Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The role of family environment in an ecological study of preschool children attending family day care Shapiro, Ellen Sara 1988

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1988_A8 S52.pdf [ 5.58MB ]
Metadata
JSON: 831-1.0053747.json
JSON-LD: 831-1.0053747-ld.json
RDF/XML (Pretty): 831-1.0053747-rdf.xml
RDF/JSON: 831-1.0053747-rdf.json
Turtle: 831-1.0053747-turtle.txt
N-Triples: 831-1.0053747-rdf-ntriples.txt
Original Record: 831-1.0053747-source.json
Full Text
831-1.0053747-fulltext.txt
Citation
831-1.0053747.ris

Full Text

THE ROLE OF FAMILY ENVIRONMENT IN AN ECOLOGICAL STUDY OF PRESCHOOL CHILDREN ATTENDING FAMILY DAY CARE By ELLEN SARA SHAPIRO B.A., The University of British Columbia, 1970 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES (Department of Counselling Psychology)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA August, 1988 © Ellen Sara Shapiro, 1988  In  presenting  degree freely  at  this  the  available  copying  of  department publication  of  in  partial  fulfilment  University  of  British  Columbia,  for  this or  thesis  reference  thesis by  this  for  his thesis  and  scholarly  or for  her  of  T h e U n i v e r s i t y o f British 1956 Main Mall Vancouver, Canada V6T  1Y3  Columbia  I  I  further  purposes  gain  the  shall  requirements  agree  that  agree  may  representatives.  financial  permission.  Department  study.  of  It not  be  that  the  Library  permission  granted  is  by  understood be  for  allowed  an  advanced  shall for  the that  without  head  make  it  extensive of  my  copying  or  my  written  ii  ABSTRACT This thesis is an exploratory ecological study of the role of the family environment as it relates to a number of variables under investigation in the Vancouver Day Care Research Project's extensive contextual study of children enrolled in family day care settings. These variables include child language scores, indices of socio-economic status, conditions of maternal employment and attitudes related to maternal employment, other measures of the home environment, and parental attitudes to childrearing. In addition, the family environments of the family day care caregivers were examined in relation to the quality of care provided. The Moos Family Environment Scale (1986) was administered to parents and caregivers enrolled in the study. Scores from its ten subscales were correlated with measures of the variables of interest and then tested for significance. Data was then analyzed for important trends, patterns and highlights. Results showed that exposure of family members to stimulating ideas and activities is facilitative of child language skills, while an emphasis on achievement seems to have a negative effect. Families from higher socio-economic status homes seemed to be more likely to provide these opportunities for their children, particularly if they are well-educated. Findings also indicate that mothers who are satisfied with their employment tend to provide more positive family environments for their children than those who are working reluctantly. Mothers who worked part-time also appeared to provide better family environments than did those who experienced the increased stress of full-time employment. Adultcentered parenting values which stressed obedience were associated with family environments which were less facilitative of child cognitive development, whereas homes with child-centered parenting values appeared to be more positive. Family day care caregivers who provided superior childcare were found to be more organized in their own families, more supportive of one another, and more able to allow their family  iii  members to function independently than were other caregivers.  There was  considerable overlap in the results for each variable of interest; many similar features were found in the environments which were considered optimal in terms of language development, socio-economic factors, conditions of maternal employment, attitudes to childrearing, and high quality care for children. The study results strongly support the importance of exposure to a wide range of intellectual and cultural stimuli, participation in activities outside the home, expression of feelings amongst family members, and well-organized family functioning in the creation of optimal family environments; an emphasis on achievement, and the use of rigid rules and doctrine were found to be deleterious to the creation of positive home environments.  iv  TABLE OF CONTENTS  ABSTRACT TABLE OF CONTENTS LIST OF TABLES ACKNOWLEDGEMENTS CHAPTER ONE  ••  INTRODUCTION Purpose of Study Conceptual Framework Nature of the Problem Central Research Question Definition of Terms Auxiliary Research Questions  ai v vii viii  1 2 3 5 6 6 8  TWO  REVIEW OF THE LITERATURE Development of the Ecological Model Ecology of Day Care Ecology of Family Environment Family Environment and Ecology of Day Care Family Environment and Child Language Development .. Family Environment and Socio-Economic Status Family Environment and Maternal Employment Summary  9 9 16 19 19 21 32 34 38  THREE  METHODOLOGY  43  Design of the Study Sample, Sampling Procedure, Population Standardized Instruments  43 43 45  V  Research Procedure Data Collection Data Processing and Analysis Limitations  51 51 52 55  FOUR  RESULTS Descriptive Information FES Subscale Intercorrelations Findings related to child language scores Findings related to indices of socio-economic status Findings related to conditions of maternal employment Findings related to HOME scores Findings related to childcare attitudes and beliefs Findings related to family environment of caregiver Summary  56 56 57 58 60 64 70 73 77 79  FIVE  SUMMARY AND DISCUSSION Summary Discussion Recommendations  81 81 87 94  REFERENCES APPENDIX  \  96 102  vi  LIST OF TABLES TABLE 1 Means and Standard Deviations of FES scores in Parent & Caregiver Total Groups  Page  56  2 FES Subscale Intercorrelations for all Parents: Pearson Correlation Coefficients  57  3 Pearson Correlation Coefficients for Parent FES Scores and Child Language Scores 4 Pearson Correlation Coefficients for Parent FES Scores and SES Variables ..  58 61  5 Pearson Correlation Coefficients for Parental FES Scores and Attitudes to Working Women  65  6 Pearson Correlation Coefficients for Parental FES scores and Maternal Employment Satisfaction, Work-Childcare Tension, and Hours of Work  66  7 Pearson Correlation Coefficients for Parent FES scores and Parent HOME Scale Scores  71  8 Pearson Correlation Coefficients for Parental FES Scores and Attitudes to Childrearing  74  9 Pearson Correlation Coefficients for Family Environments of Caregivers and Quality of Care  78  vii  ACKNOWLEDGEMENTS I would like to thank the members of my thesis committee for their availability and assistance, as well as their ongoing interest in this research. My appreciation goes to> Dr. John Friesen for his encouragement, time, and invaluable guidance as chairperson of the committee; to Dr. Hillel Goelman for his enthusiasm and expertise, as well as his exceptional support throughout this research project; to Dr. Walter Boldt for his extensive consultative role in the methodological aspects of the study. My special thanks also go to the staff of the Vancouver Day Care Research Project: notably to Warren Weir, for his commitment to this research and tireless assistance in the data analysis, and to Robin White, for her valued contribution to the data collection process. In addition, I would like to thank my family and friends for their support and encouragement throughout this project; I particularly thank my daughter Sarah for her patience, understanding, and independence, which have enabled me to pursue this endeavour.  1  CHAPTER ONE INTRODUCTION  Increasingly, child development is viewed in a context that extends beyond the individual to a larger societal perspective.  Psychologists, social workers,  pediatricians, and educators are looking beyond the narrow scope of their own field to see broader and more complex forces at work in shaping an individual than had previously been supposed. Research in human development has been marked by an expanding interest in the role played by the environment. A growing debate regarding the relative import of genetic and environmental influences on an individual's development has characterized much of this research. Whatever weight a particular theorist might ascribe to each of these factors, Lidz (1970) points out that the family is undoubtedly the mediating unit between the child's unique genetic endowment and the cultural heritage of society. The tremendous influence of the family unit is reflected in a growing interest in viewing human development in the context of the family. This trend has become particularly marked in the social climate of modern times, which has seen vast transformations in the family unit.  The prototypic nuclear family which has  characterized North American culture is becoming increasingly less prevalent. The rising divorce rate has led to a new concept of family wherein large numbers of children live in homes headed by a single parent, with step parents, or in a wide range of 'shared custody' arrangements. Traditional male-female roles have been radically altered. Women are much less likely to be financially dependent on men, and are entering the workplace in record numbers. Men are becoming increasingly involved in household and childcare responsibilites.  As Keniston (1977) notes, "we live in a  2  society where parents must increasingly rely on others for help and support in raising their children" (p. 12). As a result, children are much more likely to enter group settings outside their families in their early years. The effects of these changes are still very much under investigation, but there is no doubt as to their powerful impact on the developing child. The family is increasingly seen to be deeply affected by the social context in which it operates. According to Powell, "the current resurgence of interest in familial influences on child behaviour and development has been accompanied by strong concern for the social and economic conditions under which families carry out their child-rearing functions" (1979, p. 1). A socio-ecological study of human development takes such an approach, based on a concept which Powell refers to as "social embeddedness....The child is seen as embedded in a family system which is in turn embedded in a society. Interactions between these interconnected systems are viewed as having a critical influence on human development" (1979, p. 2). Bronfenbrenner, who has developed a model for the ecological study of development, suggests that the family be studied as a small group nested within larger social structures, and that the family unit be considered to be one of many systems that are important for the developing child (1979).  Purpose of this Study It is within the framework of this contextual understanding of human development that the Vancouver Day Care Research Project (Goelman & Pence, 1986) embarked on an extensive ecological study of preschoolers in their families and the family day care settings in which they were enrolled. While focusing to a large extent on language development, the project is particularly concerned with the ways in which different factors in the child's life interact with and influence each other. It  3  examines individual characteristics of the child in the context of family, day care, and the larger social world affecting the parents. This thesis intends to contribute to an understanding of this complex picture by exploring the inter-relationships between family environment and a select number of variables including: child language scores, socio-economic status, conditions of maternal employment, childcare attitudes and beliefs in the child's family, and the quality of care offered by the family day care home.  Conceptual Framework Central to a socio-ecological approach to the study of family functioning and human development is a systems orientation, which is based on: the notion that elements within a bounded unit of some type, whether it is a family, a corporation, or a nation-state, are interdependent. The activities of one cannot help but have a direct or indirect influence on other elements in the system, and through some feedback loop, on the first unit itself at a later point in time. (Conger, 1981, p. 201) Powell (1979, p. 2) refers to the family as a complex, open system, responsive to and dependent upon its environment. As such, the family is seen as "the primary social context of parent-child interaction and child development, with several layers of interdependent social and environmental influences interacting with the family system" (Powell, p. 3). In this context the family is not a static entity, but an evolving organism which both influences and is influenced by the environment. A systems approach attempts, according to Conger, (1981) to "incorporate information concerning the individual, the family, and outside social factors into an interactive framework" (p. 202). Minuchin (1970), in developing an orientation for studying and treating disturbed children, similarly proposed that the child, the child's  4  environment, and the "linkages between these two elements" (p. 42) be studied interdependently. These concepts are central to the ecological model. In setting forth this model, Bronfenbrenner (1979) has argued that studies of children should regard the family unit as one of many systems that are relevant to the developing child. To this end he has constructed a comprehensive theoretical schema, which posits four systems or social structures which touch the developing person; these are nested inside each other. At the core is the microsystem which is the immediate setting containing the individual.  The mesosystem comprises  "interrelations among two or more settings in which the developing person actively participates" (1979, p. 25). The exosystem involves settings in which the individual does not actively participate but which nevertheless are important to that individual. The macrosystem refers to overriding beliefs and values of the larger culture in which the individual lives. Critical to Bronfenbrenner's theory is the concept of reciprocity, or a mutual accomodation between the person and the environment, which is present in each of the four systems, and which also exists between the systems. Based on this model, ecological research does not focus on hypothesis-testing but on discovery. "Instead of trying to focus on a single variable and 'control out' all others, ecological research seeks to 'control in' as many theoretically relevant ecological contrasts as possible" (Bronfenbrenner, 1979, p. 38). This approach is critical to the design of the Vancouver Day Care Research Project, (Goelman & Pence, 1986) which focused on the world of the individual child and the immediate settings of influence: notably family and day care setting. In addition to the study of the child's microsystem, this project was also designed to study the mesosystem, or the reciprocal influences between these principle settings directly affecting the child. Arenas in the child's exosystem, in which the child is not involved directly but which undoubtedly affect him, are studied; notable here is parental employment. Beliefs and attitudes of parents and caregivers, derived from  5  the macrosystem, are a further focus. None of these variables is seen in isolation; instead the complex interactions between them are the focus of the study. The intent is to develop a comprehensive picture of the rich fabric that contributes to a child's development in the context of a larger social world. In this sense the study follows Bronfenbrenner's recommendation that "in ecological research the properties of the person and of the environment, the structure of the environmental setting, and the processes taking place within and between them must be viewed as interdependent and analyzed within systems terms" (1979, p. 41).  Nature of the Problem Comprehensive ecological studies delving into the young child in his world are a relatively new phenomenon. The complex role played by the family in such a context is receiving growing attention. As Powell notes: Increasingly there are suggestions that the socio-ecological contexts of family functioning need to be given serious consideration in the design of research pertaining to the development of young children. The interplay between the social environment and family childrearing processes is a relatively unchartered domain. (1979, p. 1) The concept of the family environment, or the realm of interpersonal and social relationships in a family that combine to create its unique social climate, has attracted increasing interest in research. It has not, however, been explored in the context of an ecological model, where its reciprocal influence on the child, other aspects of family functioning, social settings affecting the child, and larger social processes impacting on the family, can all be studied. Bronfenbrenner (1986), and Belsky (1978), have particularly noted the importance of the connections between family and day care settings in the young child's life.  The family environment, including the domain of  6  social relationships, is certainly an element that underlies all other aspects of family life and is a mediating influence in other settings and structures touching the family. In addition, no research has examined the family environment of the caregivers themselves in family day care homes; certainly this would be relevant to the care provided. A clearer understanding of the role of the family environment will undoubtedly shed further light on the complex nature of the familial factors which contribute to a child's development. In addition, such knowledge will supply one more piece of the puzzle in determining the complex interactions between individual, family dynamics, and social settings which are the focus of ecological research.  Central Research Question HOW DOES THE FAMILY ENVIRONMENT INTERACT WITH CHILD LANGUAGE  SCORES,  MEASURES  OF SOCIO-ECONOMIC  STATUS,  CONDITIONS OF MATERNAL EMPLOYMENT, PARENTAL CHILDCARE ATTITUDES, OTHER MEASURES OF HOME ENVIRONMENT, QUALITY  OF FAMILY DAY CARE  AND THE  IN AN ECOLOGICAL STUDY OF  PRESCHOOLERS ATTENDING FAMILY DAY CARE/PRESCHOOL SETTINGS?  Definition of Terms Operational definitions for the purpose of this study: Family Environment: Social-environmental characteristics of families: specifically Cohesion, Expressiveness, Conflict, Independence, Achievement Orientation, Intellectual-Cultural Orientation, ActiveRecreational Orientation, Moral Religious Emphasis, Organization and Control as measured by the Moos Family Environment Scale (Moos, 1986).  7  Child Language Scores: as measured by Peabody Verbal Picture Test (Dunn, 1979) and Expressive One-Word Picture Vocabulary Test (Gardner, 1979). Socio-economic Status: as determined by annual family income, parental education, parental occupational status, and family resources. Conditions of Maternal Employment: work satisfaction, hours of work and attitudes to work. Parental Childcare Attitudes: as measured by the Vancouver Day Care Research Project Parent questionnaire. Other Measures of Home Environment: scores from Bradley and Caldwell's Home Observation for the. Measure of the Environment (HOME) Inventory (Caldwell & Bradley, 1979). Quality of Family Day Care Setting: as measured by Day Care Home Environment Rating scale, (Harms and Clifford, 1980), a quality of care variable derived from factor analysis, and the Caregiver HOME scale scores. Ecological Study: Research design which focuses on reciprocal interactive influences within and between major settings and social structures of import to the developing child. Preschoolers: Children aged 3-5. Family Day Care: Refers to provision of day care in a home environment. Subjects are enrolled in licensed and unlicensed day care settings in the Lower Mainland. Licensed Family Day Care: Where licensing of a family day care setting has been determined by the B.C. Provincial Community Care Facilities Licensing Board; this setting can include up to five preschool children including the caregiver's own chidren. Unlicensed Family Day Care: Refers to a family day care setting operating without a license from the B.C. Community Care Facilities Licensing Board.  8  Preschool: A comparison group is enrolled in U.B.C.'s Child Study Center preschool facility; this is a provincially licensed part-time early childhood education program run by trained staff.  Auxiliary Research Questions 1)  What aspects of family environment are associated with higher and lower child language scores?  2) What aspects of family environment are associated with socio-economic status as determined by parental education, income level and parental occupation? 3) How do parents' FES scores compare with the environmental measures of their homes obtained using the HOME Inventory? 4) Is there a relationship between family environment and conditions associated with maternal employment such as attitudes to working women, job satisfaction, hours of work and employment-related stress? 5) Is there a relationship between family environment and parental childrearing attitudes and beliefs? 6) Are any aspects of the caregiver's family environment associated with higher or lower quality of day care setting?  9  CHAPTER T W O REVIEW OF T H E LITERATURE  As noted in the introduction, it is the intention of this thesis to explore the role played by the family environment in the Vancouver Day Care Research Project's ecological study of preschoolers. This literature review will outline the development of ecological research as a methodology, the use of the ecological approach in studying children attending day care, and the ecology of the family environment as it pertains to the questions of interest delineated in Chapter One.  Development of the Ecological Model In their review, Bronfenbrenner and Crouter (1983), trace the growing interest in the role played by the environment in human development over the last century. The earlier part of the nineteenth century was marked by the influence of Charles Darwin and his theory of evolution. The biological determinism central to Darwin's theory was largely responsible for a marked emphasis on the genetic aspects of human development amongst theorists and researchers. An interest in the role of the environment did not appear until approximately 1870. At that time Francis Galton initiated the nature-nurture debate, when, as part of his famous research on the importance of heredity and individual differences, he published society's first research on twins. Although his work strongly emphasized the contribution of heredity, he introduced the concept of the environment as a potential mediating factor in human development, conceiving that "the environment becomes a portion of the variance that cannot be explained by genetic factors" (Bronfenbrenner & Crouter, 1983). Galton was the first to identify family structure as an important aspect of environmental  10  influence, looking at the possible impact of birth order and parenting in his subjects. He was also aware of the possible influence of larger social factors such as socioeconomic status. These contributions, while not comprising the bulk of his work, nevertheless laid the groundwork for much of the research on the role of the environment in development that was to take place over the next century. His twin studies provided an important model for much of this research. At about the same time as Galton was working, other research being done in Europe also began to take environmental factors into account. This included a number of studies which were based on what Bronfenbrenner and Crouter refer to as the 'social address model', or "the comparison of children living at differing social addresses" (1983, p. 361).  These studies operated on the assumption that the  location in which a child lived could have considerable impact on the matter under investigation.  Although an influential model and an important development  historically, it should be noted that early designs such as these were relatively simple; they observed differences in children from one setting to another and did not look at factors in the environment that could account for these differences. In America, G. Stanley Hall extended the social address model in an 1883 study which looked at cross-cultural differences between groups of children as well as differences in locale.  As Bronfenbrenner and Crouter (1983) point out, research  designs were becoming more complex, looking at social factors of relevance to children's lives as well as differences in geographic locations. By the end of the century, although many important aspects of the environment that might be significant to child development had been introduced, there was no comprehensive theory expanding on what this relationship might be. Studies which were investigating the role of environment in human development did not look at the processes involved, but remained essentially descriptive of differences in developmental outcomes between children from different geographic locations or social backgrounds.  11  The two trends which dominated studies of the relationship between environment and human development in the late nineteenth century persisted until approximately 1930. During the early twentieth century, the social address model was prominent, as was the nature-nurture debate. Terman continued in the tradition of Francis Galton in his extensive studies of individuals with superior abilities completed in 1925. He took factors linked to family background into account in his study, but concluded that the considerable differences in family variables attributed to gifted children were the result of heredity rather than environment. Other studies focused on a comparison between adopted vs. biological children. By 1930, environmental influences on development were being studied both in terms of the immediate setting (family) of the child, and broader social spheres of influence (such as ethnic background and socio-economic status). New trends included a growing interest in the effects of parent-child relationships on child behaviour and development.  The social-address model  expanded to include an interest in parents as well as children from differing locations, ethnic backgrounds or social classes. Of early interest were the sex of the child and the sex of the parent as important variables.  World War II brought early  investigations into the effects of father absence and maternal employment on child development.  The study of children in the context of groups also emerged as a  research interest. Studies relating parenting to family size and birth order did not appear until the 1950's. At about this time ideas about child development were being organized into theories which were beginning to have a major impact on research. Many important theories were considering the possible relationship between various aspects of the environment and the developing organism.  Freud's ideas about parent-child  relationships and family dynamics had an important influence and played a role in the new interest in parenting and the family in developmental research. Bowlby (1969)  12  further emphasized the tremendous impact of the mother-child relationship in child development in his work on the effects of attachment and bonding. Piaget's theory stressed the child's own impact on the environment. He put forward the notion of cognition as an "adaptive system....it's functioning allows the organism to adapt to its environment" (Lerner, 1976, p. 160). Learning theory, which was increasingly dominant in the psychological world, began to influence the arena of developmental psychology as well. In contrast to the work of Piaget, which saw the child as the architect of his learning experiences, learning theory focused on the role of the environment in shaping these experiences. Clark Hull was a major figure, not only in pioneering many of the concepts central to learning theory, such as contingent reinforcement, but also in his systemic approach. As Miller (1983) points out, Hull was interested in integrating principles of psychoanalysis, learning theory and sociology. In emphasizing cultural aspects of development as well as psychological, he made a significant contribution to environmental research. Furthermore he set the stage for a more interdiscipinary approach in studying human development. A major impact on developmental psychology was felt with the publication of Hunt's Intelligence and Experience in 1961. His work contradicted the arguments of the heredity position, particularly the notion of fixed intelligence.  Instead he put  forward an interaction point of view, which saw development as a function of the interaction between heredity and environment. His emphasis on the role of the environment, particularly of early experience, on the developing individual, was extremely influential. Research derived from biology also had an impact on developmental psychology. The naturalistic studies of Konrad Lorenz were particularly important. Lorenz was an ethnologist who saw "animals as active organisms within a particular ecological niche, not as passive organisms prodded by stimuli, as in the tradition of learning theory" (Miller, 1983, p. 304). Partially due to the influence of such models,  13  the child was being seen increasingly in a similar way, as "an active organism seeking stimulation, provoking responses from, altering and even creatings its own surroundings" (Bronfenbrenner & Crouter, 1983, p. 270). The growing field of sociology was also making its influence felt in the emerging theories of human development. An interest in the process of socialization was to be found in the writings of psychologists such as Freud, Erikson, Jung and Bandura. In the 1920's and 30's, sociologists associated with the Chicago school developed a theory of socialization related to an individual's interactions with others over a life-span. Baltes and Schaie later focused on the developmental contexts affecting human beings at different stages of their lives and developed a methodological framework for the study of what they have called 'lifespan developmental psychology' (Baltes & Schaie, 1973). This framework allowed sociologists such as Elder (1974) to pursue context-oriented studies of groups at various stages of their lives. Elder's longitudinal study of the impact of the depression on childrens' development at critical points in their lives used a good deal of psychological data and was important in its interdisciplinary approach. Research designs based on the social address model, or the comparison of children living in contrasting environments, began to increase in complexity in the 1960's. Bronfenbrenner and Crouter (1983, p. 374) note the emergence of the "personprocess-context" model in the late 1960's and early 1970's. This new approach noted not only differences between parents and children of different social or geographic locations, but also differences in the processes or linkages between them. The model addressed the possiblity of reciprocal influences between child and environment, took into account belief systems as important factors in childrearing practices, and considered the importance of development over time. Designs such as this enable researchers to "determine the environmental circumstances that best help individuals achieve their genetic potential" (Bronfenbrenner & Crouter, 1983, p. 395). An  14  example is the work of Tulkin (1973) who looked in depth at the phenomenon of social class as it is manifested in the family; his work was especially important as it looked at differences in social class in terms of process as well as outcome, and also considered the relevance of childrearing beliefs and attitudes. Studies which looked at children in their immediate settings of influence, or "microsystems" as Bronfenbrenner (1983, p. 378) refers to them became popular. Research in this tradition provided evidence to show that settings in which children developed had a major impact on them over time.  Both physical and social  environment were objects of study. An example is Bradley and Caldwell's (1975) study relating cognitive development to home environment in young children. Implications of such studies were that persons of influence in a child's important settings, such as parents and teachers, could have considerable impact on a child's development. Models based on what Bronfenbrenner (1983, p. 382) refers to as the mesosystem are more complex and assume that "two or more settings can simultaneously affect the behaviour of the child". Some of these studies took the form of parent vs. peer influence of children in various circumstances. More recently, studies have been undertaken to show the differences between children at home and in school or day care settings. Other studies looked at the developmental significance of the child's transition from one setting to another. Still others looked into linkages between settings and of the role such linkages might play in development. Research increased further in complexity when it began to take into account influences from the exosystem, or external settings which do not actually contain the developing person. Examples are the proliferation of studies on the effects of maternal employment on children. Elder's (1974) longitudinal study of children during the depression years is another example of exosystem influences on development. In this  15  case the impact comes from historical and social forces outside the child's immediate setting.  Children's lives are strongly affected by social change and research is  increasingly documenting this phenomenon.  Parents' social world and support  systems are also considered to be relevant exosystem influences on child development. Research designs are increasingly examining children's lives over time rather than at a static point in their development. Increasingly the environment is conceived in a complex fashion.  It does not consist of separate attributes such as socio-  economic status or ethnic background which once defined the environment in research. Instead the environment becomes conceived at least implicitly as a series of interdependent systems.  Some of these systems contain the  developing person and directly influenced by that person's perceptions, behaviour and development.  Other systems are more remote but no  less powerful in their ultimate effect on developmental processes and outcomes. (Bronfenbrenner & Crouter, 1983, p. 360) An ecological approach is being increasingly pursued in a variety of contexts in psychology. Minuchin (1979) has called for an ecological approach in the treatment of disturbed children. Conger (1981) has developed his own schema for conceptualizing the ecology of the family for use in the assessment of dysfunctional families. Ultimately however it is Bronfenbrenner (1979) who has developed the most comprehensive theoretical framework for understanding ecological research. In . addition, he has consolidated and added immeasurably to the body of learning to date in this area, and pointed out key directions for future research.  16  Use of the Ecological Approach in the Study of Children in Day Care A research interest in day care is relatively new but developing rapidly with current social trends. Of particular relevance is the tremendous rise in numbers of women in the work force. In the United States, the number of working women increased eight times between 1940 and 1975, and in 1980, 42% of mothers of children under six were working outside the home (Belsky, 1984). In addition, the rapidly rising numbers of single parent families, increasing numbers of children under five, and erratic employment situations for many families have necessitated increased reliance on childcare outside the home. Bronfenbrenner (1983) has identified day care as a target area for ecological research. He emphasizes the need to obtain systematic information about the settings in which children are placed when parents are at work since the nature of this experience is likely to be even more consequential for the child's development that the circumstances of parental employment, (p. 399) Belsky, Steinberg & Walker (1982), in their review of day care research, note its relative recency, with the earliest systematic study being completed in 1970. Belsky comments on the "pervasive narrowness of scope in day-care research that has virtually limited studies of the consequences of day care to investigations of the direct effects of the day care experience on the individual child" (1978, p. 930). These earlier studies generally fall into the following categories: effects of day care on children's intellectual development, effects of day care on emotional development (particularly the mother-child bond) and effects of day care on social aspects of development such as aggression.  Findings in all areas are contradictory and  inconclusive. Goelman and Pence (1985) point out the limitations of such fragmented research. The bulk of it has been carried out in high quality group day care facilities which in fact constitute only a small proportion of the available day care spaces; such settings do not in fact represent the wide range of licensed and unlicensed care  17  including family day care which a large proportion of children attend. In addition most day care research tends to be fragmented and does not reflect the complexity of day care; in fact, the type and conditions of care, family dynamics and environment of the child, worklife of the parent, qualities of the caregiver, and community and government support of day care are all relevant. As a result, according to Goelman and Pence, "the scope and content of much research of day care is conceptually and methodologically limited" (p. 324). Goelman and Pence (1985) as well as Belsky et al (1982) make a strong case for an ecological approach to day care, based on the model of Bronfenbrenner. Goelman and Pence (1985) have traced the development of such research. In studies focusing on the microsystem (or immediate setting of the child) researchers are increasingly interested in the processes that are occurring in that setting, particularly interpersonal relationships and interactions. In addition, qualities of the environment have become a research focus.  Prescott and her colleagues (1973)  examined the difference between home and day care environments and identified aspects of the physical and social environments that were optimal for child development.  Caldwell's (1979) HOME scale established criteria for the  measurement of physical environments, social environments and social interactions between parents and children. In her work with Bradley using the HOME scale (1976), Caldwell found that there may be relevant differences between homes that may ultimately influence a child's development. In Sweden, Cochran, (1977) compared children in regular homes, family day care homes and day care centers and reported that caregivers in the first two settings spent more time interacting with children and engaging in stimulating verbal activities. Research on the mesosystem of day care has begun to examine such issues as continuity between parents and caregivers.  Powell (1980) has made a major  contribution in this area, calling for an approach which stresses "interrelatedness"  18  (p. 215). He cities such influences as social roles, cultural context, and relevant social and economic forces as critically important in studying the mesosystem of day care. Long, Peters and Garduque (1985) have further conceptualized the methodological issues involved in studying the relationship between home and day care settings. Fowler (1978), in his work on high quality intervention systems, touched both on the microsystem and mesosystem of day care. He looked at the day care setting, home setting and the relationship between the two. Fowler used Caldwell's HOME scale to observe characteristics of the child's home environment. A limitation of his work is that it is based on high quality programs which are not typical of most children's day care experience, and thus the results are not highly generalizable. This limitation was overcome in the study conducted by Project Child Care (Johnson, 1978), which was primarily based on data collected from parents and caregivers in unlicensed settings. Studies on the exosystem and macrosystem of day care have focused on interrelated aspects of parental employment, government funding, and day care availability. In the Victoria Day Care Research Project, one of the most extensive ecological studies of children in day care, (Pence & Goelman, 1987a, 1987b; Goelman & Pence, 1987a, 1987b) studied children, parents and caregivers in group day care centers and both licensed and unlicensed family day care homes. In addition to developmental measures of the children, the study involved extensive caregiver and parent interviews. The intent was to look at relevant child, family and day care variables in interaction with one another,- rather than in isolation as had been the case in so many previous studies. The Vancouver Day Care Research Project (Goelman & Pence, 1986), of which this thesis is a part, actually emerged from the Victoria study. Subjects included children, caregivers and parents involved in licensed and unlicensed family day care homes, as well as a comparison group at a child development research  19  preschool.  This study sought to further understand the complex linkages and  interrelationships involved in the world of the child attending day care. The Vancouver Day Care Research Project involved a detailed study of the reciprocal influences within and between the microsystem, mesosystem, exosystem and macrosystem of a child attending day care. In addition, as its data is derived primarily from family day care homes, it is based on childcare settings which reflect the reality of the day care experience for a majority of children.  The Ecology of Family Environment Family Environment in the Ecoloeical Study of Children Attending Day Care Family variables have not been a strong focus in ecological day care research. Prescott et al (1972) and Bradley and Caldwell (1976) focused on aspects of home environment that were optimal for development.  Caldwell's HOME scale was  particularly developed to measure factors in the home environment, and focused mainly on physical environment, parent-child interaction and such variables as language stimulation. Carew (1981) made a significant contribution in her longitudinal study of children of children at home and in their family day care settings; she used interview methods as well as a version of the HOME inventory to measure home environment, and found significant correlations between aspects of home environment and child cognitive development. The Victoria Day Care Research Project (Goelman & Pence, 1987a) added the dimensions of family structure and maternal education level, finding that poorer child performance on standardized language tests was associated with such variables as lower maternal education. In addition, it seemed that children with lower levels of language development were more likely to attend poor quality day care facilities. In an attempt to uncover some of the complex factors that are responsible for this phenomenon, the Vancouver Day Care Research Project (Goelman & Pence,  20  1986) has been more extensive in its emphasis on relevant family variables. An ecological framework has been used to focus on maternal attitudes to work, marital status, childrearing attitudes and practices, and the relationship of these variables to each other, child language scores, and other relevant factors in the home, day care and parental work setting. The HOME scale was administered to parents and caregivers, interviews conducted with parents and caregivers, and obervations done of the child at home and in the day care setting. In addition, the Family Environment Scale was administered to parents and caregivers, measuring the dimensions of family relationships, personal growth orientation and system maintenance. The FES would hopefully provide background on family dynamics and social relationships that has been notably absent from other ecological studies of children attending day care. There has been a growing interest in the characteristics of family day care caregivers. In an early study of family day care, Emlen, Donoghue and Laforge (1971) looked at caregiver-parent interactions, and also examined caregiver demographics, attitudes, strain levels and adaptability in interpersonal relationships. Wattenberg (1977) looked at different styles of caregiving which she found were frequently linked to caregiver training, values and lifestyle. The Victoria Day Care Research Project (Pence & Goelman, 1987a) had a strong focus on caregivers, assessing such factors as personal characteristics, attitudes to childrearing, job satisfaction, and support systems. However, little attention has been paid to date to the effects that the caregiver's own family dynamics might have on the level of care she provides. The Vancouver Day Care Research Project is one of the first to examine the home environment of the caregivers in family day care settings and seek information as to the caregiver's own family social relationships. Considering the potential impact of the caregiver's own family environment on the quality of care when that care is provided in her own home, this is a major gap in related research.  21  To summarize, while family environment has been identified as relevant to child development, it has not been explored in depth as an integral part of the experience of a day care child as seen from an ecological viewpoint.  Family Environment and Child Cognitive Development A surge of interest in the importance of early experience in cognitive development has marked child development literature in increasing proportions over the past two decades. With this has come a specific interest in the role played by home or family environment in cognitive growth, as the family is clearly the immediate unit of influence for the very young child. The ecological influence of Bronfenbrenner has further emphasized the need to study children in their naturalistic settings (Parke, 1981). Home environment has been associated in the literature with such factors as socio-economic status, parental intelligence or education level, parental beliefs and attitudes, parent-child interactions, physical space or crowding, social support systems, number of siblings, the use of television in the home, and family dynamics or social relationships.  This review will concentrate on those aspects of home  environment most closely associated with social interaction, and family social relationships and dynamics. Measurement of these variables has been achieved through a variety of methods including the use of standardized instruments, interview schedules and home observation. Standardized instruments are preferred by some researchers because of the generalizability and comparability across studies afforded, and more direct observation of the environment is preferred by others because of the increased complexity of information obtained from more naturalistic data. Cognitive development has been ascertained with the use of standardized instruments of intelligence or mental performance, achievement tests, and tests of language useage.  22  Socio-economic status has been clearly favored as the measure of home environment in a large proportion of studies in this area. SES, as measured by parental education or income level, has been widely found to be positively related to intellectual development.  Walberg and Marjoribanks (1976) report that studies  investigating the relationship between the chosen measure of SES and the child's IQ generally report a correlation of approximately .40. Longstreth et al (1981) describe the typical conclusion to this finding: It is usually taken for granted by most American psychologists that the superior IQs of upper and middle class children are completely explained by the superior environment in which they are reared and that, conversely, the more deprived or disadvantaged conditions of lower SES homes account for the poorer average IQ of their children, (p. 532) Longstreth et al (1981) point out that such thinking is fallacious as it does not take into account the possible role of heredity in such findings. In addition, the authors point out that more direct measures of the environment are found to correlate more highly with IQ that does SES. Walberg and Marjoribanks (1976) also point out that families from different classes may be subject to other life stresses or social conditions that may in themselves be critical factors in influencing child cognitive development. Thus high correlations between family SES and child cognition may be the result of some outside factors not taken into consideration. Gottfried (1984) points out that studies such as these provide an index of a family's relative standing in regard to demographic factors, but can provide no direct evidence regarding proximal environmental variables in the child's life that are influencing cognitive development. Few such studies explore what it is about the environment that is responsible for the differences. As a result of these limitations, until recently little evidence has existed regarding the kinds of specific environmental  23  circumstances relevant to the early development of intelligence. According to Bradley and Caldwell, (1980) many investigations "have not provided sufficient clarification regarding the specific mechanisms whereby the environment exerts its influence on development" (p. 1141). Gotttfried (1984) does note however that as early as 1929, research was being conducted that constituted early exploration in a more direct observation of the role of the home environment on learning. At that time Van Alstyne, in studying the relationship of home environment to the intellectual development of three year olds, assessed home environment through a maternal interview schedule that contains many social and physical aspects of contemporary home environment measures. Van Alstyne found a positive relationship between the degree of verbal stimulation and encouragment which a child receives and her cognitive development. Since the late 1960's, researchers have turned more fully to understanding what it is about home environments that influences human development. Moore (1968) discovered positive relationships between general IQ and language skills on the one hand and several measures of home climate on the others. These measures included availability of toys and books, parental encouragement, and emotional atmosphere of the home. This was was one of the first studies to explore the possible role of emotional aspects of home climate in cognitive development. Wachs et al (1971) also used a more process-oriented research strategy in conducting a cross-sectional study using an early version of Caldwell's HOME scale. The authors investigated the relationship between these specific environmental processes and scores from the Infant Psychological Development Scale. They found that these cognitive measures were negatively correlated with such factors as noise ratings, excessive activity and frequent visitors, and positively correlated with verbal stimulation and presence of books. In addition they found that maternal involvement with the child correlated highly with the child's IQ at 15, 18 and 22 months.  24  Marjoribanks (1972) used a similar methodological approach, using a home environment measure to interview the parents of 185 eleven year old boys and reported that the score on the home interview was a better predictor of the boys' verbal and numerical abilities than were SES variables. Tulkin and Kagan (1972), in comparing the home environments of ten-month old girls from white middle class and working class families, found that middle class mothers made significantly more verbal contacts and engaged in more cognitively stimulating behaviours with their children than did working class mothers; non-verbal contact did not differ significantly. Although the authors focused on social class as a measure of differing home environments, they also examined such factors as beliefs and values that may have contributed to the different results. The work of Clarke-Stewart (1973) provided detailed longitudinal observational data on the relationship between the quality of a broad range of motherchild interactions in the home and and the intellectual development of young children. Clarke-Stewart found that all measures of infant competence, including cognitive and language skills, were highly intercorrelated with qualities associated with optimal caregiving. In a series of important studies, Caldwell and associates employed processoriented research strategies to go beyond the simple designation of environment as SES and to characterize the stimulation potential of individual homes. These studies used a version of the HOME Inventory developed by Caldwell, Huder and Kaplan in 1966, which uses a combined observation-interview procedure. It consisted of six subscales: 1) emotional and verbal responsivity of the mother, (2) avoidance of restriction and punishment, (3) organization of the environment, (4) provision of appropriate play materials and (5) opportunities for variety in daily stimulation. In Bradley and Caldwell's 1975 longitudinal study of black and white children of mixed SES, (Elardo, Bradley & Caldwell, 1975), the authors found that "maternal  25  involvement with the child", assessed at 6 or 11 months, correlated highly with the child's IQ at 36 months. The 24 month assessment indicated that the "emotional and verbal responsivity of the mother" correlated highly with the three-year IQ score. A follow-up study of these infants at 54 months indicated that maternal involvement and the mother's emotional and verbal responsivity, assessed at 24 months, were both significant predictors of IQ at 54 months. In addition, the authors found that the "organization of the physical and temporal environment" over the first two years was related to IQ at age 3. However this dimension was less strongly associated with IQ at age 54 months, suggesting that different types of environmental stimulation may be differentially important at various developmental points (Bradley & Caldwell, 1976). In a later study using a similar sample, Elardo, Bradley and Caldwell (1977) focused on language development as a measure of child IQ, and found high correlations between several HOME subscales and the Illinois Test of Psycholinguistic Abilities. The HOME inventory was administered at 6 and 24 months, and the language test administered at age three. The scales most highly associated with higher language competence were: mother's emotional and verbal responsivity, mother's involvement with the child, provision of appropriate play materials and opportunity for variety in daily routine. The complex interpretations in this study led the authors to conclude from their results that "children's general cognitive development appears to be associated with a responsive and organized environment in which opportunities are provided for appropriate learning activities during the first year or so of life" (Elardo, et al, 1977, p. 602).  Once again the authors found that the relationship between  environmental organization and cognitive measures declined when the child was older. Findings of this sudy and Bradley and Caldwell's 1976 study seem to confirm Parke's (1978) observation that "home environments are organized by sets of social rules that aid in regulating in an orderly fashion the interactions between the occupants...These  26  organizational patterns that serve to regulate social interaction among family members shift as the child develops" (p. 65). In another study, Bradley and Caldwell (1977) made a comparison between an environmental measure of family environment (the HOME scale), and a measure of socio-economic status in predicting IQ in separate groups of blacks and whites and males and females. They found that the HOME scale was a better predictor of IQ than SES in the racially mixed group, and that it predicted IQ as well as a combination of environmental process and SES measures. They also found that there was a loss of predictive power when SES was used by itself. In an attempt to discover directionality of the relationship between home environment and cognition, Bradley, Caldwell and Elardo performed a cross-lagged panel analysis (1979). They looked at three subscales of the HOME scale and their relationship to Bayley Mental Development Index scores of children at six, twelve and twenty-four months. They found that "more capable children tended to elicit higher levels of maternal involvement and provision of more appropriate play materials during the 6-12 month period, whereas higher levels of maternal involvement tended to produce more capable children during the 12-24 month period" (p. 246).  These  findings demonstrate a reciprocal influence between child and environment. In a more recent study, Bradley and Caldwell (1980) focused on the relationship between home environment assessment (using the HOME scale) of a group of children at 6 and 12 months, and their scores on the Bayley at age one and the Binet at age three. They reported substantial correlations between HOME and Binet scores for boys and girls, but noted there were some sex differences which the authors suggest indicate that optimum home environment variables may vary for boys and girls. Other studies have sought to ascertain the differing contributions of genetic and environmental family influences in a child's intellectual development. In defending  27  the rationale for such research, Willerman (1979) points out that high correlations between home environment measures and IQ cannot rule out the potential contribution of heredity "particularly since more intelligent parents are more likely to provide intellectually more stimulating home environments" (p. 926). Scarr and Weinberg, (1978) in their study of adoptive and biologically-related families, found that parental education correlated highly with child IQ, and the authors conclude that "genetic differences account for the major part of the long term effects of "family background on IQ" (p. 674). They further caution that: "social scientests should be very wary of interpreting the causes and effects of class differences in studies of biological families. We should be sensitive to the genetic transmission of family chracteristics" (p. 691). Longstreth et al (1981) attempted to measure the separate contributions of maternal IQ and home environment as they predict child IQ. Parent and child IQ were measured with Raven's Standard Progressive Matrices and the Peabody Picture Vocabulary Test.  Home intellectual environment was assessed by maternal  interview. The authors found that home environment ratings did not add significantly to the prediction of child IQ from that predicted by maternal IQ, leading them to conclude that "the correlation of home environment and child IQ is considerably overestimated when maternal IQ is allowed to covary" (p. 532). By contrast, Carew's (1981) longitudinal study of children in home and family day care settings found strong correlations between home environment measures the (HOME scale) and several child cognitive measures, including the Binet. The author concludes that: the child's early human environment in the home seemed to play a key role in influencing his intellectual development. This role was especially influential with respect to language stimulation. Successful performance on the...Binet may be facilitated by certain caregiver-child interactions in the first two and a half years of life, especially those in which the  28  caregiver plays a direct, unilateral or a reciprocal role in creating language-mastery experiences for the child, (p. 46) Factors related to family ecology were considered in a study relating family characteristics and mother-infant interaction to prediction of IQ and language skills done by Bee et al., (1982). This study also examined such family-related variables as maternal education, social support systems, level of stress and parental perceptions of the child, and found that they were strongly related to IQ and language in the low education subsample, but not amongst more highly educated mothers. The HOME scale, which was also administered, proved to be the strongest predictor of IQ of all family data. In the 1980's more attention has been paid to family social relationships and dynamics in assessing the family impact on cognitive development. The use of the Moos Family Environment Scale (FES) has played a role in several such studies. Garfinkle (1981), in a study of genetic and environmental influences on the development of mathematical concepts in young children, used the FES in conjunction with the Piagetian Mathematical Concepts Battery (PMBC) and the Peabody Verbal Picture Test (PVPT). Garfinkle found significant relationships between the FES Intellectual-Cultural Orientation subscale and both the PMBC and the PPVT (.16 and .20 respectively). A positive relationship was found between the FES Cohesion scale and the PPVT, and a negative relationship between the FES Achievement Orientation subscale and and the PMBC. The author also reports correlations of .47 and .46 between the FES Intellectual-Cultural scale and paternal and maternal education respectively. This finding suggests that strong correlations between the IntellectualCultural subscale and cognitive measures may be mediated by a common relationship with parents' education or socio-economic status.  29  The FES was also used in a study by Harris (1982) on genetic and environmental influences on reading achievement in twin children. Cognitive measures included reading achievement tests, estimates of general intellectual ability and language scores derived from the Illinois Test of Psycholinguistic Abilities. Although the study emphasised genetic relationships, the FES and a questionnaire on attitudes toward education were included to assess possible environmental factors. Findings included a significant relationship between the Intellectual-Cutural subscale with reading tests; 4% to 9% of the variance in the reading tests could be accounted for by this subscale. The author concluded that the findings of the study "are consistent with the idea that exposure to and emphasis on intellectual and cultural activites will promote, to a limited extent, achievement in reading" (p. 110). In an attempt to rectify a weakness of previous studies, Gottfried and Gottfried (1984) controlled for SES, maternal intelligence and nursery school attendance in their attempt to study the relationship between home environment and cognitive development. In addition to the HOME scale, the FES was used as a measure of the environment because an aspect of the home environment given little attention in the recent literature in this area is "the quality of social relationships among family members" (p. 66). The authors note that social climate would certainly affect the amount of stimulation present in the household.  Cognitive development was  measured using the mental scale of the Bayley and MacCarthy scales, and the test of Early Language Development. social histories.  Maternal IQ measures were administered as were  Results showed significant correlations between the Cohesion,  Expressiveness and Intellectual-Cultural subscales of the FES and the cognitive measures; these correlations were low however (from .16 to .36).  All significant  correlations were with the sensorimotor and McCarthy scales. Cohesion related with object permanance, the Bayley and McCarthy mental indices; family Expressiveness correlated with object permanance, McCarthy verbal and memory index and the  30  perceptual-performance and general cognitive indices; the Intellectual-Cultural scale correlated with object permanance, the Bayley and McCarthy indices. In addition, FES scores were correlated with subscales of the HOME inventory. The largest number of significant correlations were with the Cohesion, Expressiveness and Intellectual-Cultural subscales of the FES.  The Cohesion and  Intellectual-Cultural subscales both correlated with the following HOME scales: Responsivity of the mother, Language Stimulation, and Pride, Affection and Warmth. Cohesion, Expressiveness and Intellectual-Cultural scales also correlated with Organization of the Environment, and Modelling of Social Maturity.  The FES  Independence subscale correlates with toys, games and material, and stimulation of academic behaviour. The Active-Recreational subscale correlates with Maternal Involvement, and Toys, Games and Materials. The Organization scale correlated positively with Physical Environment. The Moral-Religious subscale correlates negatively with Language Stimulation. Gottfried and Gottfried (1984) concluded that positive relationships were found between cognitive development and family climate as assessed by "the cohesion, expressiveness and intellectual-cultural subscales of the FES" (p. 104).  Thus  commitment, help and support amongst family members, encouragement of family members to act openly and express their feelings, and family interest in political, social, intellectual and cultural activities are all associated with higher child cognitive ability. The profile of the family environment which is more favorably stimulating for child cognitive growth is, according to these results: one where relationships are characterised by more cohesion, more expressiveness and less conflict, where personal growth is stressed predominantly by greater independence, intellectualcultural and active-recreational orientations, and system maintenance is conducted by  31  more organization in planning and structuring activities and less control by rigid rules and authoritarian techniques. The authors further conclude that "mother's intelligence does not significantly predict children's cognitive development whereas home environment is a significant predictor" (Gottfried & Gottfried, 1984, p. 109). They also point out that mother's intelligence and demographic factors relate "not only to children's cognitive development but also pervasively to the provision of their social and physical home stimulation" (p. 109). Mothers of higher intelligence, education and SES are seen to provide a social and physical environment that is generally more enriched and conducive to enhancing children's cognitive development. In assessing the role of family environment as measured by the FES, the authors conclude that a major role or influence of the social climate of the family may reside in its indirect rather than its direct relationship with cognitive development.  The quality of social relationships in the home may  influence the level and quality of home stimulation provided for the child, which in turn may account for individual differences in children's developmental status, (p. 110) Family climate and social relationships are increasingly being seen as a worthwhile aspect to explore in determining the influence of family environment on child cognitive development. The Moos Family Environment Scale is providing a useful measurement tool which allows some generalizability across studies. Although findings using this scale have varied, there is some consistency in reporting strong correlations between child cognitive development and the Intellectual-Cultural subscale of the FES. While studies using this instrument are becoming increasingly ecological in orientation (particularly Gottfried and Gottfried's 1984 study), this is an area which is only just beginning to be tapped. Clearly such factors as parental education, attitudes to childrearing, parental employment and attitudes to  32  employment, all play a mediating role in family social climate and its reciprocal influence on the family. There is considerable room for the study of the FES in specific populations, taking into account the impact of as many of these other variables as possible. Such information will enable the researcher to get further insight into the complex and changing nature of the family climate and its relationship to child cognitive development.  Parental education, socio-economic status and family environment According to Bronfenbrenner and Crouter (1983), "the most powerful environmental factors influencing both the effectiveness of socialization processes and the impact of biological factors are likely to be those associated with the family's socio-economic status" (p. 395). As already noted, many studies have used socioeconomic status as a measure of the family environment in assessing such issues as the relationship between family environment and child cognitive development just reviewed. Parental education (in particular maternal education) has often been used as an indicator of socio-economic status, as has maternal IQ, income level and occupational status. As Bronfenbrenner points out, (1986) parental education has emerged in research as an "important source for parents' conceptions of the nature and capacities...of the child" (p. 736), and has the advantage of being a variable which precedes the formation of the family unit and birth of children, and is thus less likely to be influenced by family processes. Parental IQ focuses on the genetic aspects associated with socio-economic status. Family income level is an important marker in western society where financial resources clearly are linked with the family's ability to provide adequate healthcare, childcare, and access to culturally and intellectually stimulating activities. Parental occupational level also has important repercussions in terms of social status, self-image, job satisfaction and stress. Clearly there is a strong and complex relationship between all of these factors and family environment.  33  Attempts to study socio-economic status have been fraught with difficulties and controversial issues. One of the most potent problems to emerge has been the debate referred to previously in this review over the potential role played by genetics in apparant class differences. Such researchers as Hunt, who explains differences in intellectual performance by the presence of "various environmentally produced psychological deficits" (Golden & Birns, 1976, p. 306), have been contradicted by other researchers including Jensen (1969) who concluded that compensatory education for underprivileged children was useless, as class differences in cognitive development were in fact inherited. This controversy has had important implictions in establishing the efficacy of programs such as Head Start which provide early enrichment for disadvantaged chidren. Other related issues have been the appropriateness of judging class differences using essentially middle class standards (Tulkin & Kagan, 1969), and the potentially culturally biased nature of standardized intelligence tests (Golden & Birns, 1976) in measuring class differences. Golden and Birns (1976) further note the importance of distinguishing between the structural indices of socio-economic status (which involves placing people into various groups based on such factors as parental education, income or occupation) and the class-related process-variables (which involve the direct experiences of an individual which tend to be associated with social class). These variables in fact may "have a more direct effect on children's development" (Golden & Birns, 1976, p. 300). Tulkin and Kagan (1972) used parental education and occupational status as measures of SES in their study of differing mother-infant behaviour in middle class and working class families described previously. Golden and Birns (1976) report the findings from their study which used maternal education as a measure of SES in 60 white males, with 90% of the mothers having graduated from college in the higheducation group; results showed a large significant mean IQ difference between the two SES groups at two years of age in cognitive measures. These findings were  34  markedly different from earlier research conducted by these authors (1971) which showed nonsignificant differences in two contrasting SES groups where the distinction in SES was made according to categories established by Hollingshead. In the lower class group all parents were on social assistance and 93% were single parent families; in the middle class group one or both parents had some schooling beyond high school but very few were college graduates. It appears from the findings in these studies that differences in cognitive measures are larger when the measure of SES is maternal education rather than an economically based determinant of SES. As already noted, studies by Marjoribanks (1972), Bradley and Caldwell (1977), and Bee et al (1982) provided evidence that indicated that a cluster of variables representing environmental processes and conditions were associated with child cognitive competence as were structural indices of socio-economic status. These contrasting findings indicate that future studies in this area need to take into account that results of any study using socio-economic status as a variable will vary considerably depending on the measure used to determine socio-economic status.  In addition, research trends are clearly looking more closely at process-  oriented variables associated with different classes to determine what it is occuring within these groups that is responsible for emerging differences.  Family social  relationships is one area that has not been examined in this context. Future research needs to identify class differences in family social processes more clearly, using as many measures of SES as is feasible. The aim of such research is to gain insight into the complex relationships between family dynamics, parental education, income level and occupational status.  Such an approach can help clarify the combination of  structure and process variables that are at work within social classes, and which must certainly account for some of the clearly established differences in cognitive achievement in children from these differing social backgrounds.  35  Maternal employment and family environment As noted previously, maternal employment is an increasingly common phenomenon, with 42% of mothers with preschoolers working outside the home in 1980 (Belsky, 1984). This is a social trend which is relatively recent, and one which has been associated with changing social and cultural values. Only a generation ago hostile attitudes towards working mothers with young children were prevalent, with working women considered to be irresponsible and neglectful of parenting responsibilities.  Employed mothers have now become the norm rather than the  exception, and are far more socially acceptable than was the case even a decade ago. The phenomenon of working mothers has, however, certainly inspired a recent but growing body of research dedicated to the effects of maternal employment on child development and behaviour. As Lamb points out (1982) the bulk of this research looks at employed mothers vs. unemployed mothers, and does not take into account that working mothers do not constitute a single homogenous group. In fact, according to Lamb, "maternal employment is likely to have different effects on family functioning and parenting depending on the individual's circumstances and motivations" (p. 46). He cites economic pressures as a major reason for maternal employment, with some mothers working out of choice (to provide a higher standard of living in a two-parent family for instance) and others working from necessity (such as single mothers attempting to keep their family incomes above the poverty line). However, increasingly women are choosing to work for personal reasons related to importance of career and enjoyment of work; this trend has paralleled the growth of the women's movement which has gained wide social acceptance in encouraging women to fulfill themselves in work situations outside of home and family. Lamb suggests that: there is a world of difference between the woman who has to work although she does not want to do so and the mother who has carefully  36  prepared herself for the combination of parenthood and career because they are intrinsically important to her. Both are employed but they are likely to behave differently toward their children and so affect their development in distinctly different ways. (p. 49) Lamb also notes the importance of the mother's working hours (part time or full time) and the amount of spousal support for her employment as critical factors in her parenting. However, as indicated, the bulk of research to date has focused on the link between maternal employment and child development, and there has been little research investigating the issues raised by Lamb until recently. The work of Hoffman (1982) is based on the author's belief that there is "a link between mothering attitudes and the extent to which employment provides achievement satisfaction" (p. 256). According to this theory, the mother's ability to parent will be affected by the level at which she is able to satisfy her own needs in other areas. The mother's satisfaction with her employment status "will have a positive effect on family relations, mothering behaviour and child outcomes" (p. 257). In some early research in this area, Hoffman (1963) embarked on a line of investigation focusing on the mediating role of maternal attitudes in influencing the impact of the mother's employment status on the child. She found that mothers who liked their work reported stronger feelings of attachment to the children and imposed less severe discipline than those not working outside the home. The children in turn expressed more positive attitudes toward their mother. In contrast, mothers who did not like their work used less power assertion in relating to their children, used physical force more often and responded to frustration in a less adaptive way. Hoffman (1984) identifies a critical factor as that of the mother's morale, which is clearly affected by stress associated with employment. She particularly notes such factors as excessive strain experienced by some individuals caused by juggling two  37  roles, stress associated with concerns about childcare, and stress related to working full time vs. part-time, or in less flexible working conditions as all having a potentially disruptive effect on the child and family. Contradictory evidence has been found in several studies which link time spent at work with time spent involved in family processes. Hill and Stafford for instance (1978) found that women employed more than 20 hours a week spend less time involved in child care, but that this varies according to maternal education. Hoffman (1982) reports that results from the few studies that have differentiated part-time from full-time employment have found part-time employment an unusually successful adaptation to the conflict between the difficulties of being a full-time housewife and the strain of combining this role with full-time employment. These mothers seem to be physically and psychologically healthy, positive towards their maternal roles, and active in recreational and community activities, (p. 264) Other studies have looked more at the quality of mother-child interaction in employed vs. unemployed mothers; Hoffman (1982) reports results of her research which indicate that working mothers make up for the less frequent interaction by increasing the quality or intensity of the interaction when it does occur. However, in general, studies which focus on "the mother's morale as the linkage between her employment status and the child's socialization in the family" (Hoffman, 1982, p. 261) are relatively rare. Stuckey, McGhee and Bell (1982) note that "research in the area of maternal employment has focused on the effects of the mother's employment on various aspects of child behaviour and development without considering how maternal employment has altered parental behaviour" (p. 635).  Their research examines family interaction  patterns in an attempt to further understand the influence of maternal employment on children. The authors stress the importance of the mother's attitude to her work in  38  assessing these patterns in their study, which examined the influence of maternal employment on parent-child interaction. They found that increased negative effects were found in families where there was incongruence between the mother's attitude to her work and her employment status, leading the authors to conclude that "certain attitudes, in combination with maternal employment status, have more influence on parental behaviours than maternal employment alone" (p. 635). Although family interaction has been targeted as an area which would almost certainly be affected by factors related to maternal employment status, little research has been done to ascertain exactly what patterns of family dynamics or social relationships are related to particular variables associated with the mother's work situation. Such factors as attitudes to work, spousal support of work, full or part time work conditions, flexibility of work, and stress associated with work, would all likely have some form of repercussions in the family social environment. This area provides a rich source for further study.  Summary A review of the literature indicates that family environment is a significant mediating factor in many aspects of the life of a young child enrolled in a family day care setting. Bronfenbrenner (1980) provides a comprehensive conceptual framework for the study of family environment in an ecological context where its mutual impact on relevant systems in an individual's life can be determined.  Such researchers as  Belsky, Steinberg and Walker (1982), Goelman and Pence (1987a, 1987b), and Pence and Goelman (1987a, 1987b) have validated the usefulness of such an approach in studying children attending day care. There is considerable evidence that familial factors are of great import in a young child's life, and that any such study needs to address family related variables, particularly those associated with family process and social environment, more fully than has been done in the past.  39  The research questions set forth in Chapter One are supported by the literature to allow the estabishment of corresponding hypotheses which for the most part remain general in nature due to the exploratory nature of this study. These hypotheses are as follows:  1) Hypothesis 1: Family Environment and Child Language Scores It is hypothesized that there will be a positive relationship between child language scores and the Intellectual-Cultural, Cohesion and Expressiveness scales of the FES. A relationship between  the family environment and child cognitive  development is supported by the work of Wachs et al (1971), Marjoribanks (1972), Clarke-Stewart (1973) and Bradley and Caldwell (1977). Studies using the FES to measure this relationship have been done by Garfinkle (1981), Harris (1982), and Gottfried and Gottfried (1984), all of which found a positive correlation between child cognitive scores and the Intellectual-Cultural subscale of the FES. In addition Gottfried and Gottfried (1984) also found high correlations with the FES Cohesion and Expressiveness scales. 2) Hypothesis 2: Family Environment and Socio-Economic Status It is hypothesized that there will be a relationship between the family environment as measured by the FES and the following indices of socioeconomic status: maternal education, maternal occupation, partner's education, partner's occupation, family income, and family resources. Measures of socio-economic status have been examined in connection with aspects of the family environment, in studies which include those by Tulkin and Kagan (1972), and Golden and Birns (1976). Most studies however have used only one measure of socio-economic status such as maternal education, income level or  40  occupational status. As results have varied somewhat according to which variable is chosen, it would seem advisable to obtain results for as many measures of socioeconomic status as possible.  Aspects of the family environment associated with  social class have dealt inadequately with family social processes, although Marjoribanks (1972), Bradley and Caldwell (1977) and Bee et al (1982) have all identified environmental processes as an important focus for research on social class differences. Relating a series of measures of socio-economic status to a family's FES scores would provide important information regarding family social processes associated with various aspects of social class. As indices of SES are strongly and consistently linked to cognitive development in the literature, it becomes especially relevant to develop a family environment profile of the high SES families that are found to provide superior environments for child cognitive development.  3) Hypothesis 3: Family Environment and Conditions of Maternal Employment It is hypothesized that there will be a relationship  between  the family  environment as measured by the FES and conditions of maternal employment, including maternal attitudes to work, job satisfaction, hours of work and jobrelated stress.  Hoffman (1982) and Lamb (1982) identify maternal attitudes to work, job satisfaction, hours of work and job-related stress as relevant to child development; Stuckey et al (1982) emphasizes the need to look at these issues in the context of family interaction patterns. Although the literature strongly supports the association of family social patterns with these aspects of maternal employment, little research has been done which actually examines and identifies these patterns. The information potentially yielded by an instrument such as the FES would provide valuable new information in this area.  41  4) Hypothesis 4: FES Scores and HOME Inventory Scores It is hypothesized that a relationship will be found between parental F E S scores and parental H O M E scale scores.  It would seem reasonable to determine the degree of association between FES scores and other measures of the home environment. The HOME scale (1979) has been used extensively by Bradley and Caldwell (1976, 1979, 1980) to measure aspects of home environment; Gottfried & Gottfried (1984) have correlated FES scores and HOME scores and found some significant relationships. It is not possible however to generalize Gottfried and Gottfried's specific findings to this study as they used another form of the HOME inventory designed for younger children. Although they allow us to predict a relationship between FES and HOME Scale scores in this study, the exact nature of that relationship cannot be predicted.  5) Hypothesis 5: Family Environment and Parental Childrearing Attitudes and Beliefs It is hypothesized that there will be a relationship between the family environment as measured by the F E S and parental attitudes and beliefs associated with childrearing.  Although such researchers as Tulkin and Kagan (1972) looked at parental attitudes and beliefs as mediating factors in the relationship between social class and cognitive development in children, no research to date has specifically linked parental attitudes and beliefs with family process and social environment. This is a potentially fruitful area of investigation, as underlying parental beliefs and values undoubtedly make an impact on family environment.  Ecological research as described by  Bronfenbrenner (1980) further supports the relevance of deriving information from the macrosystem, which comprises beliefs and values, in a contextual study such as this.  42  6) Hypothesis 6: Family Environment of the Family Day Care Caregiver It is hypothesized that there will be a relationship between the family environment of the family day care caregiver, as measured by the FES, and the quality of care offered by that setting.  While Emlen et al (1971), Wattenberg (1977), and Pence and Goelman (1987a, 1987b) have all emphasized the characteristics of family day care caregivers, there has been no investigation into the caregiver's own family dynamics and social relationships and their relationship to the quality of care provided. The FES has a clear role to play in filling in this gap in the research.  43  CHAPTER THREE METHODOLOGY  This chapter presents the design of the study, the sample and the population from which it was drawn, the research design, instruments utilized, and procedures for research and data analysis.  Design of the Study The present study was an exploratory study, looking at the role played by family environment in an extensive ecological study of preschool children enrolled in family day care and preschool settings; this larger study was done by the Vancouver Day Care Research Project. Family environment was measured by the Moos Family Environment Scale. This was administered to parents and caregivers in the Vancouver Day Care Research Project. Results were correlated with: scores from other instruments administered by the larger study, demographic information related to parents and caregivers, and some responses from a written questionnaire that was given to parents and caregivers. Correlation coefficients were derived and t-tests conducted where applicable to determine the strength and direction of any association as well as significant differences between groups. Sample, Sampling Procedure and Population Data was derived from subjects living in the Vancouver Lower Mainland who were enrolled in the Vancouver Day Care Research Project. The sample was derived  44  from the population of children and parents of children enrolled in both licensed and unlicensed family day care centers in the Lower Mainland, as well as the population of caregivers at these homes.  Subjects included 31 children and the parents of children  enrolled in licensed family day care centers, and caregivers at the centers in which they were enrolled; subjects also include 16 children and the parents of children enrolled in unlicensed day care facilities, and the caregivers in these settings. In addition the study included data from a comparison group of 20 children and the parents of children who were enrolled at The University of British Columbia's Child Study Center preschool. Sampling procedures for the unlicensed group involved initial telephone contact with licensed family day care operators in Vancouver, and other Lower Mainland locations including Tsawassen, Delta, and Surrey. Lists of these facilities were obtained from government offices administering day care services. Licensed family day care caregivers were informed about the study and asked if they were willing to participate; those who expressed an interest were sent information on the study for themselves and for a parent whose child was in their care who would like to participate in the study. The final group that was enrolled in the study was finalized with follow-up phone-calls confirming the decision of both the day care operator and a parent at the center to participate in the study, and clarifying requirements for participation. Unlicensed family day care centers were solicited partially through the collection of advertisements in local newspapers offering childcare in the caregiver's home; the researcher contacted these individuals by phone to determine if they were in fact operating unlicensed family day care homes and would be interested in participating in the study.  Once again written material was sent to interested  caregivers and to potential parent participants, with follow-up phone-calls finalizing the group which ultimately took part in the study. Notices were also posted at  45  community centers and educational institutions asking unlicensed family day care operators or parents of children in unlicensed day care who might be willing to take part in the study to call the project office, but these generated very little response. Parents whose children were enrolled in a preschool class at the Child Study Center were contacted individually and asked to participate in the study. Problems involved in procuring the sample included a difficulty in recruiting unlicensed family day care caregivers who were willing to participate in the study; a possible reason for this is the reluctance to be part of an official research endeavor when the facility is in fact operating outside the licensing system.  Unlicensed  operators are fearful of outside interference and perceived professional judgment of the standard of care which they provide. As a result the number of unlicensed day care centers participating in the study was low, and considerably lower than the number of licensed facilities, making it difficult to contrast findings in licensed and unlicensed groups. Standardized Instruments Family Environment Scale The FES was administered for the purpose of this study to all parents and caregivers involved in the Vancouver Day Care Research Project. It is the central instrument in this study and is used to test all research hypotheses set forth in Chapter Two. It is one of ten Social Climate Scales developed by Dr. R. Moos and his associates at Stanford University's Social Ecology Lab and measures the social environmental characreristics of families. It consists of ten subscales which assess three underlying domains: the Relationship dimension, which includes the Cohesion, Expressiveness and Conflict subscales; the Personal Growth or Goal Orientation dimension which includes the Independence, Achievement Orientation, IntellectualCultural Orientation, Active-Recreational Orientation and Moral-Religious subscales;  46  the last dimension is that of System Maintenance, which is measured by the Organization and Control subscales.  The authors define all the subscales and  dimensions summarily in their manual. (See Appendix 1 for these definitions). The test is a 90 item scale, answered in a true-false manner, which contains clear instructions in the test booklet so that it is easily self-administered. Scoring is a simple procedure using the transparent scoring key provided. Test items are arranged so that each column of responses constitutes one of the subscales. Individual scores can be easily converted to standard scores using the conversion table in the manual's appendix. The test can be administered either individually or to all family members, which enables the interpreter to devise a family average. Test development included generating a pool of items related to family social environment, conducting structured interviews with different kinds of families, and adapting items from other social climate scales. The final scale has good psychometric properties. Normative data is derived from well over 1500 families from a variety of locations in the U.S., and sampling procedures are adequate and clearly described. Reliability data is well documented, with internal consistencies for the 10 subscales ranging from .61 too .78 (Cronbach's Alpha). The corrected average item-subscale correlations range from .27 to .44.  In terms of test-retest reliability, the author  reports reliability coefficients ranging from .68 to .86 at eight weeks, .54 to .91 at four months, and in the .52 to .89 range at the twelve-month period. Validity data is presented for content and face validity as well as for construct validity. The face validity is good; the wording of each item reflects the subscale as well as the underlying domain.  Evidence for  construct validity is based on  correlations with other studies and tests purporting to measure the same things. Although the reporting of these results is somewhat vague, personal experience with the FES indicates that the constructs are in fact valid. Despite some limitations in the reporting of validity data, Buros Ninth Mental Measurement Yearbook reports that  47  the "psychometric properties of the FES make it one of the best measures available for assessing families" (Busch-Rossnagel, 1985, p. 408).  The Peabody Picture Vocabulary Test The PPVT was administered to all children participating in the study, including those at licensed and unlicensed family day care homes and the Child Study Center. Scores from this instrument were used to test Research Hypothesis 1. The PPVT is a norm-referenced test of receptive language, applicable to a wide age range (from 2 years 6 months to 40 years). It is an individually administered test in which the examinee looks at four pictures and then points to the picture that is most closely related to the pronounced word. The manual reports respectable reliability using the standardization sample. Internal consistency was established with split-half procedures resulting in coefficients from .67 to .88, and alternate form reliability values were reported of .71 to .91. Other independent studies report alternate form reliability coefficients in the moderate to high range: .66 to .88 for standard scores and .79 to .86 for raw scores (McCallum, 1985). The validity evidence supports content, construct and concurrent validity. Its congruent validity with other IQ tests varies, as it is high for other vocabulary tests (.70 to .80) and somewhat lower for performance tests (.50 to .70). While correlations are quite high with other language measures, somewhat lower correlations (from .40 to .60) have been found between the PPVT and standardized measures of intelligence, indicating that, although PPVT scores are frequently considered indicative of cognitive functioning, they should not be equated with intelligence test scores. In particular, its correlation with the Wechsler Intelligence Scale for Children (WISC) Performance is subscale is in the .50 to .60 range, while its correlation with the WISC Verbal Subscale is higher.  48  In general, the national standardization, wealth of reliability and validity studies and availability of an alternate form all contribute to the high standard of psychometric properties to be found in this test.  The Expressive One-Word Picture Vocabulary Test The EOWPVT was administered to all children participating in the study, including children enrolled at licensed and unlicensed family day care homes and the Child Study Center.  Scores from this instrument were used to test Research  Hypothesis 1. This test is an untimed individually administered test of expressive vocabulary, designed for children from two to twelve years of age. It can be administered in 15 minutes or less, and consists of two demonstrations and 110 test plates, each containing hand-drawn figures. Answer sheets indicate starting points according to chronological age as well as correct responses. The authors have made a point of providing unambiguous scoring criteria. Reliability was determined through the use of split-half methods, in which the author computed internal consistency coefficents using a corrected Spearman-Brown formula; the resulting coefficents ranged from .87 to .96, and seemed quite stable across age ranges.  The standardisation sample, though adequate in size, was  restricted to the San Fransisco area, raising concerns as to the generalizability of findings to the population at large, and also a concern that there might be various cultural or other biases (Spivack, p. 566). Content and criterion-related validity are presented in the manual. Content validity was determined partly through test development which involved sending letters to parents of children between 18 months and 2 years asking them to generate lists of words commonly used by their children. Face validity appears good, with test items appearing to represent concepts which are commonly encountered in a child's  49  environment. Criterion-related validity includes correlations of .67 to .78 with the Peabody Picture Vocabulary Test. Again it should be noted that this is a test of expressive functioning, and should not be considered a measure of a child's total verbal intellectual functioning. However it does present a relatively high correlation of .73 with the Wechsler Preschool and Primary Scale of Intelligence (WPPSI) Verbal Scale IQ scores. Another limitation is that it does not provide an opportunity to assess elaborate verbal response as it consists of only one-word answers. The EOWPVT provides an excellent measure of expressive vocabulary, but does not have the same level of psychometric qualities as does the PPVT. As a result it is most effectively used in conjunction with other psychometrically sound verbal measures which measure additional aspects of verbal and cognitive functioning. It would be an excellent choice to use along with the PPVT in measuring child verbal ability.  Home Observation for Measurement of the Environment The HOME scale was used to measure the home environment of all parents and family day care caregivers in the study, and is used to address Research Hypothesis 3 as set forth in Chapter Two. Developed by Caldwell and Bradley (1979), the HOME scale for three to six year olds is an inventory designed to describe types of stimulation in the child's home environment that foster cognitive development.  It consists of nine subscales:  stimulation through toys and games and reading materials, language stimulation, physical environment, pride and affection and warmth, stimulation of academic behaviour, modeling and encouragement of social maturity, variety of stimulation, physical punishment, and total score. Another edition of the HOME scale is used for younger children.  50  HOME is administered during a home visit while the child is present so that the observer/interviewer can note interactions between child and primary caregiver. Items on each subscale were developed to focus primarily on observation but about one third involves parent report based on an interview format with items scored as simply 'yes' or 'no'. Subscales and items on this edition of the HOME scale are based on earlier versions of the scale. Point biserial correlations of items with their factor scores and total score are adequate. Reliability was determined by internal consistency which ranged from .53 to .88 for subscales and .93 for the total score. A test-retest study showed that scores were generally stable over time. Validity data was based on association between the HOME subscales and total scores with five measures of socio-economic status, and also with measures of child achievement. The HOME inventory is summarised in the current Mental Measurements Yearbook as a "useful and well-researched tool for identifying and understanding stimulation aspects of the home environment related to later cognitive functioning as assessed by traditional IQ measures and achievement tests" (Boehm, 1985, p. 663).  The Day Care Home Environment Rating Scale The DCHERS was administered at all family day care settings, and used to address Research Hypothesis 6 as set forth in Chapter Two. This scale provides the user with an instrument to measure the quality of early childhood education settings. It consists of 37 items divided into seven subscales: personal care routines for chidren, furnishings and display for children, languagereasoning experiences, fine and gross motor activities, creative activities, social development and adult needs. Raters provide scores for each item based on personal observation. Total scores for each subscale can then be put on a graph to provide a profile of that facility.  51  Inter-rater reliability data was obtained from three independent tests where two observers rated each of a series of classrooms independently produced correlations of .89, .79 and .84. Inter-rater reliability by item produced a rank order correlation of .93.  Internal consistency reliability tests produced coefficients  (Cronbach's Alpha) ranging from .32 to .79 for subscales and .83 for the total scale. Validity data was based on responses of early childhood experts who were asked to rate items on the scale in terms of importance; 78% of items on the DCHERS received high ratings and many that did not were modified.  Research Procedure Data Collection This study which is described in this thesis involved the administration of the Family Environment Scale (FES) to all parents and family day care caregivers who were subjects in the Vancouver Day Care Research Project. As part of this project, the following data had already been obtained: Child language scores: PPVT and EXPOWVT scores for children participating in the study, including those attending licensed and unlicensed day care facilities and the Child Study Center preschool; HOME scale scores for each family participating in the study and family day care caregivers; interview schedules for parents and family day care caregivers (the interview schedules differed); DCHERS ratings for each family day care home; videotaped sequences of the child interacting with a parent. The FES forms were left with family day care caregivers and parents participating in the study, with a self-addressed envelope. Follow-up phone-calls were made to those whose forms were not returned after approximately one month. The total number of parents' FES received was 58, with 24 of these from licensed family day care and 16 from unlicensed homes. A total of 44 caregiver FES were received, with 28 of these being from licensed and 16 from unlicensed facilities.  52  Some forms were not returned; two parents found the form inapplicable to them or objectional. Several others could not be contacted; the remainder simply chose not to send the form.  Data Processing and Analysis In order to provide as comprehensive a picture as possible of the role played by family environment in an ecological study of children attending family day care, measures of association were determined between the FES scores and as many variables as were deemed relevant to the research questions. Pearson product correlation coefficients were determined for each association measured, and were then tested for significance. For the purposes of this study, parents from licensed and unlicensed family day care homes were considered as one group, due to the small numbers in the unlicensed group, the unrepresentative nature of that sample, and a desire to increase the numbers involved in statistical procedures. Further support for combining the groups in this fashion was derived from a series of t-tests done between the mean FES scores for each subscale in the licensed and unlicensed groups; no significant differences were found in any of the scales, indicating that the groups did not differ significantly in terms of family environment as measured by the FES. Similarly family day care caregivers were considered as one group rather than as separate licensed and unlicensed day care groups. A) Parental FES scores (N=58) were correlated with the following variables in accordance with the research hypothesis set forth in Chapter Two, which have been repeated for the convenience of the reader:  53  1) Measures of child language: Hypothesis: It is hypothesized that there will be a positive relationship between child language scores and the Intellectual-Cultural, Cohesion and Expressiveness scales of the FES. a. EOWPVT scores (N=48) b. PPVT scores (N=67) 2) Items from the questionnaire (N=70); these included: a.) Items associated with socio-economic status: Hypothesis: It is hypothesized that there will be a relationship between family environment as measured by the FES and indices of socio-economic status. Indices of SES included: i. maternal years of education ii. maternal occupation (Blishen occupational code) iii. partner's years of education iv. partner's occupation (Blishen occupational code) v. family income vi. a Family Resource Variable constructed by factor analysis which includes household size, work hours, Blishen occupational code, years of schooling, household income, how far to nearest relative, how long in Lower Mainland, how long at this address. b) Items associated with maternal employment: Hypothesis: It is hypothesized that there will be a relationship between family environment as measured by the FES and conditions of maternal employment. Conditions of maternal employment included: i.  Responses to 'Attitudes to working women' (see appendix 2 for specific items)  ii. Maternal employment satisfaction  54  iii. Employment satisfaction (maternal) iv. Tension between work and childcare responsibilities (maternal) h) Responses to 'Child-rearing Attitudes and Practices' (see Appendix 2 for specific items) Hypothesis: It is hypothesized that there will be a relationship between the family environment as measured by the FES and parental attitudes and beliefs associated with childrearing. 3) Parent's HOME scale scores (N=44) Hypothesis:  It is hypothesized that significant relationships will be found  between FES scores and HOME scale scores. Parent FES scores (N=58) were correlated with parent HOME scores. B) Caregiver FES scores (N=44) were correlated with the following: Hypothesis: It is hypothesized that there will be a relationship between the family environment of the caregiver, as measured by the FES, and the quality of care offered by that setting. 1) DCHERS (Day Care Home Environment Rating Scale) ratings of their center (N=46) - total score only. 2) A Quality of Care Variable constructed through factor analysis which includes: Caregiver HOME score total, number of chidren in fulltime care, number of children in part-time care, largest number of children in care at once, number of hours per week that children are in care, hours of t.v., and caregiver household income. 3) Caregiver HOME scores (N=46) Correlation matrices were derived, and all Pearson correlation coefficients tested for significance.  Associations that emerged were studied with a view to  determining highlights, trends and patterns.  55  Limitations Limitations associated with the sampling procedure have been referred to; licensed family day care homes were difficult to recruit. Those that did respond and ultimately become enrolled in the project included a large number of atypical parents with above-average indices of socio-economic status. This possibly occurred because better educated families were more interested in participating in the project. However the unusual nature of the group presented some difficulties in data analysis. The researcher is aware that interpretation of correlations of the FES with a series of closely related questions (such as the scales Attitudes to Working Women and Childrearing Attitudes which are found in Appendix 2) must be done with caution as the dependency between variables has not been accounted for, giving rise to the possibility that some significant relationships may be due to chance. This study is a preliminary correlational study of the data which relates to the research questions of interest; more complex statistical operations such as analysis of variance, multiple regression and canonical analysis would provide further dimension to the results, but are beyond the scope of the present study. The number of variables examined in relation to family environment have been necessarily limited by the scope of this study. Therefore, while essentially ecological in design, this study does not present a complete picture of the ecology of family environment in relation to children attending family day care. While the psychometric properties of the Family Environment Scale are excellent, it is important to keep in mind that family social environment is a complex and changing phenomena. Although the FES provides one of the best measures available of this domain, it is still subject to error, and any interpetations must be made with the awareness that ultimately a standardized instrument can never provide a complete picture of such a complicated, evolving, and subjective domain.  56  CHAPTER FOUR RESULTS  The findings of the study will be presented in the following way: essential descriptive data, subscale intercorrelations, results in tabular form followed by general discussion and interpretation for each research question, and summary. Please note that the full form of all FES subscale abbreviations as well as definitions for each subscale are to be found in APPENDIX 1. Descriptive Information Table 1: Means and Standard Deviations of FES Scores in Parent & Caregiver Total Groups FES SCORES COH EXP CON IND PARENT FES N=60  ACH I/C  A/R  M/R ORG CONT  mean 59.3  55.3  45.4  48.0  44.7  57.0  53.8  47.0  55.6  50.7  s.d.  11.0  13.7  12.4  10.0  9.1  11.4  11.0  12  10.2  mean 55.9  54.1  44.6  51.7  44.6  53.8  49.5  51.2  53.5  50.3  s.d.  10.9  11.6  11.9  13.00 10.0  15.3  12.9  11.6  10.0  10.6  CAREGIVER FES N=48  13.3  All above scales passed the K-S test for normality, except the following scales, which underwent a square root transformation: Parent Scales: Cohesion, Expressiveness, Active-Recreational, Organization Caregiver Scales: Cohesion, Expressiveness  57  Subscale Intercorrelations The following intercorrelations are for FES subscales for total parent and caregiver groups. Table 2: FES Subscale Intercorrelations for All Parents: Pearson Correlation Coefficients FES SCALES EXP PARENT FES N=58 COH EXP  -.33*  CONF IND ACH  VIC  A/R  .06  -.19  .08  -.29*  .24  -.07  -.2  .07  .22  -.32*  -.20  .18  .15  .03  CON IND ACH  M/R  -.12  ORG  CONT  -.36*  -.21  .11  -.01  .09  .07  -.12  -.04  .23  -.04  -.21  .10  -.09  -.10  .18  .23  .23  .20  .40*  .60*  .07  .16  .00  .20  .16  .04  -.46*  .45*  vc A/R M/R ORG  -.41*  *p < .05 Findings Once again related research hypotheses as set forth in Chapter Two are repeated for the reader's convenience.  58  1) Findings related to child language scores: It is hypothesized that there will he a positive relationship between child language  scores  and the  FES Intellectual-Cultural, Cohesion and  Expressiveness scales. Parent FES scores were correlated with Peabody Picture Vocabulary Test (PPVT) scores and Expressive One Word Picture Vocabulary Test (EOWPVT) scores. Table 3: Pearson Correlation Coefficients for Parent FES Scores and Child Language Scores PPVT SCORES N=67  EOWPVT SCORES N=48  FES SCORES N=60 COHESION  .37** ,  .17  EXPRES.  .29**  .02  CONFLICT  .33**  .15  INDEPEND.  -.31**  -.17  ACHIEV.  -.48***  -.51***  INTELL./CUL.  .30**  .22  ACTIVE / REC.  .24*  .25*  MORAL /REL  -.30**  -.31**  ORGAN.  -.10  -.11  CONTROL  -.13  -.19  * p < .1 ** p<.05 *** p<.01  59 Discussion The number of significant correlations between the PPVT (Peabody Picture Vocabulary Test)) and the FES scores indicate a stronger relationship between that test and the FES than between the EOWPVT (Expressive One Word Picture Vocabulary Test) and the FES. In light of both the larger N and the more adequate psychometric properties associated with the PPVT, the results for that test would actually have more weight than the results for the EOWPVT. The EOWPVT results are however useful in the degree to which they support the PPVT results, and hence provide further evidence for any trends that emerge. As predicted in the literature, (Gottfried & Gottfried, 1984), the Cohesion, Expressiveness and Intellectual-Cultural subscales are all associated with higher PPVT scores, with reported coefficients of .37, .29 and .30, all significant at the .05 level. In addition higher language scores are linked positively with FES Conflict (.33) and Active Recreational (.24) scores and negatively with the following FES subscales: Independence (.31), Achievement (-.48), and Moral-Religious (.30). Of particular note here is the strong negative association between the Achievement subscale and the language measures. Negative correlations of -.48 and -.51, both significant at the .01 level, are reported for the PPVT and EOWPVT scores respectively. This negative relationship is also reported in the literature (Garfinkle, 1981), and may suggest a deterimental role to be played by parental pressure in influencing children's cognitive growth. The negative correlations of the language scores with the Conflict subscale differs from Gottfried and Gottfried's 1984 results. The positive association between family conflict and child language skills may in fact speak to the value of expressing anger openly in the family arena. These results allow us to draw a tentative profile of the family environment of a child with high PPVT scores. These families tend to be strong in the relationship domain: family members are close and provide help and support for each other; in  60  addition, family members are encouraged to relate openly and express feelings directly. They express anger and conflict openly in the family setting. Families that expose their children to political, social, intellectual and cultural activities, and encourage participation in social and recreational activities, also appear to be more facilitative of child language development. By contrast, results from both language tests indicate that families which tend to cast their children's activities into an achievement-oriented framework tend to be less facilitative of child language competency. In other words, it seems from these results that it is the exposure to more activities rather than the emphasis on performance which is conducive to the development of language skills. Families which foster better language skills also seem less likely to emphasize strong principles associated with religious belief and practice. 2) Findings related to indices of socio-economic status It is hypothesized that there will be a relationship between family environment as measured by the FES and several indices of socio-economic status.  Parent FES scores were correlated with: maternal Blishen code, Maternal years of education, Family Income, Partner's Blishen code, Partner's years of education, and a Family Resources Variable derived by factor analysis. available in Chapter 3.  Exact descriptions are  61 Table 4: Coefficents for Parent FES scores and SES Variables SES VARIABLES. N=70 FES SCORES N=60  COH  MATERNAL OCCUP.  -.04  MATERNAL PARTNER YRS. EDUC. OCCUP.  .09  PARTNER YRS EDUC.  FAMILY INCOME  FAMILY RES.  .08  .11  -.00  .26**  .35**  .37**  -.04  .15  .14  .17  -.01  EXP  .32**  .24**  CONF  .04  .22*  IND  .14  .02  .11  .06  .08  .05  ACH  -.16  -.14  -.08  -.12  -.10  -.06  .06  INT/CUL  .26**  .50***  .39***  .57***  .21*  .32**  ACT./REC  .35**  .60***  44***  .42 * *  .30**  .44**  MOR./REL.  -.20*  -.25*  -.12  -.11  -.12  -.25**  ORGANIZ  -.11  -.10  -.30**  -.13  -.02  -.11  CONTROL  -.29**  -.25**  -.24**  -.25**  -.18**  -.24**  * P<.1 ** p<.05 *** p<.01 Related data It should be noted that child language scores were found to correlate with the Family Resource Factor derived by factor analysis in the following ways: Family Resource Factor and PPVT : r=.38 Family Resource Factor and EOWPVT: r=.28  62  Discussion Those FES scales yielding the bulk of significant correlations related to socioeconomic status have consistent results across all six measures of SES; this consistency as well as the strength of the correlations involved increases the generalizability of these results.  The subscales which produced significant  correlations across nearly all SES indices include: the Expressiveness (Exp), Intellectual-Cultural (Int-Cul), and Active-Recreational (Act-Rec) subscales which are all associated positively with higher measures of SES, and the Control (Con) subscale which is associated negatively with high SES factors. Consistent though non-significant trends are to be found in the Conflict subscale which is mainly associated positively with high SES, and the Achievement and Moral-Religious subscales which appear to be associated negatively with high SES. It is of note that the Intellectual-Cultural subscale shows significantly high correlations with all indices of SES. This subscale is the one consistently noted in the literature as having a strong relationship with child cognitive growth (Garfinkle, 1981, Harris, 1982 and Gottfried & Gottfried, 1982), raising the possibility that this strong relationship might be mediated by a common relationship with socio-economic status, particularly parental education, which has produced correlation coefficents of .50 and .57 (p<.01) for maternal and paternal education respectively. , The correlation between the child language scores as measured by the PPVT and SES as measured by the Family Resource Factor in this study was .38, closely resembling the typical .4 correlation between cognitive measures and SES reported by Walberg & Marjoribanks (1976). This study then supports the strong relationship found between measures of SES and child cognitive development that has been consistently found in the literature. In addition, Table 4 provides important details as to the family social processes in these higher SES families which have been found to be more facilitative of optimal language development.  63  A family environment profile of the high SES family begins to emerge from this data. The strongest features of such a family are the ability of family members to act openly and express their feelings directly, the exposure to political, intellectual and cultural activities, involvement in social and recreational activities, and a lack of rigid rules which control family life. Less pronounced but consistent trends in such families include the ability to openly express angry and hostile feelings, a tendency to deemphasize an achievement orientation, and a lack of emphasis on moral and religious values. It may be noted that the profile of the high SES family is remarkably similar to the profile of the families whose children have high PPVT (child language) scores. The high PPVT families, however, are also noteably high in family support (Cohesion) which is not true of the SES families, who show a clearer absence of rigid rules in their functioning (Control). This data indicates that the two environmental factors measured by the FES which are strongest in high SES families are the exposure to intellectual and cultural factors and participation in social and recreational activities. Both these subscales produce consistently strong positive correlations, many of which are significant at the .01 level. These findings are consistent with the relationship between FES and child language scores in Table 2, and strengthen the idea that it is the exposure to more activities and ideas rather than the casting of performance in an achievement-oriented framework that most strongly characterizes those families that are more facilitative of cognitive development.  This is compatible with the idea that families with more  economic resources have access to more enriching activities for their children. However these results do indicate that it is likely the parental education level which is the strongest factor in determining which families will seek out these enriching activities; it is noteworthy that there are substantial correlations of .50 and; .57 (p<.01) between maternal and paternal education and the Intellectual-Cultural subscale, and a very high correlation of .60 (p<.01) between maternal education and  64  the Active-Recreational subscale.  This information suggests that, while genetic  factors may be partially responsible for higher cognitive skills in children whose parents are better educated, as is indicated in the literature (Longstreth et al, 1981), factors associated with family process and social environment are also in operation. These results also indicate that, while alternate measures of socio-economic status differ somewhat in the strength of their associations with family environmental process, they in fact produce very similar results, suggesting that the relationships between family social environment and indices of SES  are fairly constant regardless  of which measure is used. 3) Findings Related to Conditions of Maternal Employment It is hypothesized that there will be a relationship between family environment as measured by the FES and conditions of maternal employment, including maternal attitudes to work, job satisfaction, hours of work and job-related stress.  Parental FES scores were correlated with: questionnaire responses to 'Attitudes to Working Women' recorded in Table 5 (see Appendix 2 for items), and employment satisfaction, tension between work and childcare responsibilities, and maternal hours of work, recorded in Table 6.  65 Table 5 : Pearson Correlation Coefficients for Parent FES Scores and Attitudes to Working Women  Attitudes to Working Women N=70 I like work Working is Mothers Fathers I'm doing economic shouldn't work more FES Scores necessity if poss. involved N=60 COH  .14  EXP  Working moms miss out on children .11  -.01  -.06  -.08  .18*  .11  .08  -.22*  CONF  .02  -.15  -.10  .06  -.13  IND  .01  -.08  -.02  ACH  .21*  -.25**  -.28**  -.25**  .25**  .05  .11  .13  VC  .20*  -.25**  .03  .09  .04  A/R  .14  -.25**  M/R  -.05  -.28**  .15  -.11  .06  .20*  -.23** .26**  ORG  .31**  .01  .13  .24**  -.10  CONT  .20*  .26**  .09  .26**  .00  The full form of the above abbreviated questionnaire items re: Attitudes to Working Women are as follows: I like the work I am doing It is necessary to our family's economic survival that I work Mothers shouldn't work unless they absolutely have to Fathers have not been involved enough in the past and deserve to be A mother who works misses the experience of seeing her children growing up All above questionnaire items are based on an interval scale where l='strongly disagree' and 7='strongly agree' with individual item. * p <.l **p <.05  66 Table 6: Pearson Correlation Coefficents for Parent FES Scores and Maternal Employment Satisfaction, Work-Childcare Tension, and Hours of Work Maternal Employment Variables N=70 Maternal Empl. Satis. FES Scores N=60  Work/ Childcare Tension  Maternal Hours of Work  COH  .15  -.1.1  -.11  EXP  .13  .07  -.08  CONF  .06  .20*  -.16  IND  -.14  .15  ACHIEV  -.07  -.03  I/C  44***  .18*  .04 24** -.25**  A/R  .32**  .07  .14  M/R  .25**  -.06  -.09  ORG  .29**  -.10  -.13  CON  .12  -.15  -.16  'Employment satisfaction' and 'Work-Childcare Tension' are measured with an interval scale where l='strongly disagree' and 7='strongly agree'. * p<.l ** p<.05 *** p<.01  67  Discussion Tables 5 and 6 each provide a measure of maternal employment satisfaction, a variable which has been consistently shown in the literature to be connected with family functioning (Lamb, 1982, and Hoffman, 1982). The employment satisfaction measure in Table 6 should be considered a stronger measure of this variable than that in Table 5; results might be confounded in the latter because it consists of a series of closely related questions which are likely to have considerable inter-dependency. In Table 5 employment satisfaction is indicated in the column "I like the work I am doing", and in Table 6 employment satisfaction is labelled as such. Both measures report positive and significant associations (.20 and .44) with the Intellectual/Cultural (I/C) subscale, and with the Organization (Org) subscale (.29 and .31).  Table 6  reports a strong significant relationship (.32) with the Active Recreational (A/R) subscale, which is supported by a positive though non-significant association (.14) with this variable in table 5. Table 6 reports a significant negative relationship between "I like the work I'm doing" and Achievement (Ach). Both scales report positive though mainly non-significant relationships with the Cohesion (Coh) FES subscale and the Expressiveness (Exp) subscale. In addition, Table 6 indicates a significant positive relationship (.25) between the Moral-Religious (M/R) subscale and employment satisfaction. Both measures of maternal employment satisfaction agree that there is an increased exposure to intellectual, cultural and political activities and increased organization in family functioning in families where the mother is happy with her work. In addition both measures indicate a tendency for these families to emphasize participation in social and recreational activities, a tendency not to cast activities in an achievement-oriented framework, a tendency to be supportive and helpful to each other, and a tendency to interact openly and express their feelings directly. It is noteworthy that these findings are very similar to those for families of children with  68  higher child language scores. The subscale which produced the strongest results was the Intellectual-Cultural subscale. This subscale has particularly been found both in this study and in the literature (Garfinkle, 1981, Harris, 1982 and Gottfried & Gottfried, 1984) to be connected with higher measures of child language development. This study also found positive associations between child language scores and the Cohesion, Expressiveness and Active-Recreational subscale, and negative associations with the Achievement subscale. Although the Organization subscale was not found in this study to be more highly associated with child language competence, family organization has been noted in the literature to be conducive to cognitive development (Bradley & Caldwell, 1976). The tentative profile that emerges of the family where the mother is satisfied with her work then is one very much like that of the family that produces higher child language scores. It would appear from these results that the mother who is more satisfied in her employment situation may be more able to provide the kind of environment that is found to be conducive to her children's language development. Data found in the column "working is an economic necessity" in Table 5 provides an interesting contrast. Those families where the mother is merely working out of economic necessity show very different trends from those where the mother is happy with her work. Noteable here is the negative relationship between working from economic necessity and the Intellectual-Cultural subscale (-.25) and the ActiveRecreational subscale (-.25). A positive association is found with the Achievement subscale (.25). In these families, the factors in the environment which seem most consistently linked with child cognitive growth are not predominant. There is instead less emphasis on the exposure to intellectual, cultural and political activities, and participation in recreational and social activities which have consistently been found to be linked to higher child cognitive skills. A mediating factor here may be socioeconomic status; these families also have many of the environmental characteristics of  69  lower SES families as depicted in Table 4, including a high degree of control or the use of rigid rules to run family life. It is conceivable that economic factors may be combining with the problem of maternal morale in relation to work cited by Hoffman (1982) as a factor that is critical to family functioning. In addition to the attitudinal aspects associated with a mother who works from economic necessity rather than choice, the family may be under severe financial constraint and thus have less practical access to stimulating activities for their children. Other findings of particular note here are the negative association in Table 5 between the responses to both "Mothers shouldn't work if possible" and "A mom who works misses out on the experience of seeing her children grow up" and the ActiveRecreational subscale (producing correlations of -.28 and -.23 respectively). This subscale has been associated with more facilitative environments for children's cognitive growth. It is consistent with the literature (Hoffman, 1982, Lamb, 1982, and Stuckey et al, 1982) and the findings already noted in this study that working mothers who have some reluctance to working or who feel that they are neglecting their children by working actually provide less stimulating environments for their children. This may be because of problems associated with morale noted by Hoffman (1982) in reluctant working mothers. Another possibility is that mothers who feel they are neglecting their children by working may tend to keep their children close to their sides when they are at home rather than encourage them to participate in outside activities. This concept is supported by the strong negative relationship (-.25) reported in Table 5 between the item "mothers should not work if possible" and the Independence subscale. Once again mothers who feel guilty about working may find it more difficult to allow their children to be self-sufficient and to individuate successfully within the family. Maternal hours of work have been linked in the literature (Hoffman, 1982) with family process, with reduced work hours being associated with a higher level of family  70  functioning. Findings recorded in table six show that longer maternal hours of work are associated negatively with the Intellectual-Cultural (-.25) and positively with the Achievement (.24) subscales.  Both attributes are connected in the literature with  poorer cognitive development in children. It would appear from these findings that mothers who are able to work part-time are able to provide a more facilitative environment for their children in terms of cognitive growth. It is most likely that part time work allows the mother more physical energy and free time to spend with her children, and also means that she is less likely to experience work-associated stress. Results related to work/childcare tension in table 6 are less clear and do not provide any strong findings. Tendencies of note here are the positive though low correlation between work/childcare stress and the Intellectual-Cultural and Conflict subscales (.18 and .20 respectively). These findings suggest that the emphasis of the family on intellectual and cultural activities is not seriously affected by stress related to work and childcare conflicts, but that increased expression of aggression and anger may be connected with this kind of strain.  However these results are highly  speculative as the correlations involved are not high. Clearly the practical conditions of maternal employment do have an impact on family process. However, the findings from Tables 5 and 6 are particularly important in the context of an ecological study because they indicate that the beliefs and values associated with the family's work situation can have as much of an impact on the family environment as do the actual employment conditions. 4) Findings Related to Parent FES Scores and HOME Inventory Scores: It is hypothesized that there will be a relationship between parental FES scores and  HOME scale scores.  Parental FES scores were correlated with parental HOME scale scores. Although the focus here is on total HOME scale scores, results from individual subscales which yielded significant correlations are also included.  71  Table 7: Pearson Correlation Coefficients for Parent FES Scores and Parent HOME Scale Scores HOME Scale scores N=44 HOME  Stimulation by Games, Toys, Reading  Physical Stim. of Environment Academic Behaviour  Social Security Modelling  Total Score  FES Scores N=60 COH  .15  .31**  -.17  -.05  .13  EXP  .16  .19  -.02  .13  .13  CONF  .20  .06  .04  .18  .19  IND  -.15  -.04  -.06  .20  -.03  ACH  -.22*  -.34**  .07  .00  -.23*  vc  .12  .09  .28**  A/R  .18  .21  .04  M/R  -.41***  .14  ORG * ** ***  p<.10 p<.05 p<.01  -.35** .31**  -.29**  .20  -.07  .00  .01  .22*  .20  -.31**  .38**  .28**  72  Discussion The most important relationship in the Table 7 is that between the FES scales and the total HOME score, which rates the overall quality of the home environment. A fairly strong positive correlation of .28 is found between the total HOME score and the Organization (Org) subscale; a positive association which is weaker but nonetheless significant at the .1 level exists with the Active-Recreational (A/R) subscale. A negative association of -.31 is reported between the HOME score total and the Moral-Religious (M/R) subscale, with a weaker negative association of -.23 reported with the Achievement subscale. These results suggest that families which have optimal home environments as measured by the HOME scale tend to be particularly well-organized in terms of the structure which they provide for family activities. These families also have more emphasis on participation in social and recreational activities, and less emphasis on achievement and on strict religious values.  These qualities have all been found in this study to be associated with  environments that are more facilitative of child language development. The Physical Environment subscale of the HOME scale produces the highest number of significant correlations with the FES. Better quality physical environments are associated with higher Cohesion scores (.31), with lower Achievement scores (-.34), with lower Moral-Religious scores (-.35) and with higher Organization scores (.31). Optimal physical environments then are more likely to be found in families whose members are supportive and helpful to each other, who tend to not emphasize achievement, who are not heavily bound by religious tradition, and who are more organized in their family functioning. Other interesting trends include the consistently negative association between the FES Achievement subscale and the HOME subscales and total score., These findings appear to corroborate the results noted earlier in this study which associate improved child cognitive functioning with lower FES Achievement scores. It may be  73  that optimal home environments as measured by the HOME instrument similarly tend not to cast their members' activities in an achievement-oriented framework. The tendency of these families to have fewer ties to religious tradition is also consistently noted in the high negative correlations between the FES Moral Religious (M/R) •  subscale and the HOME scores. This subscale in fact produces the largest number of significant correlations between the FES and the HOME inventory. The Organization (Org) subscale consistently records quite strong positive associations with the Physical Environment subscale (.31), with the Modelling of Social Security subscale (.38) and with the total HOME scale score (.28). Organization has been linked in the literature (Bradley & Caldwell, 1976 & Parke, 1978) with better family functioning. These results also indicate that families which provide more adequate structure and planning tend to have better quality physical environments and also may be more successful in modelling pro-social behaviours. It is also of note that the HOME subscale Stimulation of Academic Behaviour is correlated .28 with the IntellectualCultural (I/C) subscale of the FES. This makes sense in light of the similar areas each subscale is purported to tap. These findings are different from those of Gottfried and Gottfried (1984) which found that the Cohesion, Expressiveness and Intellectual-Cultural subscales produced the highest number of significant correlations with the HOME scale, but their study used a different version of the HOME developed for younger children and therefore their results are not fully applicable to this study. 5) Findings Related to Parental Childrearing Attitudes and Beliefs It is hypothesized that there will be a relationship between the family environment as measured by the FES and parental attitudes and beliefs associated with childrearing.  FES scores were correlated with items from the parent questionnaire referring to childrearing attitudes and beliefs; a complete list of these items is located in  74  Appendix Two. Only those items from the questionnaire which report significant correlations with FES scales are reported in Table 8. Table 8 : Pearson Correlation Coefficents for Parent FES Scores and Attitudes to Childrearing  Attitudes to Childrearing  FES Scores N=60  Child knows Obedience/ best authority  Children Grateful  COH  .04  .09  .13  EXP  .04  -.02  .04  CONF  .13  -.11  .13  IND  .09  -.05  -.04  ACHIEV  -.04  .08  .16  vc  -.05  -.22* -.21**  A/R  .04  ORG  -.50***  M/R CONT  SelfExpression  -.10 .28**  -.06 .26**  Parental Permission  .08 .14 -.15 -.02  .05  .28**  -.11  .28**  .00  -.27**  .14  .18*  -.21**  .14  .12  .12  -.02  .29**  .11  .12  .31**  -.13  .19*  .33***  .11  .32***  The full forms of the abbreviations for Attitudes to Childrearing found in the table are as follows: Often it isn't the parent but the child who knows what's best Obedience and respect for authority are the most important virtues children can learn Children should always feel grateful for what their parents provide  75 It's important to encourage a child's self-expression even if one doesn't like what comes out Children shouldn't take responsibility for anything without the permission of their parents The questionnaire items on Attitudes to Childcare are measured on an interval scale, where l='strongly disagree', and 7='strongly agree'. * ** ***  p<.10 p<.05 p<.01  Discussion Caution must be used in interpreting Table 8. Results may be confounded due to the similarity and potential overlap of the items, raising the possibility of dependency between the variables. Nevertheless the table produces some interesting findings.  The item  "Obedience and respect for authority are the most important virtues children can learn" is associated negatively with the Intellectual-Cultural subscale (-.22), negatively with the Organization subscale (-.21), positively with the Moral-Religious subscale (.29) and positively with the Control subscale (.19). This data indicates that families which are more adult-centered and oriented towards obedience tend to expose their children to less in the way of intellectual, cultural and social activities, are less organized in family functioning, and are more bound by strict religious values and rigid rules. These qualities are in direct opposition to the attributes of families which facilitate higher child language scores (Table 1) and families with better quality home environments (Table 7). Another item which focuses on more adult-oriented households is "Children should always feel grateful for what their parents provide" which is correlated negatively with the Active-Recreational subscale of the FES (-.27) and positively with the Control subscale (.33). These families are less likely to expose their children  76  to the stimulating activities which are found to be associated with better learning, and are also more likely to be bound, by rigid rules. A similar item, "Children should not take responsibility for anything without the permission of their parents", produces positive correlations of .28 with the Achievement subscale, .31 with the MoralReligious subscale, and .32 with the Control subscale of the FES. The suggestion here is that these families tend to cast activities in an achievement-oriented framework (which has been found to be negatively connected with child cognitive growth in this study), and are bound by strong religious values and controlling rules. By contrast, an item clearly associated with child-centered parenting values, "It's important to encourage a child's self-expression even if one doesn't like what comes out", has very different results. This item is correlated positively with the following FES subscales: Expression (.28), Independence (.26), and IntellectualCultural (.28). It is suggested by these results that families which are more childcentered and which encourage self-expression are more able to act openly and directly, have members who are more self-sufficient, and expose their children to more intellectual, cultural and social activities than families which discourage such selfexpression. A more extreme item tapping the dimension of child-centered parenting attitudes is "Often it isn't the parent but the child who knows best". Results here indicate that families who believe that it is the children's opinions which should dominate in the family report a very strong negative association with the Organization subscale (-.50). A cautious interpretation of these results suggests that parental attitudes which reflect a healthy respect for the child's needs for self expression, and which favour parenting situations where the adults are not overly dominant and authoritative are associated with the most beneficial family environments for children. However the results also indicate that a child-centered parenting attitude can be taken to an extreme, leading to disorganization in family functioning.  77  6) Findings Related to Family Environment of Caregivers It is hypothesized that there will be a relationship between the family environment of the family day care caregiver, as measured by the FES, and the quality of care offered by that setting.  FES scores of caregivers were correlated with three measures used to assess the quality of day care provided by the family day care home: the DCHERS (Day Care Home Environment Rating Scale), which is a standardized instrument designed especially to measure family day care environments, a quality of care variable derived by the Vancouver Day Care Research Project using factor analysis which is described fully in Chapter Three, and the HOME (Home Observation for the Measurement of the Environment) scale, which has been used extensively in the literature to measure the quality of the home environment.  Discussion An important feature of these results is the consistency across all three quality of care variables in those FES subscales producing significant correlations. The Cohesion (Coh) subscale produced significant correlations of .24, .24 and .25 with the DCHERS, Quality of Care Variable, and HOME score total respectively.  High  positive associations of .33, .45 and .33 were reported between these three scales and the FES Independence (Ind) subscale.  High positive correlations are also  reported between these scales and the FES Active-Recreational (A/R) subscale; these are .29, .33 and .40. Similarly the Organization (Org) subscale of the FES produces strong positive correlations of .50, .54, and ,40 (all significant at the .01 level) with the quality of care measures. The Intellectual/Cultural subscale, though not producing coefficients which are significant at the .05 level, nevertheless is fairly high in the three quality of care measures, with coefficients of .15, .27 and .21. The consistency of these results across all three measures strengthens their  78  Table 9: Pearson Correlation Coefficients for Caregivers FES Scores and Quality of Care  N-46  Oualitv of Care N=70  HOME Scale total N=44  .24*  .24*  .25*  DCHERS Total  FES Scores N=44 COH EXP  -.02  .00  -.22*  CONF  -.10  .04  -.12  IND  .33**  .45***  .33**  ACH  .00  .01  .04  JVC  .15  .27*  .21  A/R  .29**  .33**  .40***  M/R  .04  .06  .02  ORG  .50***  .54***  .40***  CON  -.01  -.01  .07  P<.1 p<.05 p<.01  interpretability and allows us to draw a fairly clear portrait of the family environments of those family day care caregivers who provide more optimal care situations. These families provide a high degree of support and help for one another; they value each other's independence and ability to be self-sufficient; they tend to emphasize active participation in social and recreational activities, and have a clear  79  organization and sense of structure in their own family functioning. They also have a tendency to expose their members to more intellectual and cultural activities than do poorer quality family daycare homes. It is noteworthy that these qualities are also associated with environments which are more conducive to child language competence (Table 1), and with those families which are found to provide optimal home environments (Table 7). The only exception is the Independence subscale; perhaps because these caregivers are in fact running a business within their home situations, they require more self-reliance on the part of their own family members. In general, however, these results support the view that the caregiver's own family social environment plays a role in the quality of care provided. The consistency in results across measures also strengthens the case for the validity of each measure in tapping the area of quality of care provided. Summary Results of this study showed significant relationships between the family environment as measured by the Family Environment Scale and all areas of interest as specified by the research questions.  Child language scores were associated  positively with the FES Cohesion, Expressiveness, and Intellectual-Cultural subscales as was hypothesized.  In addition there were positive associations  between child language scores and the Conflict subscale, and negative correlations with the Independence, Achievement and Moral-Religious subscales. All six measures of socio-economic status produced consistent positive correlations with the Expressiveness, Intellectual-Cultural, and Active-Recreational subscales, and consistent negative correlations with the Control subscale. The strongest FES scores in high socio-economic status families were found in the Intellectual-Cultural and Active-Recreational subscales.  80  Measures of maternal employment satisfaction were positively associated with the Intellectual-Cultural, Organization and Active-Recreational subscales, and negatively associated with the Achievement subscale. Families where mothers were working from economic necessity produced lower scores on the Intellectual-Cultural and Active-Recreational subscales.  Increased maternal hours of work were  negatively associated with the Intellectual-Cultural and postively associated with the Achievement subscales. High total HOME score were positively associated with the FES Organization subscale, and negatively associated with the Moral-Religious subscale. Correlations with attitudes related to childrearing revealed that an adult-centered parenting approach which emphasizes obedience and authority is associated positively with the Moral-Religious and Control subscales and negatively with the Intellectual-Cultural and Active-Recreational subscales.  By contrast, the child-centered attitude that  self-expression in children should be encouraged is positively related to the Expression, Independence and Intellectual-Cultural subscales. Caregivers' home environments show consistently over three measures of quality of family day care provided, that high quality care is associated with positive scores on the Active-Recreational, Independence and Organization scales of the FES. All relationships referred to in this summary are based on Pearson Correlation Coefficients which are significant at the .05 level. Results related to the different research questions indicate that there is considerable overlap, and that environments which are considered optimal in terms of language development, socio-economic factors, conditions of maternal employment, HOME scale ratings, attitudes to childrearing and optimal quality of care for children, have many similar features. These results speak to the complex nature of the family environment, and the validity of studying this concept in an ecological framework where each variable of interest can be examined contextually rather than in isolation.  81  CHAPTER FIVE SUMMARY AND DISCUSSION  This study represents a preliminary exploratory study of the role played by the family environment in an ecological study of children attending family day care; it focuses on select variables of interest as set forth in the research questions developed in Chapter One. This final chapter presents a summary of the findings of the study, followed by a discussion of the implications of the findings and recommendations for further research.  Summary 1) Hypothesis One The first hypothesis of this study was that there would be a positive relationship between child language scores and the Intellectual-Cultural, Cohesion and Expressiveness scales of the FES. As predicted, these subscales all produced significant correlations with child language scores. In addition, significant positive correlations were reported between the language scores and the FES Conflict subscale. Negative associations were found with the Independence, Achievement and Moral-Religious subscales. These findings are important in that they support research reported in the literature. In particular, all studies using the FES in this context (Garfinkle, 1981, Harris, 1982, and Gottfried & Gottfried, 1984) have found positive associations between cognitive measures and the FES Intellectual-Cultural subscale; this study confirms that the exposure of family members to intellectual, cultural and social  82  activities seems to have an important influence on child cognitive competence. In addition, emphasis on actual participation of family members in social and recreational activities (as measured by the Active-Recreational subscale), though not as strong an association, also seems to be connected with higher child language scores. By contrast, a fascinating finding of this study is that the tendency to cast children's activities in an achievement-oriented framework has a strong negative association with child language skills. This supports Garfinkle's 1981 research which found a similar negative correlation between cognitive measures and the FES Achievement subscale. In addition, families which tend to express conflict and anger more openly appear to have higher child cognitive abilities. Families where members are bound by strict religious values seem to produce lower child language scores. These findings suggest that families which emphasize the relationship aspects of family life, including support for family members and encouragement of open expression of feelings, are more facilitative of optimal language development. In addition, exposure to more stimulating ideas and activities appears to be more important than impressing the need to achieve on family members. 2) Hypothesis Two The second hypothesis predicted a relationship between family environment as measured by the FES and six indices of socio-economic status. These were: maternal occupation, maternal years of education, partner's occupation, partner's years of education, family income and a combined family resource factor.  Significant  correlations were for the most part found consistently across all SES measures, decreasing the possibility that any one significant correlation occurred due to chance alone. The strongest findings were between the indices of socio-economic status and the following subscales of the FES, which all yielded strong positive correlations:  83  Intellectual-Cultural, Active-Recreational and Expressiveness. High socio-economic status families also reported a negative correlation with the Control subscale. Other trends associated high SES families with low scores on the Achievement and MoralReligious subscales. This study found a strong positive correlation of .38 between language scores and SES measures. Many similarities in family environment were found between the family which is facilitative of high language scores and the high SES family. The emphasis on exposure to stimulating intellectual, cultural, and social activities, and actual participation in social and recreational activities seem to characterize both the high language score families and the high SES families. Other common denominators included encouragement of open interactions and expression of feelings, and less emphasis on strict religious values or the need to achieve. These findings indicate the complex nature of the family environment as it applies to socio-economic status, and indirectly to cognitive growth. It appears that parents who are better educated expose their families to more of the stimulating activities that seem to be associated with the development of optimum language skills. Although these families may actually have more resources to provide these opportunities for their children, it seems that education level is more highly suggestive than is family income of those families which will provide the enriched environments which have been found to enhance cognitive growth. These results, however, do provide important information as to the nature of the social processes that are associated with high socio-economic status families. These findings are even more significant in light of the ongoing positive association in the literature between socioeconomic status and cognitive development, a relationship which is confirmed by this study.  84  3) Hypothesis Three The third hypothesis predicted that there would be relationships between family environment as measured by the FES and conditions of maternal employment, noteably: maternal attitudes to working mothers, employment satisfaction, hours of work and stress associated with work-childcare conflicts. Results supported findings in the literature (Lamb, 1982, and Hoffman, 1982) that maternal employment satisfaction has an important connection with family functioning.  Strong positive associations were found between high maternal  employment satisfaction and exposure of family members to more stimulating intellectual activities, and participation in a range of activities outside the home. There is also an increased sense of planning and organization in family life, and less of a tendency to emphasize achievement in family members. The intellectual-cultural subscale of the FES, which has been linked consistently in the literature with high child cognitive scores (Garfinkle, 1981, Harris, 1982, and Gottfried & Gottfried, 1984), produces the strongest correlation with maternal employment satisfaction. Other similiarities also exist between high language score families and families with a high level of maternal employment satisfaction, including high scores on the ActiveRecreational and low scores on the Achievement subscales. The implication is that mothers who are satisfied with their work may be able to provide more enriching environments for their children. By contrast, families where the mothers work out of economic necessity report lower correlations with the Intellectual-Cultural and Active-Recreational subscales of the FES, and a stronger emphasis on Achievement. This combination of factors is directly contrary to the type of environment that this study has found to be facilitative of children's cognitive growth. , With high scores on the Control subscale, these families also have many of the characteristics of families of low socio-economic status. It is possible that economic constraint is a critical factor here in making it  85  difficult for such families to provide stimulating activities for their children. The issue of morale, which Hoffman (1982) has found to be critical in families where the mother is working unwillingly, may also be operating in these circumstances. A combination of low morale and increased economic strain could both contribute to the creation of family environments which are not conducive to optimal child development. The findings also suggest that mothers who feel that they are neglecting their children by working tend to provide less facilitative home environments. Maternal hours of work, as predicted in the literature (Hoffman, 1982), seem to be related to family functioning. Longer work hours are associated negatively with the Intellectual-Cultural subscale, and positively with the Achievement subscale. Both these results are indicative of poorer environments for children's cognitive growth. These findings indicate that part-time work allows the mother time and energy to provide an environment which is more conducive to learning. Findings related to stress associated with work-childcare conflict were inconclusive. It is noteworthy that both the practical conditions of maternal employment as well as beliefs and attitudes associated with work are intricately connected with the family social environment. 4) Hypothesis 4 The fourth hypothesis was that there would be a relationship between parental HOME scale scores and FES scores.  As the HOME scale has been used  extensively in the literature to measure home environment, (Bradley & Caldwell, 1976; Elardo et al, 1977; Gottfried & Gottfried, 1984) it was considered relevant in this study to compare results of the HOME scale and the FES. The total HOME score is most important, and is considered a reliable measure of optimal home environment. A strong positive association is found between the total HOME score and the Organization subscale of the FES, and a weaker positive association with the  86  Active-Recreational subscale.  Negative associations were found with the HOME  score total and the Achievement and Moral-Religious subscales.  It appears that  superior home environments as measured by the HOME inventory, are characterised by more planning and organization, more participation in activities outside the home, and less emphasis on achievement and on religious doctrine. 5) Hypothesis 5 The fifth hypothesis was that there would be a relationship between parental attitudes and beliefs associated with childcare and family environment as measured by the FES. A distinction can be made here between adult-centered childrearing values and child-centered values. Adult-centered values included an emphasis on obedience and respect for authority.  Families where this attitude dominated reported  significantly less exposure to intellectual and cultural activities, were more highly disorganized in family functioning, and more bound by rigid rules and strict religious doctrine. These qualities have all been found in this study to be negatively associated with optimal home environments. By contrast, families which adopt a more childcentered approach tend to encourage open interaction and self-sufficiency, and expose their children to more stimulating ideas and activities. A tentative interpretation of these results suggests that families which focus on adult-centered childrearing values may also have a tendency to provide less stimulating and organized environments for their children. Child-centered parenting attitudes seem to be connected with the establishment of more facilitative environments for child development, as long as these views are not so extreme that the children dominate in family functioning.  87  6) Hypothesis 6 The sixth hypothesis examines the relationship between the family environment of the family day care caregiver and the quality of care offered by that setting. The quality of care was measured by three variables, and the consistency in findings across the three measures increases the generalizability of the results. The most important findings relevant to those families which provide the best quality of care are: the strong emphasis on help and support family members provide for each other, the importance of self-sufficiency in all family members, an emphasis on participation in activities outside the home, and a strong tendency to be well organized in their own family functioning. It seems that the family social environment of the caregivers does have an impact on the quality of care they provide. It is noteworthy too that the family environment features of the homes providing higher quality care have many similarities to families which tend to be facilitative of language development and also to those families which have been found to have optimal home environments as measured by the HOME scale.  Discussion These findings are best considered within a larger ecological context. Although the results have been organized according to specific research interests, each indivdual area is related to the others and is also part of a broader picture. As in Bronfenbrenner's model, each developing child is seen as embedded in a family setting, and in turn that setting is embedded in society. At the heart or microsystem are the immediate settings of influence in the lives of the children in this study: family and day care setting. At the mesosystem level is the interaction between family and day care. The exosystem comprises areas which are important to the child though not  88  of immediate impact; notable here are conditions of maternal employment. Factors in the macrosystem or larger social world are also of import; these include massive social changes and economic factors which are encouraging women to return to the workplace in record numbers. These changes have resulted in the increased need for day care which has virtually made day care settings partners with parents in child rearing.  Beliefs derived from the macrosystem, including widespread attitudes  towards working mothers and towards childrearing, also have an impact on family functioning. The family environment is a complex phenomenon which is well suited to study in an ecological or systems framework where the mutual influences between related variables can be considered. Deriving both from the tradition which has stressed the role to be played by the environment in development, and the growing tradition which emphasizes the familial aspects of child development, the family environment is best analyzed in a contextual manner. Bronfenbrenner (1979), Powell (1979) and Conger (1981) all stress the interdependency between child, family, social settings, and influences from the larger society. This concept is clearly applicable to the results of this study which have focused on the role played by the family environment in an ecological study of preschoolers enrolled in family day care settings. Clear associations between child cognitive scores and specific family processes as measured by the FES can best be interpreted in relation to other variables under investigation in the study. Significant positive correlations were found between child language measures and the Intellectual-Cultural, Active-Recreational and Expressiveness subscales, and negative correlations with the Achievement subscales and Moral-Religious subscales.  An examination of the relationships between six measures of socio-  economic status and the FES subscales produced similar results.  This overlap  suggests that socio-economic status may be a mediating factor in the relationship  89  between child cognition and the FES subscales, as suggested by Garfinkle (1981). The correlation of .38 between SES and the child language scores in this study, which supports the estimation of a .4 correlation between cognitive measures and socioeconomic status reported in the literature (Walberg and Marjoribanks, 1976), provides another piece of the puzzle in understanding the complex relationship between child cognition, family environment and socio-economic factors. Here Bronfenbrenner's notion of reciprocity is useful. One factor which has clearly been identified, both in this study and in the literature, as having a positive influence on child language skills is the exposure to intellectual, cultural, social and political activities (as measured by the Intellectual-Cultural subscale of the FES). The quality of social relationships in the home may influence the amount of such stimulation provided for the child. However, socio-economic factors may also be at play. Low income families seem to emphasize this type of activity less than other families, indicating that economic factors may be limiting the access of these families to enriching activities for their children. Parental education also seems to be a strong indicator of those families which ultimately tend to provide more in the way of stimulation for their children; the suggestion here is that better educated parents are more aware of the importance of such factors in a child's development and may provide them more naturally as a matter of course. Hence family environment can be seen as both influencing child cognition and being influenced by socio-economic factors in a circular relationship, where no clear cause and effect can be definitely established. Heredity and parental intelligence undoubtedly play a role as well in this complicated set of relationships, but are beyond the scope of this study. Other results of the study provide further related information. Maternal employment satisfaction also produces significant positive correlations with the Intellectual-Cultural subscale and the Active Recreational subscale, and negative correlations with the Achievement subscale.  Mothers who are working out of  90  economic necessity, however, have precisely the reverse relationships with these three subscales. It appears that the mother's attitude to her work is a critical factor in her ability to establish home environments which are of optimum quality for children's cognitive growth. Once again socio-economic factors may be at play where mothers are working out of economic necessity; in these circumstances, financial restriction may limit the amount of stimulating activities which mothers can make available to their children. It is apparant from the results of this study however, that maternal morale is also associated with the quality of family environment. The reciprocal and non-linear nature of this relationship must be noted: maternal morale may be influenced by as well as influence family environment, and both these factors may be influenced by socio-economic conditions, deriving from the macrosystem, which are beyond the parent's control. Economic recession or sudden unemployment of a partner may be contributing factors in the mother's need to work. In addition, beliefs associated with working mothers have been shown to be related to the quality of family environment provided by the parents, reinforcing Bronfenbrenner's emphasis on the potential impact of beliefs and values deriving from the culture at large in human development. Beliefs are also relevant in considering the relationship of childrearing values to family social environment. Results of this study indicate that families which tend to be more child-centered in their parenting approach have high positive correlations with the Intellectual Cultural subscale of the FES, and that families which emphasize obedience and adult control indicate a significant negative correlation with that subscale.  As this scale has been consistently shown to be associated with better  environments for children's cognitive growth, these results suggest that families who emphasize child-centered values in their parenting provide more stimulating environments for their children. A mediating factor here may be parental education, which also produces very high scores on the Intellectual-Cultural subscale; it is  91  possible that parents with higher education levels may also be more aware of effective parenting strategies, and that these are often the same parents who are more conscious of providing intellectually and culturally stimulating environments for their children. These results also indicate that family environment is an important concept in the family day care setting as well as the child's home; in this case it is the family social environment of the caregivers that has been shown to have a direct impact on the quality of care available to the children. It is important in an ecological context such as this to link the findings for each research questions, in an attempt to discover similarities, patterns and discrepancies. Certain FES subscales have produced consistent results across many of the variables under investigation in this study.  The Intellectual-Cultural subscale produced  amongst the strongest results in the study and is a feature of families which provide superior environments for child language development, and families which have higher income levels, more sophisticated occupations, more resources, and high maternal and paternal education. This subscale is strongly associated with families which indicate a high level of maternal employment satisfaction, is positively associated with parents who have child-centered parenting values and negatively associated with parents who emphasize obedience and authority. A clear picture begins to emerge indicating that the exposure to a wide range of intellectual, cultural, social and political activities is linked not only with child cognitive competence, as indicated in the literature, but also with higher family socio-economic status, higher parental education, higher maternal employment satisfaction, and parenting values which emphasize respect for children rather than obedience and authority. The Active-Recreational Scale is also quite strong throughout the results, and is linked positively with child language scores, and with family income, resources, occupation level, and particularly parental education level. It's linked positively with  92  maternal job satisfaction and a standardised measure of optimal home environment (total HOME score). Families which have more adult-centered values tend to have lower Active-Recreational scores. High scores on this subscale are a feature of the family environments of family day care homes which have been given high quality ratings. It seems that the emphasis on actual participation in social and recreational families is an important aspect of family process which has implications for child cognitive development, effective childrearing, and the quality of care offered by family day care caregivers. It seems to be an important feature of homes with higher income levels, more family resources, and higher parental education levels. The Achievement subscale produced consistent negative correlations throughout the results, corroborating results found by Garfinkle in her 1981 study. Strong negative associations were found between this subscale and the child language scores. Negative associations were also found between the Achievement subscale and the measures of socio-economic status, total HOME scale score, and measures of maternal employment satisfaction. It seemed that the emphasis on achievement increased considerably with the hours of maternal employment, indicating that mothers who had less time and energy for their children displayed a stronger achievement orientation. Findings in this study then consistently support the notion that parental pressure and the casting of activities in an achievement-oriented framework has a deleterious effect on cognitive development and the optimal quality of home environment provided. It is possible that parents who are under more pressure themselves (with increased hours of work, or high job dissatisfaction) place more pressure on their children. Ultimately it seems that, as reported elsewhere in this study, it is in fact the exposure to more stimuli and activities that is beneficial to children rather than a press for achievment. Another scale which has consistent results across the study is the Expressiveness subscale.  Expressiveness is positively associated with child  93  language scores, with most measures of socio-economic status, particularly family resources and income, with maternal employment satisfaction, and with childcentered parenting values. Thus the ability to act openly and express feelings direcdy is a process which is connected with the establishment of a family environment which is conducive to child cognitive competence, and also tends to be associated with families of higher socio-economic status. It is possible that families that experience less financial strain also have more ease in their communications; this may be particularly so if the mother is satisfied with her own employment situation. The Moral-Religious subscale, which reflects an emphasis on ethical and religious values, quite consistently produced negative correlations with the variables under investigation. It is associated negatively with child language scores, some of the measures of socio-economic status, and the HOME scale total score and subscales. It is associated positively with the attitude that mothers are neglecting their children if they work and with parenting attitudes which reflect strict submission to adult authority. A related FES subscale, Control (its high intercorrelation with the Moral-Religious subscale is noted in Table 2), is also positively associated with these same parenting attitudes. Control scores also produced significant negative correlations with all indices of socio-economic status.  This subscale, which  emphasizes the use of rigid rules to run family life, clearly has some connection with the strict emphasis on religious values measured by the Moral-Religious subscale. It appears from these findings, as well as the results related to obedience-centered parenting values, that an emphasis on strict regulation of family life or of the lives of individual members has an essentially deleterious effect on the quality of the family environment established. By contrast, the Organization subscale, which measures organization and structure in family functioning (as opposed to the use of rigid rules) tends to be positively associated with optimal home environments. Positive correlations are  94  found between this subscale and the HOME scale, and measures of maternal employment satisfaction. It addition the Organization scale is very highly correlated with all three measures of day care quality in family day care caregivers' homes. It appears that well organized homes are superior environments for children, and that mothers who are happier in their employment situations tend to be more organized. This characteristic is especially important for family day care caregivers, where effective organization in their own families is carried over to the day care environment and clearly contributes to a higher quality of care. These results indicate the complex nature of family environment as it influences the individual developing child in her immediate family setting, extends to the family day care setting, and is at the same time influenced by factors in the larger world including parental employment, beliefs, values and attitudes as well as social changes and events. The notion of reciprocity, or an ongoing process of mutual accomodation between the individual and the environment, is everpresent within this ecological understanding of human development. Change is another feature of such an approach, as is the concept of a non-linear causality, where the variables affecting a young child's life are in fact interdependent, and cannot be clearly separated into a cause-and-effect framework. By shedding light on some of the processes that are associated with the family environment, this study has revealed one more layer in an emerging understanding of the complex pattern of interaction between individual, family process and social settings that contribute to a child's development.  Recommendations Future studies of this nature would do well to use a larger N to facilitate the use of multivariate statistical analyses with the data obtained. The correlational data presented in this study can really only be considered preliminary to the kind of information a more sophisticated analysis would reveal. In light of the highly complex  95  nature of the variables under consideration and their relationships to each other, multiple regression, cluster analysis and canonical analysis are recommended as future steps. Any single finding from this thesis might constitute the central focus of a future study. Examples are: the relationship between the parental press for achievement and child cognitive development;  the relationship between socio-economic factors,  child language skills, and the family environment; the role of family organization and rules in establishing optimal environments for child development, and the relationship between maternal  stress associated with work, maternal morale, and family  dynamics. A follow-up study might also focus on other variables that are relevant to family social environment.  Parenting situations (single parent, two-parent and  blended families), ethno-cultural background and family stress are all potentially interesting areas to explore in a similar ecological context. Once again, a larger N would be preferable in such a study. Although the FES has excellent psychometric properties, any quantitative measure of a complex and changing social phenomenon such as family environment has built-in limitations.  A future study might utilize direct observation of the  environment to provide further support for FES-derived data, as well as additional qualitative information. The results of this study have provided important information on the family social processes associated with child cognition, level of socio-economic status, conditions of maternal employment, and attitudes related to childrearing and maternal employment. However, naturalistic data based on direct observation of the environment would go one step further in determining precise family interaction patterns that are associated with the processes that have been identified by the FES. The use of other standardized instruments associated with family environment, in addition to the FES, is also recommended.  96  REFERENCES Baltes, P.B. and Schaie, K.W. (1973). Life-span developmental psychology: Personality and socialization. New York: Academic Press. Bee. H.E., Barnard, K.E., Eyres, S.J., Gray, C. A.,Hammond, M.A., Spietz, A.L., Snyder, C , & Clark, B. (1982). Prediction of IQ and language skill from perimatal status, child performance, family characteristics and mother-infant interaction. Child Development. 53. 1134-1156. Belsky, J. (1978). The effects of day care: A critical review. Child Development 49. 929-949. Belsky, J. (1981). Early human experience: A family perspective. Developmental Psychology 17 (1), 3-33. Belsky, J. (1984). Two waves of day care research: developmental effects and conditions of quality. In R.C. Ainslie, (Ed.), The child and the day care setting, (pp 1-34). Toronto : Praeger. Belsky, J., Steinberg, L.D., and Walker, A. (1982). The ecology of day care. In M.E. Lamb (Ed.), Nontraditional families: Parenting and child development, London: Lawrence Erlbaum Associates Publishers. Boehm, A. (1985) . Review of home observation for measurement of the environment. In J.V.Mitchell, (Ed.) The ninth mental measurements yearbook. Vol 1. Lincoln Nebrasks: University of Nebraska Press Bowlby, J. (1969) .Attachment and loss Vol. 1: (Attachment). London: Hogarth Press. Bradley R.H. and Caldwell, B.M. ( 1976) Early home environments and changes in mental test performance in children from 6 to 36 months. Developmental Psychology. 12 (2). 93-97. Bradley, R.H. and Caldwell, B.M. (1976) The relation of infants home environments to mental test performance at fifty-four months: a follow-up study. Child Development 47 (4). 1172-1174). Bradley, R.H.and Caldwell, B.M. (1979) Home environment and cognitive development in the 2 years: A cross-lagged panel analysis. Developmental Psychology. 15. (31 246-250. Bradley, R.H. and Caldwell, B.M. (1980). The relation of home environment, cognitive competence and IQ among males and females. Child Psychology. 51.. 11401148.  97 Bradley, R.H., Caldwell, B. and Elardo, QR. (1977). Home environment, social status and mental test performance. Journal of Educational Psychology. 69. (6). 697701. Bronfenbrenner, U. (1979). The ecology of human development. Cambridge , Mass: Harvard University Press. Bronfenbrenner, U. (1986). Ecology of the family as a context for research perspectives. Developmental Psychology. 22 (6), 723-742. Bronfenbrenner, U. & Crouter, A.C. (1983). The evolution of environmental models in developmental research. In P.H. Musson (Ed.), Handbook of child psychiatry. (4th ed., Vol. 1, pp.357-414). New York : John Wiley & Sons. Bronfenbrenner, U. and Moen, P. (1982). Child, family and community. In Parke, R.D. (Ed.) Review of child development research. (Vol 7, pp 283-320). Busch-Rossnagel, N.A. (1985) Review of family environment scale. In Mitchell, J.V. (Ed.), The Ninth Mental Measurement Yearbook. Vol 1. Lincoln Nebraska: University of Nebraska Press, pp.408-409. Caldwell, B.& Bradley, R.(1979). Home observation for measurement of the environment. Little Rock, AK: Center for Child Development and Education, University of Arkansas. Carew, J.C. (1980) Experience and the development of intelligence in young children at home and in day care. Monographs of the Society for Research in Child Development.45. (6-7, Serial No. 187), 1-89. Clarke-Stewart, K.A. (1973).Interactions between mothers and their young children: characteristics and consequences. Monographs of the Society for Research in Child Development. Vol. 38. (6-7. Serial No. 153.) Cochran, M. (1977). A comparison of group day and family child-rearing patterns in Sweden. Child Development. 48.. 702-707. Conger, R.D. (1981). The assessment of dysfunctional family systems. In B.B. Lahey and A. E. Kazdin (Eds.), Advances in clinical child psychology. (Vol. 4, pp 199204.) Cummings, J.A. (1985). Review of expressive one word picture test. In Mitchell, J.V. (Ed.) The Ninth Mental Measurement Yearbook. Vol 1. Lincoln , Nebraska: University of Nebraska Press, pp 565-566. Dunn, L.M. (1979). Peabody picture vocabulary test-revised. Circle Pines, MN: American Guidance Services. Elardo, R. , Bradley, R. and Caldwell, B.(1975) The relation of infants' home environments to mental test performance from six to thirty-six months. Child Development. 46. 71-76  98 Elardo, R., Bradley, R & Caldwell, B.(1977). A longitudinal study of the relation of infants' home environments to language development at age three. Child Development. 48. 595-603. Elder, G.H. (1974). Children of the great depression. Chicago: University of Chicago Press. Emlen A.C., Donoghue, B.A., and LaForge, R. (1971). Child care by kith: A study of the family day care relationships of working mothers and neighbourhood caregivers. Portland: Children's Bureau, Office of Child Development. Fowler, W. (1978). Day care and its effects on early development. Toronto: Ontario Institute for Studies in Education. Gardner, M.F. (1979). Expressive one-word picture vocabulary test. Academic Therapy Publications. Garfinkle, A.S. (1981). Genetic and environmental influences on the development of Piagetian logico-mathematical concepts and other specific cognitive abilities: a twin study. Acta Geneticae Medicae Gemellologiae. 31. 10-61. Goelman, H. (In Press). A study of the relationships between structure and process variables in home and day care settings on children's language development. In A.R. Pence (Ed.) The Practice of Ecological Research: From Concepts to Methodology. Teacher's College Press. 30 manuscript pages. Goelman, H. & Pence, A. (1985). Toward the ecology of day care in Canada: a research agenda for the 80's. Canadian Journal of Education. 10 (4) , 323-343. Goelman, H. (1986). The ianguage environments of family day care. In S. Kilmer (Ed.) Advances in Early Education and Day Care. Vol. 4. New Haven: JAI Press, 153-179. Goelman, H. and Pence, A. (1986). Vancouver Day Care Research Project. Grant proposal submitted to the Social Sciences & Humanities Research Council of Canada. Goelman, H. and Pence, A. (1987a). The relationship between family structure and child language development in three types of day care. In: S. Kontos & D. Peters (Eds.) Advances in Applied Developmental Psychology. Volume 2. Norwood, NJ: Ablex, 129-146. Goelman, H. and Pence, A.R. (1987b). Effects of child care, family and individual characteristics on children's language development. In D. Philips (Ed.), Quality in Child Care: What Does Research Tell Us? NAEYC Monograph Series. Washington, DC: National Association for the Education of Young Children, 89-104.  99 Golden, M. and Birns, B. (1976) .Social class and infant intelligence. In M .Lewis (Ed.), Origins of intelligence in infancy and early childhood. New York: Plenum Press Gottfried, A.W. (Ed.). (1984) . Home environment and early cognitive development, longitudinal research. Toronto: Academic Press, Inc. Gottfried, A.W. and Gottfried, A.E.(1984).Home Environment and cognitive development in young children of middle-socioeconomic-status families. In A.W. Gottfried (Ed.) Home environment and early cognitive development, longitudinal research. Toronto: Academic Press, Inc. Harms, T. and Clifford, R. (1980). The early childhoood environment rating scale. New York: Teacher's College Press. Harris, E.H. (1982) . Genetic and environmental influences on reading achievement: a study of first and second grade twin children. Acta Geneticae Medicae Gemellologiae. 31. 64-116. Hoffman, L.W. (1963) Effects on children: Summary and discussion. In F.I. Nye and L.W. Hoffman (Eds.) The employed mother in america. Chicago:L Rand-Nally. Hoffman, L.W. (1982). Work , family , and the socialization of the child. In R.D. Parke, (Ed.) Review of child development research. Vol. 7: The family. Chicago: University of Chicago Press. Hunt, J. McV. 1961. Intelligence and experience. New York: Ronald Press. Jensen, A.R. (1969). How much can we boost IQ and scholastic achievement? Harvard Educational Review. 39. (T). Johnson, L. (1978). Taking care. Part II. Report on Project Child Care. Toronto, Social Planning Council of Metropolitan Toronto. Kagan, J. (1979). Family experience and the child's development. American Psychologist. 34 (10), 886- 891. Keniston K., & Carnegie Council on Children. (1977). All our children. New York: Harcourt, Brace Javonovich. Lamb, M.E. (1982). Maternal employment and child development: a review. In M.E. Lamb (Ed.) Nontraditional families: Parenting and child development. Hillsdale, N.J.: Lawrence Erlbaum Associates, Publishers. Langdon, E.L., Davis, B. Carter, L., Flint, D., Owen, J-, Rickert, M., & Taylor, E. (1981). Separation of home intellectual environment and maternal IQ as determinants of child IQ. Developmental Psychology. 17 (5), 532-541. Lerner, R.M. (1976). Concepts and theories of human development. London: AddisonWesley Publishing Co.  100 Lidz, T. (1970). The family as the developmental system. In E.J. Anthony and C. Kroupernik (Eds.) The child in his family. (Vol 1) . New York: John Wiley and Sons. Long, F., Peters, D.L., and Garduque,L. (1985). Continuity between home and day care: A model for defining relevant dimensions of child care. In I.E. Sigel (Ed.), Advances in applied developmental psychology. (Vol. 1) . Norwood. N.J.: Ablex Publishing Co. Longstreth, AE.L., Davis, B., Carter, L., Flint, D., Owen, J., Rickert, M., and Taylor, E. Separation of home intellectual environment and maternal IQ as determinants of Child IQ. Developmental Psychology. Vol. 17 (5), 532-541. MacPhee, .D., Ramey, C.T., & Yeats, K.O. (1984) Home environment and early cognitive development; implications for intervention. In Gottfried, A.W. (Ed.), Home environment and early cognitive development, longitudinal research. Toronto: Academic Press. Marjoribanks, K. (1972). Environment, social class and mental abilities. Journal of Educational Psychology. 63. (2). 103-109. McCall, R.B., Applebaum, M.L, Hogarty, P.S. (1973) . Developmental changes in mental performance. Monographs of the Society for Research in Child Development. 38. (3\ 1-83. McCallum, S. (1985). Review of Peabody Picture Vocabulary Test. In J.V. Mitchell (Ed.), The ninth mental measurements yearbook. Vol. 2. Lincoln Nebraska : University of Nebraska Press. Miller, P.H. (1983). Theories of developmental psychology- San Fransisco: W.H. Freeman and Co. Minuchin, S. (1970). The use of an ecological framework in the treatment of a child. In E.J. Anthony and C. Kroupernik (Eds.), The child in his family. (Vol 1). New York: John Wiley and Sons. Moore, T. (1964); Exclusive early mothering and its alternatives: the outcome to adolescence; Scandinavian Journal of Psychology. 16. 255-272. Moore, T. (1968) Language and Intelligence: a longitudinal study of the first eight years II; Environmental correlates of mental growth. Human Development. 11. 1-24. Moos, R.H. (1986) .Family environment scale manual. 2nd edition . Palo Alto Calif.: Consulting Psychologists Press. Parke, R.D. 1978. Children's home environments: social and cognitive effects, in I. Altman and J.F. Wohlwill (Eds.) Children and the environment . New York: Plenum Press.  101 Pence, A.R. & Goelman, H. Who cares for the child in day care? An examination of caregivers from three types of care. Early Childhood Research Quarterly. Vol. 2, 315-334. Pence, A. and Goelman H. ( In Press). Silent Partners: The parents of children in three types of daycare. Early Childhood Research Quarterly. Vol. 2. (2), 103118. Plomin, R. , DeFries, J.C., & Loehlin, J.C. Genotype-environment interaction and correlation in the analysis of human behaviour. Psychological Bulletin 84. (2). 309-322. Powell, R.D. (1979). Family-environment relations and early childrearing: The role of social networks and neighbourhoods. Journal of Research and Development in Education. 13 (I). 1-11. Powell , R.D. (1980). Towards a socioecological perspective of relations between parents and child care programs. In S. Kilmer (Ed.) Advances in early education and day care (Vol I). Greenwich, Conn.: JAI Press. Prescott, E., Jones, S., Kritchevsky, S. Milich, C. and Haselhoef, E. (1973). Who thrives in group day care? Assessment of child-rearing environments: An ecological approach. (Part 1 of the final report). Pasadena, Ca: Pacific Oaks College. (ERIC Document Reproduction Service No. ED 076 228) Scarr, S. & McCartney, K. (1983). How people make their own environments: A theory of genotype - environment effects. Child Development 54. 424-435. Scarr, S. and Weinberg, R.A. (1978). The influence of 'family background' on intellectual attainment. American Sociological Review. 43. 674-692. Spivack, G.M. (1985). Review of expressive one-word picture vocabulary test. In J.V. Mitchell (Ed). The Ninth Mental Measurements Yearbook. Vol 1. Lincoln. Nebraska, University of Nebraska Press. Stuckey, M.F. , McGhee, P.E. and Bell, N.J. Parent-child interaction: the influence of maternal employment. Developmental Psychology. 18. (4). 635-644. Tulkin, S.R. (1973). Social class differences in attachment behaviours of ten-month-old infants. Child Development. 44. 171-174. Tulkin, S.R. and Kagan, J. (1972). Mother-child interaction in the first year of life. Child Development. 43. 31-41. Walberg, H.J. and Marjoribanks, K. (1976) Family environment and cognitive development: twelve analytic models. Review of Educational Research. 46. (41. 527-551) Wattenberg, E. (1977). Characteristics of family day care providers: implications for training. Child Welfare. 16. (4). 211-227.  102  Willerman. L. (1979). Effects of families on intellectual development. American Psychologist. 10. 923-97.9  103  APPENDIX 1  FES SUBSCALES A N D DIMENSION  DESCRIPTIONS  DEFINITIONS OF FES SUBSCALES, ACCORDING TO THE MANUAL, ARE AS FOLLOWS:  RELATIONSHIP DIMENSIONS  1) COHESION  the degree of commitment, help and support family members provide for one another.  2) EXPRESSIVENESS  the extent to which family members are encouraged to act openly and to express their feelings directly.  3) CONFLICT  the amount of openly expressed anger, aggression and conflict among family members.  PERSONAL GROWTH DIMENSIONS  4) INDEPENDENCE  the extent to which family members are assertive, are self-sufficient, and make their own decisions.  5) ACHIEVEMENT ORIENTATION  the extent to which activities (such as school and work) are cast into an achievement -oriented framework.  104  6) INTELLECTUAL-  the degree of interest in political, social, intellectual and  CULTURAL  cultural activities.  ORIENTATION  7) ACTIVE-  the extent of participation in social and recreational  RECREATIONAL  activities.  ORIENTATION  8) MORAL-RELIGIOUS EMPHASIS  the degree of emphasis on ethical and religious issues and values  SYSTEM MAINTENANCE DIMENSION  9) ORGANIZATION  the degree of importance of clear organization and structure in planning family activities and responsibilities.  10) CONTROL  the extent to which set rules and procedures are used to run family life. (Moos, 1986, p.2.)  105  APPENDIX 2  THE FOLLOWING ARE ITEMS FROM THE PARENT  QUESTIONNAIRE  ADMINISTERED BY THE VANCOUVER DAY CARE RESEARCH PROJECT:  MATERNAL ATTITUDES TO WORK a. I like the work I am doing. b. I like the daily routine of working. c. It is necessary to our family's economic survival that I work. d. In order to get a little bit more that the basic necessities, I work. e. It gives me a feeling of independence and self-satisfaction to work. f. My spouse-partner fully supports me working outside the home. g. It doesn't really do most children harm to spend the day away from their mother. h. Mothers shouldn't work unless they absolutely have to. i. It is important for a child to have as a role model, a mother who works outside the home. j. On the whole, I think that most women can be better mothers if they go out to work, k. Fathers have not been involved enough in parenting in the past and ought to be. 1. A mother who works misses the experience of seeing her children growing up. m. Mothers who work neglect their children as a result.  CHILDREARING ATTITUDES AND PRACTICES a. Often it isn't the parent, but the child who knows what's best for him or her. b. Children who can't take being teased are over-sensitive. c. When babies cry, it is always because they need something.  106  d. Obedience and respect for authority are the most important virtues children should learn. e. Children should learn early that their parents' needs come first. f. You can't spoil a child by trying to understand him or her. g. Children should always feel grateful for what their parents provide. h. It is always important to encourage a child's self-expression, even if one doesn't like what comes out. i. Children should never take responsibility for something without their parents' permission.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0053747/manifest

Comment

Related Items