AN EXPERIMENTAL INVESTIGATION. OF THE EFFECT OF SELECTED FACTORS ON THE SHORT-TERM RETENTION OF PITCH SEQUENCES Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY KERMIT WELLS HOLLY, JR 1977 .I o)..’,: Alln/I‘tu. ”my: .. “A u-‘L .q ._( u, I::§I:uadil:§:.331szwhim .. I. ,. .h—a .... LIBPK. 'ZY Michigszt 3 rate Umversrty This is to certify that the thesis entitled AN EXPERIMENTAL INVESTIGATION OF THE EFFECT OF SELECTED FACTORS ON THE SHORT-TERM RETENTION OF PITCH SEQUENCES presented by Kermit Wells Holly, Jr. has been accepted towards fulfillment of the requirements for Ph.D. Music degree in Major professor Date August 12, 1977 0-7639 @QTZ/Z 2/1 2% d ABSTRACT AN EXPERIMENTAL INVESTIGATION OF THE EFFECT OF SELECTED FACTORS ON THE SHORT-TERM RETENTION OF PITCH SEQUENCES BY Kermit Wells Holly, Jr. The primary purpose of this study was to investi- gate the effect of changes of timbre in a stimulus item, length of retention interval, and performance of a task (during the retention interval) on short-term recognition memory for pitch. A secondary concern of the study was the response performance on test items in which the second sequence was different from the first. Procedures A 48-item pitch sequence memory test - The Pitch- Timbre Memory Test - was developed to collect the data for the study. Each item consisted of two pitch sequences. The second sequence was either the same or different from the first. For twenty-four of the items, the same sound quality (a sine wave) was used in both sequences. The second sequence of the remaining twenty-four items used either a sawtooth, pulse or triangle wave form. The re- tention intervals were 5 seconds, 15 seconds, and 30 seconds. All test items were randomized. Kermit Wells Holly, Jr. The twenty-eight subjects used in the study were randomly assigned to one of two groups. Group I received the memory test and performed a task during each of the retention intervals. The task was one in which the sub- jects counted backward by three from a number specified on the memory test tape. To insure that they were counting, subjects were required to write the numbers as they counted. Group II only took the memory test. Conclusions On the basis of the data analysis of the results of this study, the following conclusions were drawn: 1. The sound quality of a pitch sequence tended to affect its perception and its retention in short-term memory. 2. The sawtooth sound quality had the most dis- ruptive effect on the short-term retention of pitch se- quence. The triangle sound quality appeared to be less consistent in terms of effect on short-term retention. 3. The accuracy of recognition response in short- term retention was influenced by the length of the re- tention interval. Response mean scores for the task group and the non-task group were considerably lower at the 15 second retention interval. 4. The interaction effect between timbre and re- tention interval suggested that the short-term retention of pitch is dependent upon the kind of timbre and the Kermit Wells Holly, Jr. length of the retention interval, at least in this study. 5. The performance of a non-musical task did not significantly affect the short-term retention of a pitch sequence. 6. The intention between retention interval and item-type suggested that the recognition of test items as the same or different was dependent on the length of the retention interval. Test items in which both pitch sequences were the same, were recognized with greater accuracy during the 5 second retention interval, than test items, in which the second sequence was different. 7. The triple interaction between timbre, reten- tion interval, and item type (same vs. different) suggested that the short-term retention of pitch sequences is de- pendent upon timbre, length of the retention interval, and whether the second sequence is the same or different. 8. The true influence of the main effects of timbre and retention interval should be qualified because of the number of interactive effects. AN EXPERIMENTAL INVESTIGATION OF THE EFFECT OF SELECTED FACTORS ON THE SHORT-TERM RETENTION OF PITCH SEQUENCES BY Kermit Wells Holly, Jr. A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Music 1977 TO MY MOTHER AND FATHER ii ACKNOWLEDGEMENTS To the members of my guidance committee, Dr. Sherburn, Dr. Erbes, and Professor Klausli: thank you for your time, interest and suggestions in the preparation of this thesis. To Dr. Robert Sidnell, my advisor and committee chairman: a special thank you for your guidance, insight. and patience, during this entire project. To Mrs. Aurelia Young, a special thank you for your tutelage and belief in me over the years. Finally, to my father and mother, Mr. and Mrs. Kermit W. Holly, Sr.: a very special thank you, for your unyielding faith, support and encouragement. iii Chapter II III IV TABLE OF CONTENTS THE PROBLEM Introduction The Problem Statement Purpose Hypotheses Significance of the Study Definition of Terms Assumptions Organization of the Study REVIEW OF LITERATURE General Theories of Memory Trace-Decay Theory Interference Theory Information Processing Short-Term Memory for Pitch Timbre Perception Summary DESIGN OF THE STUDY Sample Instrumentation Design Procedures The Pilot Study Treatment of Data Testable Hypotheses ANALYSIS OF THE DATA Data Procedures Results Summary SUMMARY AND CONCLUSIONS Review of the Study Findings Conclusions iv Page 10 10 12 13 16 16 17 17 18 19 21 22 45 49 52 52 52 57 58 59 60 60 63 63 66 78 78 81 85 Chapter Discussion Implications of the Study Suggestions for Further Research BIBLIOGRAPHY APPENDICES Appendix Appendix Appendix Appendix The Pitch-Timbre Memory Test Description of the Twenty-Four Test Questions Raw Data for the Twenty-Four Test Questions Test-Retest Scores Page 86 89 92 94 102 104 106 107 LIST OF TABLES Pitch-Timbre Memory Test Format Data Matrix for the Split-Plot Factorial Design ANOVA Summary for Main Effects and Interactions ANOVA for Timbre Effect (Factor A) Results of the Neuman-Kuels Test of Timbre Means ANOVA for Retention Interval Effect (Factor B) Results of the Newman-Kuels Test of Retention Interval Means ANOVA for Task and Item-Type Effect (Factors C and D) ANOVA for the Interaction (AB) of Timbre (A) and Retention Interval (B) ANOVA For the Interaction Effect of Timbre of Timbre/Task (AC) and Timbre/ Item (AD) ANOVA for the Interaction Effects be- tween Retention Interval and Task (BC), Retention Interval and Item Type (BD) and Task and Item Type (CD) ANOVA for Interaction Effects (ABC) be- tween Timbre (A), Retention Interval (B) and Task (C), and the Interaction Effects (ABD) between Timbre (A), Re- tention Interval (B), and Item Type (C) vi Page 55 64 65 66 67 67 68 69 69 71 72 73 Table Page 4.12 ANOVA for the Interaction Effects (ACD) between Timbre (A), Task (C), and Item (D), and Interaction Effects (BCD) be- tween Retention Interval (B), Task (C), and Item (D) 75 vii Figure 1.1 LIST OF FIGURES Information Flow Through the Memory System Structure of the Test Tapes Timbre/Retention Interval Inter- action (AB) The Interaction (BD) between Retention Interval (B) and Item Type (D) Interaction (ABD) between Timbre (A), Retention Interval (B) and Item Type (D) Interaction (BCD) between Retention Interval (B), Task (C), and Item (D) viii Page 59 70 72 74 76 CHAPTER I INTRODUCTION Memory and learning are interrelated processes; factors that affect the former also affect the latter. An examination of current memory literature reveals that there are several theories of memory. One of these describes the human memory system as an information processing system involving the acquisition, retention, and retrieval of information. This view of memory as information processing has led to the development of a number of 'models of memory.‘ Of these, one that seems highly applicable to a theory of music memory is the model proposed by Atkinson and Shiffrin (1968). This model postulates that memory is comprised of three components -- a sensory register, a short-term store, 1 and a long-term store. Initially, information from the environment is accepted and processed by the various sensory systems (sensory register) and is entered into the short- term store, where it is either maintained through rehearsal, or lost. Information is lost from the sensory register in from one to two seconds. Thedecay rate for information in 1R. Atkinson and R.M. Shiffrin, "The Control of Short-Term Memory," Scientific American, 1971, Vol. 25, No. 2, p. 83. 1 the short-term store is approximately thirty seconds, while in short-term store, information, if rehearsed, can be transformed to long-term storage. The long-term store is assumed to be a relatively permanent memory, from which information is not lost.2 The central feature of the Atkinson and Shiffrin model is the control process that operates within and between each of the structural components of the memory system. In the sensory register, the first control process is a decision as to what information to attend to, and where to scan within the system. Control processes inherent in the short-term store are a choice of which information to scan and a choice of what and how to rehearse. The short-term store is conceived of having two functions: (1) it acts as a buffer between the sensory register and long-term storage by utilizing rehearsal strategies; and (2) it acts as a processor of information in which an item passes through various stages of encoding until it is made permanent and becomes a part of long-term store. This model assumes an interaction between short- term memory and long-term memory and a system of control processes that keep information flowing through memory. Figure 1 illustrates the flow of information through the human memory system according to this model. 2Ibid., p. 83. FIGURE 1 Short-Term Store (STS) Sensory Registers Temporary Working Memory Long-Term Environ- Visual Control NETS? Irrrliegtéal “H Auditory ___) Processes: _9(‘_ p Permanent Rehearsal Memory . Coding Tactile Decisions Store I Retrieval . Strategies /// \I/ e . / «1 Response Output J Information Flow Through the Memory System (after Atkinson and Shiffrin, 1968) Music Memory Memory has been long considered an important component of music aptitude. This is evident by the number of music aptitude tests (such as those by Seashore, Drake, Wing, Bentley, and Gaston) that include a memory subtest. Shuter states that the importance of memory in musical ability is undeniable and that an appreciation of form, or qualities of performance would be virtually impossible without memory.3 Memory, being an essential factor in music aptitude is also an important factor in music learning. Perhaps the 3Rosamunde Shuter, The Psychology of Music Ability (London: Methuen and Co., Ltd., I968): P. 188. 4 role of memory in music learning has been best expressed in the following statement by Carl Seashore: The learning process in music involves two primary aspects; acquisition and retention of musical information and experience, and the development of musical skills. Both of these are included in the common term memory...4 Mursell, a theoretical and phiIOSOphical Opposite of Seashore, also recognized memory to be important in the music learning process. He was also astonished at the (then) lack of research involving music memory.5 A more recent indication of the significance of music memory is found in Richard Colwell's construction of the Music Achievement Test battery. Dr. Colwell's inclusion of aural discrimination tests, and tests of pitch recognition, melodic recognition, instrument recognition, and tonal memory stress the role of music memory in music achievement. Further evidence of the awareness of the prominence of memory in music learning is the emphasis placed on memory development by Carl Orff in his Schulewerk. The numerous instances provided by Dr. Orff's method for rote singing, improvisation, and performing on instruments from memory illustrate an acknowledgement of the importance of music memory to musical growth. 4Carl Seashore, Psycholo of Music (New York: McGraw-Hill, 1938, Dover Edition, 1957 , p. 149. 5James Mursell, Psychology of Music (New York: W.W. Norton and Co., 1937), p. 256. Closely related to music memory is the factor of music or aural imagery. Imagery refers to the retention of, or the reproduction of an aural stimulus after actual stimulus has terminated. Gordon6 postulates that imagery may be more important to music aptitude then music memory. (At present this position is not entirely dominant among music educators.) Imagery is, however, considered .7 important in music learning. Hoffer suggests that imagery is a necessary component in music reading. Sidnell and Reed8 feel that, for the listener, aural imagery is a requisite skill in perceiving musical structure. There is little disagreement about the prominent role that aural imagery plays in music perception. Music Memory Theory As early as 1919, Seashore was aware that there were several kinds of music memory and stages of music memory, including an immediate (STM) memory and a longer more permanent memory.9 Early writers on the subject 6Edwin Gordon, The Psychology of Music (Englewood Cliffs, New Jersey: PrentiCe-Hall, Inc., 1971), PP. 22-29. Charles Huffer, Teaching Music in_the Secondary Schools (Belmont, California: WadSworth PuEliEhing Company, 1973), p. 16. 8Owen Reed and Robert G. Sidnell, The Materials of Music Composition (Reading, Mass.: Addison-Wesley Publishers, in press, 1977). p. 164. 9Carl Seashore, The Psychology of Musical Talent (Boston: Silver Burdett and Co., 1919), p. 238. Identified three kinds of music memory: pitch memory (memory for a single pitch), tonal memory (memory for a sequence of tones, implying tonal relationships), and memory for musical phrases. Today they are considered different types of an overall memory. According to Williams, current research indicates insufficient data to construct a complete model of music memory.10 It can be postulated, however, that the processing of music information through the memory system does involve three stages similar to those in the model suggested by Atkinson and Shiffrin. It can also be postulated that those stages are subject to the same control processes. lost researchers and writers generally agree on a three-stage model of music memory. Williams (1973) has theorized that one could control the input of musical stimuli in short- term memory and its subsequent transfer of information to long-term memory9J' Deutsch (1965), a prolific researcher on the topic, subscribes to a multiprocess concept, but 10David B. Williams, "Short-Term Retention of Pitch Se— quence," Journal of Research in Music Education, Vol. 23 (1975), p. 54. ll . . . . DaVid B. Williams, "The Short-Term Retention of Pitch Sequence." Ph.D. Dissertation. Seattle: University of Washington (1973), p. 169. believes that evidence points to a highly specialized system for pitch retention different from the verbal 12 memory system. The Problem Many learning theorists have turned to the study of short-term memory in an effort to investigate the causes 13 of forgetting (loss of information). Mohs states that the contents of short-term memory affect the retrieval of 14 information in long-term store. According to Atkinson and Shiffrin: . . .the short-term memory system has been given a position of pivotal importance. That is because the processes carried out in the short-term store are under the immediate con- trol of the subject and govern the flow of information in the memory system.15 Most of the research pertaining to short-term memory has been concerned with verbal memory. Many of the findings, how- 12Diana Deutsch, Short-Term Memory (New York: Academic Press, 1975), p. 122. 13Michael Posner and Andrew Konick, "On the Role of Inter- ference in Short-Term Memory," Journal of Experimental Psychology. Vol. 72 (1966), pp. 221-231. 14R.C. Mohs, "The Effect of Short-Term Contents on Long- Term Memory Search," Memory and Cognition, No. 4 (1973), p. 443. 15Atkinson and Shiffrin, p. 82. ever, seem applicable to memory for music or sound stimuli. Studies of short—term memory for music have dealt primarily with pitch discrimination. Most of the early studies concen- trated on theories of absolute pitch, the improvement of pitch discrimination, and memory as a component of music aptitude. Recent research has dealt with melodic perception and more recently with the loss of information in short-term memory. Studies by Wickelgren (1966), Massard (1970), Deutsch (1969, 1972), and Williams (1973) indicate that factors affecting retention in verbal memory (i.e., serial effect, similarity of interpolated material and storage capacity) also have similar effects on short-term memory for pitch. A review of short-term pitch memory research reveals that there has been very little research on the influence of timbre on retention. One explanation for this is the assumption that timbre has little or no effect on recogni- tion or recall of music. This may be true for long-term store, but not necessarily so for short-term store. A study by Mull found that subjects felt that some timbres were an aid to the improvement of pitch discrimination”:16 Greer states that a number of researchers believed that timbre has an influence on perception and retention}.7 16Helen K. Mull, "The Acquisition of Absolute Pitch," The American Journal of Psychology: Vol. 26 (1925). l7Douglas Greer, "The Effect of Timbre on Brass Wind Intona- tion," Experimental Research in the Psychology of Music. 19' 18and Petzold, however, The studies of Mainwaring found that different timbres did not significantly affect the recall of tonal patterns. Petzold, although accepting the conclusions of his study, felt that further research was needed. Recent timbre research has resulted in the formulation of new theories concerning the relationship between timbre and pitch. Where timbre was once a carrier of sound, it is now thought to be a sound object aswell.20 Current theories concerning the importance of short- term memory, and new developments in timbre research in- dicated a need to investigate the effect of timbre on short-term retention. Other factors of interest were rehearsal processes and the effect of the length of the retention interval on recognition memory for pitch sequences. 18James Mainwaring, "Kinesthetic Factors in the Recall of Musical Experience," British Journal of Psychology, Vol. 23 (1933), pp. 284-307. 19Robert G. Petzold, Auditory Perception of Musical Sounds by Children in the First Six Grades, Cooperative Research Projéct, No. 1051 TMadison: University of Wisconsin, 1966). 20Robert Erickson, Sound Structures in Music (Berkeley: University of California Press, 1975), p. 18. 10 PURPOSE The primary purpose of this study was to investigate the effect of selected factors on short-term retention of pitch. Specifically this study sought to answer the following questions: 1. Does timbre affect perception and retention in short-term memory? 2. Does the length of the retention interval affect the cognition in short-term memory? 3. Does the performance of a cognitive-motor task affect retention in short-term memory? The possibility that the type of test item could have an influence on retention led to the inclusion of the following question: 4. Does the type of test item, i.e., one in which the second pitch sequence is the same as the first, and one in which the second sequence is different from the first have any affect on perception and retention in short-term memory? On the basis of the research questions, the follow- ing hypotheses were advanced for this study. H1: Recognition response mean scores will show a difference among the four types of timbre used in the second se- quence of each item on the pitch-timbre memory test. H : Recognition response mean scores on the pitch-timbre memory test will show a difference when the length of the retention interval is increased. o O. :13 In 10‘ 11 11 Recognition response mean scores will show a dif- ference between the task group and the non-task group (p. 15). Recognition response mean scores will show a dif- ference between test items in which the second pitch sequence was the same and test items in which the second sequence was different. (This hypothesis was not of experimental interest). An interaction (A B) in memory performance will be shown between timbre (A) and retention interval (B). An interaction effect (A C) in memory performance will be shown between timbre (A) and task (C). An interaction effect (A D) in memory performance will be shown between timbre (A) and same and dif- ferent responses (D). An interaction effect (B C) in memory performance will be shown between retention interval (B) and task (C). An interaction effect (B D) in memory performance will be shown between retention inteval (B) and same and different responses (D). An interaction effect (C D) in memory performance will be shown between task (C) and same-different responses (D). An interaction effect (A B C) in memory performance will be shown between timber (A), retention interval (B), and task (C). 12 H12: An interaction effect (A B D) in memory performance will be shown between timbre (A), retention interval (B) and same and different responses (D). H13: An interaction effect (A C D) in memory performance will be shown between timbre (A), task (C) and same and different items (D). H14: An interaction effect (B C D) in memory performance will be shown between retention interval (B), task (C) and same and different items (D). 315‘ An interaction effect (A B C D) in memory performance will be shown between timbre (A), retention interval (B), task (C), and same and different items (D). SIGNIFICANCE OF THE STUDY Contributions to a theory of memory are significant when it is realized that a theory of memory is a portion of the domain of a theory of learning. By analogy, con- tributions to a theory of music memory are also contribu- tions to a theory of music learning. The primary impor- tance of this study lies in the information that it may provide concerning pitch perception and short-term reten- tion (recognition) as components of a music memory system. That memory is important to music listening is un- deniable. Without a system that retained music informa- tion it would be virtually impossible to comprehend, 13 appreciate, enjoy or respond to a musical experience. 21 Paul Haack has written that: One cannot be a very sensitive perceiver of music, or any type of effective musician for that matter, without the basic ability to re- tain and recall timbres and tonal—rhythmic patterns... On a much broader level of consideration Reimer22 notes that an important change in the new conception of general education in music, is the realization that listen- ing is the primary music activity of our culture and is also the foundation for all other music involvements. Given the prominence of listening in the music experience, and the role of memory in the listening process, an investigation of factors that may affect retention would seem to be of significance. Finally, if the assumption that timbre perception is important in music learning, is valid, the assumption that timbre is important in perception may also be valid. The significance of this study also lies in the informa- tion that it may provide concerning timbre and perception. DEFINITION OF TERMS Definitions for the following terms are provided to assist the reader in understanding some of the concepts 21Paul A. Haack, "Thanks for the Memory," Music Educators Journal, 1975, March, p. 46 22Bennet Reimer, "Patterns for the Future," Music Educators Journal, 1976, December, p. 22. l4 discussed in this study. These definitions are thought to be adequate for the purposes of this study. Buffer - A block of memory set aside for the processing of input/output data. Short-term Memory (STM) - The part of the memory system that has a limited capacity and is of short duration, lasting only seconds. Other designations for STM are immediate memory, primary memory, and short-term store. Long3term Memory (LTM) - The part of the memory system that is relatively permanent and generally unlimited in capacity. Other designations for LTM are secondary memory and long-term store. Retention - The persistence of an item over time after one presentation. Information processing - The input, processing, storage, and retrieval of information. The concept has been borrowed by psychologists from computer technology to describe the human memory system. Forgetting - The loss of information in memory. This definition also implies the inability to identify or recall an item or an event. There are two compe- ting theories of forgetting: decay (time) and interference. These will be discussed in Chapter II. Encoding - The process of modifying information so that it is in a usable form for the memory system. 15 Memory Trace - A hypothetical modification of the neural system, which is postulated to account for memory. Model of Memory - A theoretical representation of the memory process. Retention Interval - The time span between the presentation of pitch sequence B in an A-B presenta- tion. For this study, the RI will be 5 sec, 30 sec, and 55 seconds. Timbre - As used in this study, refers to the sound quality of a particular wave form. Music Memory - As used in this study, refers to the input, storage and retrieval of musical information (a composite of the parameters of music - rhythm, melody, harmony, structure, and color). Pitch Memory (tonal memory) - As used in this study refers to the input, storage, and retrieval of pitch information only. Task group - refers to the group of subjects that perform a cognitive-motor task while taking the music memory test. Non~task group refers to the group of subjects that take the music memory test. 16 ASSUMPTIONS l. The basic premise of this study is that one of the primary goals of music instruction is the facilitation of music learning. 2. It is assumed that the model of memory presented here is valid and can be adapted to a theory of music memory. ORGANIZATION OF THE STUDY Chapter I has presented a statement of the problem, it's background, and it's significance. The remainder of the study is organized as follows: Chapter II contains a review of the literature con- sidered relevant to the topic. Included is dis- cussion of general theories of memory, and pertinent studies of short-term retention of pitch. Chapter III contains a presentation of the research design used in the study and a discussion of the data gathering instrument. Chapter IV contains an analysis of the data. Chapter V contains a summary, conclusions, a dis- cussion of the findings, and implications for further research. CHAPTER II REVIEW OF RELATED LITERATURE The purpose of this chapter is to review selectively the literature relevant to this study. The following topics are discussed: a) general theories of memory, and b) short- term memory for pitch. General Theories of Memory The process of memory has been of interest to man since the period of the ancient Greek philosophers. However, it was not until the beginning of psychology that quantified study could be applied to the processes of memory. Since the latter part of the nineteenth century, a number of theories have evolved to explain memory and the reasons for its failure.1 William James2 was one of the first psychologists to propose more than one kind of memory. He made a distinction 1Donald A. Norman, Memory and Attention (New York: John Wiley and Sons, Inc., 1969), pp. 1-3. 2William James, The Principles of Psychology, Volume I (New York: Henry Holt and Company, 1880), p. 420. 17 18 between a rather temporary memory, which he labeled primary, and a more latent memory, identified as secondary. These observations, however (with the exception of Carl Seashore), were not considered by psychologists until the 1950's.3 Beginning with Ebbinghaus in 1885, early approaches to memory and forgetting have involved two basic theories: trace-decay theory, and interference-theory. Ebbinghaus suggested that events in memory weaken over time if they are not maintained by rehersal.4 Trace-Decay Theory The trace-decay theory contends that when something is learned, a memory trace is formed. If the learned infor- mation is not practiced, it will decay over time, causing forgetting. The emphasis is on changes in memory storage as a result of time.5 Two hypotheses associated with the trace-decay theory are the consolidation hypotheses, and the spontaneous decay hypothesis. Consolidation assumes that if a memory trace is allowed to strengthen, or "set," 3Norman, 92, cit., p. 81 4H. Ebbinghaus, Memory, Trans. by H. A. Rogers and C. E. Bussenivs (New York: Teachers College, 1913. Paperback ed., 1964). 5David Horton and Thomas Tornage, HumanLearning (Engle- wood Cliffs, N.J.: Prentice-Hall, 1976), p. 122. 19 little forgetting will occur.6 Spontaneous decay hypothesis proposes that information that is presented briefly is for- gotten quickly if not rehersed. This aspect of trace theory is important in theoretical distinctions between temporary and relatively permanent memory systems.7 Interference Theory Interference theory maintains that a memory trace is a permanent entity, and not influenced by time decay. For- getting is explained as a difficulty of retrieval rather than storage.8 Stronger memories block retrieval of the information being recalled. The retrieval process is hampered due to the similarity of the information being recalled and other information stored in memory.9 In an attempt to describe the processes that charac- terize forgetting, interference theorists have examined two sources of interference: retroactive inhibition and pro- active inhibition. Retroactive inhibition occurs when new memories displace older memories. Proactive inhibition occurs when new memories are displaced by older memories . 61bid. 71bid. 8Laird S. Cermak, Human Memory: Research and Theory (New York: The Ronald Press, 1972), P. 6. 9Horton and Turner, gp. cit., p. 73. 20 theories had been abandoned as explanations of memory and forgetting until the mid-1950's when new research findings began to indicate that there may be more than one kind of memory process and several possible causes for forgetting, taking into account both interference and decay theory.10 Both trace-decay and interference theories of memory assumed that all events in memory were acquired, stored, and retrieved according to a single, all-encompassing set of principles, i.e., memory involved a unitary process.11 As various facts about memory began to accumulate from clinical, biological, and experimental sources, it became increasingly difficult for psychologists to maintain a single process theory of memory. The collected data indicated that human memory might involve:12 1. At least two qualitatively different systems, one operating according to trace theory, the other according to associationist (interference) theory. 2. Distinct storage and retrieval processes at both the physiological and the psychological level. 10Ibid., p. 147. llIbid.. pp. 151-152. 1211mm. IH 21 3. Complex coding of events in terms of both physical attributes (e.g., visual and auditory) and psychological attributes. As a result, psychologists began to develop complex models of memory based on the explicit assumption that memory was not a unitary process. Information Processing Whereas the earlier theories of memory had been placed in a stimulus-response framework, the newer models (theories) were and are based on a cognitive approach-~that of infor- mation theory and processing. Information theory is based on a mathematical theory of communication that offered a means of measuring infor- mation in abstract units called bits. A bit is defined as the amount of information that distinguishes between two 13 In 1949, the theory was equally likely alternatives. presented by Shannon and Weaver14 in a form accessible to psychologists. It had implications for the interpretation of cognitive processes ranging from perception to language. The information processing approach is derived from information theory and assumes that perception, memory, and 13Geoffrey R. Loftus and E. Loftus, Human Memory: The Processing of Information (New York: John Wiley and Sons, 1976‘)! p- 3. 14C. E. Shannon and W. Weaver, The Mathematical Theory of Communication (Urbana, Ill.: UniversIty of Illinois Press, 1949). 22 the other by Peterson and Peterson,20 inaugurated intensive research activity which has continued to the present. As stated in Chapter I (page 6), most of the research pertaining to memory and to short-term memory in particular has dealt with verbal memory. Non-verbal auditory memory research has only recently begun to garner similar interest. A review of the literature involving experimental research in music memory reveals three primary areas of investigation: a) studies of absolute pitch and the development of pitch discrimination; b) studies of tonal memory as a component of musical aptitude; and c) specific studies of the short- term retention of pitch. The distinction between (a) and (c) is that the latter studies are investigations of pitch retention as it is related to specific theories of short- term memory, and the former studies were primarily inter- ested in the development of pitch discrimination. Short-term memory for pitch Most of the studies of absolute pitch and pitch dis- crimination or pitch improvement have been reviewed else- where (New, 1957; Shuter, 1968); consequently, they will not be discussed here, except in instances where they relate specifically to the present study. Although they have not been classified as such, many of the early studies of music 20Lloyd R. Peterson and E. Peterson, "Short Term Reten- tions of Individual Verbal Items," Journal of Experimental Psychology, 1959, Vol. 58, pp. 193-198. 23 memory and pitch discrimination are actually studies of short-term retention, in that they involved the presen- tations of brief tonal stimuli and responses to those stimuli within a brief time span, i.e., a few seconds. One of the earliest studies of pitch retention was by Wolfe,21 who investigated the effect of time on pitch retention. Using retention intervals up to 180 seconds, he found that there was very little loss of retention from one to five seconds. However, as the retention interval was lengthened, a gradual decline was noted. Subsequent studies by Angell and Harwood,22 and Whipple23 confirmed Wolfe's findings. Anderson24 studied interstimulus intervals of 1/16 1/8, 1/4, 1/2, 1, 2, 3, and 4 seconds to determine the most effective time interval for pitch discrimination. He re- ported all of the time spans as satisfactory interstimulus intervals. 21H. K. Wolfe, "Untersuchungen fiber das Tongedfichtniss," Philosophy Studies, 1886, Vol. 3, pp. 534-571. 22F. Angell and H. Harwood, "Experiments on Discrimination of Clangs for Different Intervals of Time," American Journal of Psychology, 1899, Vol. 10, pp. 67-71. 23G. M. Whipple, "An Analytic Study of the Memory Image and the Process of Judgement in the Discrimination of Clangs and Tones,: American Journal of Psychology, 1901, Vol. 12, pp. 409- 457. 24D. A. Anderson, "The Duration of Tones, the Time Interval, The Direction of Sound, Darkness, and Quiet, and the order of Stimuli in Pitch Discrimination," Psychological Monographs, 1914, Vol. 16, pp. 150-156. 24 In 1945, Koester,25 using various retention intervals from one to ten seconds, found no loss in pitch retention due to time. In a later experiment, however, using retention intervals of 0, 5, 15, and 47 seconds, he reported no signif- icant loss of retention of to 15 seconds, but with the 47 second interval, there was a significant retention loss. Harris,26 reviewing the literature for pitch discrim- ination and interstimulus time intervals concluded that: The effect on pitch discrimination of elapsed time between stimuli has been shown . . . to depend to a considerable extent on whether the standard sequence is always the same frequency. When it is, subjects can quickly build up from the preceding stimuli a subjective standard upon which he bases his judgement of succeeding stimuli. This subjective standard remains stable for many seconds. Conse- quently, discrimination with a single standard stimulus is relatively little affected by the time between standard and comparison stimuli. If, on the other hand, the standard stimulus is changed in frequency for every judgement, then the range of the standard and variable stimuli is considerably greater, the preceding sequence of stimuli furnishes only a relatively coarse anchor for judging, and the subject is forced to consider only the two particular stimuli of a comparison pair.27 Accordingly, Harris proceeded with an experiment using both a fixed standard pitch and a varying standard pitch (from 950 Hz to 1050 Hz). Using two groups, he 25Thomas Koester, "The time Error in Pitch and Loudness Discrimination as a Function of Time Interval and Stimulus Level," Archives of Psychology, 1945, No. 297. 26J. Donald Harris, "The Decline of Pitch Discrimination with Time," Journal of Experimental Psychology, 1952, Vol. 43, pp. 96-99. 27Ibid., p. 98. 25 tested one group with a fixed standard pitch at intervals of 3, 7, 15, and 25 seconds. For the second group, the standard tone was varied. The interstimulus intervals were .1, l, 3, 7, and 15 seconds. Harris reported that with the fixed standard stimulus, no decline in discrimination occurred up to 25 seconds; a decline of 8 cps occurred after 15 seconds; a decline of 3 cps occurred after 25 seconds. On the other hand, there was a retention loss at delays of 3, 7, and 15 seconds with the group having the standard pitch.28 In 1954, Bachem29 investigated the effect of time lapse between standard and comparison tones utilizing a two group design in which one group's subjects possessed absol- ute pitch and the other group's subjects, relative pitch. The time-lapse intervals were 1 second, 3 seconds, 15 seconds, 60 seconds, 1 hour, 1 day and 1 week. Results showed that there was no observable difference in pitch retention between the two groups. However, errors increas- ed with the longer time intervals, with few errors being re- ported by the subjects possessing absolute pitch.30 281bid., p. 99. 29A. Bachem, "Time factors in Relative and Absolute Pitch Determination," Journal of the Acoustical Society of America, 1954, Vol. 26, pp. 751-753. 30Ibid., p. 753. thé de in fr N 1 4. I'll ll Ill I? 26 The studies that have been cited are important because they show that retention for a single pitch is subject to decay over time. The alternative hypothesis for loss of information in short-term memory is that of interference from competing stimuli. Wickelgren31 designed an experiment to determine whether pitch retention was also subject to interference from material interpolated during the reten- tion interval. In this experiment, subjects listened to a standard tone of either 2, 4, or 8 seconds in duration, followed by an interference tone (a tone sounded during the retention interval) of 2, 4, or 8 seconds duration. A comparison tone of 2 seconds was then played, followed by a 4-second re- sponse interval. Subjects rated their confidence of response using a scale from 1 to 5. Standard tones used in the exper- iment were either 800 Hz, 820 Hz, or 840 Hz. The comparison tones differed by either 0 Hz, +10 Hz, or +15 Hz-- all diff- erence conditions occurred equally. Wickelgren found that:32 1. Interpolating a tone between the standard and comparison tones does affect pitch recognition; the longer the duration of the interpolated tone, the poorer the pitch recognition. 2. Increasing the duration of the standard tone facilitated pitch retention. 31Wayne A. Wickelgren, "Consolidation and Retroactive Interference in Short-Term Recognition Memory for Pitch," Journal of Experimental Psychology, Vol. 72, No. 2, 1966, pp. 250-259. 321bid. 27 3. Trace strengths for the standard tone increased during presentations of the standard pitch that were over 4 seconds. He also reported that the higher pitched tones were remem- bered better than lower pitched tones.33 In a subsequent study, Wickelgren34 examined the interference tone to determine what effect its frequency and intensity similarity to the standard tone had on pitch re- tention. The tones for this experiment were randomly selec- ted in 10 Hz steps over the interval from 400 Hz to 590 Hz. Interference tones were T15 Hz, 120 Hz, :40 Hz, 1100 Hz, :200 Hz, +300 Hz, or +800 Hz from the standard (S) tone. The comparison (C) tone differed by 0 Hz or +10 Hz. Two groups were used for the study. One group was instructed to attend to the interpolated tone, and not to reherse dur- ing the retention interval. The second group was instructed to ignore the interpolated tone and to reherse the S tone. Wickelgren reported that:35 l) The similarity of the interpolated tone to the standard tone appeared to have no effect on recog- nition performance (beyond a difference of 40 Hz between the standard and interference tone) regard- less of the instructional condition. 33Ibid. 34Wayne A. Wickelgren, "Associative Strength Theory of Recognition Memory for Pitch," Journal of Mathematical PsycholOgy, 1969, Vol. 6, pp. 13-61. 3SIbid., p. 38. 28 2) There seemed to be a facilitory effect on per- formance when the interpolated tone differed by 15 Hz or 20 Hz, and the comparison tone was on the same side as the standard and interference tones. The results of the intensity experiment indicated that the interpolated tones' intensity did not affect pitch recog- nition (at least within the levels used in the study). 36 that the duration of It was concluded by Wickelgren the interpolated tone was the primary cause of interference, since neither similarity nor the intensity of the interpol- ated tone had any significant effert on the recognition performance. In 1970, Massaro,37 employing various kinds of inter- polated material, also reported that interference was a factor in the deterioration of short-term memory for pitch. He used tones, Gaussian noise and a blank time interval as the interpolated material: the retention intervals were 1 second, 2 seconds, and 4 seconds. An analysis of the data from the experiment indicated that the retention of pitch decreased as the retention interval increased, and that the decrease in retention over time is highly dependent on the interpolated material.38 36Ibid., p. 15. 37Dominic W. Massaro, "Retroactive Interference in Short-Term Recognition Memory for Pitch," Journal of Experi- mental Psychology, 1970, V01. 83, pp. 32-39. 381bid., p. 39. >1.» .- a... 29 It was also found, as in the Wickelgren study, that interpolated tones nearer in Hz to the standard tone had a facilitory effect on retention. In a series of experiments, Massaro39 sought to determine the effect of a masking tone, presented retro- actively (after the standard tone) and proactively (before the standard tone) on short-term pitch retention. Subjects in the first experiment were asked to identify two 20 milli- second tones as high (870 Hz) or low (770 Hz). A masking tone of 820 Hz and 500 Milliseconds followed after a vary- ing time interval of 0, 20, 40, 80, 160, 250, 350 or 500 milliseconds. All conditions were randomized within a given session (two per day during a four-day period). Results of the experiment indicated that pitch identification perfor- mance increased as the time interval between the test tone and the masking tone increased, up to approximately 250 milliseconds when the masking tone was presented immediately after the test tone, it interfered with the processing of the test tone.40 The procedures for the second experiment were the same as in the first experiment except that the masking tone was presented proactively. It was found that the test tone was 39Dominic W. Massaro, "Preperceptual Auditory Images," Journal of Experimental Psychology. 1970, Vol. 85, pp. 411-417. 4°Ibid., p. 412. 30 identified at a minimum performance level of 94 percent at 41 Experiments III and IV were replications each time level. of the first experiment except that the masking was present- ed dichotically i.e. the test tone was presented to one ear, and the masking tone was presented to the other ear. In experiment III, the test tone was always presented to the same ear. The masking tone was always presented to the oppo- site ear. In experiment IV, the test tone was presented to either ear randomly. The masking tone was presented to ear contralateral (opposite) to the test tone presentation. The results obtained were similar to those obtained in the first experiment. Massaro theorized that these results provided evidence for a central auditory store, since a dichotic mask- 42 ing tone was as effective as one presented binaurally. A fifth experiment was undertaken to determine whether the similarity of the masking tone to the test tone had any effect on the processing of pitch information. The data revealed no significant difference in identification or per- formance as a function of the frequency of the masking tone.43 Massaro interpreted the findings of this series of experiments as evidence for a sensory storage system that retains an auditory image(retention of a stimulus presenta- tion after the stimulus has terminated) of a tone for about 411bid., p. 413 42Ibid., p. 415 431bid., p. 416 31 250 milliseconds. He also concluded that the image could be processed while it lasts.44 This study is also important because it provides the first quantitative evidence for pitch imagery. Prior studies such as those by Whipple (1901) were based on introspective reports of subjects. In 1972, Massaro 45 sought to determine the effect of stimulus information and processing time in making absol- ute pitch judgements. He was also interested in determin- ing whether perceptual processing during a silent interstim- ulus interval was as effecient as in processing a continuous tone. Two test tones, an 800 Hz sine wave tone and an 800 Hz sawtooth wave tone, were to be identified as sharp (saw- tooth wave) or dull (sine wave). The two conditions of the study were (1) silent processing, and (2) continuous pro- cessing. In the silent processing condition, the test tone was followed by a silent time interval of 30 milliseconds, 50, 80, 130, 190, 250, 340 or 440 milliseconds. The contin- uous condition had the test presented at durations of 40, 60, 90, 140, 200, 260, 350 or 440 milliseconds, followed by a 10 millisecond silent interval, which was followed by a masking tone. 441616., p. 416 45Dominic W. Massaro, "Stimulus Information Vs. Processing Time in Auditory Pattern Recognition," Perception and Psychophysics, 1972, Vol. 12, pp. 50-59. 32 Massaro reported that processing time seemed to be more important than stimulus duration; and that the rate of processing was almost equal for the silent and the con- tinuous conditions. Performance was near chance with 50 milliseconds processing time; however, it improved to 70 percent with processing time of 270 millisedonds.46 Massaro considered these findings as further support for a Preperceptual Auditory Store (PAS). The PAS, he suggests, contains information about the acoustic charac- teristics of the stimulus (tone), which decays at a very rapid rate. In a review of literature of auditory memory, Horton and Turnage47 notes that Massaro's findings are con- tradictory to most of the research reported. To account for this discrepancy, Massaro proposed an intermediary stage between the PAS and Short-term memory, which he identifies as the Synthesized Auditory Memory (SAM). Accordingly, the recognition process involves the trans- formation of preperceptual information into a synthesized percept via the SAM, which then channels it into short- term memory.48 46Ibid., p. 52. 47Horton and Turnage, Human Learnipg, p. 200. 48Dominic W. Massaro, "Auditory Information Processing," in Handbook of Learnipg and Cognitive Processes Volume 4 -- Attention and Memory. Edited by W.K. Estes (Hillsdale, N.J.: Lawrence Erlbaum Assoc., Publishers, 1976), pp. 276-277. 33 Research in verbal memory, particularly that by Crowder and Morton,49 tend to support Massaro's theory of a preperceptual auditory memory. They, however, identify their preperceptual store as a Precategorical Acoustic Store. Deutsch50 investigated the effects of spoken numbers vs. tones interpolated during the retention. She found thattflmainterpolated numbers had little effect on pitch retention while the tones were highly disruptive.51 The studies of interference in short-term memory for pitch cited thus far have involved interpolation of single tones only. In 1972, Deutsch52 investigated the effect of interpolating a sequence of four tones (which included the standard tone, or the comparison tone, both or neither) in the retention interval. In this experi- ment the subjects listened to a standard tone which was followed by a sequence of four interpolated tones (300 milliseconds apart), a 2 second pause, and then a 49R.G. Crowder and J. Morton, "Precategorical Acoustic Storage," Perception and Psychophysics, 1969, Vol. 5, pp. 365-373. 50Diana Deutsch, "Tone and Numbers Specificity in Short Term Memory," Science, 1970, Vol. 168, pp. 1604-1615. 51Ibid., p. 1604. 52Diana Deutsch, "The Effect of Repetition of Standard and Comparison Tones on Recognition Memory for Pitch," Journal of Experimental Psychology, 1972, Vol. 93, pp. I56-162. "3.1“. A I I" 34 comparison tone. The subjects were instructed to remember the standard tone, ignore the 4 interpolated tones, and then judge whether the comparison tone was the same as or different than the standard tone. There were ten condi- tions. They were: 1. Neither standard nor comparison tone were pre- sent in the interpolated tones. 2. The standard tone was present as the second interpolated tone. 3. The standard tone was present as the third interpolated tone. 4. Neither the standard or comparison tones were present. The comparison was different than the standard. 5. The standard tone was the second interpolated tone. The comparison tone was different. 6. The standard tone was the third interpolated tone. The comparison tone was different. 7. The comparison tone was the second interpolated tone. 8. The comparison tone was the third interpolated tone. 9. The comparison tone was the second, and the standard tone was the third interpolated tone. 10. The standard tone was the second interpolated tone and the comparison tone was the third interpolated tone . 35 The analysis of the data for the study indicated that in- * serting a tone that was the same as the S tone reduced errors in pitch judgment: inserting a tone that was the ** same as or similar to the C tone increased errors in judgement.S3 In a more recent, but related study, Deutsch,54 re- ported that including a tone that is identical in pitch to the S tone (in an interpolated sequence) has a faciliatory effect on pitch memory. This facilatory effect, however, was found to be extremely sensitive to serial position, i.e., its position relative to the 8 tone. The closer the identical tone was placed to the standard pitch, the better the recognition performance. In a subsequent discussion of this serial effect, Deutsch theorized that the repeated tone produces memory consolidation by trace strengthening.55 A suggested alternate hypothesis was that the subject might have been adopting a particular strategy in which the newer tone replaces the 8 tone, forming a newer trace.56 Deutsch * Standard tone. ** Comparison tone. 53Ibid., p. 156. 54 Diana Deutsch, "Facilitation by Repetition in Recognition Memory for Tonal Pitch," Memory and Cogpition, 1975, Vol. 3, pp. 263-266. 55Diana Deutsch and J. Anthony Deutsch, Short Term Memory (New York: Academic Press, 1975), pp. 135-136. 56Ibid., p. 136. 36 performed a second experiment to test this alternative hypothesis. She concluded from the results that her memory consolidation hypothesis was the correct one.57 In 1974, Deutsch58 conducted an experiment to de- termine the interference effect of interpolated tones taken from octave ranges above and below that of the S and C tones. The S and C tones were taken from the octave c4 to b5. Neither the S nor C tones were included in the inter- polated sequence. There were four conditions in the experi- ment: 1. Condition S - the interpolated was in the same octave as the S and C tones. 2. Condition H - the interpolated tones were taken from the octave above. 3. Condition L - the interpolated tones were from the octave below the S and C tones. 4. Condition H-L - the tones were taken from the octaves above and below the S and C tones. An analysis of the data indicated that the greatest amount of interference occurred in condition H-L, followed by conditions S, H and L, respectively. Deutsch concluded that interference can be produced by tones from a relatively wide range.59 57Ibid. 58Diana Deutsch. 59Ibid., p. 232. 37 In a discussion of the results of this experiment and the other experiments cited, Deutsch offered the follow- ing hypothesis:60 When we listen to a sequence of tones, we pro- cess not only the tones themselves, but also the relationships between them, i.e., the melodic sequence of intervals... She suggests, therefore, that the greater error rate in the condition in which the interpolated tones were drawn from both the higher and lower octave ranges (Deutsch, 1974) was due to the subjects' being less able to utilize pitch re- lationships as compared to the other conditions.61 In the early 1970's researchers began to use pitch sequences and short melodic patterns as stimuli in short- term pitch retention studies. One such study by Dowling and Fujitani62 sought to determine whether melodic sequences are recognized by their contour, interval relationship, or as discrete pitches. Melody, in this study, was defined as a series of intervals between successive pitches. The study consisted of two experiments. The first experiment involved short-term memory with a standard melody, 60Ibid., p. 233. 611bid., p. 234. 62W.J. Dowling and Diane Fujitani, "Contour, Interval and Pitch Recognition in Memory for Melodies," Journal of the Acoustical Sociery of America, 1971, Vol. 49, No. 2, pp. 524-431. 38 and comparison melodies that were either untransposed or transposed from the key of the standard melody. Sub- jects were randomly assigned to one of six groups. The groups were divided equally between the transposed and untransposed conditions. Group 1 (untransposed condition) and Group 2 (transposed condition) determined whether the second of two melodies was the same or different from the first (task 1). Group 3 (untransposed) and Group 4 (trans- posed) determined whether the second of two melodies was the same or had the same contour as the first (task 2). Group 5 (untransposed) and Group 6 (transposed) compared the standard melody with a same or different contour melody (task 3). Subjects in Groups 1, 2, 3, and 4 were instructed to identify the comparison melody as same only if it were identical to the standard melody. Dowling and Fujitani reported that the effects of transposition and task and their interaction were signi- ficant (p < .01). Recognition performance for tasks 1 and 2 were equal in the untransposed condition. It was somewhat more difficult for the transposed conditions. Performance on the third task was better for the trans- posed condition than for the untransposed condition. It was suggested by the investigators that the subjects in the untransposed condition misunderstood the task. . .- I 'w'vra—v ' 39 The conclusions drawn by Dowling and Fujitani were that:63 Subjects utilize discrete pitch information when recognizing brief untransposed melodies: and melodic contour when reCOgnizing trans- posed melodies, i.e., melodic contour is more important than discrete pitch. The second experiment was categorized as long-term memory experiment since the melodies used were familiar folk songs which were considered stored in long-term memory. Subjects were instructed to listen to several melodies and then listen to distorted versions of the same melodies. The distorted versions preserved the harmonic outline, the melodic contour, or the intervallic relationships and con- tour. The results of the study showed that recognition performance for the undistorted version of the songs was 0.99. Recognition performances for the distorted versions were 0.66 for the version which preserved contour and interval size, 0.59 for the version with contour only pre- served, and 0.28 for the version that preserved the harmonic outline. Dowling and Fujitani concluded that: (a) melodic contour was important in the recognition of familiar melodies, and (b) recognition memory (long-term) appears to involve more than just storing melodic contour and relative 63Ibid., p. 531. 40 interval size. Specific pitch relationships seem to be stored as well.64 Serial Effects A review of literature for short-term pitch re- tention yields two experimental studies that dealt with special serial effects. Ortmann,65 in an often cited study, alluded to serial effect in tonal memory when he listed 'order of pitch in a sequence' as a miscellaneous variable in melodic perception. Deutsch (1972)66 noted a possible serial effect with interpolated tones. Taylor67 examined serial effect in an investiga- tion of 25 melodic intervals within melodic contexts (4 and 6 pitch, melodic sequences). One of the questions asked by Taylor was: Is each melodic interval perceived differently as a function of that interval within its melodic context? Four- and six-pitch melodies with test intervals imbedded in the first, center or last position of the 64Ibid., pp. 530-532. 65Otto Ortmann, "Some Tonal Determinants of Melodic Memory," Journal of Educational Psychology, 1933, Vol. 24, pp. 454- 467. 66Deutsch, "Effects of Repetition," p. 161. 67James Taylor, "Perception of Melodic Intervals within Melodic Context" (Doctoral dissertation, University of Washington, 1972). 41 melody, were presented to the subjects. After each pre- sentation subjects were instructed as to which interval to identify. Two numbers were given to indicate the location of the target interval, i.e., 1-2 indicated the interval between the first and second tones of the melody. A 7- second interval followed in which the name of the target interval was written down. Taylor reported that:68 Each interval was perceived differently as a function of its position within its melodic context. 1? intervals were per- ceived less accurately in the center posi- tion than in first and last position. 15 intervals were perceived less accurately in first position than in the last position. A significant interaction (p < .05) was found for the length of melody, and serial positions of the test interval. Taylor concluded that:69 The position of an interval, i.e., first, center or last...within a pitch sequence is a strong variable that significantly influences the perception of all the intervals used in the study. In 1973, Williams,70 using a three-factor repeated measures paradigm, made a more intensive investigation of of the effect of sequence length and serial position on 68Ibid., p. 168. 691bid., p. 154. 70David B. Williams, "Short-Term Retention of Pitch Sequence" (Seattle: Ph.D. dissertation, University of Washington, 1973). 42 short-term pitch retention. The pitch sequences were 3, 5, and 7 pitches in length. There were three serial posi- tions: primary (first), center, and last. The third factor, delay time, had six levels: 0, .5, 1, 3, 5, 7.5, and 15 seconds. Serial position and delay time were randomly ordered across the three sequence lengths, giving a total of 630 trials.71 Each of the 13 subjects were ex- posed to all treatment combinations. The memory task involved a forced-choice procedure (the subject is required to recall or recognize a desig- nated item). A pitch sequence was presented, followed by a delay time, followed by a visual response cue (lighted number), which indicated the position of the pitch that ‘was to be recalled. The subject responded by singing the designated pitch. A judgment was then made (by the experi- menter) as to whether the response matched the designated pitch. Because the six replications of a treatment combi- nation were pooled, a subject's score could range from 0 to 6. An analysis of the data indicated a significant' difference (p < .01) for the main effects of sequence length, serial position, and delay time. All two-factor 711bid., p. 31. 43 :Lnteractions were noted to be significant (p < .05). There was no significant three-factor interaction. concluded that : 72 On the basis of the results of his study, Williams 73 1. Loss of retention in short-term memory occurs as a result of increased delay time before recall, the pcsition of a pitch within a sequence (recency primacy center), and the increased length of a sequence . 2. Retention of a pitch sequence is a func- tion of delay time before recall, serial position apd sequence Iength, and tfieir combined effects. In regard to the specific effects of serial position, Williams asserted that: 74 The loss of information in the recency and primacy positions of a pitch sequence is dependent upon both time and item decay, and their combined effects. Loss of infor- mation in the center position, on the other hand, is a function of item decay, not time decay. Furthermore, the behavior of the recency and primacy positions in not the same. Loss of information in the recency position is equally one of time and item decay, whereas, loss of information in the primacy position is more one of item decay rather than time decay. 72 73 74 Ibid., pp. 142-143. Ibid., Ibid., p. 144. p. 144. 44 In a most recent study of short-term pitch memory, Long75 examined the effect of melody length, tonal structure, melodic contour, and music perception ability on memory for pitch. She also investigated the relation- ship between melodic memory and the amount of tonal in- formation contained in a melody. Melody lengths were 7, 11, and 15 pitches. Subjects were placed in one of three groups accord- ing to their perceptual ability. Group 1 consisted of graduate music majors; Group 2 was composed of undergraduate music majors: and Group 3 were non-music majors. Subjects in each group listened to 12 specially composed melodies. Each melody was heard twice. On the first hearing, each melody was followed by two seconds of silence, and a one- second test tone. The instructions were to remember the melody and determine if the test tone had occurred in the melody. Each melody was presented once followed by a test tone (correct one) thatlunibeen heard in the melody. On the second hearing the melody was followed by a test tone that deviated by a half-step (above or below) the first tone. This tone had not occurred in the melody. Subjects responded, indicating the confidence of their decision on 7SPeggy Ann Long, "Pitch Recognition in Short Melodies" (Tallahassee: Unpublished Ph.D. dissertation, The Florida State University, 1975). In Dissertation Abstracts, 1976, No. 4083). 45 a six-point rating scale (+3 very sure, yes, -3 very sure, no). Long reported that the results of the analysis of variance showed the factors of groups and tonal structure to be significant (p < .01). Memory performance by the groups composed of music majors was better than that of the non-music major group. Tonal melodies were remembered better by all three groups. As the length of melodies in— creased, memory performance decreased. All other tests were found to be not significant (p < .01.). The findings of Long tend to support Taylor (1972) and Williams (1973) in regards to pitch length. Shorter melodies are remembered better than melodies of longer duration. Timbre Perception According to the literature, there has been little study of the influence of timbre on perception and reten- tion. Mainwaring,76 in an early study of recall memory, decided that the performance medium of a melody was unim- portant in its recall. Usingtflmaintrospective report of subjects to the question" "Whether it was easier to recall 76James Mainwaring, "Kinesthetic F-ctros in Recall of Musical Experience," British Journal of P§ycholqu: 1933, Vol. 23, p. 295. 46 a tune that had been presented on one instrument rather than another," Mainwaring found that subjects responded indecisively both times that question was asked. He con- cluded that the subjects' indecision provided evidence that the medium of tone is forgotten or completely ignored.77 The five-year longitudinal study of auditory per- ception by Petzold78 is one of the most extensive inves- tigations of music memory undertaken. The major purpose of the study was to determine the differences between children at each of the first six grade levels, in the manner in which they perceive and respond to the auditory presentation of musical sounds. A secondary aspect of the study dealt with the in- fluence that combinations of the basic music elements, i.e., rhythm, timbre, and harmony, might have upon the auditory perceptions of melodic sequences.79 Since the findings of Dr. Petzold's study are well known and are frequently reported in the literature, Only the timbre study will be discussed. Although the timbre study was a pilot project within the overall longitudinal study, it is perhaps the 77Ibid. 78Robert G. Petzold, Auditory Perception of Musical Sounds by Children in the First Six Grades (Madison: University of Wisconsin Press, 1966). 791bid., p. 253. 47 most extensive investigation of the effects of timbre on auditory perception to date. Petzold stated that the pur- pose of the timbre study was to seek answers to the follow- ing question: What kinds of relationships exist between the type of performance medium utilized for the aural pre- sentation of melodic items, and the accuracy with which children perceive and reprodue these items? Subsidiary questions were concerned with determining whether any single performance medium seemed to be most appropriate for use at a given grade level, and whether one medium of performance led to error responses that were markedly different from those of other performance media. Originally eight timbre qualities were to be used in the study: piano, soprano voice, flute, violin, tenor voice, French horn, trombone, and cello, representing SOprano, tenor and bass ranges. An exploratory study found that the children were unable to make octave trans- POSitiODS With their voices; therefore, the tenor and bass range timbres were elimianted in the actual study. Four forms of the Timbre Test were recorded, utilizing the piano, soprano voice, flute, and violin. The results of the study indicated that all three main effects (timbre treatment, grade, and sex) were significant (p < .01). Thus, the use of different per- formance media for presenting melodic items produced significant differences in test scores, regardless of grade 48 80 . . level or sex. There was, however, no interaction between the main variables. The data also showed that melodic items performed on the violin were responded to with greater accuracy, followed by the voice and then the piano and flute, in that order. Petzold could not account for this result.81 In summarizing the results of this study, Petzold suggested that the four tests did not cover as broad a range of possibilities as they should have and that further research should be undertaken.82 Recently, Williams83 found that changes from a 'bland' to a 'rich' timbre affected children's perception of melodic motion. Thirty-two second-grade and thirty-two fifth-grade students took part in an experiment in which the task consisted of identifying the unidirectional motion of two-pitch melodic patterns (unisons, ascending-descending minor seconds, thirds, and perfect fourths). The timbre of the initial tone of each pattern was varied either as a rich timbre (strong in harmonics) or a bland timbre 80Petzold, p. 129. 81Ibid., p. 134. 82Ibid., p. 144. 83David B. Williams, "An Interim Report of a Programmatic Series of Music Inquiry Designed to Investigate Melodic Pattern Identification Ability in Children," Council for Research in Music Education, 1976, pp. 78-83. 49 (deficient in harmonics). The second tone of all the patterns was a rich timbre. After listening to a pattern, the child, by pushing one of three buttons, indicated whether the pattern ascended, descended or remained stationary. Williams reported that changes from the rich timbre to the bland timbre were perceived as ascending motion. The effect was particularly dominant with descending minor seconds.84 SUMMARY The review of literature supports evidence for time decay and interference as causative factors for loss of in- formation in short-term pitch retention. The studies of Solf (1896), Angell (1899), Whipple (1901), Anderson (1914), Koester (1945), Harris (1952), Pollack (1952), and Bachem (1954) have shown that memory for pitch gradually dete- iorates over time. The research of Wickelgren (1966, 1969), Massaro (1970, 1971, and 1972), and Deutsch (1970-1974) has shown that short-term memory for pitch is also subject to interference. In addition, the research of Massaro has provided evidence for a sensory storage system -- the Preperceptual Auditory Store (PAS) which holds auditory information for 84Ibid. 50 a short time (about 250 msec) after a stimulus has been presented. And, what is perhaps more important, Massaro's studies have provided a quantitative measurement of auditory imagery. Serial effects in pitch memory were implied by Ortmann (1933) and reported by Taylor (1972) and Williams (1973). Dowling and Fujitani indicate that contour is important in the preservation of transposed melodies. Taylor, Williams (1973) and Long (1975) have shown that the length of a melody or pitch sequence also affects the short-term retention of that melody, longer melodies being less well remembered. The review of literature for timbre perception re- veals inconclusive results, particularly for timbral effects on immediate perception and retention. The findings of Mainwaring, Petzold and Williams (1975) suggest, however, that timbre may affect perception. Finally, the review of literature has shown that there are similarities between pitch memory and verbal memory. Shuter85 suggests that memory for music is analogous to verbal memory. The research literature does provide evidence that pitch memory and verbal memory are subject to the same effects, i.e., decay over time, inter- ference, and serial position. 85Rosamund Shuter, Psychology of Musical Ability (London: Methuen and Co., Ltd., 1968), P. 202. 51 Deutsch, on the other hand, interprets the findings of her research to indicate that memory for pitch and ver- bal memory are different. She asserts that the two are distinct systems.86 Bower supports this contention, stating that:87 We can remember something about nonverbal events even though we can't describe them ...They seem to be represented in terms of analogical structures that are not nec- essarily connected to the verbal system. 86Diana Deutsch, "Short-Term Memory," pp. 145-146. 87Gordon H. Bower, "Introduction to Concepts and Issues," in Handbook of Learning and Cognitive Processes. Edited by W.K. Estes (New York: John W. Wiley and Sons, 1975), p. 57. CHAPTER III DESIGN OF THE STUDY Sample R The sample for this study consisted of twenty- eight Michigan State University undergraduate students enrolled in Music 271, a fundamentals of music class for non-music majors. The subjects were from various schools and departments within the university. There were 11 female students and 17 male students. All academic classi- fications (freshman, sophomore, junior and senior) were represented in the class structure. Data obtained from student information cards (completed during the first week of the term) indicated that the members of the class possessed varying degress of experience with music, ranging from little or no experience to several years of prior study. Instrumentation An investigation designed pitch-memory test - the Pitch-Timbre Memory Test (P-TMT) was used to collect the data for this study. For several reasons it was deemed necessary to construct a test rather than use one of the existing published tests. One reason was to provide a 52 53 control for prior experience with music memory tests. Another reason was to control for other variables such as tonality, rhythm, and the number of pitches used in each sequence. The actual development of the test took approxi- mately a year and a half. At the outset of the study, the investigator had planned on using acoustic instruments (Flute, clarinet, trumpet, French horn, cello, and voice) as the sound sources for the test items. These instru- ments were used in the preparation of the initial version of the test. The test itself contained 20 items. Each item consisted of two, six-pitch sequences composed by the investigator. Six Michigan State University undergraduate music students performed and recorded the test. It was discovered during the playback of the recorded test that the items were not performed equally, i.e., there were differences in articulation, intonation, dynamics, and attacks and releases, between the items. After several subsequent recordings it was concluded that the problem of controlling for performer differences was of such magnitude as to make the use of acoustic instruments im- plausible. It was also determined that the pitch sequences were tonally biased because of the investigator's un- conscious preference for certain intervallic combinations. The decision therefore, was made to (a) have the test items programmed by a computer and (b) use a syn- thesizer as the sound source for the test items. The 54 revised version of the test was completed during the Winter term of 1977. In its final form, the Pitch-Timbre Memory Test contained 48 items. Each item consisted of two, four- pitch sequences. The atonal pitch sequences were de- veloped by the CDC-6500 computer system located in the Computer Center at Michigan State University from a pro- gram written especially for this study. The pitch sequences were generated by an Arp 2600 synthesizer and recorded on a TEAC 33440-8 tape recorder. The automatic trigger of attack and release was accomplished by the internal clock oftimasample and hold. The four-pitch sequences were played at a rate of four sixteenth notes in a quarter note, m.m. J = 60. The four wave forms (sine, sawtooth, pulse and triangle), which constituted the timbres of the pitch sequences, came out of voltage control oscillator No. 2 of the enve10pe generator. The sine wave was used as the standard timbre. All four timbre qualities were randomly assigned as comparison timbres. The test tapes were prepared and edited (under the supervision of the investigator) by professional recording engineers at a professional recording studio in East Lansing, Michigan. Two members of the Michigan State University music faculty listened to the pitch sequences in order to certifytfimfiursuitability as atonal pitch sequences. The 55 Table 3.1 Pitch—Timbre Memory Test Format Item. Sequence A Retention Timbre Sequence B No. Interval 1 sine 5 sec. sine same 2 sine 30 sec. sine different 3 sine 15 sec. sawtooth different 4 sine 15 sec. sine different 5 sine 5 sec. sine different 6 sine 15 sec. sine same 7 sine 5 sec. pulse same 8 sine 15 sec. sawtooth different 9 sine 15 sec. sine same 10 sine 15 sec. triangle same 11 sine 5 sec. pulse different 12 sine 30 sec. pulse same 13 sine 5 sec. sine different 14 sine 15 sec. sine different 15 sine 5 sec. sine different 16 sine 30 sec. sine same 17 sine 15 sec. triangle different 18 sine 15 sec. pulse different 19 sine 5 sec. sine same 20 sine 15 sec. triangle same 21 sine 5 sec. sawtooth different 22 sine 5 sec. triangle same 23 sine 30 sec. sine same 24 sine 5 sec. sine different 25 sine 30 sec. pulse different 26 sine 30 sec. sawtooth different 27 sine 15 sec. sine same 28- sine 15 sec. sine different 29 sine 15 sec. sawtooth same 30 sine 30 sec. triangle different Table 3.1 (continued) 56 Item Sequence A Retention Timbre Sequence B No. Interval 31 sine 30 sec. triangle same 32 sine 15 sec. pulse same 33 sine 5 sec. sine same 34 sine 5 sec. sine same 35 sine 30 sec. sawtooth different 36 sine 5 sec. triangle same 37 sine 30 sec. sine different 38 sine 30 sec. sine different 39 sine 30 sec. sine same 40 sine 5 sec. pulse different 41 sine 15 sec. sine different 42 sine 30 sec. sawtooth same 43 sine 5 sec. triangle different 44 sine 30 sec. sine different 45 sine 30 sec. sine same 46 sine 30 sec. pulse same 47 sine 15 sec. sine ,same 48 sine 5 sec. sawtooth Same 57 pitch sequences were also judged atonal by a class of non-music major undergraduate students enrolled in Music 271. (The distinction between tonal and atonal had been explained to the class.) One of the three retention intervals (5 seconds, 15 seconds, or 30 seconds) was used between the first and second pitch sequences of each item. There was a 5 second response interval between each item. Two test tapes were used in the study. Both tapes contained the memory test: however, there were differing sets of instructions at the beginning of each tape. Test-retest reliability was established with r = .92 using the Pearson Product Moment 9 Correlation formula. Table 3.1 shows the test format. Design A 2 x 2 x 3 x 4 split-plot factorial design (Winer, 1971)1 was selected as the basic research design fortflxastudy. The independent variables were Timbre, Retention Interval, Task and Same/Different Items. Factor A, Timbre had 4 levels (sine wave, sawtooth wave, pulse wave, and triangle wave). Factor B, retention interval, had three levels: 5 seconds, 15 seconds and 30 seconds. Factor C, task, had two levels, same and different. Factor D, item-type had two levels, same sequence and different lB.J. Winer, Statistical Principles in Experimental Design, 2nd Edition (New York: McGraw-Hill, 1971, pp. 367-371). 58 sequence. The dependent variable was response performance (test score), as measured by the Pitch-Timbre Memory Test. Procedures Subjects were randomly assigned to one of two groups. Each group contained 14 subjects. Group I, took the Pitch-Timbre Memory Test, and performed a task during the retention interval of each item. The task consisted of counting backwards by threes from a specified number announced immediately after the first pitch sequence. The task is an interference procedure developed by Peterson and Peterson,2 to study forgetting in short-term memory. To insure that they were really counting, subjects were instructed to write the numbers as they counted. Group II, the no-task group, simply took the test. Figure 3.1 shows the structure of the test tapes. Although the instructions and practice items were recorded, the procedures were also explained and dia- grammed on the blackboard prior to the playing of the test tape. Both test tapes were played on a Sony TC 155 tape recorder connected to a Scott Solid State Amplifier Model 299-1, and two large Jensen Speakers. The ampli- fiers and speakers were part of the classroom's audio system. 2Lloyd Peterson and M.J. Peterson, "Short-Term Retention of Individual Verbal Items," Journal of Experimental Psycholong 1959, Vol. 58, pp. 193-198. 59 a. Conditions for the Non-Task Group Seconds A B . 5/15/30 ORGSPOHSG b. Conditions for the Task Group A Seconds B ~\ 5/15/30 Number Cue for“"//> Cue to Second Begin Sequence Counting Figure 3.1 The Pilot Study A pilot study was undertaken during the term prior to the actual testing to determine procedural efficacy. The sample size and the conditions were the same as in the actual study. All procedures were found to be gen- erally satisfactory. It was found, however, that some subjects in the ra§§_group did not understand the pro- cedures. It was also discovered that once the test had begun that several of the students did not always perform the task. Therefore, during the study itself, a careful but unobtrusive observation was maintained for the entire testing period. 60 Treatment of Data The tests were scored by the Michigan State Univer- sity Testing Service, whose services included transferring the scores to data cards. The actual data analysis was done by a CDC 6500 computer system housed in the Computer Center at Michigan State University. ANOVAH, a special analysis of variance program that treats nested classifica- tions, was used to examine the main effects and interac- tions. Testable Hypotheses For experimental purposes the hypotheses advanced in this study were stated in their 2211 form. H01: No difference will be found in the recognition response mean scores among the four types of timbre (wave forms) used in the second sequence or each item of the pitch-timbre memory test. H 2: No difference will be found in recognition re- sponse mean scores as a result of change in the length of the retention interval. H03: No difference will be found in the recognition response mean scores between the task group and the non-task group. H04: No difference will be found in recognition re- sponse mean scores between test items in which the second pitch sequence is the same and test items in which the second sequence is different. H 10: 11: 12: 13: 61 No significant interaction (AB) will be found be- tween timbre (A) and retention interval (B). No significant interaction (AC) will be found be- tween timbre (A) and task (C). No significant interaction (AD) will be found be- tween timbre (A) and same sequence items vs. different sequence items (D). No significant interaction (BC) will be found be- tween retention interval (B) and task (C). No significant interaction (BD) will be found be- tween retention interval and same sequence items vs. different sequence items. No significant interaction (CD) will be found be- tween task (C) and same sequence item vs. dif- ferent sequence item (D). No significant interaction (ABC) will be found between timbre (A), retention interval (B) and task (C). No significant interaction will be found between timbre (A), retention interval (B), and same sequence items vs. different sequence items (D). No significant interaction (ACD) will be found between timbre (A), task (C), and same sequence item vs. different sequence item (D). H 14: H 15: 62 No significant interaction (BCD) will be found between retention interval (B), task (C) and same sequence items vs. different sequence items (D). No significant interaction (ABCD) will be found between timbre (A), retention interval (B), task (C) and same sequence items vs. different sequence items (D). CHAPTER IV ANALYSIS OF THE DATA The data for this study were processed and analyzed by the computer system located at Michigan State Univer- sity. An SPSS (Statistical Package for the Social Sciences) program was utilized to determine the means for the 24 types of test items for each subject. These were coded into a special format and instrumented through the ANOVAH program which examined the main effects and interactions by analysis of variance techniques. The ANOVAH1 program is a FORTRAN IV Program designed to perform N-way analysis of variance. It was revised for use (by the Office of Research Consultation, College of Education) on the CDC 6500 computer at Michigan State University. The routine will perform the calculations necessary for balanced, fully replicated or nested factorial designs and can handle any design up to 9 factors (counting replications). Table 4.1 illustrates the split-plot factorial designed data matrix with descriptive statistics for the 24 types of question. 1ANOVAH was originally programmed by Tom Houston in FORTRAN- 63 for the CDC 1604 at the Laboratory of Experimental of the Wisconsin ResearchanuiDevelopment Center for Cognitive Learning. 63 64 .Hawo some Ga «H fl 2 « M swam. ooom. smam. same. I amen. mmqa. meme. ames. .666 on mm W meme. awe». omen. smmm. I amen. Heme. mmm~. smms. .omm ma mm mamcmspa _ omam. same. mmmm. some. 1 mass. meas. seem. same. .666 m an «4 l - m pmam. «Hem. mums. smmv. amps. mean. Hmom. mmeo. .uom om mm w swam. ooom. some. smmm. A «has. mmsm. «has. Hemm. .omm ma mm amend ” swam. coon. oooo. oooo.H Name. damn. mess. mean. .666 m Hm mm m mmmm. some. smam. meme. vemm. Hemm. amen. asmm. .666 on mm M . . m Hmmm. emmm. mmms. meNm. seam. seam. Hmsm. «mes. .omm ma mm spoouzmmm _ smam. same. mmme. smms. maps. emmm. amen. Heme. .666 m an ma” Hmwm. some. mosm. boas. mmam. mmes. seem. mamm. .666 on mm m _ . smem. eflem. HGHN. emmm. mspm. sasm. Hemm. Heep. .omm ma «m was» _ n ommd. mass. mmsm. when. Amps. mess. mamm. Hmms. .666 m Hm HA _ Cm cmwz Om Cflmz 0m .3002 am cmwz HM>H$HCH _ ucoummmao mfimm ucmummMmm, mfium c0wucmuom N Amoum3xomnmceucaoov A _ U xmmaucoz O xmma _ cmwmmn Hmwuouomm uon uwamm on» MOM xfluumz mama H.e manna i 65 Table 4.2 ANOVA Summary for Main Effects and Interactions (Timbre = A Time = B Task = C Item = D) Sum oIT Mean Source of Variation Squares df Squares F Timbre (A) 1.574 3 .524647 3.76* within 10.863 78 .139270 Retention Interval(B) 2.018 2 1.009022 7-97* within 6.582 52 .126579 Task-Groups (C) .227 1 .226968 2.73 within 2.155 26 .082868 Items (D) 1.146 1 1.145926 4.03 within 7.387 26 .284104 A x B 6.402 6 1.067057 7.74* within 21.511 156 .137892 A x C .360 2 .120123 .86 within 10.863 78 .138270 A x D .380 3 .126581 .745 within 13.236 78 .169689 B x C .380 2 .190026 1.50 within 6.582 52 .126574 B x D 6.754 2 3.376953 17.63** within 9.960 52 .191543 C x D .635 1 .634557 2.23 within 7.387 26 .284104 A x B x C 1.669 6 .278121 2.01 within 21.511 156 .137892 A x B x D 1.922 6 .320406 2.19* within 22.787 156 .146072 A x C x D .678 3 .225426 1.33 within 13.236 78 .169689 B x C x D 1.908 2 .953865 4.97* within 9.960 52 .191543 A x B x C x D .964 6 .160711 1.10 within 22.787 156 .146072 * p < .05 *‘k p < .01 66 Results Table 4.2 shows a summary of the ANOVA results for the main effects and interactions. The presentation of data will continue with a con- sideration of the results of the tests of the individual null hypotheses. H01: No difference will be found in the recognition response mean scores between the four types of timbre (wave forms) used in the second sequence of each item of the pitch-timbre memory test. A primary concern of this study was the effect of changes in timbre on perceptions and retention. The re- sults of the ANOVA tests on this main effect show that F = 3.76 is significant beyond the .05 level (Table 4.3). Therefore null hypothesis 1 was rejected. Table 4.3 ANOVA for Timbre Effect (Factor A) Sum of Source of Variation Squares df Mean Square F Timbre (A) 1.574 3 .524647 3.76* within 10.863 78 .139270 * p < .05 A Neuman-Kuels test was performed to determine the difference between the means found that the mean score at the sawtooth wave differed significantly (p < .05) from those of the other wave forms, i.e., sine, pulse, and triangle (Table 4.4). 67 Table 4.4 Results of Neuman-Kuels Test of Timbre Means Timbre .640 .582 .560 .426 (triangle) . (sine) rpulse) (sawtooth) .640 Vi? - I .022 .044 1.28* m .582 ' - .022 1.06* g .560 - .084* Q) I e .472 g 1'_ * significant at the .05 level H 2; No difference will be found in recognition mean response scores as a result of change in the length of the retention interval. Table 4.5 shows the results of the ANOVA tests for null hypothesis two. The F ratio 7.97 was significant at the .05 level. Null hypothesis two was therefore rejected. Table 4.5 ANOVA for Retention Interval Effect Factor (B) Sum of I Source of Variation Squares df Mean Squares F Retention Interval (B) 2.018 2 1.009022 7.97* within 6.582 52 .126529 * p < .05 A post-hoc test (Neuman-Kuels) performed to deter- mine the difference between the means for the main effect of retention interval found that the 15 second retention 68 interval and the 30 second interval mean scores differed significantly (p < .05) from that of the 5 second interval (Table 4.6). Table 4.6 Results of the Neuman-Kuels Test of Retention Interval Means 0.625 (5 sec.) 0.550 (30 sec.) 0.491 (15 sec.) 0.625 - 0.075* 0.134* 0.550 - 0.059 0.491 - * significant at the .05 level 3: No difference will be found in the recognition response mean scores between the ragk gropp and the pop-ragk group. H 4: No difference will be found in the recognition response mean scores between test items in which the second pitch sequence is the gamg and test items in which the second pitch sequence is gr:- ferent. The results of the ANOVA tests for the main effects of task and item-type are shown in Table 4.7. The F = 2.73 for task effect and the F = 4.03 for the item- type effect were found to be non-significant at the .05 level. Null hypotheses 3 and 4, there, failed to be re- jected. 69 Table 4.7 ANOVA for Task and Item-Type Effect (Factors C and D) Sum of Source of Variation Squares df Mean Square F Task (C) .227 1 .226968 2.73 within 2.155 26 .082868 Item-type (D) 1.146 1 1.145926 4.03 within 7.387 26 .284104 H05: No significant interaction (AB) will be found between timbre (A) and retention interval (B). Interaction effects between timbre and retention were found to be significant F = 7.74 at the .05 level. The ANOVA results are shown in Table 4.8. Null hypothesis 5 was rejected. Table 4.8 ANOVA for the Interaction (AB) of Timbre (A) and Retention Interval (B) Sum of Source of Variation Squares df Mean Square F Timbre x TI (A x B) 6.402 6 1.067057 7.74* within 21.511 156 .137892 * p < .05 Figure 4.1 shows the interaction (AB) of timbre (A) and retention interval (B). Means .90 .30 .70 .60 .30 .40 .30 .20 .10 70 Triangle sawtooth seconds Figure 4.1 Timbre/Retention Interval Interaction (AB) No significant interaction (AC) will be found be- tween timbre (A) and task (C) No significant interaction (AD) will be found be- tween timbre (A) and item-type (D). The ANOVA tests for interaction effects between timbre and task, and timbre and item type indicated non- significance for both interactions (Table 4.9). Null hypotheses 6 and 7 failed to be rejected - p < .05. 71 Table 4.9 ANOVA for the Interaction Effect of Timbre/Task (AC) and Timbre/Item (AD) Sum of Source of Variation Sguares df Mean Sguare F Timbre/Task (AC) .360 3 .120163 I .86 within 10.863 78 .139270 I Timbre/Item (AD) .380 3 .126581 I .75 within 13.236 78 .169689 I H08: No significant interaction (BC) will be found be- tween retention interval (B) and task (C). The results of the ANOVA indicated no significant interaction between retention and task (Table 4.7). There- fore the null hypothesis failed to be rejected. H09: No significant interaction (BD) will be found between retention interval (B) and item-type (D). A significant interaction, F = 17.63 at the .05 level, was found between retention interval and item-type (Table 4.10). Figure 4.2 graphs the (BD) interaction. H010: No significant interaction (CD) will be found between task (C) and item type (D). As indicated by Table 4.10 the interaction between task and item type were found to be non-significant. Null hypothesis 10 failed to be rejected. Means 72 .90 .80 \\ .70 \\ ..o \ \ .50 \—------¢ Same pitch sequence /-:——---- different pitch sequence .40 .30 .20 .10 5 15 30 Retention interval in seconds Figure 4.2 The Interaction (BD) between Retention Interval (B) and Item Type (D) Table 4.10 The ANOVA for the Interaction Effects between Retention Interval and Task (BC), Retention Interval and Item Type (BD) and Task and Item Type (CD) Sum of Source of Variation Squares df Mean Square F RI - Task (BC) .380 2 .190026 1.50 within 6.582 52 .126579 RI - Item (BD) 6.754 2 3.376953 17.63* within 9.960 52 .191543 Task - Item (CD) .635 1 .634557 2.23 within 7.387 26 .284104 i: p < .05 73 H011: No significant interaction (ABC) will be found between timbre (A) retention interval (B) and task (C). The ANOVA indicated that no significant inter- action occurred between timbre, retention interval and task. Table 4.11 shows that the obtained F ratio was not significant. Null hypothesis 10 failed to be rejected. H012: No significant interaction (ABD) will be found be- tween timbre (A) retention interval (B) and item type (D). The triple interaction (ABD) between timbre, reten- tion interval and item type was shown to be significant F = 2.19 (Table 4.11) at the .05 level. Null hypothesis 12 was rejected. Table 4.11 ANOVA for Interaction Effects (ABC) between Timbre (A), Retention Interval (B) and Task (C), and the Interaction Effects (ABD) between Timbre (A), Retention Interval (B) and Item Type (C) Sum of Source of Variation Squares df Mean Square F Timbre - RI - Task (ABC) 1.669 6 .278121 2.01 within ' 21.511 156 .137892 Timbre - RI - Item . (ABD) 1.922 6 .320406 2.19* within 22.787 156 .146072 * p < .05 74 Figure 4.3 shows a line plot of the triple interaction (ABD) between timbre, retention interval and item type. Means Me an S .80 .70 .60 .50 .40 .30 .20 .10 .80 .70 .60 .50 .40 .30 .20 .10 5 15 30 Same sequence \ Triangle 5 15 30 Different sequence Figure 4.3 75 H013: No significant interaction (ACD) will be found between timbre (A), task (C) and item type (D). The results of the ANOVA indicated a non-significant interaction effect between timbre, task and item type (Table 4.12). Null hypothesis 13 failed to be rejected. H014: No significant interaction (BCD) will be found between retention interval (B), task (C), and item type (D)- The interaction effect between retention interval, task, and item-type was found to be significant (F = 4.97) at the .05 level. Null hypothesis 14 was rejected. Table 4.12 shows both (ACD) and (BCD) ANOVA results. Table 4.12 ANOVA for the Interaction Effects (ACD) between Timbre, Task, and Item, and Interaction Effects (BCD) between Retention Interval, Task and Item Sum of Source of Variation Squares * df Mean Square F Timbre x Task x Item (ACD) .678 3 .225926 1.33 within 13.236 78 .169689 RI x T x I (BCD) 1.908 2 .953865 I 4.97* within 9.960 52 .160711 * p < .05 Figure 4.4 shows a graph of the interaction between retention interval, task and item type. Means Means .90 .80 .70 .60 .50 .40 .30 .20 .10 .90 .80 .70 .60 .50 .40 .30 .20 .10 76 5 15 30 seconds Gp l (task) 0 \ \ \ \ \ \ \ \ I \ ,I’ ‘,” O 5 15 30 seconds Gp 2 (non-task) Figure 4.4 77 H015: No significant interaction (ABCD) will be found between timbre (A), retention interval (B), task (C) and item type (D). The interaction between timbre (A), retention in- terval (B), task (C), and item type (D) was found to be non-significant. Null hypothesis 15 was, therefore failed to be rejected. Summary As a result of the analysis of data null hypotheses l, 2, 5, 9, 12, and 14 were rejected. The remaining null hypotheses failed to be rejected. The main effects of timbre (A) and retention in- terval (B) were found to be significant at the .05 level Interaction effects of timbre and retention interval (AB), retention interval and item type (BD); timbre, retention interval and item (ABD): and retention interval, task and item type (BCD) were also found to be significant (p < .05). CHAPTER V SUMMARY AND CONCLUSIONS The primary purpose of this study was the investiga- tion of the effect of changes of timbre, length of reten- tion interval (time) and performance of a task (during the retention interval) on short-term recognition memory for pitch. A secondary concern was the response perfor- mance on test items in which the second sequence was the same as the first, and those test items in which the second sequence was different from the first. A review of short-term memory for pitch reveals that there has been little research on the influence of timbre on perception and retention. Research by Mull,1 Main- waring,2 Petzold,3 and Williams,4 indicate a possible lHelen K. Mull, "The Acquisition of Absolute Pitch," The American Journal of Psychology, 1925, p. 489. 2James Mainwaring, "Kinesthetic Factors in Recall of Musical Experience," British Journal of Psychology, 1933, Vol. 23, p. 295. 3Robert G. Petzold, Auditory Perception of Musical Sounds py Children in the First Six Grades (Madison: UniversiEy of Wisconsin Press, 1966). 4David B. Williams, An Interim Reprot of a Programmatic Series of Music Inquiry Designed to Investigate Melodic Pattern Identification Ability in Children," Council for Research in Music Education, 1976, pp. 78-83. 78 79 timbre effect. Petzold,5 whose research involved a more direct investigation of timbre influence, concluded that timbre did not significantly affect the recall of tonal patterns, suggested, however, that further research was needed. The literature on short-term memory for pitch is replete with investigations of the effect of time decay, and interference on the retention of a single pitch. Studies by Anderson,6 Harris,7 and Bachem,8 show that short term memory for pitch deteriorates over time. Later re- search by Wickelgren,9 Massaro,10 and Deutsch,11 indicated that short-term memory for pitch is also subject to inter- ference of these effects on the retention of pitch sequences. 5Petzold, p. 129. 6D.A. Anderson, "The Duration of Tones, the Time Interval, the Direction of Sound, Darkness, and Quiet, and the Order of Stimuli in Pitch Discrimination," Psychological Mono- graphs, 1914, Vol. 16, pp. 150-156. 7J. Donald Harris, "The Decline of Pitch Discrimination with Time," Journal of Experimental Psychology: 1952, Vol. 43, pp. 96-99. 8A. Bachem, "Time Factors in Relative and Absolute Pitch Determination," Journal of the Acoustical Society of America, 1954, V01. 26. pp. 7512753. 9Wayne A. Wickelgren, "Consolidation and Retroactive Inter- ference in Short-Term Recognition for Pitch," Journal of Experimental Psychology, 1966, Vol. 72, No. 2, pp. 250-259. 10Dominic W. Massaro, "Retroactive Interference in Short- Term Recognition Memory for Pitch," Journal of Experimental PsycholQflX, 1970, Vol. 83, pp. 32-39. 11Diana Deutsch, "The Effect of Repetition of Standard and Comparison Tones on Recognition Memory for Pitch," Memory and Cognition, 1975, Vol. 3, pp. 263-266. 80 A four factor split-plot design (Winer, 1971, p. 367-371; Kirk, 1968, p. 308-311) was adopted for this study. The independent variables were timbre (Factor A), retention interval (Factor B), task (Factor C) and item type (Factor D). The dependent variable was accuracy of recognition response measured by performance on a pitch-sequence memory test. The data gathering instrument was a 48 item pitch- sequence memory test. The Pitch-Timbre Memory Test was designed by the investigator especially for this study. Each item consisted of two pitch sequences. The second sequence was either the same or different from the first. For twenty-four of the items, the same sound quality (a sine wave) was used for both pitch sequences of the item. The second sequence of the remaining twenty four items was either a sawtooth pulse or triangle wave form. The re- tention intervals were either 5 seconds, 15 seconds, or 30 seconds. All items were randomized. The twenty-eight subjects used in the study were randomly assigned to one of the two groups. Group I re- ceived the memory test and performed a task during each of the retention intervals. The task consisted of count- ing backwards by three from a number specified on the memory test tape. To insure that they were counting subjects wrote the numbers as they counted. Group II only took the memory test. 81 Findings Four primary hypotheses and eleven secondary hypo- theses were examined in their null form. ANOVAH, an analysis of variance computer program which handles vested factorial and repeated measure designs, was used to analyze the data. According to the data analysis, the following re- results are reported. H01: No difference will be found in the recognition response mean scores among the four types of timbre (wave forms) used in the second sequence of each item of the pitch-timbre memory test. A significant difference was found between the mean socre of the sawtooth wave and the mean scores of each of the other wave forms (p < .05). Null hypothesis one was rejected with an F ratio of 3.76 which is significant be- yond the .05 level. H02: No difference will be found in recognition response mean scores as a result of change in the length of the retention interval. A significant difference was found between the mean scores of the 16 second and 5 second retention intervals, and between the mean scores of the 30 second and 5 second retention interval (p < .05). Null hypothesis two was rejected F = 7.97 which is significant beyond the .05 level. 82 H 3: No difference will be found in the recognition re- sponse mean scores between the task group and the non-task group. Null hypothesis three failed to be rejected. There was no significant difference between the means of the task group and the non-task group. H04: No difference will be found in recognition response mean scores between test items in which the second pitch sequence is the same and test items in which the second sequence is different. Null hypothesis four failed to be rejected. There was no significant difference between the means of the two types of items. H05: No significant interaction (AB) will be found be- tween timbre (A) and retention interval (B). Interaction effects between timbre and retention interval were found to be significant F = 7.74 at the .05 level. Null hypothesis five was rejected. H06: No significant interaction (AC) will be found be- tween timbre (A) and task (C). There was no significant interaction effect be- tween timbre and task. Null hypothesis six failed to be rejected. H07: No significant interaction (AD) will be found be- tween timbre.(A) and same sequence items vs. dif- ferent sequence items (D). 83 There was no significant interaction between timbre and items. Null hypothesis seven failed to be rejected. H08: No significant interaction (BC) will be found between retention interval (B) and task (C). Null hypothesis eight failed to be rejected. There was no significant interaction between retention interval and task. H09: No significant interaction (BD) will be found between retention interval and same sequence items vs. dif- ferent sequence items. A significant interaction was found between retention interval and items. The F ratio of 17.63 was found to be significant at the .05 level. Null hypothesis nine was rejected. H010: No significant interaction (CD) will be found be- tween task (C) and same sequence item vs. different sequence item (D). Null hypothesis ten failed to be rejected. There was no significant interaction between task and item. H 11: No significant interaction (ABC) will be found be- tween timbre (A), retention interval (B) and task (C). Null hypothesis eleven failed to be rejected. There was no significant interaction between timbre, re- tention interval, and task. 84 H012: No significant interaction will be found between timbre (A), retention interval (B) and same sequence items vs. different sequence items (D). The triple interaction (ABD) between timbre reten- tion interval, and task was found to be significant F = 2.19 at the .05 level. Null hypothesis twelve was rejected. H013: No significant interaction (ACD) will be found be- tween timbre (A), task (C), and same sequence item vs. different sequence item (D). There were no significant interactions between timbre, task, and item. Null hypothesis thirteen failed to be rejected. H014: No significant interaction (BCD) will be found be- tween retention interval (B), task (C) and same sequence items vs. different sequence items (D). A significant interaction effect F = 4.97 (p < .05) was found between retention interval, task and items. Null hypothesis fourteen was rejected. H015: No significant (ABCD) will be found between timbre (A), retention interval (B), task (C) and same sequence items vs. different sequence items (D). There was no significant quadruple interaction be- tween timbre, retention interval, task, and items. Null hypothesis fifteen failed to be rejected. 85 Conclusions An examination of the findings of this study suggest the following conclusions: 1. The sound quality of a pitch sequence tended to affect its perception and its retention in short-term memory. 2. The sawtooth sound quality had the most dis- ruptive effect on the short-term retention of pitch sequence. The triangle sound quality appeared to be less consistent in terms of effect on short-term retention. 3. The accuracy of recognition response in short- term retention was influenced by the length of the re- tention interval. Response mean scores for the EEEE group and the non-task group were considerably lower at the 15 second retention interval. 4. The interaction effect between timbre and re- tention interval suggested that the short-term retention of pitch is dependent upon the kind of timbre and the length of the retention interval, at least in this study. 5. The performance of a non-musical task did not significantly affect the short-term retention of a pitch sequence. 6. The intention between retention interval and item-type suggested that the recognition of test items as the same or different was dependent on the length of the retention interval. Test items in which both pitch 86 sequences were the same, were recognized with greater accuracy during the 5 second retention interval, than test items, in which the second sequence was different. 7. The triple interaction between timbre, reten- tion interval, and item type (same vs. different) suggested that the short-term retention of pitch sequences is de- pendent upon timbre, length of the retention interval, and whether the second sequence is the same or different. 8. The true influence of the main effects of timbre and retention interval should be qualified because of the number of interactive effects. Discussion Taylor12 has proposed a list of perceptual determin- ants in melodic recognition. Included in the list are: l. Gestalt of melody 2. Rhythm 3. Tempo 4. Dynamics 5. Timbre 6. Transposition 7. Melodic contour 8. Interval (size, direction, etc.) He suggests that it is possible that an interval, for instance a p5, may be perceived and remembered differently 12James Taylor, "Perception of Melodic Intervals Within Melodic Context," (Doctoral dissertation, University of Washington, 1972), p. 6. 87 when played on a trumpet, as compared to a piano.13 The results of this study would seem to give some support to that hypothesis. This may be particularly true in the case of non-music majors, who, for the most part, have ruyt been exposed to the specialized training of the musician. Furthermore, the findings of this study appear 1:) agree with those of Williamsl4, who noted that children tended to musjudge intervals when presented by rich timbre (as compared to presentation by a bland timbre. With ref- erence to this investigation the sawtooth wave (rich timbre) had a detrimental influence on perception and re- tention. The findings in respect to retention interval were somewhat contrary to expectation. The investigator had theorized a greater retention loss with the 30 second re- tention interval rather than the 15 second retention in- terval. These results, however, may be explained by the possibility that by the 15 second time interval the limits of short-term retention; and the improved perfor- mance during the 30 second interval was due to the trans- ferral of pitch information into long term memory. 131hid., p. 11. 14David B. Williams, "An Interim Report of a Programmatic Series of Music Inquiry Designed to Investigate Melodic Pattern Identification Ability in Children," Council for Research in Music Education," 1976, pp. 78-83. 88 though this explanation is somewhat inconsistent with other findings, it is plausible. Furthermore, the research literature on short-term retention seems to be inconsistent in regards to the length of the short-term memory: report ing time-spans from 15 seconds to 60 seconds. Another puzzling finding of this investigation was the lack of difference in performance between the task group and the non-task group. One explanation for the lack of difference is that the task of counting backwards (while writing) was not difficult enough to interfere with the processing of the pitch information. An alternative possibility is that two different kinds of information processing systems were operating simultaneously - one processing pitch information, the other mathematical - verbal information. The counting procedures did not interfere with the pitch rehersal procedures. The research of Deutsch15 would seem to support such a thesis. Her theory that pitch information retains its character- istics in short-term memory without being recoded seems to be operable in this instance. 15Diana Deutsch and J. Anthony Deutsch, Short Term Memory (New York: Academic Press, 1975), pp. 141-146. 89 Implications of the Study One of the implications of this study is that some consideration should be given to the kind of timbre used in a music learning situation. The richness or blandness of a particular timbre might affect the perception of the 16 found a difference between mean sound stimuli. Petzold scores of items that were presented on the violin as compared to those presented on the flute. He also noted mean differences between test items presented on the piano and those presented by voice. Both of these mean comparisons were between relatively rich timbres (violin and voice) and relatively bland timbres (flute and piano). He found no difference between the mean scores of the two rich timbres, or between the means of the two bland timbres. The greatest number of correct responses occurred when the violin was the presenting timbre. The findings of Petzold's study and those of this study imply that certain timbres have facilatory effects in memory. What this means to music teaching is that the music educator in both learning situations or using echo type exercises might improve student response by using a particular timbre. The findings of this study would seem to have a direct implications for music testing. In light of the 16Petzold, op. cit., p. 144. 90 results, the assumption that timbre does not affect memory, does not seem to hold true for short-term memory. Since most published music tests are tests of short-term memory, it might be well to re-evaluate such tests for possible timbre effects. A test battery such as Gordon's 17 Music aptitude Profile might be highly susceptible to timbre effect. Authors of new tests should experiment with various timbre before their selection of a sound source for test items. On a much broader level the findings of this study point to a theory of music memory that is more inclusive. Earlier concepts of music memory are based on a tonal system of music, i.e., tonality, and do not account for atonality. During the first half of the twentieth century, when most of the research in music memory took place, those involved did not understand or recognize atonality. Atonal music was considered by many of these researchers to be nonsense music and invalid as a means of expression. Therefore, music organized on atonal principles was not included in a theory of music memory. This has been found to be an erroneous conclusion, since atonality is a dominant principle in organizing pitch today. l7Edwin Gordon, Musical Apritude Profile (Boston: Houghton Miflin Company, 1965). 91 The theory of music memory based primarily on tonal- ity, however, has had enormous consequences for music learn- ing. Music teaching on all academic levels, the elementary grade through senior college, has been relatively success- ful in getting students to perceive, respond to, and in many instances understand music embracing a tonal system. This has not been true for twentieth century music, especially that utilizing atonal organization. One of the problems is the difference in the prin- ciples of organization between tonal and atonal music. Tonal music is based on repetition; melodic, harmonic, rhythmic and structural repetition. In many instances this repetition is quite obvious. A hierarchy of tonal re- lations is another feature of tonal music. This hierarchy leads to both expectation and anticipation in the percep- tion of tonal music. Essentially, what this means in terms of music memory is that rehearsal mechanisms are inherent in the music itself, providing a maximum retention of such music. Atonal music, on the other hand, involves less repet- tion, or a much more subtle kind of repetition which pro- vides a minimum amount of rehearsal in the music itself. Therefore it is less easily perceived or retained. Since it is now acknowledged that music organized according to principles other than those of tonality represents a valid form of expression, it would seem prudent to re-examine 92 a music memory theory based primarily on tonal hierarchies. For, in essence, such a theory has had the consequence on music learning of maximizing the amount of information available in tonal music; and minimizing the information available in atonal music. Teachers of music theory, and some music educators have come to realize that any linear or vertical combina- tion of two or more tones has information potential. This realization, however, has not yet filtered into music education for the general student. Given the in- creasing recognition that listening is the principal music activity or means of involvemnt for the non-musican, perhaps a theory of music learning that is evolved from a music information theory of memory would achieve greater success with the teaching of all kinds of music. Such a theory of music memory would be inclusive of all composi- tional practices: and the information available would be relative to the particular compositional style. For instance, a composition by Webern could convey the same amount of information as a composition by Mozart. This is not implicit in current memory theory. Implications for Further Research The findings of this investigation suggests replica- tions of this study with subjects at lower academic levels (elementary, junior high school or senior high school 93 students). In addition the following research is re- commended. l. A more direct investigation of timbre effect in which descriptive responses are possible. For instance, same - higher - lower. 2. A study similar to this one except that three groups are utilized. The third group would have conditions in which timbre and retention interval are held constant. 3. A replication of this study using music majors as well as non-music majors. 4. A replication of this study using a different task during the retention interval. For instance, writing a musical phrase. 5. A replication using a larger sample. BIBLIOGRAPHY BIBLIOGRAPHY Books Campbell, Donald T. and Stanley, Julian C., Experimental and Quasi-Experimental Designs for Research, Chicago: Rand McNally and Co., 1969. Cermak, Laird A., Human Memory: Research and Theory, New York: The Ronald Press Company, 1972. Deutsch, Diana and Deutsch, J.H., editors, Short-Term Memory, New York: Academic Press, 1975. Erickson, Robert, Sound Structures in Music, Berkeley: University of CaIifornia Press, 1975. Estes, W.K., editor, Handbook of Learning and Cognitive Processes, New York: John W. Wiley and Sons, 1975. Gordon, Edwin, The Psychology of Music Teacher, Englewood Cliffs: Prentice-Hall, 1971. Horton, David L. and Turnage, Thomas W., Human Learning, Englewood Cliffs: Prentice-Hall Inc., 1976. Klatsky, Roberta, Human Memory: Structures and Processes, San Francisco: W.H. Freeman and Co., 1975. James, William, The Principles of Psychology, Volume I, New York: Henry Holt and'Company, 1880. Mursell, James, Psychology of Music, New York: W.W. Norton and Company, 1937. ‘Norman, Donald A., Memory and Attention, New York: John W. Wiley and Sons, Inc., 1969. Reed, Owen and Sidnell, Robert, The Materials pf Music Composition, Reading: Addi§on~Wesiey Publishers, in press, 1977. Seashore, Carl E., The Psychology of Musical Talent, Boston: Silver Burdett and Company, 1919} 94 95 Shuter, Rosamunde, The Psychology of Music Ability. Winer, B.J., Statistical Principles in Experimental Desrgn, New York: McGraw-Hill,21962. Periodicals Agnew, Marie, "Comparison of Auditory Images of Children, Psychologists and Musicians," Psychological Review, Vol. 31, 1922, pp. 268-278. Agnew, Marie, "The Auditory Images of Great Composers," Psychological Monographs, Vol. 31, 1922, pp. 274- 282. Anderson, D.A., "The Duration of Tones, the Time, the Interval, the Direction of Sound, Darkness, Quiet, and the Order of Stimuli in Pitch Discrimination," Psychological Monographs, Vol. 16, 1914, p. 3. Angell, E.,-"Discrimination of Clangs for Different Intervals of Time," Journal of Psychology. Atkinson, R.C. and Shiffrin, "The Control of Short-Term Memory," Scientific American, Vol. 225, No. 2, 1971, PP. 82-90. Bachem, A., "Genesis of Absolute Pitch," Journal of the Acoustic Society of America, Vol. 2, 1940, pp. 434-439. Bachem, A., "Time Factors in Relative and Absolute Pitch Determination," Journal of theAgoustic Society of America, Vol. 26, 1954, pp. 751-753. Baird, J.W., "Memory for Absolute Pitch," Studies in Psychology: Titchener Commemorative Volume: Worcester, Mass.: Louis N. Wilson, 1917. Bergan, J.R., "Factors Effecting Pitch Discrimination," Bulletin of the Council of Research in Music Educa- tiOn, No._8, 1966, pp. 15-2I} Bergan, J.R., "The Relationship Among Pitch Identification, Imagery for Musical Sounds and Musical Memory," Journal of Research in Music Education, Vol. 15, 1967. pp. 99-109. Boggs, L.P., "Studies in Absolute Pitch," Amerisan Journal of Psychology, Vol. 18, 1970. pp. 194-203. 96 Broadbent, D.E., "Flow of Information within the Organism," Journal of Verbal Learninq_and Verbal Behavior, Vol. 2, 1963, pp. 34-39. Brown, J., "Some Tests of the Delay of Immediate Memory," Quarterly Journal of Experimental Psychology, Vol. 10, 1958. Bugg, E.G. and Herpel, L., "The Significance of Tonal Memory for Musicality," Journal of General Psychology, Vol. 35, 1946, pp. 3:15. Bugg, E.G. and Herpel, L., "An Analysis of Conditions In- fluencing Consonance Judgements," Journal of Experimental Psychology, Vol. 24, 1939, pp. 54-72. Burroughs, G.E.R., "Factors Involved in Learning a Simple Musical Phrase," British Journal of Educational Psychology, Vol. 13, 1962, pp. 83-98. Buttram, J.B., "The Influence of Selected Factors on Interval Identification," Ph.D. Dissertation, Dissertation Abstracts, January 1968, 28: 2707A. Carlsen, James C., "Deve10ping Aural Perception of Music in Context," Journal of Research in Music Educa- tion, 1969. Carlsen, J.C., Divenyi, P.I. and Taylor, J.A., "A Pre- liminary Study of Perceptual Expectancy in Melodic Configurations," Council for Research in Music Education Bulletin, VoIT’22, 1970, pp. 4-12. Cole, R.A., "Different Memory Functions for Consonants and Vowels," Cognitive Psychology: Vol. 4, 1973, pp. 39-54. Cohen, J.F., "Information Theory and Music," Behavioral Science, Vol. 7, 1962, pp. 137-163. Crowder, R.G., "The Sound of Vowels and Consonants in Immediate Memory," Journal of Verbal Learning and Verbal Behavior, Vol.T10, 197i, pp. 587-59 . Crowder, R.C. and Morton, J., "Precategorical Acoustic Storage," Perception and Psychophysics, Vol. 5, 1969. Deutsch, D. and Deutsch, J.A., editors, Short-Term Memory, New York: Academic Press, 1975. 97 Deutsch, D. and Deutsch, J.A., "Music Recognition," Psychological Review, Vol. 76, 1969, pp. 300-307. Deutsch, D. and Deutsch, J.A., "The Deterioration of Pitch Information in Memory," Unpublished Ph.D. Dissertation, University of California at San Diego, 1970. Deutsch, D. and Deutsch, J.A., "Tones and Numbers: Specificity of Interference in Short-Term Memory," Science, Vol. 168, 1970, pp. 1604-1605. Deutsch, D. and Deutsch, J.A., "The Effect of Repetition of Standard and Comparison Tones on Recognition Memory for Pitch," Journal of Experimental Psychology, Vol. 93, 1972, pp. 166-162. Deutsch, D. and Deutsch, J.A., "Octave Generalization and Tune Recognition," Perception and Psychophysics, Vol. 11, 1972, pp. 411-412. Deutsch, D. and Deutsch, J.A., "Interference in Memory Be- tween Tunes Adjacent in the Musical Scale," Journal of Experimental Psychology, Vol. 100, 1973, pp.’228-231. Deutsch, D. And Deutsch, J.A., "Octave Generalization of Specific Interference Effects in Memory for Tonal Pitch," Perception and Psychophysics, Vol. 13, 1973, pp. 271-275. Deutsch, D. and Deutsch, J.A., "Generality of Interference by Tonal Stimuli in Recognition Memory for Pitch," Quarterly Journal of Experimental Psychology, Vol. 26, 1974, pp. 229-234. Deutsch, Diana, "Facilitation by Repetition in Recognition Memory for Tonal Pitch," Memory and Cognition, Vol. 3, 1975. Dowling, W.J. and Fugitan, D.J., "Contour, Interval, and Pitch Recognition in Memory for Pitch," Journal of the Acoustical Society of America, Vol. 49, No. 2, Pt. 2, 1971, pp. 524-531. Drake, R.M., "Four New Tests of Musical Talent," Journal of Applied Psychology: VOl. 17, 1933, pp. 136-147. Elliot, L.L., "Pitch Memory for Short Tones," Perception and Psychgphysics, Vol. 8, 1970, pp. 379-384. 98 Elliot, L. L., "Auditory Memory for One and Two Tones," The Journal of the Acoustic Society of America. Gough, Evelyn, "The Effects of Practice on Judgments of Absolute Pitch," Archives of Psychology, V01. 7, 1922, pp. 642-653. Greer, Douglas R., "The Effect of Timbre on Brass-Wind Intonation," Research in the Psychology of Music, Edwin Gordon Ed., Iowa City: University of Iowa Press, Vol. 6, 1970, pp. 65-94. Haack, Paul, "The Influence of Loudness on the Discrimina- tion of Musical Sound Factors," Journal of Researgh in Music Education, Vol. 73, No. 1, 1975, pp. 67-77. Harris, Donald J., "Discrimination of Pitch: Suggestions Towards Method and Procedure," American Journal of Psychology, Vol. 61, 1948, pp. 309-322. Harris, Donald J., "The Decline of Pitch Discrimination with Time," Journal of Experimental Psychology, Vol. 43, 1952, pp. 96-99. Heinlein, C.P., "A Brief Discussion of the Nature and Func- tion of Melodic Configuration in Tonal Memory with Critical Reference to the Seashore Tonal Memory Test," Journal of Genetic Psychology; Vol. 35, 1928, pp. 45-61. Kimura, D., "Left-Right Differences in the Perception of Melodies," Quarterly Journal of Experimental Psychology, Vol. 16, 1964, pp. 355-358. Koester, Theodore, "The Time Error and Sensitivity in Pitch and Loudness Discrimination as a Function of Time Interval and Stimulus Level," Archives of Psychology, Vol. 41, 1945, p. 297. Lowery, H., "Musical Memory," British Journal of Psychology, Vol. 19, 1929, pp. 397-404. Mainwaning, James, "Experiments on the Cognitive Processes in Musical Ability," British Journal of Psychology, Vol. 22, 1931, pp. 180-203. Mainwaning, James, "Kinesthetic Factors in the Recall of Musical Experience," British Journal of Psychology, Vol. 23, 1933, pp. 284- 307. 99 Massaro, D.W., "Consolidation and Interference in the Perceptual Memory System," Perception and Psycho- Massaro, D.W., "Retroactive Interference in Short Term Memory for Pitch," Journal of Experimental Psychology: Vol. 83, 1970, pp. 32-39. Massaro, D.W., "Perceptual Auditory Images," ggurnal of Experimental Psychology, Vol. 85, No. 3, 1970, pp . 411-417. Massaro, D.W., ”Effect of Masking Tone Duration on Per- ceptual Auditory Images," Jourpsl of Experimgntal Psychology, Vol. 87, No. 1, 1971, pp. 146-148. Massaro, D.W., ”Stimulus Information vs. Processing Time in Auditory Pattern Recognition," Percgptign and Psychophysics, Vol. 12, No. 1A, 1972, pp. 50-56. Miller, G.A., "The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information," Psychological Review, Vol. 2, 1966, pp. 81-97. Mohs, R.C., "The Effect of Short-Term Contents on Long- Term Memory Search,” Memory and Cognition, Vol. 4, 1973. Moss, S.M., Myers, J.L. and Filmorie, T., "Short Recogni- tion Memory of Tones," Perce tion and Ps cho- physics, Vol. 7, No. 6, I970, pp. 369-373. Mull, Helen K., "The Acquisition of Absolute Pitch," American Journal of Psychology, Vol. 36, 1925, pp. 469-493. Neu, D.M., "A Critical Review of the Literature on Absolute Pitch," Psychological Bulletin, Vol. 44, 1947, Ortmann, Otto, "Some Tonal Determinants of Melodic Memory," Journal of Educational Psychology: V01. 24, 1933, pp. 620-625. Peterson, Lloyd R. and E. Peterson, "Short-Term Retention of Individual Verbal Items," Journal of Experi- mental Psychologx, V01. 58, 1959: Pflederer, Marilyn, "How Children Conceptually Organize Sounds," Council of Research in Music Education, Bulletin, V61. 7) 1967, pp. 1-12. 100 Posner, Michael and Andrew Konick, "On the Role of Inter- ference in Short-Term Memory," Journal of Experi- mental Psychology, Vol 72, 1966. Rubin-Rabson, Grace, "Studies in the Psychology of Memoriz- ing Music,” Journal of Educational Psychology: 1937- 1941. Vernon, P.E., "Auditory Perception," "The Gestalt Approach," British Journal of Psychology, Vol. 60, 1934, pp. 123-139. Vernon, P.E., "Auditory Perception, Evolutionary Approach," British Journal of Psychology, Vol. 61, 1935, pp. 265-281. Whipple, Guy, "An Analytical Study of Memory Image and the Process of Judgment in the Discrimination of Clangs and Tones," The American Journal of Psychology, 1902, pp. 219-268. Whitley, Paul, "The Influence of Music on Memory," Journal of General Psychology, Vol. 10, 1934, pp. 137-151. Wickelgren, Wayne A., "Consolidation and Retroactive Inter- ference in Short-Term Recognition for Pitch," Journal of Experimental Psychology. V01. 72, 1966, pp. 250-259. Wickelgren, Wayne A., Associative Strength Theory for Recognition of Pitch," Journal of Mathematical Psychologx, Vol. 6, 196972pp. 13-61. Williams, David B., "Short-Term Retention of Pitch Sequence," Journal of Research in Music Education, Vol. 23, No. 1, 1975, pp. 53-65. Williams, David B., "An Interim Report of a Programmatic Series of Music Inquiry Designed to Investigate Melodic Pattern Identification Ability in Children," Council for Research in Music Education, 1976. Wolfe, H.K., "Untersuchungen fiber das Tongedachtniss," Philosophnytudies, V01. 3, 1886. Zenatti, Arlette, "Melodic Memory Tests: A Comparison of Normal Children and Mental Defectives," Journal of Research in Music Education, Vol. 23, No. 1, 1‘9'7'5 pp. Tl-SZ. 101 Reports Petzold, Robert G., Auditory Perosption of Musical Sounds 9y Children in the First Six Grades. Cooperative Researcfifiproject, No. 1051, Madison: University of Wisconsin Press, 1966. Dissertations Long, Peggy Ann, "Pitch Recognition in Short Melodies," Tallahassee: Unpublished Ph.D. dissertation, The Florida State University, 1975. In Disserta- tion Abstracts, 1976, No. 4083. Taylor, James, "Perception of Melodic Intervals within Melodic Context," Unpublished Ph.D. dissertation, University of Washington, 1972. Williams, David Brian, "Short-Term Retention of Pitch Sequence," Unpublished Ph.D. dissertation, Univer- sity of Washington, 1973. APPENDICES APPENDIX A THE PITCH-TIMBRE MEMORY TEST Appendix A The Pitch-Timbre Memory Test A B A B 102 The Pitch-Timbre Memory Test ,4. 8 4 F. 103 APPENDIX B THE TWENTY FOUR TEST QUESTIONS 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. Appendix B Description of Test Questions Sine Wave Sawtooth Wave Pulse Wave Triangle Wave Sine Wave Sawtooth Wave Pulse Wave Triangle Wave Sine Wave Sawtooth Wave Pulse Wave Triangle Wave Sine Wave Sawtooth Wave Pulse Wave Triangle Wave Sine Wave Sawtooth Wave Pulse Wave Triangle Wave Sine Wave 15 15 15 15 30 30 30 30 5 5 15 15 15 15 30 seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds seconds 104 same same same same 8 ame S ame S ame same same same same 8 ame different different different different different different different different different sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence sequence 105 22. Pulse Wave - 30 seconds - different sequence 23. Pulse Wave - 30 seconds - different sequence 24. Triangle Wave - 30 seconds - different sequence APPENDIX C RAW DATA FOR THE 24 QUESTIONS macho gusslsoz ac macho xmma c .106 00. OO. Om. mh. OO.H OO.H OO. Om. OO.H OO. OO.H om. OO.H Om. OO.H Om. 0m. 00. OO. Om. OO.H OO.H OO.H mN. ON cc OO.H OO.H OO.H mm. 00. OO. OO.H OO. OO.H OO.H 00. CV. 00. Om. OO.H mN. Om. 00. 00. mb. Om. OO.H 00. m5. 5N cc OO.H OO. Om. OO.H OO.H OO.H Om. Om. OO. OO.d OO. om. OO. OO.H OO.H 0m. 00. OO. 00. 0m. 0m. OO.H OO.H mh. ON cc 00. OO. OO.H Om. OO.H OO. Om. OO.H OO. 00. OO. 0?. OO.H OO.H 00. mm. OO.H OO. 00. OO.H OO.H OO.H OO.H OO.H mN cc OO.H OO.H Om. OO.H OO. 00. 00. mn. 00. OO. 00. OV. CO. CO. OO. 00. OO.H CO. CO. OO.H OO.H OO.H OO.H OO.H vN cc OO. OO.H OO.H OO.H OO.H 00. cm. mb. OO.H OO.H OO. 00. OO. 00. OO. mm. 00. OO. 00. mm. OO.H OO.H OO.H OO.H MN cc OO.H OO.H OO.H Om. OO. OO.H Om. OO.H OO. OO.H OO.H 00. 00. OO. 00. mN. Om. 00. OO. Om. OO.H OO.H OO.H OO.H NN cc OO.H OO.H OO.H mN. OO.H OO.H OO. mN. OO.H OO.H OO.H 0v. OO. 00. OO. 00. 0m. 00. 00. mm. Om. OO.H OO.H Om. HN cc OO.H OO. Om. mh. OO. OO.H Om. mN. OO. 00. OO.H ON. OO.H CO. CO. mN. Om. OO. OO.H mN. OO.H OO.H OO.H mh. ON cc 00. OO. OO.H mh. OO.H 00. OO. Om. OO. 00. OO.H 0v. OO.H CO. CO. mh. OO.H OO.d OO. Om. Om. OO.H OO.H mh. mH cc OO.H 00. 00. mm. OO.H OO. 00. OO.H OO.H OO.H OO. 0?. 00. cm. OO.H mN. OO. 00. OO. Om. Om. OO.H OO.H OO.H OH cc OO. OO.H OO.H mm. OO.H OO.H OO. mN. 00. OO. 00. OW OO.H OO.H OO.H Om. Om. OO.H OO. OO.H OO.H OO.H OO.H OO.H ha cc OO. OO.H OO.H Om. OO.H OO. 00. mN. OO.H OO. 00. om. OO.H Om. OO.H mN. Om. OO.H OO.H mN. OO.H OO.H OO. Om. 0H cc OO. OO.H Om. Om. OO. OO.H Om. OO.H OO. OO.H OO.H 00. OO. OO.H 00. m5. OO.H OO.H OO.H Om. OO. OO.H OO. Om. mH cc OO. OO.H Om. mN. OO.H OO.H OO.H mN. OO.H OO.H 00. ON. 00. Om. OO.H mN. Om. 00. OO. OO.H OO.H OO. OO.H mh. VH c 00. OO.H Om. mm. OO.H OO. Om. Om. OO.H OO. 00. 0v. OO. Om. OO.H mN. Om. 00. 00. mm. OO.H OO.H OO.H mh. MH c 00. OO.H Om. mm. OO.H OO.H Om. OO.H OO.H OO. OO.H 0v. OO. 00. OO.H OO.H OO.H OO. 00. Om. OO.H OO.H OO.H Om. NH c OO.H OO.H OO. OO.H OO.H OO.H 00. mp. OO.H 00. OO. 00. OO. OO.H OO.H mN. OO.H OO.H 00. mm. OO.H OO.H OO.H Om. HH c 00. OO. Om. mm. OO.H OO. 00. Om. OO. OO.H OO.H 0v. OO.H Om. 00. mm. OO.H 00. 00. mm. OO.H OO.H OO.H mh. OH c 00. OO.H Om. mh. OO.H 00. cm. mN. OO.H OO.H 00. CV. OO.H CO. CO. OO.H Om. OO.H 00. mm. OO.H OO.H OO.H mh. m c 00. OO. Om. mm. 00. OO.H 00. mm. 00. 00. CO. CO. CO. OO.H OO.H Om. OO.H OO. 00. mm. OO.H OO.H OO.H mm. m c 00. OO. Om. Om. OO.H OO.H Om. OO.H OO.H OO. OO.H 0v. OO.H OO.H OO.H mm. OO.H OO.H OO. Om. Om. OO.H OO.H OO.H h c 00. OO.H 00. mm. OO.H OO.H Om. Om. OO. 00. OO.H 0v. OO.H om. OO.H mm. Om. OO.H OO. mN. OO. 00. OO. OO.H m c 00. OO. Om. mN. OO. OO.H 00. mm. 00. 00. 00. ON. OO.H OO.H OO.H mb. 0m. 00. OO. Om. OO.H OO.H OO.H OO.H m c 00. OO.H 00. cm. OO.H 00. 00. mm. OO.H OO. 00. ON. OO.H OO.H OO.H mb. OO.H OO. OO.H m5. OO.H OO.H OO.H mh. v c 00. OO.H Om. h. OO.H OO.H OO. mN. OO.H OO.H OO. 0?. OO.H OO.H OO.H mN. OO.H OO. 00. OO. 00. OO.H OO. OO.H m c 00. OO.H 00. cm. OO.H OO. Om. mN. OO.H 00. 00. OO. OO.H Om. OO.H mh. OO.H OO.H OO.H mh. 0m. 00. OO.H Om. N c OO.H OO.H Om. mm. OO.H OO.H Om. Om. OO.H OO.H 00. CV. OO.H Om. OO.H mN. 0m. 00. OO. Om. OO.H OO. OO.H mN. H c VN MN NN HN ON ma ma NH 0H mH QH MH NH HH OH 0 m h m m v m N H mcoflumwsm ucocsum mcowummso vm on» new name 3mm 0 xflosmmmé APPENDIX D TEST RETEST SCORES Test-Retest Scores Student Test 1 Retest 1 33 34 2 31 30 3 31 33 4 29 30 5 29 29 6 28 30 7 28 27 8 27 28 9 27 27 10 26 27 11 26 25 12 26 28 13 25 24 14 24 26 15 32 33 16 32 30 17 32 32 18 29 29 19 28 29 20 26 28 21 26 . 27 22 26 26 23 25 25 24 25 24 25 24 25 26 24 27 27 22 23 28 22 22 r = .92 107 lllllllrl wumwu» 111 93 030 12 I "I All H "I. "