INFLUENCE OF MEANIN‘GFULNESS AND FAMILIARIZATION TRAINING 0N BASIC SIGHT VOCABULARY LEARNING WITH FIRST-GRADERS Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY ERNEST LOYD ADAMS 1970 cf," THF 5 I This is to certify that the thesis entitled \ Influence of Meaningfulness and Familiarization Training on Basic Sight Vocabulary Learning with First-Graders presented by Ernes t L. Adams has been accepted towards fulfillment of the requirements for Ph.D. degree in Elementary Education Mainr hfn‘nccnr Date 7’20 " '70 0-169 \ alumna av. ¥ .. mm; & sous -—f In amnm me. “I!!!” myosns ABSTRACT INFLUENCE OF MEANINGFULNESS AND FAMILIARIZATION TRAINING ON BASIC SIGHT VOCABULARY LEARNING WITH FIRST—GRADERS By Ernest Loyd Adams Paired-associate learning research has examined various factors that influence the rate at which a verbal stimulus is associated with a verbal response. Two such factors are the meaningfulness of the materials to be learned and the learner's familiarity with the materials. Studies investigating the influence of meaningfulness on the rate of learning have consistently shown that highly meaningful material is learned more rapidly than less meaningful material. These studies were conducted with adult subjects learning paired-associate lists consisting of nonsense syllables. Similar subjects and lists were used to investigate the effect of familiarization training on the rate of acquisition. Experimenters familiarized their subjects Ernest Loyd Adams with the paired-associate units. Response familiarization was shown to influence the rate of learning significantly. The purpose of this study was to determine whether meaningfulness and familiarization training have a similar influence on word recognition learning with first-graders. Thirty words were selected from the Dolch basic sight vocabulary list. The meaningfulness value of each word was determined by Noble's production method for scaling meaningfulness values. Each word was scaled from high to low meaningfulness on the basis of the number of different associ- ations that the word elicited from a group of subjects. Two groups of 20 subjects each participated in the scaling experi- ment. A Pearson product-moment correlation coefficient was computed to determine the agreement between Group A and Group B in the ranking of the Dolch words from high to low meaningfulness. A correlation of .93 indicated agreement between the groups. Four high meaningfulness words and four low meaning- fulness words were selected for learning. Four experimental groups were established. The subjects in the control group learned the experimental words without familiarization train- ing. The second group received response familiarization prior to learning, and group three received stimulus familiarization prior to learning the words. The fourth Ernest Loyd Adams group of subjects received stimulus and response familiari~ zation before learning the words. The data were analyzed using a 2X2X2 factorial design with repeated measures on the meaningfulness factor. The results obtained from analysis indicated that high meaning- fulness words were learned more rapidly than low meaningful- ness words (significant at p<.OOl). Further analysis showed that response familiarization facilitated the learning of low meaningfulness words. The group that received response familiarization and the group that received stimulus and response familiarization learned the low meaningfulness words more quickly than the control group (significant at p<.Ol). Two factors, meaningfulness and response familiariza- tion, influenced the first-graders' rate of learning the basic sight words. The third factor, stimulus familiarization, did not influence the subjects' rate of learning to recognize the experimental words. These findings indicate that teachers may be more successful in teaching pupils the basic sight words if they will provide response familiarization exercises on the words to be learned prior to learning. INFLUENCE OF MEANINGFULNESS AND FAMILIARIZATION TRAINING ON BASIC SIGHT VOCABULARY LEARNING WITH FIRST-GRADERS By Ernest Loyd Adams A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Elementary and Special Education 1970 /-— 2 “7-- ‘7/ To George Sherman, my friend and professor ii ACKNOWLEDGMENTS A thesis represents the concerted efforts of many peOple and although the researcher accepts full responsibility for any weaknesses which remain, he must share the credit for all strengths. It is not possible to mention the name of each person who contributed to this study, but this does not reduce the writer's appreciation of their efforts. The writer is particularly grateful to the chairman of the doctoral committee, Dr. Byron H. van Roekel, for his constant encouragement, sound counsel, and his willingness to invest hours of his time throughout this study and the entire graduate program. This study would not have been possible without the scholarly assistance of Dr. Clessen J. Martin. The writer extends his most sincere thanks for Dr. Martin's guidance and patience throughout this research effort. Grateful adknowledgment is extended to committee member Dr. Glen 0. Cooper for his support and encouragement. His example as an outstanding teacher will not be soon for- gotten. iii The writer also extends his appreciation to Dr. Ruth Brend for her invaluable suggestions during the final stages of writing. To the members of his family, the writer offers his most affectionate appreciation for their understanding and support throughout the duration of his graduate program. The greatest contributor to the writer's efforts in this accomplishment was LaDonna, his wife. For her patience and support as well as the many sacrifices she made, the writer is eternally grateful. iv TABLE OF CONTENTS Page DEDICATION . . . . . . . . . . . . . . . . . . . . . ii ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . iii LIST OF TABLES . . . . . . . . . . . . . . . . . . . vii LIST OF FIGURES . . . . . . . . . . . . . . . . . . ix Chapter I. INTRODUCTION . . . . . . . . . . . . . . . . 1 II. REVIEW OF LITERATURE . . . . . . . . . . . . 9 Methods of Scaling Meaningfulness . . . . . 9 Familiarization Techniques . . . . . . 18 Theoretical Role of Familiarization . . . . 20 Research Hypotheses . . . . . . . . . . . . 26 III. METHODOLOGY . . . . . . . . . . . . . . . . . 28 Scaling of Materials . . . . . . . . . . . 28 materials 0 O O O O O O O O O O O O O O O O 29 Subjects . . . . . . . . . . . . . . . . . 32 Procedures . . . . . . . . . . . . . . . . 32 IV. RESULTS . . . . . . . . . . . . . . . . . . . 4O Scaling Meaningfulness Values . . . . . . . 40 Experimental Design . . . . . . . . . . . . 42 .Analysis of Data . . . . . . . . . . . . . 42 Test Trial Data . . . . . . . 44 Analysis of Complex Retention Data . . . . 46 Analysis of Simple Retention Data . . . . . 48 Chapter Page V. DISCUSSION . . . . . . . . . . . . . . . . . 53 Introduction . . . . . . . . . . . . . . . 53 Word Recognition . . . . . . . . . . . . . 53 Summary and Conclusion . . . . . . . . . . 6l BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . 64 APPENDICES A. DIFFERENT ASSOCIATIONS . . . . . . . . . . . 70 B. LEARNING TRIAL DATA . . . . . . . . . . . . . 76 C. TEST TRIAL DATA . . . . . . . . . . . . . . . 78 D. RETENTION DATA . . . . . . . . . . . . . . . 80 E. RESPONSE FAMILIARIZATION PILOT STUDY DATA . . 82 F. INSTRUCTIONS GIVEN TO SUBJECTS . . . . . . . 83 vi Table l. l. 00 4: u: t» L» l 2 I—‘Ul-P-‘UJ LIST OF TABLES Origin of Dolch List . . . . . . . . . . Percentage of Dolch Words Found in Elementary Textbooks . . . . . . . Stimulus Response Learning Tasks . . . . Pilot Study Data . . . . . . . . . . . . Paired-Associate and Word Recognition Stimulus Learning . . . . . . . . . . . Words Selected for Scaling . . Basic Sight Words Selected for Familiarization Training and Learning Trials . . . . . . . . Response Familiarization Lists Stimulus Familiarization Identification List. Learning Trial Lists . . . . . . . . . . . . Number of Different Associations for Each Word . . . . . . . . . . . . Summary Table for Analysis of Variance on Test Trial Data . . . . . . . . . . Means of Test Trial Data . . Summary Table for Analysis of Variance on Complex Retention Data . . . . Means of Complex Retention Data . . . vii Page 18 23 3O 32 33 36 37 41 45 45 47 49 Table 4.6 Summary Table for Analysis of Variance on Simple Retention Data . . . . . . . . 4.7 Means of Simple Retention Data . . . viii Figure 2.1 4.1 4.2 4.3 4.4 LIST OF FIGURES Paired-Associate-Word Recognition Response Learning . . . . . . . . . . . . Model of Experimental Design Interaction Effects of Association Value by Treatment Conditions . . . . . . . . . Interaction Effects of Association Value by Words Recalled on Complex Retention Task . Interaction Effects of Association Value .by Words Recalled on Simple Retention Task ix Page 21 43 46 50 CHAPTER I INTRODUCTION. University libraries have been well stocked with research in the area of reading instruction and the early development of reading skills. The bulk of these efforts have investigated various traits of the learner and his level of reading achievement or the desirability of one method of instruction as Opposed to another. The present exploration has attempted to investigate some of the psychological processes which Operate as a first-grader learns to recognize written words. The materials most commonly used to teach reading to children are basal readers. These materials are designed to teach students the skill of word recognition. very often, the approach used to teach word recognition involves the learning of phoneme-grapheme correspondences. That is, the student attempts to learn the sound that is represented by each written letter and then blend these individual sounds together as they are presented graphemically to form words. 2 In conjunction with this instruction, pupils are also expected to begin recognizing at sight a group of words lthat constitute a high percentage of all the words used in materials written in English. Dolch identified these words in 1935 (Zintz, 1966). He compiled this list by taking words from three other previously organized vocabulary lists. The first list was that published by the Child Study Committee of the International Kindergarten Union (Zintz, 1966). That study investigated the vocabulary of pre-first- grade children from a list of words which was assembled through the detailed observations of kindergarten classrooms. From that list, Dolch chose only words which occurred one hundred or more times. The second list consisted of the first five hundred words of the Gates List, a list which has been used as a basis of many studies in reading vocabulary. This list has been recognized as containing words of first importance in children's reading. The third list, formed by Wheeler and Howell, con- sists of 453 words found in ten primers published between 1922 and 1929. When the words common to all three lists were identified, they totaled 193 words. Dolch then selected twenty-seven words that appeared on two of the three lists. 3 The origin of the List is summarized in Table 1.1. TABLE l.l.--Origin of the Dolch List Number of List Source Words I.K.U. Spoken Vocabulary (5 yr. olds) 510 Gates Reading Vocabulary (Primary Grades) 5OO Wheeler & Howell Ten Primers and Ten First Readers 453 Common to all lists 193 Common to two lists 27 220 The Dolch Basic Sight Words are, as the name implies, words that the reader should recognize at sight. Complete mastery of these words is essential. They make up over fifty per cent of the words found in materials written in English. Table 1.2 shows the results of a vocabulary study that determined the percentage of words found in various elementary school textbooks that are Dolch words. THF‘SIS 0-169 if L;F:i<.ai;{y :- A1;. Itigan State 4 . I D L? Umvermty .3: This is to certify that the thesis entitled k Influence of Meaningfulness and Familiarization Training \" on Basic Sight Vocabulary ‘ Learninngith First-Graders \ presented by Ernest 1... Adams has been accepted towards fulfillment of the requirements for Ph.D. degree in Elementary Education Major professor Date 2’20" 26) ABSTRACT INFLUENCE OF MEANINGFULNESS AND FAMILIARIZATION TRAINING ON BASIC SIGHT VOCABULARY LEARNING WITH FIRST-GRADERS BY Ernest Loyd Adams Paired—associate learning research has examined various factors that influence the rate at which a verbal stimulus is associated with a verbal response. Two such factors are the meaningfulness of the materials to be learned and the learner's familiarity with the materials. Studies investigating the influence of meaningfulness on the rate of learning have consistently shown that highly meaningful material is learned more rapidly than less meaningful material. These studies were conducted with adult subjects learning paired-associate lists consisting of nonsense syllables. Similar subjects and lists were used to investigate the effect of familiarization training on the rate of acquisition. Experimenters familiarized their subjects Ernest Loyd Adams with the paired-associate units. Response familiarization was shown to influence the rate of learning significantly. The purpose of this study was to determine whether meaningfulness and familiarization training have a similar influence on word recognition learning with first-graders. Thirty words were selected from the Dolch basic sight vocabulary list. The meaningfulness value of each word was determined by Noble's production method for scaling meaningfulness values. Each word was scaled from high to low meaningfulness on the basis of the number of different associ— ations that the word elicited from a group of subjects. Two groups of 20 subjects each participated in the scaling experi— ment. A Pearson product-moment correlation coefficient was computed to determine the agreement between Group A and Group B in the ranking of the Dolch words from high to low meaningfulness. A correlation of .93 indicated agreement between the groups. Four high meaningfulness words and four low meaning- fulness words were selected for learning. Four experimental groups were established. The subjects in the control group learned the experimental words without familiarization train— ing. The second group received response familiarization prior to learning, and group three received stimulus familiarization prior to learning the words. The fourth Ernest Loyd Adams group of subjects received stimulus and response familiari- zation before learning the words. The data were analyzed using a 2X2X2 factorial design with repeated measures on the meaningfulness factor. The results obtained from analysis indicated that high meaning— fulness words were learned more rapidly than low meaningful- ness words (significant at p<.OOl). Further analysis showed that response familiarization facilitated the learning of low meaningfulness words. The group that received response familiarization and the group that received stimulus and response familiarization learned the low meaningfulness words more quickly than the control group (significant at p<.Ol). Two factors, meaningfulness and response familiariza- tion, influenced the first-graders' rate of learning the basic sight words. The third factor, stimulus familiarization, did not influence the subjects' rate of learning to recognize the experimental words. These findings indicate that teachers may be more successful in teaching pupils the basic sight words if they will provide response familiarization exercises on the words to be learned prior to learning. INFLUENCE OF MEANINGFULNESS AND FAMILIARIZATION TRAINING ON BASIC SIGHT VOCABULARY LEARNING WITH FIRST-GRADERS By Ernest Loyd Adams A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Elementary and Special Education 1970 To George Sherman, my friend and professor ii ACKNOWLEDGMENTS A thesis represents the concerted efforts of many people and although the researcher accepts full responsibility for any weaknesses which remain, he must share the credit for all strengths. It is not possible to mention the name of each person who contributed to this study, but this does not reduce the writer's appreciation of their efforts. The writer is particularly grateful to the chairman of the doctoral committee, Dr. Byron H. Van Roekel, for his constant encouragement, sound counsel, and his willingness to invest hours of his time throughout this study and the entire graduate program. This study would not have been possible without the scholarly assistance of Dr. Clessen J. Martin. The writer extends his most sincere thanks for Dr. Martin's guidance and patience throughout this research effort. Grateful adknowledgment is extended to committee member Dr. Glen 0. COOper for his support and encouragement. His example as an outstanding teacher will not be soon for- gotten. iii The writer also extends his appreciation to Dr. Ruth Brend for her invaluable suggestions during the final stages of writing. To the members of his family, the writer offers his most affectionate appreciation for their understanding and support throughout the duration of his graduate program. The greatest contributor to the writer's efforts in this accomplishment was LaDonna, his wife. For her patience and support as well as the many sacrifices she made, the writer is eternally grateful. iv TABLE OF CONTENTS Page DEDICATION . . . . . . . . . . . . . . . . . . . . . ii ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . iii LIST OF TABLES . . . . . . . . . . . . . . . . . . . vii LIST OF FIGURES . . . . . . . . . . . . . . . . . . 1X Chapter I. INTRODUCTION . . . . . . . . . . . . . . . . 1 II. REVIEW OF LITERATURE . . . . . . . . . . . . 9 Methods of Scaling Meaningfulness . . . . . 9 Familiarization Techniques . . . . . . 18 Theoretical Role of Familiarization . . . . 20 Research Hypotheses . . . . . . . . . . . . 26 III. METHODOLOGY . . . . . . . . . . . . . . . . . 28 Scaling of Materials . . . . . . . . . . . 28 Materials . . . . . . . . . . . . . . . . . 29 Subjects . . . . . . . . . . . . . . . . . 32 Procedures . . . . . . . . . . . . . . . . 32 IV. RESULTS . . . . . . . . . . . . . . . . . . . 4O Scaling Meaningfulness Values . . . . . . . 40 Experimental Design . . . . . . . . . . . . 42 _Analysis of Data . . . . . . . . . . . . . 42 Test Trial Data . . . . . . . 44 Analysis of Complex Retention Data . . . . 46 Analysis of Simple Retention Data . . . . . 48 Chapter V. DISCUSSION . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . Word Recognition . . . . . . . . . . . . . Summary and Conclusion BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . APPENDICES A. DIFFERENT ASSOCIATIONS B. LEARNING TRIAL DATA . . . . . . C. TEST TRIAL DATA . . . . . . . . D. RETENTION DATA . . . . . . . . . . . . . . E. RESPONSE FAMILIARIZATION PILOT STUDY DATA . F. INSTRUCTIONS GIVEN TO SUBJECTS vi Page 53 53 53 61 64 7O 78 8O 82 83 Table 1.1 1.2 1.3 2.1 2.2 3.1 3.2 3.3 3.4 3-5 4.1 4.3 4.4 4.5 LIST OF TABLES Origin of Dolch List . . . . . Percentage of Dolch Words Found in Elementary Textbooks . . . . . . . . . . . . Stimulus Response Learning Tasks Pilot Study Data . . . . . . . . . . . . Paired-Associate and Word Rec0gnition Stimulus Learning . . . . . . . . . . Words Selected for Scaling . . . . . . . . . Basic Sight Words Selected for Familiarization Training and Learning Trials . . . . . . . . Response Familiarization Lists Stimulus Familiarization Identification List. Learning Trial Lists . . . . . . . . . . . . Number of Different Associations for EaCh word I O O O O O O O O O O O O O O O 0 0 Summary Table for Analysis of variance on Test Trial Data . . . . . . . . . Means of Test Trial Data . . Summary Table for Analysis of Variance on Complex Retention Data . . . . Means of Complex Retention Data . vii Page 18 23 3O 32 33 3.6 37 41 45 45 47 49 Table 4.6 Summary Table for Analysis of Variance on Simple Retention Data . . . . . . . 4.7 Means of Simple Retention Data viii Page 49 51 Figure 2.1 4.1 4.2 4.3 4.4 LIST OF FIGURES Paired-Associate-Word Recognition Response Learning . . . . . . . . . . . . Model of Experimental Design Interaction Effects of Association Value by Treatment Conditions . . . . . . . . . Interaction Effects of Association Value by Words Recalled on Complex Retention Task . Interaction Effects of Association Value _by Words Recalled on Simple Retention Task ix Page 21 43 46 48 50 CHAPTER I INTRODUCTION. University libraries have been well stocked with research in the area of reading instruction and the early development of reading skills. The bulk of these efforts have investigated various traits Of the learner and his level of reading achievement or the desirability Of one method of instruction as Opposed to another. The present exploration has attempted to investigate some Of the psychological processes which operate as a first-grader learns to recognize written words. The materials most commonly used to teach reading to children are basal readers. These materials are designed to teach students the skill of word recognition. Very often, the approach used to teach word recognition involves the learning of phoneme-grapheme correspondences. That is, the student attempts to learn the sound that is represented by each written letter and then blend these individual sounds together as they are presented graphemically to form words. 2 In conjunction with this instruction, pupils are also expected to begin recognizing at sight a group of words ”that constitute a high percentage of all the words used in materials written in English. Dolch identified these words in 1935 (Zintz, 1966). He compiled this list by taking words from three other previously organized vocabulary lists. The first list was that published by the Child Study Committee Of the International Kindergarten Union (Zintz, 1966). That study investigated the vocabulary of pre-first— grade children from a list of words which was assembled through the detailed Observations of kindergarten classrooms. From that list, Dolch chose only words which occurred one hundred or more times. The second list consisted of the first five hundred words of the Gates List, a list which has been used as a basis of many studies in reading vocabulary. This list has been recognized as containing words of first importance in children's reading. The third list, formed by Wheeler and Howell, con- sists of 453 words found in ten primers published between 1922 and 1929. When the words common to all three lists were identified, they totaled 193 words. Dolch then selected twenty-seven words that appeared on two of the three lists. 3 The origin of the List is summarized in Table 1.1. TABLE l.l.--Origin of the Dolch List Number of List Source Words I.K.U. Spoken Vocabulary (5 yr. olds) 510 Gates Reading Vocabulary (Primary Grades) 5OO Wheeler & Howell Ten Primers and Ten First Readers 453 Common to all lists 193 Common to two lists 27 220 The Dolch Basic Sight Words are, as the name implies, words that the reader should recognize at sight. Complete mastery of these words is essential. They make up over fifty per cent Of the words found in materials written in English. Table 1.2 shows the results of a vocabulary study that determined the percentage of words found in various elementary school textbooks that are Dolch words. 4 TABLE 1.2.-—Percentage of Dolch Words Found in Elementary Textbooks Grade Number of Textbook Series I II III IV V' VI Reading 4 7o 66 65 61 59 59 Arithmetic 2 , 62 63 57 57 Geography 2 60 59 5LI History 2 57 53 52 A successful reader learns to recognize many of these words quickly and accurately. This is necessary if he is to begin reading soon, since most of the words found in primary materials are Dolch words. Generally, students are expected to learn these words by memorization. Two factors suggest this mode of learning. First, beginning first—graders do not possess the necessary word analyzation skills to attack unknown words. Second, many Of the Dolch words cannot be recognized by using analyzation skills since there is not a one to one phoneme-grapheme correspondence. For example, gaid contains four letters (s-a-i-d) representing only three phonemes and kggw_has four letters (k-n-O-w) representing only two phonemes--no. 5 A systematic method of teaching these words has not been devised. Therefore, pupils often spend a considerable amount of time learning these words through rote drill exercises such as flash card games. The problem investigated in this research effort examined some of the psychological processes that occur as first-grade pupils learn to recognize basic sight words in isolation. Educators have neglected this area of research, and little is known about the processes that influence the word recognition learning task. However, the task of learn— ing to associate a written symbol with its oral representa- tion is believed to be similar to paired-associate learning, an area which has been extensively investigated by several verbal learning psychologists. Paired-associate learning tasks contain two elements, a stimulus and a response. In psychological research these components are usually nonsense syllables consisting of two consonants separated by a vowel. These syllables are scaled according to their meaningfulness. A syllable's meaningful- ness is determined by the number of associations it elicits when presented to a subject. In a paired—associate task the verbal unit on the left is designated as the stimulus member while the member on the right is considered the reSponse. The subject's 6 objective is to learn to recall the response member when pre- sented the stimulus member. An example Of a paired—associate item is big - £93, In this instance, the stimulus is big and the response is Egg, Table 1.3 presents two stimulus- response learning lists, a paired-associate list and a word recognition list. TABLE l.3.--Stimu1us-ReSponse Learning Tasks Paired-Associate Learning Word Recognition Learning §_Member R Member S Member 3 Member rac — bod (down) - down cip — rOf (give) - give nat - waf (said) - said fam - toz (went) - went pic - nib (had) - had tet - neg (not) - not kof - zap (many) - many A similar environment is present in a word recogni- tion task where the subject is to look at the written symbol (down) and respond with the sound Of QQEE: In this instance, the written symbol is the stimulus and is expected to elicit the sound as the response. Although there is an existing similarity between paired-associate learning and word recognition tasks, several differences also exist. For example: 7 l) The elements which make up a paired-associate task might be letters or nonsense syllables. In the word recognition task, the components are meaningful words. 2) The stimulus-response elements contained in a paired-associate task are constructed arbitrarily. Any nonsense syllable may be used in either the stimulus position or the response position. The bik (S) - rOf (R) example discussed above could be reversed rof (S) - bik (R). A word recogni- tion task does not have this flexibility because the written symbol (S) dictates the oral response (R) to be elicited. Word recognition learning and paired-associate learn- ing are undoubtedly similar; but if learning to recognize a word is influenced by the same variables that Operate in paired-associate learning, it has not been reported in edu- cational literature. One objective of the research recorded here was to determine the extent to which paired-associate findings are applicable to word recognition learning. The paired—associate literature shows that a verbal unit's mean- ingfulness value influences rate Of learning. In addition, familiarization training has also been shown to facilitate the learning of paired-associate units. The relationships of meaningfulness and familiarization training to the learn- ing of basic sight words in isolation are the specific vari- ables examined in the present experiment. There were no meaningfulness values available for the Dolch sight words. Thus, it was necessary to scale the 8 meaningfulness values of a selected number of words which were used to complete the experiment. Psychologists scaled the elements of a paired-associate list in several ways. These are reviewed in Chapter II along with the experiments concerning familiarization training and its influence on paired-associate learning. CHAPTER II REVIEW OF LITERATURE Methods of Scaling Meaningfulness Paired-associate research has shown that highly meaningful material is easier to learn than less meaningful material. Determining the meaningfulness values of various types of verbal units has necessarily preceded the experi- ments that examined the influence of meaningfulness on learning. Three distinct methods have been used to rank verbal units from high to low in meaningfulness. These are the association method, the production method and the rating scale method. Early experimenters employed the association method to scale the meaningfulness values of nonsense syllables. Each §_was shown a list of syllables. The syllables were presented one at a time on a tachistOSOOpe for a given period of time. Some experiments exposed the verbal units for two seconds. Others used as much as seven seconds. Subjects were asked to give orally an association for each nonsense syllable presented. The association method 9 10 ranked a nonsense syllable from high to low meaningfulness based On the number of §§_that gave at least one association to the syllable within the time limit. This approach was used by Glaze (1928), Hull (1934), Witmer (1935), Hilgard (1951), and Arclier (1960). Using this approach to gather association values on the Dolch list from six year old subjects presents a serious limitation because of the short period of time given to respond. The responses of a group of second—grade subjects which were recorded in a pilot project indicated that a time limit less than sixty seconds is probably not workable .with primary school subjects. Noble (1952) is primarily responsible for the produc- tion method of determining association values of verbal units. He presented a written unit to his subjects, and they were to respond by writing as many associations as possible in two minutes. The units scaled were dissyllables and non- sense syllables. In determining the meaningfulness of the items, NOble calculated the average number of responses given for each item in the time period to determine its meaningfulness. Mandler (1955) also used this method. The production method was not suitable for scaling meaningfulness values with first-graders because they can neither read nor write. But this manner of determining 11 meaningfulness values can be used, however, if the verbal units are pronounced to the six year Old and he is permitted then to produce his associations to the items orally. The rating scale method required §§_to order items according to ease of learning, familiarity or pronunciation. This method, yielding a familiarity value, was used by Haagen (1949), Noble, Stockwell, and Pryer (1957), and Underwood and Shultz (1960). The rating scale method for determining meaningful- ness values did not seem apprOpriate for first-grade students since it assumes that the subjects can make accurate judg- ments concerning an item's ease Of learning, pronounceability and its familiarity. This method is also dependent upon the subject's ability and willingness to verbalize this information. The writer modified Noble's production method to accommodate the six year Old's communication skills and used it to determine the meaningfulness values of the words to be learned. Several studies have shown that the three methods discussed above produce similar results. A correlation coefficient of .65 between the results of the production method and the number of associations method was found by Mandler (1955). In 1957, Noble, Stockwell l2 and Pryer found a correlation coefficient of .81 and .86 between the meaningfulness values derived from the rating scale method and the values obtained by the number of asso— ciations method for 100 syllables. Correlation coefficients of .90, .92 and .78 were found in three independent studies done by Underwood and Schultz (1960) investigating the rela- tionship between the production method and the rating scale method. The materials that have been sealed for meaningful- ness by those researchers mentioned above have been used in many studies to point out that rate of learning is function- ally related to meaningfulness. Study after study has shown that material of H_meaningfulness value is significantly easier to learn than material of L meaningfulness value. Some of the studies showing this relationship are: Reed (1929), Davis (1930), McGeoch (1930), Sisson (1938), Noble (1952), Underwood and Richardson (1956), Sowling and Brown (1957), Sarason (1957), Braun and Heyman (1958), Kimble and Dufort (1955), Mandler (1956), Noble (1957), and Noble, Stockwell and Pryer (1957). These studies have dealt with scaled nonsense syllables, numbers, nonsense figures using adult subjects. No studies have scaled the meaningfulness values of the 13 basic sight words using beginning first-grade pupils as subjects. The investigations concerning meaningfulness have utilized four types of paired-associate units. These are: (1) high Q stimulus-high m response, (2) low E stimulus- high m response, (3) high m stimulus-low m_response, and (4) low m stimulus-low m_response (H—H, L-H, H-L, and L-L). Experiments have been carried out to determine which of these types is learned most readily. Stoddard in 1929 had one group of §§_learn from French words to English words (L to H), and another group from English to French (H to L). The English words were considered to be more meaningful than the French. Learning was measured in terms of number of correct responses recalled immediately after acquisition. The mean score for the group learning the L-H list was 15.1 words of the 25 presented. Those learning the H—L list had a mean score of 8.0 for the 25 items presented. A 1933 effort by Cason employed l8 paired-associate lists. Each contained 16 pairs. Units are referred to as familiar words and unfamiliar words (F-F, U-F, F-U, and U-U). Two groups were involved in Cason's study; each was given lists with mixed types of verbal units. One group was given four to eight minutes to study the list; the 14 other group had the pairs read to them. The recall was given just after acquisition with the experimenter spelling and pronouncing the stimulus. Subjects were to recall the response. N0 difference was found between the auditory and visual treatments. However, Ss recalled a significantly higher number of F-F pairs than of U-U pairs with the learning of the U—F and F-U types falling between the F-F and U-U types. Sheffield's 1946 study was similar except for the addition of a memory drum to control the time factor per unit. His study showed that H-H material is learned faster than L-L materials. Also L-H units were learned at a slower rate than were H-H units, and H-L were learned some- what more quickly than L-L units. He concluded that differences in meaningfulness Of stimulus has less influence on learning rate than does meaningfulness of response. Kimble and Dufort (1955) constructed lists of ten paired—associates using as stimuli, items from Noble's dissyllables. These items represented the entire range of meaningfulness according to Noble's scale. Common three letter words were paired in the response position. Each list contained the following types of S-R components: H-H, L-H, and H-L. 15 The researchers asked one group to learn a list with dissyllables in the stimulus position and another group to learn a list with the same dissyllables in the response posi- tion. They concluded that the units with the dissyllables in the stimulus position were more difficult to learn. In 1958 Noble, Stockwell and Cieutat made paired- associate lists using four types (H-H, L-H, H-L, L-L); however, each list was limited to one type only. Pairs in the L—H list were identical with the H-L list except for a reversal of position. Each subject was given twelve learning trials. A test trial was administered at the end of each learning trial. The percentage of correct answers for each test trial on each list was used as the criterion for learning. Ease of learning came in the following order: H-H, L-H, H-L, L-L. The finding has been replicated many times under varying conditions by Goss (1965), Nodine (1963), Harleston (1963), Martin, Cox, Boersma (1965), Underwood and Schultz (1960), Lambert and Paiva (1956), and Weiss (1958). Results are the same when the material is presented to individual subjects or to groups; or when material is presented in a constant order or varied order. It appears that meaningful- ness of response has a greater influence on learning than does meaningfulness of stimulus. 16 In order to examine the extent to which factors in paired-associate learning are present in word—recognition learning, the basic sight words to be learned were scaled for meaningfulness. No attempt was made to scale the mean— ingfulness values of the written words (S) because the §§ could not read the words. However, the oral representations of the words (R) were scaled from high to low meaningfulness. Prior to the scaling of association values for the basic sight words, a pilot study was conducted to determine the feasibility of scaling Dolch words with primary pupils. Twelve second-grade pupils were selected at random from a lower middle-class school for the pilot study. Noble's production method was altered to accommodate the §§L. communication skills. Twenty-five Dolch words were selected for scaling after the pupils were asked to read the Dolch list. Any word recognized by the §§_was eliminated from the study. Subjects were given two minutes to emit orally as many associations as possible to each word. The pupils were given practice with a free association exercise before being asked to give associations to the Dolch words. Example: "John, what do you think of when I say dog?" Response--"Spot." "What else?" Response-—"B1ack and white." "What else?" Response--"Cat." etc. As soon as the subject understood the free association "game," the 17 "service" words were presented in place of the common nouns such as dog, food, house, animal, etc. The words were ranked from H—L according to the number Of different asso- ciations given for each word by the twelve subjects. The number of different associations ranged from 11 for the word bagwg to one association for the word gggs, Eight of the words were presented to the §§_in a series Of learning trials. The four words with the highest number of different associations and the four words having the lowest number of different associations were selected. The words chosen for the learning trials occur with the same frequency according to Thorndike's word list. The words were also controlled for length. No word was longer than five letters or shorter than three. The words were also of similar configuration. The learning trials were presented on a memory drum. Subjects were shown each word for ten seconds. The instruc- tor pronounced the word as it was shown and asked the subject to say the word. Each subject was asked to read the words at sight following each learning trial. Accurate recognition on two successive trials was the criterion for learning each word. Test trials and learning trials were alternated until the §_learned all the words. I The result of this brief effort indicated that there was a negative correlation between the meaningfulness 18 value of a word and the number of trials necessary to learn it. The number Of trials required for learning to recognize the written symbol (stimulus) decreased as the number of different associations given for the oral representation (response) increased. TABLE 2.1.--Pilot Study Data X Trials to NO. of Different Word Criterion Associations 1. does 15 l 2. come 14 3 3. been 11 2 4. for 8 3 5. carry 6 5 6. not 6 5 7. once 54 7 8. brown 4 11 r = -.84. This experiment apparently indicated that meaning— fulness does affect the rate at which basic sight words are learned. However, it was realized that this notion needed to be demonstrated with a greater number of subjects with the variables carefully controlled. Familiarization Techniques Several researchers feel that H_meaningful material can be learned more rapidly than L meaningful material 19 because the H_material is more familiar to the learner. It has also been suggested that material which is encountered most frequently becomes the most meaningful. Hence, it is learned more rapidly. Experiments have been conducted that examined the effect of familiarization training prior to learning, to determine the influence of familiarity and frequency on meaningfulness. Familiarization training has been presented in various ways. Some researchers ask the §§ to repeat silently the verbal units to be learned. Gannon and Noble (1961). Others have asked their §§_to repeat the materials to be learned aloud for a period Of time. Cieutat (1960) had §§_simply look at the verbal units to be learned. Underwood and Schulz (1960) asked their §§_to spell the items prior to learning. It seems quite Obvious that the technique used as well as the amount Of preétraining will influence the rate of learning. The skills and abilities of the §§ dictate the type of familiarization training that is feasible. Familiarization training that requires SE to use skills they do not possess can not facilitate learning. Familiarization exercises requiring the learner to read could not be employed with first—grade pupils. However, six year old §§_can speak and hear. Therefore, it was 2O reasonable to assume that the gs could listen to the words to be familiarized and repeat them orally. In addition, primary school §§_could trace or COpy the word to be learned and become familiar with the written symbol or stimulus prior to learning. The experiment reported in Chapter IV reports the influence of stimulus and reSponse familiariza— tion on learning prior to word recognition learning. Theoretical Role of Familiarization Several researchers support the notion that becoming familiar with materials to be learned prior to acquisition trials facilitates learning. Underwood and Schulz (1960) felt that the familiari— zation process made responses more available during the learning trials that followed. Mandler (1954) sees the pre-training on the response component as an exercise that decreases the probability of error as the responses to be learned become more integrated or familiar to the subject. Underwood and Schulz (1960) concluded that familiarization has two purposes: (1) to reduce the number of alternatives to the learned one and (2) to reduce the separate informa— tion elicited by the components of the familiarized item, make them more integrated and available during learning. An illustration of this idea is presented in Figure 2.1 using the paired-associate response items and 21 Paired-Associate Response Learning Sal tiv gur gup cug dah wij fon faw fip roc pol buf dup gen bep naf lil zom pik huf nac goe pak zot sab mos hok max kap cip fam nan aor eof tor tod bot fen dag joz sab fuh ric nok gad Word-Recognition Response Learning own Old at my is am into play fun happy for air home look sit like this time come nice go two cold when ride both was pretty brown with done let food can then they in were so the saw me could an these very but the about to once if little run wash six do Figure 2.l.-—Paired-Associate and Word-Recognition Response Learning 22 word-recognition response items shown in Table 3. The items shown within the circles are the responses to be learned by the subject. Isolation of these items within the learner's memory to the exclusion of all other possible responses seems to have a facilitating influence upon response learning. This has been shown in paired-associate studies. A second type Of learning, stimulus learning, has been facilitated by familiarization also. Gibson (1940) feels familiarization aids in the discrimination between items to be learned. Stimulus items become more distinc- tive as they become familiar. Familiar stimuli have become distinctive in the learner's memory, and are more readily joined with the apprOpriate response during acquisition trials. This conclusion leads one to hypothesize that a learner would learn the stimulus member (written word) of a word recognition task more readily if he were familiar with the shape or configuration of the stimulus prior to acquisition trials. It seems logical that being familiar with the shape Of a word would decrease the difficulty of associating that word with the prOper oral response. If a subject were familiar with the paired-associate and word recognition stimuli shown in Table 2.2, the ease of associ- ating each stimulus with the prOper response is increased. 23 TABLE 2.2.--Paired-Associate and Word-Recognition Stimulus Learning Paired—Associate Stimulus Learning (r1) bod (r1) zap (r2) rof (r2) rof (Sl) rac (r3) waf (SQ) pic (r3) toz (r4) neg (r4) nib (r1) /not/ . (r1) /give/ (r2) /had/ ‘ (r2) /down/ (Si) (not) (r3) /many/ (32) (had) (r3) /had/ (r4) /down/ (r4) /said/ There are two positions regarding the effect of familiarization on the learning of a paired-associate list. It is held by some researchers that familiarization facili— tates learning of low meaningfulness units. However, Iambert and Jakobvits (1960) and Kanungo, Lambert and Mauer (1963) feel that familiarization can have a prohibitive effect. This is very likely to be the case if the material being familiarized is of H_meaningfulness value. The pro- ponents of this idea state that too much repetition decreases the meaningfulness value causing the item to become L_in meaningfulness. If a H meaningfulness item becomes L_as a result of familiarization, then it would follow that such pre-training retards the learning of H meaningfulness 24 material. This theory also states that the opposite is true. Familiarization of a L meaningfulness item will cause it to become H and easier to learn. The following reports the results of studies that support these ideas concerning the effect of familiarization exercises. A 1946 experiment by Scheffield examined the learn— ing of a H-H list without response familiarization and H-L list which received response familiarization. The treatment given the H—L list had an effect of facilitation. A com— parison of H-H, H-L, and L-L which had received response training and similar lists receiving no treatment was made by Weiss in 1958. He found the average number of trials to criterionon the familiarized units to be less than those receiving no treatment. Other research efforts have found no difference in the acquisition of familiarized response members and those receiving no treatment prior to learning. For example, in 1960 Cieutat arranged two mixed lists containing four pairs of L dissyllables. The same §§_were utilized for all treat- ments. Familiarization was given by having the SE look at the items for sixty seconds. The finding in this instance was that familiarization of the response component inhibits learning when paired with an unfamiliarized component. Rank- ing the four combinations in terms of ease of learning from 25 easiest to most difficult was found to be familiarized- familiarized, unfamiliarized—unfamiliarized, familiarized— unfamiliarized, and unfamiliarized-familiarized. A 1961 study by Kanungo, Lambert and Mauer indi- cated that response familiarization did not facilitate learning. Two groups were used in this study. The researchers constructed a paired-associate list containing H-H components. The groups were given learning on the same list with one group receiving response familiarization prior to learning. This group's acquiSition rate was inferior to the group receiving no treatment. The research that has been primarily concerned with stimulus familiarization is also inconsistent. In 1961 Gannon and Noble showed that familiarization of the stimulus element improved learning. This finding was supported by Martin (1963) and Martin and Schulz (1964). However, no significant difference in learning with familiarization of either component was found by Bailey and Jeffery in 1958. In any case, there appear to be three learning processes which include stimulus learning, response learning and the learning which occurs between stimulus and response learning. Several of the studies discussed above suggest that meaningfulness facilitates these learning processes as does familiarization training prior to learning. 26 The present study concerned itself with two factors, meaningfulness and familiarization. If these factors operate in learning to recognize words in isolation as they have influenced paired-associate learning, the following predic- tions can be logically made. Research Hypotheses H01 The number of trials to criterion on high m words receiving no familiarization training is significantly less than the number of trials to criterion on low m words receiving no familiarization training. H02 The number of trials to criterion on low m_words receiving high response familiarization training is significantly less than the number of trials to criterion on low m_words receiving no familiarization training. H03 The number Of trials to criterion on low m words receiving high stimulus familiarization training is significantly less than the number of trials to criterion on low Q words receiving no familiarization training. H04 The number of trials to criterion on low Q words receiving high stimulus and high response familiari- zation training is significantly less than the number of trials to criterion on low m_words receiving no familiarization training. ‘ H05 The number Of trials to criterion on high m words receiving high response familiarization training is significantly less than the number of trials to criterion on high m_words receiving no familiarization training. H06 The number of trials to criterion on high m_words receiving high stimulus familiarization training is significantly less than the number of trials to criterion on high m_words receiving no familiarization training. 27 H07 The number of trials to criterion on high m_words receiving high stimulus and high response familiariza- tion training is significantly less than the number of trials to criterion on high m words receiving no familiarization training. CHAPTER III METHODOLOGY Scaling of Materials Several methods of scaling the meaningfulness of verbal units were discussed in Chapter 11. One of the modes used to scale the meaningfulness values of such material was devised by Noble in 1952. His system of determining the meaningfulness of a verbal unit consisted of presenting the written unit to a subject, and the subject was expected to write as many associations as possible in two minutes. Noble counted the average number of responses given for each unit to determine its meaningfulness value. This manner of scaling meaningfulness required subjects who could read and write. In scaling the material to be studied in this experi- ment, Noble's production method was modified to accommodate the communication skills of first-graders at the beginning of the year. Instead of presenting the subject with a written symbol, the experimenter pronounced a word to the subject. The subject was instructed to give orally as many 28 29 associations to the word as possible in 150 seconds. For example: Experimenter--"What do you think of when I say rug?" Subject--"fast." Experimenter--"What else?" Sub- ject--"go." Experimenter--"What else?" Subject--"car." This continued for 150 seconds or until the subject failed to give any response for 60 seconds. The only encouragement given by the experimenter was to ask "What else do you think of when I say run?" or "What else?" Materials The words scaled in this experiment were selected from the Dolch list of basic sight words. Two criteria were used for their selection. Each word has been identified by Thorndike's research as having a similar frequency of occurrence in the English language (AA rating). The words were also controlled for a length of two to six letters. Words with homOphonus forms such as at§_and g1ght_were excluded. These words were presented to each subject in the order shown in Table 3.1. Forty first-grade subjects were selected for this experiment. Using standard randomization procedures, the subjects were selected from a population of 120 first- graders attending a lower-middle class public school in Tallahassee, Florida. Subjects found to be repeating first 30 grade were eliminated from the study. The subjects were randomly assigned to two groups of 20 each. TABLE 3.l.--Words Selected for Scaling l brown 7 saw 13 six 19 them 25 were 2 walk 8 if 14 wash 20 very 26 then 3 little 9 was 15 cold 21 yellow 27 about 4 ride 10 many 16 run 22 go 28 how 5 play 11 went l7 done 23 can 29 where 6 could l2 seven 18 once 24 pretty 30 into The subjects were interviewed individually by the experimenter between 9:00 A.M. and 11:30 A.M. on regular school days. Before asking a subject to give associations to the words in Table 3.1, he received instructions on how to free associate. These instructions were in the form of a simple game. Example: Experimenter--"Johnny, what do you think of when I say Eggd?" Subject-~"a hot dog." Experi- menter--"What else?" Subject--"breakfast" and etc. This was continued until the subject consistently gave expected associations to such common nouns as food, animal and house. When a subject had demonstrated an understanding of the free-association technique, the basic sight words were introduced in place of the common noun.. Each subject responded with as many associations as possible for each of the thirty words. This procedure involved approximately 31 one hour with each subject. Therefore, it was necessary to stOp the activity at fifteen minute intervals allowing the first-grade subject a rest. A word was assigned a value of high or low meaning— fulness on a basis of the number Of different associations it received from a group of subjects. Example: If every sub- ject inia group gave the same association to the word i333, that word received a score of Egg different association. Several subjects, when asked, "What do you think of when I 1! ll say brown?" responded with a dog. This response counted as gng_different association for the word brgwn, Since brown received 24 different associations from Group A and 20 differ- ent associations from Group B, it was considered to be of high meaningfulness value. The word w§r§_received one different association from each group and was considered to be low in meaningfulness value. A Pearson product—moment correlation coefficient statistic was used to determine the agreement of the order of the scaling between Group A and Group B. The results are shown in Chapter IV. After the words were scaled from high m4to low Q, four of the high m_words and four of the low Q words were chosen for use in further exploration. The eight words chosen are shown in Table 3.2. These words were used to 32 determine if meaningfulness or familiarization training influences the rate of learning the words. TABLE 3.2.—-Basic Sight Words Selected for Familiarization Treatment and Learning Trials i ‘——‘ ’ _f Number Of Different Associations Word A B m Value brown 24 20 H play 23 22 H run 20 24 H seven 19 17 H if 1 2 L could l 2 L done 1 l L were 1 l L Subjects .Sixty first-grade subjects were selected at random from the pOpulation described previously. Students repeating first-grade were eliminated. In addition, any student who could recognize any of the eight words was eliminated from the study. These restrictions reduced the total pOpulation to 106 first-graders. Procedures Subjects were randomly assigned to four treatment groups of fifteen §§_each. The treatments of these groups 33 were designed to examine the influence of meaningfulness and familiarization training on the learning of the basic sight words listed in Table 3.2. One group received no treatment or familiarization training. This particular group of subjects served as a control group (C-group). After the experiment was completed and the statistical procedures were applied, a comparison of the number of trials necessary for learning high m_words and low m_words was made to examine the effect of meaningfulness on learning. Another experimental group of subjects were given response familiarization training (R-group). This group's performance was used to test the predictions that learning is facilitated when subjects have memorized the oral response of the Sight words prior to the word recognition task. A third group received stimulus familiarization (S-group). The results of this group's performance were employed to test predictions made at the conclusion of Chapter II regarding the effect of pre-training on the stimulus member of a word recognition task on the learning of that word. Finally, a group of subjects were given both stimulus and response training (S-R-group). The two types of familiarization were given separately. The subjects were 34 given stimulus training prior to response training. The two research predictions made concerning the effectiveness of both treatments prior to learning were examined by comparing this group's performance with that of the control group. Response familiarization required §§_to memorize the oral representation of the eight words to be learned. The words were tape recorded in five different random orders (see Table 3.3). four second intervals. each word after it was voiced on the recorder. The words in each list were recorded in Subjects were instructed to say Subjects attempted to recall each word from memory after each familiarization exercise. was the criterion for response familiarization. Accurate recall of all the words Response familiarization trials and recall trials were alternated TABLE 3.3.--Response Familiarization Lists A B C D E brown done run if could could seven play done run seven brown seven run done done run could brown if if were were were play run if brown could brown were play done seven seven play could if play were 35 until the S reached criterion. Recall periods were a maximum of 30 seconds. Learning trials began immediately after the S was familiar with the experimental responses. Stimulus familiarization training involved two activities. Subjects receiving this treatment traced and cepied the eight words (stimuli) to be learned. The words were presented in isolation on 4x6 cards. Each word was printed in lower case letters. Without telling the §§ the word, each was asked to look carefully at the word and trace it three times. After training, the subject then copied each word one time with a crayon. After the stimulus familiarization exercise, the S was shown a list of twelve words containing the eight words involved in this study (see Table 3.4). The 3 attempted to underline the eight words being considered. Subjects reached criterion for stimulus familiarization when the experimental words (stimuli) could be identified. The familiarization exer- cise was repeated if the subject could not accurately identify all the experimental words. Complete stimulus familiarization trials were alternated with identification trials until criterion was reached by the subject. Since these §§_were first-graders, attention span problems had to be considered. Therefore, it was Often necessary to meet with a subject more than one time. However, no 36 subject failed to reach criterion after three twenty minute sessions. TABLE 3.4.——Stimulus Familiarization Identification List brown ran many were cat done play so could if seven three Learning trials were presented to each S individually. All familiarization training, learning trials, and test trials occurred between 9:00 A.M. and 11:30 A.M. on regular school days. I The learning trials were presented on a memory drum. The five lists used during response familiarization were reordered and typed on a memory drum in lower case letters (see Table 3.5). Each word was presented for four seconds. As the word appeared, the experimenter pronounced it and then the S pronounced it. This procedure continued until each word in the list had been exposed to the subject. This procedure was repeated three times on the same list prior to each test trial. For example: Subjects began with list A, After looking at each word individually as it appeared on the drum, the subject pronounced it after hearing the experimenter say the word. After going through list A three times a test trial was attempted. Following 37 a test trial, three more learning trials were given using list B, at the end of which another test trial occurred. The lists shown in Table 3.5 were used repeatedly from A through E until criterion for learning was reached by the subject. TABLE 3.5.-—Learning Trial Lists A B C D E brown were play done ran done done ran if seven ran seven done seven were were brown if could if could if could were play play ran brown ran done seven could were brown could if play seven play brown Projector which was equipped with a flash meter. Test trials were presented on a Keystone Overhead The words were projected onto a white screen six feet from the subject. Projections were adjusted to the subject's eye level. As a word appeared on the screen, the subject was instructed to look at the word and say it. Each word was exposed for two seconds.1 The subject was given only one Opportunity to recognize each word during a test trial. were included on each test trial. were learned on test trial number four, All the words For example, if a word it was not eliminated 38 from subsequent learning trials or test trials. Total exposure time on the eight words for a test trial was sixteen seconds. A word was considered learned when the subject accurately recognized it on two successive test trials. If a S accurately recognized the word rag_on test trials numbers five and six, the experimenter considered the word learned on trial number six. The dependent variable was the sum of the number of trials to reach criterion on each word. Every word was included in each learning trial and test trial. Learning trials and test trials were continually alternated until each word had been recognized successfully on two successive test trials or until twenty test trials were completed. Any word not learned after twenty test trials was given a score of twenty. Two days after acquisition, §§ were asked to read the words at sight. Two types Of grammatical structures were constructed to test the §§_recall of these words. One task required the Ss to read the complex sentence, "If you were done, you could run and play with the seven brown dogs." All words in this sentence except If were printed in lower case letters on standard notebook ' paper. It required the Ss to learn the additional words 39 you, and, the, with and dogs. The classroom teachers pro- vided instruction on these words. Each §_demonstrated that he knew these words before attempting the recall sentence. The second recall task required the §s_to read the following: I run, I play, I could, seven dogs, brown dogs, you 323g, I am gggg, ;£_I go . . . . All the words except I’in these structures were printed in lower case letters on standard notebook paper. The §§ were required to learn the additional words I, gggg, ygu_and am, The §§ classroom teachers provided this additional instruction also. The §§_were given 30 seconds to read the complex sentence. The number of experimental words read correctly was recorded. The §§_were allowed five seconds to read each of the simple structures. The number Of experimental words read correctly was recorded. The highest possible score on either task was eight, with the lowest possible score being zero. CHAPTER IV RESULTS Scaling Meaningfulness Values Following the procedure described in Chapter III, two groups of 20 subjects each were employed in the scaling of the meaningfulness values of the thirty basic sight words listed in Table 3.1. The words were scaled from high mbto low m_according to the number of different associations given by an entire group. Table 4.1 presents a summary of the number of different associations given by each group on the individual words. Appendix A summarizes the associa- tions given by both groups. A Pearson product-moment correlation coefficient was computed to determine the agreement between Group A and Group B in the ranking of the Dolch words from high m_to low m. A correlation of .93 indicated a high degree of agreement between the groups (significant at .001 level). The high significance of the correlation indicates both that the two student groups are from the same pOpula- tion (measure of reliability) and that these words have been 40 41 consistently ranked from high m4to low m_by both groups. This agreement between the groups of §§_used for sealing provided a defensible base for the latter phase of the study. TABLE 4.l.--Number of Different Associations for Each Word — ‘ ~— _7 .— Groups . Groups Groups Word A B Word A B . Word A B brown* 24 20 six 15 11 how 5 6 play* 23 22 yellow 15 14 was 5 4 run* 20 24 wash l4 16 very 3 2 seven* l9 17 can 11 ll them 3 3 cold 19 2O saw 10 8 where 3 4 ride l7 15 many 10 10 then 2 1 walk l7 13 once 9 7 if* 1 2 little l7 18 went 9 10 could* l 2 go 15 15 into 6 7 done* 1 1 pretty 15 17 about 5 5 were* 1 l *indicates experimental words Two complexities were noted during the collection of the scaling data. First, the first-grade §§ Often became restless, thus it was necessary to provide a short period Of rest every fifteen minutes. Secondly, §§ often abandoned the task of giving associations to the "word" being scaled and began a rhyming game. An example Of this would be: Experimenter: "What do you think of when I say gun?" Subject: "fast." Experimenter: "What else?" Subject: "far." Experimenter: "What else?" Subject" "around the 42 house." Experimenter: "What else?" Subject: "go." Experimenter: "What else?" Subject: "poe." Experimenter: "What else?" Subject: "row." Experimenter: "What else?” I! Subject: "doe. The subject described here Obviously ceased to give associations to run and began to play a rhyming game. These types of associations were discounted. It is inter- esting that this type of behavior occurred only on words scaled as high m and never on the words that were eventually scaled as low m, Therefore, ignoring the rhyming responses in no way influenced the scaling of the list. Experimental Design The data used to examine the null hypotheses were analyzed using a 2x2x2 factorial design with repeated measures on the third factor. TheSe factors were (1) response familiarization (R), (2) stimulus familiarization (S), and (3) meaningfulness (m). A model of the experimental design is shown in Figure 4.1. Using four high m_words and four low E words from Table 3.5 an experiment to study the influ- ence of the factors shown in the experimental model on the rate of learning high m and low m words was conducted. Analysis of Data The raw data used to examine the statistical hypothe- ses are recorded in Appendices C and D. Appendix G contains 43 Group 2 Group 4 Group 4 I High Group 1 Group 3 Famil- o ,3 iarization g c o rt '3 LO Grou. l Grou 3 O ‘3 w p p esponse O U) (0 <1: 0 + Stimulus Familiari- zation Figure 4.l.--Model of Experimental Design a record of the number of test trials to criterion for each subject. Appendix D is a compilation Of the number of words accurately recognized on each of the retention tasks by each subject. Each of these three sets of data was treated with an analysis of variance. Summary tables for these analyses are included in the following discussion along with interaction 44 graphs which have been plotted for each set of data. Test Trial Analysis Only the following F values were found to be signifi- cant. The main effect of meaningfulness Of words within subjects was significant at p.<.001 supporting the assump— tion that words assigned to high m and low E categories do differ significantly in mean number of trials to criterion and therefore are not from the same level of difficulty. The main effect of response familiarization between §§ was found to be significant at p.<.01. This result shows that prior response memorization in the R and S-R groups significantly influenced the rate at which the words were recognized during original learning. The interaction effect of response familiarization and meaningfulness within subjects was found significant at p.OJ Ulh~k r1 r1 b-S- W3C) h— Hd> “sluis- I>OJoI c301w1 r4«3c> rid‘C> «3 «I ca 01 co Ch:TCU a~wi rlfllfll rlfllrl o: 01 a: an «i midis: @363“) rdd‘ri rid‘fil on cu a: a: a: l I h1oia- «id» tacnco ran-rt cu a: 03 r1 \0 our: \ocmcu rdd‘rl r: (n on cu an a) Exoin- .::¢ rdfilri r4:tri 01 c: a: a: d‘OJd‘ :rarr carers (fin-GI cu cu a) a: a: a: «10101 01“) dqfllfll :rcu a) a) a: a) ‘ cu:ro1 and HM ash“? a a PIN)?! rtwiwi K401r1 \ornra O. on GI 01 a) g a) a: h (5 c>aiol c>aiwi UWPTC) u401c> 01 cu a) a: s a: an .9, .p O\«i«1 «L: :1G3c> d1fllrl m r: .4 a) a: ‘3 a: on L. m d>r1c> «>610: «qr-r: :rcu r1 .x rt .4 U) .3: U) H 02 C0 :4 to :3 s a £3 E: m b-Olrl E! bJOJRi oqwirt (flflfiri h. x rt Fl 03 '3 co m .3 ca '3 co m Q gI G HI m>owoJ SI m>didi talolc> r1 0 O r! rt 03 a) rapnco a. <3 co 0: a: ‘0 z: tut—l tut—1 a: tut-SI Din—l Stimulus Familiarization Group Complex Task 4 2 4 l 3 2 336 S37 538 $39 540 4 3 2 0 S 3 O fiqW301 a) giviri U) qufri a) amt—I U) Simple Task NOON \o-sm mam m U) I-l-I‘I'N m U) 81 Stimulus and Response Familiarization Group Complex Task S 2 2 SIM (\Jr-I S46 S41_ S48 S49 350 Simple Task S46 347 $48 $49 350 851 $52 353 S54 I]. ‘IN N'IIJI I APPENDIX E RESPONSE FAMILIARIZATION PILOT STUDY DATA Subjects Trials to Criterion Time to Criterion SI 15 8:00 32 20 10:40 S3 14 7:28 SA 16 8:32 S5 22 11:44 85 13 6:56 S7 15 ' 8:00 88 21 11:12 89 19 10:08 310 14 7:28 811 22 11:44 $12 17 9:04 $13 18 9:36 $14 17 9:04 S15 19 10:08 S16 16 8:32 $17 17 9:36 818 23 12:16 319 26 13:52 820 22 11:44 X = 18.30 X = 9:47 82 I . a I . VVVV APPENDIX F INSTRUCTIONS GIVEN TO SUBJECTS Scaling of Materials "Tell me what you think of when I say A ." When a subject gave an association, he was asked, "What else do you think of?" Stimulus Familiarization Each subject receiving this treatment was asked, "Look at this word carefully and trace it three times with your crayon." After completion of this task he was told, "Now copy the word on this paper with your pencil." Response Familiarization Subjects receiving this treatment were told, "You are going to hear a few words on the recorder. Listen very carefully and say each word after you hear it." This same procedure was used for the response familiarization pilot data collection. 83 l.ll‘lll . 1‘ 84 Learninngrial Data Learning trials were presented on a memory drum. Sub- jects were told, "Look at each word carefully and listen as I say each word. After I have pronounced a word you say it, too." Test Trial Data Subjects were asked, "Look at the word on the screen and say it." Retention Task Data Instructions for both retention tasks were the same. subjects were instructed, "Look at these words and read as many as you can." ‘4. hmi‘dfl‘ r.