) study with respect to question placement e f f e c t s . Question p o s i t i o n and race were the only independent v a r i a b l e s considered i n analyzing scores on Relevant Question Set One. Test expectancy was not introduced u n t i l a f t e r the subjects had encountered e i t h e r the Mathematical Computation Test or Incidental Test One. Therefore the following hypothesis and research questions must be answered using only subjects i n Conditions 12 and 22. The hypothesis f o r Incidental Test One was the same as that postulated f o r I ncidental Test Two with respect to question place. On Incidental Test One a s i g n i f i c a n t main e f f e c t f o r question placement would be found such that the post-adjunct question group would score higher than the pre-adjunct question group. The following research questions were also asked: 44 Would a s i g n i f i c a n t main e f f e c t f o r race be found on Incidental Test One? Would a s i g n i f i c a n t i n t e r a c t i o n between question placement and race be found on Incidental Test One? The acceptance l e v e l f o r the s t a t i s t i c a l t e s t of t h i s hypothesis and these research questions was set at p = .05. C o e f f i c i e n t s of the i n t e r n a l consistency r e l i a b i l i t y of each group's scores on Inci d e n t a l Test One are presented i n Table J of Appendix H. The K-R20 c o e f f i c i e n t s f o r Condition 12 Indian and Conditions 22 non-Indian and Combined were at acceptable l e v e l s . However, the r e l i a b i l i t i e s f o r a l l other groups were low. These low r e l i a b i l i t i e s place considerable c o n s t r a i n t on the i n t e r p r e t a t i o n of r e s u l t s . The r e s u l t s of the analysis of variance are presented i n Table V I I I . None of the e f f e c t s was s i g n i f i c a n t . The main e f f e c t f o r question placement was not s i g n i f i c a n t (F = 0.44, d.f. = 1/26, p = 0.51), though the mean f o r the post-adjunct group (X = 4.00) was higher than f o r the pre-adjunct group (X = 2.95). The main e f f e c t f o r race was not s i g n i f i c a n t (F = 0.37, d.f. = 1/36, p = 0.55) , though the mean score f o r the non-Indian group (X = 3.56) was higher than the Indian group (X = 3.00). The i n t e r a c t i o n between question placement and race was also not s i g n i f i c a n t (F = 0.01, d.f. = 1/36, p = 0.87). Analyses of variance f o r the Indian and non-Indian subjects' scores on Incidental Test One supported the data noted previously. A comparison of the variances on Incidental Test One i s presented i n Table L of Appendix H. This comparison suggested that the assumption of homogeneity of variance may have been v i o l a t e d . This v i o l a t i o n would make the true l e v e l of s i g n i f i c a n c e d i f f e r e n t from the nominal alpha of .05. However, i t would seem u n l i k e l y that t h i s v i o l a t i o n would s i g n i f i c a n t l y a l t e r the i n t e r p r e t a t i o n of these f i n d i n g s . 45 TABLE VIII ANALYSIS OF VARIANCE OF THE INDEPENDENT VARIABLES: QUESTION PLACEMENT, AND RACE ON INCIDENTAL TEST ONE Source Sum of Squares D.F. Mean Square F Prob Mean 208.68 1 208.68 17.52 0.00 Q 5.24 1 5.24 0.44 0.51 R 4.41 1 4.41 0.37 0.55 QR 0.16 1 0.16 0.01 0.87 E r r o r 428.90 36 11.91 N = 40 *p ^  .05 Q = Question placement R = Race 46 In conclusion, the analysis of variance of scores on Incidental Test One indicated no significant differences. However, the r e l i a b i l i t y of these findings i s questionable. The failure to find a significant main effect for question placement on Incidental Test One was inconsistent with the findings of other investigators of incidental learning (Frase, 1967; Rothkopf, 1966; Rothkopf & Bisbicos, 1967). The most plausible explanation of the failure to replicate Rothiopf (1966) would seem to l i e in the unreliability of scores on Incidental Test One. Incidental Test One was similar to the test employed by Rothkopf (1966) and contained half the questions from Rothkopf*s (1966) Incidental Test. The remaining questions were constructed by the experimenter. The K-R20 co-efficients on this test ranged from 0.56 to 0.68 for the combined groups (N = 20). If Rothkopf's (1966) test was no more reliable than the test used in this study, the discrepancy i n results could be a function of the unreli-a b i l i t y of the dependent measure. Another plausible explanation of the failure to replicate Rothkopf (1966) would seem to l i e in the age level of the subjects and the consequent d i f f i c u l t y of the test. In the experimental conditions of this study the highest mean score on Incidental Test One was only 3.53 based on 40 subjects. This would indicate that Incidental Test One was extremely d i f f i c u l t for these subjects. As Rothkopf (1966) used paid volunteer university subjects, i t seems feasible that they were highly motivated and found the test much easier than did the subjects in this study. Thus the d i f f i c u l t y level of the test may have influenced the a b i l i t y of the test to discriminate between groups. This would pa r t i a l l y account for the low r e l i a b i l i t y of scores. The failure to find a significant main effect for question placement on Incidental Test One was consistent with the failure to find significant 47 differences for question placement u n t i l the 12th question reported by Morasky & Willcox (J.970) . However, for this finding to support Morasky & Willcox's (1970) interpretation of a shaping procedure taking place during the reading process i t would be necessary to find a significant main effect on Incidental Test Two for question placement. This was not the case. Incidental Test Two Incidental Test Two was used in this study as the main dependent variable. It was upon Incidental Test Two that the research hypotheses and thesis of this study were based. Question placement test expectancy, and race were considered as independent variables for the analysis of raw scores on Incidental Test Two. The hypotheses based on Incidental Test Two were stated as follows: 1 . On Incidental Test Two a significant main effect for question placement would be found, such that the post-adjunct question group would score higher than thethe pre-adjunct question group. 2. On Incidental Test Two a significant main effect for test expectancy would be found, such that the test expectancy guoud would score higher than the no-test expectancy group. 3. On Incidental Test Two a significant ordinal interaction would be found between question placement and test expectancy such that test expectancy and question placement would be differentially related. as asked: In addition to these hypotheses, the following research questions were 4. Would a significant main effect for race be found on Incidental Test Two? 5. Would a significant interaction be found between question placement and race on Incidental Test Two? 6. Would a significant interaction be found between test expectancy and race on Incidental Test Two? 43 7. Would a significant three-way interaction be found between question placement, test expectancy, and race on Incidental Test Two? The acceptance level for the s t a t i s t i c a l test of these hypotheses and research questions was set at p = .05. Coefficients of the internal consistency r e l i a b i l i t y of each group's scores on Incidental Test Two are presented i n Table J of Appendix H. The K-R20 coefficients were quite low for the Indian subjects across conditions. However, the low r e l i a b i l i t i e s for the Indian subjects' scores are based on such a small sample size that they may be sampling artifacts. The results of the analysis of variance are presented in Table VI. A significant main effect for question placement was not found (F =0.04, d.f. = 1/72, p = 0.81). The mean for the post-adjunct group (X = 5.53) was only slightly higher than for the pre-adjunct group (X = 5.20) . A significant main effect for testing was not found (F = 1.55, d.f. = 1/72, p = 0.21). The mean score for the test expectancy group (X = 5.58) was slightly higher than the mean score for the no-test expectancy group (X = 5.15). A significant main effect for race was not found (F = 0.58, d.f. = 1/72, p = 0.45). The mean score for the non-Indian group (X = 5.49) was slightly higher than the mean score for the Indian group (X = 4.58). Only one of the interactions tested was found to be significant. This interaction was between test expectancy and race (F = 5.10, d.f. = 1/72, p = 0.02). The mean scores for this interaction are presented in Figure 1. A comparison of the variances on Incidental Test Two i s presented in Table L of Appendix H. A large discrepancy in variance was noted between Conditions 12 (.Indian) and 21 (non-Indian) . This discrepancy suggested that the assumption of homogeneity of variance may have been violated. It was thought highly unlikely that this violation would significantly alter the 49 i n t e r p r e t a t i o n of findings reported as non - s i g n i f i c a n t . However, t h i s v i o l a -t i o n might have tended to make the true l e v e l of s i g n i f i c a n c e d i f f e r e n t from the nominal alpha l e v e l of .05. This change could a l t e r the s i g n i f i c a n c e of the i n t e r a c t i o n between t e s t expectancy and race. As w i l l be noted i n the following chapter, a two way analysis of v a r i -ance was performed on Incidental Test Two scores within the non-Indian group. This analysis i s presented i n Table M of Appendix H. The assumption of homo-geneity of variance appeared j u s t i f i e d within t h i s group. The findings agreed completely with those c i t e d f o r the main e f f e c t of question placement and t e s t expectancy on Incidental Test Two. A two way analysis of variance was also performed on Incidental Test Two scores within the Indian group. This analysis i s presented i n Table N of Appendix H. The assumption of homogeneity of v a r i -ance was not v i o l a t e d within t h i s group. The findings i n d i c a t e d a s i g n i f i c a n t main e f f e c t f o r t e s t expectancy (F = 15.63, d.f. = 1/8, p = .00). The scores i n the no t e s t expectancy group (X- = 6.67) were higher than the scores i n the t e s t expectancy group (X = 2.50). This r e s u l t may help to explain the i n t e r -a ction discussed and i l l u s t r a t e d on pages 49 and 50. In conclusion, the analysis of variance of scores on Incidental Test Two indic a t e d only one major s i g n i f i c a n t f i n d i n g . When a t e s t was expected, non-Indian subjects d i d somewhat bet t e r (though not s i g n i f i c a n t l y better) while Indians d i d s i g n i f i c a n t l y and s u b s t a n t i a l l y l e s s w e l l . However, the low r e l i a b i l i t i e s of the Indian groups' t e s t scores and the large discrepancies i n variance noted f o r the non-Indian groups'test scores placed some constraint on the i n t e r p r e t a t i o n of t h i s f i n d i n g . This c o n s t r a i n t was further emphas-ized by the small sample s i z e f o r the Indian subjects. The analysis of variance of scores on Incidental Test Two f a i l e d to support the hypothesized main e f f e c t s f or question placement, t e s t expect-5Q TABLE IX ANALYSIS OF VARIANCE OF THE INDEPENDENT VARIABLES: QUESTION PLACEMENT, RACE, AND TEST EXPECTANCY ON INCIDENTAL TEST TWO Source Sum of Squares D.F. Mean Square F F Prob Mean 1034.05 1 1034.05 71.66 0.00 Q 0.61 1 0.61 0.04 0.82 T 22.35 1 22.35 1.55 0.21 R 8.30 1 8.30 0.58 0.45 QT 13.88 1 13.88 0.96 0.33 TR 73.60 1 73.60 5.10 0.02 * RQ 0.06 1 0.07 0.00 0.90 RQT 4.53 1 4.53 0.31 0.58 Error 1038.90 72 14.43 N = 80 *p < .05 Q = Question placement T = Test expectancy R = Race FIGURE I (6.67) non-Indian (N=68) NO- TEST TEST EXPECTANCY EXPECTANCY 'Fig. I. Interaction of Race by Test expectancy f o r I n c i d e n t i a l Test Two. 52 ancy. As the i n t e r n a l consistency r e l i a b i l i t y c o e f f i c i e n t s f o r the combined groups'scores on Incidental Test Two ranged from 0.66 to 0.82, i t would seem u n l i k e l y that the f a i l u r e to f i n d s i g n i f i c a n t e f f e c t s was e n t i r e l y due to the u n r e l i a b i l i t y of the dependent measure. The most p l a u s i b l e explanation of the f a i l u r e to obtain s i g n i f i c a n t e f f e c t s on Incidental Test Two would seem to be i n the d i f f i c u l t y l e v e l of the t e s t . (The d i f f i c u l t y of the t e s t probably also contributed to the low i n t e r n a l consistency r e l i a b i l i t i e s . ) The seventeen subjects i n the Baseline Data Group achieved a mean score of .94 (S.D. = .83) on Incidental Test Two. This score was obtained by chance alone as these subjects had not previously read the prose s e l e c t i o n . The sixteen subjects i n the R e l i a b i l i t y Group achieved a mean score of 9.50 (S.D. = 4.99). They had previously had the opportunity to read the prose s e l e c t i o n . Rothkopf & Bisbicos (1967) reported that under c e r t a i n conditions ( s i m i l a r to those used i n t h i s study) the incorporation of questions and other t e s t - l i k e features i n written i n s t r u c -t i o n a l material increased the amount learned from the text (Hershberger, 1964; K e i s l a r , 1960; Rothkopf 1963, 1965, 1966; and Rothkopf & Coke, 1963). These findings suggested that subjects who received questions should have scored higher on a t e s t than subjects who d i d not receive questions. However, i n the experimental treatments involved i n t h i s study the mean scores varied from 4.65 to 5.60 based on sample s i z e s of twenty. While di f f e r e n c e s between schools might account f o r t h i s v a r i a t i o n , c l e a r l y the means obtained from the experimental treatments were extremely low. I t was therefore concluded that the t e s t was overly d i f f i c u l t f o r the experimental subjects. On the Debriefing Questionnaire the\>majority of the subjects i n a l l conditions i n d i c a t e d that they had read the d i r e c t i o n s and had t r i e d to answer the questions. However, subjects i n a l l conditions were s p l i t evenly as 53 to whether they thought the reading was easy. The majority of subjects i n a l l conditions, with, the exception of Condition 11, ind i c a t e d that they d i d not read the prose passage c a r e f u l l y . This f a i l u r e to read c a r e f u l l y may have p a r t i a l l y accounted f o r the d i f f i c u l t y l e v e l of the t e s t and may i n part account f o r the t e s t ' s f a i l u r e to discriminate between treatment e f f e c t s . The assumption of homogeneity of variance was not v i o l a t e d f o r the two by two analysis of variance of the scores on Incidental Test Two f o r the non-Indian subjects. The i n t e r n a l consistency r e l i a b i l i t y c o e f f i c i e n t s ranged from .66 to .85 f o r the seventeen subjects i n each experimental condition. The f a i l u r e to f i n d a s i g n i f i c a n t main e f f e c t f o r question placement or t e s t expectancy i n t h i s analysis s e r i o u s l y questioned the adequacy of Rothkopfs (1966) theory. Question 17 of the Debriefing Questionnaire administered immediately a f t e r the subjects had completed Incidental Test Two added support to t h i s contention. Subjects reported that they read the second section of the prose more c a r e f u l l y under the condition of t e s t expectancy f o r the post-adjunct question group. This reported reading behaviour was i n d i r e c t contra-d i c t i o n to the concept that post-adjunct questions maintain optimal reading behaviours. The Debriefing Questionnaire i s presented i n Tables 0 and P of Appendix H. CHAPTER V 54 SUMMARY, FINDINGS, AND CONCLUSION Statement of the Problem The purpose of t h i s study was to i n v e s t i g a t e the e f f e c t s of the i n t e r -a ction between t e s t expectancy and question placement on i n c i d e n t a l l earning. The general hypothesis underlying t h i s study was that t e s t i n g and post-adjunct questions would r e s u l t i n s i m i l a r reading behaviours on the part of the subjects, while t e s t i n g and pre-adjunct questions would r e s u l t i n d i s -s i m i l a r reading behavours on the part of the subjects. T h e o r e t i c a l Framework and Model A basic assumption o f t h i s study was that the a s s o c i a t i v e process of learning may be i n f e r r e d from studies of eye movements. This assumption was based on the t h e o r e t i c a l framework- provided by Rothkopf (1966, 1970) f o r i n c i d e n t a l l e a r n i n g and a comparison of the v i s u a l f a t i g ue studies conducted by Carmichael & Dearborn (1947) and Hoffman (1946) Procedures The sample of t h i s study was l i m i t e d to the Grade Ten and Eleven students e n r o l l e d i n the Mission Senior Secondary School during the 1972-73 session. The study used a two by two by two f a c t o r i a l design. The factors were: a) question placement (pre-adjunct or post-adjunct), b) t e s t expect-ancy (test expectancy or no-test expectancy, and c) race (Indian or non-Indian) . The major hypotheses of t h i s study were based on a two by two by two a n a l y s i s of variance of scores on Incidental Test Two. The f a c t o r s i n t h i s 55 analysis were: a) question placement, b l t e s t expectancy, and c) race. P o s s i b l e confounding e f f e c t s - o f t h i s study were investigated by analy-ses of variance of scores on Relevant Question Set One, Relevant Question Set Two, and Incidental Test One. m addition, data on the Debriefing Question-naire was viewed i n terms of the independent experimental v a r i a b l e s question placement and t e s t expectancy. Findings The f i n d i n g s of t h i s study are presented i n t h i s section, i n summary form, as follows: 1. The research hypothesis. 2. The s i g n i f i c a n t s t a t i s t i c a l r e s u l t s . 3. The conclusions based on the fi n d i n g s . General conclusions are also presented following the discu s s i o n re-garding each s p e c i f i c research question. Hypothesis 1. On Incidental Test Two a s i g n i f i c a n t main e f f e c t f o r question place-ment would be found, such that the post-adjunct question group would score higher than the pre-adjunct question group. The analysis of variance d i d not r e s u l t i n a s i g n i f i c a n t main e f f e c t f o r question placement (F = 0.04, d.f. = 1/72, p = 0.82). The mean d i f f e r e n -ces i n question placement were i n the predicted d i r e c t i o n , but only n e g l i g i b l y d i f f e r e n t . The mean score f o r the post-adjunct group (X = 5.25) was s l i g h t l y higher than f o r the pre-adjunct group (X = 5.20). In other words, i n terms of the measures used, the scores on post-adjunct questions were not s i g n i f i c a n t l y d i f f e r e n t from the scores on pre-adjunct questions. Other i n v e s t i g a t o r s (Frase, 1967; Rothkopf, 1966; 56 Rothkopf & Bi s b i c o s , 19671 had reported that they were s i g n i f i c a n t l y d i f f e r e n t . Hypothesis 2 On Incidental Test Two a s i g n i f i c a n t main e f f e c t f o r t e s t expectancy would be found, such that the "knowledge of t e s t i n g " group would score higher than the "no-knowledge of t e s t i n g " group. The an a l y s i s of variance d i d not r e s u l t i n a s i g n i f i c a n t main e f f e c t f o r t e s t expectancy CF = 1.55, d.f. = 1/72, p = 0.21). However, the data was i n the predicted d i r e c t i o n . The mean score f o r the "knowledge of t e s t i n g " group (X = 5.58) was s l i g h t l y higher than f o r the "no-knowledge of t e s t i n g " group (X = 5.15). In terms of the measures used, t e s t expectancy f a i l e d to act i n the manner predicted from the t h e o r e t i c a l model. Hypothesis 3 On Incidental Test Two a s i g n i f i c a n t o r d i n a l i n t e r a c t i o n would be found between question placement and t e s t expectancy such that t e s t expectancy and question placement would be d i f f e r e n t i a l l y r e l a t e d . The analysis of variance d i d not r e s u l t i n a s i g n i f i c a n t i n t e r a c t i o n between question placement and t e s t expectancy (P = 0.96, d.f. = 1/72, p = 0.33). In terms of the measures used, t e s t expectancy and question placement f a i l e d to i n t e r a c t i n the manner predicted from the t h e o r e t i c a l model. Research Question 4 On Incidental Test Two would a s i g n i f i c a n t main e f f e c t f o r race be found? The analysis of variance d i d not r e s u l t i n a s i g n i f i c a n t main e f f e c t f o r race (F = 0.58, d.f. - 1/72, p = Q.45X. The mean score f o r the non-Indian group (X = 5.49) was s l i g h t l y higher than f o r the Indian group 57 (X = 4.58). In terms of the measures used, the Indian subjects d i d not d i f f e r s i g n i f i c a n t l y from the non-Indian subjects i n scores on Incidental Test Two. Research Question 5 On Incidental Test Two would a s i g n i f i c a n t i n t e r a c t i o n be found between question placement and race? The analysis of variance d i d not r e s u l t i n a s i g n i f i c a n t i n t e r a c t i o n between question placement!; and race (F = 0.004, d.f. = 1/72, p = 0.90). In terms of the measures used, t h i s f i n d i n g implied that the Indian and non-Indian subjects d i d not d i f f e r n s i g n i f i c a n t l y i n t h e i r scores on Incidental Test Two with respect to question placement. Research Question 6. On Incidental Test Two would a s i g n i f i c a n t i n t e r a c t i o n be found between t e s t expectancy and race? The a n a l y s i s of variance r e s u l t e d i n a s i g n i f i c a n t i n t e r a c t i o n between t e s t expectancy and race (F = 5.10, d.f. = 1/72, p = 0.02). Under the "no-knowledge o f t e s t i n g " condition the mean score f o r the Indian subjects (X = 6.67) was higher than f o r the non-Indian subjects (X = 4.88). Under the "knowledge of t e s t i n g " condition, the mean score f o r the Indian subjects (X = 2.50) was lower than f o r the non-Indian subjects (X = 6.09). In terms of the measures used, t h i s f i n d i n g implied that the Indian subjects reacted d i f f e r e n t l y to t e s t expectancy than the non-Indian subjects. When a t e s t was expected, non-Indian subjects d i d somewhat better (though not s i g n i f i c a n t l y better) while Indian subjects d i d s i g n i f i c a n t l y and s u b s t a n t i a l l y l e s s w e l l . Research Question 7 On Incidental Test Two, would a s i g n i f i c a n t three-way i n t e r a c t i o n be 58 found between t e s t expectancy, question placement, and race? The analysis, of variance d i d not r e s u l t i n a s i g n i f i c a n t three-way i n t e r a c t i o n between question placement, t e s t expectancy, and race (F = 0.31, d.f. = 1/72, p = 0.58). The mean scores f o r t h i s i n t e r a c t i o n are presented i n Table K of Appendix H. In terms of the measures used, t h i s f i n d i n g implied that t e s t expect-ancy, question placement, and race d i d not i n t e r a c t . Conclusions The major conclusion of t h i s study was that when a t e s t was expected, non-Indian subjects d i d somewhat b e t t e r (though not s i g n i f i c a n t l y better) while Indian subjects d i d s i g n i f i c a n t l y and s u b s t a n t i a l l y l e s s w e l l . This conclusion was rendered somewhat uncertain because of low r e l i a b i l i t y of measurement f o r some of the treatment groups and by some doubt as to the t e n a b i l i t y of the assumption of homogeneity of variance. In add i t i o n , question placement and t e s t expectancy d i d not support the t h e o r e t i c a l contention of Rothkopf (1966, 1970). Moreover, findings with regard to question placement d i d not support the findings of previous i n v e s t i g a t o r s (Frase, 1967; Rothkopf, 1966; Rothkopf & Bisb i c o s , 1967). An examination of the data was c a r r i e d out i n order to as c e r t a i n whether the unexpected findings of t h i s study were a t t r i b u t a b l e to i n e f f e c t i v e manipulations of the independent v a r i a b l e s . A two by two an a l y s i s of variance was conducted on the scores on Relevant Question Set One. The factors i n t h i s a n alysis were: a) question placement, and b) race. This analysis revealed a s i g n i f i c a n t main e f f e c t f o r question placement (F = 12.23, d.f. = 1/76, p = .00) such that the mean score f o r the post-adjunct group (X = 5.80) was higher than the pre-adjunct question group (X = 2.85). 59 In terms of the measures used, placement of relevant questions seemed an e f f e c t i v e manipulation of the independent v a r i a b l e . The manipulation of t h i s v a r i a b l e was further confirmed as e f f e c t i v e by the subjects' responses i n Conditions 11 and 12 on the Debriefing Questionnaire that they d i d not expect the type of question they received. A two by two by two analysis of variance was also conducted on scores on Relevant Question Set Two. The factors i n t h i s analysis were: a) question placement, b) t e s t expectancy, and c) race. This analysis revealed a s i g n i f i -cant main e f f e c t f o r question placement (F = 2G.91, d.f. = 1/72, p = .00) such that the mean score f o r the post-adjunct group (X = 6.17) was higher than f o r the pre-adjunct group (X = 3.00). This f i n d i n g supported the v a l i d -i t y of the manipulation of question placement as an independent v a r i a b l e reported for Relevant Question Set One. The a n a l y s i s of scores on Relevant Question Set Two also r e s u l t e d i n a s i g n i f i c a n t main e f f e c t f o r t e s t expectancy (F = 7.49, d.f. = 1/72, p = 0.01). The " t e s t expectancy" condition (X = 5.33) scored higher than the "no-test expectancy" condition (X = 3.85) In terms of the measures used, t e s t expectancy was an e f f e c t i v e gianipulation of the independent v a r i a b l e . Subject responses on the Debriefing Questionnaire d i d not support t h i s conclusion, however i n a l l conditions subjects in d i c a t e d that they expected to be tested. Unfortunately i t was not c l e a r as to what type of t e s t they expected. An examination of the data was c a r r i e d out i n order to a s c e r t a i n whether the v i o l a t i o n s of the assumption of homogeneity of variance on Incidental Test Two had r e s u l t e d i n spuriously non-signiBcant r e s u l t s . A two by two a n a l y s i s of variance was conducted, on the scores on Incidental Test Two for the non-Indian subjects. The f a c t o r s were question placement and t e s t 60 expectancy. This analysis d i d not v i o l a t e the assumption of homogeneity and confirmed the findings reported f o r Incidental Test Two with respect to question placement and t e s t expectancy. A two by two analysis of variance was also conducted on the scores on Incidental Test Two for the Indian sub-j e c t s . The f a c t o r s were question placement and t e s t expectancy. This analysis revealed a s i g n i f i c a n t main e f f e c t f o r t e s t expectancy (F = 15.63, d.f. = 1/8, p = 0.00). The mean f o r the no t e s t expectancy group (X = 6.67) was higher than f o r the t e s t expectancy group (X = 2.50). This f i n d i n g suggested that the i n t e r a c t i o n between t e s t expectancy and race was l a r g e l y due to the Indian subjects. Implications of the Study The only s t a t i s t i c a l l y s i g n i f i c a n t f i n d i n g of t h i s study was the i n t e r a c t i o n between t e s t expectancy and race noted for Incidental Test Two data. This f i n d i n g suggested that Indian subjects achieved greater i n c i d e n t a l learning under conditions where t e s t s were not expected, while non-Indian subjects achieved greater i n c i d e n t a l learning under conditions where t e s t s were expected. However, t h i s f i n d i n g should be cautiously interpreted to the classroom s i t u a t i o n . Low r e l i a b i l i t i e s on Incidental Test Two, the small sample f o r the Indian population and a p o s s i b l e v i o l a t i o n of the assumption of homogeneity of variance may have confounded t h i s f i n d i n g . However, undue importance should not be attached to reported v i o l a t i o n s of the assumption of homogeneity i n t h i s study. The s i z e of the variances increased as a function of sample s i z e . Boneau (1960) reported that with variances of one and four and sample si z e s of f i v e and f i f t e e n , the F s t a t i s t i c s required f o r s i g n i f i -cance at the .05 l e v e l were exceeded i n only .01 per cent of the cases i n v e s t i g a t e d . 61 I t would appear that the s i g n i f i c a n t i n t e r a c t i o n between t e s t expect-ancy and race reported i n the two by two by two analysis of variance on Incidental Test Two data was l a r g e l y a function of d i f f e r e n c e s i n t e s t expect-ancy a t t r i b u t a b l e to the Indian subjects. This was because a s i g n i f i c a n t main e f f e c t was found f o r the Indian and not the non-Indian subjects on separate two by two a n a l y s i s of variance on Incidental Test Two data. The f a i l u r e to f i n d a s i g n i f i c a n t main e f f e c t f o r question placement reported by Rothkopf (1966) f o r s i m i l a r materials tends to confirm Rothkopf's (1970) acknowledgment that question placement manipulations are not necessari-l y f a c i l i t a t i v e by nature. Due to the complexity of the r e l a t i o n s h i p among the questions, prose, and reader, and due to the lack of an adequate t h e o r e t i -c a l base to p r e d i c t performance; fur t h e r i n v e s t i g a t i o n of these v a r i a b l e s i s necessary before d i f f e r e n t i a l question placement i s used to manipulate prose learning i n school s i t u a t i o n s . The f a i l u r e to f i n d a s i g n i f i c a n t main e f f e c t f o r t e s t expectancy i n t h i s study suggests that the use of t e s t s f o r i n s t r u c t i o n a l rather than evaluative purposes should be further examined. The incentive associated with t e l l i n g students that they w i l l be tested on material they are reading i s not n e c e s s a r i l y f a c i l i t a t i v e . This f i n d i n g was p a r t i c u l a r l y emphasized f o r the Indian subjects. Test expectancy f o r the Indian subjects r e s u l t e d i n lower t e s t performance than not expecting a t e s t . I t would therefore appear that text expectancy may be detrimental to the performance of Indian subjects. Areas f o r Further Study While the trend of the data was i n the predicted d i r e c t i o n i n t h i s study, the r e s u l t s were not s i g n i f i c a n t with respect to the t h e o r e t i c a l formulations .62 concerning question placement, test expectancy, and the interaction between question placement and test expectancy. The question arises as to whether this was a function of the d i f f i c u l t y level of the test or inadequate r e l i a b i l i t y of the dependent measure, or whether the theoretical formulations of this study were incorrect.. The subjects' responses to Question 17 of the Debriefing Questionnaire suggest that perhaps the theoretical formulation of the study was incorrect. The non-Indian subjects reported that they read the second section of the prose more carefully under conditions of test expect-ancy even in the post-adjunct question group. This reported change in reading behaviour was in direct contradiction to the concept that post-adjunct questions maintain optimal reading behaviour. A study needs to be designed i n which the eye movements of the subjects are directly correlated with relevant question placement and performance on tests of incidental and relevant learning. A study of this nature would greatly help to c l a r i f y and test the theoretical speculations of Rothkopf (1966). A study also needs to be done which can explain the failure to replicate the results of Rothkopf C1966) in this study for question placement. The failure to replicate Rothkopf (1966) suggests that factors other than those investigated in this study may influence test performance. A factor worth consideration i s the motivational level of the subjects. Frase, Patrick & Schumer (1970) found a significant main effect for monetary incentive. Frase (1971) attempted to replicate this result but failed to find a significant main effect and he attributed this lack of a significant main effect to differences in the motivational level of the subjects in the two studies. It i s therefore conceivable that differences in the motivational level of subjects in studies of incidental learning may obscure results. A means for measuring at least achievement motivational levels of subjects 63 would seem u s e f u l f o r further i n v e s t i g a t i o n of t h i s area. Another factor to be considered i s reading time. Morasky & Willcox (1970) reported that the post-adjunct question group took longer to read the prose than d i d the pre-adjunct question group. This f i n d i n g was supported by Frase, P a t r i c k & Schumer (1970) and Rothkopf (1966). I t was not supported by Rothkopf & Bisbicos (1967). I t would seem pos s i b l e that the o v e r a l l time taken to read the prose may have a l t e r e d the r e s u l t s of these studies. I t would therefore seem u s e f u l to design a study to in v e s t i g a t e d i f f e r e n c e s i n o v e r a l l reading time with pre-adjunct and post-adjunct question placement on passages of d i f f e r e n t length and with subjects of d i f f e r e n t ages. This study might consider blocking d i f f e r e n t periods of time f o r the same passage or measuring time as a dependent v a r i a b l e f o r the pre-adjunct, post-adjunct, t e s t expectancy, and no t e s t expectancy conditions. The s i g n i f i c a n t main e f f e c t f o r t e s t expectancy reported i n the two by two analysis of variance of scores on Incidental Test Two and the s i g n i f i c a n t i n t e r a c t i o n between t e s t expectancy and race reported i n the two by two analysis of variance of scores on Incidental Test Two point to the need to further i n v e s t i g a t e the e f f e c t s of t e s t i n g on Indian subjects. 64 BIBLIOGRAPHY Boneau, C.A. The effects of violations of assumptions underlying the t test. Psychological Bulletin 1960, 57, 49-64. Carmichael, L., & Dearborn, W.F. Reading and visual fatigue. Boston: Houghton M i f f l i n , 1947. Frase, L.T. Learning from prose material: length of passage, knowledge of results, and position of questions. Journal of Educational  Psychology, 1967, 513, 266-272. Frase, L.T. Some data concerning the mathemagenic hypothesis. American Educational Research Journal, 1968a, 5_, 181-189. Frase, L.T. Effect of question location, pacing, and'mode upon retention of prose material. Journal of Educational Psychology, 1968b, 59, 244-249. Frase, L.T. Boundary conditions for mathemagenic behaviours. Review of Educational Research, 1969, 40_, 337-347. Frase, L.T. Effect of incentive variables and type of adjunct question upon text learning. Journal of Educational Psychology, 1971, 62, 371-375. Frase, L.T., Patrick, E.M. & Schumer, H. Effect of question position and frequency on learning from text under different levels of incentive. Journal of Educational Psychology, 1970, 61, 52-56. Glass, G.V., & Stanley, J.C. S t a t i s t i c a l methods i n educational psychology. New Jersey: Prentice-Hall, 1970. Hoffman, A.C. Eye movements during prolonged reading. Journal of Experimental Psychology, 1946, 36, 95-118. Hopkins, K., & Chadbourn, R. A scheme for proper u t i l i z a t i o n of multiple comparisons in research and a case study. Review of Educational Research, 1967, 4_, 407-412. Ladas, H. The mathemagenic effects of factual review questions on the learning of incidental information: a c r i t i c a l review. Review of Educational Research, 1973, 43, 71-82. Natkin, G., & Stahler, E. The effects of adjunct questions on short and long-term r e c a l l of prose materials. American Educational Research Journal, 1969, 6_, 425-432. Nichols, E.D., Collins, B., & MacPherson, D. Modern elementary algebra. Toronto: Holt, Rinehart & Winston, 1963. 65 Petrinovich, L., & Hardyck, C. Error rates for multiple comparison methods: some evidence concerning the frequency of erroneous conclusions. Psychological Bulletin, 1969, 71, 43-45. Rothkopf, E.Z. Some conjectures about inspection behavior i n learning from written sentences and the response mode problems in programmed self-instruction. Journal of Programmed instruction, 1963, 2, 31-46. Rothkopf, E.Z. Some theoretical and experimental approaches to problems in written instructions. In J.D. Krumboltz (Ed.), Learning and the  educational process. Chicago: Rand McNally, 1965. Rothkopf, E.Z. Learning from written materials: an exploration of the control of inspection behavior by test-like events. American  Educational Research Journal, 1966, J3, 241-249. Rothkopf, E.Z. The concept of mathemagenic a c t i v i t i e s . Review of Educational Research, 1970, 40, 325-336. Rothklpf, E.Z. Toward a conceptual model of learning from written discourse: a review of experimental findings on the mathemagenic effects of adjunct questions. Proceedings. (79th annual convention of the American Psychological Association), 1971, 507-508, Rothkopf, E.Z., & Bisbicos, E. Selective f a c i l i t a t i v e effects of inter-spersed questions on learning from written material. Journal of  Educational Psychology, 1967, 58, 56-61. Rothkopf, E.Z., & Bloom, R.D. Effects of interpersonal ineraction on the instructional value of adjunct questions in learning from written material. Journal of Educational Psychology, 1970, 61, 417-422. Schroeder, S.R. & Holland, J.G. Operant control of eye movements. Journal of Applied Behavior Analysis, 1968, l _ f 161-166. Tinker, M.A. Eye movement in reading. Journal of Educational Research, 1936, 30_, 241-277. APPENDIX A PS1 THE SUNLESS SEA Where great whales come s a i l i n g by s a i l and s a i l , with unshut eye. - Matthew Arnold Between the s u n l i t surface waters of the open sea and the hidden h i l l s and v a l l e y s of the ocean f l o o r l i e s the l e a s t -known region of the sea. These deep, dark waters, with a l l t h e i r mysteries and t h e i r unsolved problems, cover a very considerable part of the earth. The whole world ocean extends over about three-fourths of the surface of the globe. I f we subtract the shallow areas o f the continental shelves and the scattered banks and shoals, where at l e a s t the pale ghost of sunlight moves over the underlying bottom, there s t i l l remains about h a l f the earth that i s covered by miles-deep, l i g h t l e s s water, that has been dark since the world began. This region has withheld i t s secrets more ob s t i n a t e l y than any other. Man, with a l l h i s ingenuity, has been able to venture only to i t s threshold. Wearing a d i v i n g helmet, he can walk on the ocean f l o o r about 10 fathoms down. He can descend to an extreme l i m i t of about 500 feet i n a complete d i v i n g s u i t , so heavily armored that movement i s almost impossible, carrying with him a constant supply of oxygen. Only two men i n a l l the h i s t o r y of the world have had the experience of descending, a l i v e , beyond the range of v i s i b l e l i g h t . These men are William Beebe and O t i s Barton. In the bathysphere, they reached a depth of 3028 feet i n the open ocean o f f Bermuda, i n the year 1934. Barton alone, i n a s t e e l sphere known as the benthoscope, descended to the great depth of 4500 feet o f f C a l i f o r n i a , i n the summer of 1949. Although only a fortunate few can ever v i s i t the deep sea, the p r e c i s e instruments of the oceanographer, recording l i g h t penetration, pressure, s a l i n i t y , and temperature, have given us the materials with which to reconstruct i n imagination these e e r i e forbidding regions. Unlike the surface waters, which are s e n s i t i v e to every gust of wind, which, know day and night, respond to the p u l l of sun and moon, and change as the seasons change, the deep waters are a place where change comes slowly, i f at a l l . Down beyond the reach of the sun's rays, there i s no a l t e r n a t i o n of l i g h t and darkness. There i s rather an endless night, as o l d as the sea i t s e l f . For most of i t s creatures, groping t h e i r way end-l e s s l y through i t s black waters, i t must be a place of hunger, where food i s scarce and hard to f i n d , a s h e l t e r l e s s place where there i s no sanctuary from ever-present enemies, where one can only move on and on, from b i r t h to death, through the darkness, confined as i n a p r i s o n to h i s own p a r t i c u l a r layer of the sea. They used to say that nothing could l i v e i n the deep sea. I t was a b e l i e f that must have been easy to accept, f o r without proof to the contrary, how could anyone conceive of l i f e i n such a place? A century ago the B r i t i s h b i o l o g i s t Edward Forbes wrote: 'As we descend deeper and deeper i n t o t h i s region, the inhabitants become more and more modified, and fewer and fewer, i n d i c a t i n g our approach to an abyss where l i f e i s e i t h e r extinguished, or e x h i b i t s but a few sparks to mark i t s l i n g e r i n g presence.' Yet Forbes urged further exploration of 't h i s vast deep-sea region* to s e t t l e forever the question of the existence? of l i f e at great depths. Even then, the evidence was accumulating. S i r John Ross, during h i s exploration of the a r c t i c seas i n 1818, had brought up from a depth of 1000 fathoms mud i n which there were worms, 'thus proving there was animal l i f e i n the bed of the ocean not-withstanding the darkness, s t i l l n e s s , s i l e n c e , and immense pressure produced by more than a mile of superincumbent water.' Then from the survey ship Bulldog, examining a proposed northern route f o r a cable from Faroe to Labrador i n 1860, came another report. The Bulldog's sounding l i n e , which at one place had been allowed to l i e f o r some time on the bottom at a depth of 1260 fathoms, came up with 13 s t a r f i s h c l i n g i n g to i t . Through these s t a r f i s h , the ship's n a t u r a l i s t wrote, 'the deep has sent,.forth the long coveted message'. But not a l l the zoologists of the day were prepared to accept the message. Some doubters asserted that the s t a r f i s h had 'convulsively embraced' the l i n e somewhere on the way back to the surface. In the same year, 1860, a cable i n the Mediterranean was ra i s e d f o r repai r s from a depth of 1200 fathoms. I t was found to be heavily encrusted with cor a l s and other s e s s i l e animals that had attached themselves at an ea r l y stage of development and grown to maturity over a period of months or years. There was not the s l i g h t e s t chance that they had become entangled i n the cable as i t was being r a i s e d to the surface. Then the Challenger, the f i r s t ship ever equipped f o r oceanographic exploration, set out from England i n the year 1872 and traced a course around the globe. From bottoms l y i n g under miles of water, from s i l e n t deeps carpeted with red clay ooze, and from a l l the l i g h t l e s s intermediate depths, net-haul a f t e r net-haul of strange and f a n t a s t i c creatures came up and were s p i l l e d out on the decks. Poring over the weird beings thus brought up f o r the f i r s t time i n t o the l i g h t of day, beings no man had ever seen before, the Challenger s c i e n t i s t s r e a l i z e d that l i f e existed even on the deepest f l o o r of the abyss. The recent discovery that a l i v i n g cloud of some unknown creatures i s spread over much of the ocean at a depth of several hundred fathoms below the surface i s the most e x c i t i n g thing that has been learned about the ocean f o r many years. When, during the f i r s t quarter of the twentieth century, echo sounding was developed to allow ships while under way to record the depth of the bottom, probably no one suspected that 69 i t would also provide a means of learning something about deep-sea l i f e . But operators of the new instruments soon discovered that the sound waver, di r e c t e d downward from the ship l i k e a beam of l i g h t , were r e f l e c t e d back from any s o l i d object they met. Answering echoes were returned from i n t e r -mediate depths, presumably from schools of f i s h , whales, or submarines; then a second echo was received from the bottom. These f a c t s were so well established by the l a t e 1930's that fishermen had begun to t a l k about using t h e i r fathometers to search f o r schools of herring. Then the war brought the whole subject under s t r i c t s e c u r i t y regulations, and l i t t l e more was heard about i t . In 1946, however, the United States Navy issued a s i g n i f i c a n t b u l l e t i n . I t was reported that several s c i e n t i s t s , working with sonic equipment i n deep water o f f the C a l i f o r n i a coast, had discovered a widespread 'layer' of some sort, which gave back an answering echo to the sound waver. This r e f l e c t i n g l a y e r , seemingly suspended between the surface and the f l o o r of the P a c i f i c , was found over an area 300 miles wide. I t l a y from 1000 to 1500 fe e t below the surface. The discovery was made by three s c i e n t i s t s , C.F.Eyring, R.J.Christensen, and R.W. R a i t t , aboard the U.S.S. Jasper i n 1942, and f o r a time t h i s mysterious phenomenon, of wholly unknown nature, was c a l l e d the ECR l a y e r . Then i n 1945 Martin W. Johnson, marine b i o l o g i s t of the Scripps I n s t i t u t i o n of Oceanography, made a further discovery which gave the f i r s t clue to the nature of the l a y e r . Working aboard the vessel E.W.Scripps, Johnson found that whatever sent back the echoes moved upward and downward i n rhythmic fashion, being found near the surface at night, i n deep water during the day. This discovery disposed of speculations that the r e f l e c t i o n s came from something inanimate, perhaps a mere p h y s i c a l d i s c o n t i n u i t y i n the water, and showed that the layer i s composed of l i v i n g creatures capable of c o n t r o l l e d movement. From t h i s time on, discoveries about the sea's 'phantom bottom' came r a p i d l y . With widespread use of echo-sounding instruments, i t has become c l e a r that the phenomenon i s not something p e c u l i a r to the coast of C a l i f o r n i a alone. I t occurs almost u n i v e r s a l l y i n the deep ocean basins - d r i f t i n g by day at a depth of several hundred fathoms, at night r i s i n g to the surface, and again, before sunrise, sinking i n t o the depths. On the passage of the U.S.S. Henderson from San Diego to the A n t a r c t i c i n 1947, the r e f l e c t i n g layer was detected during the greater part of each day, at depths varying from 150 to 450 fathoms, and on a l a t e r run from San Diego to Yokosuka, Japan, the Henderson's fathometer again recorded the l a y e r every day, suggesting that i t e x i s t s almost continuously across the P a c i f i c . During July and August 1947, the U.S.S. Nereus made a continuous fathogram from Pearl Harbor to the A r c t i c and found the s c a t t e r i n g layer over a l l deep waters along t h i s course. I t d i d not develop, however, i n the shallow Bering and Chuckchee seas. Sometimes i n the morning, the Nereus fathogram showed two l a y e r s , responding i n d i f f e r e n t ways to the growing i l l u m i n a t i o n of the water; both descended i n t o deep water, but there was an i n t e r v a l of twenty minutes between the two descents. Despite attempts to sample i t or photograph i t , no one i s sure what the layer i s , although the discovery may be made any day. There are three p r i n c i p a l theories, each of which has i t s group of supporters. According to these theories, the sea's phantom bottom may c o n s i s t of small planktonic shrimps, of f i s h e s , or of squids. As f o r the plankton theory, one of the most convincing arguments i s the well-known f a c t that many plankton creatures make regular v e r t i c a l migrations of hundreds of feet, r i s i n g toward the surface at night, sinking down below the zone of l i g h t penetration very e a r l y i n the morning. This i s , of course, exactly the behavior of the s c a t t e r i n g l a y e r . What-ever composes i t i s apparently strongly r e p e l l e d by sunlight. The creatures of the layer seem almost to be held prisoner at the end r - or beyong the end - of the sun's rays throughout the hours of daylight, waiting only f o r the welcome return of darkness to hurry upward into the surface waters. But what i s the power that repels; and what the a t t r a c t i o n that draws them surfaceward once the i n h i b i t i n g force i s removed? Is i t comparative safety from enemies that makes them seek darkness? Is i t more abundant food near the surface that lures them back under cover of night? Those who say that f i s h are the r e f l e c t o r s of the sound waves usually account f o r the v e r t i c a l migrations of the layer by suggesting that the f i s h are feeding on planktonic shrimp and are following t h e i r food. They believe that the a i r bladder of a f i s h i s , of a l l structures concerned, most l i k e l y from i t s construction to return a strong echo. There i s one outstanding d i f f i c u l t y i n the way of accepting t h i s theory: we have no other evidence that concentrations of f i s h are u n i v e r s a l l y present i n the oceans. In f a c t , almost everything else we know suggests that the r e a l l y dense populations of f i s h l i v e over the continental shelves or i n c e r t a i n very d e f i n i t e l y determined zones of the open ocean where food i s p a r t i c u l a r l y abundant. I f the r e f l e c t i n g layer i s eventually proved to be composed of f i s h , the p r e v a i l i n g views of f i s h d i s t r i b u t i o n w i l l have to be r a d i c a l l y revised. The most s t a r t l i n g theory (and the one that seems to have the fewest supporters) i s that the layer consists of concentrations of squid, 'hovering below the illuminated zone of the sea and awaiting the a r r i v a l of darkness i n which to resume t h e i r r a i d s i n t o the plankton-rich surface waters.' Proponents of t h i s theory argue that squid are abundant enough, and of wide enough d i s t r i b u t i o n , to give the echoes that have been picked up almost everywhere from the equator to the two poles. Squid are known to be the sole food of the sperm whale, found i n the open oceans i n a l l temperate and t r o p i c a l waters. They a l s o form the exclusive d i e t of the bottlenosed whale and are eaten extensively by most other toothed whales, by s e a l s , and by many sea b i r d s . A l l these f a c t s argue that they must be p r o d i g i o u s l y abundant. I t i s true that men who have worked close to the sea surface at night have received v i v i d impressions of the abundance and a c t i v i t y of squids i n the surface waters i n darkness. Long ago Johan Hjort wrote: One night we were hauling long l i n e s on the Faroe slope, working with an e l e c t r i c lamp hanging over the side i n order to see the l i n e , when l i k e l i g h t n i n g flashes one squid a f t e r another shot towards the l i g h t . . . In October, 1902 we were one night steaming outside the slopes of the coast banks of Norway, and f o r many miles we could see the squids moving i n the surface waters l i k e luminous bubbles, resembling large milky white e l e c t r i c lamps being constantly l i t and extingui shed. Thor Heyerdahl reports that at night h i s r a f t was l i t e r a l l y bombarded by squids; and Richard Fleming says that i n h i s oceanographic work of the coast of Panama i t was common to see immense schools of squid gathering at the surface at night and leaping upward toward the l i g h t s that were used by the men to operate t h e i r instruments. But equally spectacular surface displays of shrimp have been seen, and most people f i n d i t d i f f i c u l t to be l i e v e i n the ocean-wide abundance of squid. Deep water photography holds much promise f o r the s o l u t i o n of the mystery of the phantom bottom. There are t e c h n i c a l d i f f i c u l t i e s , such as the problem of holding a camera s t i l l as i t swings at the end of a long cable, t w i s t i n g and turning, suspended from a ship which i t s e l f moves with the sea. Some of the p i c t u r e s so taken look as though the photographer has pointed h i s camera at a s t a r r y sky and swung i t i n an arc as he exposed the f i l m . Yet the Norwegian b i o l o g i s t , Gunnar Rollefson had an encouraging experience i n c o r r e l a t i n g photography with echograms. On the research ship Johan Hjort o f f the Lofsten Islands, he p e r s i s t e n t l y got r e f l e c t i o n of sound from schools of f i s h i n 20 to 30 fathoms. A s p e c i a l l y constructed camera was lowered to the depth indi c a t e d by the echogram. When developed, the f i l m showed moving shapes of f i s h at a distance, and a large and c l e a r l y recognizable cod appeared i n the beam of l i g h t and hovered i n f r o n t of the lens. D i r e c t sampling of the layer i s the l o g i c a l means of discovering i t s i d e n t i t y , but the problem i s to develop large nets that can be operated r a p i d l y enough to capture swift-moving animals. S c i e n t i s t s at Woods Hole, Massachus-e t t s , have towed ordinary plankton nets i n the layer and have found that euphausiid shrimps, glassworms, and other deep-water plankton are concentrated there; but there i s s t i l l a p o s s i b i l i t y that the l a y e r i t s e l f may a c t u a l l y be made up of l a r g e r forms feeding on the shrimps - too large or s w i f t to be taken i n the presently used nets. New nets may give the answer. T e l e v i s i o n i s another p o s s i b i l i t y . Shadowy and i n d e f i n i t e though they be, these recent i n d i c a t i o n s of an abundant l i f e at mid-depths agree with the reports of the only observers who have a c t u a l l y v i s i t e d comparable depths and brought back eyewitness accounts of what they saw. William Beebe's impressions from the bathysphere were of a l i f e f a r more abundant and v a r i e d than he had been prepared to f i n d although, over a period of s i x years, he had made many hundreds of net-hauls i n the same area. More than a quarter of a mile down, he reported aggregations of l i v i n g things 'as t h i c k as I have ever seen them'. At h a l f a mile - the deepest descent of the bathy-sphere - Dr. Beebe r e c a l l e d that 'there was no i n s t a n t when a mist of plankton ... was not s w i r l i n g i n the path of the beam'. The existence of an abundant deep-sea fauna was discovered, probably m i l l i o n s of years ago, by c e r t a i n whales and a l s o , i t now appears, by seals. The ancestors of a l l whales, we know by f o s s i l remains, were land mammals. They must have been predatory beasts, i f we are to judge by t h e i r powerful jaws and teeth. Perhaps i n t h e i r foragings about the deltas of great r i v e r s or around the edges of shallow seas, they discovered the abundance of f i s h and other marine l i f e and over the centuries formed the habit of following them fart h e r and f a r t h e r i n t o the sea. L i t t l e by l i t t l e t h e i r bodies took on a form more s u i t a b l e f o r aquatic l i f e ; t h e i r hind limbs were reduced to rudiments, which may be discovered i n a modern whale by d i s s e c t i o n , and the forelimbs were modified i n t o organs f o r steering and balancing. Eventually the whales, as though to divide the sea's food resources among them, became separated i n t o three groups: the plankton-eaters, the f i s h - e a t e r s , and the squid-eaters. The plankton-eating whales can e x i s t only where there are dense masses of small shrimp or copepods to supply t h e i r enormous food requirements. This l i m i t s them, except f o r scattered areas, to a r c t i c and a n t a r c t i c waters and the high temperate l a t i t u d e s . Fish-eating whales may f i n d food over a somewhat wider range of ocean, but they are r e s t r i c t e d to places where there are enormous populations of schooling f i s h . The blue water of the t r o p i c s and of the open ocean basins o f f e r s l i t t l e to e i t h e r of these groups. But that immense, square-headed, formidably toothed whale known as the cachalot or sperm whale discovered long ago what men have known f o r only a short time - that hundreds of fathoms below the almost untenanted surface waters of these regions there i s an abundant animal l i f e . The sperm whale has taken these deep waters f o r h i s hunting grounds; h i s quarry i s the deep-water population of squids, i n c l u d i n g the giant squid A r c h i t e u t h i s , which l i v e s p e l a g i c a l l y a t depths of 1500 f e e t or more. The head of the sperm whale i s often marked with long s t r i p e s , which c o n s i s t of a great number of c i r c u l a r scars made by the suckers of the squid. From t h i s evidence we can imagine the b a t t l e s that go on, i n the darkness of the deep water, between these two huge creatures - the sperm whale with i t s 70-ton bulk, the squid with a body as long as 30 f e e t , and writhing, grasping arms extending the t o t a l length of the animal to perhaps 50 f e e t . The greatest depth at which the giant squid l i v e s i s not d e f i n i t e l y known, but there i s one i n s t r u c t i v e piece of evidence about the depth to which sperm whales descend, presumably i n search of the squids. In A p r i l 1932, the cable r e p a i r ship A l l America was i n v e s t i g a t i n g an apparent break i n the submarine cable between Balboa i n the Canal Zone and Esmeraldas, Ecuador. The cable was brought to the surface o f f the coast of Colombia. Entangled i n i t was a dead 45-foot male sperm whale. The submarine cable was twisted around the lower jaw and was wrapped around one f l i p p e r , the body, and the daudal fl u k e s . The cable was r a i s e d from a depth of 540 fathoms, or 3240 feet . Some of the seals also appear to have discovered the PS11 hidden food reserves of the deep ocean. I t has long been something of a mystery where, and on what, the northern f ur seals of the eastern P a c i f i c feed during the winter, which they spend o f f the coast of North America from C a l i f o r n i a to Alaska. There i s no evidence that they are feeding to any great extent on sardines, mackerel, or other commercially important f i s h e s . Presumably four m i l l i o n seals could not compete with commercial fishermen f o r the same species without the f a c t being known. But there i s some evidence on the d i e t of the f u r seals, and i t i s highly s i g n i f i c a n t . Their stomachs have y i e l d e d the bones of a species of f i s h that has never been seen a l i v e . Indeed, not even i t s remains have been found anywhere except i n the stomachs of s e a l s . Ichthyologists say that t h i s 'seal f i s h ' belong to a group that t y p i c a l l y inhabits very deep water, o f f the edge of the continental s h e l f . How e i t h e r whales or seals endure the tremendous pressure changes involved i n dives of several hundred fathoms i s not d e f i n i t e l y known. They are warm-blooded mammals l i k e ourselves. Caisson disease, which i s caused by the rapid accumulation of nitrogen bubbles i n the blood with sudden release of pressure, k i l l s human divers i f they are brought up r a p i d l y from depths of 200 f e e t or so. Yet, according to the testimony of whalers, a baleen whale, when harpooned, can dive s t r a i g h t down to a depth of h a l f a mile, as measured by the amount of l i n e c a r r i e d out. From these depths, where i t has sustained a pressure of h a l f a ton on every inch of body, i t returns almost immediately to the surface. The most p l a u s i b l e explanation i s that, unlike the d i v e r , who has a i r pumped to him while he i s under water, the whale has i n i t s body only the l i m i t e d supply i t c a r r i e s down, and does not have enough nitrogen i n i t s blood to do serious harm. The p l a i n t r u t h i s , however, that we r e a l l y do not know, since i t i s obviously impossible to confine a l i v i n g whale and experiment on i t , and almost as d i f f i c u l t to d i s s e c t a dead one s a t i s f a c t o r i l y . At f i r s t thought i t seems a paradox that creatures of such great f r a g i l i t y as the glass sponge and the j e l l y f i s h can l i v e under the conditions of immense pressure that p r e v a i l i n deep water. For creatures at home i n the deep sea, however, the saving f a c t i s that the pressure i n s i d e t h e i r t i s s u e s i s the same as that without, and, as long as t h i s balance i s preserved, they are no more inconvenienced by a pressure of a ton or so than we are by ordinary atmospheric pressure. And most abyssal creatures, i t must be remembered, l i v e out t h e i r whole l i v e s i n a comparatively r e s t r i c t e d zone and are never required to adjust themselves to extreme changes i n pressure. But of course there are exceptions, and the r e a l miracle of sea l i f e i n r e l a t i o n to great pressure i s not the animal that l i v e s i t s whole l i f e on the bottom, bearing a pressure of perhaps f i v e or s i x tons, but those that r e g u l a r l y move up and down through hundreds or thousands of f e e t of v e r t i c a l change. The small shrimps and other planktonic creatures that descend i n t o deep water during the way are examples. F i s h that possess a i r bladders, on the other hand, are v i t a l l y a f f e c t e d by abrupt changes of pressure, as anyone knows who has seen a trawler's net r a i s e d from a hundred fathoms. Apart from the accident of being captured i n a net and hauled up through waters of r a p i d l y diminishing pressures, f i s h may sometimes wander out of the zone to which they are adjusted and f i n d themselves unable to return. Perhaps i n t h e i r p u r s u i t of food they roam upward to the c e i l i n g of the zone that i s t h e i r s , and beyond whose i n v i s i b l e boundary they may not stray without meeting a l i e n and inhospitable conditions. Moving from layer to layer of d r i f t i n g plankton as they feed, they may pass beyond the boundary. In the lessened pressure of these upper waters the gas enclosed within the a i r bladder expands. The f i s h becomes l i g h t e r and more buoyant. Perhaps he t r i e s to f i g h t h i s way down again, opposing the upward l i f t with a l l the power ,/of h i s muscles. I f he does not succeed, he ' f a l l s ' to the surface, injured and dying, f o r the abrupt release of pressure from without causes distension and rupture of the t i s s u e s . The compression of the sea under i t s own weight i s r e l a t i v e l y s l i g h t , and there i s no basis f o r the o l d and picturesque b e l i e f that, at the deeper l e v e l s , the water r e -s i s t s the downward passage of objects from the surface. According to t h i s b e l i e f , sinking ships, the bodies of drowned men, and presumably the bodies of the la r g e r sea animals not consumed above by hungry scavengers, never reach the bottom, but come to r e s t at some l e v e l determined by the r e l a t i o n o f t h e i r own weight-.to the compression of the water, there to d r i f t forever. The f a c t i s that anything w i l l continue to sink as long as i t s s p e c i f i c g r a v i t y i s greater than that of the surrounding water, and a l l large bodies descend, i n a matter of a few days, to the ocean f l o o r . As mute testimony to t h i s f a c t , we br i n g up from the deepest ocean basins the teeth of sharks and the hard ear bones of whales. Nevertheless the weight of sea water - the pressing down of miles of water upon a l l the underlying layers - does have a c e r t a i n e f f e c t upon the water i t s e l f . I f t h i s downward com-pression could suddenly be relaxed by some miraculous suspension of natural laws, the sea l e v e l would r i s e about 93 f e e t a l l over the world. This would s h i f t the A t l a n t i c c o a s t l i n e of the United States westward a hundred miles or more and a l t e r other f a m i l i a r geographic outlines a l l over the world. 79 Immense pressure, then, i s one of the governing conditions of l i f e i n the deep sea; darkness i s another. The unrelieved darkness of the deep waters has produced weird and i n c r e d i b l e modifications of the abyssal fauna. I t i s a blackness so divorced from the world of the sunlight that probably only the few men who have seen i t with t h e i r own eyes can v i s u a l i z e i t . We know that l i g h t fades out r a p i d l y with descent below the surface. The red rays are gone at the end of the f i r s t 200 or 300 f e e t f and with them a l l the orange and yellow warmth of the sun. Then the greens fade out, and at 1000 f e e t only a deep, dark, b r i l l i a n t blue i s l e f t . In very c l e a r waters the v i o l e t rays of the spectrum may penetrate another thousand f e e t . Beyond t h i s i s only the blackness of the deep sea. In a curious way, the colo r s of marine animals tend to be re l a t e d to the zone i n which they l i v e . Fishes of the surface waters, l i k e the mackerel and herring, often are blue or green; so are the f l o a t s of the Portuguese men-of-war and the azure-t i n t e d wings of the swimming s n a i l s . Down below the diatom meadows and the d r i f t i n g sargassum weed, where the water becomes ever more deeply, b r i l l i a n t l y blue, many creatures are c r y s t a l c l e a r . Their glassy, ghostly forms blend with t h e i r surroundings and make i t easier f o r them to elude the ever-present, ever-hungry enemy. Such are the transparent hordes of the arrow-worms or glassworms, the comb j e l l i e s , and the larvae of many f i s h e s . At a thousand f e e t , and on down to the very end of the sun's rays, s i l v e r y f i s h e s are common, and many others are red, drab brown, or black. Pteropods are a dark v i o l e t . Arrowworms, whose r e l a t i v e s i n the upper layers are c o l o r l e s s , are here a deep red. J e l l y f i s h medusae, which above would be transparent, at a depth of 1000 fee t are a deep brown. At depths greater than 1500 f e e t , a l l the f i s h e s are black, deep v i o l e t , or brown, but the prawns wear amazing hues of red, s c a r l e t , and purple. Why, no one can say. Since a l l the red rays are strained out of the water f a r above t h i s depth, the s c a r l e t raiment of these creatures can only look black to t h e i r neighbors. The deep sea has i t s s t a r s , and perhaps here and there an eerie and t r a n s i e n t equivalent of moonlight, f o r the mysterious phenomenon of luminescence i s displayed by perhaps h a l f of a l l the f i s h e s that l i v e i n dimly l i t or darkened waters, and by many of the lower forms as w e l l . Many f i s h e s carry luminous torches that can be turned on or o f f a t w i l l , presumably helping them f i n d or pursue t h e i r prey. Others have rows of l i g h t s over t h e i r bodies, i n patterns that vary from species to species and may be a s o r t of recognition mark or badge by which the bearer can be known as f r i e n d or enemy. The deep-sea squid e j e c t s a spurt of f l u i d that becomes a luminous cloud, the counterpart of the 'ink' of h i s shallow-water r e l a t i v e . Down beyond the reach of even the longest and strongest of the sun's rays, the eyes of f i s h e s become enlarged, as though to make the most of any chance i l l u m i n a t i o n of whatever so r t , or they may become t e l e s c o p i c , large of lens, and protruding. In deep-sea f i s h e s , hunting always i n dark waters, the eyes tend to lose the 'cones' or c o l o r - p e r c e i v i n g c e l l s of the r e t i n a , and to increase the 'rods', which perceive dim l i g h t . Exactly the same modification i s seen on land among the s t r i c t l y nocturnal prowlers which, l i k e abyssal f i s h , never see the sunlight. In t h e i r world of darkness, i t would seem l i k e l y that some of the animals might have become b l i n d , as has happened to some cave fauna. So, indeed, many of them have, compen-sating f o r the lack of eyes with marvelously developed f e e l e r s and long slender f i n s and processes with which they grope t h e i r way, l i k e so many, b l i n d men with canes, t h e i r whole knowledge of f r i e n d s , enemies, or food coming to them through the sense of touch. The l a s t traces of p l a n t l i f e are l e f t behind i n the t h i n upper layers of water, f o r no p l a n t can l i v e below about 600 f e e t even i n very c l e a r water, and few f i n d enough sunlight f o r t h e i r food-manufacturing a c t i v i t i e s below 200 f e e t . Since no animal can make i t s own food, the creatures o f the deeper waters l i v e a strange, almost p a r a s i t i c existence of u t t e r dependence on the upper l a y e r s . These hungry carnivores prey f i e r c e l y and r e l e n t l e s s l y upon each other, yet the whole community i s u l t i m a t e l y dependent upon the slow r a i n of descending food p a r t i c l e s from above. The components of t h i s never-ending r a i n are the dead and dying p l a n t s and animals from the surface, or from one of the intermediate l a y e r s . For each of the h o r i z o n t a l zones or communities of the sea that l i e , i n t i e r a f t e r t i e r , between the surface and the sea bottom, the food supply i s d i f f e r e n t and i n general poorer than f o r the la y e r above. There i s a h i n t of the f i e r c e and uncompromising competition f o r food i n the saber-toothed jaws of some of the small, dragonlike f i s h e s of the deeper waters, i n the immense mouths and i n the e l a s t i c and d i s t e n s i b l e bodies that make i t p o s s i b l e f o r a f i s h to swallow another several times i t s s i z e , enjoying swift r e p l e t i o n a f t e r a long f a s t . Pressure, darkness, and - we should have added only a few years ago - s i l e n c e , are the conditions of l i f e i n the deep sea. But we know now that the conception of the sea as a s i l e n t p lace i s wholly f a l s e . Wide experience with hydrophones and other l i s t e n i n g devices f o r the detection of submarines has proved that, around the shore l i n e s of much of the world, there i s an extraordinary uproar produced by f i s h e s , shrimps, porpoises, and probably other forms not yet i d e n t i f i e d . There has been l i t t l e i n v e s t i g a t i o n as yet of sound i n the deep, offshore areas, but when the crew of the A t l a n t i s lowered a hydrophone in t o deep water o f f Bermuda, they recorded strange mewing sounds, shrieks, and ghostly moans, the sources of which have not been traced. But f i s h of shallower zones have been captured and confined i n aquaria, where t h e i r voices have been recorded f o r comparison with sounds heard at sea, and i n many cases s a t i s f a c t o r y i d e n t i f i c a t i o n can be made. During the Second World War the hydrophone network set up by the United States Navy to protect the entrance to Chesapeake Bay was temporarily made useless when, i n the spring o f 1942, the speakers a t the surface began to give f o r t h , every evening, a sound described as being l i k e a 'pneumatic d r i l l t e a r i ng up pavement'. The extraneous noises that came over the hydrophones completely masked the sounds of the passage of ships. Eventually i t was discovered that the sounds were the voices of f i s h known as croakers, which i n the spring move i n t o Chesapeake Bay from t h e i r offshore wintering grounds. As soon as the noise had been i d e n t i f i e d and analyzed, i t was p o s s i b l e to screen i t out with an e l e c t r i c f i l t e r , so that once more only the sounds of ships came through the speakers. Later i n the same year, a chorus of croakers was discovered o f f the p i e r of the Scripps I n s t i t u t i o n at La J o l l a . Every year from May u n t i l September the evening chorus begins about sunset, and 'increases gradually to a steady uproar of harsh froggy croaks, with a background of s o f t drumming. This continues unabated f o r two o r three hours and f i n a l l y tapers o f f to i n d i v i d u a l outbursts at rare i n t e r v a l s . 'Several species of croakers i s o l a t e d i n aquaria gave sounds s i m i l a r to the 'froggy croaks', but the authors of the s o f t background drumming - presumably another species of croaker - have not yet been discovered. One of the most e x t r a o r d i n a r i l y widespread sounds of the undersea i s the c r a c k l i n g , s i z z l i n g sound, l i k e dry twigs burning or f a t f r y i n g , heard near beds of the snapping shrimp. This i s a small, round shrimp, about h a l f an inch i n diameter, with one very large claw which i t uses to stun i t s prey. The shrimp are forever c l i c k i n g the two j o i n t s of t h i s claw together, and i t i s the thousands of c l i c k s that c o l l e c t i v e l y produce the noise known as 1 shrimp cr a c k l e ' . No one had any idea the l i t t l e snapping shrimps were so abundant or so widely d i s t r i b u t e d u n t i l t h e i r signals began to be picked up on hydrophones. They have been heard a l l over a broad band that extends around the world, between l a t i t u d e s 35°N and 35°S, (for example, from Cape Hatteras to Buenos Aires) i n ocean waters l e s s than 30 fathoms deep. Mammals as we l l as f i s h e s and crustaceans contribute to the undersea chorus. B i o l o g i s t s l i s t e n i n g through a hydro-phone i n an estuary of the St.Lawrence River heard 'high-pitched resonant whistles and squeals, varied with the t i c k i n g and clucking sounds s l i g h t l y reminiscent of a s t r i n g orchestra tuning up, as well as mewing and occasional c h i r p s ' . This remarkable medley of sounds was heard only while schools of the white porpoise were seen passing up or down the r i v e r , and so was assumed to be produced by them. The mysteriousness, the eerieness, the ancient unchang-ingness of the great depths have l e d many people to suppose that some very o l d forms of l i f e - some ' l i v i n g f o s s i l s ' - may be l u r k i n g undiscovered i n the deep ocean. Some such hope may have been i n the minds of the Challenger s c i e n t i s t s . The forms they brought up i n t h e i r nets were weird enough, and most of them had never before been seen by man. But b a s i c a l l y they were modern types. There was nothing l i k e the t r i l o b i t e s o f Cambrian time or the sea scorpions o f the S i l u r i a n , nothing reminiscent of the great marine r e p t i l e s that invaded the sea i n the Mesozoic. Instead, there were modern f i s h e s , squids, and shrimps, strangely and grotesquely modified, to be sure, f o r l i f e i n the d i f f i c u l t deep-sea world, but c l e a r l y types that have developed i n rather recent geologic time. Far from being the o r i g i n a l home of l i f e , the deep sea has probably been inhabited f o r a r e l a t i v e l y short time. While l i f e was developing and f l o u r i s h i n g i n the surface waters, along the shores, and perhaps i n the r i v e r s and swamps, two immense regions of the earth s t i l l forbade invasion by l i v i n g things. These were the continents and the abyss. As we have seen, the immense d i f f i c u l t i e s of surviving on land were f i r s t overcome by c o l o n i s t s from the sea about 300 m i l l i o n years ago. The abyss, with i t s unending darkness, i t s crushing pressures, i t s g l a c i a l c o l d , presented even more formidable d i f f i c u l t i e s . Probably the successful invasion of t h i s region - at l e a s t by higher forms of l i f e -occurred somewhat l a t e r . Yet i n recent years there have been one or two s i g n i f i -cant happenings that have kept a l i v e the hope that the deep sea may a f t e r a l l , conceal strange l i n k s with the past. In December 1938, o f f the southeast t i p of A f r i c a , an amazing f i s h was caught a l i v e i n a trawl - a f i s h that was supposed to have been dead f o r at l e a s t 60 m i l l i o n years! This i s to say, the l a s t known f o s s i l remains of i t s kind date from the Cretaceous, and no l i v i n g example had been recognized i n h i s t o r i c time u n t i l t h i s lucky net haul. The fishermen who brought i t up i n t h e i r trawl from a depth of only 40 fathoms r e a l i z e d that t h i s f i v e - f o o t , b r i g h t blue f i s h , with i t s large head and strangely shaped sc a l e s , f i n s , and t a i l , was d i f f e r e n t from anything they had ever caught before, and on t h e i r return to port they took i t to the nearest museum. This s i n g l e specimen of Latimeria, as the f i s h was christened, i s so f a r the only one that has been captured, and i t seems a reasonable guess that i t may inhabit depths below those o r d i n a r i l y f i s h e d , and that the South A f r i c a n specimen was a stray from i t s usual h a b i t a t . Occasionally a very p r i m i t i v e type of shark, known from i t s puckered g i l l s as a ' f r i l l s h a r k ' , i s taken i n waters between a quarter of a mile and h a l f a mile down. Most of these have been caught i n Norwegian and Japanese waters -there are only about 50 preserved i n the museums of Europe and America - but recently one was captured o f f Santa Barbara, C a l i f o r n i a . The f r i l l s h a r k has many anatomical features s i m i l a r to those of the ancient sharks that l i v e d 25 to 30 m i l l i o n years ago. I t has too many g i l l s and too few dorsal f i n s f o r a modern shark, and i t s teeth, l i k e those of f o s s i l sharks, are three-pronged and b r i a r l i k e . Some i c h t h y o l o g i s t s regard i t as a r e l i c derived from very ancient shark ancestors that have died out i n the upper waters but, through t h i s s i n g l e species, are s t i l l c a r r y i n g on t h e i r struggle f o r ear t h l y s u r v i v a l , i n the quiet of the deep sea. Possibly there are other such anachronisms l u r k i n g down in these regions of which we know so l i t t l e , but they are l i k e l y to be few and scattered. The terms of existence i n these deep waters are f a r too uncompromising to support l i f e unless that l i f e i s p l a s t i c , molding i t s e l f constantly to the harsh conditions, s e i z i n g every advantage that makes po s s i b l e the s u r v i v a l of l i v i n g protoplasm i n a world only a l i t t l e l e s s h o s t i l e than the black reaches of interplanetary space. 86 APPENDIX B RQS1 1. The very deep portions of the ocean, into the depths of which very l i t t l e light ever penetrates, cover about th) of the sur-face of the earth. 2. The British biologist, Edward (Forbes) , writing early in the 19th century, doubted that much l i f e existed i n the greatest ocean depths. 3. The survey ship Bulldog , which recovered starfish from a depth of 1260 fathoms in 1860, was exploring a route for a cable between Faroe and (Labrador) 4. In the year 1860, a cable in the Mediter-ranean was raised from a depth of 1200 fathoms. It was found to be heavily en-crusted with corals and other Csessile) animals. 5. In the late 1930's, fishermen began to consider finding schools of herring by use of the ship's UdifafoteieiX-( • 6. During the 1947 voyage of the U.S.S. Henderson from San Diego to the Antarctic, the reflecting layer (phantom bottom) was detected during the greater part of the day at depths ranging from 150 - (450) fathoms. 7. One of the theories of the phantom bottom is that i t i s composed of a dense layer of sea animals. Three different kinds of creatures have been thought to be involved. These were: 1. planktonic shrimp, 2. fishes, and 3. (squids) 8. Johan Hjort, Thor Heyerdahl and Richard Flemming report seeing immense schools of squid gathering on the surface at night. Equally spectacular surface dis-plays of (shrimp) have also been seen. 87 9. The Norwegian b i o l o g i s t Gunnar Rollefson on the research ship Johan Hjort o f f the Lofsten Islands photographed f i s h i n 20 to 30 fathoms. One recognizable f i s h that he photographed was atn) (cod) . 10. Some evidence of the depth to which sperm whales descend was provided by the cable r e p a i r ship A l l America. A dead 45-foot whale was found entangled i n a cable r a i s e d from (540) fathoms. RQS11 1. Rapid ascent from depths of 200 feet or so i s almost i n v a r i a b l y f a t a l to human di v e r s . Death r e s u l t s from (Caisson) disease which i s the r a p i d accumulation of nitrogen bubbles i n the blood when pressure i s sud-denly reduced. 2. At f i r s t thought i t seems a paradox that creatures of such great f r a g i l i t y as the glass sponge and the ( j e l l y f i s h ) can l i v e under the conditions of immense pressure that p r e v a i l i n deep water. 3. I f the downward compression of the sea was suddenly relaxed by some miraculous suspension of natural laws, the sea l e v e l would r i s e about (93) fe e t a l l over the world. 4. L i g h t fades r a p i d l y with descent below the surface of the sea. What co l o r rays have disappeared from the l i g h t which reaches 200 or 300 fe e t below the sea's surface? (red) 5. At depths greater than 1500 f e e t , a l l the f i s h e s are black, deep v i o l e t , or brown; but the (prawns) wear amazing hues of red, s c a r l e t , and purple. 6. Like some cave fauna, many animals of the deep sea have become b l i n d . With which other sensory f a c u l t y do they tend to compensate f o r t h e i r lack of sight? (touch) 88 7. The hydroplane network set up by the Navy to protect Chesapeake Bay during World War II was temporarily made useless by noise made by f i s h (Spring, 1942). The species of f i s h which produced the noise are c a l l e d (croakers). 8. One of the most widespread underwater sounds i s shrimp cr a c k l e . This sound, which has been described as c r a c k l i n g or s i z z l i n g i s produced by the snapping shrimp. With what part of h i s body does the shrimp make t h i s noise? (claws) 9. When l i f e was developing i n ea r l y times, two regions of the earth presented enormous d i f f i c u l t i e s f o r s u r v i v a l , the continents and the abyss of the sea. How many years ago d i d successful l i f e on the continents probably s t a r t ? (300 m i l l i o n ) 10. A p r i m i t i v e type of shark generally found i n Norwegian and Japanese waters i s known as the " f r i l l s shark" because of the str u c -ture of i t s ( g i l l s ) . 89 APPENDIX C NAME: I T l F i l l i n the blanks with, the correct wordCsl or number(s). Place your answer i n the space provided i n the column at the r i g h t . ANSWER The surface waters of the ocean are s e n s i t i v e to every gust of wind, know day and night, respond to the sun and moon and change with the (seasons) The sperm whale weighs approximately ? tons. (70) During an exploration of the A r c t i c Seas i n 1818, S i r John ? dredged mud containing worms from a depth of 1000 fathoms, thus proving the existence of animal l i f e at great depths. CRoss)_ 4. In h i s descent i n the bathysphere Dr. Beebe r e c a l l e d that 'there was no ins t a n t when a mist of••••? w a s n o t s w i r l i n g i n the beam. ' (plankton) 5. The ECR l a y e r was discovered by the s c i e n t i s t s Eyring, Christensen, and R a i t t working aboard the U.S.S. Jasper i n the year ? . (1942)  6. Those who hold that the s c a t t e r i n g l a y e r (phantom bottom) consists of a dense concentration o f f i s h , p o i n t out that the ? of f i s h i s (are) of a l l structures concerned, the most l i k e l y to return a strange echo. ( a i r bladders) 7. The giant squid has a t o t a l body length of approximately ? feet . (50) 90 B. What i s another name f o r the cachalot? ? 9. Deep water ? holds much promise f o r the s o l u t i o n of the mystery of the phantom bottom. 10. Whatever composes the phantom bottom i s apparently strongly r e p e l l e d by 11. In 1946 the United States Navy issued a b u l l e t i n which reported that the r e f l e c t i n g l a y e r , seemingly suspended between the surface and the f l o o r of the P a c i f i c was found over an area ? miles wide. 12. In 1945, Dr. Martin W. Johnson of the Scripps I n s t i t u t i o n of Ocean-ography, determined that the sound r e f l e c t i n g l a y e r i n the ocean, the s o - c a l l e d phantom bottom, moved rythmically up and down. He i n f e r -red from t h i s that the phantom bottom was composed of ? 13. During the f i r s t quarter of the twentieth century a new method f o r recording the depth of the ocean bottom was developed. I t could be used by ships while underway and was c a l l e d ? sounding. 14. Just wearing a d i v i n g helmet, man can s a f e l y walk the ocean f l o o r at a depth of about ? fathoms. 15. Long ago Johan Hjort wrote: 'In October, ? we were one night steaming outside the slopes of the coast banks of Norway,...' 16. Although the phantom bottom has been reported throughout the P a c i f i c Ocean, i t i s not found i n the Bering and Chuckchee Seas. What c h a r a c t e r i s t i c s of the Bering and Chuckchee Seas are responsible f o r the absence of the mobile r e f l e c t i n g layer? ? (sperm whale) (photography) (sunlight) (300) ( l i v i n g creatures) (echo) (10) (1902) (shallow depths) -91 17. Sperm whales and bottle-nosed whales feed e x c l u s i v e l y on ? 18. A d i v e r , wearing a heavily armored d i v i n g s u i t , can s a f e l y descend to an extreme l i m i t of ? f e e t into the ocean. (squid) (500) 19. The giant squid A r c h i t h e u t i s l i v e s at depths of ? fee t or more. (1500) 20. Whales which do not eat fish- or squids can only e x i s t where there are dense masses of small shrimp or ? to supply t h e i r enormous food requirements. (copepods) 21. During J u l y and August, 1947 the U.S.S. Nereus made a continuous fathogram from Pearl Harbor to the ? and found the s c a t t e r i n g l a y e r over a l l deep water along t h i s course. (Arctic) i n a s t e e l apparatus known ? descended to 4500 22. Barton, as the feet o f f the C a l i f o r n i a coast i n 1949. (benthoscope) 23. In 1934, Beebe and Barton descended to a depth of 3028 fee t i n the open ocean o f f the shore of ? (Bermuda) 24. The f i r s t ship ever equipped f o r oceanographic exploration s t a r t e d from England on a globe-wide c r u i s e i n 1872. The ship's name was ? (Challenger) 25. S c i e n t i s t s at Massachusetts have found that d i r e c t sampling of the phantom bottom with nets may not be accurate because the nets can not be operated r a p i d l y enough to capture swift moving animals. (Woods Hole) APPENDIX D NAME: IT11 F i l l i n the blanks with the correc t word ( s i or number (s)• Place your answer i n the space provided i n the column at the r i g h t . ANSWER 1. The r e a l miracle of sea l i f e i n r e l a t i o n to great pressure i s not thetanimal that l i v e s i t s whole l i f e on the bottom, bearing a pressure of f i v e or s i x tons, but those that r e g u l a r l y move up and down through hundreds of thousands of fee t of v e r t i c a l change. Animals that do t h i s are the whale, j e l l y f i s h , squid, plankton and ? (shrimp) The phenomenon of luminescence i s displayed by perhaps ? of a l l the f i s h e s that l i v e i n dimly l i t or darkened waters, and by many of the lower forms as w e l l . (.h) The h i n t that there i s a f i e r c e un-compromising competition f o r food i n the deep waters l i e s i n the e l a s t i c and d i s t e n s i b l e bodies of many f i s h and t h e i r immense ? . (mouths) 4. Since no animal can make i t s own food, the creatures of the deeper waters l i v e a strange, almost p a r a s i t i c e x i s t -ence i n u t t e r dependence on the ? (upper waters) 5. A s i n g l e specimen of f i s h c a l l e d Latimeria was caught i n 1938. This species dates from the Cretaceous period. I t has a large head, strangely shaped scal e s , and i s of a b r i g h t ? c o l o r . (blue)  6. ? say that the 'seal f i s h ' belongs to a group that t y p i c a l l y inhabits very deep water, o f f the edge of the co n t i n -e n t a l s h e l f . (ichthyologists) 93 7. What structures tend to be missing i n the r e t i n a of deep sea fishes? ? Ccones) 8. One of the transparent species that l i v e i n the upper layers of the sea i s the arrowworm. Another name f o r t h i s creature i s the ? . Cglassworm) 9. S c i e n t i s t s who study the ' f r i l l s h a r k ' are c a l l e d ? (ichthyologists) 10. The baleen-.vwhale can dive to the depths of one-half a m i l e and r e -turn immediately to the surface without the f a t a l increase of nitrogen i n the blood experienced by human d i v e r s . At such depths the whale sustains a pressure of about ? on every inch of i t s body. 11. What co l o r i s the most prevalent i n sea animals which l i v e near the surface of the ocean? ? . (blue or green) 12. The white ? makes a high-pitched resonant whistle and squeal v a r i e d with the t i c k i n g and clucking sounds s l i g h t l y reminiscent of a s t r i n g orchestra tuning up, as well as mewing and occasional c h i r p s . (porpoise) (h pound) 13. Animals l i v i n g i n deep water become ? as they r i s e . (lighter) 14. Immense pressure i s one of the governing conditions of l i f e i n the deep sea; ? i s another. (darkness) 15. Many s c i e n t i s t s hope to f i n d o l d forms of l i f e i n the great depths of the sea. Although weird forms have been caught i n nets, there have been no signs of ' l i v i n g f o s s i l s ' reminiscent of such ancient forms of l i f e as the great marine r e p t i l e s of the ?  period. (Mesozoic) 16. At depths of about a thousand f e e t , f i s h are commonly of s i l v e r y c o l o r . However, the angle worm which l i v e s at t h i s depth i s a deep ?' (red) 94 17. The noise of shrimps have been heard a l l over a broad band that extends around the world between l a t i t u d e s 35°N and 35°S (for example, from Cape Hatteras to ? I• (Buenos Aires) 18. A specimen of Latimeria, a species of f i s h which, was supposed to have been e x t i n c t f o r a t l e a s t 60- m i l l o n years was caught a l i v e o f f the southeast coast of ? i n 1938. (Africa) 19. The crew of the ? recorded strange mewing sounds, shrieks, and ghostly moans i n deep water o f f the coast of Bermuda, the sources of which have not been traced. (Atlantis) 20. Sinking ships, the bodies of drowned men, and presumably the bodies of large sea animals not consumed above by hungry scavengers ? reach the bottom of the sea. (do) 21. J e l l y f i s h medusae, which above would be transparent, at a depth of 1000 fe e t are a deep ? (brown) 22. At depths below ? fe e t there are no traces of plant l i f e since no plant can f i n d enough sunlight f o r food-manufacturing below t h i s l e v e l . (600) 23. At the Scripps I n s t i t u t i o n at La J o l l a , the underwater voices of f i s h known as croakers has been detected i n the even-ing and have been described as a "steady uproar of harsh froggy croaks with a background of s o f t strumming." This chorus of croakers begins i n May and ends l a t e i n what month of the year? ? / (September) 24• ? m i l l i o n seals could not compete with commercial fishermen f o r the same species without the f a c t being known. (4) 25. The northern fur s e a l which- l i v e s i n the eastern P a c i f i c , spends i t s winters feeding o f f the coast of ? (North America) APPENDIX E NAME: Answer each of the following, the column at the r i g h t . MCT Place your answers i n the space provided i n ANSWER Write the simplest number symbol f o r : 1. ((3 + 4) x 5 - 2) x h 2. (3 x 7 + 5 x 4) x 6 Using the r e a l numbers, determine the sol u t i o n sets of the following: 3. 4(7 - 3x) = 5 3 4. 5(3*lx) - 4) <_13 5. 7x + 5/3 = 8 h 6. 3 " l r l < 4 (16^5) (246) (13/12) (-33/15 ...+33/15) (1/3) (-8/3 ...+8/3) 4v - 7 + y = 8 2 3 = 9 y - 3 -4 3 - y 2 x - 3 3 - x 10. -4z + 9 < -13 11. /3r + 5/ < 16 12. x + 3 > 5 4 Name the p r i n c i p l e s i l l u s t r a t e d i n the following: 13. 3 x 0 = 0 14. 7 + 0 = 7 (6 1/5) (40/9) ( z ^ 11/2) (-7 i> x £ 11/3) ( x > 17) (Zero f o r Mult.) (Zero f o r Add.) MCT-2 15. 2(x + 2) = 2x + 2*2 16. x-y = y x 17. (5-6)-7 = 5(6-7) 18. 3 - 1 / 3 = 1 19. 9-1 = 9 20. 7 + (-7) = 0 Simplify: 21. 2x 3•• x" 1 4 x -3 4 22. 5y -y -2 y x • X 24. (5x y J) 4 2 h 25. (25x y P ( D i s t r i b u t i v e ) (Comm. f o r Mult.)  (Ass, f o r Mult.)  (Mult. Inverse)  (Mult. Identity)  (Additive Inverse) (2/x 2)  (5y 3) (3x 2) (25x 4/y 6)  (- 5x 2y) 97 APPENDIX F EXPERIMENTAL DIRECTIONS WHAT YOU ARE TO DO TODAYI We are t r y i n g to f i n d out what makes some books d i f f i c u l t while others are easy to study. In ou t l i n e here i s what you w i l l be doing today. You w i l l be asked to study a s p e c i a l l y edited passage on explorations i n the deep waters o f the oceans. Later we w i l l give you a written t e s t to f i n d out how much you learned from your reading. A f t e r receiving the f i r s t t e s t you w i l l be asked to study more of the same passage and w i l l receive a second t e s t . THE SECOND TEST YOU RECEIVE WILL BE SIMILAR TO THE FIRST  TEST YOU RECEIVE. Some of you w i l l receive an arithmetic t e s t while others w i l l receive a t e s t on the prose s e l e c t i o n being read. The information which we obtain i n t h i s way may be of help i n the design of better textbooks and other t r a i n i n g materials. Here are some d e t a i l s . Y o u ' l l f i n d a mimeographed chapter from Rachel Carson's book about the world's oceans further on i n t h i s notebook. The chapter i s c a l l e d "The Sunless Sea." Study each page of the chapter c a r e f u l l y . PLEASE PAY CLOSE ATTENTION TO THE FOLLOWING: While you are reading the prose s e l e c t i o n you w i l l encounter questions. These questions are samples of the kind of things which can be asked about the study materials on the sea. Try to guess the answer f o r each question. Write i t on the question s l i p and then tear the s l i p from the notebook r i n g s . As soon as you have written down your guess and torn out the question s l i p , place the s l i p i n t o the box provided f o r t h i s purpose. When you have f i n i s h e d reading the prose s e l e c t i o n , replace i t i n i t s envelope and deposit the envelope i n the answer box. Repeat t h i s procedure f o r a l l four envelopes. IT IS EXTREMELY IMPORTANT THAT YOU WORK THROUGH THE MATERIALS IN ALPHABETICAL ORDER AND THAT YOU DO NOT LOOK BACK TO A PRECEDING PAGE ONCE YOU HAVE READ IT. You may now s t a r t Package A. 98 APPENDIX G NAME: QUESTIONNAIRE You have been engaged i n a study concerned with how questions and t e s t s can be used to help people learn more information from reading. Your a t t i -tudes towards the study might provide a d d i t i o n a l information which would be h e l p f u l . I t would be appreciated i f you would f i l l out the following questionnaire. Answer by p l a c i n g a check mark i n the space provided under YES, NO, or OTHER. Please give reasons f o r your answer i n the space provided beneath each question. YES NO OTHER 1. Did you read the di r e c t i o n s ? 2. Did you believe the d i r e c t i o n s ? 3. Did you follow the d i r e c t i o n s ? Did you expect questions about what you were reading? 5. Did you t r y to answer the questions? 99 YES NO OTHER 6. Were the questions f a i r ? 7. Were the questions i n the f i r s t p a rt more d i f f i c u l t than the questions i n the second part? 8. Did you expect to be tested on what you were reading? 9. Did you t r y to answer the t e s t questions? 10. Were the tes t s f a i r ? 11. Were the questions on the f i r s t t e s t more d i f f i c u l t than the questions on the second test? 100 Q--3 YES NO OTHER 12. . Was the reading i n t e r e s t i n g ? 13. Was the reading easy? 14. Did you read c a r e f u l l y ? 15. Was the f i r s t p a rt of the reading more i n t e r e s t i n g than the second part? 16. Was the f i r s t part e a s i e r to read than the second part? 17. Did the way you read the second part change from the way you read the f i r s t part? 101 Q—4 YES NO OTHER 18. Did you work through the envelopes i n alphabetical order? 19. Did you open any of the envelopes out of turn? 20. Did you look at other pages to help answer the questions you found while reading? 21. Did you look back at an envelope to help answer questions on a test? 22. Were you able to see any other student's answers? 23. Were you able to guess what the study was about? 102 Q--5 YES NO OTHER 24, Have you previously read about l i f e i n the deep portions of the ocean? 25. I f you have any other comments which you think might be h e l p f u l , please respond i n the following space. THANK YOU FOR YOUR CO-OPERATIONI 103 APPENDIX H TABLE A CHARACTERISTICS OF THE PRE-STUDY POPULATION Number of Persons by Sex and Grade Level Selected f o r the Baseline Data Group a f t e r A t t r i t i o n . Male Female Grade 10 6 3 Grade 11 _ J -L T o t a l 9 8 Number of Persons by Sex and Grade Level Selected f o r the R e l i a b i l i t y Group a f t e r A t t r i t i o n . Male Female Grade 10 4 6 Grade 11 __i .—L T o t a l 5 1 1 Number of Persons by Sex and Grade Level l o s t through A t t r i t i o n i n the Baseline Data Group and the R e l i a b i l i t y Group. Baseline Data Group R e l i a b i l i t y Group Male Female Male Female Grade 10 0 3 1 2 Grade 11 __1 0 0 0 _ To t a l 1 3 1 2 104 TABLE B BASELINE DATA FOR TESTS Mean Number Standard Correct Deviation Relevant Question Set One .18 .53 Relevant Question Set Two .29 .59 Incidental Test One .47 .72 Incidental Test Two .94 .83 Mathematical Computation Test 8.82 7.16 N = 17 TABLE C TABLE OF SPECIFICATIONS FOR EXPERIMENTAL QUESTIONS ITII ITI RQSI RQSII Number of Words i n Question M SB 26.60 14.47 22.40 10.29 27.90 6.19 31.40 5.30 Word Location of Stimulus i n Question Word Location of Response i n Question Page Location of Question Line Location of Answer on Page Line Location of Answer i n Prose Word Location of Answer on Page Word Location of Answer i n Prose Number of Words i n Answer i n Sentence Word Location of Stimulus i n Answer Word Location of Response i n Answer M SD M SD M SD M SD M SD M SD M SD M SD M SD M SD 16.76 14.86 21.56 16.30 5.08 2.80 14.64 9.62 136.08 86.55 1462.32 908.67 138.92 98.20 35.40 20.69 14.48 16.65 17.84 13.28 11.48 10.84 16.84 10.99 5.36 3.26 16.40 7.94 149.80 95.56 1552.12 989.96 165.64 83.06 35.08 14.57 12.28 10.09 16.36 12.43 16.50 12.89 24.90 9.62 4.70 2.91 14.80 9.45 131.40 93.12 1350.90 956.19 150.90 100.35 39.10 22.95 13.50 14.03 23.00 19.45 19.90 8.62 27.10 8.02 5.40 3.03 13.40 9.02 146.20 91.73 1551.30 953.31 137.20 100.64 30.60 8.15 12.70 8.67 17.30 13.95 M Mean SD = Standard Deviation ITII = Incidental Test Two ITI = Incidental Test One RQSI = Relevant Question Set One RQSII = Relevant Question Set Two 106 TABLE D COMPARISON OF STIMULI AND RESPONSE ITEMS IN QUESTIONS I T U ITI RQSI RQSII S R S R S R S R Number of Proper Nouns 2 6 10 5 2 2 2 0 Number of Common Nouns 22 19 15 20 8 8 8 10 Number of Verbs 1 0 0 0 0 0 0 0 Number of Numerical Items 1 4 0 8 0 3 0 2 Number of Mul t i p l e Word Items 5 3 7 4 0 0 1 0 ITII = Incidental Test Two ITI = Incidental Test One RQSI = Relevant Question Set One RQSII = Relevant Question Set Two S = St i m u l i R = Response TABLE E DISTANCE IN WORDS BETWEEN QUESTIONS IH RELEVANT QUESTION SET ONE AND INCIDENTAL TEST ONE Relevant Incidental Test one Questions uestion Set te Questions 1 2 3 4 5 6 7 8 9 10 11 12 13 14 IS 16 17 18 19 20 21 22 23 24 25 • ^ 1 249 2925 503 2532 1117 1701 2925 2818 2156 1247 1085 1204 896 65 2026 1401 1894 80 2864 2732 1371 163 143 747 2355 2 "173 2497 75 2104 689 1273 2497 2390 1728 846 657 776 468 "363 1598 973 1466 "348 2436 2304 943 "265 "285 319 1927 3 "317 2359 "63 1966 551 1135 2359 2252 1590 708 519 638 330 "501 1460 835 1328 "486 2298 2166 805 "403 "423 181 1789 4 -433 2243 "179 1850 435 1019 2243 2136 1474 592 403 522 214 "617 1344 719 1212 "602 2182 2050 689 "519 "539 65 1673 5 "736 1940 "482 1547 132 716 1940 1833 1171 289 100 219 "89 "920 1041 416 909 "905 1879 1747 386 "822 "842 "238 1370 6 "1058 1618 "804 1225 "190 394 1618 1511 849 "33 "222 "103 "411 "1242 719 94 587 "1227 1557 1425 64 "1144 "1164 "560 1048 7 "1237 1439 "983 1046 -369 215 1439 1332 670 "212 "401 ~282 "590 "1421 540 "85 408 "1406 1378 1246 "115 "1323 "1343 "739 869 8 "1882 795 "1627 402 "1013 "429 795 688 26 "856 "1045 "926 "1234 "2065 "104 "729 "236 "2050 734 602 "759 "1967 "1987 "1383 225 9 "1969 707 "1715 314 -1101 "517 707 600 "62 "944 "1133 "1014 "1322 "2153 "192 "817 "324 "2138 646 514 "847 "2055 "2075 "1471 137 10 "2838 "162 "2584 "555 "1970 "1386 "162 "269 "931 "1813 "2002 "1833 "2191 "3022 "1061 "1686 "1193 "3007 "223 "355 "1716 "2924 "2944 "2340 "732 Sign of numbers indicates directionality. TABLE F DISTANCE IN WORDS BETWEEN QUESTIONS IN RELEVANT QUESTION SET TWO AND INCIDENTAL TEST TWO Incidental Test Two Questions Relevant 3uestion S e t _ Vro Questions 8 10 11 13 14 15 16 17 13 19 20 21 22 23 24 25 3 4 5 6 7 8 9 10 289 1175 1639 1523 2665 -53 1351 1034 2781 66 964 22E1 486 790 2389 1083 2192 2601 1771 646 1099 1480 2005 -121 -178 118 1004 1468 1352 2494 "224 1180 863 2610 "105 793 2110 315 619 2218 912 2021 2430 1600 475 928 1309 1834 -292 "349 -450 436 900 784 1926 -792 612 295 2042 "673 225 1542 "253 51 1650 344 1453 1862 1032 -537 299 763 647 1789 "929 475 158 1905 "810 88 1405 "390 "86 1513 -327 59 523 407 1549 -1169 235 "82 1665 "1050 "152 1345 "630 "326 1273 93 360 741 1266 "860 "917 107 1316 1725 895 "230 223 604 1129 "997 -1054 •33 1076 1485 655 "470 -17 364 869 "1237 "1294 94 1236 "1482 "78 "395 1352 "1363 "465 352 "943 "639 960 "346 763 1172 342 "7 83 "330 51 576 "1550 "1607 -1140 -254 -210 -1643 -757 "293 "409 733 "1985 "581 "898 849 "1866 "968 349 "1446 "1142 457 "849 -1839 -953 "489 "605 537 "2181 "77 "1094 653 "2062 "1164 153 "1642 "1338 261 "1045 -2223 -1337 "873 "989 153 "2565 "1161 "1478 269 "2446 "1548 "231 "2026 "1722 "123 "1429 -2481 -1595 "1131 "1247 "105 "2823 "1419 "1736 11 "2704 "1806 "469 "2284 "1980 "381 "1687 "578 "169 "999 "2124 "1671 "1290 "765 "2891 "2948 260 669 "161 "1286 "833 "452 73 "2053 "2110 64 473 "357 "1452 "1029 "648 "123 "2249 "2306 "320 89 "741 "1866 "1413 "1032 "507 "2633 "2690 Sign of numbers indicates d i r e c t i o n a l i t y 109 TABLE G NUMBER OF STIMULI AND RESPONSE TERM REPLICATIONS IN QUESTIONS Number of S t i m u l i Term Replications i n Questions ITII ITI RQSI RQSII ITII 4 ITI 0 5 RQSI 1 1 1 RQSII 2 0 0 0 Number of Response Term Replications i n Questions ITII ITI RQSI RQSII ITII 3 ITI 0 0 RQSI 2 1 0 RQSII 1 1 0 0 Number of St i m u l i Term Replications i n Responses of Questions ITII ITI RQSI RQSII ITII 0 ITI 1 1 RQSI 0 0 0 RQSII 1 1 1 0 ITII Incidental Test Two ITI Incidental Test One RQSI = Relevant Question Set One RQSII = Relevant Question Set Two 110 TABLE H CHARACTERISTICS OF THE PARENT AND SAMPLE POPULATIONS Number of Persons by Sex, Race and Grade Level i n the Grade Ten and Eleven Student Population of the Mission Senior Secondary School Non-Indian Indian Male Female Male Female Grade Ten 101 98 Grade Eleven 128 124 8 11 4 4 Tota l 229 222 12 15 Number of Persons by Sex, Race and Grade Level i n Population the Sample Non-Indian Indian Male Female Male Female Grade Ten 25 26 Grade Eleven 32 25 3 5 3 1 To t a l 57 51 6 6 TABLE I NUMBER OF SUBJECTS BY GRADE SELECTED FOR EACH CONDITION OF THE EXPERIMENTAL DESIGN, ATTRITION, AND NUMBER OF SUBJECTS CAST OUT OF EACH CONDITION OF THE DESIGN BY GRADE Condition Number of A t t r i t i o n Number of Number of Subjects of Subjects Subjects Used Selected Subjects Cast Out i n Experiment Condition 11 Grade Ten 18 3 3 12 Grade Eleven 12 4 0 8 Total 30 7 3 20 Condition 12 Grade Ten 14 1 3 10 Grade Eleven 16 5 1 10 Total 30 6 4 20 Condition 21 Grade Ten 17 3 0 14 Grade Eleven 13 7 0 6 Total 30 10 0 20 Condition 22 Grade Ten 10 3 1 6 Grade Eleven 20 5 1 14 Tota l 30 8 2 20 TABLE J RELIABILITY OF RELEVANT QUESTION SET ONE, RELEVANT QUESTION SET TWO, THE MATHEMATICAL COMPUTATION TEST, INCIDENTAL TEST ONE, AND INCIDENTAL TEST TWO Condition RQSI RQSII MCT ITI ITII Condition 11 Indian 0.96 0.37 Non-Indian 0.77 0.30 Combined 0.81 0.81 0.64 0.82 0.82 -1.04 0.77 0.72 Condition 12 Indian 0.98 0.92 Non-Indian 0.78 0.78 Combined 0.71 0.80 0.80 0.55 0.56 .40 .66 .66 Condition 21 Indian -0.32 0.58 Non-Indian 0.56 0.72 Combined 0.56 0.82 0.78 0.90 0.89 -0. 0. 0. 16 85 82 Condition 22 Indian 0.79 Non-Indian 0.70 Combined 0.70 0.53 0.74 0.72 -0.15 0.72 0.68 0.30 0.78 0.80 N = 3 f o r Indian N = 17 f o r Non-Indian N = 20 for Combined * Interpret as 0.00 TABLE K MEAN SCORES ON RELEVANT QUESTION SET ONE, RELEVANT QUESTION SET TWO. THE MATHEMATICAL COMPUTATION TEST, INCIDENTAL TEST ONE, AND INCIDENTAL TEST TWO Condition RQSI RQSII MCT ITI ITII Condition 11 Indian 3.00 1.00 2.67 5.67 Non-Indian 2.82 2.58 4.29 4.47 Combined 2.85 2.30 4.00 4.65 Condition 12 Indian 3.33 2.33 3.33 Non-Indian 3.76 3.06 6.18 Combined 3.70 3.05 5.75 Condition 21 Indian 5.67 4.33 2.67 7.67 Non-Indian 5.82 5.59 3.41 5.29 Combined 5.80 5.40 3.30 5.65 Condition 22 Indian 8.00 Z.'bl 3.67 1.67 Non-Indian 6.76 4.06 6.00 Combined 6.95 4.00 5.40 N of Indian = 6 3 3 3 3 N of Non-Indian = 34 17 17 17 17 N of Combined = 40 20 20 20 20 TABLE L COMPARISON OF VARIANCE ON RELEVANT QUESTION SET ONE, RELEVANT QUESTION SET TWO, THE MATHEMATICAL COMPUTATION TEST, INCIDENTAL TEST ONE, AND INCIDENTAL TEST TWO Condition RQSI RQSII MCT ITI ITII Condition 11 Indian 12.67 .67 2.89 1.56 Non-Indian 5.54 2.01 12.09 12.48 Combined 6.63 6.63 11.59 11.03 Condition 12 Indian 12.67 11.56 4.22 2.90 Non-Indian 7.04 7.44 4.69 11.20 Combined 6.44 8.09 4.70 10.99 Condition 21 Indian 1.56 4.22 6.22 2.90 Non-Indian 4.96 6.60 20.11 22.83 Combined 4.61 7.89 18.39 15.83 Condition 22 Indian 4.67 2.00 1.56 1.55 Non-Indian 5.87 6.30 9.18 15.53 Combined 5.79 5.85 8.05 18.43 N = N = N = 3 f o r Indian 17 f o r Non-Indian 20 f o r Combined 115 Source TABLE M ANALYSIS" OP VARIANCE OF THE INDEPENDENT VARIABLES QUESTION PLACEMENT AND TEST EXPECTANCY ON INCIDENTAL TEST TWO FOR NON-INDIAN SUBJECTS Sum of Squares D.F. Mean Square Prob. Mean Q T QT Error 2046.01 1.78 24.72 4.25 1012.24 1 1 1 1 64 2046.01 1.78 24.72 4.25 15.82 129.36 0.11 1.56 0.27 0.0 0.73 0.21 0.61 N = 68 *P < .05 Q = Question Placement T = Test expectancy 116 TABLE N ANALYSIS OF VARIANCE OF THE INDEPENDENT VARIABLES QUESTION PLACEMENT AND TEST EXPECTANCY ON INCIDENTAL TEST TWO FOR INDIAN SUBJECTS Source Sum of Squares D.F. Mean Square F Prob. Mean 252.08 1 252.08 75.63 0.0 Q 0.08 1 0.08 0.03 0.85 T 52.08 1 52.08 15.63 0.00* QT 10.08 1 10.08 3.03 0.08 E r r o r 26.67 8 3.33 N = 12 *P <C .05 Q T = Question Placement = Test Expectancy 117 TABLE 0 DEBRIEFING QUESTIONNAIRE: FREQUENCY OF ANSWERS Y N O Y N 0 Y N O 1. Did you read the d i r e c t i o n s ? 11 3 0 0 13 4 0 16 4 0 20 12 0 0 2 16 1 0 16 1 2 19 21 3 0 0 15 1 0 18 1 0 19 22 3 0 0 13 1 2 16 1 2 19 2. Did you be l i e v e the d i r e c t i o n s ? 11 0 3 0 7 9 1 7 12 1 20 12 1 1 0 11 5 1 12 6 1 19 21 0 2 1 15 1 0 15 3 1 19 22 3 0 0 15 1 0 18 1 0 19 3. Did you follow the d i r e c t i o n s ? 11 1 0 2 11 5 1 12 5 3 20 12 0 1 1 9 7 2 9 8 3 20 21 3 0 0 10 3 3 13 3 3 19 22 1 2 0 11 3 2 12 5 2 19 4. Did you expect questions about what you were reading? 11 2 1 0 13 4 0 15 5 0 20 12 1 0 1 12 5 0 13 5 1 19 21 2 1 0 12 4 0 14 5 0 19 22 3 0 0 12 4 0 15 4 0 19 5. Did you t r y to answer the questions? 11 3 0 0 12 3 2 15 3 2 20 12 1 0 1 16 1 0 17 1 1 19 21 1 2 0 13 3 0 14 5 0 19 22 1 0 1 12 3 1 13 3 2 18 6. Were the questions f a i r ? 11 1 2 0 2 14 1 3 16 1 20 12 0 1 1 3 11 3 3 12 4 19 21 1 2 0 10 4 2 11 6 2 19 22 2 0 1 7 7 2 9 7 3 19 118 Condition Indian Non-Indian Combined NR Y N O Y N O Y N O 7. Were the questions i n the f i r s t p a rt more d i f f i c u l t than the questions i n the second part? 11 0 0 3 6 10 1 6 10 4 20 12 0 0 2 8 7 2 8 7 4 19 21 1 2 0 5 8 2 6 10 2 18 22 2 0 1 7 7 2 9 7 3 19 88. Did you expect to be tested on what you were reading? 11 2 0 0 12 5 0 14 5 0 19 12 2 0 0 12 5 0 14 5 0 19 21 2 0 0 12 3 0 14 3 0 17 22 2 1 0 13 2 1 15 3 1 19 9. Did you t r y to answer the t e s t questions? 11 2 1 0 13 2 2 15 3 2 20 12 2 0 0 17 0 0 19 0 0 19 21 2 1 0 13 3 0 15 4 0 19 22 2 0 0 15 0 1 17 0 1 18 10. Were the t e s t s f a i r ? 11 0 1 2 4 13 0 4 14 2 20 12 1 0 1 7 8 2 8 8 3 19 21 1 0 2 9 4 2 10 4 4 18 22 1 1 0 8 7 1 9 8 1 18 11. Were the questions on the f i r s t t e s t questions on the second test? 11 0 0 2 5 12 0 0 2 8 21 2 1 0 9 22 0 2 0 3 12. Was the reading i n t e r e s t i n g ? 11 2 1 0 7 12 2 0 0 6 21 0 1 2 7 22 0 2 1 6 more d i f f i c u l t than the 10 2 5 10 4 19 9 0 8 9 2 19 4 2 11 5 4 20 12 1 3 14 1 18 9 1 9 10 1 20 8 3 8 8 3 19 8 1 7 9 3 19 7 3 6 9 4 19 13. Was the reading easy? 11 2 1 0 9 6 2 11 7 2 20 12 2 0 0 7 8 2 9 8 2 19 21 1 1 1 5 8 3 6 9 4 19 22 1 2 0 7 8 0 8 10 0 18 119 0 7 5 5 10 5 5 20 0 7 10 0 7 12 0 19 1 3 10 3 4 11 4 19 0 3 13 0 4 15 0 19 reading more i n t e r e s t i n g than the second part? 1 3 12 2 3 14 3 20 2 3 12 1 3 12 3 18 0 3 11 2 4 13 2 19 1 4 9 3 5 9 4 18 1 1 2 13 2 3 14 3 20 1 1 5 12 0 5 13 1 19 1 1 3 11 3 4 12 4 20 1 1 5 9 2 6 10 3 19 Condition Indian Non-Indian Combined NR Y N O Y N O Y N 14. Did you read c a r e f u l l y ? 11 3 0 12 0 2 21 1 1 22 1 2 15. Was the f i r s t part of t l 11 0 2 12 0 0 21 1 2 22 1 0 16. Was the f i r s t p a rt e a s i e r to read than the second part? 11 1 " 12 0 21 1 22 1 17. 17. Did the way you read the second part change from the way you read the f i r s t part? 11 12 21 22 18. Did you work through the envelopes i n al p h a b e t i c a l order? 11 12 21 22 19. Did you open any of the envelopes out o f turn? 11 0 3 0 5 12 0 12 1 0 1 0 17 0 21 1 2 0 0 15 1 22 0 2 0 0 16 0 20. Did you look at other pages to help answer the questions you found while reading? 11 12 21 22 3 0 0 3 13 1 6 13 1 20 0 3 0 9 6 1 9 8 1 18 2 1 0 9 7 1 11 8 1 20 1 1 0 12 • 4 0 13 5 0 18 3 0 0 12 3 2 15 3 2 20 2 0 0 17 0 0 19 0 0 19 2 1 0 15 0 1 17 1 1 19 3 0 0 15 2 0 18 2 0 20 5 15 0 20 1 17 1 19 1 17 1 19 0 18 0 18 1 1 1 5 11 1 6 12 2 20 0 1 1 6 8 3 6 9 4 19 0 3 0 5 11 1 5 14 1 20 2 1 0 5 10 1 7 11 1 19 120 Condition Indian Non-Indian Combined NR Y N O Y N O Y N O 21. Did you look, back at an envelope to help answer questions on a test? 11 0 3 0 3 14 0 3 17 0 20 12 0 2 0 3 14 0 3 16 0 19 21 1 2 0 1 15 0 2 17 0 19 22 0 3 0 1 15 0 1 18 0 19 Were you able to see any other student's answers? 11 2 1 0 9 8 0 11 9 0 20 12 0 1 1 9 7 1 9 8 2 19 21 0 3 0 5 10 1 5 13 1 19 22 0 3 0 7 9 0 7 12 0 19 Were you able to guess what the study was about? 11 1 2 0 9 3 0 10 10 0 20 12 0 1 1 4 10 3 4 11 4 19 21 1 2 0 6 10 0 7 12 0 19 22 0 3 0 5 10 1 5 13 1 19 Have you previously read about l i f e i n the deep portions of the ocean? 11 2 1 0 6 9 2 8 10 2 20 12 0 2 0 1 12 0 1 14 0 15 21 0 3 0 6 9 1 6 12 1 19 22 0 3 0 4 10 2 4 13 2 19 N = 3 f o r Indian N = 17 f o r Non-Indian N = 20 f o r Combined Y = Yes N = No , 0 = Other NR = Number of responses 121 TABLE P DEBRIEFING QUESTIONNAIRE: VERBAL ANSWERS Verbal Answer Condition Condition Condition Condition 11 12 21 22 1. Did you read the d i r e c t i o n s ? to know what to do 7 0 3 5 were t o l d to 3 8 2 2 to help survey 1 1 1 0 most 0 0 1 1 h a l f 0 1 Q 1 confusing 0 1 1 1 too l a z y 0 0 0 1 2. Did you bel i e v e the d i r e c t i o n s ? no reason not to 4 4 3 4 were reasonable 0 2 1 1 p a r t i a l l y 1 3 0 0 confusing 5 2 1 0 didn't t r y 0 0 0 1 3. Did you follow the d i r e c t i o n s ? no reason not to 2 3 3 2 were easy to follow 1 0 0 0 wanted to do properly 1 1 1 1 as w e l l as I could 3 1 2 0 p a r t i a l l y 1 2 1 1 only a survey 1 0 1 0 forgot 0 3 0 0 too lazy 0 0 0 2 4. Did you expect questions about what you were reading? u s u a l l y get questions 2 4 3 4 d i r e c t i o n s t o l d us 1 1 1 0 some 1 0 0 0 rot type of question 3 1 0 0 no 1 1 1 1 forgot 0 2 0 0 122 Verbal Answer Condition Condition Condition Condition 11 12 21 22 5. Did you t r y to answer the questions? see how much I remember 0 3 0 0 best I could 2 1 0 2 were t o l d to 3 4 1 1 help with survey 1 0 0 0 not l i k e an exam 0 0 0 2 some 1 0 2 2 d i f f i c u l t but t r i e d 1 0 0 0 waste of time 1 0 4 0 sheet a l l wrecked 1 0 0 0 what questions 0 0 0 1 couldn't because answer on page a f t e r 3 2 0 0 6. Were the questions f a i r ? were r e l a t e d to story 0 1 0 3 not t r i c k y 0 0 0 1 too much to study 0 1 2 1 too d i f f i c u l t 3 2 1 1 asked before story 9 7 0 0 7. Were the questions i n the f i r s t part more d i f f i c u l t than the questions i n the second part? f i r s t p a rt harder 0 4 2 2 same 9 5 2 4 can't do Math 1 0 1 0 can't remember .0 1 0 1 didn't read c a r e f u l l y 0 1 0 0 didn't f i n i s h 0 1 1 0 8. Did you expect to be tested on what you were reading? u s u a l l y are 4 6 2 1 d i r e c t i o n s said so 1 2 0 1 couldn't t e l l 0 1 1 1 t h i s a i n ' t a t e s t 1 0 0 0 123 Verbal /Answer Condition Condition Condition Condition 11 12 21 22 9. Did you t r y to answer the t e s t questions? as much, as p o s s i b l e 3 to help with survey 1 d i r e c t i o n s s a i d to 2 h a l f of them 0 don't know 0 what questions 0 too uninteresting 1 10. Were the t e s t s f a i r ? 2 0 0 0 0 0 1 0 0 0 0 1 2 0 0 0 1 1 1 2 1 a l l on what we read 4 too d i f f e r e n t 1 didn't expect them 1 too hard 2 too much to study 1 couldn't r e c a l l 0 3 1 1 0 0 0 1 1 1 1 2 2 1 1 0 1 0 1 11. Were the questions on the f i r s t t e s t more d i f f i c u l t than the questions on the second test? f i r s t t e s t e a s i e r 1 same 3 forgot math 1 can't remember 0 1 1 0 5 2 2 0 1 0 2 0 0 12. Was the reading i n t e r e s t i n g ? l i k e d i t 1 learned something 2 i n t e r e s t i n g 1 so so 2 too long 1 too much d e t a i l 1 boring 1 too d i f f i c u l t 0 not i n t e r e s t i n g 1 2 0 3 1 0 0 3 1 1 1 1 0 1 0 0 0 0 0 0 1 0 0 0 1 2 2 2 1-2.4 Verbal Answer Condition Condition Condition Condition 11 12 21 22 13. Was the reading easy? no d i f f i c u l t y 1 1 1 1 some d i f f i c u l t y 1 2 0 1 too long 1 a 1 0 wo<nds too d i f f i c u l t 4 l 2 0 too d i f f i c u l t 2 2 1 0 14.. Did you read c a r e f u l l y ? t r i e d to 4 0 2 1 d i r e c t i o n s s a i d to 1 0 0 0 wanted to lea r n 1 1 0 0 looked f o r answers 2 0 0 0 skimmed 0 2 2 2 rushed through i t 1 2 2 2 not r e a l l y 0 1 0 0 too d i f f i c u l t 0 1 0 0 too lazy 0 0 0 1 15. Was the f i r s t p a rt of the reading more i n t e r e s t i n g than the second part? f i r s t p a rt harder 1 2 1 1 same 5 3 5 3 got bored 2 1 0 0 don't know 0 0 0 1 16. Was the f i r s t part easier to read than the second part? f i r s t p a rt harder 1 1 0 0 p a r t l y 1 0 0 0 same 2 4 5 5 math too d i f f i c u l t 1 0 0 0 17. Did the way you read the second part change from the way you read the f i r s t part? read slower 0 0 1 0 read more c a r e f u l l y 0 5 1 2 t r i e d to remember more 0 1 0 2 same 5 3 0 0 never noticed 0 0 0 1 read f a s t e r 2 0 1 1 read l e s s c a r e f u l l y 1 0 0 0 125 Verbal Answer Condition Condition Condition Condition 11 12 21 22 18. Did you work through the envelopes i n - a l p h a b e t i c a l order? the l o g i c a l way 3 2 1 0 d i r e c t i o n s s a i d to 4 3 0 0 to help survey 0 0 0 1 skipped one 0 0 1 0 19. Did you open any of the envelopes out of turn? didn't see any poi n t 4 3 0 1 d i r e c t i o n s s a i d not to 0 3 0 0 wasn't p o s s i b l e 1 Q 0 0 would have been caught 0 0 0 1 j u s t the f i r s t one 1 1 0 0 was curious 0 1 1 0 20. Did you look at other pages to help answer the questions you found while reading? didn't see any point 0 2 1 0 d i r e c t i o n s s a i d not to 4 1 1 0 a f t e r I answered them 0 1 0 0 at f i r s t 1 1 2 0 once 1 3 0 0 sometimes 3 1 1 2 couldn't 0 1 1 0 21. Did you look back at an envelope to help answer questions on a test? didn't see the poi n t 4 5 2 1 d i r e c t i o n s s a i d not to 1 0 0 1 to help the survey 0 0 0 0 couldn't 3 0 1 1 22. Were you able to see any other student's answers? no need to 3 1 2 2 wasn't i n t e r e s t e d 1 2 0 0 am near sighted 0 1 0 0 box was i n way 1 0 1 0 working at d i f f e r e n t speed 0 1 1 0 didn't use them 0 0 1 1 couple of times 0 1 0 0 why not 0 1 0 0 she was seated next to me 1 1 1 0 126 Verbal Answer Condition Condition Condition Condition 11 12 21 22 23. Were you able to guess what the study was about? att e n t i o n 1 0 0 0 comprehension 1 1 0 0 d i r e c t i o n s t o l d us 1 1 0 0 not r e a l l y 4 2 2 2 disc o v e r i e s i n the sea 0 0 0 1 f i s h 0 0 1 0 24. Have you previously read about l i f e i n the deep portions o f the ocean? i n elementary school 1 0 0 1 i n science 0 1 1 0 National Geographic 0 0 0 1 20,000 Leagues under the sea 1 0 0 0 a l i t t l e 1 1 1 2 no time 0 0 0 1 not inte r e s t e d 4 0 2 1 too spaced to rememberl 0 0 0 1 never 1 0 1 1 25. I f you have any other comments which you think might be h e l p f u l , please respond i n the following space. more i n t e r e s t i n g t o p i c 0 2 6 1 too d i f f i c u l t 0 0 0 1 too many questions 0 0 1 0 didn't l i k e pre-questions 3 3 0 0 didn't l i k e math t e s t 2 0 1 0 questions i r r e l e v a n t 0 0 0 1 exams too long 0 1 0 0 questions interrupted 1 1 0 0 reading 1 1 0 0 .127 APPENDIX I Rothkopf (1966) Rothkopf (1966), using one hundred f i f t y - n i n e paid volunteers from c Fairleigh-Dickenson U n i v e r s i t y i n Madison, N.H. and a f i f t y - t w o hundred ( s i c ) * word s e l e c t i o n from Rachael Carson's book, The Sea Around Us, i n s t i g a t e d the f i r s t i n v e s t i g a t i o n i n t o the e f f e c t s of adjunct questions under conditions such that the d i r e c t i n s t r u c t i v e consequences of questions were eliminated. The prose s e l e c t i o n was divided i n t o twenty pages of approximately two hundred and s i x t y words each and presented i n seven sections. The f i r s t section contained two pages and the remaining s i x sections contained three pages each. Whether the number of words i n each zone va r i e d was not reported by Rothkopf (1966). Of the t h i r t y - n i n e questions used i n the experiment, s i x were constructed on each o f the f i r s t four sections and f i v e on each of the three remaining sections. A l l experimental materials, i n c l u d i n g d i r e c t i o n s and t e s t s , were pre-sented i n the form of a l o o s e - l e a f notebook. Each subject worked through the material at h i s own rate and recorded the s t a r t i n g and f i n i s h i n g times from a clock located on h i s t a b l e . When a subject f i n i s h e d reading a page or question, he tore i t from the notebook and dropped i t i n t o a b a l l o t box con-t a i n e r . This was to prevent a l t e r a t i o n of answers and re-reading of the prose s e l e c t i o n pages. The treatments included pre-adjunct questions with answers, pre-adjunct questions without answers, pre-questions with answers, post-adjunct questions with answers, and post-adjunct questions without answers. Three control 9 The prose s e l e c t i o n that Rothkopf (1966) used, "The Sunless Sea", was a c t u a l l y 6288 words long. This represents 314 words on each page rather than the 260 claimed by the o r i g i n a l experimenter. 128 groups were used. The f i r s t control group, the "control group", read the same prose selec-tion as the experimental groups, without questions, and took the same tests. This group was used to provide a baseline to compare reading without questions to reading with questions. The second control group, a "directional reference" group, was used to determine the relative magnitude of question effects com-pared to direction effects. This group was directed to read carefully and slowly. Otherwise, i t was similar to the f i r s t control group. The third con-t r o l group, a "transfer evaluation" group, was used to assure that there was no transfer of specific training between the relevant and incidental tests. This group took the incidental test, did not read the material, learned the relevant test item answers to a criterion of one hundred percent regardless of order, and then were tested on a comparative form of the incidental test. The dependent measures consisted of a general test of twenty-five incidental questions and a relevant test of the fourteen items used in the experimental treatments. Reading time was also recorded by the subjects. The analysis of variance showed between treatment effects to be significant (F = 2.44, d.f. = 6,139, p = .05). The post-adjunct question group which did not receive answers and the directional reference group pro-duced significantly higher overall test scores than the f i r s t control group (t = 2.21, d.f. = 139, p = .01 and t = 2.09, d.f. = 139, p = .05, respectively). However, as Hopkins & Chadbourn (1967) point out, the use of multiple t tests exploits chance differences and introduces inextricable patterns of dependency and non-orthogonality. This leads to an inaccurate probability model upon which to base Ho. According to Ladas (1973) , Hopkins & Chadbourn (1967) re-analyzed Rothkopfs (1966) data using the Newman-Keuls technique of multiple compari-129 sons. This method seemed to be more s e n s i t i v e given Rothkopf's (1966) design. Hopkins & Chadbourn (1967) found no s i g n i f i c a n t d i f f e r e n c e s even allowing one-t a i l e d t e s t s . The use of one-tailed t e s t s by Rothkopf (1966) was also questionable. A one-tailed t e s t i n f e r s a d i r e c t i o n a l hypothesis. Rothkopf's (1966) study did not state any s p e c i f i c hypothesis although he d i d r e f e r to Rothkopf (1963, 1965) i n which the f a c i l i t a t i n g e f f e c t s of t e s t - l i k e events were hypothesized. However, i t i s not c l e a r how these hypotheses apply to Rothkopf's (1966) study. Rothkopf (1966) should have used two-tailed t e s t s . The r e s u l t s of t h i s study may have also been confounded by the p o s s i -b i l i t y that relevant and i n c i d e n t a l questions d i f f e r e d s i g n i f i c a n t l y i n terms of d i f f i c u l t y and d i s c r i m i n a b i l i t y . There was also the p o s s i b i l i t y that the unequal d i s t r i b u t i o n of questions introduced a l o c a t i o n a l bias i n terms of emphasis on c e r t a i n portions of the material. Unequal c e l l s i z e s and the small number of subjects (six) i n the pre-question group may also have influenced the f i n d i n g s . Likewise, the use of paid volunteers and the lack of random assignment to groups severely l i m i t s the g e n e r a l i z a b i l i t y of r e s u l t s . Informing the subjects that they would be tested and the lack o f information provided on the r e l i a b i l i t y and v a l i d i t y of the dependent measure also detracted from the i n t e r p r e t a t i o n of the r e s u l t s of t h i s study. However, the f a c t that Hopkins & Chadbourn (1967) found no s i g n i f i c a n t d i f f e r e n c e s does not i n v a l i d a t e the s i g n i f i c a n t F r a t i o s which support the conclusion that question placement i n the pre-adjunct or post-adjunct con-d i t i o n can make a d i f f e r e n c e to i n c i d e n t a l l earning. Rather, t h i s f i n d i n g questions whether the d i f f e r e n c e s noted by Rothkopf (1966) were a t t r i b u t a b l e to the pre-adjunct questions or the post-adjunct questions. 130 Rothkopf & Bisbicos (1967) Rothkopf & Bisbicos (1967) used two hundred f i f t y - t w o paid volunteers from Grades Ten, Eleven and Twelve at Governor Livingston Regional High School of Berkeley Heights, New Jersey and a nine-thousand word passage from Rachael Carson's, The Sea Around Us, investigated the use of r e s t r i c t e d categories of questions on i n c i d e n t a l learning. A two by two by three f a c t o r i a l design was used with the factors being: a) question type (relevant or i n c i d e n t a l ) , b) question p o s i t i o n (pre-adjunct or post-adjunct), and c) question category (a quantitative term or name, a common Engli s h or t e c h n i c a l word, and a mixture of the two). Two c o n t r o l groups were used. One consisted of t h i r t y - s i x subjects who read the prose s e l e c t i o n and took the t e s t s without encountering any relevant questions. The second control group was a "transfer evaluation" group as previously described (Rothkopf, 1966), and consisted of an a d d i t i o n a l twenty-three subjects from the same population. Subjects were assigned i n a non-systematic manner to each of the seven experimental treatments. The dependent measure was a retention t e s t administered at the end of the prose s e l e c t i o n . This t e s t consisted of twenty-four relevant and twenty-four i n c i d e n t a l questions. Four forms of t h i s t e s t were used, varying only i n order of presentation. The completion time of each page of the prose s e l e c -t i o n was recorded by the subjects. The passage of prose consisted of t h i r t y - s i x pages divided i n t o twelve three page zones. Eight completion type questions were composed on each zone. These questions were di v i d e d according to t h e i r response requirements i n t o four categories with one i n c i d e n t a l and one relevant question i n each category. The categories were described by Rothkopf & Bisbicos (1967) to include: a) common phrases, b) t e c h n i c a l E n g l i s h words, c) measures, and d) names. The 131 procedure was s i m i l a r to that used by Rothkopf (1966). The three way analysis of variance indicated s i g n i f i c a n t treatment d i f -ferences (F = 3.09, d.f. = 6,245, p = .01). However, Rothkopf & Bisbicos (1967) then made an i d e n t i c a l analysis as d i d Rothkopf (1966) using multiple t t e s t s . While the r e s u l t s of t h i s analysis d i d support Rothkopf (1966) , the use of multiple t t e s t s d i d not provide any confidence i n which factors were responsible f o r the d i f f e r e n c e s i n i n c i d e n t a l learning. Bias may have been introduced into t h i s study i n terms of differences i n d i f f i c u l t y l e v e l with the various question categories. V a l i d i t y and r e l i a b i l i t y information was not presented f o r the questions. The use of paid volunteers l i m i t e d the g e n e r a l i z a b i l i t y of r e s u l t s . Systematic differences may have also been introduced between the experimental and c o n t r o l groups as randomizing procedures were not implemented. Like Rothkopf (1966), Rothkopf & Bisbicos (1967) also informed t h e i r subjects that a t e s t would follow the prose s e l e c t i o n . Frase (1967) Frase (1967) used seventy-two introductory psychology students at the U n i v e r s i t y of Massachusetts and a two thousand word biographical passage on William James i n M i l l e r ' s (1962) Psychology: The Science of Mental Health. He investigated the e f f e c t s of length of passage, knowledge of r e s u l t s and p o s i t i o n of questions on i n c i d e n t a l l e a r n i n g . The prose s e l e c t i o n and relevant questions were presented on twenty, four by eleven inch sheets of paper with each sheet containing a multiple choice question having f i v e a l t e r n a t i v e answers and r e q u i r i n g r e c a l l of s p e c i f i c f a c t u a l information. Each sheet was approximately one hundred words i n length. The subjects were i n s t r u c t e d to read a page, answer the question and turn to the next page. The subjects were informed as to the nature of the 132 experiment and knew that a t e s t would follow the i n s t r u c t i o n a l materials. A two by three by two f a c t o r i a l design was used with the factors being: a) question placement (pre-adjunct or post-adjunct), b) question c o n t i g u i t y (ten, twenty, or f o r t y l i n e s ) , and c) knowledge of r e s u l t s (present or absent). The dependent measures included an i n c i d e n t a l t e s t c o n s i s t i n g of twenty ques-t i o n s , p e r t a i n i n g to the f i r s t h a l f of each paragraph, and a relevant t e s t c o n s i s t i n g of twenty questions p e r t a i n i n g to the second h a l f of each paragraph. The study was conducted on two consecutive days. The subjects were randomly assigned to one of the twelve groups f o r one of the two days. Half of the subjects i n each condition were run on t h e . f i r s t day and the remainder were run on the second day. A no question c o n t r o l group was used. The analysis of variance revealed a s i g n i f i c a n t main e f f e c t f or a l l three f a c t o r s upon the relevant t e s t . Post-adjunct questions were found superior to pre-adjunct questions (F = 16.10, d.f. •• 1/60, p = .001). The twenty l i n e passage was found superior to the ten and f o r t y l i n e passages (F = 3.94, d.f. = 2/60, p = .02). Providing subjects with knowledge of r e s u l t s was found superior to not providing subjects with knowledge of r e s u l t s (F = 52.20, d.f. = 1/60, p = .001). A s i g n i f i c a n t i n t e r a c t i o n was found between the p o s i t i o n of questions and whether knowledge of r e s u l t s was a v a i l a b l e or not (F = 4.22, d.f. = 1/60, p = .05) such that i f knowledge of r e s u l t s was not provided, then questions were more e f f e c t i v e following the passage. A s i g n i f i c a n t main e f f e c t f o r question placement was found on the i n c i d e n t a l t e s t such that post-adjunct relevant questions produced superior scores to pre-adjunct relevant questions (F = 22.00, d.f. = 1/60, p = .001). This f i n d i n g supported Frase (1967), Rothkopf (1966) and Rothkopf & Bisbicos (1967). 133 While t h i s study included controls f o r possible differences i n d i f f i -c u l t y between relevant and i n c i d e n t a l questions with a p r e - t e s t , i t introduced a v a r i a b l e i n terms of question placement within paragraphs which could pos-s i b l y have confounded the r e s u l t s . Relevant questions were based on the f i r s t h a l f of each paragraph while i n c i d e n t a l questions were based on the second h a l f of each paragraph. Presentation of the materials over a two day period may have l e d to s e l e c t i v e d i f f e r e n c e s between groups due to i n t r a subject communication. More-over, the population used i n t h i s study was highly s e l e c t and probably highly motivated. Subjects were informed that a t e s t would be found at the end of the reading m a t e r i a l . Likewise, the absence of r e l i a b i l i t y and v a l i d i t y data on the dependent measure confounds the i n t e r p r e t a t i o n of r e s u l t s . The s i g n i f i c a n t d i f f e r e n c e on relevant t e s t items between the pre-adjunct and post-adjunct question groups reported by Frase (1967) contradicts Rothkopf's (1966) e a r l i e r f i n d i n g . Rothkopf (1966) reported that pre-adjunct relevant questions produced greater relevant learning than post-adjunct r e l e -vant questions. Frase (1967) reported that post-adjunct relevant questions produced greater relevant learning than pre-adjunct relevant questions. This co n t r a d i c t i o n of findings may p o s s i b l y be a function of the d i f f i c u l t y l e v e l of the questions. I f Frase's (1967) questions were more d i f f i c u l t than Rothkopf's (1966) questions, they would be expected to engender more proactive interference i n the pre-adjunct question condition and subsequently reduce relevant learning. This c o n t r a d i c t i o n i n findings may also p o s s i b l y be a function of the motivational l e v e l of the subjects. Rothkopf's (1966) sub-j e c t s were paid volunteers, while Frase's (1967) subjects were unpaid c o n s c r i p t s . Under conditions of high motivation, question placement e f f e c t s may be overridden. (Frase, 1971; Frase, P a t r i c k , & Schumer, 1970). 134 While studies of i n c i d e n t a l learning (Frase, 1967; Rothkopf, 1966; Rothkopf & Bi s b i c o s , 1967) point out that greater i n c i d e n t a l learning occurred from prose when post-adjunct questions were used than when pre-adjunct ques-tions were used, the studies by Rothkopf (1966) and Rothkopf & Bisbicos (1967) are l i m i t e d i n terms of the use of paid volunteers and non-random assignment to experimental and control groups. While Frase (1967) d i d not use paid volunteers, and d i d use random assignment to experimental groups, h i s popu-l a t i o n , l i k e that of Rothkopf (1966) and Rothkopf & Bisbicos (1967) was s e l e c t i v e . Each in v e s t i g a t o r introduced the p o s s i b i l i t y of t e s t expectancy d i f f e r -e n t i a l l y i n t e r a c t i n g with question placement by informing the subjects that a t e s t was to follow the reading assignment. Likewise, each i n v e s t i g a t o r f a i l e d to provide r e l i a b i l i t y and v a l i d i t y data on the dependent measures. As such, the reported s i g n i f i c a n t findings may have been a function of poor r e l i a b i l i t y and large e r r o r of measurement. However, the trend^of these i n v e s t i g a t i o n s does support the hypothesis that greater i n c i d e n t a l learning occurs when ques-tions accompanying highly f a c t u a l prose are presented i n the post-adjunct format, as compared to the pre-adjunct question format. Frase (1968a) Frase (1968a) investigated the e f f e c t s of continguity i n terms of pre-adjunct and post-adjunct questions, using s i x t y - f o u r introductory psychology students from the Uni v e r s i t y of Massachusetts and the same materials and procedures as he previously used (Frase, 1967). The design was two by two by two by four f a c t o r i a l with repeated measures on the l a s t two f a c t o r s . The factors were: a) co n t i g y i t y y (ten or twenty sentences), b) question placement (pre-adjunct or post-adjunct), c) question type (relevant or i n c i d e n t a l ) , and d) blocks of paragraphs (four 135 blocks of f i v e paragraphs each). The dependent measure was a f o r t y item reten-t i o n t e s t covering a l l twenty paragraphs and c o n s i s t i n g of twenty relevant and twenty i n c i d e n t a l questions grouped according to blocks. Two questions were constructed on each paragraph. One question pertained to the f i r s t h a l f of the paragraph and one question pertained to the second h a l f . Half the subjects received the questions on the f i r s t h a l f of the para-graph as relevant, while the other h a l f of the subjects received questions on the second h a l f of the paragraph as relevant. The remaining questions f o r each group were i n c i d e n t a l and were only seen on the post t e s t . This procedure enabled Frase (1968a) to co n t r o l f o r the question placement within paragraphs e f f e c t i n Frase (1967). Each paragraph was approximately one hundred words i n length and repro-duced on four by eleven inch sheets of paper along with the questions. Subjects e i t h e r saw one, or two questions before, or a f t e r a paragraph, depending on the c o n t i g u i t y and question placement condition. The subjects were randomly assigned the experimental materials. They were informed that a retention t e s t would be administered when they had com-pleted the reading. Knowledge of r e s u l t s was not given. The analysis of variance ind i c a t e d a s i g n i f i c a n t main e f f e c t f o r ques-t i o n p o s i t i o n (F = 10.93, d.f. = 1, p = .05) such that the group which viewed the post-adjunct questions performed at a higher l e v e l than the group which received the pre-adjunct questions on both the relevant and i n c i d e n t a l t e s t s . For both the pre-adjunct question group and the post-adjunct question group, the relevant material was retained better than the i n c i d e n t a l material (F = 25.03, d.f. = 1, p = .001). The i n t e r a c t i o n of co n t i g u i t y with question p o s i t i o n was s i g n i f i c a n t (F = 16.52, d.f. = 1, p = .05) and opposite f o r the two question groups. The nearer post-adjunct questions were to the material 136 to which they applied^ the greater the i n c i d e n t a l learning. The f a r t h e r pre-adjunct questions were from the material to which they applied, the greater the i n c i d e n t a l learning. A s i g n i f i c a n t e f f e c t was found f o r blocks of paragraphs (F = 3.37, d.f. = 3, p = .025). A trend analysis of the blocks revealed a s i g n i f i c a n t cubic component (F = 6.40, d.f. = 1180, p = .025). Since the pre-adjunct and post-adjunct questions were af f e c t e d i n the same manner across blocks, the cubic trend might r e f l e c t primacy or recency e f f e c t s or perhaps di f f e r e n c e s i n item d i f f i c u l t y . I t was i n t e r e s t i n g to note that t h i s study d i d f i n d a s i g n i f i c a n t i n t e r -a c t i o n between co n t i g u i t y and question placement. A defect i n the experimental design should have decreased the p r o b a b i l i t y of such a f i n d i n g . Within each block there were f i v e questions, one on each paragraph, that were relevant. In the ten sentence c o n t i g u i t y condition, one question preceded or followed each paragraph within the block. However, i n the twenty sentence c o n t i g u i t y con-d i t i o n , two questions preceded or followed each of the f i r s t four paragraphs. This l e f t only one question to follow the l a s t paragraph i n each block. As such, the l o c a t i o n of the l a s t question i n each block i n the twenty sentence c o n t i g u i t y condition was only ten sentences from the relevant portion of the prose s e l e c t i o n . The l o c a t i o n of t h i s l a s t question decreased the d i s t i n c t i o n between the ten and twenty sentence c o n t i g u i t y condition and consequently should have decreased the p r o b a b i l i t y of obtaining s i g n i f i c a n t d i f f e r e n c e s . The confounding e f f e c t s associated with Frase (1968a) were s i m i l a r to those noted f o r Frase (1967). The one notable d i f f e r e n c e was that Frase (1968a) co n t r o l l e d f o r a p o s s i b l e question placement e f f e c t within paragraphs by using the relevant questions as i n c i d e n t a l f o r h a l f h i s subjects. The s i g n i f i c a n t main e f f e c t s f o r question p o s i t i o n support the findings of Frase (1967) f o r both relevant and i n c i d e n t a l l earning. The f i n d i n g that the relevant material was retained b e t t e r than the i n c i d e n t a l material also supports Frase (1967). The i n t e r a c t i o n of question pacing with question pos-i t i o n suggests that the optimal distance relevant questions should be from the material to which they p e r t a i n was ten l i n e s rather than the twenty l i n e s reported by Frase (1967). Frase (1968b) Frase (1968b) again used one hundred twenty-eight introductory psycho-logy students at the Univ e r s i t y of Massachusetts and the same prose passage used previously (Frase 1967, 1968a) to inve s t i g a t e the e f f e c t of question placement, pacing, and mode upon the retention of prose. A two by four by two by two f a c t o r i a l design was used i n t h i s study with repeated measures on the l a s t f a c t o r . The factors were: a) question l o c a t i o n (pre-adjunct or post-adjunct), b) question c o n t i g u i t y (ten, twenty, f o r t y , or f i f t y sentences), c) content l o c a t i o n ( f i r s t h a l f or second h a l f of each ten sentence paragraph), d) question mode (multiple choice or constructed response), and e) question form (relevant or i n c i d e n t a l ) . The dependent mea-sure was a f o r t y item post t e s t administered immediately upon completion of the task. The passage was divided i n t o twenty paragraphs of ten l i n e s each. Two multiple choice questions, one on the f i r s t h a l f of each paragraph and one on the second h a l f of each paragraph, were constructed. The questions were pres-ented using a cloze technique with f i v e a l ternate answers f o r multiple choice items. The same questions were reworded f o r constructed responses. The f o r -mat and presentation of the materials was s i m i l a r to that used previously (Frase 1967, 1968a). The analysis of variance indicated a s i g n i f i c a n t main e f f e c t f o r question l o c a t i o n (F = 10.43, d.f. 1/96, p = .001) such that questions f a c i l i -138 tated retention more when they were placed a f t e r the prose passage. Retention of the relevant information was s i g n i f i c a n t l y higher than retention of the i n c i d e n t a l information (F = 28.6, d.f. = 1/96, p = .001). /An i n t e r a c t i o n was found between question p o s i t i o n and c o n t i g u i t y of questions (F = 3.2, d.f. = 3/96, p = .05) such that the advantage of the post-adjunct question group became la r g e r with more frequent questioning. An i n t e r a c t i o n of pacing with retention items (F = 3.24, d.f. = 3/96, p = .05) was found such that the reten-t i o n of the i n c i d e n t a l material was depressed with frequent questioning. An i n t e r a c t i o n was also found between the retention item and the l o c a t i o n of the material (F = 5.6, d.f. = 1/96, p = .025) such that regardless of pacing or l o c a t i o n of questions, higher i n c i d e n t a l learning was achieved i f the i n c i d -e n t a l material followed the relevant material. Frase (1968b) d i d c o n t r o l f or the confounding e f f e c t of pacing evidenced i n Frase (1968a). However, t h i s study maintained the other confounding e f f e c t s noted previously (Frase, 1968a). Frase (1968b) supports h i s e a r l i e r findings (Frase, 1967, 1968a) i n that post-adjunct questions increased relevant learning. This study also sup-ported Frase (1968a) i n that the ten sentence c o n t i g u i t y condition produced optimal i n c i d e n t a l learning. This f i n d i n g was again contrary to that reported by Frase (1967). Frase (1968b) also supported the findings of previous i n v e s t i g a t o r s (Frase, 1967, 1968a; Rothkopf, 1966; Rothkopf & B i s b i c o s , 1967) i n that post-adjunct questions r e s u l t e d i n greater i n c i d e n t a l learning than pre-adjunct questions. Frase, P a t r i c k , S Schumer (1970) Frase, P a t r i c k , & Schumer (1970) studied the e f f e c t of question place-ment and frequency under d i f f e r e n t l e v e l s of incentive. They used two hundred seventy educational psychology students from the U n i v e r s i t y of Massachusetts 139 and the same passage as previously used by Frase (1967, 1968a, 1968b). A two by three by two f a c t o r i a l design was used. The f a c t o r s were: a) question placement (pre-adjunct or post-adjunct), b) incentive l e v e l (zero, three, or ten cents f o r each cor r e c t answer), c) co n t i g u i t y (ten or f i f t y l i n e s ) , and d) question form (relevant or i n c i d e n t a l ) . A no-question c o n t r o l group was used. The dependent measures were a f o r t y item post t e s t and the completion time of the post t e s t recorded by the subject. The materials used and the procedures implemented i n the study were s i m i l a r to those used by Frase (1968a, 1968b). The subjects were randomly assigned to the experimental packages which contained the text, t e s t , and IBM scoring sheets. The subjects were informed that they would receive zero, three, or ten cents f o r each cor r e c t response, according to the number p r i n t e d on t h e i r package. The incentive v a r i a b l e was introduced i n an attempt to determine to what degree the r e s u l t s obtained by previous i n v e s t i g a t o r s (Frase, 1967, 1968a, 1968b; Rothkopf, 1966; Rothkopf & Bisbicos, 1967) were generalizable and comparable. Duncan's multiple range t e s t i n d i c a t e d that post-adjunct questions were superior to pre-adjunct questions (F = 4.61, d.f. = 2/261, p = .01). Time was found to be influenced by question condition (F = 5.28, d.f. 2/261, p = .01). The post-adjunct question group took s i g n i f i c a n t l y longer than the con t r o l group. Incentive e f f e c t s were also apparent i n time scores (F = 43.37, d.f. = 2/261, p = .001). The ten cent condition took longer than the three cent condition, which i n turn, took longer than the zero cent condition. Relevant lea r n i n g was s i g n i f i c a n t l y greater than i n c i d e n t a l learning (F = 15.3, d.f. = 1/168, p = .001) and an i n t e r a c t i o n was found between question frequency and type of learning (F - 15.3, d.f. = 1/168, p = .001). Subjects scored s i g n i f i -c a n tly higher on relevant items and lower on i n c i d e n t a l items when questioning 140 was frequent as opposed to infrequent. A four-way i n t e r a c t i o n among question p o s i t i o n , frequency, incentive, and item type was- found (F=4.2, d.f. = 2/168, p = .0251. This i n t e r a c t i o n revealed that the expected ordering of incentive factors only occurred when questions were infrequent. Frase, P a t r i c k , & Schumer's (1970). findings support those of previous i n v e s t i g a t i o n s (Prase, 1967, 1968a, 1968b,• Rothkopf, 1966; Rothkopf & Bis b i c o s , 1967) i n that conditions having post-adjunct questions y i e l d e d superior i n c i d e n t a l learning to pre-adjunct questions and that subjects scored higher on relevant items than i n c i d e n t a l items. However, t h i s study also i n d i c a t e d that the advantage of relevant items over i n c i d e n t a l Items may have been a function of the frequency of relevant questions. Frase, P a t r i c k , & Schumer (1970) also pointed out that time was influenced by question condition. The post-adjunct question group took s i g -n i f i c a n t l y longer than the c o n t r o l group. Likewise, incentive e f f e c t s were also apparent i n time scores. Thus, post-adjunct questions appear to have caused the subjects to engage i n a c t i v i t i e s with s i m i l a r r e s u l t s to those found f o r incentive. However, t h i s study maintained the methodological d i f f i c u l t i e s a t t r i b u t e d to Frase (1968b) which placed some constraint on the i n t e r p r e t a t i o n of r e s u l t s . Natkin & Stahler (1969) Natkin & Stahler (1969) r e c r u i t e d twenty-eight students i n an introduc-tory educational psychology course at Buchnell U n i v e r s i t y , Lewisburg, Pennsylvania. Using two 2500 word passages from two standard biology t e x t s , one dealing with plants and the other dealing with c i l i a t e s , Natkin & Stahler (1969) investigated the e f f e c t s of post^adjunct questions on short and long term r e c a l l . Two versions of each passage were prepared, one without questions and 141 the other with post-adjunct questions. The passages were presented on eight and one h a l f by f i v e and one h a l f inch pages. Each page contained approxi-mately one hundred words. The selec t i o n s were presented i n separate booklets. Twenty-five questions were prepared on each booklet. A two by two by two f a c t o r i a l design was used with repeated measures on the l a s t f a c t o r . The fa c t o r s were: a) text (plants or c i l i a t e s ) , b) question condition (post-adjunct or control) and c) r e c a l l (short term or long term). The dependent measure was a post t e s t of twenty-five short answer questions on the c i l i a t e s , i d e n t i c a l to the relevant questions used i n the c i l i a t e booklet, but covering d i f f e r e n t d e t a i l s . A l l groups were requested to read the plant booklet, answering any questions they might encounter, without looking back to previous pages. Half of the subjects received post-adjunct questions while the other h a l f received no questions. A f t e r f i f t e e n minutes they were given the c i l i a t e booklet with the same d i r e c t i o n s . Again h a l f of the subjects received post-adjunct ques-tions while the other h a l f received no questions. When the subjects had f i n i s h e d reading the c i l i a t e booklet, the post t e s t was administered. Subjects were requested not to discuss the experiment and one week l a t e r were again given the post t e s t on the c i l i a t e s . The three way analysis of variance revealed a s i g n i f i c a n t main e f f e c t for questions on c i l i a t e s (F = 4.625, d.f. = 1, p = .05) and a s i g n i f i c a n t three way i n t e r a c t i o n between questions on c i l i a t e s , plants and measures (F = 22.088, d.f. = 1, p = .01). These r e s u l t s i n d i c a t e that under conditions of high pre-exposure to questions, there was a decline from immediate to delayed t e s t i n g . When questions were introduced with no pre-exposure, they r e s u l t e d i n a marked increase i n delayed performance. Unfortunately, t h i s study incorporated a rather p o t e n t i a l l y serious 142 confounding factor in equating incidental questions with relevant questions. While the post-adjunct questions used in the study were, relevant, the ques-tions on the f i r s t administration of the post test were incidental. By definition, the post test questions, which were incidental on the f i r s t administration, became relevant for the delayed criterion test. This test measured relevant, rather than incidental learning. This resulted in the com-parison between short and long term retention being a comparison of relevant to incidental learning. Natkin & Stahler (1969) also encountered similar confounding effects as did Frase (1967) with respect to their design. However, they did provide some information on validity. They reported a personal communication from Rothkopf indicating that the procedure for differentiating between incidental and rele-vant questions was about ninety-five percent effective for ensuring different content. Natkin & Stahler (1969) provided the f i r s t study which did not directly inform the subjects that they would receive a test. Unfortunately, they did not include a pre-adjunct question condition which would have enabled a comparison of this study with previous studies. While Natkin & Stahler's (1969) study did encounter some rather serious d i f f i c u l t i e s and while this study was not directly comparable to other studies in this review, i t was included here as the main effect reported supported previous investigations (Frase, 1967, 1968a, 1968b; Frase, Patrick, & Schumer, 1970; Rothkopf, 1966; Rothkopf & Bisbicos, 1967). This study was also included because the finding that post-adjunct questions produced superior incidental short term recall and superior incidental long term recall was a novel finding and one worth further investigation. Morasky & Willcox (1970) Morasky & Willcox (1970) conducted a series of two experiments using 143 eighty-two f i r s t and second year students at the State U n i v e r s i t y of New York at Plattsburg and a two thousand word a r t i c l e . They investigated the time r e -quired to process information as a function of question placement. The two thousand word a r t i c l e was divided into twenty paragraphs, rang-ing from forty-nine to two hundred and three words each. I t also included two graphs. Forty-two questions were prepared. Two questions were prepared on each paragraph. One question pertained to the f i r s t h a l f of the paragraph and one question pertained to the second h a l f of the paragraph. Two questions were prepared on the graphs. The questions were multiple choice, having three response a l t e r n a t i v e s . No feedback was given. Twenty-one questions were ran-domly chosen as relevant. The remaining questions were classed as i n c i d e n t a l . A two by two f a c t o r i a l design was used. The factors were: a) question placement (pre-adjunct or post-adjunct), and b) question type (relevant or i n c i d e n t a l ) . The dependent measure was a forty-two question c r i t e r i o n t e s t . Twenty-one of the questions on t h i s t e s t were relevant and twenty-one of the questions were i n c i d e n t a l . The reading completion time of the prose s e l e c t i o n was also recorded as a dependent v a r i a b l e . The subjects were requested to read the a r t i c l e as quickly as p o s s i b l e . Re-reading was discouraged. Completion time was ind i c a t e d by the subject r a i s i n g h i s hand and was recorded by the experimenter. The c r i t e r i o n t e s t was then admin i s tered. Morasky & Willcox's (1970) analysis of t h e i r f i r s t study revealed that the mean completion time for the pre-adjunct group was s i g n i f i c a n t l y l e s s than fo r the post-adjunct group (t = 4.92, p = .001). The pre-adjunct question group's rate of comprehension was s i g n i f i c a n t l y greater than the post-adjunct question group on relevant t e s t scores ( t - 4.49, p = .001), i n c i d e n t a l t e s t scores (t = 2.99, p = .001), and t o t a l t e s t scores (t = 4.13, p = .001). 144 The finding that the post-adjunct question group did not score higher than the pre-adjunct question group supported a similar finding by Rothkopf & Bisbicos (1967). Rothkopf & Bisbicos (1967) reported that when the experi-mental questions required a common word, or technical word or phrase for the answer, post-adjunct questions enjoyed no advantage over pre-adjunct questions (t = .65, p = .50). This finding was based on a Type VI analysis of variance (Lindquist, 1963). However, Morasky & Willcox's (1970) and Rothkopf & Bisbicos's (1967) analyses were subject to the d i f f i c u l t i e s associated with the use of multiple t tests. This weakness in these studies cast some doubt on the interpretability of the results. Morasky & Willcox's (1970) finding that the mean completion time for the pre-adjunct question group was significantly less than for the post-adjunct question group supported a similar finding by Frase, Patrick, & Schumer (1970). Rothkopf (1966) noted that the directional reference group and the post adjunct group resulted in markedly slower reading than the other conditions, however, he did not put the means to a s t a t i s t i c a l test. Rothkopf & Bisbicos (1967) reported no reliable differences in mean inspection time per page among the various treatments. (F = .80, d.f. = 6,245). It was interesting to note that Morasky & Willcox (1970) calculated the rate of comprehension"^ and found i t to be greater for the pre-adjunct question group on relevant test scores, incidental test scores, and total test scores. On the basis of this finding, they concluded that i f the rate of comprehension were considered important to reading then pre-adjunct common-word questions were most advantageous to the reader. Morasky & Willcox's (1970) second study investigated whether the Rate of comprehension was equal to the per cent of test items correct multiplied by the completion time. 145 a d d i t i o n a l time spent by the post-adjunct question group was used i n process-ing information i n the paragraphs or on the questions. They used thirty-two subjects drawn from the same population, s i m i l a r materials and s i m i l a r pro-cedures with the exception that the subjects were informed that the experiment was a speed t e s t and that each subject was timed by the experimenter. This technique for timing was evaluated by three judges and resulted i n a .98 i n t e r -judge c o r r e l a t i o n . Morasky & Willcox (1970) found that the t o t a l paragraph completion time required by the pre-adjunct question group was s i g n i f i c a n t l y l e s s than that required by the post-adjunct question group (Wilcoxon Rank Sum Test, T = 196.5, p = .01). However, the post-adjunct question group took s i g n i f i c a n t l y l e s s time to complete the twenty-one questions than d i d the pre-adjunct question group (p = .005, Wilcoxon Rank Sum Test, T = 158). An analysis of the scores f o r each paragraph indicated that the d i f f e r e n c e s between the pre-adjunct and the post-adjunct question groups were not evident u n t i l the twelfth, f i f t e e n t h , seventeenth, eighteenth, nineteenth, twentieth, and t w e n t y - f i r s t paragraphs. These r e s u l t s were interpreted to i n d i c a t e that a shaping procedure was taking place during the reading, rather than a constant d i f f e r e n c e e x i s t i n g from the f i r s t to the l a s t paragraph. Perhaps the major problem with these two studies, aside from the use of multiple t t e s t s , l i e s i n the d i f f i c u l t y of i n t e r p r e t i n g the r e s u l t s of the second study. In the f i r s t study, subjects were instruc t e d to read as f a s t as p o s s i b l e and were informed that any questions not answered would be scored as i n c o r r e c t . S i g n i f i c a n t differences were not found between the pre-adjunct and post-adjunct conditions, with respect to i n c i d e n t a l l earning. In the second study, Morasky & Willcox (1970) changed the d i r e c t i o n s . The subjects were in s t r u c t e d twice that the study was a speed t e s t , and that every question must 146 be answered. This d i f f e r e n c e i n d i r e c t i o n s may have resulted i n differences between; the pre-adjunct and post-adjunct question conditions with respect to i n c i d e n t a l learning. Morasky & Willcox (1970) d i d no:, report having tested for d i f f e r e n c e s i n i n c i d e n t a l learning a t t r i b u t a b l e to question placement i n t h e i r second study. Rothkopf & Bloom (1970) Rothkopf & Bloom (1970) used sixty-three p a i d volunteers from Grades eleven and twelve at Highland Park High School, Highland Park, New Jersey, and a sixteen thousand word s e l e c t i o n from Fletcher & Wolfe's, Earth Science, a high school geology text. They investigated whether o r a l post-adjunct ques-tions were more e f f e c t i v e than written post-adjunct questions. Subjects were randomly assigned to treatments and were run three at a time with each subject i n a separate booth. The treatments involved: post-adjunct written questions, post-adjunct o r a l questions, and a no-question co n t r o l group. Presentation of the text was by i n d i v i d u a l s l i d e projectors with each of the one hundred and eight pages of the prose s e l e c t i o n on a separate s l i d e . The materials d e a l t with r i v e r s , g l a c i e r s , and t h e i r respec-t i v e production of geological features. The subjects' progress through the s l i d e s was recorded by automatic cumulating counters and the time spent on each s l i d e was recorded by means of an Esterline-Angus operation recorder. A l l subjects read the f i r s t twenty-four s l i d e s without exposure to questions. In the written treatment, one question was included a f t e r every s i x t h s l i d e between s l i d e s twenty-five and one hundred and eight. Subjects responded to the questions on a s l i p of paper and placed t h e i r answer i n a b a l l o t box. This procedure was implemented to prevent the subject from r e -reading or changing an answer. Knowledge of r e s u l t s were not provided. In the o r a l treatment, the experimenter asked the questions, standing behind the 147 subject. The question was repeated once i f the subject paused i n answering. The subjects responded i n the same manner as i n the written presentation. The time f o r questioning t h i s group ranged from ten to f o r t y seconds per question. No questions were asked of the c o n t r o l group. A break of approximately two minutes was introduced a f t e r the s i x t y -s i x t h s l i d e f o r the experimenter to change s l i d e trays. The subjects l e f t the room during t h i s period. No reported contact occurred between the subjects. A t h i r t e e n item reading i n t e r e s t questionnaire was used to provide a ten min-ute break between a c q u i s i t i o n and t e s t i n g . The c r i t e r i o n t e s t consisted of a set of twelve questions on s l i d e s one to twenty-four and a set of twenty-two questions on s l i d e s twenty-five to one hundred and eight. These questions were reported to have no overlap with the questions used i n the experimental treatments. Following t h i s t e s t , the sub-j e c t s were given a second t e s t on the fourteen questions used i n the experimental treatment. Rothkopf & Bloom (1970) tabulated the number of co r r e c t responses separately i n the two question sets. The f i r s t set consisted of the twelve questions from the f i r s t 'untreated' twenty four s l i d e s . The second set con-s i s t e d of the twenty-two questions from the remaining eighty-four s l i d e s . Scores on the second set were used as the main dependent v a r i a b l e i n an analysis of covariance, while the scores from the f i r s t set were used as the c o n t r o l v a r i a b l e i n t h i s a n a l y s i s . The experimental treatments were found to have a s i g n i f i c a n t e f f e c t on r e c a l l (F = 6.87, d.f. = 2/59, p = .01). Oral questions resulted i n the highest r e c a l l performance, followed by written questions and then the no-question c o n t r o l group. Treatments i n v o l v i n g questions resulted i n higher learned performance than reading without questions (t = 3.108, p = .01) and 148 o r a l questions were found superior to written questions (t = 2.005, p = .05). The o r a l question treatment and the written question treatment r e s u l t e d i n a drop i n inspection rate between the pages j u s t preceding an experimental question and the pages immediately following an experimental q u e s t i o n ^ . With performance under the no-question condition taken as the expected outcome, the chi-square calculated f o r o r a l questions was found to be s i g n i f i -2 cant (x = 25.92, p = .001) as was the chi-square calculated f o r written 2 questions (x = 10.5, p = .01). When diff e r e n c e s i n average learning a b i l i t y among treatments were corrected by the analysis of covariance, the treatment e f f e c t was found to be s i g n i f i c a n t (F = 7.14, d.f. = 2/59, p = .01) and expo-sure to the experimental questions appeared to have su b s t a n t i a l e f f e c t s on the performance i n subsequent confrontation with experimental questions (t = 3.57, d.f. = 59, p = .001). However, the r e s u l t s of t h i s study may have been confounded by the d i f f e r e n t i a l o v e r a l l time sequence f o r groups. The o r a l question group, by manner of presentation, took longer to process than the other two groups. Rothkopf & Bloom (1970) also indicated that differences i n the experimental d i r e c t i o n s f o r the three groups may have confounded the r e s u l t s . There was also the p o s s i b i l i t y that in t e r - s u b j e c t communication occurred. Subjects may have communicated during the two minute break a f t e r the s i x t y - s i x t h s l i d e . More l i k e l y , however, was the p o s s i b i l i t y that subjects having f i n i s h e d the study communicated with subjects waiting t h e i r turn. This study also r e l i e d h e a v i l y upon multiple t t e s t s and consequently was subject to the same c r i t i -cism l e v e l e d at Rothkopf (1966). Rate was defined as the number of s y l l a b l e s i n the text s l i d e divided by the inspection time on that s l i d e . 149 Frase (1971) Frase (1971) used f o r t y - e i g h t p a i d volunteer college undergraduates from Montclair State Teachers College, New Jersey, and three constructed passages. He investigated the e f f e c t of incentive v a r i a b l e s and type of adjunct question upon text learning. A two by two by two by two f a c t o r i a l design was used with repeated mea-sures on the l a s t f a c t o r . The factors included: a) question p o s i t i o n (pre-adjunct or post-adjunct), b) incentive l e v e l (half the subjects were informed that they could earn four d o l l a r s and twenty cents f o r answering a l l the questions c o r r e c t l y on the post t e s t ) , c) incentive p o s i t i o n (subjects were informed, e i t h e r before or a f t e r reading the prose s e l e c t i o n , about the incen-t i v e condition), and d) type of r e c a l l (fact or inference). Separate analyses were c a r r i e d out on r e c a l l f o r assertions i n c i d e n t a l to the adjunct relevant questions and f o r r e c a l l on the adjunct relevant questions. A two by two by two analysis of variance was also performed on the correct solutions to the questions. Three passages were used. They described: people i n a foreign country, production of automobiles, and astronomical d i s c o v e r i e s . Each passage had four experimental text sentences embedded i n i t . Two adjunct questions were presented e i t h e r before or a f t e r each paragraph i n the form of a sentence. Subjects responded ' v a l i d ' or ' i n v a l i d ' to each question. Factual questions described r e l a t i o n s h i p s within sentences, while i n f e r e n t i a l questions described r e l a t i o n s h i p s between sentences. Each subject was exposed to the same type of question on each passage. The t e s t s involved having the subjects write a l l that they could remember concerning each passage, along with a l l the v a l i d inferences that could be derived. These te s t s were administered immediately upon the comple-150 t i o n of each passage. The subjects were randomly assigned to the eight experi-mental conditions and progressed through the prose s e l e c t i o n at t h e i r own rate. The analysis of variance indicated that incentive information before reading produced more corre c t solutions on the adjunct questions than when i t was given a f t e r reading. (F = 5.0, d.f. = 1/40, p = .01). Pay produced more co r r e c t solutions to adjunct questions than no pay (F = 8.6, d.f. = 1/40, p = .01). I n f e r e n t i a l adjunct questions produced higher r e c a l l than f a c t u a l questions (F = 37.7, d.f. = 1/40, p = .001). Incentive information before reading r e s u l t e d i n higher r e c a l l than a f t e r reading (F = 9.3, d.f. = 1/40, p = .005). Incentive p o s i t i o n interacted with the type of adjunct question (F = 9.2, d.f. = 1.40, p - .005) such that the Inference group r e c a l l e d more f a c t s and inferences when incentive i n f o r -mation was given before reading than a f t e r reading. Overall r e c a l l f o r fact s was found to be greater than f o r inferences (F = 4.14, d.f. = 1/40, p = .05). Incentive information given before reading re s u l t e d i n higher r e c a l l of the relevant adjunct questions than when i t was given a f t e r reading (F = 6.76, d.f. = 1/40, p = .025). R e c a l l f o r the f a c t u a l relevant questions was higher than f o r the i n f e r e n t i a l relevant questions (F = 91.04, d.f. = 1/40, p = .001) with the exception that the Fact group r e c a l l e d more f a c t s than they had seen as relevant adjunct questions and the Inference group remembered more i n f e r -ences than they had seen as relevant adjunct questions (F - 5.24, d.f. = 1/40, p = .05). I t was i n t e r e s t i n g to note that t h i s study d i d not f i n d a s i g n i f i c a n t main e f f e c t f o r incentive as noted by Frase, P a t r i c k , & Schumer (1970). Frase attempted to explain t h i s r e s u l t i n terms of a confounding e f f e c t induced by the high incentive of paid volunteers as compared to captive subjects. Conse-quently, Frase's (1971) use of paid volunteers severely l i m i t e d the g e n e r a l i z a b i l i t y of t h i s study. However, these findings d i d support those of previous i n v e s t i g a t i o n s (Frase, 1967, 1968a, 1968b; Frase, P a t r i c k & Schumer, 1970; Natkin & Stahler, 1969; Rothkopf, 1966; Rothkopf & Bisbicos, 1967; Rothkopf & Bloom, 1970) of question p o s i t i o n on i n c i d e n t a l learning. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

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.831.1-0054407/manifest

Comment

Related Items