[NiTiAL AND DELAYED RETENTION OF RELEVANT AND INCIDENTAL. ' INFORMATEON FROM A PROSE E PASSAGE PRESENTED ,AURALLY T0 * TRAENABLE MENTAL REI'ARDATES Thesis for the Degree of Ph. D. MlCHiGAN STATE UNEVERSITY SUE A. WiSENBERG ' 1970 «(Ht-.93“ a.“ _ ‘ t.) g L I' .1: RA R Y I Mzchigzm State “.5 LIE"! . a. 7? L s a. a. ' ‘ . y .. .( This is to certify that the thesis entitled INITIAL AND DELAYED RETENTION OF RELEVANT AND INCIDENTAL INFORMATION FROM A PROSE PASSAGE PRESENTED AURALLY TO ’. TRAINABLE MENTAL RETARDATES \ presented by L f SUE A . WISENBERG has been accepted towards fulfillment j . /\ of the reqmrements for \ Ph . D . degree in EDUCAT ION Date_ 0-7639 smSma av 1: “MB & SUNS' 800K BINDERY INC. uamnv among ABSTRACT INITIAL AND DELAYED RETENTION OF RELEVANT AND INCIDENTAL INFORMATION FROM A PROSE PASSAGE PRESENTED AURALLY TO TRAINABLE MENTAL RE TARDATES By Sue A. Wisenberg This study was conducted in order to investigate the facilitation of immediate and delayed recall of factual information from prose materials presented aurally to trainable mental retardates. Trainable mentally retarded individuals receive most of their verbal information by auditory means due to their inability to read. Therefore, an investigation of methods designed to facilitate the retention of prose materials presented aurally to trainable mental retardates could contribute useful information for the instruction of these children. It was also assumed that this kind of investigation could lend itself to an existing theory concerning the learning processes of trainable mentally retarded children. Further- more, it was assumed that such studies would disclose some of the Sue. A. Wisenberg variables which influence the recall of factual information of aurally presented prose materials. Seventy -two trainable mentally retarded students were individually read a 10-paragraph prose passage via one of six pre- sentation methods which manipulated the location of adjunct questions and a procedure designed to gain the attention of the retarded §s. These manipulations were based upon the inspection behavior theory of Rothkopf (1963, 1965) and the attention theory of retardate learning (House and Zeaman, 1960; Zeaman and House, 1963). The same 30-item criterion test was administered to each _S_ immediately . following each of the three readings of the prose passage. The criterion test consisted of 15 three -foi1 multiple choice relevant questions and 15 three -foil multiple choice incidental questions. The relevant questions pertained to information in the passage to which the fi' 3 attention had been directed by means of a special procedure and were identical to adjunct questions which some §s heard during the reading of the passage. The fi' 3 attention was not directed toward the information in the passage which answered the incidental questions and these questions were not heard by any _S_ during passage presentation. Immediate, 100 per cent reinforcement in the form of M 8: M candies was used for all §s during the first three administrations of the criterion test. A measure of delayed Sue A. Wisenberg retention was also obtained one week following the three initial administrations of the passage and criterion tests. No reinforcement was administered during the delayed retention segment of the study. Analysis of covariance was used to analyze the 2 X 3 X 5 X 2 factorial design with repeated measures on the last two factors. The first factor, cueing, consisted of two levels--no cueing and cueing. The second factor was question location and consisted of three levels—- no questions, questions after each paragraph, and questions before each paragraph. The third factor represented the five administra- tions of the same criterion test. The fourth factor assessed two types of information, relevant and incidental. The covariable was the total number of correct answers tO questions immediately following the reading of each of three short pre ~passages. The dependent variable was the number of correct relevant and incidental questions for each administration of the criterion test. The repeated measures analysis of covariance and subse- quent post hoc investigations indicated that trainable mentally retarded §s presented with an aural prose learning task via a pro- cedure designed to gain attention and/or adjunct questions did not benefit from any of the six presentation methods. Furthermore, these _S_s were not facilitated in the retention of factual information via the procedure designed to gain attention or adjunct questions, Sue A. Wisenberg regardless of location. The trainable mentally retarded _S_s did retain relevant information better than incidental information. And finally, these §s retained as much relevant and/or incidental information one week following the original presentations as they did immediately following the original presentations. INITIAL AND DELAYED RETENTION OF RELEVANT AND INCIDENTAL INFORMATION FROM A PROSE PASSAGE PRESENTED AURALLY TO TRAINABLE MENTAL RETARDATES By \ Sue AfVWisenberg A THESIS Submitted to Michigan State University in partial fulfillment Of the requirements for the degree of DOC TOR OF PHILOSOPHY Department of Educational Psychology 1970 CHM/Z ACKNOWLEDGMENTS The author would like to take this opportunity to express her sincere gratitude to Dr. Clessen J. Martin, chairman, for his guidance throughout her doctoral program. Dr. Martin' s sugges - tions and time during the conduction and completion of this dissertation were, also, greatly appreciated. Gratitude is also extended toward the other committee members, Dr. Donald Burke, Dr. Andrew Porter, and Dr. William Stellwagen, for their direction and suggestions. The assistance of Mr. David Wright during the analysis stage of the dissertation was invaluable. Mr. Wright' s suggestions and insight were greatly appreciated. Finally, my husband, family and friends are to be thanked for their continued faith and support throughout my entire graduate program. Successful completion of my graduate studies would not have been possible without the unselfish assistance which they lent. ii CHAPTER TABLE OF CONTENTS LIST OF TABLE S LIST OF FIGURE S THE PROB LE M Purpose , Theory , Hypotheses REVIEW OF THE LITERATURE Attention as a factor affecting learning by the mentally retarded , Influential variables in written prose learning studies Question location Question pacing Question mode . . . . . Repetition in written prose learning studies Repetition as a factor in retardate learning. . Definition of terms . Hypotheses DESIGN OF THE STUDY Sample . . . . . . . Design and experimental treatments Materials iii Page vi . viii 01 10 10 13 13 17 22 23 25 29 3O 34 34 35 39 CHAPTER Procedure . . P—re -passages . Vocabulary checklist Test passage . IV. RESULTS OF THE ANALYSES . Main Analysis Exact Hypotheses V. DISCUSSION . Introduction Conclusions Discussion . Engineering the attention of the retardate The inducement of mathe - magenic behavior Initial versus delayed retentiOn Implications for Future Research VI. SUMMARY . LIST OF REFERENCES APPENDIX A. INDIVIDUAL CHRONOLOGICAL AGES AND INTELLIGENCE QUOTIENTS . B. THE THREE PRE -PASSAGES C. THE TEST PASSAGE: "A TRIP TO THE BEACH" D. QUESTIONS FOR THE PRE -PASSAGES E. CRITERION TEST QUESTIONS iv Page 42 43 44 45 49 49 63 73 73 74 75 75 79 82 83 86 93 96 98 . 101 103 105 APPENDIX F. ORDERS OF ANSWERS FOR PRE - PASSAGE ONE: "THEY WALK AT NIGHT" ORDERS OF ANSWERS FOR PRE - PASSAGE TWO: "THE KING" . ORDERS OF ANSWERS FOR PRE - PASSAGE THREE: "FUNNY TREES" VOCABULARY CHECKLIST INSTRUCTIONS FOR EACH TREATMENT GROUP ORDERS OF ANSWERS FOR INITIAL CRITERION TESTS ORDERS OF ANSWERS FOR THE DELAYED CRITERION TEST DEPENDENT VARIABLES AND COVARIA BLE Page 107 108 109 110 111 114 118 122 _ TAB LE LIST OF TABLE S Means and Standard Deviations of the Chronological Ages and Intelligence Quotients of the Six Treatment Groups Results of Item Analyses on the Pilot of the Criterion Test for Group I (P) . Results of Item Analyses on the Pilot of the Criterion Test for Group IV (PQ*) , Correlations Between the Covariate and Number Correct Relevant and Incidental Questions on Each Admin- istration of the Criterion Test for the Six Treatment Groups . Mean Number of Correct Answers for Relevant and Incidental Questions at Each Administration of the Criterion Test for Each of the Six Treatment Groups Summary Table for 2 X 3 X 5 X 2 Analysis of Covariance with Repeated Measures on the Last Two Factors Matrix of Differences in Mean Total Number Correct Questions Between Administrations of the Criterion Test Mean Number of Correct Answers for Relevant and Incidental Questions for Each Administration of the Criterion Test vi Page 35 41 42 50 52 56 58 60 TABLE Page 9. Matrix of Differences in Mean Number Correct Relevant Questions Between Administrations of the Criterion Test . . . . 61 10. Matrix of Differences in Mean Number Correct Incidental Questions Between Administrations of the Criterion Test . . . . 62 11. Matrix of Differences Between Mean Number Correct Relevant and Incidental Questions at Each Administration of the Criterion Test . . . . . . . . . . . . . 64 12. Summarized Restatement of Hypotheses One through Nine . . . . . . . . . . . . . 65 13. Mean Total Number Correct Relevant and Incidental Questions for Each Treatment Group . . . . . . . . . . . . . 66 14. 1 X 6 Analysis of Covariance Summary Table for Total Number Correct Relevant Questions for Each Treatment Group . . . . . . . . . . . . . . 67 15. Weighted Scheffé Contrasts for Hypotheses Concerned with Retention of Relevant Information . . . . . . . . . . . . 68 16. 1 X 6 Analysis of Covariance Summary Table for Total Number Correct Incidental Questions for Each Treatment Group . . . . . . . . . . . . . 71 17. Weighted Scheffé Contrasts for Hypotheses Concerned withRetention of Incidental Information . . . . . . . . . . . 72 18. Variances of Relevant and Incidental New Correct Responses and Incorrect Responses Previously Correct After the Third Administration of the CriterionTest...............78 Vii LIST OF FIGURES FIGURE Page 1. Data Matrix of the 2 X 3 X 5 X 2 Repeated Measures Design . . . . . . . . . . . . . 37 2. Mean Number Correct Relevant Questions for Each Treatment Group at Each Administration of the Criterion Test . . . . 53 3. Mean Number Correct Incidental Questions for Each Treatment Group at Each Administration of the Criterion Test . . . . . . . . . . . . . . 54 4. Means for Relevant and Incidental Information Across All Treatment Groups..................59 viii CHAPTER I THE PROBLEM The school setting is one in which information to be learned and retained is administered through written instruction, auditory instruction and various combinations of the two. Until recently, investigators paid little attention to written instruction. However, the last decade of research in verbal learning has evidenced an increased concern with written prose learning. The general char- acteristic of these prose studies has been the facilitation of the recall of factual information. Various presentation methods have been used in an attempt to find variables which facilitate the recall of factual information in written prose. Those variables which have facilitated the recall of factual information in written prose materials presented to normal subjects might also be expected to facilitate the recall of factual information in prose materials presented aurally to mentally retarded subjects. This distinction between written instruction of the normal student and oral instruction of the retarded student is of major importance. Both written and oral means of instruction are used for the normal student. However, the majority of non -deaf mentally retarded children obtain most Of their verbal information through the auditory channel. This is due to difficulties they encounter in learning to read. Furthermore, almost all trainable mentally retarded children receive their entire verbal instruction by means of some type of aural presentation. This is due to the trainable mental retardate' s increased inability to read in comparison to educable mentally retarded children. Most educable mentally retarded children can read at primary reading levels and some of them can read at a third and fourth grade reading level. However, most trainable mentally retarded children cannot read. They encounter difficulties at the earliest reading stages, such as letter discrimination and word recognition. Therefore, auditory instruction is the primary means for the administration of information to be learned and retained by the trainable mentally retarded. Aural presentation methods of prose materials are important to the trainable mentally retarded since they rely upon the aural reception of the verbal information. However, little information is available concerning optimal methods of presenting auditory prose to these children. Investigation of various aural presentation methods designed to facilitate the recall of factual information in prose materials would contribute useful information for the instruc - tion of trainable mentally retarded children. Such studies could possibly provide suggestions for facilitating the aural recall of factual information by these children. In addition, this kind of investigation could lend itself to an existing theory concerning the learning processes of such children. And furthermore, such studies would disclose some of the variables which influence the recall of factual information of aurally presented prose materials. Purpose The purpose of this study is to investigate the facilitation of immediate and delayed recall of factual information from prose material presented aurally to trainable mental retardates. Various presentation methods will be used in an attempt to facilitate aural recall and, thereby, provide guidelines for enhancing recall of factual information of aurally presented material. Some of the presentation methods will be based upon the attention theory of retardate learning as proposed by Zeaman and House (1963). As such, these methods will manipulate a signal designed to gain the attention of the subject during the aural presen- tation of the prose material. Other methods will be based upon the inspection behavior theory of Rothkopf (1963, 1965). This theory has been investigated in studies of the facilitation of the recall Of factual information from written prose. Influential variables found to facilitate recall in written prose studies will be manipulated in the present study. This will be done in order tO ascertain if the same variables facilitate the recall of factual information in prose material presented aurally. In addition, both of the above mentioned theories will be used as a basis for still other presentation methods. These methods will manipulate various combinations of a signal designed to gain the attention of the subject and influential variables found to facilitate the recall of factual information from written prose. Finally, the present study will investigate the assumption that the mentally retarded benefit from repetition. This assumption is put into daily practice by teachers of the mentally retarded and is continuously employed in research paradigms concerning the learning of the mental retardate. Both educators and researchers give multi -presentations of the learning material to retarded sub- jects. The basis for this methodological procedure would seem to be that mental retardates perform in an incremental fashion when engaged in a learning process. Theory Two theoretical positions form the basis for the development of this study. Both of these positions are based upon different learn- ing tasks than those employed in this investigation. However, both theoretical positions are applicable to the purpose of this study. The first of these two theoretical positions is the attention theory of retardate learning as proposed by Zeaman and House (1963) and House and Zeaman (1960). It was derived from investigations in two-choice, simultaneous, visual discrimination learning tasks using normals and mental retardates as subjects. House and Zeaman (1960) and Zeaman and House (1963) concluded that the deficit of retardates as compared to normals was an attention deficit. Until the attention of the retarded subjects was focused on the relevant characteristics of the learning material, retardates made more errors than normal subjects. When the retarded subjects began to pay attention to the relevant characteristics of the learning material, the learning of the normal and retarded subjects was comparable. Therefore, retar- dates did not suffer from a learning deficit but, instead, a deficit in the ability to attend to relevant characteristics of the learning mate- rials. This would suggest that retardates must first learn to attend to the relevant characteristics of the material to be learned before learning of the material can occur. Attempts to direct the attention Of retarded subjects to the relevant characteristics of the learning material would, hopefully, decrease their attention deficit and, thereby, facilitate their learning of the material. In fact, Zeaman and House (1963) suggest ways of engineering the attention of retar- dates in order to draw their attention to the relevant characteristics of the learning material. These suggestions are made as an attempt to overcome the attention deficit of retarded subjects in two—choice, visual, discrimination learning tasks. It is plausible to suggest that the attention of retarded sub- jects could also be engineered in a prose learning task which involves aural preSentation of the material to be learned. Such a methOd would direct the attention of the retardate to the relevant character- istics of the learning material. This would, hopefully, aid the retardate in reduction of an attention deficit and might facilitate his learning of the material. The second theoretical position used as a basis for this study is the inspection behavior theory of Rothkopf (1963, 1965). Rothkopf (1963, 1965) presents a conceptual model of learning from written sentences in which he proposes three classes of inspection behaviors. These behaviors are defined as those activities which subjects engage in during reading. Accordingto RothkOpf (1963, 1965) there are two categories of stimuli in learning from written materials. The first category consists of stimulus objects of a physical nature. Stimulus objects are the printed symbols to which subjects are exposed. In other words, stimulus objects are any part of the written material pre - sented to the subject. The second category consists of effective stimuli of a psychological nature. Effective stimuli refer to the psychological consequences of being exposed to the stimulus objects. The formation of associations between the stimulus and response components of a sentence are necessary in order for learn- ing to occur. RothkOpf (1963, 1965) proposes that the formation of these associations depends upon the character of effective stimula- tion. Furthermore, the character of effective stimulation depends upon the activities subjects engage in during reading. As mentioned before, RothkOpf (1963, 1965) calls these activities ”inspection ' He proposes three classes of inspection behaviors, of behaviors. ' which "mathemagenic" behavior is one. The term mathemagenic ”is derived from the Greek roots mathema (that which is learned) and gignesthai (to be born)" (Rothkopf, 1965, p. 199). Therefore, mathemagenic behaviors are activities subjects engage in which produce learning during the reading process. RothkOpf (1965) suggests that mathemagenic behaviors can be induced and brought under control by "test —like events" or questions interspersed within the reading passage. He presents the results of an unpublished paper which verify his suggestion. Results of later research (RothkOpf, 1966) also indicate that when students are questioned over material just read, their recall of the factual infor- mation is facilitated if questions have been interspersed within the reading passage. It is plausible that the benefits to be gained from the use of questions in learning from prose materials are not confined to written prose. Questions interspersed within a prose passage presented aurally may also facilitate the recall of factual information. Hypotheses Based upon research to be reviewed in the next chapter, it is expected thatthe use of questions in the auditory presentation of prose materials to trainable mentally retarded subjects will yield results similar to those obtained in studies of written prose materials presented to normal subjects. It is also expected that the use of a signal during auditory presentation of prose material will direct the attention of the trainable mental retardate to the relevant sentence within the passage. It is hypothesized that questions located within the passage will facilitate the immediate recall of factual information relevant to the interspersed questions. It is also hypothesized that a particular location of questions within the passage will facilitate the immediate recall of factual information which is not relevant to those questions. It is further hypothesized that the same particular location of ques - tions will facilitate the delayed recall of factual information relevant to those questions. Furthermore, it is hypothesized that a signal designed to gain attention will facilitate the immediate recall of factual information found in each sentence which follows the signal. CHAPTER II REVIEW OF THE LITERATURE The comprehension of prose materials presented aurally to trainable mentally retarded subjects might be facilitated by devising appropriate methods for presentation of the materials. These methods could take into account manipulation of the attention of the trainable mental retardate and influential variables in the learning Of prose materials. Repetition of prose material pre - sented to trainable mentally retarded subjects may or may not be of importance in facilitating their recall of such materials. Attention as a factor affecting learning by the mentally retarded Attention is of the utmost importance to mental retardates in a learning situation. McCarthy and Scheerenberger (1966) present a discussion of attention as a factor affecting the learning of the mentally retarded in a ten -year review of applied and basic research applicable to the classroom of the mentally retarded. Their discus - sion centers around the attention theory of Zeaman and House (1963). 10 11 The attention theory of Zeaman and House (1963) and House and Zeaman (1960) grew out of two-choice, simultaneous, visual discrimination learning problems presented to normals and trainable mental retardates. Typically, the stimuli, "junk objects" or colored geometric forms, were presented to one subject at a time on a modi- fied Wisconsin General Test Apparatus. Candy was found under the correct stimulus. The general finding was that normals made sig — nificantly fewer errors than the trainables. When backward learning curves were constructed, an ogival shaped learning curve resulted. This type of learning curve was initially flat at chance level and increased sharply near the criterion. Zeaman and House (1963) found that the initial flat part of the curve varied with IQ while the final approaches to criterion were the same for both groups. Slow learners were found to remain at chance level for a longer period than normals but moved quickly toward criterion when their performance began to improve. According to Zeaman and House, the dimensional aspects of the stimuli elicit attentional responses which then elicit specific cues leading to the eventual elicitation of instrumental responses. It was felt that the initial flat chance section of the curve was controlled by an attention process and the sharp increase was due to the learning of the correct instru— mental response. Therefore, Zeaman and House concluded that the 12 retardate deficit in discrimination learning was not due to an inability to learn. Instead, the retardate learning deficit was due to an inability to direct and maintain attention to relevant stimulus dimensions. The difference between the fast and slow learners was considered to be the number of trials for learning to begin. House and Zeaman (1960) proposed that the attention of the mental retardate could be engineered or directed toward relevant stimulus dimensions. Zeaman and House (1963) presented specific suggestions for drawing the attention of retardates to the relevant stimulus dimensions. Some of these suggestions were the use of novelty, three -dimensional stimulus Objects and large stimulus cues. Although these suggestions were for visual discrimination tasks, specifically, the concept of engineering the attention of the retardate may profitably be extended to other learning tasks. The compre - hension of prose materials presented aurally to retardates is a learning task which lends itself to such an extension. Presentation methods can be devised to direct the attention Of the retardates to relevant sentences within the passage. These sentences would con- tain the necessary information to correctly answer a question con- cerning the passage. l3 Influential variables in written prose learning studies Most prose learning studies have used normal subjects, the majority of which have been college students or adults. Furthermore, such studies have concentrated on written prose materials. These studies have been consistent in their findings and the identification of influential variables which facilitate learning of written prose. Three of these variables are question location, question pacing, and question mode. These variables may also be influential in the facilitation of the comprehension of prose materials presented aurally to trainable mental retardates. Question location. --Rothkopf (1966) was the first to manipu- late the location of questions within a written prose passage. His interest in the manipulation of this variable was based upon earlier conceptions (RothkOpf, 1963) that "test-like" events could induce and control mathemagenic behavior. Adjunct questions and instructions were thought of as "test-like" events. RothkOpf (1966) individually presented college subjects with a 5200 -word written prose passage divided into seven sections. Thirty -nine short answer questions were devised. Two questions from each section were used as experi- mental questions (EQs) during reading and for the experimental question relevant test (EQRT). The other 25 questions were used for the general test (GT). Four treatment groups differed in whether 14 the EQs came before or after each section, with or without-knowledge of results. A fifth treatment group (LBA) received all 14 EQs and their appropriate answers before reading the passage. A control group (C) received no EQs or knowledge of results and were told to remember as much as they could. A direction reference group (DRG) received no EQs or knowledge of results and were told to read carefully and slowly for factual information. Each subject, except those in the transfer evaluation group, was administered the general test followed by the EQRT. The transfer evaluation group took the general test, studied the EQs with knowledge of results and when perfect, took another form of the general test. No specific transfer from EQs to the general test was found for the transfer evaluation group. The results on the EQRT scores indicated that all groups having EQs did better than the control group or the direction reference group. Neither of the latter two groups received EQs. Also, the groups re- ceiving knowledge of results, regardless of question location, did better than the groups not receiving knowledge of results. Further- more, there were no differences in EQRT scores due to location of EQs in the text. And finally, the direction reference group received significantly higher EQRT scores than the control group. The results for the general test.(GT) scores indicated that only the direction reference group and the group receiving post -questions without 15 knowledge of results did better than the control group. Pre -question groups did not evidence general facilitative effects. Rothkopf and Bisbicos (1967) further explored mathemagenic behavior concepts by using restricted categories of questions in the text and increasing the number of pre - and post -experimental ques - tions. High school subjects were given a 9000 -word written prose passage cOncerning animals and minerals in the ocean. This passage was divided into 12 three -page zones. Two short-answer questions for each of four restricted categories-were devised for each three - page zone. The four restricted categories were common phrases (C), technical phrases (T), measures of size, date, distance (M), and names (N). Four short -answer questions, one per category, from each zone were used as EQs during reading. The other four short- answer questions, one per category, were used for the criterion test. Thus, the study was measuring the recall of incidental information. All subjects in six treatment conditions were given two EQs per three- page zone which differed in type and location. The EQs dealt with names and measures (NM), common phrases and technical phrases (CT), and a mixture of all four restricted category types (MX). The EQs were located either before or after each three -page zone. A seventh treatment condition was a control group which received no EQs during reading of the passage. The results showed that subjects 16 whoreceived post -questions in each of the restricted category groups (NM, CT, and MX) produced significantly higher criterion test scores than the control group. There were no differences between any of the pre -question restricted category groups and the control group. Also, there were significant differences in favor of post -questions as opposed to pre -questions for the mixture (MX) and name -measure (NM) restricted category groups. This result was not found for the common -technical phrase (CT) group. Furthermore, the facilitative effect of post-questions was found to be better for the second half of the prose passage than for the first half. To summarize, the results indicated that post —questions had a general facilitative effect when restricted categories of questions were used and that the fre- quent use of post —questions shaped inspection behavior. Question location has been included as one of several inde - pendent variables under investigation in each of three separate studies by Frase (1967, 1968a, 1968b). These studies will be reviewed in the next subsection since they have been primarily con— cerned with the variable of question pacing. However, it should be noted at this point that the findings of RothkOpf and Frase related to question location are in agreement and disagreement. Both authors have found post -questions to facilitate the retention of incidental information. In terms of the facilitation of the retention of information 17 Rothkopf (1966) finds both pre - and post -question groups to retain more relevant information than a control group. But, he finds no difference between pre - and post -question groups in retention of relevant information. Whereas, Frase (1967, 1968a, 1968b) finds post -question groups to retain more relevant information than pre - question groups. These consistent and inconsistent results have been found when the length and content of material as well as response mode have differed in the investigations of Rothkopf and Frase. Question pacing. —- The investigation of question pacing and the interaction of this variable with question location and the type of retention items has been a major concern of Frase (1967, 1968a, 1968b). Question pacing refers to the frequency of questions to which subjects have been exposed during reading. Frase has been interested in the effect of this variable on subsequent retention of information and the effect of this variable on the location of questions. Frase (1967) presented 12 groups Of college subjects with written prose material which differed in passage length, question location, and whether or not knowledge of results was given. Questions were paced every 10, 20, or 40 lines within the passage. Questions (EQS) were located either before or after each of these designated passage lengths. A biographical prose passage of 20 ten -line paragraphs was used. Two multiple -choice factual questions 18 were devised for each paragraph. One concerned information found in the first part of the paragraph and the other concerned information found in the second part of the paragraph. The ques - tions relating to the second part of each paragraph were used as EQs during reading and on the retention test. The other 20 questions were used to measure incidental information on the retention test. Each subject took the retention test immediately after reading the prose passage. In terms of relevant information, it was found that post-questions were more facilitative than pre -questions. Also, the 20-line passage was found to have the most facilitative effect on relevant information. In addition, the presence of knowledge of results was more facilitative than no knowledge of results. Finally, a significant interaction of question location and knowledge of results was found for relevant information. This result indicated that no differences existed between post -question and pre —question groups when knowledge Of results was given. However, post -questions were more effective than pre -questions when no knowledge of results was given. In terms of incidental information, it was found that post -questions were more facilitative than pre -questions. There were no differences for the main effects of passage length or knowledge of results on the facilitation of incidental information. l9 Frase (1968a) included question pacing, question location, retention items, and blocks in a four -factor design with repeated measures on the last two factors. Questions were paced every 10 or 20 sentences and were located before or after each pacing segment. Retention items consisted of relevant and incidental information. Four blocks of five paragraphs each were used in order to ascertain if post—questions entail a learning to learn phenomenon or elicit. a previously acquired general problem solving skill. College subjects were given the same written prose passage as in the previous study by Frase (1967). The passage was administered under the various conditions designated by the 2 X 2 X 2 X 4 design. No knowledge of results was given and upon immediate completion of the reading material all subjects were given a multiple -choice test of 40 items. Twenty of these items had been used as EQs during reading and were designated as relevant retention. items. The other 20 items had not been seen by the subjects during reading and were designated as incidental retention items. Frase found that post -question groups retained both relevant and incidental information better than pre - question groups. Furthermore, relevant information was better retained than incidental information for pre- and post-question groups. A significant Question Location X Question Pacing interaction was found which evidenced an opposite effect of pacing for pre - and 20 post—question groups. Questions with every paragraph were found to be advantageous for post -question groups whereas questions with every other paragraph were advantageous for pre -question groups. Finally, a significant main effect for blocks was found but the inter— action of Blocks X Question Location was not significant. These results indicated that pre - and post -question groups were affected in the same way across blocks. Frase then plotted the percentage of subjects in post-question groups who passed items not passed by pre -question groups. This was done for those groups who received questions with every paragraph. He found 10 to 40 per cent of the post -question groups above the pre —question groups on each paragraph except the seventeenth. He concluded that post -question groups were consistently above pre -question groups when frequent questioning was used. Furthermore, Frase concluded that the facilitation effect of post-questions was not acquired during the reading task but was due to the fact that they reinforce and maintain a previously acquired general problem solving skill. Frase (1968b) further explored the variable of question pacing in a five -factor design with repeated measures on the last factor. The five factors were question location, question pacing, content location of question relevant material, question mode, and retention items. College studentswere used as subjects and were 21 administered the same 2000 word written prose passage as used previously by Frase (1967, 1968a). In this experiment the EQs were paced before or after every 10, 20, 40, or 50 sentences. Multiple- choice questions and constructed response questions were used as EQS. In addition, half of the subjects received EQs relating to the second half of the paragraph. A criterion test of 20 relevant and 20 incidental multiple -choice questions was taken by each subject immediately following completion of the reading material. This was the same test as used in the previous two studies. Frase found a main effect for question location which showed that post —questions were more facilitative than pre -questions. He also found retention of relevant information to be significantly higher than retention of incidental information. A significant interaction between question pacing and question location indicated that the more frequent the ques - tions, the more advantage for post-questions and the more disadvantage for pre —questions. A significant interaction between question pacing and retention items indicated a high degree Of relevant retention regardless of the frequency of questions and a depression of incidental retention due to frequent questioning. A significant interaction between content location of question relevant material and retention item also occurred. This result indicated that incidental information was better retained when it was located in the second half of the 22 paragraphs or, in other-words, when it followed relevant information. Question mode. -- The form or construction of questions used as test-like events in the studies reviewed so far have been both multiple -choice and short -answer. It would appear that question mode is not an influential variable in the facilitation of the compre- hension of written prose materials since consistent results have been found for the variables of question location and question pacing. Only one of the previously cited studies included question mode as an independent variable and systematically varied two fixed levels of this factor (Frase, 1968b). Multiple -choice and constructed response were used. Frase (1968b) found no significant differences for question mode. In a study related to question mode, Frase (1968c) investi— gated the orientation effects of three different types of introductory questions. College subjects were given a general, specific, or comparative introductory question and the same written prose para - graph. This was immediately followed by a nine -item multiple - choice test concerning the material read. Frase found that the specific question group had the highest per cent correct on the test item relevant to their introductory question and the highest number of correct responses on the total test. These findings were 23 attributed to the fact that specific questions allow subjects to rehearse the stimulus, response, and the association of the two. Repetition in written prose learning studies Slamecka (1959, 1960) investigated the effects of repetition on the rote learning of prose materials via a serial anticipation pro- cedure. His objective was to determine whether the manipulation of classic variables in the retention Of unconnected discourse evidenced the same results when they were varied in a study Of the rote reten- tion of prose materials. Slamecka (1959), in one of three experiments, explored verbatim recall of prose material via the classic variable of degree of original learning. College subjects were given either 1, 3, 5, 6, 7, or 9 presentations of the original passage followed by original recall. Then three presentations Of an interpolated passage and, finally, delayed recall of the original passage. Both original recall and delayed recall Of the original passage were found to increase as degree of original learning increased. Slamecka (1960) studied the phenomenon of retroactive inhibition in rote retention of connected discourse. College subjects were given four pairs of sentences with the first sentence as original learning and the second sentence as interpolated learning. Each sub- ject was given various combinations of 2, 4, or 8 original learning 24 trials followed by 0, 4, or 8 interpolated learning trials. In addition, four relearning trials of the original learning task and three relearn— ing trials of the interpolated learning task were also administered. The acquisition of the original learning task was found to signifi- cantly increase with increasing original learning trials at every level of interpolated learning. I A related study by Rothkopf (1968) measured the textual constraint of written prose passages. College and/or high school subjects were given one of two passages for 0, 1, 2, or 4 times. Ten minutes after reading the assigned passage, subjects were given a completion test which omitted nonfunction words. The correct number of deleted words increased with exposures for both passages in a negative accelerating fashion. Asymptote occurred at two exposures. Rothkopf and Coke (1963, 1966) performed two studies with adult females which investigated the effect of immediate and delayed repetition of sentences during acquisition on subsequent recall of those sentences. The dependent variable was the number of correct deleted terms in the previously presented sentences. Rothkopf and Coke (1963) gave eight pairs of sentences to all subjects under various rehearsal methods for the second sentence of each pair. Subjects rehearsed the second sentence of each pair when it was 25 presented for the second time. In addition, the rehearsal of the second sentence of each pair was either immediately after its initial presentation or following interpolated sentences. Rothkopf and Coke (1966) presented a prose passage to all subjects in which particular sentences were or were not rephrased when repeated later in the passage. In addition, repetition of the test sentence was either immediate or followed interpolated sentences. In both studies, delayed repetition of the material to be learned produced better recall than immediate repetition. Ausubel and Youssef (1965) investigated spaced repetition of a written prose passage. An experimental group of college sub- jects was administered the learning passage on the first and third days of the experiment. A control group of college subjects received another passage on the first day and the learning passage on the third day. Both groups were given a multiple -choice retention test on the fifth day. Retention scores were significantly higher for the experimental group than for the control group. Repetition as a factor in retardate learning For years it has been assumed that retardates benefit from repetition of the material to be learned. Therefore, repetition of the material to be learned is often times included as a methodological 26 procedure regardless of the learning material or major intent of the study. Two recent reviews of basic research with suggested appli- cations for the classroom situation (Vergason, 1968; McCarthy & Scheerenberger, 1966) suggest the use of repetition in learning tasks administered to mental retardates in order to overcome their short- term memory deficit. However, a recent study found repetition to be of little benefit to trainable mental retardates in a short -term recall task. Baumeister, Hawkins and Holland (1967) conducted a study of retroactive inhibition in short —term recall of a digit sequence by normals and trainable mental retardates of the same chronological age. A five place digit span was used for the trainables and a nine place digit span was used for normal subjects. Levels of interpolated tasks were 0, 1, 2, or 3. They found that trainable mental retardates were below chance for levels 1, 2, and 3 of the interpolated tasks. Furthermore, trainables were just above chance after 25 repetitions of the same digit sequence with no interpolated task. Two studies differing in major intent have used repetition with educable mentally retarded subjects in learning tasks involving prose passages. Brown and Sellin (1967) investigated whole versus part presentation methods and meaningful activities versus non- meaningful activities in learning poetry. Meaningful activities were 27 defined as use of colorful pictures, recordings, discussion, slides, and drawings after oral presentation of the poem. Nonmeaningful activities were defined as a limited discussion after oral presenta- tion of the poem. All four possible treatment groups were given 10 minutes of practice per day over a three -week period. The dependent variable was individual pupil errors during recitation one week after treatment. Whole presentation was found to be better than part presentation and whole —meaningful was the best combination. Neisworth, Smith and Deno (1968) investigated the use of advanced organizers in facilitating the learning and retention of written prose materials by educable mental retardates and normals of comparable reading achievement. The learning passage was a 1200—word passage on sound. The advanced organizer was a 200- word conceptual passage which was pertinent tothe learning passage. As such, it was an overview of sound and did not contain any infor- mation found in the learning passage. A 200 -word control introductory passage was also used which contained examples of sound. Both subject populations contained experimental and control groups. All subjects received the advanced organizer or the control introductory passage for four days. The learning passage was also administered to all subjects on the third and fourth days. On the fourth day, a 28 30-item criterion test was given to all subjects following their appropriate introductory passage and the learning passage. The same test was administered to all subjects on the eighteenth day. There were no significant main effects due to subject status,. treat- ment or retention time. However, a significant interaction between treatment and retention time occurred. This result indicated that organizer groups had higher initial retention scores than delayed retention scores. Whereas, the control groups performed better on the delayed retention test than on the initial one. A second -order interaction between treatment, subject status and retention was also significant. This result indicated that normal organizer and control groups had lower delayed retention scores than initial retention scores. However, the educable mentally retarded control group had higher delayed retention scores than initial retention scores. The normal organizer groups performed significantly better than the normal control groups on both retention measures. There were no significant differences between the educable mentally retarded organizer and control groups on either retentiOn test. It may be that repetition in short recall tasks and prose tasks does not facilitate the learning of such tasks by mentally retarded subjects. 29 Definition of terms (The studies of Frase and Rothkopf have used the term "retention" to refer to the immediate assessment of the recall of factual information from written prose materials. Inaddition, none of the studies investigating influential variables in learning from written prose materials have included a recall measure of relevant information fOllowing a specified length of time after the first reten- tion test. The author prefers the terms "initial retention" and ”delayed retention" in order to distinguish between these two reten- tion measures in the present study. They are defined as follows: "initial retention" refers to the assessment of the recall of factual information immediately after presentation of the prose materials, and "delayed retention" refers to the assessment of the recall of factual information one week after presentation of the prose mate - rials. Delayed retention as well as initial retention of factual information from prose materialswould seem to be of importance in the learning process. It may be that the same variables which influence the initial retention of factual information from prose materials also have an effect upon the delayed retention of that information. To be more precise, the use of questions after each paragraph and/or a procedure designed to gain attention may 30 facilitate the delayed retention of relevant factual information. For these reasons, a delayed retention measure will be included in the present study in order to assess the effects of question location and) a signal designed to gain attention on delayed retention as well as initial retention of factual information from aurally presented prose materials. Hypotheses Based upon the importance of attention in retardate learning and previous research which has investigated the effect of question location and question pacing on the facilitation of the retention of written prose materials, it is hypothesized that: Hypothesis 1: Questions after each paragraph will facilitate initial retention of relevant factual information in a prose passage presented aurally to trainable mentally retarded subjects when compared to a condition where questions precede each paragraph or no questions accompany each paragraph. Hypothesis 2: Questions after each paragraph will facilitate initial retention of incidental factual information in an aurally presented prose passage when compared to Hypothesis 3: Hypothesis 4: Hypothesis 5: Hypothesis 6: 31 a condition where questions precede each paragraph or no questions accompany each paragraph. Questions after each paragraph will facilitate delayed retention of relevant factual information in an aurally presented prose passage when compared to a condi- tion where questions precede each paragraph or no questions accompany each paragraph. A procedure designed to gain attention will facilitate initial retention of relevant factual information when compared to conditions where the procedure is not present Questions before each paragraph will facilitate initial retention of relevant factual information when com - pared to a condition where no questions accompany each paragraph. The most facilitative condition for initial retention of relevant factual information will be combined treatment of questions aftereach paragraph and a procedure designed to gain attention. Hypothesis 7: Hypothesis 8: Hypothesis 9: 32 A combined treatment of questions after each paragraph and a procedure designed to gain atten- tion will facilitate the initial retention of incidental factual information when compared to a condition where no questions accompany each paragraph, with or without a procedure designed to gain attention, or a condition where questions precede each para- graph, with Or-without a procedure designed to gain attention. The most facilitative condition for delayed retention of relevant factual information will be a combined treatment of questions after each paragraph and a procedure designed to gain attention. A combined treatment of questions before each para- graph and a procedure designed to gain attention will facilitate initial retention of relevant factual informa - tion when compared to a condition where no questions accompany each paragraph, with or without a pro- cedure designed to gain attention, or a condition where questions precede each paragraph. 33 Hypothesis 10: Retention of relevant factual information will be better than retention of incidental factual informa - tion. CHAPTER III DESIGN OF THE STUDY Sample The sample consisted of 72 male and female traihable mentally retarded students enrolled in the secondary program of a midwestern training center for trainable mental retardates. The training center was a nonresidential school in session the five days of the normal school week. The secondary program of the center provided classes in art, music, physical education, homemaking, vocational skills, and learning skills for the trainable mentally retarded student. Chronological ages of the students in this sample ranged from 11 -8 years to 21 -6 years. Intelligence quotients of these _S_s ranged from 30 to. 63 on the Stanford -Binet Intelligence Scale, Form L-M, 45 to 56 on the Wechsler Intelligence Scale for Children, and 43 to 61 on the Wechsler Adult Intelligence Scale. Individual chronological ages and intelligence quotients for each treatment grouprare presented in Appendix A. Table 1 presents the means and 34 35 standard deviations of the chronological ages and intelligence quotients for each treatment group. Table 1 Means and Standard Deviations Of the Chronological Ages and Intelligence Quotients of the Six Treatment Groups CA IQ Treatment Group Mean SD Mean SD I (P) 17.47 3.05 42.50 5.85 II (P*).. 17.06 2.45 43.42 9.48 III (PQ) 15.07 1.76 46.17 8.73 IV (PQ*) 16.29 2.20 47.33 6.62 V (QP) 15.65 1.99 45.67 5.41 VI (QP*) 15.44 2. 55 48. 00 6. 40 NO _S_with an uncorrected hearing defect was included in the sample. Design and experimental treatments A 2 X 3 X 5 X 2 factorial design with repeated measures on the last two factors was used. The first factor, 'cueing, consisted of two levels-—no cueing and cueing. The second factor was question 36 location and consisted of three levels —-no questions, questions after, and questions before. The third factor represented five administra- tions of the same criterion test. The levels of the last factor were relevant and incidental information. Figure 1 presents a schemata of the data matrix showing the four independent variables. Six presentation methods resulted from combinations Of the levels of the first two factors. The six presentation methods included one control condition and five experimental treatment con- ditions. The five eXperimental treatment conditions involved the manipulation Of relevant questions and/or a procedure designed to gain the attention of the _S_. This procedure consisted of the E grasp- ing the S' s shoulder and saying, "Now listen.‘ A previous pilot study had used taped stories for presentation and had used an inserted voice on the tape which said, "Now listen" in the appropriate places. The pilot study had indicated that the simple insertions of "now listen" on the tape were not a prominent enough means of gaining the attention of the §S. It was also felt that the procedure used in the pilot study was not classroom related because it did not typify teacher behavior. Therefore, the present procedure to gain the _S_' s attention was devised. It was thought to be classroom related, allowed for eye contact with the S, and was felt to be an effective means of gain- ing the S' s attention. 37 Data Matrix of the 2 X 3 X 5 X 2 Repeated Measures Design T 83 2 R2 I2 1 NQ 12 1 3 NC QA 24 2 5 QB 36 37 NQ 48 49 C QA 60 . 61 QB 72 Figure 1 38 The six presentation methods of the present study were as follows: Treatment I: The control condition in which each of the 10 paragraphs of the test passage was presented without the use of relevant questions or a grasp of the shoulder and the words, ”Now listen. " (P) Treatment II: A treatment condition in which the sentence(s) containing an answer(s) to the relevant question(s) for each paragraph was (were) preceded by a grasp of the shoulder and the words, "Now listen. " (P*) Treatment 111: A treatment condition in which the relevant question(s) for each paragraph was (were) read to the 83 immediately after each paragraph. (PQ) Treatment IV: A treatment condition in which the relevant question(s) for each paragraph was (were) read to the SS immediately after each paragraph and a grasp of the shoulder and the words, "Now listen" preceded the sentence(s) containing an answer(s) to the relevant question(s) for each paragraph. (PQ*) Treatment V: A treatment condition in which the relevant question(s) for each paragraph was (were) read to the 88 immediately before each paragraph. (QP) 39 Treatment VI: A treatment condition in which the relevant question(s) for each paragraph was (were) read to the _S_s immediately before each paragraph and a grasp of the shoulder and the words, ”Now listen" preceded the sentence(s) containing an answer(s) to the relevant question(s) for each paragraph. (QP*) Materials Three pre —passages and one test passage were used in this study. The three pre -passages were taken from Getting the Facts: Specific Skill Series, Book A by Richard A. Boning. All pre -passages were on a first grade reading level. The three pre -passages are presented in Appendix B. The test passage was constructed by the E. The test passage is presented in Appendix C. The three pre -passages consisted of 148, 150, and 157 words, respectively. Each of the three pre -passages contained five para- graphs and information to answer five factual three -foil multiple - choice questions found in the listening skills booklet. The five ques- tions for each pre -passage are presented in Appendix D. The test passage consisted of 317 words. The test passage contained 10 distinct paragraphs which differed in the factual informa- tion found in each of them. The number of words per paragraph ranged from 24 to 46 words, with a mean word length of 31. 7 words. 40 The test passage concerned a family who went on a one -day trip to the beach. In general, the test passage related the family's activi- ties in preparation for the trip, on their way to the beach, while at the beach, and on their way home from the beach. In particular, the test passage described the activities and responsibilities of each family member during the course of the events preceding, during and following the time spent at the beach. Thirty three -foil multiple -choice questions were constructed which pertained to the information found in the test passage. Three questions were constructed from the information found in each para - graph. Fifteen of the 30 questions were designated as relevant ques- tions. These questions were heard during passage presentation by some of the SS depending on the experimental treatment group to which _S_s had been randomly assigned. Each paragraph contained information to answer one or two Of the relevant questions. The other question or questions in each paragraph were incidental ques - tions. The incidental questions were not heard by any _S_ during passage presentation. Each paragraph contained information to answer one or two of the incidental questions. The 15 relevant and 15 incidental questions were used as the criterion test. The order of the relevant and incidental questions on the criterion test was sequential to the order of the information in the test passage which 41 answered each question. The criterion test is presented in Appendix E. The criterion test was piloted in order to ascertain if relevant and incidental questions were of equal difficulty. Ten _S_s were administered the control condition (P) and the condition pro- posed as the most facilitative (PQ*). Attest was performed on the reported item difficulties for the control condition (P) fOr the first administration of the criterion test. This analysis showed that relevant and incidental items were of equal difficulty (_t_ = -1. 3057, 28 93’ p > . 05). Table 2 presents the results of an item analysis on the three administrations of the criterion test for the control condi- tion (P) . Table 2’ Results of Item Analyses on the Pilot of the Criterion Test for Group I (P) Order of Administration Ite'm Statistics 1 2 3 Mean item difficulty . 48 . 48 . 41 Mean item discrimination . 48 . 48 . 45 Mean point biserial correlation - . 44 . 42 . 39 Kuder-Richardson #20 .83 .85 . 83 Standard error of measurement 2. 37 2. 27 2. 16 N = logs in Group I (P) 42 Table 3 presents the results of an item analysis on the three administrations of the criterion test for the condition proposed as the most facilitative (PQ*). Table 3 Results of Item Analyses on the Pilot of the Criterion Test for Group IV (PQ*) Order of Administration Item Statistics 1 2 3 Mean item difficulty . 43 . 71 . 38 Mean item discrimination . 58 . 05 . 63 Mean point biserial correlation . 47 . 32 . 51 Kuder -Richardson #20 . 90 . 95 . 93 Standard error of measurement 2. 16 1. 59 2. 03 N = 10 Ss in Group IV (PQ*) Procedure Seventy -seven students enrolled in the secondary program of the training center were given the three pre -passages. Two S3 with uncorrected hearing defects were eliminated from the study. Seventy- two 53 were then randomly selected from the remaining 75 students. These 72 SS were then randomly assigned to experimental treatment groups for administration of the test passage. 43 Pre -passages. --Each of the pre -passages was read to small groups of four to logs on each of three consecutive days. The}; read the pre -passages. The same order of passage presentation was maintained for all groups. One pre -passage was presented three times to each group on each of the three consecutive days. Subjects were asked the same five three -foil multiple -choice questions immediately following each presentation of the pre -passage. The E read the question and the three choices for each question. The E made certain all _S_s had responded before reading the next question and its three choices. Subjects circled their answer to each Of the five questions for the pre -passage on an answer sheet which contained the three choices for each question. Subjects were clOsely monitored by the _E_3, classroom teacher, and anaide to make sure the verbalized answers of the _S_s coincided with their circled answers. The answer sheets were distributed immediately following each presentation of the pre -passage and collected immediately after the completion of each testing situation. Question order'was maintained for each test administration which occurred after each presentation of the pre -passage. How— ever, the order of the three -foil answers on the answer sheet was uniformly different for allfis for each presentation of a question. 44 This was done in an attempt to eliminate any response set tendency due to position cues. The different orders Of the three -foi1 answers to the five questions for the first, second, and third pre -passage are presented in Appendix F, G, and H, respectively. The following instructions were read to the _S_s by the E immediately before the first presentation of a pre -passage on any particular day: Today I am going to read you a story. I will read the story three times. After I read the story each time, I am going to ask you some questions about the story. Listen carefully to the story so you can answer the questions. Now, I will read the story for the first time. The E said, "Now, I' 11 read the story again" immediately before the second and third presentations of the pre -passage. No knowledge of results was given. Vocabulary checklist. --During the two weeks required for presentation of the three pre -passages, the 77 students enrolled in the secondary program of the training center were individually administered a vocabulary checklist. This was done in order to identify any students who were nOt familiar with the vocabulary con- tent of the test passage. The E asked each student a series of questions which dealt with the conceptual meaning of the vocabulary content of the test passage. The E subjectively ascertained from the verbal reports of the students whether they did nor did not understand 45 the meaning of particular words within the test passage. If the _S_ could give a definitive explanation and/ or exemplars of the vocabu- lary concepts of interest, the E concluded that the E knew the con- ceptual meaning of the vocabulary content of the test passage. No student was eliminated from the study on the basis of his or her verbal report to the vocabulary checklist. The vocabulary checklist is presented in Appendix I. Test passag_e_. - - Two days after all Es had been given the three pre -passages, the administration of the test passage began. The test passage was individually administered to all Es in one of two small rooms which permitted the elimination Of extraneous noise and/ or interruptions. Each E was brought into the testing room by the E and read the appropriate instructions for the particular treatment condition to which he or she had been randomly assigned. The six sets of instructions are presented in Appendix J. The test passage was read three times to each _S_. After each presentation of the test passage the same criterion test was administered to all Es. The E read each question, read the three answers for each question while pointing to them, and circled the reply of the E. Question order after each test passage presentation 46 was the same for all Es. However, the order of the three -foil answers on the answer sheet was uniformly different for all Es for each question presentation. This was done in an attempt to eliminate any response set tendency due to position cues. The orders of the answers for the three initial administrations of the criterion test questions are presented in Appendix K. The following instructions were read to each _S immediately after the first reading of the test passage: O.K. (child' 5 name), now I'm going to read you the questions. (Child' 8 name), each time you get a right answer I'm going to give you a piece of candy. We'll see how many pieces of candy you can get. The criterion test was then administered for the first time. Immediately following the first administration Of the criterion test the E said, "Now (child' 8 name), I'm going to read the story again. Are you ready?" The test passage was then read for the second time. After the second reading of the test passage the E said, "O.K. (child' s name), now I'm going to ask you the questions again" and administered the criterion test for the second time. Following this, the E said, "O.K. (child's name), now we're going to listen to the story one more time. Ready?" The test passage was then read by the E for the third and final time. After the third presentation of the test passage the E said, "O.K. (child' 3 name), now I'm going 47 to ask you the questions one more time" and administered the criterion test for the third and final time. Knowledge of results was given during the three initial administrations of the criterion test. Each time a _S_ correctly answered a question, the E gave him an M & M and said, "That' 3 right. " Two delayed retention tests were individually administered to each E one week following the three‘initial presentations of the test passage and criterion tests. Each _S_was brought by the E into one of two small testing rooms and read the following instructions: (Child' 8 name), remember the story I read to you last week? It was about a family who went to the beach. They did many things while they were at the beach. The family stayed at the beach all day and got home when it was dark. Today I am going to ask you the same questions about the story that I asked you last week. Are you ready? O.K. , here we go. Immediately following the instructions the first delayed retention test was administered. The original criterion test was used. The E read each question, read the three answers for each question while pointing to them, and circled the reply of the E. Question order was the same as that used in the original criterion test. How - ever, the three -foil answers for each question were in a different order than any used during the initial presentations Of the criterion test. This order was uniform for all ES and was done in an attempt to eliminate any response set tendency due to position cues. 48 The E then said: Now (child' 3 name), I am going to read the story to you again. After I read the story, I am going to ask you to answer the questions one more time. Listen carefully to the story so you can answer the questions. TheE read the test passage one time via the standard presentation condition (P). The second delayed retention test was administered immediately after this single delayed presentation of the test pas - sage. The original criterion test was used under the same procedure as in the administration of the first delayed retention test. Question order remained the same as in the original criterion test. The three -foil answers for each question were in a different order than any used during the initial presentations or the first delayed presen- tation of the criterion test. This order was uniform for all Es and attempted to eliminate any response set tendency due to position cues. The orders of the answers for the two delayed administra- tions of the criterion test questions are presented in Appendix L. No knowledge of results was given on the first or second delayed retention test. CHAPTER IV RESULTS OF THE ANALYSES Main Analysis The main analysis was an analysis of covariance of the 2 X 3 X 5 X 2 design with repeated measures on the last two factors. Cueing, the first factor, consisted of cueing and no cueing. The second factor was question location and consisted of no questions, questions after, and questions before. The third factor was time and represented five administrations of the same criterion test. The levels of the fourth factor were relevant and incidental informa- tion. The decision to use analysis of covariance instead of analysis of variance was based upon (1) correlations between the covariate and dependent variables and (2) the result of a test of the assumption of homogeneity of within -group regression coefficients. Table 4 presents the correlations between the covariate and each of the 10 dependent variables for each of the six treatment groups. The covariate was the total number of correct answers on the three pre -passages. The 10 dependent variables were the 49 50 number of correct answers to relevant and incidental questions on each administration of the criterion test. The covariate and 10 dependent variables for eachE in each of the six treatment groups are presented in Appendix M. Table 4 Correlations Between the Covariate and Number Correct Relevant and Incidental Questions on Each Administration of the Criterion Test for the Six Treatment Groups Time Treatment T1 T2 T3 T4 T5 Group _ R I R I R I R I R I 1(P) .65 .74 .69 .79 .75 .82 .56 .78 1.68 .73 II(P*) .39 .46 .61 .55 .52 .54 .53 .62 .59 .63 III(PQ) .56 .69 .44 .67 .47 .49 .53 .54 .51 .54 IV(PQ*) .41 .37 .47 .50 .41 .57 .57 .54 .48 .49 V(QP) .63 .75 .955 .73 .60 .74 .69 .74 .58 .80 VI(QP*) .60 .65 .61 .68 .50 .71 .65 .66 .49 .70 R Relevant questions I Irrelevant questions As can be seen from the data in Table 4, the correlations between the covariate and each of the 10 dependent variables tend to 51 be high. The correlations of . 50 or above are significantly greater than zero. However, none of the 60 correlations are < .30: the recommended absolute value of the correlation coefficient in order to increase precision by means of analysis of covariance (Elashoff, 1969, p. 386). The mean correlations for treatment groups I, II, III, IV, V, and VI are .72, .54, .48, .68, and .62, respectively. In order to test the assumption of homogeneity of within- group regression coefficients, the procedure discussed by Kirk (1968, pp. 469-470) was used. The E test was not significant (E = .0146, 5/60, p > .25). This result indicated that the within- group regression coefficients for the six treatment groups were homogeneous. The means and standard deviations of the number of correct answers to relevant and incidental questions for each treatment group at each administration of the criterion test are presented in Table 5. Figure 2 shows the mean number of correct answers to relevant questions for each treatment group at each administration of the criterion test. Figure 3 shows the mean number of correct answers to incidental questions for each treat- ment group at each administration of the criterion test. Mean Number of Correct Answers for Relevant and Incidental Questions at Each Administration of the Criterion Test for Each of the Six Treatment Groups 52 Table 5 Order of Type Treatment Group Adminis- ‘ of ' tration Information P P* PQ PQ* QP QP* T1 Mean 10. 00 8. 92 10. 75 10. 00 11.00 ' 10.08 SD 3.34 2.29 2.83 2.74 2.74 3.12 T1 Mean 9.08 7.75 9.00 8.50 8.75 7.58 SD 4.09 2.68 3.56 3.50 2.80 3.20 T2 Mean 9. 92 9.42 12.08 10.33 11.83 10.42 SD 3.17 3.40 3.25 3.30 3.44 3.04 T2 Mean 9.58 8.75 10.50 8.75 10.25 8.83 SD 3.95 2.05 3.62 2.98 3.32 3.48 T3 Mean 10. 58 8.17 12.08 10.75 12.00 10.83 SD 3.75 3.53 2.90 3.22 3.14 2.91 T3 Mean 10.42 8.08 10.92 10.00 11.17 10.08 SD. 3.38 3.86 3.10 2.58 3.80 2.66 T4 Mean 10.50 8.92 11.33 9.83 11.83 9.92 SD 3.48 3.82 3.68 3.18 2.67 3.33 T4 Mean 10.42 8.58 9.83 8.50 10.67 8.67 SD 3.57 3.52 3.67 2.78 3.52 3.35 T5 Mean 11.17 8.83 11.83 11.08 12.08 10.42 SD 3.69 4.24 3.31 2.87 3.30 3.28 T5 Mean 10.83 8.00 10.33 9.58 10.58 9.58 SD 3.51 3.70 3.97 3.15 4.13 3.01 N = 12 in each treatment group 53 12.0 _ QP PQ P 11.0 t. A PQ, I , A / QP* *5 810.0 _ L. O U La 0) .Q g 9.. _ IN ‘0 13* 8.0 — 0—. P H PQ A—A QP 7.0 - O—O P* H PQ* H 013* \\ l l l I 1 T1 T2 T3 _ T4 T5 Figure 2 Mean Number Correct Relevant Questions for Each Treatment Group at Each Administration of the Criterion Test K Number Correct 54 //R \\ /.P / \ *rt‘QP r/ I’D-y /IPQ // / / // / / \ ‘I/ // _ // \\ PQ* g: PQ* /// // \\ // // \\ // / \\ // at >87 0/ // \\ / \ // \ / \ // \0/ \OP* / A/ .-—-. P '—-. PQ ‘4 QP O—-O P* 0-D PQ* A-A QP* / / 1 l l l 1 T1 T2 T3 T4 T5 Figure3 Mean Number Correct Incidental Questions for Each Treatment Group at Each Administration of the Criterion Test 55 A 2 X 3 X 5 X 2 analysis of covariance with repeated measures on the last two factors was performed on the data summarized in Table 5. The four independent variables and the levels of each were cueing (no cueing and cueing); question location (no questions, questions after and questions before); time (five administrations of the same criterion test) and information (relevant and incidental). Only one covariable was used and, therefore, only between Es sources of variation were adjusted for the covariable (See Kirk, 1968, pp. 482 - 485 and Winer, 1962, p. 607). The results of this analysis of covariance are presented in Table 6. The results of the 2 X 3 X 5 X 2 analysis of covariance with repeated measures on the last two factors indicated that the main effect of Time, the main effect of Information, and T X I were sig - nificant sources of variation when the Geisser-Greenhouse conserva- tive E test (See Kirk, 1968, pp. 262 -263) was applied to the data presented in Table 6. The four independent variables were cueing (no cueing and cueing); question location (no questions, questions after and questions before); time (five administrations of the same criterion test) and information (relevant and incidental). Therefore, only these three sources of variation were considered for post hoc investigations following the over-all main analysis. 56 Table 6 Summary Table for 2 X 3 X 5 X 2 Analysis of Covariance with Repeated Measures on the Last Two Factors Source . SS df MS F Between Subjects C (adj.) 339.172 1 339.172 3. 48 Q (adj.) 129. 467 2 64. 734 . 66 ‘CQ (adj.) 10.223 2 5.112 .05 S:CQ (adj.) 6342.217 65 97. 573 Within Subjects T 119.690 4 29.920 13.42* CT 12.270 4 3.070 1. 38 QT 34. 690 8 4. 340 1. 95* CQT 34. 870 8 4. 360 1. 96* S: TCQ 588. 280 264 2.230 I 222.220 1 222.220 36. 79* CI 0.000 1 0.000 0.00 QI 34.540 2 17.270 2. 86 CQI 1. 460 2 .730 .12 S:ICQ 398. 380 66 6.040 TI 21. 650 4 5. 410 4. 36* CTI .570 4 . .140 .11 QTI 5. 420 8 .680 .55 CQTI 1.810 8 .230 .19 S:TICQ 327. 950 264 1. 240 * p < . 05 Note: Only T, I, and T X I were significant (p < .05) using the Geisser-Greenhouse conservative E test. See Kirk (1968, pp. 262 -263). 57 Hypothesis 10 stated that retention of relevant information would be better than retention of incidental information. The mean number of correct answers for relevant questions was 10. 564 and the mean number of correct answers for incidental questions was 9. 453. Therefore, hypothesis 10 was supported. The retention of relevant information was better than the retention of incidental information. The significant main effect of Time was investigated by means of pairwise and complex Scheffé contrasts of the number of correct questions at each administration of the criterion test. The degrees of freedom for the Geisser-Greenhouse conservative Etest were used for all Scheffé contrasts involving the main effect of Time. The pairwise Scheffé contrasts revealed two significant results. More questions were answered correctly on both the third adminis - tration and the fifth administration of the criterion test than on the first administration. There were no other significant pairwise differences. Table 7 presents these results. A complex Scheffé contrast of initial (T1 + T2 + T3) versus delayed (T4 + T5) retention was performed. The mean total number correct for eachadministration of the criterion test are presented in Table 7. The average of the mean total number correct for initial retention was 9. 92 and 10. 14 for delayed retention. The average mean difference between initial and delayed retention (-. 2167) was 58 not significant (critical value = . 6406, 1/66 d_f, p > .05). This result indicated that there was no difference between initial and delayed retention. Table 7 Matrix of Differences in Mean Total Number Correct Questions Between Administrations of the Criterion Test T1 T4 T2 T5 T3 T1 = 9.29 ---- .63 .77 1.07* 1.13* T4 = 9.92 ---- .14 .44 .50 T2 = 19.06 ---- .30 .36 T5 = 10.36 --—- .06 T3 = 10.42 _-__ Critical value at .05 level (1, 66) = . 99 Figure 4 shows the interaction between Time and Information. The means and standard deviations of the number of correct answers for relevant and incidental questions for each administration of the criterion test are presented in Table 8. Scheffé contrasts were used in order to investigate pairwise differences in relevant information between each administration of the criterion test. The MS error term for these contrasts (S:TCQ) X Number Correct 59 11.0 .. 10.0 - /A\ / \ / \ /’ / \ // / \ / / v / / 9.0 u / / / / / J R 8.0 -— ____ I /9 ‘ L l I l 1 T1 T2 T3 T4 T5 Figure 4 Means for Relevant and Incidental Information Across All Treatment Groups T X I 60 Table 8 Mean Number of Correct Answers for Relevant and Incidental Questions for Each Administration of the Criterion Test Order of Type of Information Administration Relevant Incidental T1 Mean 10.13 8.44 SD 2. 92 3. 40 T2 Mean 10. 67 9. 44 SD 3. 40 3. 38 T3 Mean 10. 74 10.11 SD 3. 49 3. 42 T4 Mean 10. 39 9. 44 SD 3. 52 3. 54 T5 Mean 10.90 9. 82 SD 5. 26 3. 71 N = 72 at each level of Time was obtained from a 2 X 3 X 5 analysis of variance of the number of correct answers to relevant questions at each administration of the criterion test. The appropriate degrees of freedom were those for the Geisser-Greenhouse conservative E test. No significant pairwise differences were found in relevant information between each admin - istration of the criterion test. Table 9 presents these results. 61 Table 9 Matrix of Differences in Mean Number Correct Relevant Questions Between Administrations of the Criterion Test TlR T4R T2R T3R T5R T1R = 10.13 --- .26 .54 .61 .77 T4R = 10.39 --- .28 .35 .51 T2R = 10.67 --- .07 .23 T3R = 10.74 --- .16 T5R = 10.90 -__ Critical value at . 05 level (1, 66) = . 86 A 2 X 3 X 5 analysis of variance of the number of correct answers to incidental questions at each administration of the criterion test was also performed. The appropriate MS error term from this 2 X 3 X 5 analysis of variance (S:TCQ) was used for Scheffé contrasts of pairwise differences in incidental information between each administration of the criterion test. The degrees of freedom were those from the Geisser-Greenhouse conservative E test. The Scheffé contrasts of differences in the retention of incidental informa - tion between administrations of the criterion test revealed four significant results. More incidental questions were answered correctly at each of the second, third, fourth, and fifth administrations 62 of the criterion test than on the first administration. These results are presented in Table 10. Table 1 0 Matrix of Differences in Mean Number Correct Incidental Questions Between Administrations of the Criterion Test T11 T21 T41 T51 T31 T11: 8.44 ---- 1.00* 1.00* 1.38* 1.67* T21 = 9.44 —--- 0.00 .38 .67 T41: 9.44 ---- .38 .67 T51: 9.82 ---- .29 T31 = 10.11 --__ Critical value at . 05 level (1, 66) = .89 Differences between relevant and incidental information at each administration of the criterion test were investigated by means of a 2 X 3 X 10 analysis of variance and subsequent Scheffé contrasts of pairwise differences. The appropriate MS error (S:TCQ) from the 2 X 3 X 10 analysis of variance of correct answers to relevant and incidental questions at each administration of the criterion test was used for the Scheffé contrasts. The degrees of freedom were those for the Geisser-Greenhouse conservativeE test. The Scheffé 63 contrasts indicated that significantly more relevant than incidental questions were answered correctly on the first administration of the criterion test. There were no significant differences between the number of correct relevant and incidental questions at any other administration of the criterion test. Table 11 presents these results. Exact Hypotheses Hypotheses one through nine stated expected differences between treatment groups in terms of the retention of initial —relevant, delayed -relevant, and initial -incidenta1 information. However, the complex Scheffé contrast of initial versus delayed retention indicated that the difference between initial and delayed retention was not sig - nificant. Therefore, those hypotheses which proposed differences between the same treatment groups, but in terms of initial -relevant retention for one hypothesis and delayed -relevant retention for the other hypothesis, were now considered as one hypothesis. Hypothesis one (initial -relevant retention) and hypothesis three (delayed -relevant retention) stated that treatment groups receiving questions after each paragraph‘would retain more than treatment groups receiving ques - tions before each paragraph or treatment groups receiving no questions. Hypothesis six (initial -relevant retention) and hypothesis 64 av A .I. 8m .3 7:3 mo . am m3m> 303:0 IIII omdfi u mosh ofi . IIII «$4.: 0 mmrfi mm. ho. IIII bag: 0 GNP Hm. mm. mm. IIII mmdfi u my? 5.. HQ. mm. mm. IIII mag: u .mrh mu. mm. mm. mm. No. IIII :dH 0 SH. moé mm. mm. pm. Hm. mm. IIII mad u EH. ové omé «NA mm. mm. pm. mm. IIII wed u HRH. wvé omé mmé mm. mm. hm. mm. cod IIII vwd u HNB *mv.m *om.m *mm.m *mmé *moé *bmé mmé ooé ooé IIII vvd u :5 mmh. mmB MNH. Maw? .mHE 58 SH. :48 gm. :8 EOE ~833qu o5 mo :ofimuumwfiacaw £08m an mcofimmna 3.28305 can ESE—cm pomuuou awn—552 G852 cmmgmm noodoummflfl mo 5.332 H H 39mg... 65 eight (delayed -relevant retention) stated that the most facilitative condition‘would be treatment group IV (PQ*). Hypotheses one and three were then considered as one hypothesis (1&3) and hypotheses six and eight were then considered as one hypothesis (6&8). Table 12 presents a summarized restatement of hypotheses one through nine in terms of retention of relevant and incidental informa- tion. Table 12 Summarized Restatement of Hypotheses One Through Nine Type of Information Relevant Incidental H1&3: PQ+PQ*>QP+QP*+P+P* H2: PQ+PQ*>QP+QP*+P+P* H4 P*+PQ*+QP*>PQ+P+QP H7: PQ°¥>P+P*+QP+QP* H5 QP+QP*>P+P* H6&8: PQ*>P+P*+PQ+QP+QP* H9 QP*>P+P*+QP In order to investigate the restated hypotheses in Table 12, the original 10 dependent variables were collapsed to form two new dependent variables. The two new dependent variables were the total number correct relevant questions and the total number correct 66 incidental questions. The means and standard deviations for the total number correct relevant and incidental questions for each treatment group are presented in Table 13. Table 1 3 Mean Total Number Correct Relevant and Incidental Questions for Each Treatment Group Relevant Incidental Treatment Group Mean SD Mean SD 1(P) 52.17 16.64 50.33 17.86 II(P*) 44.25 16.43 41.17 14.53 III(PQ) 58.08 15.40 50.58 17.26 IV (PQ*) 52. 00 14.21 45. 33 13.83 V (QP) 58.75 14.29 51.42 16.85 VI (QP*) 51.67 14.68 44.75 14.77 Hypotheses one and three, four, five, six and eight, and nine were investigated by means of a 1 X 6 analysis of covariance of the total number correct relevant questions for each treatment group and subsequent complex Scheffé contrasts. The results of the 1 X 6 analysis of covariance for relevant information are presented in Table 14. 67 Table 14 1 X 6 Analysis of Covariance Summary Table for Total Number Correct Relevant Questions for Each Treatment Group Source SS df MS F Between groups, adj. 1080.20 5 216.04 1.26 Within groups, adj. 11133.20 65 171.28 Total, adj. 2213.40 70 3.05; 5,65=2.36 These results indicated that there were no between -group differences in the total number of correct relevant questions. Nevertheless, complex Scheffé contrasts were used to investigate the specific informationcontained in the exact hypotheses. The weighted Scheffé contrasts for hypotheses one and three, four, five, six and eight, and nine are presented in Table 15. These weights were assigned so that the sum of the weights would equal zero for each of the hypotheses one and three, four, five, six and eight, and nine. Hypotheses one and three stated that treatment groups receiving questions after each paragraph would score. higher on retention of relevant questions than treatment groups receiving 68 questions before each paragraph or groups not receiving questions. The average adjusted mean difference of the Scheffé contrast for hypotheses one and three (3. 79) was not significant (critical value = 11.24, 5/65 if, p > . 05). This result indicated that post-questions did not facilitate the retention of relevant information when compared to pre -questions or no questions. Hypotheses one and three were not supported . Table 1 5 Weighted Scheffé Contrasts for Hypotheses Concerned With Retention of Relevant Information Hypothesis _ Weighted Scheffé Contrast 1 1 1 1 1 1 1&3 [(+2—) PQ+ (+2—) PQ*] - [ (-Z)QP+ (--4—)QP*+ (-;1-)P+ (- Z)P*] 4 I (+%-)P* + (+%-)PQ*+ I+§IQP=I=I - [ (- ng+ (- §IPQ+ (- .13—)QP] 1 1 1 1 5 [(+2-lQP+(+2—)QP*I '[(-2-)P+(-§')P*] 6&8 (+1)PQ* -[(-«§1)-)P+(-;—)P*+(-%—)PQ+(-%)QP+(-;—)QP*] 9 (+1)QP* - [(-;—)P+(-%)P*+(-%)QP] Hypothesis four stated that treatment groups who were given a procedure designed to gain attention-would score higher on reten- tion of relevant questions than treatment groups who were not given a procedure designed to gain attention. The average adjusted mean 69 difference of the Scheffé contrast for hypothesis four (-5. 36) was not significant (critical value = 10. 64, 5/65 E, p > .05). This result indicated that the procedure designed to gain attention did not facilitate the retention of relevant information. Hypothesis four was not supported. Hypothesis five stated that treatment groups who received questions before each paragraph would score higher on retention of relevant questions than treatment groups who did not receive ques - tions. The average adjusted mean difference of the Scheffé contrast for hypothesis five (5. 12) was not significant (critical value = 13.03, 5/65 _di, p > .05). This result indicated that pre -questions did not facilitate the retention of relevant information when compared to no questions. Hypotheses six and eight stated that the group who received the combined treatment of questions after each paragraph and a procedure designed to gain attention would score higher on retention of relevant questions than any other treatment group. The average adjusted mean difference of the Scheffé contrast for hypotheses six and eight (.437) was not significant (critical value = 14.24, 5/65£i_f, p > .05). This result indicated that the combination of post-questions and a procedure designed to gain attention was not the most facilitative condition for retention of relevant information. Hypotheses six and eight were not supported. 70 Hypothesis nine stated that the group who received the combined treatment of questions before each paragraph and a pro- cedure designed to gain attention would score higher on retention of relevant questions than treatment groups who only received questions before each paragraph or no questions, with or without the procedure designed to gain attention. The average adjusted mean difference of the Scheffé contrast for hypothesis nine (-1. 31) was not significant (critical value = 15.00, 5/65 21;, p > , 05). This result indicated that the combination of pre -questions and a procedure designed to gain attention did not facilitate the retention of relevant information when compared to pre -questions only or no questions regardless of whether the procedure involved an attempt to gain _S_s attention. Hypotheses two and seven were investigated by means of a 1 X 6 analysis of covariance of the total number correct incidental questions for each treatment group and subsequent complex Scheffé contrasts. Hypothesis two stated that treatment groups who received questions after each paragraph would score higher on retention of incidental questions than treatment groups who received questions before each paragraph or did not receive questions. Hypothesis seven stated that the group who received the combined treatment of questions after each paragraph and a procedure designed to gain attention would score higher on retention of incidental questions 71 than treatment groups who received questions before each paragraph, with or without a procedure designed to gain attention, or treatment groups who received no questions, with or without a procedure designed to gain attention. The results ef the 1 X 6 analysis of covariance for incidental information is presented in Table 16. Table 1 6 1 X 6 Analysis of Covariance Summary Table for Total Number Correct Incidental Questions for Each Treatment Group Source SS df MS F Between groups, adj. 525. 55 5 105.11 . .71 Within groups, adj. 9685. 65 65 149. 01 Total, adj. 10211.20 70 505; 5,65 = 2.36 These results indicated that there were no between group differences in the total number of correct incidental questions. Nevertheless, complex Scheffé contrasts were used to investigate the specific information contained in the exact hypotheses. The weighted Scheffé contrasts for hypotheses two and seven are presented in Table 17. These weights were assigned so that the sum of the weights would equal zero for each of hypotheses two and seven. 72 Table 17 Weighted Scheffé Contrasts for Hypotheses Concerned with Retention of Incidental Information Hypothesis Weighted Scheffé Contrast 2 I<+lIPQ+I+lIPQ>I=I - II-lIQP+I--1-IQP*+I-lIP+<--1-)P*I 2 2 4 4 4 4 7 (+1)PQ* - [(-%)P+(-%)P*+(-;11—)QP+(-%)QP*l The average adjusted mean difference of the Scheffé contrast for hypothesis two (1. 60) was not significant (critical value = 10. 49, 5/65fl, p > .05). This result indicated that post—questions did not facilitate the retention of incidental information when com - pared to pre -questions or no questions. Hypothesis two was not supported. The average adjusted mean difference of the Scheffé contrast for hypothesis seven (. 05) was not significant (critical value = 13. 55, 5/65 if, p > .05). This result indicated that the combination of post -questions and a procedure designed to gain attention did not facilitate the retention of incidental information when compared to pre -questions or no questions regardless of whether the procedure involved an attempt to gain Es attention. CHAPTER V DISCUSSION Introduction The results of the present study indicated that trainable mental retardates presented with an aural prose learning task do not benefit from a procedure designed to gain attention or adjunct questions, regardless of location, within the passage. The repeated measures analysis of covariance revealed no significant differences due to the main effect of Cueing, the main effect of Question Loca- tion, or the interaction of Cueing and Question Location. However, within Es sources of variation due to the main effect of Time, the main effect of Information, and the interaction of Time and Informa- tion were significant. Post hoc investigations of the significant main effect of Time indicated that (1) more questions were answered correctly on both the third and fifth administration of the criterion test than on the first administration and that (2) there was no difference between initial and delayed retention. The significant main effect of 73 74 Information showed that the retention of relevant information was better than the retention of incidental information. Post hoc investigations of the significant interaction of Time and Information indicated (1) no pairwise differences in relevant information between each administration of the criterion test, (2) more incidental questions were answered correctly at each of the second, third, fourth, and fifth administrations of the criterion test than on the first administration, and (3) more relevant than incidental questions were answered correctly only on the first administration of the criterion test. Conclusions The major conclusions of this study were: 1. None of the six methods of aurally presenting prose materials to trainable mental retardates proved to be superior to any of the others. 2. A procedure designed to gain the attention of the trainable mental retardate in an aural prose learning task did not facilitate retention of information found in the passage. 75 3. Adjunct questions, regardless of location, did not facilitate the retention of relevant or incidental information found in a prose passage presented aurally to trainable mental retardates. 4. Trainable mental retardates retained relevant information better than incidental information from an aurally presented prose passage. 5. Trainable mental retardates retained as much relevant and/or incidental information from an aurally presented prose passage one‘week following the original presentations as they did immediately following the original presentations. Discussion Engineering the attention of the retardate. -- The fact that cueing was an ineffective means of facilitating the retention of relevant information in a prose passage presented aurally to train- able mentally retarded children does not support the attention theory of retardate learning (House and Zeaman, 1960; Zeaman and House, 1963). House and Zeaman (1960) and Zeaman and House (1963) hypothesized an attention deficit instead of a learning deficit on the part of the mentally retarded from their investigations of 76 two-choice, simultaneous, visual discrimination learning tasks. It may be that the attention theory of retardate learning is appropriate only for visual discrimination learning tasks, and the same findings are not to be found in other types of learning tasks. Another explanation for the obtained results is the procedure designed to gain the attention of the retardate in the present study may have served as a distractor instead of an attention -getting device. The grasp of the _S_' s shoulder and the E' 3 words, "Now listen, " may have caused a startle response instead of gaining the attention of the E. Instead of directing the E' s attention to the relevant characteristics of the learning material, the E may have, unknowingly, simply succeeded in distracting the E. If so, the _S_' s attention would have been diverted from the passage instead of being directed toward the relevant sentences within each paragraph. There is some evidence to suggest that _S_s were aware of the procedure that was used to gain attention. During the delayed retention segment of the study, when the attention -getting device was not used, several Es said to the E, "Oh, you' re not going to poke me?" or "Now listen. " This would indicate that at least some Es were aware of the procedure designed to gain attention. In addition, the trainable mental retardate who received cueing may only have been able to attend to a limited amount of 77 information at one time. He may not have been able to attend to all 15 relevant sentences during the course of one reading of the test passage. Therefore, his attention may have been focused on a particular sub —set of the 15 relevant sentences during one reading of the test passage and a different sub -set of the 15 relevant sentences may have received his attention on a subsequent reading(s) of the test passage. Furthermore, the trainable mentally retarded indi- vidual may possess a limited capacity for retention of correctly recalled information. When he elicits a new correct response toward which his attention has been directed, the trainable mental retardate may incorrectly answer a question which was previously correct. In order to investigate the above propositions and others to be discussed later, four new dependent variables were tallied and analyzed. The new dependent variables were the number of (1) relevant new correct responses, (2) relevant incorrect responses previously correct, (3) incidental new correct responses, and (4) incidental incorrect responses previously correct. Table 18 presents the variances of these four new dependent variables after the third administration of the criterion test for each of the six treatment groups. It was hypothesized that the variability of the relevant new correct responses and relevant incorrect responses previously 78 Table 18 Variances of Relevant and-Incidental New Correct Responses and Incorrect Responses Previously Correct After the Third Administration of the Criterion Test Relevant . Incidental Relevent InCIdental Incorrect Incorrect Treatment New New Response Response Group Correct . Correct . Res onse PreVIOusly Res onse PreVIOusly p Correct p Correct 1(P) 1.29 2.25 2.46 1.69 II(P*) 1.52 2.26 2.40 2.69 III(PQ) 3.57 2.90 3.92 2.08 IV (PQ*) 4.43 2.29 3.56 2.07 V (QP) 1.92 1.46 4.75 3.25 VI (QP*) 4.06 2. 83 5. 54 1. 57 correct would be higher for cued groups than for non-cued groups (1) if the procedure designed to gain attention had diverted attention instead of directing attention, (2) if the cued _S_s changed the sub —set of questions they focused on, and (3) if cueing interacted with a limited capacity for correct responses. An E ratio of the variability for relevant new correct responses for cued versus non -cued groups was not significant (E = 1. 48, 33/33 .93 p > .05). An E ratio of variability for relevant incorrect responses previously correct for 79 cued versus non -cued groups was also not significant (E = 1. 12, 33/33 d_f, p > . 05). Therefore, cueing did not result in more variability than non -cueing for-relevant new correct responses and relevant incorrect responses previously correct. It seems that the only conclusion to be-drawn from the thorough investigation of the data is that the present procedure designed to gain the attention of the trainable mental retardate in an aurally presented prose learning task did not facilitate the retention of the retardate. No explanation can, as yet, be offered for this result. The inducement of mathemagenic behavior. --Questions, regardless of location, did not facilitate the retention of information in prose materials presented aurally to trainable mental retardates. This fact does not support the mathemagenic behavior theory of RothkOpf (1963) or the general findings of written prose learning studies (Rothkopf, 1966; Rothkopf and Bisbicos, 1967; Frase, 1967, 1968a, 1968b). However, these findings and the theory of mathema- genic behavior-were found and developed from studies of written prose using college students as §s. It may be that trainable mentally retarded individuals do-not engage in mathemagenic behavior as a result of adjunct questions or that adjunct questions do not facilitate 80 the retention of information in a prose passage presented aurally to trainable mental retardates. In the present study there is some evidence to suggest that both questions before each paragraph and questions after each para - graph diverted the §_' 3 attention rather than focusing or directing the §' 5 attention on relevant material within the passage. The _l: ratios of variability indicated significantly more variability in relevant new correct responses after the third administration of the criterion test for post-question treatment groups than for no -question treatment groups (5 = 2. 75, 22/22 _df, p < .05) and for pre -question treatment groups than for no -question treatment groups (E = 2.13, 22/22 _d_f_, p < .05). Another E ratio of variability also indicated significantly more variability in incidental new correct responses after the third administration of the criterion test for pre -question treatment groups than for no -question treatment groups (_F_‘ = 2. 18, 22/22 d_f, p < .05). There were no significant differences between pre - and post-question treatment groups in the variability of relevant or incidental new cor- rect responses and relevant or incidental incorrect responses previously correct. Therefore, both pre- and post-questions may have served as distractors for relevant information and pre- questions may have diverted the §' 3 attention from incidental information. .__._J I a... m.~.“.-L ‘fiym- .‘J. “4"‘ 81 It seems that the only conclusions to be drawn from these findings is that the use of adjunct questions in prose material presented aurally to trainable mental retardates does not induce mathemagenic behaviors. In fact, pre - and post-questions may serve as distractors in the retardate' s learning of the material. In a recent article, Rothkopf (1970) more clearly defines the mathemagenic effects of human activities in specified instructional situations. He specifies four categories of mathemagenic effects as (1) mathemagenic positive, (2) mathemagenic negative, (3) mathe- magenic neutral, and (4) mathemagenic unknown. Mathemagenic positive effects are defined as those effects which result from activities which are "conducive to the attainment of the specified instructional objectives" (Rothkopf, 1970, p. 327). Mathemagenic negative effects result from activities which "interfere" with the attainment of specified instructional objectives. The results of the present experiment are an example of mathemagenic neutral effects. The use of immediate, 100 per cent reinforcement may also have contributed to the results of the present study. Most of the pre -1970 studies have not included motivation and have not manipu - lated an incentive variable. It may be that a high level of motivation or incentive overrides or masks the effects of pre- and post- questions interspersed within a prose passage. Frase (1970) . w 4 'M-‘u-fl' WU" WW .._-_§J 82 discusses incentive or the motivational level of the learner as one of three major boundaryconditions for mathemagenic behaviors. He reports a recent study (Frase, Patrick, and Schumer, 1970) in which incentive level, frequency of questions, and question location were manipulated. The findings suggest that (1) post -questions work best at a low incentive level, (2) groups receiving pre -questions are below no ~question groups and post —question groups at moderate incentive levels and that (3) no differences occur between the three question groups at high levels of incentive. Perhaps the highlevel of incentive used in the present study was a contributing factor to the results which showed no differences between no -question, pre - question, and post-question groups. Initial versus delayed retention. —- The fact that there were no significant differences between initial and delayed retention is encouraging. In fact, there were no significant differences between the third administration of the criterion test and the fourth admin— istration for relevant information, incidental information, and relevant and incidental information combined. It should be remem- bered that the fourth administration of the criterion test took place one week after the third administration and was not prompted by another reading of the passage. 83 These findings suggest that information which trainable mental retardates learn from a prose passage presented aurally several times is retained after a relatively long period. The finding of no difference in initial and delayed retention under the conditions of the present study are similar to those general findings of reminiscence in pursuit-motor learning tasks. The similarity suggests that trainable mental retardates who receive massed, reinforced practice on an aural prose learning task remain at the same level of retention performance after a one week delay. This would indicate that once the trainable mental retardate learns information from repeated aural prose material he retains that information over a relatively long period of time. Implications for Future Research The fact that cueing and adjunct questions were ineffective in facilitating the retention of information from prose materials presented aurally to trainable mentally retarded individuals suggests several investigations which should be conducted. Hopefully, these investigations would, clarify the present findings and the attention theory of retardate learning. These investigations might also extend the mathemagenic behavior theory in aural presentations of prose material. 84 The exact methodology and materials used in the present study should be used in two separate studies. One of these would have normal first grade children as §s and the other would have educable mental retardates as _S_s. These two studies would, hope- fully, indicate whether the present findings were due to the method- ology presently employed or the intellectual level of the §s. An aural prose learning study should also be conducted using college students as _S_s. The methodology of this study would be identical to the present one with two exceptions. The procedure designed to gain attention would not be included and the material would be presented to the gs only one time during the original pre- sentation. This situation, plus the use of a delayed retention measure, would extend our knowledge concerning mathemagenic behavior. It would allow us to see if influential variables operate in the same manner in aural prose material as they do in written prose material. The use of a delayed retention measure would permit one to investigate the effects of mathemagenic behavior on longer retention intervals. Current theorizing with respect to retardate learning of an aural prose task is limited. The present study does indicate that repetition of the materials to be learned is beneficial and that repeated, aural, prose materials are retained by the retardate over 85 a relatively long period of time. However, specific methodologies cannot be suggested from the results of this study. A few words of caution are herein directed toward those who would engage in future research endeavors concerned with the aural prose learning of mentally retarded individuals. Specific method- ologies and the manipulations of specific independent variables found to facilitate prose learning of college sophomores should be viewed with caution when one attempts to adopt these in retardate learning of an aural prose task. In fact, generalizations from specific methodologies and the manipulations of independent vari- ables in written prose learning tasks with college sophomores may not be able to be made in aural prose learning tasks with college students as _S_s. With this in mind, only future research efforts will allow us to more clearly delineate the retardate' 5 learning of aural prose material. CHAPTER VI SUMMARY This study was conducted in order to investigate the facilitation of immediate and delayed recall of factual information from prose materials presented aurally to trainable mental retar— dates. Trainable mentally retarded individuals receive most of their verbal information by auditory means due to the inability to read. Therefore, an investigation of methods designed to facilitate the retention of prose materials presented aurally to trainable mental retardates could contribute useful information for the instruc- tion of these children. It was also felt that this kind of investigation could lend itself to an existing theory concerning the learning processes of trainable mentally retarded children. And furthermore, it was thought that such studies would disclose some of the variables which influence the recall of factual information in aurally presented prose materials. Presentation methods were devised which were based upon the inspection behavior theory of Rothkopf (1963, 1965) and the 86 87 attention theory of retardate learning (House and Zeaman, 1960; Zeaman and House, 1963). According to Rothkopf (1965), the use of adjunct questions in written prose materials serves to induce mathemagenic behavior. House and Zeaman (1960) and Zeaman and House (1963) have emphasized the importance of engineering the attention of the retardate toward the relevant characteristics of the learning material before learning can occur. They have preferred to hypothesize an attention deficit instead of a learning deficit on the part of the retardate in two-choice, simultaneous, visual discrimina- tion learning tasks. Therefore, two major extensions occurred in the present study. One was the use and investigation in aural prose of those variables found to be influential in the facilitation of the retention of information in written prose. The other major extension was the use and investigation of a signal designed to gain the attention of the trainable mental retardate in an aural prose learning task as opposed to a two-choice, simultaneous, visual discrimination learning task. Seventy -two trainable mentally retarded students were individually read a 10 paragraph prose passage via one of six presentation methods which manipulated the location of adjunct questions and a procedure designed to gain the attention of the 88 retarded §_s. The same 30item criterion test was administered to each §_ immediately following each of the three readings of the prose passage. The criterion test consisted of 15 three -foil multiple choice relevant questions and 15 three -foil multiple choice incidental ques - tions. The relevant questions pertained to information in the passage to which the §' 5 attention had been directed by means of a special procedure and were identical to adjunct questions which some _S_s heard during the reading of the passage. The ss‘ attention was not directed toward the information in the passage which answered the incidental questions and these questions were not heard by any _S_ during passage presentation. Immediate, 100 per cent reinforce- ment, in the form of M & M candies, was used for all §s during the first three administrations of the criterion test. A measure of delayed retention was also included. The criterion test, the passage via the control condition, and the criterion test were individually administered to each §_ one week following the three initial administrations of the passage and criterion tests. No reinforcement was used during the delayed retention seg- ment of the study. Analysis of covariance was used to analyze the 2 X 3 X 5 X 2 factorial design with repeated measures on the last two factors. The first factor, cueing, consisted of two levels--no cueing and cueing. 89 The second factor was question location and consisted of three levels -—no questions, questions after each paragraph and questions before each paragraph. The third factor represented the five administrations of the same criterion test and the levels of the fourth factor, information, were relevant and incidental. The covariable was the total number correct answers to questions immediately following the reading of each of three short pre- passages. The dependent variable was the number of correct relevant and incidental questions for each administration of the criterion test. The repeated measures analysis of covariance revealed no significant differences due to between _S_s sources of variation. The main effect of Cueing, the main effect of Question Location, and the interaction of Cueing and Question Location were not significant. However, within §s sources of variation due to the main effect of Time, the main effect of Information, and the interaction of Time and Information were significant. Post hoc investigations of the significant main effect of Time indicated that (1) more questions were answered correctly on both the third and fifth administration of the criterion test than on the first administration and that (2) there was no difference between initial and delayed retention. The significant main effect of 90 Information showed that the retention of relevant information was better than the retention of incidental information. A posteriori investigations of the significant interaction of Time and Information indicated (1) no pairwise differences in relevant information between each administration of the criterion test, (2) that more incidental questions were answered correctly at each of the second, third, fourth, and fifth administrations of the criterion test than on the first administration, and (3) that more—relevant than incidental questions were answered correctly only on the first administration of the criterion test. Therefore, it was concluded that trainable mentally retarded students presented with an aural prose learning task via a procedure designed to gain attention and/or adjunct questions (1) did not benefit from any of the six presentation methods; (2) were not facilitated in the retention of factual information via the procedure designed to gain attention; (3) were not facilitated in the retention of factual information via adjunct questions, regardless of location; (4) retained relevant information better than incidental information and (5) retained as much relevant and/or incidental information one week following the original presentations as they did immediately following the original presentations. 91 The finding that the procedure designed to gain attention did not facilitate the retardate' s retention of factual information was discussed in association with the attention theory of retardate learn— ing (House and Zeaman, 1960; Zeaman and House, 1963). It was suggested that (l) the procedure designed to gain attention may have diverted the §' 8 attention instead of directing it toward the relevant aspects of the material to be learned, (2) the trainable mental retardate may only be able to attend to a limited amount of informa- tion at one time, and (3) the trainable mentally retarded individual may possess a limited capacity for retention of correctly recalled information. However, further analyses of the trainable mental retardate' s variability on relevant new correct responses and relevant incorrect responses previously correct did not support the aforementioned hypotheses . The finding that adjunct questions, regardless of location, did not facilitate the retardate' s retention of factual information was discussed in association with the behavior theory of Rothkopf (1963). It was suggested that both pre - and post-questions may have diverted the trainable mental retardate' s attention rather than focusing his attention on relevant material within the passage. Further analyses indicated that both pre - and post-question treatment groups were more variable in relevant new correct responses than no -question F.‘r"‘ IAL- - rm, 1!: .‘M' V...“ we“. ,._-_.~!" 5 92 treatment groups and that pre -question treatment groups were more variable than no -question treatment groups in incidental new correct responses. Recent research (Frase, Patrick and Schumer, 1970) has shown no differences to occur between pre -, post-, and no -question groups due to high inventive levels. This was discussed in lieu of the fact that 100 per cent reinforcement was used in the present study. Three specific studies were implicated for future research efforts in aural prose learning tasks. Two of these would involve use of the exact methodology and materials found in the present study. One of these two studies would have normal first grade children as _S_s and the other would have educable mental retardates as §_s. The third proposed research effort in aural prose learning would be conducted with college students and involve a delayed reten— tion measure. Finally, the limitations of current theorizing with respect to retardate learning of an aural prose task were discussed. Future researchers were cautioned against generalizing from specific methodologies and the manipulations of specific independent variables found to be influential in the learning of written prose materials by college §s. LIST OF REFERENCES LIST OF REFERENCES Ausubel, D. P. , & Youssef, M. The effect of spaced repetition on meaningful retention. Journal of General Psycholog_y, 1965, 73, 147 -150. Baumeister, A. A., Hawkins, W. F., & Holland, J. M. Retroactive inhibition in short-term recall in normals and retardates. American Journal of Mental Deficiency, 1967, 72, 253 -256. Boning, R. A. Getting the Facts; Specific Skill Series, Book A. Rockville Centre, New York: Barnell Loft, Ltd. , 1966. Brown, N. J. , & Sellin, D. F. Learning a school-like task and educational procedures for EMR children. American Journal of Mental Deficiency, 1967, 71, 772 -775. Elashoff, J. D. Analysis of covariance: a delicate instrument. American Educational Research Journal, 1969, 6, 383 -401. Frase, L. T. Learning from prose material: length of passage, knowledge of results, and position of questions. Journal of Educational Psycholog_y, 1967, 58, 266 -272. Frase, L. T. Some data concerning the mathemagenic hypothesis. American Educational Research Journal, 1968, 5, 181 -189. (a) Frase, L. T. Effect of question location, pacing, and mode upon retention of prose material. Journal of Educational Psy- chology, 1968, 59, 244-249. (E; Frase, L. T. Some unpredicted effects of different questions upon learning from connected discourse. Journal of Educational Psychology, 1968, 59, 197-201. (c) 93 ‘..1. ‘flifiigiifh .2“: thing's-U- “my” _J \_ .. .- i 94 Frase, L. T. Boundary conditions for mathemagenic behaviors. Review of Educational Research, 1970, 40, 337 -347. 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, in press. _ House, Betty, & Zeaman, D. Learning and Transfer in Mental Defectives. Progress Report No. 2, NIMH, USPHS, 1960 (Research Grant M-1099 to University of Connecticut). Kirk, R. E. Experimental Desigfin: Procedures for the Behavioral Sciences. Belmont, California: Wadsworth Publishing Co. , 1968. McCarthy, J. J. , & Scheerenberger, R. C. A decade of research on the education of the mentally retarded. Mental Retarda- tion Abstracts, 1966, 3(4), 481-501. Neisworth, J. T., Smith, R. M., & Deno, S. L. Influences of an Advance Organizer on the Verbal Learning and Retention of Educable Mental Retardates. Project No. 6-2122, U.S. Dept. of Health, Education and Welfare, 1968. (Grant No. OEG-1-6-06122 -157o) Rothkopf, E. Z. Some conjectures about inspection behavior in learning from written sentences and the response mode problem in programmed self—instruction. Journal of Programmedlnstruction, 1963, 2(4), 31 ~46. Rothkopf, E. Z. Some theoretical and experimental approaches to problems in written instruction. In J. D. Krumboltz (Ed. ), Learninfimd the Educational Process. Chicago: Rand McNally, 1965. Rothkopf, E. Z. Learning from written instruction material: an exploration of the control of inspection behavior by test- like events. American Educational Research Journal, 1966, 3, 241-249. Rothkopf, E. Z. Textual constraint as a function of repeated exposures. Journal of Educational Psychology, 1968, 59, 20 -2 5. 95 Rothkopf, E. Z. The concept of mathemagenic activities. Review of Educational Research, 1970, 40, 325-336. Rothkopf, E. Z. , & Bisbicos, Ethel E. Selective facilitative effects of interspersed questions on learning from written materials. Journal of Educational Psycholog_y, 1967, 58, 56 -61. Rothkopf, E. Z. , & Coke, Esther U. Repetition interval and rehearsal method in learning equivalences from written sentences. Journal of Verbal Learning and Verbal Behavior, 1963, 2, 406-416. ‘ Rothkopf, E. Z. , and Coke, Esther U. Variations in phrasing, repetition intervals, and the recall of sentence material. Journal of Verbal Learning and Verbal Behavior, 1966, 5, 86 -91. Slamecka, N. J. Studies of retention of connected discourse. American Journal of Psychology, 1959, 72, 409-416. Slamecka, N. J. Retroactive inhibition of connected discourse as a function of practice level. Journal of Experimental Psychology, 1960, 59, 104-108. Vergason, G. A. Facilitation of memory in the retardate. Excep- tional Children, 1968, 34, 589-594. Winer, B. J. Statistical Principles in Experimental Design. New York: McGraw-Hill, 1962. Zeaman, D. , & House, Betty. The role of attention in retardate discrimination learning. In N. R. Ellis (Ed.), Handbook of Mental Deficiency. New York: McGraw -Hill, 1963. APPENDICES U) (qumo'llthJND-‘lm HHH NHO INDIVIDUAL CHRONOLOGICAL AGES AND INTELLIGENCE QUOTIENTS 1 (In (3A. _ISL. 14.17 48 17.92 37 19.00 46 21.25 51 14.17 35 14.25 46 17.75 41 19.83 30 20.58 43 17.58 47 11.67 46 21.42 40 APPENDIX A 11 CP*) (Mt .;Ei_ 17.58 36 18.92 36 15.67 32 20.83 30 18.42 61 14.75 54 21.50 40 14.75 37 12.83 44 16.50 54 16.83 52 16.17 45 96 111 (13(2) cut _Jgg_ 14.92 51 15.00 53 13.58 44 13.42 45 16.17 34 18.67 30 15.25 55 12.58 50 15.75 63 12.50 39 17.00 44 16.00 46 rub-e- .I-Ll— “'1.- l' h.‘ -. _ coooqcncntpoomr-a HHH NHO IV (PQ*) cut .J§g_ 12.25 47 14.08 56 15.17 50 19 00 50 18.17 51 18.33 30 17.67 42 13.83 46 17.75 56 16.50 49 18.58 47 14 17 44 97 V (QP) cut .Jgg_ 16.67 51 14.75 37 18.17 54 13.67 40 18 92 46 15.33 46 13.33 .53 15.67 44 18.50 51 15 17 38 15 00 44 12.67 44 VI (QP*) (Ml .lEL. 11.67 51 13.58 44 18.33 46 12 00 50 18.08 34 20.83 48 13 67 45 14.75 53 16.17 60 15.67 55 14.75 42 15.75 48 APPENDIX B THE THREE PRE -PASSAGES Pre -passage one: "They Walk at Night" Some people walk in their sleep. They get out of bed and walk around the house. Then they go back to bed. People who talk in their sleep are called sleepwalkers. Sleepwalkers do funny things. One woman got out of bed and began to cook food. She mad a fine meal while still asleep. Some sleepwalkers even walk out of their homes. One little boy walked out of his house. He walked until he came to a river. Into the water he jumped. One man got up in the night. He went out and cut down a tree. Back to bed he went. In the morning he got up again. There was the tree on the ground. People say that sleepwalkers do not get hurt. This is not so. Sleepwalkers 'can fall. One man fell from the t0p of his house. He got up on the roof in his sleep! 98 99 Pre massage two: rfiThe King" The king of all animals is not the big elephant. It is not the lion. It is an animal that is not big at all. It is an animal that is very small. It is the little fly called a flea. There is no such thing as a lion circus or an elephant circus. There is a circus with just fleas in it. It is called a flea circus. In two weeks time the flea can get ready for the circus. It can learn to pull wagons. It can learn to kick a ball. It can even learn to dance. The flea is dressed. Sometimes it is dressed like a boy or girl. Sometimes a gold collar is put around its neck. It must look its very best. The flea cannot fly, but it is the best jumper there is. If a boy could jump as well for his size, he could jump right over his house. Prejassage three: "Funny Trees" No one can get up the trunk of the Banana Tree. No one can try to do it. Do you know why? The Banana Tree has no trunk. One tree looks like a puff of smoke. This tree has no leaves. It's gray in color. Guess what it is called. It is called the Smoke Tree. 100 Can you picture a tree with more than one trunk? Can you picture a tree with ten trunks? There is a tree with 200 trunks. It is called the Banyan Tree. There is one tree that looks like it' s going to sleep. It doesn't grow up. It only grows sideways. It looks like it is going to fall down. It is called the Tired Tree. You can' t put your arms around one tree. It takes 30 men. All must hold hands to reach around its trunk. We call it the Fat Tree. APPENDIX C THE TEST PASSAGE: "A TRIP TO THE BEACH" One Saturday the sun was shining. **Father did not have to work that day. He was home. So, the Blackman family decided to take a trip. **Father asked mother to make sandwiches to take with them. **They were going to have a picnic. But, there was not enough food. So, mother went to the grocery store. Mother drove the car. **The store was six blocks away. She bought bread, baloney and apples. Mother also bought cupcakes to eat after lunch. She spent three dollars. Mother began to make lunch. **Carol helped make the lunch. **They also made lemonade to drink. They put the lunch in a box. **Father carried the lunch to the car. The family got in the car and drove away. They only drove one block. **Bill had forgotten his baseball. So, father turned the car around. It took two minutes to go back to the house. 101 102 The family drove to a beach. They ate their lunch first. **During lunch, Bill dropped his apple in the sand. But, mother had another one for him. After lunch, they played baseball. All of the family played. ** Father threw the ball to the batter. Bill and Carol took turns batting. Bill hit a home run. ‘1 Carol did not hit a home run. **Mother caught the ball. ‘1 After playing baseball, the family went swimming. **They swam for a long time. Mother was the best swimmer. Carol was just learning to swim. The family walked up and down the beach after swimming. **They looked for shells. They put them in a bucket. Carol found a pink one. She kept it in her pocket. It was time to go home. They had stayed at the beach all day. **Bill fell asleep on the way home. But, he woke up when they stopped. They stopped at a restaurant. **All of the family had cokes. They got home when it was dark. ** --Asterisks indicate the places where the experimenter grasped the shoulder of the subject and said, "Now listen" to subjects in groups II(P*), IV (PQ*) and VI (QP*). APPENDIX D QUESTIONS FOR THE PRE -PASSAGES Questions for pre -passage one: "They Walk at lfight" 1. People who walk in their sleep are called: 2. When asleep one woman: 3. One boy jumped into a: 4. A man cut a: 5. Sometimes those who walk when asleep get: Questions for pre -passage two: "The King" 1. The kind of all animals is the: 2. There is a circus just for: 3. The flea learns to: 4. The fleas learn many things in two: 5. The flea cannot: 103 104 Questions for pre -passa_g_e_ three: "Funny Trees" 1. The Banana Tree has no: 2. The Smoke Tree is: 3. A tree with many trunks is the: 4. One tree looks like it is going to: 5. To reach around the Fat Tree it takes: *2. *4. *5. *7. *10. *11. 12. *13. *14. 15. 16. APPENDIX E CRITERION TEST QUE STIONS What day'was the sun shining? Who did not have to work that day? 4 What was the name of the family? What did father ask mother to make? What were they going to have ? Where did mother go? How far away was the store? What did mother buy to eat after lunch? How much money did mother spend? Who helped mother make the lunch? What did they make to drink ? What did they put the lunch in? Who carried the lunch to the car? .What had Bill forgotten? How long did it take to go back to the house? What did they, do first? 105 *17. 18. *19. 20. *21. *22. 23. 24. *25. 26. 27. *28. 29. *30. 106 Whad did Bill drop during lunch? When did they play baseball? Who threw the ball to the batter? Who hit a home run? Who caught the ball? How long did they swim? Who was the best swimmer? Who was. just learning to swim? What did they look for on the beach? What did they put them in? What was the color of the one Carol found? Who fell asleep on the way home? Where did they stop? What did all of the family have? * -—Asterisks indicate relevant questions. Question 1. Administration 1 .APHEEDHIEK]? "THEY WALK AT NIGHT" Administration 2 ORDERS OF ANSWERS FOR PRE -PASSAGE ONE: Administration 3 nightwalkers sleepwalkers sleepers sleepers sleepwalkers nightwalkers read ate cooked cooked ate read boat river car car river boat tree cake rope rope cake tree hurt lost sick sick lost hurt 107 sleepers sleepwalkers nightwalke rs cooked ate read car river boat rope cake tree sick lost hurt Question 1. APPENDIX G ORDERS OF ANSWERS FOR PRE -PASSAGE TWO: Administration 1 "THE KING" Administration 2 Administration 3 lion elephant flea birds dogs fleas talk kick sing days years weeks fly jump eat flea elephant lion fleas dogs birds sing kick talk weeks years days eat jump fly 108 elephant flea lion dogs fleas birds kick sing talk years weeks days jump eat fly APPENDIX H "FUNNY TREES" ORDERS OF ANSWERS FOR PRE -PASSAGE THREE: Question - Administration 1 .Administration 2 Administration 3 1 . trunk leaves bananas bananas bananas leaves leaves trunk trunk 2 . green gray blue blue blue gray gray green green 3 . Banyan Elephant Baby Baby Baby Elephant Elephant Banyan Banyan 4. play sleep run run run sleep sleep play play 5. 30 men 5 men 2 men 2 men 2 men 5 men 5 men 30 men 30 men 109 APPENDIX I VOCABU LARY CHE CK LIST doesn't know s know Make believe you forgot your coat this morning. Where did you leave it? Make believe you' re playing baseball and you hit a home run. What does that mean? (How far around the bases would you get to run?) What do you do in a restaurant? What do you wear when you go swimming? What do you do on a picnic? You' re walking along the beach and you find some shells. What are shells? What do they look like ? Color? Size ? What do you buy in a grocery store? What is a bucket? What are sandwiches ? What do you do with lemonade? Note: The underlined words indicate the vocabulary con- cepts of interest. 110 APPENDIX J INSTRUCTIONS FOR EACH TREATMENT GROUP Group I (P) "Today we are going to listen to a story. I will read the story to you three times. ”After I read all of the story, I am going to ask you questions about the story. Listen carefully to the story so you will be able to answer the questions I ask you. Are you ready, (child' 8 name)? O.K. Now I will read the story for the first time. Remember to listen carefully so you will be able to answer the questions I ask you at the end of the story. " Group II (P*) "Today we are going to listen to a story. I will read the story to you three times. Sometimes I will say, ' Now listen. ' When I say ' Now listen, ' I want you to listen carefully to the next sentence I read. "After I read all of the story, I am going to ask you questions about the story. Listen carefully to the story so you will be able to answer the questions I ask you. Are you ready, (child' 3 name)? O.K. Now I will read the story for the first time. Remember to listen carefully so you will be able to answer the questions I ask you at the end of the story. " 111 112 Group III (PQ) "Today we are going to listen to a story. I will read the story to you three times. I will also read some questions after each paragraph. When I read these questions, I want you to try to answer them. "After I read all of the story, I am going to ask you questions about the story. Listen carefully to the story so you will be able to answer the questions I ask you. Are you ready, (child' s name)? O.K. Now I will read the story for the first time. Remember to listen carefully so that you will be able to answer the questions I ask you at the end of the story. " Group IV (PQ*) "Today we are going to listen to a story. I will read the story to you three times. Sometimes I will say, 'Now listen. ' When I say ' Now listen, ' I want you to listen carefully to the next sentence I read. I will also read some questions after each paragraph. When I read these questions, I want you to try to answer them. ”After I read all of the story, I am going to ask you questions about the story. Listen carefully to the story so you will be able to answer the questions I ask you. Are you ready, (child' s name)? O.K. Now I will read the story for the first time. Remember to listen carefully so you will be able to answer the questions I ask you at the end of the story. " Group V (QP) "Today we are going to listen to a story. I will read the story to you three times. I will also read some questions before each paragraph. When I read the paragraph, I want you to try to answer the questions. ”After I read all of the story, I am going to ask you questions about the story. Listen carefully to the story so you will be able to answer the questions I ask you. Are you ready, 113 (child' 3 name)? O.K. Now I will read the story for the first time. Remember to listen carefully so you will be able to answer the questions I ask you at the end of the story. " Group VI (QP*) "Today we are going to listen to a story. I will read the story to you three times. Sometimes I will say, 'Now listen. ' When I say ' Now listen, ' I want you to listen carefully to the next sentence I read. I will also read some questions before each paragraph. When I read the paragraph, I want you to try to answer the questions. "After I read all of the story, I am going to ask you questions about the story. Listen carefully to the story so you will be able to answer the questions I ask you. Are you ready, (child' 8 name)? O.K. Now I will read the story for the first time. Remember to listen carefully so you will be able to answer the questions I ask you at the end of the story. " APPENDIX K ORDERS OF ANSWERS FOR INITIAL CRITERION TESTS t1 Question Administration 1 Administration 2 Administration 3 1 . Friday Saturday Sunday 2 Saturday Friday Saturday Sunday Sunday Friday 2. Father Mother Uncle Jim Mother Father Mother Uncle Jim Uncle Jim Father 3. Carson Daniels Blackman Daniels Carson Daniels Blackman Blackman Carson 4. a cake sandwiches cookies sandwiches a cake sandwiches cookies cookies a cake 5. a talk a new dog a picnic a new dog a talk a new dog a picnic a picnic a talk 6. grocery store gas station post office gas station grocery store gas station post office post office grocery store 7. 1 block 10 blocks 6 blocks 10 blocks 1 block 10 blocks 6 blocks 6 blocks 1 block 114 Question 8. 10. 11. 12. 13. 14. 15. 16. 17. Administration, 1 115 Administration 2 Administration 3 cookies cupcakes pies $5 $1 0 $3 Carol Father Bill ice tea lemonade kool -aid a basket a paper bag a box Father Bill Carol his swimsuit his baseball his jacket 10 minutes 5 minutes 2 minutes ate lunch went swimming played baseball his sandwich his cookie his apple cupcakes cookies pies $10 $5 $3 Father Carol Bill lemonade ice tea kool -aid a paper bag a basket a box Bill Father Carol his baseball his swimsuit his jacket 5 minutes 10 minutes 2 minutes went swimming ate. lunch played baseball his cookie his sandwich his apple pies cupcakes cookies $3 $1 0 $5 Bill Father Carol kool -aid lemonade ice tea a box a paper bag a basket Carol Bill Father his jacket his baseball his swimsuit 2 minutes 5 minutes 10 minutes played baseball went swimming ate lunch his apple his cookie his sandwich Question 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Administration 1 116 Administration 2 Administration 3 after lunch before lunch after supper Mother Father Bill Bill Carol Mother Carol Mother Bill a little while an hour a long time Mother Bill Father Bill Carol Mother shells rocks pebbles a box a bucket a basket pink orange white before lunch after lunch after supper Father Mother Bill Carol Bill Mother Mother Carol Bill an hour a little while a long time Bill Mother Father Carol Bill Mother rocks shells pebbles a bucket a box a basket orange pink white after supper before lunch after lunch Bill Father Mother Mother ‘ Carol Bill Bill Mother Carol a long time an hour a little while Father Bill Mother I l I Mother l Carol 1 Bill ' | pebbles rocks shells a basket a bucket a box white orange pink Question 28. 29. 30. Administration 1 Carol Mother Bill at a restaurant at a friend' 8 house at church cokes sundaes hotdogs 117 Administration 2 Mother Carol Bill at a friend' 3 house at a restaurant at church sundaes cokes hotdogs Administration 3 Bill Mother Carol at church at a friend' s house at a restaurant hotdogs sundaes cokes APPENDIX L ORDERS OF ANSWERS FOR THE DELAYED CRITERION TEST Question Administration 1 Administration 2 1. Friday Sunday Sunday Friday Saturday Saturday 2. Father Uncle Jim Uncle Jim Father Mother Mother 3. Carson Blackman Blackman Carson Daniels Daniels 4. a cake cookies cookies a cake sandwiches sandwiches 5. a talk a picnic a picnic a talk a new dog a new dog 6. grocery store post office post office gas station grocery store gas station 7. 1 block 6 blocks 6 blocks 1 block 10 blocks 10 blocks 118 Question 8. 10. 11. 12. 13. 14. 15. 16. 17. 119 Administration 1 Administration 2 cookies pies cupcakes $5 $3 $1 0 Carol Bill Father ice tea kool -aid lemonade a basket a box a paper bag Father Carol Bill his swimsuit his jacket his baseball 10 minutes 2 minutes 5 minutes ate lunch played baseball went swimming his sandwich his apple his cookie pies cookies cupcakes $3 $5 $1 0 Bill Carol Father kool -aid ice tea lemonade a box a basket a paper bag Carol Father Bill. his jacket his swimsuit his baseball 2 minutes 10- minutes 5 minutes played baseball ate lunch went swimming his apple his sandwich his cookie Question 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 120 Administration 1 after lunch after supper before lunch Mothe r Bill Father Bill Mother Carol Carol Bill Mother a little while a long time an hour Mother Father Bill Bill Mother Carol shells pebbles rocks a box a basket a bucket pink white orange Administration 2 after supper after lunch before lunch Bill Mother Father Mother Bill Carol Bill Carol Mother a long time a little while an hour Father Mother Bill Mother Bill Carol pebbles shells rocks a basket a box a bucket white pink orange Question 28. 29. 30. 121 Administration 1 Carol Bill Mother at a restaurant at church at a friend' 8 house cokes hotdogs sundaes Administration 2 Bill Carol Mother at church at a restaurant at a friend' 5 house hotdogs cokes sundaes DEPENDENT VARIABLES AND COVARIABLE APPENDIX M T1 T2 T3 T4 T5 .88 Covariable R I R I R I R I R I Treatment Group I (P) 1 9 7 7 8 8 8 9 10 9 8 14 2 5 3 6 5 6 6 5 4 8 8 19 3 14 13 11 11 .11 10 11 9 » 13 11 20 4 15 14 14 15 15 15’ 15 15 15 15 37 5 7 4 7 7 5 7 7 6 9 6 15 6 12 13 14 15 15 14 15 15 15 15 32 7 7 5 9 4 10 7 9 8 6 8 16 8 5 5 5 4 6 6 5 8 4 5 20 9 9 7 8 8 8 11 11 9 12 12 21 10 13 12 14 12 14 13 15 13 15 14 22 11 11 12 12 12 14 13 13 13 13 14 22 12 13 14 12 14 15 15 11 15 15 .14 41 Treatment Group II (P*) 1 8 5 5 6 7 .3 6 7 3 3 16 2 4 5 6 8 4 7 4 3 4 2 12 3 7 5 6 _ 5 5 6 5 6 4 3 17 4 10 9 8 8 7 3 10 8 7 7 13 5 10 10 15 12 11 13 14 13 14 13 31 6 9 9 11 9 10 9 10 10 10 9 35 7 9 8 11 10 9 12 10 12 12 11 18 8 9 4 6 8 6 4 5 6 8 8 18 9 12 11 11-11 10 11 11 10 ll 11 21 10 12 12 15 10 15 13 15 11 14 10 18 11 6 5 7 7 2 4 4 3 4 6 17 12 11 10 12 11 12 12 13 14 15 13 26 123 T1 T2 T3 T4 T5 83 Covariable R I R I R I R I R I Treatment Group III (PQ) 1 14 13 14 14 14 12 15 11 14 12 43 2 13 10 14 13 .13 14 12 11 13 11 13 3 10 9 13 13 15 12 14 12 15 15 29 4 9 5 13 8 11 7 9 7 11 7 18 5 6 5 5 5 6 7 5 4 5 4 18 6 7 3 8 5 11 8 9 6 10 8 10 7 14 12 . 15 12 15 13 15 14 15 14 17 8 11 8 14 7 13 7 14 7 ll 7 17 9 14 13 15 14 15 14 15 14 15 14 42 10 7 5 7 7 7 8 4 5 6 4 17 11 11 12 13 13 11 14 11 12 12 13 25 12 13 13 14 15 14 15 13 15 15 15 27 Treatment Group IV (PQ*) 1 7 4 5 6 7 5 5 7 8 5 15 2 10 11 12 10 13 12 12 11 12 11 26 3 10 5 10 5 7 9 7 6 6 6 17 4 7 9 9 8 9 7 6 6 8 6 18 5 10 8 11 9 11 11 9 10 13 13 16 6 13 7 13 8 12 10 11 6 11 9 25 7 8 5 7 6 7 9 8 6 10 9 ‘22 8 13 14 15 11 14 12 15 ll 15 9 23 9 13 12 14 14 14 14 14 14 15 15 40 10 14 14 14 14 15 13 14 12 15 14 16 11 5 4 5 5 6 7 8 7 10 7 19 12 10 9 9 9 14 11 9 6 10 11 19 124 T1 T2 T3 T4 T5 Ss Covariable R. I R. I R. I R. I R. I Treatment Group V (QP) 1 14 13 14 14 12 14 13 14 12 15 34 2 5 6 6 6 6 4 5 , 4 5 5 10 3 13 11 13 13 14 14 14 14 15 15 44 4 10 6 4 7 5 4 9 8 6 5 14 5 14 12 15 13 15 15 14 14 15 15 26 6 12 8 .15 12 13 14 .14 13 15 13 19 7 11 8 12 10 13 11 12 10 15 10 16 8 12 13 13 14 14 14 13 14' .13 14 29 9 11 10 14 14 15 15 14 '14 14 15 33 10 13 6 13 5 12 9 13 6 13 6 23 11 6 5 9 7 11 11 9 7 10 7 18 12 11 7 14 8 . 14 9 12 10 12 7 16 Treatment Group VI (QP*) 1 11 9 12 8 11 10 12 11 11 9 26 2 8 6 9 10 10 .11 8 10 13 10 20 3 7 8 10 9 9 9 6 5 8 7 16 4 12 11 12 ,11 12 12 10 11 10 10 15 5 5 4 5 5 9 10 7 7 10 9 18 6 12 ‘4 12 6 13 7 12 7 13 8 17 7 13 8 ~12 10 13 9 13 10 13 10 19 8 10 5 9 8 11 9 11 6 10 11 25 9 15 14 15 .15 15 15 15 15 15 15 43 10 14 12 15 15 15 15 14 13 13 15 40 11 6 5 6 4 7 7 4 4 3 7 17 12 8 5 8 5 5 7 7 5 6 4 26