IsmmmlHHWHWWW'MI 3 1293 01107 4303. RETURNING MATERIALS: )VIESI_J Place in book drop to LIBRARIES remove this checkout from .5213. your record. FINES Will be charged if book is returned after the date stamped below. Kama W93» L44 (w; ’ ‘_; I THE USE OF MELODIC-RHYTHMIC MNEMONICS WITH LEARNING DISABLED AND NORMAL STUDENTS AS AN AID TO RETENTION By Kay E. Gfeller A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Music 1982 ABSTRACT THE USE OF MELODIC-RHYTHMIC NNEMONICS WITH LEARNING DISABLED AND NORMAL STUDENTS AS AN AID TO RETENTION By Kay E. Gfeller This study examined the effectiveness of one particular strategy, the melodic-rhythmic mnemonic, as an aid to retention of academic facts--specifically, multiplication facts. According to some re- searchers in learning disabilities, using such strategies helps learn- ing disabled students, who may exhibit poor short-term memory, in keeping pace with successful learners who spontaneously initiate strategies. Subjects included 30 learning disabled and 30 normal male stu- dents, ages 9.0 to ll.9 years with 90 to l20 1.0.5, as measured by a Slosson Intelligence Test. All subjects participated in two experi- ments conducted over three days. Experiment I consisted of a pretest, a single rehearsal of multi- plication facts, and, immediately following rehearsal, a posttest for short-term memory. This test examined the influence of the indepen- dent varibles, group membership (learning disabled or normal), and rehearsal mode (musical or verbal) on the dependent variable, short— term memory. Kay E. Gfeller Immediately following Experiment I, all subjects participated in Experiment II: four additional rehearsals spaced over three days. Treatment consisted of further exposure to multiplication facts in order to examine the effect of group membership, rehearsal mode, and extended time. The independent variable of teaching method was added: (a) repetition only, and (b) repetition with modeling and cuing. The Results Data from Experiment I were analyzed through univariate analysis of covariance. Main effects of group membership and rehearsal mode were statistically significant at the .05 level, indicating that nor- mals showed greater recall than learning disabled subjects, and that greater recall for both groups resulted from verbal as opposed to musical rehearsal. There was no significant interaction effect of group by rehearsal mode. Data from Experiment 11 were analyzed using multivariate analy- sis of covariance for repeated measures. Interaction of rehearsal mode by teaching method was statistically significant at the .000l level. This supported musical rehearsal in conjunction with modeling and cuing as an effective aid to retention for both groups. Further hypotheses for (a) main effect of group, (b) interac- tion effects of group by rehearsal mode, and (c) group by rehearsal mode by teaching method were statistically nonsignificant at the .05 level. ACKNOWLEDGEMENTS I would like to express my appreciation and thanks to my disser- tation advisor, Dr. Dale Bartlett, and to the members of the disserta- tion committee, Drs. Dale Bonge, Linda Patriarca, Charles Ruggiero, Melanie Stuart, and Robert Unkefer, for their counsel and construc- tive criticism concerning the preparation of this thesis. A special acknowledgement is due Dale Bartlett and Linda Patri- arca for their support and invaluable suggestions through each step of this endeavor. I also wish to thank Denise Joseph-Enders for her steadfast as- sistance throughout the data collection procedures. Her professional excellence, conscientiousness, and friendly enthusiasm were crucial to the completion of this research project. Additionally, I would like to thank Gabby Belli for her clear and thorough statistical consultation and Judy Jankowski, Jane Cook, and Juan Olivarez for facilitating access to various school systems in the greater Grand Rapids, Michigan, area. Without their help and the cooperation of the many administrators, teachers, parents, and students contacted throughout the research process, this study would not have been possible. I am also deeply greatful to my family and friends for their en- couragement and support throughout my doctoral program. ii Finally, I wish to express my love and heartfelt appreciation to my husband, Kyran Cook, who provided unfailing support, patience, en- couragement, understanding, and a sense of humor, even in the most trying times of my graduate endeavors. TABLE OF CONTENTS Page List of Tables ......................... vi List of Figures ........................ vii CHAPTER ONE: OVERVIEW OF THE PROBLEM ............. l Purpose .......................... l Background of the Problem ................. 2 The Problem ........................ 8 Importance of the Study .................. lO Assumptions ........................ l3 Limitations ........................ 16 Theoretical Background .................. l6 Definitions ........................ 23 CHAPTER TWO: SURVEY OF RELATED LITERATURE ........... 28 Normal Memory Storage and Retrieval ............ 29 Musical Memory as It Relates to Normal Memory ....... 33 Psychoacoustic Research on Musical and Musical/ Verbal Memory of Normal Subjects ........ 37 Research on Musical/Verbal Memory with the Mentally Retarded .................... 40 Special Characteristics and Needs of the Learning Disabled .................... 45 CHAPTER THREE: PROCEDURE ................... 54 Subject Selection ..................... 54 Subject Characteristics ................ 61 Overview of the Study ................... 65 Pilot Studies ....................... 66 The Major Study ...................... 7O Preparations and Materials .............. 70 Data Collection .................... 75 Data Analysis--Experiment I .............. 80 Data Analysis--Experiment II ............. 86 iv Methodological Assumptions ................ 87 Limitations ........................ 88 Statement of the Testable Hypotheses ........... 89 The Hypotheses for Experiment I ............ 89 The Hypotheses for Experiment II ........... 90 Summary .......................... 9l CHAPTER FOUR: FINDINGS .................... 93 Experiment I ....................... 93 Interaction Effects .................. 94 Main Effects ..................... 94 Experiment II ....................... 97 Interaction Effects .................. 100 Main Effects ..................... 105 Interpretation ...................... l06 Experiment I ..................... lO6 Experiment 11 ..................... lO8 Additional Findings .................... llO Effect of Musical Environment ............. llO CHAPTER FIVE: SUMMARY AND CONCLUSIONS ............. ll3 Summary .......................... ll3 Experiment I ..................... ll3 Experiment 11 ..................... ll6 Conclusions ........................ l27 Recommendations ...................... l28 FOOTNOTES ........................... 129 APPENDICES A. DATA COLLECTION FORMS ................. l32 B. TESTING SCHEDULE AND PROBLEM ORDER .......... l35 C. MELODIC-RHYTHMIC PATTERNS ............... 137 D. MUSICAL ENVIRONMENTAL QUESTIONNAIRE .......... l39 E. STANDARDIZED INSTRUCTIONS ............... l4l F. TAPE SCRIPT ...................... 149 G. PARENTAL CONSENT FORMS . . H. UNIVARIATE STEPDOWNS--EXPERIMENT II .......... l53 REFERENCES ........................... 155 LIST OF TABLES 1221?. 329.9. 1. Table of Specifications ................. 6O 2. Frequencies and Central Tendencies for Learning Disabled Students ............... 62 3. Cross Tabulation of Frequencies and Chi Square Analysis .................... 63 4. Experiment I and Experiment II--the Procedural Design . . 77 5. Experiment I ...................... 78 6 Breakdown of Subgroups--Experiment II .......... 82 7 Experiment II Treatment Design ............. 83 8. ANCOVA--Findings for Experiment I ............ 95 9 Obtained Group Means Derived from Gains Between Pretest and Posttest .............. 96 IO. Multivariate Analysis of Covariance--Experiment II . . . lOl ll. Musical Environment and Its Effects on Recall Accuracy .................... lll vi Figure l. 2. LIST OF FIGURES Symmetric Matrix Formed from the Repeated Measures within Experiment II ......... Group Means-~Rehearsal Mode by Teaching Method Interaction .................. vii CHAPTER I OVERVIEW OF THE PROBLEM Purpose The purpose of this study was to examine melodic-rhythmic mnemon- ic devices as an aid to retention for learning disabled* and normal children. The effectiveness of this mnemonic, which will be called musical rehearsal, was measured in a two-part experiment examining the following variables: (a) subject classification (learning dis- abled versus normal students), (b) mode of rehearsal (musical versus normal speech rehearsal), and (c) method of teaching (repetition ver- sus repetition with modeling and cues). While many studies for short- term memory use mnemonic devices in conjunction with serial lists, this study investigated the effectiveness of such a device with actual academic information which necessitated a closed-ended response -- in this case, multiplication facts. *According to law R.340.l7l3 of the State of Michigan's law for special education, specific learning disability means a disorder in one or more of the basic psychological processes involved in under- standing or in using language, spoken or written, which may manifest itself in an imperfect ability to listen, think, speak, read, spell, or to do mathematical calculations. The term includes conditions such as perceptual handicaps, brain injury, minimal brain dysfunc- tion, dyslexia, and developmental learning problems which are pri- marily the result of visual, hearing, or motor handicaps, of mental retardation, or emotional disturbance, or of environmental, cultural, or economic disadvantage. Background of the Problem Finding effective and efficient teaching methods for learning disabled children is an ongoing concern of parents, educators, and school administrators. Teachers and parents most directly involved with the students are acutely aware of the career limitations and so- cial handicaps that a student with academic problems faces. State legislators and administrators are confronted continually with budge- tary concerns that arise from the high cost of special education pro- gramming. For these reasons, finding effective and efficient teach- ing strategies present a significant problem in special education. While public sensitivity to the educational needs of the learn- ing disabled has grown rapidly within the past two decades, ophtha- mologists and neurologists have been aware of specific reading prob- lems as early as l9l7. Hinshelwood (l9l7) and Orton (l928) both viewed "word blindness" as a neurological condition which affected the individual's ability to read despite normal visual acuity and nor- mal intelligence. In l947, Strauss and Lehtinen's description of the brain injured child resulted in an entirely new category of exceptional children. This category included a collection of children with disturbances in perception, thinking, and emotional behavior caused by some type of organic impairment which impeded normal learning processes, despite normal acuity and 1.0. Since 1947, causal labels (brain injury, mini- mal brain dysfunction) have been discarded in favor of descriptive labels (attention deficit disorders, learning disabilities) which more directly describe the resulting behaviors and problems. This dis- ability has been approached as a perceptual-motor problem (Cruickshank and Kephart) and as a cognitive or language problem (Johnson and Myklebust); but, despite the varying approaches, these professionals have worked toward the common goal of increased academic success for the learning disabled student. Like other special education personnel, music educators and music therapists are seeking more effective remediation strategies for the learning disabled. These professionals use music as a medium for re- inforcing appr0priate behavior, academic concepts, and motor develop- ment. Past research by music therapists has documented the use of music as an effective motivational tool for reinforcing various beha- vioral and academic goals; however, the bulk of this research has been with either mentally retarded or normal individuals. Research with retardates forms the basis for many of the current musical interven- tions used with the learning disabled. Yet the learning disabled constitute a population with characteristics and needs unique from either mentally retarded or normal children. In the search for effective musical strategies for the learning disabled, there are two problems that arise: (a) there exists very limited methodological literature on using music with learning dis- abled students, and even less research on the effectiveness of musi- cal methods with this population; and (b) much of the extant program- ming literature is developed either from research on the mentally retarded or from perceptual or motor remediation theories such as those of Kephart, Getman, Cratty, and Doman and of Delcato (Phipps, l975; Flick, l975). While it appears that many learning disabled children do have some percpetual or motor problems, research find- ings are inconclusive at best, regarding the effectiveness of isolating and remediating perceptual or motor skills. Furthermore, research data showing that academic improvement results from train- ing is lacking (Lerner, l976; Reid and Hresko, l98l; Hammill and Lar- sen, l974). If the validity of isolating and remediating perceptual-motor processes is questionable, and if the goal is to enhance academic learning, the use of music therapy interventions based on those theories should be closely reexamined for actual effect on academic proficiency. Additionally, music therapists have a professional re- sponsibility to reevaluate the role of music interventions now used in view of current thinking and to maintain valid methodology. One of the current trends in research and programming for the learning disabled is the cognitive approach, which includes develop- ment of teaching strategies for various academic tasks (Reid and Hresko, l98l). Cognition is a complex, process of knowing and per- ceiving, consisting of various thinking skills including memory. Memory capacity and recall facilitation has been a current focus of numerous investigations with the learning disabled: It is widely recognized that the learning disabled have deficits* in memory. Two of the six major areas of dis- ability described by Myers and Hammill (l976) and all of the basic learning abilities described Valett (l969) involve memory. Many other writers (Chalfant and Scheffe- lin, l969; Doering, l968; Johnson and Myklebust, l967; Kirt, l966) have also reported that the learning dis- abled may experience difficulties with one or another memory tasks. *Disorder may be a more appropriate term, since deficit implies the lack of memory, as opposed to ineffective use of memory. Reid and Hresko (l98l) report that problems may occur in visual or auditory memory, ability to transfer information from short-term to long-term store, and in organization of incoming information. They stress that memory as an end in itself is not an appr0priate educa- tional goal. However, memory as it interacts with mastery of educa- tional content is an important process which affects academic success. Thus, while memory exercises in and of themselves are of doubful value, memory strategies (based on theories of the memory process) incorporated into presentation or rehearsal of classroom information may enhance academic performance. Some of the strategies which may enhance storage of information developed from our current understand- ing of the memory process include (a) organization or chunking of in- coming stimuli; (b) effective rehearsal methods, including verbal re- hearsal and mnemonic strategies; and (c) appropriate presentation and review rates (Reid and Hresko, l98l; Lerner, l976). Considering these suggestions, in what way can music act as an effective intervention in aiding retention of academic information? For years music therapists, working with the mentally retarded, have used songs to aid retention of selected concepts or knowledge of colors or numbers. While it is true that the mentally retarded have many characteristics which are different from the learning dis- abled, such as subnormal I.Q., many retardates also have memory prob- lems (Kirk and Gallagher, l979), though not necessarily with the same etiological base as learning disabled children. While the research results on music to aid retention have been inconclusive, several studies do support its use. For example, research by Isern (1958) and Lathom (l970) support the use of music as an effective educational aid with the retarded. Some researchers consider music to be an effective educational mediation because it functions as a motivator or because it is a novel and, thus, interesting stimulus; other researchers, however, see mu- sic's potential embodied in the organizational elements of rhythm and tonality, which may, in and of themselves, enhance meaningful communi- cation and increased retention (Lathom, l97l; Seashore, l9l9; Restle, l970; Robinson and Solomon, l974; and Bloomslitter and Creel, l977). Various psychoacoustical studies have shown that rhythmic organi- zation and tonally based melodies do, indeed, aid meaning and reten- tion of music without words. In addition, such musical structures superimposed upon verbal information have shown effective increase of recall in a short-term memory task using serial lists, such as unre- lated digits (Jellison, l976; Prickett, l974). Although the organizational components of music have proven ef- fective in aiding retention of serial lists, do these same components facilitate recall when utilized in an actual academic task? Research- ers in the memory field believe that memory processes can differ de- pending on the nature of information to be recalled (Norman, l976). It is, therefore, important when developing memory strategies based on findings from research with serial tasks to reexamine the effective- ness of those strategies when used in conjunction with meaningful academic information. Further, in view of the heterogeneous nature of the learning dis- abled population,* it is important to examine critically a strategy considering individual student needs as opposed to viewing a strategy as an all-effective panacea. Unfortunately, researchers too often face one of the following research problems: (a) there is little con- trol over heterogeneity of wide ranging characteristics in the learn- ing disabled population; (b) the research has included so little de- scriptive information on the subject sample that replication of the study is thwarted and generalizability of findings is clouded; or (c) the strategy under examination has been analyzed only in conjunc- tion with group trends, with little attention to specific subject characteristics within the group. In view of these research problems, Keogh et al. (l98l) emphasize the importance of detailed subject de- scription along with research findings or methodological recommenda- tions. In summary, while music therapists have recommended the use of music as an educational aid for the learning disabled (Phipps, l975; Flick, l975; Roskam, 1977), its value must be examined systematically in view of current research findings in the field of learning disa- bilities. Additionally, musical methods must be examined for actual benefit to academic growth, always keeping in mind the importance of individual educational needs within the learning disabled population. *This population represents a conglomeration of learning problems despite normal 1.0. and sensory acuity. The Problem This study investigated whether melodic-rhythmic organization acted as an effective mnemonic device to aid learning disabled and normal students in retention during a short term memory task. The problem was broken into the following subproblems: I. Can the use of melodic-rhythmic structure facilitate encoding and retention* in a short term memory task? Can selected melodic-rhythmic patterns introduced dur- ing the rehearsal process provide an effective retrie- val cue during recall? Is the effectiveness of a melodic-rhythmic mnemonic strategy increased by modeling its use and providing retrieval cues? (See the definitions for these terms at the close of this chapter.) Do learning disabled and normal students exhibit greater or lesser retention in interaction with specific methods of verbal or musical rehearsal? When provided with memory strategies and their appro- priate use over time, will learning disabled stu- dents exhibit similar growth in retention as normal students? Within the heterogeneous learning disabled population, do any particular subgroups (categorized by severity *Darley, Glucksberg, Kamin, and Kinchla (l98l) describe en- coding as the way information is first stored in the memory process, retention describes the preservation of information over time, and retrieval is the manner in which information is recovered from memory. of academic or attentional problems, activity level, or area of academic disability) respond differently to various treatment methods? These subproblems were examined through two experimental proce- dures, Experiment I and Experiment II. Experiment I consisted of a pretest, a single rehearsal of the memory task, followed immediately by a posttest to assess recall. Testing examined the effects of two independent variables, group level (normal or learning disabled), and rehearsal mode level (musical or verbal rehearsal) on the dependent variable, posttest recall of the memory task. The following hypotheses were generated about the expected find- ings in Experiment I: l. Normal students will show significantly greater im- provement in recall accuracy than learning disabled subjects in both musical and verbal rehearsal. 2. All subjects will show significantly greater improve- ment in recall accuracy from musical as opposed to verbal rehearsal. 3. A significant interaction effect will result from group level (normal or learning disabled) by rehear- sal mode level (musical or verbal). Experiment II, which consists of three repeated measures of the dependent variable recall accuracy, examined the effect over time of the following independent variables: group level (normal and learning disabled); rehearsal mode level (musical and verbal rehearsal); and teaching method (a) posttest only for control purposes, (b) method I l0 which is simple repetition of the memory task, and (c) method 11 which included modeling and cuing along with the memory task. The following hypotheses are related to research procedures in Experiment II: 1. Normal students will show significantly greater im- provement in recall accuracy than learning disabled students in all treatment methods. All subjects will show greater improvement in recall accuracy following musical rehearsal compared to ver- bal rehearsal. All subjects will show significantly greater im- provement in recall accuracy between levels of teach- ing methods (posttesting only versus methods I and 11 combined and method I versus method II). A significant interaction effect will result from the following combinations across time: a. group level by rehearsal mode level, b.’ group level by teaching method level, c. rehearsal mode level by teaching method alone, and d. group level by rehearsal mode level by teaching method level. Importance of the Study The field of music therapy, that is, using music to achieve spe- cific therapeutic goals, is a relatively young discipline in which continued and more thorough research efforts are needed greatly. ll Music therapists are becoming increasingly aware of the need to re- search and document actual effectiveness of utilized music therapy methodology. As music therapists expand their practice to include work with the learning disabled, increased research efforts become even more vital to success and credibility. Currently, there is much methodo- logical reliance on research resulting from work with the mentally retarded,and very little available research exists on the use of music as an educational aid for the learning disabled. In addition, several issues in the field of learning disabilities require further study. First, there is a critical need for more effective teaching techniques. With a shift away from isolated remediation of percep- tual motor skills and a movement toward improved cognitive perfor- mance (Reid and Hresko, l98l), existing teaching techniques must be examined for effectiveness, and further approaches developed. Lerner (l976) and Reid and Hresko (l98l) have identified memory problems as one area of cognitive functioning that requires further understanding and investigation. One example of a specific academic task in which memory problems surface is illustrated by Moyer and Mayer (1978). They discuss a problem that may occur in a mainstreamed situation in which a class is expected to learn multiplication tables: Difficulty recalling the basic facts is not limited to learning disabled children: memorizing the facts is a laborious process for many children. However, for those children with accompanying memory problems, this task may be overwhelming.2 It is worth noting that during the data collection process of this study, it was not at all unusual to find l0 and ll year old learning disabled students still struggling with memorization of the same facts 12 that the majority of their normal counterparts had mastered at age nine. Second, currently available or newly developed teaching strate- gies must be scrutinized for actual effectiveness. Rather than assum- ing a method to be effective, it should be examined with students in an actual academic task. For example, Reid and Hresko (l98l) have re- commended the use of rhythmic grouping as a tool for aiding memory: "Notice, for example, how often television commercials and pop songs take advantage of rhymes and rhythms . . . . Children are . . . re- 3 While this recommendation is found in sponsive to these techniques." a general text on learning disabilities, this particular strategy sug- gestion is not supported by specific research findings. Another example of currently utilized musical memory aids are existing educational tapes and records which present multiplication facts with a musical format. Several teachers contacted during the study stated that they use music-related methods; but when asked whether these methods actually aided retention, they were uncertain. It is possible as well that a method which might be motivating and thus effective for a normally achieving students, might actually be over-stimulating or distracting for the learning disabled student. When examining the effectiveness of a teaching strategy, the heterogeneous make up of the learning disabled population creates ad- ditional research questions. Frequently, researchers fail to report specific details of subject selection, forcing the reader to speculate about the characteristics of the test sample (Keogh et al., "Summary Report--UCLA Marker Variable Project," l98l, unpublished). As a re- sult of vague reporting, several problems hamper interpretation of_ l3 research results: (a) the reader has little or no information con- cerning the effectiveness of a teaching strategy as it interacts with specific subject characteristics, (b) the generalizability of the study becomes clouded, and (c) replication of the study becomes next to impossible. In short, research studies must provide adequate docu- mentation of subject characteristics as they interact with specific teaching strategies. To conclude, in order to provide viable programming for the learning disabled, music therapists should critically examine the ef- fectiveness of methodology in view of specific learning needs and sub- ject characteristics. The truly professional music therapist cannot assume interventions to be effective. In addition, because memory is a cognitive skill which affects so many areas of learning (Donahoe and Wessells, l980), the develOpment of effective remediation techniques which will enhance retention deserves more thorough consideration. Assumptions For the purpose of this study, the following assumptions were made: l. Those individuals who have been identified as learn- ing disabled by an individual educational program re- port, prepared in accordance with state and federal regulations, are actually of that disability. In ad- dition, information on diagnosis and severity of im— pairment, as provided by special education personnel, is assumed accurate. I.Q. measurement, as measured by the Slosson Intelligence Test, is considered an l4 adequate screening device to help assure presence of normal intelligence level among selected subjects, a key criterion. Individuals randomly drawn from regular classrooms with no official record of learning disabilities or handicapping conditions constitute normal subjects. Precautions taken to increase this probability in- clude the exclusion of any students with academic problems which might reflect undiagnosed learning disabilities or borderline retardation as reported by classroom teachers. In addition, none of the stu- dents had been referred for special education or had been retained due to academic difficulties. Those learning disabled and normal subjects selected are a representative sample of the 9.0 to ll.9 year old male population for the greater Grand Rapids, Michigan, area with comparable reported characteris- tics. The wide range of learning and behavioral character- istics associated with the broad term learning dis- abilities is accounted for through randomized assign- ment of subjects to specific treatment and control group. The memory task used in this study constitutes a short- term memory task due to the brief intervening time between rehearsal and posttesting and due to the transient active state of the rehearsal and posttest l5 process. The format of the questions most closely re- sembles the paired-associates type of memory task since it requires an association of the item stem (ex., 8x9) with the accurate matching response (ex., =72). It should be noted that because rehearsals are repeated over five sessions, the type of memory process involved probably shifts between strict short-term memory and the indefinite transitional stage which links information to long-term store. Currently, researchers are un- able to conclude with any certainty the exact point at which informa- tion enters long-term store. This hotly debated question requires further investigation; but for the purposes of this study, the follow- ing explanation by Anderson (1980) will be utilized to clarify the contrast of the two phases: The term short-term memory as it contrasts with long-term memory can be confusing. The phrase short-term indicates that some of the information in the active state may be lost when it becomes inactive. Thus, information in short-term memory may not become a permanent part of long— term memory. However, short-term memory can contain in- formation from long-term memory . . . (while drawing on a permanent fact in long-term memory) the fact is active in short-term memory and long-term memory are not oppo- sites; rather, they overlap. Short-term memory refers to a small amount of information that is in a transient, ac- tive state. Long-term memory refers to a large body of information that is relatively permanently encoded. In other words the two termi can refer to different aspects of the same information. Based on this discussion and for purposes of clarification in this study, the memory task within the experimental procedure will be re- ferred to as short-term memory. 6. The melodic-rhythmic patterns used in this study con- stitute musical rehearsal, and the normal speech pat- terns (reviewed by two speech pathologists for l6 normalcy) consitute what will from this point be called verbal rehearsal. Limitations 1. Results of this study cannot be generalized beyond the following group characteristics: male, ages 9.0 to 11.9, 90-120 1.0., and residents of the greater Grand Rapids, Michigan, area. 2. This study does not intend to examine causal factors or specific etiological groupings such as language problems, motor deficits, dyslexia, or minimal brain dysfunction found in the learning disabled population. 3. While attention and motivation are important factors in accurate and efficient recall, these factors are not examined directly. 4. This study does not examine mathematical competency or remediation of a conceptual or procedural nature. It is limited to rehearsal methods of a closed-ended memory task--specifica11y, multiplication facts of the numbers 7 and 8 (single digit multiplication). While mnemonic devices may be used with materials in other academic subjects, this study is limited to retention of multiplication facts. Theoretical Background In the investigation of music as an aid to retention, supporting theoretical premises encompass not only the structure and processing of musical stimuli, but also the memory process itself. In addition, 17 in an investigation of learning disabled subjects, an extra dimension is added to the memory process since these subjects often experience memory problems and require special interventions to ameliorate poor retention. In developing an appropriate teaching strategy to aid retention of the learning disabled, one must consider the following: (a) In the normal memory process, what types of mental activity or manipula- tion of stimuli enhance retention? (b) What are probable sources of memory breakdown among the learning disabled, and how can classroom information be presented or rehearsed in order to facilitate a more effective memory process? and (c) What components of music might aid efficient encoding or retrieval in the memory process? First, consider the normal memory process and methods which fa- cilitate retention. The storage of information is believed to take place in three stages: sensory store, short-term store, and long- term store. In order for information to enter sensory store, atten- tion is an important prerequisite, for without it the entire memory process is short circuited. In this first stage of memory, informa- tion is not retained, keeping a memory trace for only a fraction of a second. During intake, selective attention is given to incoming stimuli and its critical features are analyzed, identified, and inte- grated in the transition process to the second stage of memory, short- term store (Williams, 1973). After stimuli are taken into sensory store, continued rehearsal or manipulation of information is necessary to keep that input alive. According to Miller (1956), short-term memory can assimilate approxi- mately 7 :_2 bits of information in 30 seconds. In order to increase 18 the short-term memory capacity for incoming information, Miller sug- gests that an organizational process called "chunking" takes place. Chunking involves the recoding of information into groups which are then perceived as a unit. This recoding can facilitate the retention of relatively large amounts of information (Donahoe and Wessels, 1980). Short-term memory is not only limited in assimilation capacity, but also in the length of memory trace. Shortly after active rehear- sal ceases (within 10 to 20 seconds) or with the introduction of 10 to 12 new items, the memory trace decays unless the information has been transferred to the next stage, long-term store (Williams, 1973). As far as is known, long-term store is not limited in the ways that short-term memory is--input capacity or trace duration. The primary difficulty is provision of adequate rehearsal or reorganiza- tion of input in order to assure storage and retrieval. It is be- 1ieved that the rehearsal process, including chunking of input, aids transfer of information from short-term to long-term store. This rehearsal process may also be the critical stage for introduction of retrieval cues which will later aid recall (Donahoe and Wessell, 1980; Wood, 1972; Peterson and Peterson, 1959). In view of this previous model, experts in memory have recom- mended various strategies for aiding retention. One suggestion is to chunk or organize bits of information in order to increase short-term store capacity (Anderson, 1980). While there are numerous methods for chunking information, one very successful and seemingly simple method is the use of a mnemonic system (a formula or aid which helps remem- bering). According to Norman (1976), the power of a mnemonic system IQ is the result of reducing long lists of unrelated material into short related liSts. Through chunking, the verbal elements are kept within the span of short-term memory. While mnemonics are often associated with learning pairs of non- sense syllables in a laboratory situation, according to Anderson (1980), the use of a mnemonic strategies has implications outside of research: In many situations we have to associate various combina- tions of terms that do not have much inherent meaning. We have to learn shopping lists, names for faces, tele- phone numbers, rote facts in a college class, vocabulary items in a foreign language, and so on. In all cases we can improve memory if we transfer the task into one of associating the items meaningfully.5 Besides initially organizing the information, Tulving and Donaldon (1979) believe that the mnemonic devices used during the encoding pro- cess may later act as an effective retrieval cue-~"Successfu1 retrie- val depends upon the presence of stimuli that provide information that was present at the time of encoding."6 There are certain drawbacks to mnemonic systems, however. The main objection is that the subject must learn the mnemonic device as well as the information to be learned. While this is true, Norman (1976) points out that through the addition of images or assertions, the original number of meaningful units may be reduced to a smaller, more manageable number. Another drawback is that certain mnemonic strategies are specific to certain types of memory tasks (Wingfield, 1979). While isolated practice of memory strategies is of questionable value, memory, and methods which will enhance retention are of great importance to educators. According to Donahoe and Wessells (1980): 20 Even momentary reflection reveals that learning and memory are in fact different aspects of the same phenomenon. Re- membering involves retaining the effects of experience over time, for learning is said to occur when the events that take place at one time during an organism's life influence the behavior of that organism at some later time. Thus memory is implicit in all learning . . . . To go one step further (sic), memory is necessary for learning. The model of normal memory processes is of great interest to those who work with the learning disabled, for it suggests possible areas of memory "breakdown." This model in turn provides clues for possible teaching strategies to ameliorate these memory problems. Lerner (1976) lists three areas in which memory problems can occur: reception, storage, or retrieval. Poor reception, associated with sensory store, may result from insufficient attending skills. In addition to fluctuating attention, the student may be less effective in filtering out irrelevant stimuli than nondisabled children. Since attention is "probably a prerequi— site for adequate learning . . . inattention and poor short-term mem- ory may be responsible for slower acquisition in learning disabled children."8 Failure of storage, associated with rehearsal methods during short-term store, and finally transmission to long-term store may be a result of poor encoding and retrieval methods. Recent studies have indicated that developmental changes occur in the self-initiation of verbal rehearsal and mnemonic strategies and that learning disabled students may have a lag in this skill (Reid and Hresko, 1981). Bauer (1979) suggests, however, that provision of such strategies, along with explicit instructions for their use, may result in significantly improved recall. 21 Finally, ability to recall stored information may be hampered by inadequate retrieval cues. According to Wood (1972), the discrepancy that sometimes occurs between information that is accessible and that which is available, supports the importance of retrieval cues in the recall process. According to Donahoe and Wessells (1980), the inter- action of encoding and retrieval is an important part of the presenta- tion of stimuli that are to be remembered. For successful retrieval to occur, the stimuli that provide information must be present at the time of encoding. Additionally, Anderson (1960), Hagen et a1. (1978), and Tulving and Osler (1968) emphasize the importance of retrieval cues in aiding recall performance. In considering musical presentation as a potential remedial meth- od to aid retention, one must briefly consider the memory process in view of melodic-rhythmic input. "Insofar as is known, the ability to recall or have memory of, the auditory sensation of music is not un- like the memory process for any other sensory experiences."9 If musical stimuli are actually processed in a manner similar to normal speech, what might justify the musical presentation of informa- tion to enhance recall? While the processing itself might be similar, the content of music has some specific structural characteristics which may facilitate the memory process: (a) rhythmic regularity and (b) tonal organization. According to Robinson and Solomon (1974), rhythm provides a pat- tern or structure which punctuates the reference points to which words or digits can be attached--a temporal cuing system. A rhythmic frame- work can be imposed on seemingly unrelated sequences of information, furnishing new integration and organization. 22 Even in normal speech patterns, a sense of order and unity is provided through inflections and pauses (such as phrases, clauses, sentences, and paragraphs). When a metrically regular pattern is im- posed upon speech, there is not only chunking of words into units or bits, but there is also increased predictability or anticipation of upcoming information through regularity and redundancy (Blooms- letter and Creel, 1977). Metric regularity limits the number of po- tentially correct responses within a particular time frame. This method describes what Bolton and Bower (1965) call the metric mnemonic --a place-keeping feature which helps the listener keep track of the number and order of stressed syllables in each line.* Increased predictability, or ability to anticipate uncoming sounds, is sometimes referred to as expectancy, a concept which plays an important part in melodic, as well as rhythmic, organization of music. According to Carlsen (l98l): Theories in aesthetics (Meyer, 1965) and in perception (Bruner and Postman, 1949; Carlsen, 1969) have been pro- posed which are based upon the concept of perceptual expectancy and its degree or lack of fulfillment. Most persons will recognize the phenomenon of expectancy as that experience of anticipating what will occur within an event even though the event may be unfamiliar one. In music, a strong preparation for an authentic cadence generates an expectancy for finality. When the cadence occurs, our expectancy is fulfilled, whereas if that preparation leads to a deceptive Sadence, our expectancy for closure is left unfulfilled.] *Bolton and Bower (1965) and Wingfield (1979) emphasize the im- portance of rhyming final syllables in each phrase in conjunction with the rhythmic grouping of verbal information. Therefore, one might question the effectiveness of a rhythmic mnemonic which does not in- clude the rhyming factor. It is possible, however, that tonal or- ganizational aids, such as cadences, may assist the rhythmic group- ing's place-keeping function in the absence of rhyming syllables. 23 Within a musical culture, certain melodic fragments become such a reg- ular part of musical vocabulary that these melodic cliches have a high level of expectancy to the listener (Radocy, 1980). According to Williams (1973), when melodies derived from such high expectancy in- terval combinations are heard, there is a high level of association with past musical experience and, hence, a higher rate of retention. Thus, when a melodic-rhythmic framework is imposed upon normal speech as a mnemonic device, two processes are at work: (a) the chunking of information into manageable units; and (b) increased ex- pectation, or ability to anticipate stimuli through metric regularity and highly idiomatic melodic fragments. In summary, in order to develop effective musical mnemonics for the learning disabled, one must consider the following: (a) in the normal memory process, input capacity and transmission to long term score can be aided by rehearsal using mnemonics which organize or chunk input; (b) while learning disabled students seem to lack initia- tive in using verbal rehearsal or mnemonic strategies, yet such stra- tegies, if used appropriately, may enhance retention; and (c) the rhythmic and tonal organization of music may act as a mnemonic device which aids the encoding and retrieval stages of memory. Definitions Cuing An operational definition for purposes of this study conforms with methods used in related memory studies (Anderson, 1960; Hagen et al., 1978; Tulving and Donaldson, 1972). It consists of a portion of 24 the initial stimulus (as it appeared during the rehearsal process) presented as a prompt at the time of desired recall. Echoic Memory Echoic memory is the name commonly used in the field of cognitive psychology for auditory sensory memory. Wingfield (1979) describes it as the experience of being able to mentally hear a sound again. Because auditory events are presented sequentially, changes in fre- quency, intensity, etc., make up the information which may vary tem- porarally. The main purpose, according to Lachman, Lachman, and But- terfield (1979), of echoic memory is to preserve temporal patterns. In the field of psychoacoustics, this auditory memory is sometimes called eidetic auditory memory (Radocy and Boyle, 1979). In the field of learning disabilities, it is often called reauditorization. Etiology The cause or origin of the learning disability, whether it be minimal brain damage, perceptual handicaps, hyperactivity, or any other causal factor is etiology. Mainstreaming According to Lerner (1976), . . mainstreaming is a delivery system that integrates handicapped children into regular classrooms. Central to the mainstreaming movement is the theme that, given the desire, facilities, and reasonable professional pre— paration, the average teacher can learn to educate ex- ceptional youngsters in the regular classroom with the support and consultative services of special education personnel (Birch, 1974, p. l).II 25 It includes the concept that each child should be educated in the least restrictive environment suitable for their educational and re- lated needs. Mathematical Facts and Calculation For the purpose of this study, this includes onlylearningcyf bas- ic facts of single digit multiplication. It does not include proce- dural methods such as borrowing and carrying. Melodic-Rhythmic Mnemonics These mnemonics display two organizing features: rhythmic regu- larity or meter and tonal melodic sequences. Melody (Structural Characteristics) In its broadest sense, melody means any succession of single tones (Ortmann, 1926) . . . the pitch and durational rela- tionships among the tones contribute to each melody's indi- viduality. A particular tonal and temporal structure cre- ates a constant melodic contour that we perceive and to which we respond. The melodic contour is a Gestalt or holistic pattern to which we respond.12 Meter According to Cooper and Meyer (1960), "Meter is the measurement of the number of pulses between more or less regularly recurring ac- cents."13 Modeling For purposes of this study, modeling refers to the process in which the experimental tester shows an example of desired cognitive behavior in order to encourage similar behavior from the subject. 26 Rehearsal This term . . denotes the repetition of a verbal item--either im— mediate or delayed, silent or overt, deliberate or volun- tary. The initial perception of a stimulus probably must also qualify as a rehearsal.14 Rote Learning The phrase rote learning refers to both a process and a product. The process is one of repetition of the same sequence of words, as in learning a poem or the lines of a play or the basic facts of addition or multiplication. The product is a ready recall of the exact word or number sequences that were learned. Rhythm The New College Encyclopedia of Music defines rhythm as "the organization of music in respect to time." According to LaRue (1970), it results from "changing combinations of duration and intensity within all elements and dimensions of (musical) growth."16 Short-Term MemorygTask Norman (1976) describes it as the presentation of a specifically ordered list of stimulus items. The use of a short distractor, which can consist of either an irrelevant task or presentation of later stimulus items (for the retention of earlier items), precedes the written or verbal posttest to determine amount and accuracy of recall. Test content often consists of serial lists (lists of digits, unrelated words, or nonsense syllables) or paired-associate tasks, where the subject must recall the correct pairing of two unrelated words. Recall follows rehearsal within a short period of time, thus calling on short-term memory skills. 27 Recall for items can consist of free recall, serial recall fitems must be reported in the same order as presented), ordered recall (in which the items can be reported in any order, but the subject must also report the correct position of each item), and probed recall in which the subject receives a cue for items (sometimes a portion of the item itself). Strategy A particular method, device, or special approach used to aid as— similation of a learning task is a strategy. Tonality While a succinct definition of tonality is inadequate, the defin- ition which appears in the Oxford Concise Dictionary of Music (1964) is "the observance of what may be called 'loyalty to a tonic,‘ i.e., to the key scheme of a composition." The concept of tonality is strongly related to cultural norms in Western music in which a type of a priori arrangement of tones in relationship to one another exists. CHAPTER II SURVEY OF RELATED LITERATURE The use of music as a retention aid for exceptional children has been the focus of numerous studies for over two decades. For example, professionals working with the mentally retarded have noted with sur- prise the ability of clients who generally exhibit limited memory ca- pacity to remember lengthy lyrics if accompanied by music. In re- sponse to this observation, music therapists have carried out syste- matic studies to investigate the potential of music as an aid to re- calling academic information. While the learning disabled have many learning characteristics different from retardates, both groups exhibit problems with memory. Based on the studies of the mentally retarded, music has been sug- gested as a potential memory aid for the learning disabled as well. However, very few studies are available which examine music as a re- tention aid with the learning disabled; thus numerous questions remain. This review of related literature will cover the following sub- jects: (a) normal memory processes, (b) musical memory as it relates to normal memory, (c) psychoacoustic research on musical and musical/ verbal memory of normal subjects, (d) research on musical/verbal mem- ory with the mentally retarded, and (e) special characteristics and needs of the learning disabled, available research findings, and need for further study. 28 29 In this study, music is not acting as an aesthetic entity, as is often a major goal in traditional music education, but rather as a re- hearsal method for aiding an extra-musical cognitive goal--specifical- 1y, improved retention. Therefore, the review of literature encom- passes not only research on musical memory, but also research which examines human memory processes. In addition to literature on normal memory activities, the selection of learning disabled subjects neces- sitates consideration of abnormal memory processes as well, since many learning disabled students may exhibit memory patterns which digress from the usual. In short, pertinent literature reflects the interac- tion of normal human memory processes in conjunction with musical stimuli and the resulting effect on retention of academic information in the learning disabled. Normal Memory Storage and Retrieval The study of human memory is both complex and multifaceted and has been the focus of extensive research by psychologists, biologists, and educators. In educational tasks, since learning involves reten- tion over time, memory is implicit in the learning process (Donahoe & Wessells, 1980). Due to the scope of cognitive research related to the study of memory, it is impossible to review the research in its en- tirety. The literature cited in this chapter will be limited to those pertinent studies of short-term memory and recall facilitation. Though many questions on memory remain, there does exist a gen- erally accepted model of the three stages of verbal memory: sensory store, short-term store, and long-term store. While long-term store has the advantage of unlimited storage capacity and length of memory 3O trace, short-term store is limited both in input capacity and length of memory trace. This limitation is an important factor in the effi- ciency of the memory process; thus, much of the research on retention focuses on short-term memory. Additionally, some researchers see the rehearsal process during short-term store as a crucial activity in transferring information to long-term store (Peterson & Peterson, 1959). Research by Miller (1956) appears at the base of numerous studies on the limitations of short-term memory. His research describes the input capacity of short-term memory as limited to seven (plus or minus two) unidimensional bits of information within a 30-second time span. Despite this seemingly severe limitation, the mind's ability to re- organize or chunk bits of information allows much greater memory capa- city than is initially apparent. Through chunking or reorganizing separate bits of information, large amounts of information can be re- structured into a more easily assimilated input format (Donahoe & Wessels, 1980). In attempts to increase memory capacity, psychologists have studied various strategies for chunking input, including metrical or- ganization. According to Robinson and Solomon (1974), rhythmic struc- ture provides reference points to which digits or words can be attached. Through such rhythmic groupings, the subject can recall ap- proximately as many small groups of information as individual items without grouping (Seashore, 1919). Because memory of auditory events, such as a spoken sequence of words, is displayed in a temporal manner (Underwood, 1976), provision of easily retained groupings can assist auditory or echoic memory. 31 Bower (1970b) describes rhythmic chunking in the following manner: One of the fundamental strategies used by people in learn- ing a long series of symbols is to segment it into several smaller chunks or groups. Even a monotonous series can be grouped by the person imposing on it his own rhythmi- cal stress or pauses . . . . This segmentation of a string may serve several useful purposes. First, the segments are small and are themselves easily learned, so the person's job is reduced to seriating a smaller number of units than was true before groupings were im- posed. Second, the size of the successive groups and their order in the string may conform to a simple, repe- titive, rhythmical pattern which helps the person to plan and execute his serial reproduction (cf. Neisser, 1967). The beats of the rhythm might serve as subjective anchor points to which digits are attached, and the pauses serve as phrase markers delineating major constituents. While chunking, or rhythmic grouping, happens in normal speech patterns (Restle, 1972), the regularity or constancy of an imposed rhythmic pattern is advantageous in facilitating recall due to what Bower (1970b) terms a "reallocation hypothesis." When repetition of input contains the same group structure as previous presentations, this input "string" is allocated to the previous memory "location," and that trace is strengthened. In contrast, inconsistent or variable input of the input "string" fails to show the same immediate recall improvement (Dowling, 1973; Waters & Waters, 1979; Bower, 1970b). According to Boomslitter (1977), metrically regular language ex- tends the capacity of language perception because it is regular and repetitious, thus making the expectation of the time scheme easier. "Rhythm's importance to educators comes from its function as organizer of meaningful thought groups . . . . In metrical poetry, language is strongly organized (and) abundantly connected."18 Besides the chunking of information, verbal rehearsal also takes place in short-term store. This activity not only prevents trace 32 decay, but may also aid transition of information to long-term store (Anderson, 1980). Whether this rehearsal is silent inner speech or overt audible speech, verbal rehearsal is a particularly effective mode when dealing with language (Norman, 1976). Norman differentiates between coding characteristics of short- term store, which are predominantly acoustic, and those of long-term store, which are largely semantic. Since letters or digit groups have little semantic content, acoustic, or articulatory coding may predomi- nate in this type of memory task. There is some disagreement as to how coding takes place in the memory trace process. Bower (1970b) suggests that the memory trace established is not of an acoustic nature (monitored by the subject's own ears), but rather a proprioceptive muscle memory in the mouth re- gion which recalls the physical properties of various phonemic shapes. The act of verbal rehearsal requires active attention to the informa- tion, thus circumventing interference of competing stimuli. The successful memory process consists not only of information intake, but also successful retrieval of that information at the de- sired moment. Tulving and Donaldson (1972) and Tulving and Osler (1968) have found that retrieval cues facilitated to-be-remembered words, but only when those cues were stored in conjunction with the information during the learning process. As Donahoe and Wessells (1980) explain this process, "Successful retrieval depends upon the presence of stimuli that provide information that was present at the time of encoding."19 In summary, facilitation of memory storage and retrieval is aided by the following procedures: (a) increased short-term memory input 33 capacity through chunking or grouping of incoming information, (b) consistent presentation of repeated input in order to strengthen available memory traces, (c) adequate rehearsal to prevent decay and to promote transfer into long-term storage, and (d) provision of ade- quate cues during the storage process which will later aid informa- tion retrieval. Musical Memory as It Relates to Normal Memory Memory for musical stimuli falls into several research categories: memory for musical stimuli without accompanying speech, memory for musical stimuli and speech or sung speech, and musical stimuli as an aid to memory. In the selection for musical stimuli which might aid retention, it is important to understand the basic organizational structures of musical stimuli without speech, focusing on the pro- cesses which enhance recall. The memory processes of stimulus input, establishment of a memory trace, and storage are believed to take place in musical as well as speech recall (Bartlett, 1980). As is the case in hearing normal speech, the various sensory impressions must be organized into mean- ingful units in order for recall to take place. This organization is facilitated in music through rhythmic groupings, melodic contours, and tonal structure. According to Radocy (1980), "Rhythm connotes some sort of recurring and, within limits, predictable event. . . .(In music, rhythm provides) the pattern of organized sounds and silences. To perceive a rhythm is to relate the respective durations in order across time.20 34 This rhythmic grouping provides the same function as in verbal material by helping the listener organize and recall pitch sequences. These metric groups tend to have more regularity and redundancy, how- ever, than the pauses or tempo fluctuations that occur in natural speech patterns. Through this metric regularity or constancy, anti- cipation of upcoming sounds across time is facilitated. In addition to rhythmic groupings, music is organized according to pitch sequences. When such pitch sequences are recognized by the listener as an aesthetic whole, or a musical Gestalt, this pitch se- quence is considered to be a melody. While melodies vary from culture to culture,Lundin (1967) notes three general melodic characteristics: (a) propinquity (closeness of pitch between successive intervals), (b) repetition (tendency to return to certain pitches), and (c) fi- nality (extensive use of authentic or half cadences for musical clo- sure). Recognition of a melodic sequence is the perception not only of specific intervals, but also pitch range and contour.* Research on recall for melodies is concerned largely with the re- call of pitch sequences, identification of a tone as belonging to a sequence, and recognition of altered melodies (Radocy, 1980). Exactly what makes certain melodies more easily recalled has proven difficult to pinpoint. In long-term memory for familiar melodies, exact inter- val as well as contour plays an important role. Contour, however, appears to be a greater facilitator of short-term memory than specific interval recognition, especially for transposed melodies (Dowling & *Contour refers to the pattern of successive pitches which makes up a melody. 35 Fujitani, 1971). In comparing the recall of various melodic contours, those melodies with M-shaped contour appear more easily recalled than melodies with V-shaped pitch relationships (Long, 1977). Other factors that affect recall for tones within a melodic se- quence include the delay time between the stimulus termination and subject response, the tone's position within the entire sequence (serial position), and the sequence length. Delay of response (Williams, 1975) or interfering tones played between the stimulus and the response (Deutsch, 1977) produce deleterious effects on recall ef- ficiency. Serial position effects demonstrate that within pitch se- quences, the most recently heard pitches are recalled most easily and middle sequence pitches are recalled with the least accuracy. In general, the longer the pitch sequence, the more difficult recognition of a single tone within the sequence becomes. This serial position effect, sometimes called primacy-recency effect occurs not only in re- sponse to musical sequences, but also in word sequence tasks as well. This phenomenon is an excellent example of how musical memory is simi- lar to memory for verbal stimuli. While many questions remain concerning processing and memory for melodic stimuli, it is suggested that tonal sequences are more easily recalled than atonal sequences (Zenatti, 1975; Long, 1977). Tonality is a concept closely related to melodic organization, since certain melodic pitch relationships or interval patterns appear over and over within a tonal framework. Zenatti describes tonality (or tonal accul- turation) as a feeling of familiarity with melodic formulas, chord progressions, and intervals frequently used in our Western music. 36 Deutsch (1977) describes these pitch relationships as organized in a hierarchical fashion with a priori probabilities established for both linear and harmonic pitch sequences. The typical musical patterns of a culture "serve as overlearned perceptual 'frameworks' in terms of which melodies are encoded.”21 This framework enhances anticipation of upcoming musical events as well as recall for melodic sequence. According to Zenatti, this tonal acculturation of melodic patterns is developed around six to eight years of age.* Therefore, in children past age eight, tonal organization becomes a potential aid for melodic recall. Based on these studies, the choice of melodic material which might aid the retention of verbal information would be metrically re- gular melodies based on atonal structure. However, the implications we can make for music in conjunction with speech are limited for the following reasons: (a) according to Mayo (1981), the processing of melodic patterns which accompany sung speech may not be exactly like either the processing of speech or music without speech; and (b) many of the studies for melodic memory actually investigate memory for iso- lated pitch as opposed to full melodic Gestalts. In view of these limitations, it becomes important to examine those studies which in- vestigate music as it interacts with speech. *Although music has at times been called "the universal langu- age," in terms of tonal acculturation, this is far from the case since major cultures have their own established melodic formulas and harmonies. 37 Psychoacoustic Research on Musical and Musical/Verbal Memory of Normal Subjects Investigation of music as an aid to recall of music with speech includes studies carried out by Prickett (1974), Jellison (1976), and Sims (1981). Prickett isolated the rhythmic component of music as a potential aid in recalling serial lists of digits. Using Miller's (1956) research in short-term memory limitations, Prickett limited digit lists to seven items. The subjects, first grade children, were asked to repeat the lists after listening to those lists with either a steady quarter note pulse or an imposed rhythmic pattern.* The im- posed rhythmic pattern resulted in significantly greater retention at the .05 level. Prickett concluded from the study results that imposed rhythmic groupings aided retention of verbal input in a serial list format. Sims' (1980) investigation is similar to Prickett's research in that both studies utilized serial lists of seven items and both stud- ies isolated rhythm as a treatment mode. Sims' research differs in several aspects: nonsense syllables were used as opposed to digits, adult subjects were selected, and the subjects were then classified as music major or nonmusic major. In addition to comparing recall following verbal or rhythmic treatment, melodic presentation was also compared. Melodic sequences consisted of a self-composed pitch pat- tern using five pitches (A through E in A major tonality) presented either in a steady quarter note pulse or superimposed upon the same rhythmic pattern which appeared in the rhythmic treatment condition.** * SNJ M JVJ N 711 “2.1 DJ DII 38 Both rhythmic and melodic-rhythmic presentation proved nonsignif- icant in comparison to verbal presentation for both musicians and non- musicians. The only statistical significance resulted from differ- ences in recall accuracy among lists.* Because the Prickett and Sims' studies differ on several variables, interpretation of the difference in results becomes difficult. It is possible that adults were less responsive to rhythmic grouping than were children. In view of re- search on memory processes, it is more probable that the rhythmic or- ganization acted more effectively as a retention aid with digits as opposed to nonsense syllables. According to Wingfield (1979), mnemon- ic systems are specific to certain types of memory tasks. Thus, it is also possible that the method used for retrieval of information was more difficult in one of the studies. For example, free recall may be less difficult than recall which requires a specific order of items (Norman, 1976). Additionally, it is possible that the rhythmic pat- tern used in Prickett's study was more effective in grouping the seri- al lists than the rhythmic pattern used in Sims' research. In sum- mary, the many variables which differ between the two studies makes comparison of studies and their outcomes extremely difficult. Jellison's (1976) research studying recall of digit lists is unique among these studies, in that verbal and musical input were carried out through a dichotic listening format. In this type of for- mat, left and right ears are presented with different input through *Despite the supposed lack of relationship among items on the lists, it appears that subjects, in order to provide some sense of meaning or organization, imposed their own relationships among these nonsense syllables. 39 the use of earphones. Through this method, cerebral dominance for verbal or musical stimuli can be examined. The memory task for this study consisted of lists of seven digits either spoken or sung (with each digit presented in a steady pulse every half second) to melodies composed by Jellison. These melodies were arbitrarily composed using a limited number of pitches (A3 to G4) in a tonal framework (C major tonality). Among the findings, sung speech did significantly facilitate digit recall for both musically trained and untrained subjects, though musicians showed a slight su- periority of recall over nonmusicians in the musical treatment condi- tion.* While these studies all examine normal subjects' recall of seven item serial lists in verbal, rhythmic, melodic, or rhythmic/melodic presentation modes, the results are inconsistent. Only Prickett's study supports the effectiveness of rhythm alone as an aid to short- term memory. While Jellison's study found melodic patterns to aid re- tention, Sims' study did not. Once again, perhaps Jellison's choice of digits for a memory task was more appropriate for a musical mnemon- ic than Sims' choice of nonsense syllables. Perhaps replication of these studies may shed further light on the results, but at present these research results remain inconclusive. Additionally, implica- tions for populations that exhibit serious memory problems must be drawn with great reserve. It is therefore important to consider re- search findings more directly involving exceptional children. *The following criteria defines a musically trained subject: (a) music major, (b) instrumental or vocal studies for at least four out of the last five years, and (c) participation in an ensemble or indi- vidual performance within the past year. 40 Research on Musical/Verbal Memory with the Mentally Retarded The bulk of research using music as an aid to retention has been carried out with the mentally retarded. As in the research with normals, serial lists have been used to examine short-term recall. Studies by Wilson (1971) and Myers (1979) both utilized paired- associate lists in verbal, melodic-rhythmic, or rhythmic presentation. Myers' study (1979) was based on the theory that intensive train- ing and rehearsal, grouping and mediational techniques, and overlearn- ing material would improve performance of memory tasks. Melodic- rhythmic presentation was used as a mediational technique and compared to rehearsal using normal speech. The results showed no significant difference between musical and verbal rehearsal. Myers added, how- ever, that since the musical presentation did not impede recall, it, therefore, might provide some variety in teaching approach. In Wilson's study (1971), paired-associate lists were presented with short, self-composed rhythmic patterns, and additionally with those same rhythmic patterns superimposed on self-composed melodic sequences. Visual aids were utilized in conjunction with the presen- tations. Subjects were not required to overtly rehearse the lists. Rhythmic presentation alone resulted in no increased retention com- pared with speech presentation. Though not statistically significant, a trend toward improved retention was reported from the melodic- rhythmic presentation. The memory task in Crowe's study (1977) differs from that of Wilson or Myers' in the choice of memory task. Rather than selecting serial lists, Crowe presented a specific sequence of colors in a song 41 and also in verbal instructions. Following this presentation, sub- jects were asked to place four different-colored geometric shapes in the proper positions based on the previous instructions. During the experiment, subjects were exposed to two presentations of the color sequence in one of the following orders: verbal/verbal, verbal/mus- ical, musical/musical, and musical/verbal. Subjects were not required to sing or repeat the information during the treatment. Neither the mode of presentation (musical versus verbal), number of presentations (one or two) nor interaction of presentation methods (musical with verbal) resulted in statistically significant results. The only fac- tor to show statistically significant results was the individual sub- ject's functional level: higher mental age of subjects correlated positively with higher recall in all presentation methods. The results of these three studies provide minimal support for use of music as an aid to retention. While musical presentation did not interfere with recall, only musical-rhythmic presentation in Wilson's study showed a trend toward improved retention. Several points should be considered, however. Repetition is a mainstay of academic programming for the mentally retarded. It is possible that such brief exposure to stimuli did not provide adequate repetition or rehearsal in order to gain measureable improvement. Secondly, because the subjects were not required to overtly rehearse the memory task, the attentional aspect of memory may have been inadequate to facili- tate initial sensory store and maintenance of information in short- term store. Thirdly, the absence of rhyming syllables may have ren- dered the musical mnemonic ineffective. Fourthly, the abstractness of the memory task (paired associate lists which have little or no 42 meaningfulness as a unit), while moderately difficult for a normal subject (who may have the cognitive skills to devise his/her own or- ganizational structure), may constitute a much more difficult memory task for a mentally retarded subject. It is worth noting that while the musical presentation did not show statistically significant improvement, neither did this method impede recall. Studies by Madsen and Forsythe (1973), Steele (1968), and Madsen, Cotter, and Madsen (1968) have shown that music can act as an effective motivator in academic situations. Therefore, if musical methods do not impede recall and in addition have a motivating effect, musical methods may still merit use as an educational tool. This issue deserves further investigation. Studies by Isern (1958) and Lathom (1970) which examine music as an aid to retention differ from the studies by Crowe, Wilson, and Myers in the type of memory task and recall response. Rather than utilizing serial lists or color names, information of a meaningful na- ture was presented through songs, stories, or poetry. Isern (1958) hypothesized that the "feeling state"* of music may produce improved memory in retarded children due to increased reinfor- cement, increased attention, and greater organization of the listening experience. In order to test this hypothesis, a song was sung to each subject. Immediate recall was tested by the number of song items recalled. A story was then presented, and recall tested in the same manner as the song. Testing was done both three days later (recent recall) and three months later (remote recall) for recall of song or *Isern provides no definition or detailed discussion of this concept. 43 story items. In all tests, musical presentation resulted in signifi- cantly greater recall than story presentation, including a greater level of accuracy among those items recalled. It is difficult to draw conclusions from these results since the researcher provides very little detail about subject selection or pro- cedural methods. Furthermore, no information on song or story selec- tion or methods for collecting responses is available to the reader. Lathom's (1970) study "Retarded Children's Retention of Songs, Stories, and Poems" also uses stories and songs as methods for presen- tation and does provide much fuller detail on procedure and subject selection. Subjects chosen for this study included retarded children of mean age 13.5 years with a Measured Intelligence of Level II or Level III (1.0. ranging from 25 to 54). All songs, stories, and poems used in the study were selected by experts in music and children's literature who verified that these materials were interesting and suitable for the mental age of the subjects. The amount of repeti- tion or redundancy in all materials were judged by two experts in in- formation theory. Treatment consisted of six half-hour periods of training with two presentations each of song, story, or poem. Visual aids were used in conjunction with all presentations. On the following day, indivi- dual tests for retention were given which included recall, chaining, and verbal concepts. The retention of information within the songs as compared to stories or poems was statistically significant at the .01 level. While the level of Measured Intelligence was an important factor, a higher level of Measured Intelligence resulted in consistently greater 44 retention scores across all groups. The degree of repetition within materials was found to be an important factor: Lathom concluded that redundancy facilitated recall. Lathom based her research hypothesis on information theory. Ac- cording to this theory, a message which contains novel or unexpected input is high in new information, as opposed to redundant or predic- tive stimuli, which adds little new information. In successful com- munication, a balance between totally novel and highly redundant in- fbrmation level is desirable. Some degree of redundancy short of monotony facilitates understanding, while a novel stimulus provides an adequate interest level. In music, redundancy results from idiomatic, culturally familiar melodic or rhythmic sequences, while the wide variety of available musical vocabulary provides new information. Ac- cording to Lathom, this balance is ideal, making music an excellent facilitator of meaningful communication. Similar to the studies reviewed on retention in normals.(Prickett, 1974; Sims, 1981; Jellison, 1976), results of those reviewed studies with the mentally retarded (Wilson, 1971; Myers, 1979; Crowe, 1977; Isern, 1958; Lathom, 1970) are inconclusive. Certainly many factors might contribute to inconsistent findings: subject selection, re— hearsal method (active or passive), amount of exposure to stimuli, chosen response mode, and the nature of the memory task itself. In reviewing these studies done with the mentally retarded, it is important to note the difference in meaningfulness among the memory tasks. In the studies by Lathom and Isern, to-be-remembered informa- tion consisted of entire songs and stories with meaningful information as opposed to the nonmeaningful serial lists in the research of Wilson 45 and Myers. Researchers in the field of memory know that meaningful information is more easily recalled than that which is not. In addi— tion, various types of mnemonic strategies are more or less appropri- ate for different memory tasks (Wingfield, 1979). Thus the teaching strategy selected for remediation should be evaluated in terms of specific task requirements. Wingfield (1979) makes the following recommendation for an effective musical mnemonic: "If a rhyme is good, a rhyme with a little melody is even better."22 The absence of rhyming in the paired-associate tasks (Wilson, 1971; Myers, 1979) may affect the strength of the musical mnemonic. In complete songs, such as those used in the studies by Isern (1958) and Lathom (1970), final rhyming syllables often appear at the end of a musical phrase. Equally important as the choice of appropriate strategies for specific tasks is the need to evaluate teaching strategies in view of individual student characteristics and needs. To exemplify, when teaching strategies are selected for the learning disabled, based on research with the mentally retarded, there is a failure to acknowledge differences in learning patterns or problems between these two popula- tions. Therefore, it is necessary to consider musical remediation methods specifically for the learning disabled student. Special Characteristics and Needs of the Learning Disabled Essentially, the term learning disabled covers a heterogeneous group of students who, despite normal 1.0. and sensory acuity, have serious learning problems. While this diagnosis covers a multitude of academic deficiencies, there are some cognitive areas which have been 46 identified as problematic to a large number of this population. One such problem area is memory. The fact that some children who are unsuccessful learners have poor memories was observed long before learning dis- ablilities was recognized as a field of study. Yet there is now a renewed interest in the role of memory in learn- 122t,'2n§“‘ink$§3§ $5 lfilfizle°£§lfifil'3§3"'th'" ' “mm" Because memory is so interrelated with learning (Donahoe & Wes- sells, 1980), problems in this area can affect a wide range of academ- ic skills. Johnson and Myklebust (1967) have related memory impair- ment to reading difficulties, since reading requires both auditory and visual memory. They note that problems of attention and memory, while differentiated, often occur together, and that memory span is often deficient in the learning disabled child. They make the following suggestions for cues: (a) cuing the child to "wait, listen" to facil- itate attending; (b) emphasizing the rhythmic sequence of words to improve auditory sequentialization; and (c) teaching the student "to organize, revisualize, or reauditorize a task before starting it."24 Memory impairment may not be limited only to reading problems, however. According to Webster (1979), there exists a significant dif- ference in short-term memory capacity between those children with ade- quate math achievement and learning disabled students who are perform- ing below grade level. He cites the ability to recall number tables as one example of recall difficulty and further suggests that the problem may stem from failure to use efficient coding mechanisms in the short-term memory process. The inefficient coding procedure discussed by Webster is also suggested by Torgesen (1979, 1980), Bauer (1979), and Tarver et a1. 47 (1976, 1977). Torgesen found learning disabled students much less likely than normal children to use verbalization and mnemonic aids spontaneously during study periods or sequential memory tasks. Torge— sen sees this inability to generate spontaneously verbal mediational strategies as a develOpmental lag which results in memory problems and failure to approach cognitive tasks in an efficient manner. Bauer (1979) concurs that deficient mnemonic processing, coupled with inattention, may be important in that these impede short-term memory in learning disabled children. Poor rehearsal strategies, which initially play an important role in short-term memory, may later hamper the transfer of coded information to long-term storage, since there is speculation that rehearsal or manipulation of sensory input helps this transfer. Bauer suggests that the short-term memory pro- cess may be aided through appropriate use of mnemonic strategies. Verbal rehearsal strategies are a focus of studies by Tarver, et al. (1976, 1977). According to Tarver, selective attention and spon- taneous use of verbal rehearsal strategies are considered basic psychological processes which develop with age. In the learning dis- abled, this process may lag several years behind. And though the learning disabled child may eventually develop equal attentional and rehearsal skills, the academic information lost in the interim hampers academic success (Tarver et al., 1977). In experimentally induced verbal rehearsal strategies, however, selective attention became more efficient, suggesting that environmental manipulation through appro- priate strategies and reinforcement may be successful tools to aid re- tention. Tarver et a1. (1977) emphasize the use of repetition, 48 rehearsal, guided practice in which the instructor models the use of the rehearsal strategy, and reinforcement with the learning disabled. According to Reid and Hresko (l98l), failure to use rehearsal strategies may constitute a production deficiency. This type of de- ficiency is characterized by a person who has an ability, but fails to use it at the appropriate time. As an example, a learning disabled child may have learned the spelling rule "j_before g_except after 9," but may fail to apply this little rhyme spontaneously at the appropri- ate time. Rather, the student will tend to use such a mnemonic appro- priately only when instructed how to do so. This same production deficiency can effect rehearsal in the mem- ory process. Conscious rehearsal is often needed to retain unfamiliar information for short periods, but learning disabled chil- dren do not always recognize that verbal repetition is a ifiéatfigittfliyt§3"h‘ésio'3.1223213?" “'9 '"f°"""'°"' y U 9 The use of verbal rehearsal can act not only as a rehearsal meth- od, but also as an aid to self-regulation (Keogh & Margolis, 1976; Weithorn & Kagan, 1979; Meichenbaum, 1977). Because many learning disabled children show poor organizational skills and lack selective or sustained attention skills or respond impulsively, remedial pro- grams cannot be limited to orders of "slow down." As Keogh and Margo- lis point out, strategies must aid in developing more appropriate techniques to use the time fruitfully. They list the following strategies as helpful in developing internal controls: verbal self- instruction, increase of critical cues, modeling or demonstration, verbal analysis and emphasis, and novel or sudden stimulus. Weithorn 49 and Kagan support the use of verbal self-regulation--especial1y for children who tend to respond to a task impulsively. While no single strategy can provide a panacea for the varied problems of the learning disabled, these studies suggest the following approaches which may aid impulsive responses, ineffective cognitive strategies, and inefficient short-term memory strategies: (a) overt verbal rehearsal as an aid to efficient coding of incoming information and as a self-monitoring device; (b) imposed organization such as special or temporal series when presenting lesson material (Reid and Hresko, 1981); and (c) provision of cues to help retrieval of that in- formation, providing reinforcement to actually use a strategy at the appropriate time. It is not sufficient (for organizational strategies) to be present. Children must be told that it is there . . . Children, especially those with memory deficits, tend not to discover and impose structure for themselves. Based on these suggestions, in what way can music fit into mem- ory strategies? Do current research findings provide any answers? Unfortunately, current research on using music as an educational aid for the learning disabled is very limited. One of the few avail- able studies includes Roskam's research (1977) incorporating music into a reading readiness program. Working from the premise that many children with language-based learning disabilities exhibit minor de- ficits in auditory discrimination and memory, Roskam assigned 36 learning disabled children (ages six to nine) to one of the following treatment groups: (a) musical activities stressing auditory sequenc- ing or discrimination, (b) the usual prescribed learning disabilities remediation for language development (as prescribed in their usual 50 classroom program), and (c) a combination of the music activities and prescribed language development remediation utilized in groups (a) and (b). Pretests and posttests for reading readiness were based on verbal and nonverbal auditory awareness, reading recognition, comprehension, and spelling. Additional data on behavioral appropriateness were also compiled, since many of the subjects also displayed some emotional or behavioral maladjustment. While no statistically significant differ- ence was found between groups, the mean scores for the musical activ- ity group were highest for behavioral improvement and all posttest areas except reading comprehension. Combined music and language de- velopment showed the highest mean scores in reading comprehension. This particular study relies heavily on the theory that percep- tual skills can be remediated in an isolated fashion. Recent re- search in learning disabilities has failed to support this approach. Rather, the most direct and closely related remediation to a learning problem seems the most effective. Besides the fact that Roskam's study is based on remediation of perceptual skills, it also does not address the cognitive area of mem- ory. Therefore, its relationship to the present study is quite lim- ited. Shehan's (1981) research does, however, give information on retention in that it focuses on the use of verbal, visual, musical, or mixed mediational techniques in order to examine whether not one mediational approach acts as a more effective aid to retention. In Shehan's study, 16 distractible learning disabled students (ages seven to ten) were exposed to four lists of seven pairs of words presented in the following formats: (a) musical (sung to the interval 51 of an ascending fourth in an even rhythmic pulse), (b) verbal (spoken at the same even pulse), (c) musical/visual (in which outlined pic- tures of the words accompanied the sung words), and (d) verbal/visual (in which the picture accompanied the spoken words). Responses were gathered by giving each subject the first word of a pair and asking him/her to state the second word. The mean scores for each presentation format were as follows: musical/visual--2.3l, verbal/visual--2.56, verbal--.69, and musical—- 1.0. The results, significant at the .01 level, indicated that com- bined use of musical/visual or verbal/visual teaching strategies best facilitated the retention of paired associates. While Shehan's study does examine music as an aid to retention, it has several procedural limitations which should be noted: (a) the use of only two pitches (an ascending fourth) as a musical stimuli is a very limited sound source. One could certainly question whether or not only two pitches provide an adequately lengthy pitch pattern or enough sense of finality (through a cadence) to form an actual melodic sequence or aesthetic whole. In fact, Shehan's recommendations for further study suggest the investigation of’more extensive melodic stimuli in conjunction with the memory task; (b) Shehan's study pro- vides little information on subject selection. Beyond the subjects' ages and their tendency toward hyperactivity and distractibility, there is minimal information on the subjects to infer the generaliz- ability of research findings; (c) the study focuses on the efficacy of visual, verbal, and musical mediation, yet provides little supporting theoretical basis to explain why musical mediation might prove to be an effective strategy; and (d) the memory task and rehearsal 52 procedures were limited to one presentation of a paired associates' list. There was no examination of the strategy's effectiveness over time, nor was there provision for overt verbal rehearsal. Thus, while the results of Shehan's study point to a slightly greater mean score for musical over verbal presentation, and signifi- cantly greater recall resulting from mixed mediational techniques, Shehan's study provides limited information on the effectiveness on the effectiveness of music as an aid to retention. In summary, while several researchers recommend the use of music as an aid to retention, the inconclusive results of many of these studies suggest the need for further investigation, particularly in view of the following issues related to the learning disabled popula- tion. 1. Further study must be done specifically with the learn- ing disabled population rather than relying on research findings gleaned from studies with the mentally re- tarded. 2. Due to the heterogeneous nature of the learning dis- abled population, there is a need for more detailed communication of specific subject characteristics. 3. In view of the suggested value of verbal rehearsal strategies as aids to encoding and self-monitoring behavior, the experimental procedure should provide for investigation of active, overt rehearsal of musi- cal mnemonic devices. 53 4. In addition to providing the musical mnemonic, the re- searcher must consider the importance of thorough and clear presentation of a strategy's appropriate use. 5. Further research must extend beyond examination of nonmeaningful information, such as serial lists, to the investigation of strategy effectiveness within ac- tual academic tasks. 6. More extended measurement of rehearsal and subsequent recall is needed in order to more clearly examine the long-term effects of a possible teaching strategy. The present study addresses these research needs, going beyond single rehearsal of serial lists to a repeated measure of recall for actual academic tasks with learning disabled students. The research procedure not only provides for the examination of musical rehearsal, and its effectiveness with both normal and learning disabled stu- dents, the manner in which a strategy is taught is also considered. In order to adequately take into account the heterogeneous nature of the learning disabled population, subject characteristics in regard to academic, behavioral, and socio-economic factors have been reported. CHAPTER III PROCEDURE The purpose of this study was to investigate the effectiveness of musical rehearsal as an aid to retention in normal and learning dis- abled students. While other research studies have examined music as an educational aid, this study attempted to provide further knowledge on music as a retention aid for learning disabled students by (a) more specifically addressing the learning disabled population and indivi- dual academic and behavioral characteristics of this group, and (b) providing procedural methods which examine the effectiveness of music for retention in conjunction with actual academic tasks as opposed to nonmeaningful serial lists. Repeated measures furnish an opportunity to examine the impact of a musical rehearsal strategy over time. Subject Selection Sixty subjects were drawn from two populations: 30 learning dis- abled male students and 30 normal male students, ages 9.0 to 11.9 (grades 3, 4, and 5). Selecting students in this age range seemed most appropriate for the memory task studied herein-~namely, the mul- tiplication tables, since single digit multiplication is often intro- duced and practiced in the school grades which include this age range. Additionally, this age range was considered to be advantageous in that several related research studies on verbal rehearsal techniques (Tar- ver et al., 1977; Hallahan et al., 1974) selected a similar age range 54 55 and developmental characteristics. Tarver and Hallahan both consider short-term memory strategies, especially verbal rehearsal, to be a developmental characteristic. Because of the predominance of male learning disabled students reported in the literature (80%--Farnham- Diggory, 1979), only male subjects were included in the study. In order for a subject to qualify as a learning disabled group member, the subject had to be officially classified as learning dis- abled by an Individual Educational Program committee report and cur- rently enrolled in either a self-contained or resource room program for the learning disabled. Normal students selected for the study included students enrolled in regular classrooms and having no official record of handicapping conditions. Furthermore, classroom teachers were asked to include only those students who, in their judgment, possessed no academic de- ficiencies or serious learning problems. In order to insure that both normal and learning disabled stu- dents met normal intelligence criteria, all subjects were screened using the Slosson Intelligence Test, an individual standardized intel- ligence screening device. This particular screening device was chosen for the following advantageous features: (a) length--adminis- tration time is brief (10-30 minutes); (b) it can be administered by teachers or responsible adults, as opposed to some tests which require a psychologist or psychometrist for administration; and (c) it pro- vides more sensitive measurement of individual 1.0. level than a group test, and for the purposes of this test provided adequate screening information. 56 Himelstein and Hunt (Seventh Mental Measurement Yearbook, edited by O. K. Buros, 1965) both describe the test as a valuable and quick screening device which is easily administered. Himelstein does sug- gest caution concerning reliability and validity, but adds that there is significant scale and item content similarity to the Stanford-Binet Intelligence Scale. Questions and response mode are verbal, which is advantageous for those learning disabled students with severe reading problems. The item content is primarily mathematical reasoning, vo- cabulary, auditory memory, memory, and information (Hunt, 1965). The publisher reports a reliability coefficient (test-retest) of .97 with a Standard Error of Measurement of 4.3. The intended mean or average for the Slosson Intelligence Test (SIT) is 100, with an aver~ age standard deviation of 17. The range for low-average to high- average 1.0. range is 80 to 119. For this study, a lower cutoff point of 90 was chosen in order to filter out possible "slow learners" who may have been inadvertently placed in a learning disabilities program. In addition, this cutoff point provides an extra margin of safety in consideration of inaccuracies due to the Standard Error of Measurement associated with this testing device. An upper cutoff point of 120 was selected in order to limit subjects to upper normal intelligence range. The test manual for the SIT includes screening of the learning disabled as one use for this test, but adds that such individuals may exhibit "scatter" or inconsistent performance in various portions of the test. In order to provide testing consistency, all screening was carried out by one test administrator--a certified teacher in learn- ing disabilities. 57 In addition to 1.0. requirements (90-120 on the SIT), all sub- jects were required to have previous exposure to multiplication con- cepts and tables as reported by the classroom or resource room teach- er. Subjects with 76% or higher accuracy on the pretest were dis- missed from the study to insure that subjects would not ceiling out. Because of the heterogeneous nature of the learning disabled pop- ulation, it was important to consider not only 1.0., age, and school grade, but also other variables such as race, academic problem area (reading, math, written, language, etc.), behavioral characteristics, and socio-economic status. These characteristics were obtained through a data collection instrument (Appendix A-B) filled out by the resource room or classroom teacher. The following characteristics were tallied for all subjects: parent occupation, race, whether or not the subject was bilingual, at- tendance at public or private school, grade, chronological age, and intelligence quotient as measured by the SIT. For learning disabled students, the following additional charac- teristics were noted by the resource room teacher through a checklist (Appendix A) of academic and behavioral characteristics: type of learning disability (i.e., reading, math, written language), educa- tional placement (resource room or self-contained room), severity of underachievement in academic problem area and mathematics (if not the area of referral), activity level (hyperactivity or hypoactivity), existing attentional problems, level of severity of attentional and secondary behavioral or emotional problems. These characteristics were selected as the most pertinent to this study from an extensive list of variables in the article by Keogh et 58 a1. entitled "UCLA Marker Variable Project." This article provides a list of descriptors or markers which can be used in reporting sample characteristics in research on learning disabilities. Characteristics were recorded in order to provide the reader with more specific infor- mation about the sample included in this study. Among the subject characteristics included on the data collection sheet was the subject's socio-economic status. This was determined through the Duncan 1970 Socio-Economic Index (SEI)* which ranks occu- pations on a scale of 0 to 100. According to Duncan (1961), this scale may be broken into any degree of coarseness desired; in this study, in which socio-economic variables are a secondary considera- tion, the decline units were arbitrarily broken into five categories: upper class (80-90), upper middle class (60-79), middle class (40-59), lower middle class (20-39), and lower class (0-19). The SEI score for each subject was established by selecting the highest ranking occupa- tion of either parent. Because the individual subject's attitude toward music might in- fluence the subject's success in using musical rehearsal techniques, a questionnaire on musical environment and attitude was developed. In designing a questionnaire appropriate for use with the learning dis- abled, several factors were considered: many learning disabled stu- dents have poor or limited attention span, or a tendency toward impul- sivity. Some learning disabled students have difficulty with word recognition or reading comprehension; and, finally, because of *The Duncan 1970 Socio-Economic Index appears in R. M. Hauser and D. L. Featherman's The Process of Stratification (New York: Academic Press, 1977). 59 underachievement in academic areas, it is important to avoid lengthy disruption of crucial classroom time. Therefore, a questionnaire was designed which would remove the students from the classroom for only a limited time, yet would be maximally informative within time constraints. An interview format was chosen to avoid reading problems, and a closed-form response was chosen to ensure more objective and easily coded data. Finally, age- appropriate language was selected to increase reliability of response. Most music educators agree that a child's attitude toward, and active involvement in, musical activity are influenced by exposure to music and the musical attitude of close family members. These factors are one area of focus in Gaston's Test of Musicality (1957), which in- cludes a questionnaire on musical experience and attitudes. Since Lehman (1968) reports that Gaston's test was normed for grades 4 through 12 (which makes it approximately age appropriate for this study) and that the test is both carefully constructed and well stan- dardized, this measurement device is a logical model for possible questionnaire items. In examining other questionnaires used in re- search with children, the importance of Gaston's test becomes ap- parent, since many of these instruments are updated or revised forms of Gaston's original questionnaire. One study of particular interest, which uses the Gaston test as a model, is reported by Roskam (1977) who includes an examination of musical environment and attitudes among learning disabled children. Roskam's questionnaire is essentially an updated version of the Gaston questionnaire. While Roskam's revision of Gaston's test appeared an adequate measurement device for her study, she provided no information 60 on questionnaire development or content validity; therefore, rather than utilizing this pre-existent questionnaire, similar item revision of the Gaston test was prepared specifically for this study. Item coverage was developed using a table of specifications, concentrating on two basic areas: musical environment and attitude toward music (Table 1). TABLE 1 TABLE OF SPECIFICATIONS Musical Environment--60% Available Points A. Personal active participation in musical events--questions 1 and 3 5 B. Family participation in musical activities--questions 2 and 4 5 C. Passive musical activities, such as listening to music--questions 5 and 6 4 Musical Attitude--40% A. Current interest in musical activities-~questions 8 and 9 5 8. Current interest in passive musical activities--questions 7 and 9 3 TOTAL POINTS: 22 In order to minimize interview time, it was important to select optimally informative items. In an effort to choose such items and to check content validity, potential items were submitted to the fol- lowing panel of experts from the music education faculty at Michigan State University: Dr. Melanie Stuart, Dr. Robert Erbes, and Dr. 61 Albert LeBlanc. Based on their suggestions for rank, weight, and validity of questions, final items were chosen (see Appendix D). The overall subject score for the musical questionnaire was tal- lied in the following manner: potential points were assigned to each item, with a total score of 22 available. For final analysis of musi- cality factors, the total score is subdivided in the following manner: high (15-22 points), medium (8-14 points), and low (0-7 points).* It should be noted that the total length of the questionnaire is a small sampling of behavior and thus in danger of low reliability. However, lengthy measurement instrument which goes byond the optimum attention span of many learning disabled students may be even less effective due to possible inaccurate or imprecise responses. In addition to characteristics of musical attitude and involve- ment, subject characteristics such as age, 1.0., and specific type and severity of learning disability were documented in order to pro- vide more detailed information on the test sample. Subject Characteristics Through the use of various measurement instruments and data col- lection forms, the following descriptive statistics of various group characteristics were tabulated (see Tables 2 and 3). In comparing the normal and learning disabled groups, only two categories show significant difference between groups: age and I.Q. While the majority of normal subjects in the study were aged nine, the majority of learning disabled subjects were aged ten and still *Of those 30 normal and learning disabled subjects in the musical treatment group, 14 subjects measured in the low category, 15 in the medium category, and only one student in the high category. 62 TABLE 2 FREQUENCIES AND CENTRAL TENDENCIES FOR LEARNING DISABLED STUDENTS* Number of Students Type of Learning Disability, Per Category Reading 5 Mathematics 1 Reading and mathematics 15 Written language 3 Reading, mathematics, and written language 5 Auditory memory 1 Severity of Underachievement in Reading and/or Written Language 1 year 4 1.5 years 9 2.0 years 5 2.5 years 1 3.0 years 2 3.4 years 4 3.5 years 1 Mean underachievement = 1.87 years Mathematics Underachievement Above grade level 0 At grade level 1 Less than one year below grade level 10 1-1.5 years below grade level 13 2-2.5 years below grade level 4 2.5 + years below grade level 2 Activity Level Hyperactive 7 Hypoactive Neither 22 63 Number of Students Attention Problems Per Category None 4 Slight 12 Moderate 14 Severe O Emotional or Behavioral Characteristics Minor emotional problems 19 Secondary emotional problems Severe emotional problems None 1 *Learning disability characteristics were evaluated by the re- source room teachers and are thus subject to their individual judg- - ments. TABLE 3 CROSS TABULATION OF FREQUENCIES AND CHI SQUARE ANALYSIS Learning Category Normal Disabled Grade Third 10 9 Fourth 15 17 Fifth 4 4 Raw Chi Square 2 df .16073 P = .9228 Age 9 22 10 10 8 17 11 0 3 Raw Chi Square 2 df 10.74 P - .OO47* 64 Learning Category Normal Disabled Rgg§_ Caucasian 24 26 Black 3 2 Hispanic 1 l Black/Hispanic 2 1 Raw Chi Square 2 df .6133 P = .8934 Iggy 90-99 9 21 100-109 11 110-120 10 Raw Chi Square 2 df 11.002 P = .OO4* Socio-Economic Index High 0 3 Medium high 7 6 Medium 7 4 Medium low 4 1 Low 12 16 Raw Chi Square 4 df 6.2665 P = .1801 *Indicates statistical significance at the .05 level struggling with multiplication facts. Most 10 year old normal sub- jects who were screened had already mastered their tables past the 76% accuracy level acceptable for subject inclusion. The learning disabled students had a significantly greater number than normals in the lower 1.0. range. This finding should be interpreted with care, since "scatter" can complicate the interpretation of an overall test score. Subject characteristics specific to the learning disabled group will be discussed in Chapter V in conjunction with experimental findings. 65 Overview of the Study The primary purpose of this study is to examine the effectiveness of musical rehearsal on short-term memory. In order to measure its effectiveness, a memory task was devised which was followed by immedi- ate posttesting for recall accuracy. Traditionally, the study of verbal memory includes the use of what is commonly called rote 1earning--word-for-word formulas or defi- nitions. Conceptual understanding may follow, but the memorization process itself is perhaps best described as the chaining of verbal associations. Wingfield (1979) adds that such study is important for two reasons. First, the memorization of an order of words verbatim . illustrates chaining of multiple associations and serves as a vehicle for further development of the prin- ciples of learning. The second reason is a practical one. Whether desirable or not, rote learning represents a large bulk of formal learning in the early school years and later.27 For this particular study multiplication tables were selected as a memory task since they require the word-for-word or verbatim recall generally associated with verbal memory research. In addition, they have a similar format to paired-associates memory tasks (in this case, the stem of the multiplication problem, such as 7 x 9, is associated with the correct response, 63), they are easily quantifiable, and yet they constitute an actual academic task learned by third, fourth, and fifth grade students in a normal educational milieu.* In order to examine the suitability of multiplication tables as a memory task, in addition to standardizing verbal instructions and data collection methods, two pilot studies were conducted. *Many of the teachers of students in the study corroborated this statement, reporting that they do encourage and drill on memor12at1on of the times tables. 66 Pilot Studies In the first pilot study, nine normal nine year old subjects were selected from the third grade class of St. Andrew's School in Grand Rapids, Michigan. The nine subjects were randomly assigned to one of three treatment groups: (a) normal speech rehearsal; (b) rhythmic speech rehearsal, with the rhythmic patterns from the song "00, Re, Mi" by Rogers and Hammerstein imposed upon the speech; (c) melodic- rhythmic rehearsal which consisted of rehearsal of the memory task to the song "Do, Re, Mi." The memory task consisted of the table of 75 presented on tape. The table of 75 was selected since the greatest number of calculation errors occurs with the number 6 and higher (Killian et al., 1980). Subjects participated in a pretest, treatment, and posttest in groups of three, with each group receiving a different treatment con- dition. The pretest for all groups consisted of 27 cards (with a separate set for each subject) containing individual presentations of the nine multiplication problems included in the memory task. Answers were completed in the written mode with five seconds allowed for re- sponse. Treatment consisted of a sequence of listening, repeating, writ- ing, and once again repeating the nine problems included in the table of 75. Each group of three rehearsed the problems together, using the specific rehearsal mode to which they were assigned. Immediately fol- lowing rehearsal for that group, subjects were postested using the same content and procedure as in the pretest. Although the n of subjects was too small to carry out statisti- cally meaningful analysis, the mean score for recall improvement was 67 slightly higher for the musical group (1': 5.3) than either the normal speech (§'= 2.3) or the rhythmic speech group (§'= 3.0). Several ob- servations were made (a) the students were able to follow the verbal instructions and carry out the treatment task appropriately, (b) the use of a pre-existent song ("00, Re, Mi") appeared to create a bias as recall accuracy tended to be favorable for those students who indi- cated familiarity with the song, and (c) the use of a single song for the entire table of 7s prevented alternative sequence of problems. Thus, the problem 7 x 9 = 63 always appeared last during the rehearsal process. This problem also showed the greatest recall accuracy. Based on these results, the following procedural decisions were made: self-composed (meaning composed by the researcher) rhythmic patterns rather than pre-existent tunes were used in order to use a musical source of equal unfamiliarity for all subjects, and separate short melodic-rhythmic patterns for each multiplication problem were used to facilitate alternative presentation sequences of the problems. Incorporating these changes, a second pilot study was conducted with nine learning disabled students aged nine and ten from a learning disabilities resource room at Wilcox Elementary School in Holt, Michi- gan. As in the first study, the subjects, selected by the resource room teacher, were randombly assigned to three different treatment groups: (a) normal speech rehearsal, (b) rhythmic speech rehearsal, and (c) melodic-rhythmic speech rehearsal. The same self-composed rhythmic patterns were used for both the rhythmic speech and melodic- rhythmic rehearsal modes. Nine different self-composed melodies were superimposed on these rhythmic patternsiin~the melodic-rhythmic pre- sentation of the memory task. A different melodic sequence was used 68 for each multiplication problem, but all sequences were limited to the pitches C4 through G4. All presentation modes were presented on tape. As in the first pilot study, each treatment group participated as a unit, with each group participating in testing and treatment at separate times. The instructions, treatment task, and pretest and posttest methods were consistent with those methods used in the first pilot study. While the learning disabled students were able to carry out the various tasks as instructed, the test results showed markedly lower recall accuracy, following a single treatment session compared to recall by normal students. Those learning disabled students in the normal speech and rhythmic speech treatment groups displayed a mean gain of 0.0 in the posttest. While the subjects in the melodic- rhythmic treatment group showed a slight mean gain of 1.1, this is a much lower gain than the recall level for the same treatment used with normal subjects. Because gains in recall were so limited following a single re- hearsal by learning disabled subjects, the use of additional rehearsal sessions (with each session followed by posttesting) was considered. Because repeated practice gives the opportunity for additional items to be learned in each trial (Wingfield, 1979), it was hoped that the learning disabled subjects might show more substantial recall gains by the end of a total of five rehearsals. Repeated rehearsals also approximate more closely actual classroom learning than does a single memory rehearsal. In addition, Reid and Hresko (1981) recommend fre- quent reviews covering rather small bits of information in each ses- sion in order to maximally facilitate retention. 69 Following the pilot studies, Dr. William Schmidt, professor of statistics and chairperson of the Counseling, Educational Psychology, and Special Education Department at Michigan State University, was consulted in order to review the research design and to establish an adequate n for sampling. Two procedural decisions resulted from this consultation: (a) in order to remove interactive effects among sub- jects, it would be necessary to carry out treatment and testing with individual subjects rather than in groups of threes (as was the case in the pilot study); and (b) in order to test all three treatment methods utilized in the pilot study, an n of 90 subjects would be re- quired (45 normal and 45 learning disabled subjects). The availabil- ity of 45 learning disabled subjects of the correct age, 1.0. range, gender (male), and level of mathematics achievement seemed unrealistic within the locational and time confines of this study. Therefore, the decision was made to examine the effect of melodic-rhythmic patterns in contrast to verbal rehearsal, omitting the variable of rhythmic patterns. This examination of two rather than three contrasting meth- ods required only 30 learning disabled and 30 normal subjects for the total sample. The use of melodic-rhythmic patterns as Opposed to rhythmic pat- terns was chosen as the treatment condition for the following reasons: 1. In both pilot studies, the melodic-rhythmic rehearsal method showed slightly greater mean gains than did rhythmic or normal speech methods. 2. According to the definition of a rhythmic mnemonic by Bolton and Bower (1969), rhyming syllables are a major part of the effectiveness of rhythmic groupings as an 7O aid to recall. While rhythm along with rhyme was con- sidered as a treatment method, there was concern that the rhyming element may provide memory aids unrelated to the actual rhythmic structure itself (extra- musical associations). 3. The use of rhythmic speech alone has had limited sup- port as an aid to auditory memory. While Prickett's study using serial lists of digits did show improved retention from rhythmic groupings, studies by Wilson (1972), Sims (1980), and Myers (1979) did not support the isolated use of rhythmic structure. Melodic- rhythmic patterns or melodic patterns have shown trends toward improved retention or significantly im- proved retention in studies by Wilson (1972), Lathom (1970), Jellison (1976), and Shehan (1981). The major drawback to use of melodic-rhythmic patterns was the inability of some subjects to accurately match the pitches within the melodic sequence. Most subjects, however, were able to approximate the basic contour of the melody. The Major Study Preparations and Materials Following the two pilot studies, final procedural decisions were made and testing and treatment materials were prepared. The memory task selected was multiplication tables for the numbers seven and eight, specifically the facts 7 x 6 through 7 x 9 and 8 x 5 through 8 x 9. These two sets of multiplication facts are subject to greater 71 number of calculation error than smaller numbers (Killian et al., 1980). In addition, the results of the pilot study showed that the lower multiplication facts for the table of 75 were not particularly difficult, while the higher tables showed high incidence of error on the pretest. Thus, to provide an adequate sampling of problems for sufficient measurement, the upper tables of 85 were included as well as the upper table of 75. During the treatment session, these multiplication problems were presented in auditory and visual fashion, using tape recordings and flash cards. This mixed mediational approach was utilized for the following reasons: (a) Shehan (1981) reports in her study with learn- ing disabled students, that musical/visual and verbal/visual media- tional strategies are both superior to either musical or verbal pre- sentation alone in enhancing recall; and (b) the combined mediational mode more nearly approximates normal classroom presentation of ma- terial. In addition, the visual presentation more nearly approxi- mates the written response mode used in pretest and posttests (visual presentation consisted of each problem, with and without answers, on white 6" x 8" flash cards). Preparation of tapes for treatment. The auditory presentation of the multiplication facts was recorded on tape for both verbal and musical treatment conditions in order to assure consistency of presen- tation and rehearsal. Recordings were made on Maxell C90 tapes, using a SONY TCM-131 recorder with condensor microphone. All spoken and sung input was provided by an adult female with adequate musical skill 72 to read music and maintain the desired pitch sequences without accom- paniment. All tapes were reviewed by Dr. Dale Bartlett, professor of music at Michigan State University, in order to assure clarity and accuracy. In order to establish that the verbal tapes did, indeed, constitute nornml speech, the verbal tapes were reviewed by two speech patholo- gists: Sherry Martin, M.A., CCC-Sph, teacher of speech and language impaired, Ingham Intermediate School District (Michigan); and Kelly Kidder, supervisor, Speech Pathology Department, Grand Rapids (Michi- gan) Public Schools. Both experts signed affadavits stating that the tapes consisted of normal adult speech. Melodic sequences for the nine multiplication problems were self- composed, using pitches C4 through G4. All intervallic movement was by step or arpeggiation (skip of a third) and implied C major tonality Self-composed tunes were used for the following reasons: (a) to avoid subject familiarity with pre-existent tunes; (b) to limit the pitch range, using the tonic and dominant poles of C major tonality (re- search supports tonal music as more easily recalled than random or atonal pitch sequences); and (c) finally, self-composed melodies with- in a restricted pitch range are commonly utilized in psychoacoustic research (Jellison, 1976). The nine melodic sequences were set alternately to three rhythmic patterns: (a) duple meter with minimal distortion of natural speech rhythms, (b) duple meter with syncopation—a deliberate distortion of natural speech rhythms, and (c) triple compound meter. Each of these rhythmic patterns was used in conjunction with each of the nine melo- dic patterns on three separate tapes A, B, and C. For example, the 73 melody accompanying the problem 8 x 5 = 40 was sung to a duple meter with minimal rhythmic distortion on Tape A. On Tape 8, the same mel- ody was sung to duple meter with syncopated rhythmic groupings. On Tape C, triple compound meter was used in conjunction with the same melody. (See Appendix C for all melodic-rhythmic patterns.) The com- bining of the nine melodic sequences with three different rhythmic patterns was to examine any trend toward increased recall due to specific types of melody or rhythm. In order to avoid primacy/recency effects, the three tapes pre- sented the nine problems in different sequence orders. A11 tapes ini- tially presented the tables of both seven and eight in ascending or-- der; but following this initial presentation, the problems were pre- sented in an alternate order based on an even or odd sequence (see Appendix B for the order of problem presentation). Both musical and verbal tapes contained the same word content in the same order of pre- sentation. While the majority of instructions appear on the tapes, the re- maining instructions for pretest, treatment, and posttests were writ- ten in a standard form to provide greater consistency of testing con- ditions. These instructions were examined for usability during the pilot study and minimally adjusted for use in the major study. All standardized instructions appear in Appendix E. Preparation of measurement devices of recall accuracy-pretest and gosttests. In order to determine each subject's pretest knowledge and posttest recall accuracy, a device for measuring accurate recall of the memory task was developed. For the pretest and all five 74 posttests, subject responses were answered on individual 3% x 3%" cards containing one of the nine unanswered multiplication problems (found in the treatment memory task) per card. Each problem was pre- sented twice during each pretest and posttest with all 22 cards pre- tested in a random order.* Responses were obtained in the written mode, since that is the mode which is used predominantly in classroom tasks. The range of correct responses was from O to 18. In addition to the two presentations of each problem with the number seven or eight appearing in the top position (ex. 7 x 9 or 8 x 9), a third presentation of the problem appeared in reverse order (ex. 9 x 7 or 9 x 8) in order to examine the subject's awareness of the commutative principle of multiplication. These additional prob- lems were not included in statistical analysis but rather are dis- cussed in the final chapter as an observational discussion. Posttest problems were presented individually for the following reasons: (a) to provide control over the amount of time available for each response (five seconds). By providing a flash presentation of one problem at a time, the subject had limited time to rely on compu- tational aids aside from rehearsal recall (such as using fingers or adding). This was particularly important to monitor since the mea- sured outcome is to represent short-term memory recall; (b) to avoid an interaction effect among problems. Since each problem was pre- sented twice, the individual presentation of problems impeded the *A minor exception to random presentation was due to deliberate avoidance of the two most recently rehearsed problems. This was done in order to let the first two posttest problems act as distractors in the manner described by Norman (1976) for short-term memory testing. 75 subject's ability to look back at earlier responses in order to check later responses; (c) to reduce visual stimuli. Some learning dis- abled students are easily distracted or become frustrated when pre- sented with a full page of mathematical problems. Thus, by presenting one problem at a time, the student could address full attention only to that response. In order to tally the number of correct posttest responses (and also to note any unusual response patterns such as digit reversals), scoring sheets were provided for each subject (see Appendix A). The number of accurate pretest and posttest responses provide the depen- dent variable for the study. Data Collection Following the development of treatment and testing materials, the collection of research data was conducted in the following elementary schools of the greater Grand Rapids, Michigan, area: East Grand Rapids Public Schools Wealthy School Breton Downs School Forest Hills Public Schools Ada School Collins School Orchard View School Pine Ridge School Thornapple School Godfrey Lee Public Schools Godwin Heights Public Schools North Godwin South Godwin West Godwin 76 Grand Rapids Public Schools Beckwith School Coit School Fountain School Hall School Ottawa Hills School Palmer School Oakdale Christian School St. Andrew's Catholic School Seymore Christian School Sylvan Christian School Wyoming Public School Oriele Park School Taft School All research procedures were carried out by the author and Denise Joseph-Enders, a certified teacher of the learning disabled. Each of the researchers tested half of the subjects for each treatment condi- tion in order to equate response differences due to testor differences. Student involvement was subject to the consent of school adminis- trators, classroom and resource room teachers, written consent of parents, and, finally, consent of the student (see parental consent forms in Appendix G). In order to assure subject anonymity, each stu- dent was issued an identification number for data collection purposes. Following parental consent, all potential subjects were screened for 1.0. and current knowledge of the memory task. Following screen- ing, all eligible subjects participated in a total of five treatment sessions, including one session for Experiment I and four additional sessions for Experiment II (see Table 4 for a description of the en- tire research design). 77 II F O gnome pmnpm>\xpco pmmppmoa gnome pmupmaz\mpco pmmppmom n we mappmmOprmapmaz 1 mx coppppmamxxpmopmzz 1 mx mappamOprmmxmp u Nx coppppmamxxpmmcm> u px mummppmom ccm pmmpmpa u p .lllm Ema *“ N Ame * p xmo “ mp No mp No pp No mp mm m mp p.x mp mx mp mx mp mx Np pmapmsz pp 2 mp mx mp mp mp mxl mp mx 2 mmwmwwwm mp pm mp pm mp po mp pm m . mp Nx mp Nx pp Nx mp Nx Np pmmca> pp m mp px mp px mp px mp pm > mp No mp No pp No mp mm m mp xx mp xx mp px mp mx Np pmapmaz pp a mp mx mp mx mp mx mp mm :2: .x oz mp pm mp pm xp pm mp po N p m mp Nx mp Nx mp Nx mp Nx p pmmcm> p m mp px mp px mp px mp px > Em .HH «mwm Em .HH .awm Em .HH .axm Emlxpp «mxm Em .H .axm mmmmmum pm>mm amocw > copmmmm >H copmmmm HHH copmmmm HH copmmmm H copmmmm onmmo 4