wags MN“ I W“ 3 M 8 Michigan S ta te University ABSTRACT EVALUATION OF A TECHNIQUE FOR IDENTIFYING COMMUNICATION PATTERNS OF MOTHERS AND PRESCHOOL CHILDREN OF TWO SOCIOECONOMIC GROUPS by Trellis Taylor Waxler This study attempted to determine if the communi- cation of children could be reliably coded. It also at- tempted to identify various communication patterns of chil- dren during mother-child interaction within the home. The subjects were nine children attending the Michigan State University Laboratory Preschool during the Fall term 1966 and Winter term 1967, and their mothers. The data used in the study were collected in a Michigan State University Agricultural Experiment Station Project 786, "Use of Space, Material and Communication Among Family Members Performing Family Activities." The mother and preschool child in each family were observed performing four selected activites. The communication data were tape recorded and later transcribed. Eight five-minute samples of communication between mother and child were coded using the procedure described in Magrabi, et a1. ("Framework for Studying Family Trellis Taylor Waxler Patterns,‘ Journal of Home Economics, Vol. 59, Nov. 1967, 714-719). The communication was coded on two dimensions-- the resource and the mode. A resource is defined as a concept a sender uses to influence another person. Resources include fact, preference, direction, and motivation. The mode is the manner in which the resource is transmitted to the receiver. Modes include offering, seeking, accepting, and not accepting. Together they form sixteen mode- resource categories. It was found that the coding procedure does dis- criminate between the sixteen categories. Offering fact, offering direction, and seeking fact are the most frequently used categories. The subjects in the middle socioeconomic group communicated significantly more than did the subjects in the lower socioeconomic group. They also seemed to offer fact more often than did the subjects in the lower socio- economic group. The subjects in the lower socioeconomic group offered direction more often. Intercoder agreement was checked by finding the agreement between two trained coders. The reliability of the instrument was also checked. The ability of the coding procedure to discriminate socioeconomic classes demon- I strated a certain level of validity of the coding procedure. EVALUATION OF A TECHNIQUE FOR IDENTIFYING COMMUNICATION PATTERNS OF MOTHERS AND PRESCHOOL CHILDREN OF TWO SOCIOECONOMIC GROUPS BY Trellis Taylor Waxler A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Home Management and Child Development 1968 ACKNOWLEDGMENTS Appreciation is expressed to Dr. Frances Magrabi for her guidance, direction, and encouragement in planning and conducting the study. Appreciation is also expressed to Dr. Martha Dale, Miss Phyllis Lueck, and Dr. Ellen Strommen for their helpful suggestions. Appreciation is also expressed to Miss Mabel Merrill and Mrs. Angela Smith, Graduate Assistants, for their sug- gestions and for their assistance in determining intercoder agreement. ii TABLE OF CONTENTS Page ACKNOWLEDGEMENTS O I O I C O O O O C O O O O O C O 0 ii LIST OF TABLES O I O I O O O O O I I O O O O I O O 0 iv LIST OF APPENDICES . . . . . . . . . . . . . . . . . Vi CHAPTER I 0 INTRODUCTION 0 O O O O O O O O O O O O I O O 1 Definition of Terms . . . . . . . . . . . . 3 Assumption 0 O I O O O I I C I O O O O O O 4 Objectives 0 O O O O O I O I O I O O O O O 4 Hypothesis . . . . . . . . . . . . . . . . 5 II. REVIEW OF LITERATURE . . . . . . . . . . . . 7 III. METHODS AND PROCEDURES . . . . . . . . . . . 20 Selection and Description of Sample . . . . 20 Collection of Data . . . . . . . . . . . . 23 ceding the Data 0 O O O O O O O O O O O O O 25 Intercoder Agreement . . . . . . . . . . . 28 Reliability 0 O O O O O O O O O O O O O O I 30 validity O O O O O O O O O O O O O O O O O I 30 Analysis of Data . . . . . . . . . . . . . 30 IV. PRESENTATION AND DISCUSSION OF RESULTS . . . 34 Reliability 0 O O I O O O O O O O O O O O O 34 validity O O O O O O O O I O O C O O O O O 3 5 Identification of Communication Patterns . 35 Relationships Tested by Analysis of Variance . . . . . . . . . . . . . . . . 46 Limitations of th Study . . . . . . . . . 47 v. SUMMARY AND CONCLUSIONS . . . . . . . . . . . 51 BIBLIOGRAPHY O O O O O I O O O O O O O O O O O O O I 54 APPENDICES I O O O O O O O O O I O O O O O O O O O C 57 LIST OF TABLES Table Page 1. Annual Income Per Family . . . . . . . . . . . 22 2. Distribution of Families According to Level of Education of Parents . . . . . . . . . . 23 3. Number of Children Per Family . . . . . . . . 24 4. Age of Parents in Each Family . . . . . . . . 24 5. Paradigm of Mode-Resource Categories . . . . . 28 6. Methods Used in Analysis of Data . . . . . . . 32 7. Reliability Between Observation Sessions and Code Categories . . . . . . . . . . . . . . 34 8. Summary of Analysis of Variance . . . . . . . 36 9. Total Number of Message Units of Mothers and Children for Each Socioeconomic Level . . . 37 10. Mean Number of Message Units for Each Socio- economic Level for Each Mode-Resource Category . . . . . . . . . . . . . . . . . . 38 11. Percentage of Message Units in the Mode- Resource Categories for Mothers and Children in Each Socioeconomic Group . . . . 39 12. Percentage of Message Units in Each Mode for Mothers and Children of Each Socio- economic Level . . . . . . . . . . . . . . . 39 13. Percentage of Message Units in Each Resource for Mothers and Children of Each Socio— economic Level . . . . . . . . . . . . . . . 40 14. Percentage of Children's Responses in Each Mode-Resource Category to Mothers' Offering Fact . . . . . . . . . . . . . . . 42 iv Table 15. l6. 17. 18. Percentage of Children's Responses in Each Mode-Resource Category to Mothers' Offering Direction . . . . . . . . . . . Percentage of Children's Responses in Each Mode-Resource Category to Mothers' seeking Fact 0 O O O O O O O O O O O O 0 Mean Number of Message Units for Mothers and Children in Each Mode-Resource Category . . . . . . . . . . . . . . . . Frequencies of Message Units in Each Mode- Resource Category for all Subjects . . . Page 43 44 48 48 LIST OF APPENDICES Appendix Page I. Original Data for Mode-Resource Categories . . . . . . . . . . . . . . . . 58 II. Original Data for Modes . . . . . . . . . . . 60 III. Original Data for the Resources . . . . . . . 62 vi CHAPTER I INTRODUCTION Sapir (1949, p. 15) stated that, "The primary func— tion of language is generally said to be communication." The development and function of language in children has been studied by many researchers including Piaget (1925), McCarthy (1930, 1954), and Templin (1957). Nevertheless much confusion and controversy exists over the interpreta- tion of the results of these studies. McCarthy (1954) feels that many of the discrepancies in the findings from various studies of the function of language in children can be accounted for in terms of (1) differences in definition and interpretation of terms, (2) the situation in which the responses are recorded, and (3) individual differences of the children being observed. Although the immediate situation in which the lan- guage of the child has been recorded has been recognized as an important variable, not much cognizance has been taken of the fact that because of the very nature of lan- guage and communication this situation is an interaction_ process (Smith, 1966). Some of the situations that have been varied and studied include conversation of child to ‘ - l!‘)_, .5 “ adult, conversation during outdoor free play, conversation during indoor free play, and outdoor play situations at home. In studying the function and development of lan- guage in children there is a need for a system that has clearly defined terms that are applicable to all communica- tion situations and that take into account the interactive nature of communication. The procedure of coding communication that was tested in this study is an adaptation of a method of coding interpersonal behavior developed by Richard Longabaugh (1963). Longabaugh (1963, p. 321) stated that ". . . inter- personal actions are dimensionalized on two criteria: (1) the resources(s) salient in the interaction, and (2) the modality used by the actor in connection with the resource." An investigator is free to define resources as he wishes. There are two requirements for the conceptualiza- tion of what resources should be included for study. First, the resources measured should correspond to some extent to those actually valued and exchanged by the participants. Secondly, the transmission of these resources must be satisfactorily measured in the situ- ation observed. In practice, the investigator's suc- cess in measuring resources present will be indicated by the degree of empirical validity of the obtained measures. (Longabaugh, 1963, p. 321) Modes in the Longabaugh system are seeking, offer— ing, depriving, accepting, and not accepting. The Longabaugh categories were modified by Magrabi et al. (1967) in order to make them more appropriate to the —. study of family activity patterns within the home. Modes are defined as seeking, offering, accepting, and not accept— ing; the salient resources were defined as fact, preference, direction or command, and motiviation or encouragement. Together they form sixteen mode-resource categories. This system was devised to code only verbal communication and was used by Smith (1968) to code the communication of mothers interacting with their preschool children. The data analyzed in this study were collected in a Michigan State University Agricultural Experiment Station Project 786, "Use of Space, Material and Communication Among Family Members Performing Family Activities." The mother and preschool child in each family were observed performing four selected activities. The communication data were tape recorded and later transcribed. Definition of Terms Communication is defined as an interaction situation in which the person speaking transmits a message to the re— ceiver with the conscious attempt to affect the latter's behavior. Communication pattern is the frequency of message units in a particular mode-resource category. A resource is the concept a sender uses to influence another person or persons. These include: (1) fact, (2) preference, (3) di- rection, and (4) motivation. Mode is the manner in which the resource is transmitted to the receiver. This include: (l) offering, (2) seeking, (3) accepting, and (4) not accept- ing. Each attempt to transmit a single thought or idea from one person to another is a message unit (Smith, 1968). Assumption It was assumed that verbal communication can be dif- ferentiated into categories for analysis. That is, communi- cation can be categorized in terms of its salient modes and resources . Objectives The first objective was to test the reliability and validity (insofar as possible) of the Magrabi gE_al. proce- dure for coding the comunication of children. As Longabaugh (1963, p. 321) stated, "If the defined resources cannot be measured reliably--or, once measured, they fail to have empirical validity--then it is likely that the conceptual- ization of resources has been inadequate." The second objective was to identify the various communication patterns of children using the Magrabi coding procedure. The following comparisons were made: 1. Comparison of mother and child on the sixteen mode- resource categories utilized in the Magrabi procedure. 2. Comparison of the two groups of children on the six— teen mode-resource categories. 3. Comparison of mother and child on the four mode cate- gories utilized in the Magrabi procedure. 4. Comparison of the two groups of children on the four mode categories. 5. Comparison of mother and child on the four resource categories. 6. Comparison of the two groups of children on the four resource categories. 7. Comparison of the responses of the two groups of children to various communication patterns of the mothers. Hypothesis The general hypothesis upon which this study was based is that communication patterns are learned by imita- tion of the model of the mother and hence communication patterns of children will be related to the communication patterns of their mothers. The communication patterns of the children and the mothers from the two socioeconomic groups were expected to be different. Using a three-factor analysis of variance seven relationships were statistically tested: 1. Difference between socioeconomic levels. 2. Difference between mother and child. 3. Interaction between socioeconomic level and family membership (mother and child). 4. Difference between the sixteen mode-resource cate- gories (codes). 5. Interaction between socioeconomic level and codes. 6. Interaction between family membership and codes. 7. Interaction between socioeconomic level, family membership, and codes. CHAPTER II REVIEW OF LITERATURE The review of literature is divided into four parts: (1) the function of language, (2) influence of the situation, (3) the influence of social class, and (4) the coding of com- munication. Function of Language Susan Ervin-Tripp in her review of language develop- ment stated that, "The greater theoretical agreement and methodological precision in the study of linguistic form has led to a natural emphasis on this aspect of children's language, in contrast to studies of function where one must devise testing methods and even social theories." (Ervin- Tripp, 1966, p. 85) Language is one of the ways in which we express our ideas (Berlo, 1960). The main purpose of human communica- tion as far as he is concerned is to affect the behavior of others. He states that, "All communication behavior has as its purpose the eliciting of a specific response from a specific person (or group of persons).' (Berlo, 1960, p. 14) He listed five verbal communication skills: (1) writing, (2) speaking, (3) reading, (4) listening, and (5) reasoning or thought. This study was concerned with spoken communication. John Dewey (1925) and Edward Sapir (1949) agree that the primary purpose of language is communication. "The heart of language is not 'expression' of something anteced- ent, much less expression of antecedent thought. It is com- munication. . . ." (Dewey, 1925, p. 178) In 1926 Piaget's book The Language and Thought of the Child was published. This book presented a new approach to the study of the function of language. Piaget was mainly interested in the child's language as a means of revealing his thought processes. Egocentric speech and socialized speech were the two types of sppech that Piaget found in the child's language. In egocentric speech, as defined by Piaget, the child does not bother to know to whom he is speaking nor whether he is being listened to. In socialized Speech the listener must be considered. The speaker must consider the point of view of the listener, and try to in- fluence him or exchange views with him. Egocentric speech is defined as (1) repetition, vocalization in which the child reports words and syllables for the pleasure of talk- ing; (2) monologue, in which the child talks to himself as though he were thinking aloud; (3) collective monologue, in which another child is present but his point of view is never taken into account. Socialized speech is defined as (1) adapted information, when the child really exchanges his thought with others; (2) criticism, which includes re- marks about the work or behavior of others; (3) commands, requests, and threats; (4) questions; and (5) answers. Socialized speech corresponds to communication as defined by Berlo (1960). Miller (1951) sees language as serving four main functions: (1) to increase uniformity of information, (2) increase uniformity of opinion, (3) to change statuses in groups, and (4) to express emotion. As we can see from the above theorists, language does serve many functions. Nevertheless, it does appear to this writer that the main function of language is that of communication. According to Gerald Miller (1966) com- munication has as its main focus ". . . those behavioral situations in which a source transmits a message to a re- ceiver(s) with conscious intent to affect the latter's be- haviors." (Miller, 1966, p. 92) Therefore, this study was designed to study communication that takes place in the home in which a mother and her preschool child are perform- ing some household activity. Influence of Situation Piaget (1926) estimated that 38 per cent of a six year old child's remarks were egocentric and 62 per cent were socialized. Whereas Piaget made records on the verbal U lO behavior of two children during free play, McCarthy (1930) in her repetition of his work stimulated the child's speech by the use of pictures of animals, illustrated Mother Goose rhymes, and toys consisting of an auto, a cat, a telephone, a mouse, a music box, and a small ball. There were 140 children in McCarthy's study ranging in age from 18 months to 54 months. McCarthy reported that only 3.6 per cent of the vocalizations of these children were egocentric in nature. One of the major differences between McCarthy's and Piaget's studies is that McCarthy's children talked with an adult. In Piaget's original work two children talked to each other. In her review of the work done on language develOp- ment of children, McCarthy (1954) found that many of the discrepancies in the various studies of the function of language in children can be accounted for in terms of (1) differences in definition and interpretation of terms, (2) the situation in which the responses are recorded, and (3) individual differences of the children being observed. Reviewing the same literature on the function of language in children, Irwin (1960) concluded that the quan— titative discrepancies in the per cent of egocentric versus socialized speech were not due to the method of data collec- tion but to the manner of categorizing the data. That is, he did not believe that the method of data collection or 11 the situation in which the data are collected was as impor- tant as they way the data were ultimately classified. He was more concerned with how an investigator defined egocen- tric or socialized speech. McCarthy would certainly agree with Irwin. She stated that differences in definition and interpretations of terms played a big part in studies dealing with the func- tion of language in children. However, she did find in her 1954 review of the literature that the situation in which the language responses were recorded did have an effect upon the results. For example, children used longer sentences in conversation with adults than they used in conversation with other children. Medorah Smith (1935) found that chil- dren used longer sentences and asked more questions when talking with adults. This writer agrees with McCarthy's three points. For this particular study a distinction between egocentric and socialized speech is not of particular importance be- cause of the nature of the situation in which the data were collected. Communication is socialized speech (Cherry, 1966). The speaker, in order to communicate must consider the position of the listener. The situation in which the language was recorded was controlled. All children were talking with their mothers in a home setting while perform- ing four selected activities. Therefore, one can conclude that the majority of the communications were socialized. 12 Influence of Social Class Although it is recognized that differences, e.g., sex and age, might be important for language development, only differences due to social class membership were con— sidered in this study. There appear to be many class and cultural varia- tions in the function and values attached to language. For example, Bernstein (1966) prOposed a social class theory of communication in which he defined two codes of communi- cation--restricted and elaborated. The first code empha- sizes social relations and the second emphasizes informa- tion and Opinion exchange. In order for an elaborated code to develop a speaker must consider the listener. The speaker is forced to expand and elaborate his meanings so that he will be understood. A person with only a restricted code cannot or does not consider the listener. Berstein noted that the middle class used both codes whereas the lower class used only the restricted code. Berstein stated that as a child progresses through school it is important that he possess, or at least be ori- ented toward, an elaborated code if he is to succeed. He stated that: "The learning generated by these systems is quite different, whether it be social, intellectual, or affective. From a developmental perspective, an elaborated code user comes to perceive language as a set of theoretical 13 possibilities available for the transmission of unique ex- perience." (Bernstein, 1966, p. 438) According to this theory language structures and conditions what the child learns and how he learns, setting limits within which future learning may take place. The elaborated code is very important for learning. In it com- munication is specific to a particular situation, topic or person. It permits experience of a wide and complex range of thought. If this theory is correct then the lower class child who uses the restricted code exclusively is doomed to failure in our school system. With this in mind Hess and Shipman (1965) have re- ported a series of studies testing derivations from Bern- stein's theory. They have compared behavior of Negro mothers and children of various social classes and communi- cation between them in experimental situations. They used 160 mothers with 4-year old children selected from four different social status levels. The groups were upper middle class, upper lower class, lower lower class, and ADC mothers. The mothers were interviewed in their homes and later brought to the university ". . . for an interaction session between mother and child in which the mother was taught three simple tasks by the staff member and then asked to teach these tasks to the child." (Hess and Shipman, 1965, p. 874) They found that middle and lower class mothers differed little in the affective —— 14 elements of their interaction with their children. The dif- ferences appeared in the verbal and cognitive environments that they presented. The middle class mothers talked more with their children. Middle class mothers gave more ex- plicit information about a task to their children and they used more abstract words with their children than did the lower class mothers. Marge (1965) also found that home background played an influential part in the development of oral communication skills in children. He used 143 11—year old white subjects, their parents, and their teachers. The children were meas- ured on 40 speech and language factors by classroom teachers and speech specialists. Marge found that permissive mothers have children who achieve high scores on language maturity. Although he had data on the socioeconomic status of each family he did not analyze social class differences. All of his data about parent-child relations were collected by means of questionnaires. This writer believes that the study would have been more informative if the child and his parent had been observed together. In a study of verbal performance of 30 middle-aand lower-class males (ages 9-12) Schwebel (1966) found that middle class children described more ideas and events, used longer and better developed sentences, gave more accurate descriptions, and grouped objects more effectively. The children were compared on four standard tasks: (1) Picture 15 Description Task, (2) Events of the Day, (3) Sentence Con— struction Task, and (4) Grouping of Objects Task. Schwebel concluded that the inadequately developed vocabularies of the lower class children resulted from a lack of parental stimulation. This conclusion does seem to be consistent with Bernstein's theory and Hess and Shipman's findings. After analysis of data gathered from the partici- pants in a natural disaster Schatzman and Strauss (1955) concluded that one of the differences between the communica- tion of the middle and lower class was that the person be— longing to the lower class in answering questions could not take the listener's role. Their explanations were not ade- quate enough to make their meaning clear to a listener. In other words, the lower class person fails to communicate because he does not or cannot adequately consider the lis- tener. In a review of the literature on the language devel- opment in socially disadvantaged children Raph (1965) found that the process of language acquisition for socially dis— advantaged children, in contrast to that of middle class children, is more subject to: (a) a paucity of experiences in conversations with more verbally mature adults in the first three or four years of life, (b) severe limitations in the opportunities to develop mature cognitive behavior, and (c) the types of emotional encounters which result in the restrictions of the children's conceptual and verbal I l.- 'ii- 16 skills. As did Berstein, Raph found that disadvantaged children have a meagerness of quantity and quality of verbal expression which serves to depress intellectual functioning as the children grow older. It does appear from the literature that there are class differences in language ability. It was expected that the coding procedure used in this study would discriminate between the two socioeconomic classes used in the study. Coding of Communication "Language is specifically a mode of interaction of at least two beings, a speaker and hearer. . . ." (John Dewey, 1925, p. 185) The problem is how to code or cate- gorize communicative acts so that the interactive nature of language is considered. One of the first social scientists to code inter- personal behavior was Bales (1950). He developed an inter- action process analysis system. In his system all spoken or gestured acts are placed in one of twelve categories. This method uses on-the-spot coding. Using a social exchange model, Longabaugh (1963) divided each communication message into (1) resources sa— lient in the interaction and (2) modes of transmission. Resources are information, support, and control. Modes are seeking, offering, depriving, accepting, and not ac- cepting. Longabaugh also used on—the-spot coding. l7 Observers were required to make judgments as to whether any bit of behavior was a social act. He stated that, "A social act was Operationally defined as any coherent bit of behavior by an actor which was judged by the observer to have the in- tent of acting as a stimulus for a response by an present other." (Longabaugh, 1963, p. 329) Longabaugh attempted to validate his category system in a pilot study using 51 mother-child dyads. The children ranged in age from 5 to 12 years of age. The dyads were ob- served in a waiting room from behind a one-way mirror. The mother was asked to fill out a short questionnaire in which she might or might not seek information from her child. The interaction of mother and child was coded while the mother filled out the questionnaire. The dyad was observed for five minutes after the mother had completed the questionnaire. The reliability of two observers in coding the inter- action was found to range from as low as 38 per cent to as high as 92 per cent. The two observers had simultaneously coded 49 of the 51 experimental sessions. The median per- centage agreement for the 49 sessions was 60 per cent. Magrabi g£_al. (1967) adapted Longabaugh's model by modifying the categories of salient resources to make them more appropriate to the study go family activity pate terns within the home. The salient resources in this sys- tem are fact, preference, direction or command, and motiva- tion or encouragement. The Magrabi coding procedure was 18 designed to code only verbal behavior collected by use of a tape recorder. It does not use on-the-spot coding. Smith (1968) in attempting to identify the communica- tion patterns of mothers used the Magrabi method to code the communication of mothers interacting with their preschool children. She related various situational variables, such as (1) number of community moves since marriage, (2) amount of time spent by mothers in household tasks, (3) amount of help with household duties, (4) mother's educational level, (5) father's daily hours away from home due to employment, (6) family income, and (7) frequency of eating dinner to- gether, to the mother's communication patterns. She found that the situations or factors most frequently related to the communication patterns were: (1) family size and com- position, (2) number of community moves since marriage, (3) amount of time spent by the mother performing household chores, and (4) amount of help with household duties. Smith used two graduate students to code the com- munication of the mothers. The agreement of the two coders using this system to code a five-minute sample of communica- tion was 85.6 per cent. Summary Language is not only important for communication, absorption, and interpretation of the environment, but it also reflects highly acculturated styles of thought and 19 ideational modes for solving and not solving problems. As one studies the background influence on qualitative vari- ables in language and language development one also studies the effects of the same influence on cognitive development and problem solving styles and abilities (Deutsch, 1965). A review of the literature has shown that social class has a great influence upon the language development of the child. Language itself has been shown to be related to the intellectual development of children. One function of language is that of communication. As humans we have a need to share our thoughts with our fellow man. Communication was described as a process of interaction that included at least two persons--a speaker and a listener--in which the speaker tries to influence the behavior of the listener. The situation in which the language of the child is recorded has been shown to be a vital factor in explaining results of language studies. Methods of analyzing or coding language that take into account the interactive nature of language were re- viewed. This method must be flexible enough to handle data that are gathered under many situations, such as the natural setting of the home. Since the literature has emphasized the importance of the influence of the parent upon the lan- guage development of the child, the method of coding used in this study had to be able to analyze the language of adults as well as the language of children. ' A l I , . CHAPTER III METHODS AND PROCEDURES Communication is described as one of the main func- tions of language. Communication as defined by this study is an interaction process in which the person speaking transmits a message to the receiver with the conscious at- tempt to affect the latter's behavior. It was assumed that verbal communication can be meaningfully differentiated into categories for analysis. The hypothesis upon which this study was based is that communication patterns are learned by imitation of a model. Several comparisons were made of the various sub— jects. Selection and Description of Sample The data analyzed in this study were collected in a Michigan State University Agricultural Experiment Station Project 786, "Use of Space, Material and Communication Among Family Members Performing Family Activities." Data for Project 786 were obtained by observing a mother and preschool child in each of ten urban families in the Lansing area performing four selected activities. 20 21 The sample was drawn from a group of thirty families with children enrolled in the Laboratory Preschool at Mich- igan State University. Five of the families were from a lower socioeconomic class and five were from the middle socioeconomic level. Data for one of the lower socioeco- nomic class families are incomplete because of a death in the family, therefore only nine families were included in this study. The socioeconomic levels of the families were indi- cated by the educational attainment of the parents and by the family's annual income. The income for the middle class families ranged from $8500 to over $15000. The lower socio- economic group had incomes ranging from $1500 to $6499 (Table 1). All of the middle class wives and husbands had some college education. Two of the husbands had Ph.D. degrees, one was a candidate for the doctorate, one had a B.A. degree, and one had some college. One of the middle class mothers was a candidate for a master's degree, one had done some graduate work, one had a B.A. degree, and two had some col- lege experience (Table 2). Of the four lower socioeconomic level families only one was a complete family in the sense that both husband and wife were present. The father in this one family had some high school education. One of the lower socioeconomic level mothers had some college, two were high school 22 .ucmflmflomu Dada .nmms pxm: new mmv.ma on oom.HHw msoocfl ooummfloflucmm name.v mmv.o mme.m mmv.m um>o mme.m ame.ma ammq.o mmv.oa ummw Ham CD on on on cam on on on on oom.mm oom.mw oom.mm oom.aw oom.mam oom.mm oom.HHm oom.wm oom.m w mEoocH m m s G m e m m a mmmHU Hm3oq mmmHU OHUUHE HBenz aoflumUAMHuamoH waflsmm MHflEmm “mm mEoocH Hmsccdll.a magma 23 Table 2.--Distribution of Families According to Level of Education of Parents Level of Education H o m .1: O - (D 0‘ 4:0) H U) 4.) -H ()D m 0 - m mra mtu o1 rim Lam H o s m a)m a)m o :4 (1)0 ,C'U (DH .CH JJH JJ (U 8:: UHU era 0!» wt» 0 4J or) -H¥4 o<3 m m mmlcoo so mcoflm ._ Inoo GeoumlcmEnmmmm tam Imom unom map you muoanSm , .ucmfloammmoo GOHDMHGHHOU Ham How mafia: mmMmmmE Hmuoe .mmflnommumo mmonod _ .HdOm “ cam oBu meowmmmm on Umumm .coflu IEOO mmunu cam wco meHmmwm nomnuoo aeoumlcmsummmm Mom mwooo ma may :0 muommndm .ucmHOAMMoOO coflumamunoo Ham How muflcs mommmmfi Hmuoe .chHmmmm mmouo¢ NUflHflQMHHmm .mmooo Ucm .mflnm uanEmE >HHEMM .Hm>ma OHEOG loomoHUOm cmmeumn scepomumucH .mmcoo cam magmHmQEmE MHHEMM cmmzumn coauomnmucH % .mmooo Ucm Hm>ma UHEocoom Iofloom cmmsbmn coauomnmucH .mmpoo cmm3umn wocmHmMMflc .mflcmumn IEmE waflemm can Hm>ma OHEOG loomoflUOm cmm3umn cowuomnmucH .mflnmumneme MHHEMM cmm3umn mocmummmao .mam>ma oHEocoom IOHUOm cmmBumn mocwumMMflo .pmmulm 6cm .mmfinomwumo mousommnlmooe ma . . mommaum> mo mfimwamce so muoanSm HON mpflcs mmMmmmz mocmofluflqmfim mo umwa poms oaumflumum mammamcd cw poms mumo mammamce mo mmomndm mama mo mwmhamc< Cw womb mwonuwzll.m magma nu .mumcuofi “amen mo mcnmpumm coapmoflcde IEoo mSOHHm> on mmmcommmu m.cmnmaflno mo mmsoum 038 .mmfluom Iwumo moHDOmmu HSOM we» so cmHUHHco mo masonm 038 .mmOHDOmmH usom may :0 maano mam Hmnpoz .mmmoE Hmom men so cmumaflno mo museum 039 .mmmoe usom men so UHHSO mam Hmcpoz .mmfluommumo we we» .mmmoE Hmom map mam .mwfluommumo so commaflzo mo mmsoum 038 Mu mOHDOmmH snow map .mmauom .mmfluommumo ma umumo moHSOmmulmmoE ma one no map so maflno mam Hmzuoz .mmmmucmonmm muomflQSm Ham How means mommmmz mGOmHHmmEou .mmfinom .ummulm mam Imumo mOHSOmwulmmoE ma co .Hm>ma UHEocoom mocmflum> mo mflmhamcd muomeSm Ham How muflss mmmmmmz |0fl00m cmmsumn mocmHmMMfln Nuwmflam> mmmb oaumflumum mammamcd ca moms mama mammamcd mo mmomnsm .emscfiucoouu.m magma CHAPTER IV PRESENTATION AND DISCUSSION OF RESULTS This chapter is divided into several sections: (1) reliability, (2) validity, (3) identification of communica- tion patterns, (4) relationships tested by the analysis of variance, and (5) limiations of the study. Reliability Table 7 shows that the reliability coefficients for sessions were as high for children as they were for mothers. These coefficients indicate that the measurements were sta- ble across sessions. Table 7 also shows that the reliability of the code was as high for children as for mothers when the Spearman-Brown correction was used. However, the non- corrected coefficient was not significantly different from a coefficient of zero for the mothers. Table 7.--Re1iability Between Observation Sessions and Code Categories Session Reliability Code Reliability Not Not _ N Corrected Corrected N Corrected Corrected * Mothers 9 .79a .89a 9 .43NS .6051 Children 9 .823 .909 9 .4810 .65a ap < .01. bp < .05. 35 Validity Table 8 shows that the socioeconomic levels differ significantly (p < .001) in the amount communicated. In addition, Table 9 shows that this difference is in the same direction as reported by Hess and Shipman (1965), i.e. the middle class communicates more than does the lower class. Moreover, the statistically significant interaction of socio- economic level and the codes (Table 8) confirms other results found by Hess and Shipman (1965), i.e. the middle class of- fered fact proportionately more than the lower class (Table 10). Identification of Communication Patterns Several comparisons were proposed in Chapter I. They are based on percentages so that each comparison can be made across socioeconomic levels and within family groups. Comparison I. The first comparison was of the mother and child on the sixteen mode-resource categories utilized in the Magrabi §t_al. coding procedure. Table 11 shows this comparison. Comparison II. This comparison is of the two groups <>f children on the 16 categories. This comparison is also contained in Table 11. As we can see from Table 11 not all of the categories are used by these families. The raw data on the mode- resource categories does show that all categories are used I L _ . u; .NH.G n Ammo. v moo .mm.m u lfioo. v men .om.m u Aaoo. v new vmm.vma on Aaflcuflz Honnmv msoum cflnufl3 .mQSm x mmfluommumo moHDOmmHImUOZ .1 szmm. Geo.moa ma mmfluommumo mousommnlmmoe x magnumnEmE maflsmm x Hm>ma UHEocoomoHoom cmom.o som.omo.a ma mmfluommumo mousOmmu Immoe x aficmHmQEmE kHHEmm ommm.b mmh.mmara ma mmfluommumu mowsommu ImmoE x Hm>ma Ufleocoomofloom nmmm.om Hmm.eon.m ma mmfluommumo GOHDOmmHImUOZ 6 3 osm muomflnsm queues mmm.mam ea Acmmzpmn Houumv dsoum canvas .flnsm mszN. mom.mm H mflcmumnEmE haflEmm x Hm>ma oHEocoomoHoom ammm.m oom.oma.m H mflnmumnsws saflsmm man.NH Hon.mmm.m H Hm>ma anocoomofloom NH mpomflnmm cmmsumm m m: MU COHumHHm> mo mousom mUGmHHm> mo mwmhamcd mo mumfifimmnl.m magma 37 Table 9.--Total Number of Message Units of Mothers and Children for Each Socioeconomic Level Socioeconomic . Number of Level Mothers Children Total Subjects Middle 1478 881 2359 10 Lower 649 298 947 8 TOTAL 2127 1179 3306 18 by someone except seeking motivation and not accepting motivation (Appendix I). Table 11 shows that there were differences in the categories utilized by the families. The children of both groups offer fact proportionately more than their mothers. The lower socioeconomic children used the category of not accepting direction proportionately more than the middle socioeconomic level children. Mothers offered direction proportionately more than the children. The lower socioeconomic status mothers offer direction proportionately more than the middle class mothers. The middle socioeconomic status mothers offer fact propor- tionately more often than the lower socioeconomic mothers. Comparisons III and IV. Comparison was made of the another and child on the four modes of offering, seeking, accepting, and not accepting. Comparison IV was made of tile two groups of children on the four modes. Table 12 contains comparisons III and IV. 38 comm o mv 0H mv N mm Hm MNH c we mm va OOH vmm Hm mvNH HmuOB Edm hwm o mm.w mm. oo.N 0 mm. mm. mh.H o om. oo.m mN.mN mh.H mm.mv mh.m mN.mm HOBOA mmmm o ow.H on. om.m om. om.v om.H om.OH o om.m om.m oo.mm om.m om.mm om.v om.mm OHUUHE HMMME S O m .m 2 Q m m 2 Q m .m E Q m .m HO>QH UZHBmMUUHN BOZ UZHBQMUU< UZHMmHmm GZHmmmmO OflEOGOOO IoHoom whommumo mousommmlmcoz whommumu mUHDOmmHImmoz comm How Hm>mH OHEocoomOHoom comm How muHCD mommmmz mo HmQEdz cmmz||.0H mHQmB 39 Table 11.—-Percentage of Message Units in the Mode—Resource Categories for Mothers and Children in Each Socioeconomic Group Mode—Resource Categories OFFERING SEEKING ACCEPTING ACC%%%1NG F P D M F P D M F P D M F P D M Total Middle Class Mothers 39 1 29 5 l4 4 l 0 4 O l O 1 0 O O 1478 Children 46 4 18 1 l4 1 2 0 5 1 3 0 l 0 l O 881 Lower Class Mothers 20 l 48 2 21 2 0 0 2 O 0 0 2 0 1 0 649 Children 45 6 19 0 15 1 1 0 l 0 2 O l O 8 O 298 Table 12.--Percentage of Message Units in Each Mode for Mothers and Children of Each Socioeconomic Level Modes OFFERING SEEKING ACCEPTING NOT Total ACCEPTING Middle Class Mothers 73 19 6 2 1478 Children 70 17 10 2 881 Lower Class Mothers 72 23 2 4 649 Children 70 17 3 10 298 40 Offering and seeking are the most frequently used modes for all groups. The middle Class children used the accepting category proprotionately more often than the mothers Of both groups and the children of the lower socio- economic group. The lower socioeconomic Children used the not accepting mode proportionately more Often than any of the other three groups of subjects. Comparisons V and VI. These comparisons are of the four groups of subjects on the four salient resources-- fact, preference, direction, and motivation. Table 13 contains these comparisons. Table l3.--Percentage Of Message Units in each Resource for Mothers and Children of Each Socioeconomic Level Resources Fact Preference Director Motivation Total Middle Class Mothers- 58 5 31 Children 67 7 25 Lower Class Mothers 44 3 50 2 649 (Shildren 62 7 30 0 298 1478 881 —— All of the salient resources are utilized by the :subjects. Motivation and preference were not used very