m1“.:¢1—;r= ‘ “‘.".L’£ , ‘ - .. .. «n» i“ “A. 4 -'x2 ‘2'“? agjlififin .. ,H m. rut; as: .4. r m. 11‘ 4v “Wu: "we“ .~ .‘..1 .n 1.'/ ""1” L».' V H Nun-y. u v .—"‘.-4u w. .ru 4‘. uu‘ng‘l u s : u c": 4 ‘ . we. mam . Q‘IK‘ ‘rw w e. n r 4: ‘ 9 “av 4:21;} 45:.” ‘ air“ i g S?!) , . w. _r. V 1 ,..~'. 5. '1‘: . ‘égt 9 r. . « 'asr1‘r « .u-wmn u‘u‘lw‘r ‘ '41 ,n a. u ‘ .JL. “ms -. 4 .(i , “man, “am An: . .4 x . ' 32!! p.453" r.“ . mm ra' - .1 [WARE . ‘ :‘n ;." fin. n‘s I; .L 4: ac: .,. . 2,9“- .' ‘4 u , ' A V r r 1 'V. 1‘ - (try, u- . p I 'w . .-. tr,“ I", ' ‘ 5- “- ‘. 1. , <: ”ans 5' 1123253 " m . ;: ~ . Va .5 . r1, ‘1'" ~ 1,. T...:-‘ I”, _ 1 . 4 ‘7'! a 1r 1 J. c 4.19.5“ 15:15” i M » 1- ‘ 11.3.»: -.<. .wfl. ‘u'Jv‘n‘ n. . . g» '2- . . ., -m «rum ...‘, . . nun!!! mv- 'Nffli ,w k;- 73 « :- '4 v u my; ”ammun‘. n- ’v‘“' a: . ..,.- ¢(-‘ I x n “7. » '. 7‘4‘. . .- mm. W. . ' "1.11" "v. V.‘ .w A m; r 3.14.“; ’- J.A!u r ¢ ..,.._ um" J. W ,. 3...“ a a. ~. . ‘ r NM, - 1 v ”u... . ., ,,..;V ,7 . 4 I L .1. w This is to certify that the dissertation entitled "Laro Tayo": Parent-child and Peer Play Activities of Filipino Children and Related Variables presented by Marita Depante Bernardo has been accepted towards fulfillment of the requirements for Ph.D. degree in Psychology 5am JAZCWM Dr. Ellen A. Strommen Major professor Date June 16, 1994 MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771 llliliiiilill LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE 2M "LARO TAYOI": PARENT—CHILD AND PEER PLAY ACTIVITIES OF FILIPINO CHILDREN AND RELATED VARIABLES BY Marita Depante Bernardo A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1994 ABSTRACT "LARO TAYO!"‘: PARENT-CHILD AND PEER PLAY ACTIVITIES OF FILIPINO CHILDREN AND RELATED VARIABLES BY Marita Depante Bernardo This study identified differences in play activities engaged in by children varying in sex, age and socioeconomic status. Frequency of parent—child play was correlated with parental acceptance and child’s self—concept. Sex, age and socioeconomic differences in parent—child play activities were described. Lastly, differences in play and other social behaviors of children varying in popularity and self— concept were identified. Four hundred thirty—seven first and fourth graders from two public and two private coeducational schools completed a Home Play Survey, a sociometric instrument, the Child Parental—Acceptance Rejection Questionnaire and the Pasao Self—Concept Scale. Based on the results of the sociometric measure, 32 selected children were observed during recess. Results of the Home Play Survey showed that there were age, sex and socioeconomic differences in play activities at home. These differences could be accounted for by social and lifestyle variations among families and by cognitive and social development of children. Among the fourth graders, more frequent parent—child play was associated with more parental acceptance indicating that parent—child play could be a cause or an indication of positive parent—child relations. Among the first graders, more frequent parent— child play was associated with more parental rejection and higher self—concept. The latter indicates that this interaction can enhance competencies and relationships. The PARQ scores could have a different interpretation for younger children and that moderate parental acceptance is associated with more frequent parent—child play. Different models for predicting self—concept, with parent—child play and parental rejection as predictors were separately presented for Grade One and Grade Four children. Sex-typing was observed in parent—child play activities. There were fewer play activities among the lower class older children than among the middle class and the lower class younger children. The data from observations showed that popular children were more social than unpopular children. Unpopular children with high self—concept tended to be isolated. The play activities and interactions of unpopular children with low self—concept were more motorically active and aggressive compared to play of popular children. Recommendations regarding methodology, applications and topics for future studies were made. 'k Filipino for "let’s play" Copyright by MARITA DEPANTE BERNARDO 1994 To my parents, Mariano and Socorro Bernardo, with all my love. ACKNOWLEDGMENTS I express my sincerest gratitude and love to the following people: I thank Dr. Ellen Strommen, the perfect adviser, for her insightful comments on drafts, for her patience in reading countless revisions, for her generosity, and most especially, for her ever—present words of encouragement. To my family, my parents, sisters and brothers—in—law, I am forever grateful for their moral and financial support. Simply by being there, they have given me courage to face difficult tasks. I would like to thank my best friend, Tito, for the restful breaks provided by him during this tiring process. I am forever indebted to my American family, the Pennings, for their gracious hospitality. Their lifestyle provided inspiration for this dissertation. To my dear friends at De La Salle University, and University Reformed Church, I thank them for their prayers. I thank my dissertation committee for an enjoyable defense, a truly generous exchange of ideas from colleagues. I especially thank Dr. Ralph Levine for imparting statistical techniques that have enriched the analysis. vi I am grateful for the financial support provided by De La Salle University (Manila, Philippines). I appreciate the patience and perseverance of the research assistants, and the cooperation of the principals, teachers and students of Seton School, Pamplona Elementary School, Pasay City Academy and Apelo Cruz Elementary School. And most of all, I would like to express my everlasting gratitude to the Lord God Almighty who was with me in this endeavor, from beginning to end. He has given me comfort and assurance in times of difficulty. He has been faithful and was a constant source of strength. To Him belongs the glory! TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES Chapter 1. Review of THE PROBLEM AND REVIEW OF LITERATURE Related Literature Significance of play Development of social skills Social relationships: Peer play . Social relationships. Parent child play Mental health: The self . . . Sex roles Summary of significance of play Categories of play Play Content and complexity of play Physical activity . Symbolic play Games with rules Peer social context Adult social context Settings Importance of categorization and contextualization patterns according to age, sex and social class and personality Age differences . . . Sex differences Social class differences Personality differences Summary of play patterns The Philippine experience Filipino children’ s play Parent— child relations Methodological issues Settings . Participants and their roles Subjects . . . . . . . viii Page xiii xvii Conceptual Framework . . . . . . . . . . . 34 Variations in play activities . . . . . . . . . . 34 Play and popularity . . . . . . . . . . . 37 Play with parents, parent—child relationship, and self—concept . . . . . . . . . . . . . . . . 38 Summary of framework . . . . . . . . . . . . . . . 39 Statement of the Problem . . . . . . . . . . . . . . . 39 Hypotheses . . . . . . . . . . . . . . . . . . 40 Definition of Terms . . . . . . . . . . . . . . . . . . 41 Scope and Limitations . . . . . . . . . . . . . . . . . 46 Significance of the Study . . . . . . . . . . . . . . . 47 Chapter 2. METHODOLOGY . . . . . . . . . . . . . . . . 49 Subjects . . . . . . . . . . . . . . . . . . . . . . . 50 Instruments . . . . . . . . . . . . . . 53 Sociometric Rating Measure . . . . . . . . . . 54 Scoring the sociometric scale . . . . . . . 55 Rationale for choosing peer rating . . . . . 57 Home Play Survey . . . 59 Child Parental Acceptance— Rejection Questionnaire. 61 Scoring the PARQ . . . 62 Reliability and validy of the Child PARQ . . 62 Pasao Pictorial Self— concept Rating Scale . . . . 65 Scoring the Pasao Scale . . . 68 Reliability and validity of the Pasao scale . 69 Peer Play Scale . . . 7O Inter— observer reliability . . . . . . . . . 72 Procedure . . . . . . . . . . . . . . 73 Pre data- gathering phase . . . . . . . . . . . . . 73 Data gathering phase . . . . . . . . . . . . . . . 73 Data Analysis . . . . . . . . . . . . . . . . . . . . . 75 Problem #1 . . . . . . . . . . . . . . . . . . . 75 Problem #2 . . . . . . . . . . . . . . . . . . . 76 Problem #3 . . . . . . . . . . . . . . . . . . . 77 Chapter 3. RESULTS . . . . . . . . . . . . . . . . . . 79 Play activities at home . . . . . . . . . . . . 79 Sex differences in play at home . . . . . . . . . 79 Age differences in play at home . . . . . . . 83 Socioeconomic/school differences in play activities at home . . . . . . . . . . . 86 Parent— child play . . . . . . . . . . . . . . . . 89 Playing with parents, parental warmth and self— concept . . . . . . . . . . . . . . . . . 9O ix Playing with parents, parental warmth and self concept: Grade Four data . Correlations between parental warmth and self- concept: Grade Four data Predicting self—concept: Grade Four data. Playing with parents, parental warmth and self— concept: Grade One data Correlations between parental warmth and self— concept. Grade One data Predicting self—concept: Grade One data Other results: Parent— child play activities Sex comparisons in parent- child play Age and school differences in parent- child play . . . . . . Popularity, play and self— concept Popular and unpopular children . Play behaviors of popular and unpopular children Other behaviors of popular and unpopular children . . . Behaviors of high vs. low self— concept children Play behaviors of children with high vs low self— concept . . Other behaviors of high vs. low self~concept children . . . . . . Play and other social behaviors of children varying in popularity and self—concept Relationship between popularity and self—concept Chapter 6. DISCUSSION Play activities at home Sex differences Age differences . Socioeconomic differences Parent- child play and related variables . . . Correlates of parent— child play: Grade Four. Parent— child play and parental acceptance Parent— child play and self— concept: Grade Four Predicting self— concept: Grade Four . Correlates of parent—child play: Grade One . Parent—child play and parental acceptance Parental acceptance and child’s self—concept Predicting self-concept: Grade One 91 96 97 104 106 106 110 115 118 126 127 127 129 130 136 137 137 138 139 141 143 144 144 145 146 148 148 151 152 Parent- child play activities . . . 153 Sex, age and socioeconomic differences in parent— child play . . . . . 153 Variety in parent— child play: Issue of defining play . . . . . . . . . . 157 Peer play, popularity and self— concept . . . . 158 Behaviors of popular and unpopular children . . . 159 Interaction between popularity and self concept. . 160 Chapter 5. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS . . 162 Summary . . . . . . . . . . . . . . . . . . . . . . . 162 Problem . . . . . . . . . . . . . . . . . . . . . 162 Methodology . . . . . . . . . . . . . . . . . . . 162 Results . . . . . . . . . . . . . 163 Play activities at home . . . 163 Parent— child play, parental acceptance and child’s self— concept . . . . . . . . . . 164 Predicting self— concept . . . . . . . . . . . 164 Parent— child play activities . 165 Popularity, self~ concept and social behaviors 165 Conclusions . . . . . . . . . . . . . . . . . . . . . 166 Play activities at home . . 166 Parent- child play, parental acceptance and self— self— concept . . . . . . . . 167 Parent- child play activities . . . . 168 Popularity, self— concept and social behaviors . . 169 Recommendations . . . . . . . . . . . . . . . . . . . . 170 Methodology . . . . . . . . . . . . . . . . . . . 170 Applications . . . . . . . . . . . . . . 171 Topics for future studies . . . . . . . . . . . . 173 LIST OF APPENDICES A Sociometric Rating Measure . . . . . . . . . . . . 175 B Home Play Survey . . . . . . . . . . . . . . . . . 176 C Child PARQ (copyright address) . . . . . . . . . . 178 D Pasao Pictorial Self—Concept Scale . . . . . . . . 179 (copyright address) E Play Observation Guide . . . . . . . . . . . . . . 180 F Consent letter to parents . . . . . . . . . . . . 181 xi G Explanation to parents H Consent request to children I Consent request to children J Categories of behaviors K Categories of play behaviors L Description of Filipino outdoor games (general study). (observational study). M Correlations between Grade and Scales of the PARQ and the Pasao Self-Concept Scale N Correlations between sex, socioeconomic status and scales of the PARQ and the Pasao self—concept scale: O Correlations between sex, Grade Four data socioeconomic status and scales of the PARQ and the Pasao self— concept scale: Grade One data P Three-way analysis of variance: age, economic status as independent variables, and PARQ and Self-concept scales as dependent variables REFERENCES xii sex and socio— 182 183 184 185 186 187 188 189 190 191 198 LIST OF TABLES TABLE TITLE 10 11 12 13 14 15 Distribution of Classes from each Grade Level per School Number of Subjects in Play Survey Data Distribution of Subjects in Observations Minimum and Maximum Scores for the PARQ Internal Consistency Reliability Coefficients (alpha) for PARQ scales and Total Scale (Grade One sample). . . . . . . . . . . . . . . . . . Internal Consistency Reliability Coefficients (alpha) for PARQ scales and Total Scale (Grade Four Sample). . . . . . . . . . . . . . . . . . . Correlations Between PARQ Scales and Self—concept in Relation to Significant Others (Grade Four only) Reliability Coefficients for the Pasao Pictorial Self—Concept Scale Validity Coefficients for the Pictorial Scale. Play Activities of Boys and Girls at Home. Play Activities of First Grade Girls at Home Play Activities of Fourth Grade Girls at Home. Play Activities of First Grade Boys at Home. Play Activities of Fourth Grade Boys at Home Grade One and Grade Four Children Reporting Ball Games and Playing House. xiii Page 50 52 53 63 64 66 67 69 7O 80 82 83 84 85 85 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Number of Children Playing Videogames by Age and Socioeconomic Status . . . . . . . . Dolls Mentioned by Public and Private School Girls. . . . . . . . . . . . . . . . . Outdoor Games of Public and Private School Children Number of Children Playing Ball Games by Age, Sex and Socioeconomic Status . . . . . . Mean Scores in the Pasao and PARQ (Grade Four data) Mean Scores in the Pasao and PARQ (Grade One data) Correlations Between Frequency of Parent-son Play and Scores in the PARQ: Grade Four data (n = 82) . . . . . . . . . . . . . . Correlations Between Frequency of Parent-daughter Play and Scores in the PARQ: Grade Four data (n = 99) . . . . . . . . . . . . . . Correlations Between PARQ scales and Self-Concept Scales: Grade Four data (n = 181). Correlations Between Parent—child Play, Parental Rejection, and Self-concept: Grade Four Girls Beta Weights/Partial Correlations Between Parent— child Play, Parental Rejection, and Self—concept: Grade Four Girls . . . . . . . . . . . . . Correlations Between Mother-child Play, Maternal Rejection, and Self-concept: Grade Four Boys Correlations Between Father—son Play, Paternal Rejection, and Self~concept: Grade Four Boys Beta Weights/Partial Correlations Between Parent- child Play, Parental Rejection, and Self-concept: Grade Four Boys . . . . . . . . . . . Correlations Between Frequency of Parent—child Play and Scores in the PARQ: Grade One data (n = 187). . . . . . . . . . . . . . xiv 86 87 88 89 92 93 94 95 98 99 101 102 103 103 105 31 32 33 34 35 36 37 38 39 4O 41 42 43 44 45 46 47 48 Correlations Between PARQ Scales and Self-concept Scales: Grade One data (n = 187) . . . . . . . . . 107 Correlations Between Parent—child Play, Parental Rejection, and Self—concept: Grade One (n=187) . . 109 Beta Weights/Partial Correlations Between Parent- child Play, Parental Rejection, and Self—concept: Grade One . . . . . . . . . . . . . . . . . . . . 110 Mother-daughter Play Activities. . . . . . . . . . 111 Mother-son Play Activities . . . . . . . . . . . . 112 Father-daughter Play Activities. . . . . . . . . . 113 Father-son Play Activities . . . . . . . . . . . . 114 Play Activities of Boys and Girls with Mothers and Fathers (condensed). . . . . . . . . . . . . . . . 115 Boardgames Played with Mothers and Fathers . . . . 116 Number of Boys and Girls Whose Mothers Play/ Do Not Play with Them . . . . . . . . . . . . 117 Number of Boys and Girls Whose Fathers Play/ Do Not Play with Them . . . . . . . . . . . . . . . . . 118 Children by Grade and School Reporting that Parents Do Not Play with Them . . . . . . . . . 119 Number of Children Observed Per Category. . . . . 120 Behaviors During Recess of Popular and Unpopular Children. . . . . . . . . . . . . . . . . . . . . 123 Play Activities of Popular and Unpopular Children. 124 Categories of Play Activities of Popular and Unpopular Children. . . . . . . . . . . . . . . . 126 Play Activities of High vs. Low Self—concept Children. . . . . . . . . . . . . . . . . . . . . 128 Categories of Play Activities of Children Varying in Self— concept . . . . . . . . . . . . . . . . 130 xv 49 50 51 52 Behaviors During Recess of High vs. Low Self— concept Children . . . . . . . . . . . . . . . . 132 Behaviors During Recess of Children Varying in Popularity and Self— concept . . . . . 133 Play Activities of Children Varying in Self— concept and Popularity. . . 134 Self— concept Scores of Popular and Unpopular Children. . . . . . . . . . . . . . . . . . . . 136 Figure LIST OF FIGURES Title Sample sociometric questionnaire. A model predicting self—concept of fourth grade girls . . . . . . A model predicting self—concept of fourth grade boys A model predicting self-concept of first graders xvii Page 55 100 102 108 "Laro tayo!"*: Parent—child and peer play activities of Filipino Children Chapter 1 THE PROBLEM AND REVIEW OF RELATED LITERATURE This study describes play patterns of male and female Filipino children of two grade levels (first and fourth grades) and of low and middle socioeconomic status. Play activities at home were determined from children’s responses to a questionnaire, and free play activities in school were determined from observations of selected children during recess. These play patterns were related to children’s opportunity to play with their parents, the parent-child relationship, the child’s self—concept, and their popularity with peers. Differential patterns of play were explored in this study. Western literature has already described clear patterns. Several factors identified in this literature were examined in the play activities of Filipino children. These factors included sex, age, and socioeconomic differences in play. They also included interpersonal or social variables as well as intrapersonal variables such as personality. i' Filipino for "let’s play" 2 Sex differences in play activities are evident in early childhood. Boys play more physical games. When they engage in symbolic play, the types of roles played also show marked sex differences (Rubin et al., 1983). Play patterns and opportunities for play also differ among the various social classes. The social forms of play appear to be more evident among the middle and upper classes. This is partly due to less interactions between parents and children of the lower class and of the fewer opportunities afforded to children by these parents (Smilansky (1968) and wain & Shmukler (1981, cited in Johnson, et al., 1983). With age, there is an increasing social component to play (Smith, 1977). This is consistent with the overall pattern of increasing socialization in childhood (Gottman, 1986 cited in Hetherington & Parke, 1986). There are also marked changes in terms of complexity and cognitive maturity again consistent with cognitive development in childhood (Kalverboer, 1977; O’Connel and Bretherton, 1984). Relationship of play to several interpersonal and intrapersonal variables was also explored. The development of social skills through play has been inferred in studies relating social competence and social types of play (Rubin et al., 1983). Social competence could be enhanced as social play further provides the opportunity for social interactions. In two types of social play, for example, 3 symbolic or make—believe play, as well as games-with—rules, players learn and practice social roles and expectations (Mendez, Jocano, Rolda & Matela, 1984; Piaget, 1972). The correlational nature of this relationship is evident in the popular child’s engagement in more social forms of play (Connoly, 1980, cited in Rubin et al.) While social play as a correlate of social forms of play has been largely explored in terms of peer interactions, another interesting context would be that of parent-child relations. In the generally authoritarian Filipino family, the context of play is the one activity where parent and child are equals. Engagement in such a presumably enjoyable and egalitarian activity could enhance the relationship between parent and child. Development of self-concept could be an indirect consequence of engaging in play in childhood. Early socialization is an important source of the development of self-concept. Children base their perception of themselves on how they believe others see them. Perception of parental warmth could produce a more positive View of oneself. If parent-child play and parent-child relationship are related to each other, it would be interesting to see how they contribute to a child’s self-concept, whether separately or in combination. In peer relations, a child’s View of oneself in relation to his or her peers would be related to his or her popularity. A child’s view of oneself could be 4 indirectly related to play behaviors as differences in play behaviors have been found to be related to popularity. This study is exploratory as hypothesized relationships between patterns of play and selected variables have largely been based on western literature. Findings of either similar or different patterns of Filipino children’s play activities would be an important contribution to the growing literature in Filipino Psychology, as well as to cross- cultural research. It is the belief of this researcher that play is a valuable and necessary activity in childhood. Relating play activities to interpersonal relationships, and with the intrapersonal dimension of self-concept could provide evidence for this belief. This study has several objectives: 1. To describe play activities of Filipino children at home, 2. To relate frequency of parent-child play with the variables: child’s self—concept and the parent-child relationship, and 3. To describe play behaviors of popular and unpopular children in school during recess. 5 Review of Related Literature This present study is aimed at describing Filipino children’s play activities. Patterns of play, variations in play, and possible relationships with selected variables are explored. To provide background for the areas of interest, the literature reviewed in this section includes significance of play, categories of play behaviors, patterns according to age, sex and socioeconomic status, play variations in the context of different settings and adult involvement, the Filipino context, and methodological issues in research on play. Significance of play It is now commonly believed that play must have some important developmental outcomes. Developmental psychologists have given this issue increasing attention, finding cognitive and social benefits. While many are still speculative, the following are among the topics for which there is empirical evidence showing benefits of play. Development of social skills Social play is believed to promote social skills such as cooperation, sharing, followership and flexibility (Athey, 1984). Sociodramatic play promotes social awareness, flexibility in approaching different situations (Smilansky, 1968), and rehearsal of adult roles (Mendez, Jocano, Rolda & Matela, 1984). 6 There is some empirical support for the social benefits of sociodramatic play. Sociodramatic play tutoring had positive effects on group cooperation (Rosen, 1974), and increased positive social interactions (Smith, Dalgleish & Herzmark, 1981, cited in Johnson et al., 1987) and empathy (Saltz, Dixon & Johnson, 1977, cited in Johnson et al., 1987). Connolly’s (1980) observational study of children at play found that incidence of sociodramatic play significantly predicted performance on measures of social competence, popularity, and role-taking activity (cited in Rubin, et al., 1983). In games with rules, children learn to accept prearranged rules and to adjust to them (Smilansky, 1968); they learn to control their behaviors and reactions within given limits. Even rough-and-tumble play has been proposed to have positive values. Hartup (1977) and Johnson et al. (1987) claim that in rough-and—tumble play children learn to control impulses so as to be able to participate appropriately within the group. Among popular children, it is a form of playful provocation, a means to elicit a response from another (Pellegrini, 1989). However, the socializing aspect of rough-and-tumble play affects children of different dispositions in different ways. For aggressive children, rough-and-tumble play progresses into aggression (Pellegrini, 1989) and seems to be an act of establishing dominance. 7 Several studies have found negative correlations between some types of play and social skills. Using teacher ratings, Christie & Johnsen (1989) found group play to be related to social maladjustment. In their study, the researchers point out the strong influence of context (1 e. teacher attitudes) on this unexpected finding. The teachers find group play to be disruptive and less directly related to academic tasks than solitary activities such as reading. Thus it was looked upon negatively by these teachers. Rubin’s study (1985) clarifies the importance of social play for social skills to appear. In his study, even dramatic play, if it is solitary, is not associated with improvement in social competence. It appears that for dramatic play to have positive social benefits, it must occur with others. Social relationships: Peer play Few studies have looked at how play promotes relationships between the child and significant others. Most studies describe how children play with peers and adults, but not how engaging in play with each other is related to the quality of their relationships. In peer relations, Sutton-Smith (1984) asserts that "play is about the struggle for identity within the dominance-subordination domains of one’s peers" (p.61). Play is seen as a medium for finding a niche in the context 8 of peer relations. These speculations point to the positive role play may have in establishing relationships. Empirical support is needed to establish the validity of such speculations. Social relations: Parent-child play Athey (1984) speculates that the body contact in infant play with adults builds a sense of security and belonging. Atienza (1982) suggests that games played by the family encourage mutual acceptance and affection, and that the child is more likely to feel that parents who play with him or her "really care". In a review of studies on parent-child play, Henderson (1984) listed some conclusions. First, a secure attachment to a caregiver provides the young child a base from which to explore. Sorce & Emde (1981, cited in Cohen, 1987) observed that when the mother is in the same room as the child but could not be "used" by the child (i.e. mother buried her head in a newspaper), the child's play and explorations were limited. Van der Kooij (1989a) also found that when highly stimulated by parents to play, children show higher play intensity. Secondly, adults may facilitate play by focusing the child on novel objects. Mothers show explicit teaching with play with objects (Dunn & Dale, 1984; Dunn & Wooding, 1977). Belsky, Goode & Most (1980, cited in Cohen, 1981) observed that children played most competently when mothers 9 focus their children’s attention to what the toys were and what could be done with them. Mothers also appeared to be sensitive to children’s capacities. With younger children, they used more physical strategies such as demonstrating an object; with older ones they were more verbal. Thirdly, adult participation facilitates play. Dunn & Wooding (1977) found that with the mothers’ joint attention, length of play increased. It also provided a forum in which verbal exchange between mother and child is rich. O’Connel & Bretherton (1984) observed facilitation of play by mothers. However, they claim that it is the child who actually determines the effectiveness of mother’s instruction. Thus while adult involvement appears to help, the child is not a passive recipient of such interaction. And lastly, adult involvement facilitates social skills. In the first year of life, mothers provide attention—maintaining stimulation (Fitzgerald, Strommen & McKinney, 1982). Caregivers actively encouraged mothering play among 2-year olds (Miller & Garvey, 1984). This is significant in that it is the first step towards adopting a social role other than one’s own. In all of these studies, very little mention is made of fathers, usually simply describing the more physical play of fathers with their children. This neglect is surprising since fathers may actually spend more time playing with their children in proportion to total time spent with them lO (Jurilla, 1986). Thus it is important that a study on parent—child play interactions and relations include the father as well. Mental health: the self Play is conceptualized to be an empowering activity. According to Johnson, Christie & Yawkey (1987), it is self— enabling; it enhances the self as an autonomous and functioning person who can control events. The individual is seen as "kept in balance" by the activity (Sutton—Smith & Kelly—Byrne, 1984). The importance of play to mental health is indirectly supported by findings relating early peer relations and adult mental health problems (Hartup, 1977). Peer rejection or poor peer relations significantly predict future problems. And as play occurs mainly in the context of peer relations, it could be said that failure to play with peers is a good indicator of poor peer relations. This, in turn, subsequently predicts future mental health problems. Sex—roles Children are socialized into their respective sex—roles very early in life. Differences can be seen in the toys bought for them (Block, 1981; Rubin et al., 1983) and in parent-child play. Dunn & Dale (1984) observed mothers and their 2—year olds. They found that mothers initiated nurturing and household themes more often to their daughters than to their sons (34% of playing time with daughters vs. 11 % with sons). Actions to vehicles on the other hand showed a reverse trend (15% with sons, % with daughters). Teachers also play a role in reinforcing sex—typed play. They are more likely to leave children alone as long as they played in the traditional way (Fagot, 1983, cited in Johnson et al., 1987). Arrangement of the play area influences the way children play. While boys already show a preference to the block area, and girls to the housekeeping area, keeping the two areas separate reinforced the division between "boy areas" and "girl areas". Kinsman & Berk (1979) found that simply removing the divider between these two areas significantly increased play between boys and girls and also encouraged play with opposite-sex-typed toys. Summary of significance of play We see that in play, children learn valuable social skills which are inherent in play activities themselves, and are, therefore, acquired and practiced through them. Effects of play on relationships with significant others are little studied to date, and the relationship between play and mental health has been only indirectly shown by looking at its opposite. In light of these observations, three areas are the special foci of this study: (1) play with parents and children’s relationships with them as measured by a parental acceptance/rejection measure, 12 (2) play and mental health (or healthy functioning), specifically play as related to the child’s self— concept, and (3) play and social skills, in particular, play as related to the child’s popularity with peers. Categories of play There are many ways of analyzing and categorizing play activities. Smith (1977) suggests that one look into the different aspects of play. A researcher can look at one or a combination of the following: content, complexity, and context of play. This study looks at the combination of content and complexity, and the social contexts of peer and adult interactions. Content and complexity of play Content refers to the kinds of behaviors to which combinatorial repetitions and variations are applied. This includes physical activity, use of objects and symbolic play. Most categorizations of play according to content are also arranged according to complexity. Thus, these two ways of categorizing play behaviors are combined. An additional type of play which has been identified when analyzing complexity, but not content, is games—with—rules. These four play categories, physical activity, use of objects, symbolic play, and games—with-rules, are discussed successively. 13 Physical activity This play category involves gross body movements. One example is rough-and-tumble play which includes the following: tease, hit at/kick at, poke, pounce, sneak up, carry child, play fight, pick on, chase, hold and push (Pellegrini, 1989). This type of play obviously involves at least one other person. Another type of activity which makes use of gross body movement is described by Piaget as the earliest form of play in infancy and early childhood. Sensorimotor exercise or practice play consists of simple repetitive muscle movements (Piaget, 1962). This type of play is done for performance of already existing schemas with no effort at adaptation. Examples are running and jumping. They could also be repetitive movements involving objects such as bouncing or throwing a ball. Use of objects Perhaps the most well—known categories of play are Piaget’s (1962). His categories are developmental and were originally used to describe cognitive levels. The earliest two levels both involve object use. The first level is functional play. It is done as a manipulatory activity but is not constructive; that is, it is done for its own sake. An example is kneading or rolling clay with no attempts to construct something out of the clay. 14 The second level is constructive play. Manipulations of objects are done in order to create something. Play with blocks, clay, paints usually end up as creations. Other categories of object play have been proposed. Kalverboer (1977) and O’Connel and Bretherton (1984) describe similar categories. Sympolic play The most frequent topic of research on play is symbolic play. Other names for this type of play are pretend (Rubin, Fein, & Vandenberg, 1983), dramatic (Piaget, 1962), and fantasy play (Smith, 1977). Shotwell, Wolf & Gardner (1980, cited by McCune- Nicolich & Fenson, 1984) define symbolic play as "the ability to represent actual or imagined experience through the combined use of objects, motion and language" (p.84). Symbolic play when elaborated in cooperation with at least one other role player is called sociodramatic play (Smilansky, 1968). Focus is on role-playing. Saltz and his colleagues (cited in Rubin et al., 1983) differentiate betwen ordinary sociodramatic play and thematic fantasy play. In the latter, the roles played are far removed from the children’s everyday experiences. An example is pretending to be fairies in another world. Because it requires more imagery and pretense, they declare that it is a more mature form of group pretense. 15 Garvey and Berndt (1977, in Rubin et al., 1983) describe even more types of dramatic roles: Functional roles are those organized by an object or activity (e.g. teacher when playing school). Relational roles are those that imply complementary relationships (e.g. mother—child). Character roles are based on stereotypic occupational or habitual activities (e.g. cowboy) or on fictional roles (e.g. Robin Hood). And peripheral roles are those discussed and addressed but not portrayed by the child himself or herself (e.g. imaginary friends). The various categories discussed above suggest that symbolic play in itself has several components. One can look at the role—playing aspect such as types of roles played. The use of objects and other people in the play activity are also important. In both cases one can see how far the symbolization extends. For example, while use of a toy plate requires little or no imagination, the use of a leaf for a plate does require imagination. Playing "mommy" which is directly imitative is not as creative as playing "space monster," for which children have no real models. In playing with others, it is interesting to note roles assigned to self and others as well as how the children cooperate to maintain the dramatization of their respective roles. Being able to maintain sociodramatic play requires important social skills such as cooperation and turn—taking. 16 Games with rules And lastly, complexity of rules varies in different types of play. While play of young children is virtually rule—less, symbolic play makes use of rules that are made up by the role players (Rubin et al., 1983). However, one type of play is characterized by the presence of formal rules. Games with rules are probably the most complex of play activities. In them, the children have to accept prearranged rules and to adjust to them. They learn to control their own actions and reactions within given limits (Smilansky, 1968). Piaget (1976) predicts that symbolic play gradually evolves into games with rules. The former becomes increasingly social and rule-governed, developing into the formal rule—governed competitive games. Rules in the latter could not be changed unless mutually agreed upon before the game is played. According to Piaget (1976), "competition is controlled by a collective discipline, with a code of honour and fair play" (p.569). Games with rules differ from organized sports in that motivation could be intrinsic rather than extrinsic (i.e. while the player plays to win, there are no external rewards and the players are not influenced by external demands to participate). In this sense, organized sports would not be classified as play, while spontaneously played informal 17 sports would be classified as play even though both are rule—bound. These different ways of categorizing play are useful in determining play patterns derived from play activities. They form a basis for initially grouping the varied play activities and for describing differences in play activities between relevant groups such as boys and girls. Peer social context One classification scheme which looks at the social dimension of play development is Parten’s social participation scale (cited in Johnson et al., 1987). It has the very simple categories of: (l) solitary play - playing alone with materials different from those of children within speaking distance; no conservation with others; (2) parallel play - playing with toys or engaging in activities similar to those of other children who are in close proximity; however, there is no attempt to play with the other children; (3) group play - playing with other children, roles may or may not be assigned. Howes (1980; in Johnson, et al., 1987) examines these social levels in greater detail. He focuses on two dimensions of peer play: (1) the complexity of the social interactions among children, and (2) the degree to which their activities are organized and integrated. He describes five levels of increasing complexity: 18 Level 1 - Simple parallel play Children, in close proximity to one another, are involved in similar activities but do not engage in eye contact or any social behavior. Level 2 - Parallel play with mutual regard Children are involved in similar activities and engage in eye contact. The children, though not socially interacting, are aware of others’ presence and activities. Level 3 - Simple social play Children direct social behaviors to one another. Typical behaviors include vocalizing, offering objects, smiling, touching, taking toys, and aggression. The children’s play activities, however, are not coordinated. Level 4 - Complementary/reciprocal play with mutual awareness Children engage in activities in which their actions reverse other children’s actions, demonstrating awareness of each other’s roles. No conversation or other social exchange takes place. Level 5 - Complementary/reciprocal social play Children engage in complementary and reciprocal activities, as in Level 4, and in social exchanges, as in Level 3. Both of these scales look at the social context in terms of peer involvement. This has been a major concern of Inany studies on play (Rubin et al., 1983). As 19 classification in these scales allows the researcher to infer a child’s social maturity, it is important to note that there are many environmental constraints which affect the child’s capacity to engage in social play. Jennings & Curry (1982, cited in Curry & Arnaud, 1984) have found that the following affect children’s social play: presence of mother, familiarity with the other child(ren), teachers, physical setting, and the length of time provided for the children to become familiar with one another. The following section further looks at adult involvement and settings. Adult social context Johnson et al. (1987) describe different types of adult involvement in play with children: 1. Parallel play — The adult plays alongside but not with the child or children. When adult participation is in this pattern, children tend to persist longer in play. And by observing how the adult is playing, they may learn new ways of playing with objects. 2. Co—play — Adult joins but lets the children control the course of the play. By asking for information, instructions, and responding to children’s actions and comments, the adult can add new elements to the children’s play. 3. Play tutoring - The adult initiates a new play episode and takes a more dominant role thereby teaching new play behaviors. This can be done in two ways: 20 a. outside intervention — adult is not involved in the play itself, b. inside intervention - adult takes on a role. 4. Thematic fantasy training — The adult helps children act out stories. This is beneficial for children with no or little experience with sociodramatic play. 5. Spokesman for reality — the adult may classify or explain roles more accurately. This helps children understand roles of others. In any of these different types of adult participation, the adult involvement is gradually phased out such that children end up playing among themselves. Also, the authors caution that while adult involvement may help children, it is not always necessary. An adult must be sensitive as to when his or her involvement is needed. They suggest that an adult intervene only when: (1) children do not engage in make—believe play on their own, (2) children have difficulty playing with other children, or (3) play becomes repetitious or appears ready to break down. Settings When categorizing according to settings, a play setting can be academic or nonacademic. Many studies look at free play in the classroom or in the laboratory. Christie & Johnsen (1989) argue that play which occurs in these settings is not really free play. Types of toys and 21 expectations of adults are different in these settings than in nonacademic settings. Schwartzman (1984) points out how these settings affect children’s play. Higher socioeconomic children show more mature forms of play in the academic settings than lower socioeconomic children. However, outside the laboratory or classroom, even lower SES children show such mature forms. Another classification scheme is that of outdoor play versus indoor play. The difference between these two settings may account for some of the social class differences in play as upper class children play indoors more frequently, while lower class children spend more time playing outdoors (Minoza, 1984). The type of outdoor environment also changes play behaviors. Traditional playgrounds with fixed, conventional equipment encourage functional play, especially large motor play (Johnson et al., 1987; Rubin et al., 1983). Naylor’s (1985) review shows that traditional playgrounds are seldom used by children and that they show a preference for "adventure" playgrounds where elements are loose and children can create their own structures. Undesignated play spaces, such as backyards, encourage more social play. Importance of categorization and contextualization The studies reviewed above are useful for this study in that they help organize the analysis of play activities. Sex, age and socioeconomic differences can better be 22 understood if one can categorize the play activities according to some dimension such as complexity. They can also be better interpreted when viewed in their proper context such as the physical and sociocultural environments wherein these play activities occur. In this study, play occured within the settings of home and school during recess. This study was done in a different culture from the one from which categories were derived. However, it is assumed by this researcher that although play activities, per se, may differ between the Philippine and Western cultures, these categories would still be useful for initial organization of the data. Should there be play activities which do not fit the categories, the categories could be reconceptualized. Plav patterns according to age. sexl social class, and personality The sections above described ways by which variations in play could be categorized. The following sections show variations in play according to children’s age, sex, social class, and personality. Age differences Age is usually related to certain types of play. Some types of play are more frequent in certain age groups and some do not emerge until a certain age. Christie & Johnsen (1989) found that for American children functional play was 23 found to be negatively correlated with age, peaking at around 4 years of age; while constructive play was positively correlated with age peaking at 6 years. Van der Kooij (1989) found similar results, with constructive play peaking at 5 years among Dutch children. Symbolic play emerges from 3-6 years (Smilansky, 1968) among advantaged preschoolers but not among disadvantaged children. Schwartzman (1984) argues that disadvantaged children did engage in symbolic play but at a later age (6-8 years). There are also changes in symbolic play with age. Children show increasing representational ability. As they grow older, they tend to use ambiguous props to engage in symbolic play (Pellegrini & Perlmutter, 1990). There is also marked maturity in terms of playing roles. As the child grows older, he or she is capable of elaborating more complex role relationships. Play of children becomes increasingly social with age (Smith, 1977). At 2 years, social interactions are infrequent and short in duration. Most play of 2-year olds is with objects or with adults. At 3 years, social play begins in the play group or nursery. Interactions are mostly dyadic. At 4—6 years, there is a marked increase in duration of social play and in number of children involved in it. 24 Thus with age, children’s play changes in terms of types of play, complexity, sociability, and cognitive maturity. Sex differences Sex differences are evident even in early childhood. This can be seen in children’s play activities. Boys seem to prefer gross motor and rough—and—tumble play. Girls, on the other hand, prefer play with objects (Smith, 1977). Both sexes engage in symbolic play. However, there are sex differences in the types of roles played. Boys are more likely to play fictional, superhero characters. Girls play familial characters (Rubin et al., 1983). Pellegrini & Perlmutter (1990) show differences in social interactions during play. Girls engage in more imitations, requests for help, and give more responses to play topic initiations. Boys exhibited more utterances, gave more commands and play topic initiations. These changes reflect sex—typed characteristics of more nurturance and submissiveness among girls, and more activity, assertiveness and achievement among boys. While we cannot say that it is in play that sex-typed behaviors are learned, it is evident that these are maintained in play. Among older children, these same trends can be seen in their games (Block, 1981). Boys’ games reward initiative, improvisation and extemporaneity. They encourage within— % 25 team cooperation as well as between—team competition. Girls’ games are mostly highly structured, turn-taking games that are strictly rule—governed and less often require contingent strategies. Thus it seems that boys’ games are more geared towards development of competence than are girls’ games. These sex—typed behaviors increase with age, and sex— typing is more evident among boys (Rubin et al., 1983). Older boys avoid playing with feminine toys more than preschool boys; and boys generally avoid feminine toys more than girls avoid masculine toys. Girls’ preferences are broader in scope. In symbolic play, for example, girls are more likely to engage also in fictional pretense than are boys to engage in familial themes. Social class differences Children from lower socioeconomic groups seem to exhibit less mature forms of play than those from middle and high SES groups. Lower SES preschoolers have been found to engage in more solitary and functional parallel play compared to their middle SES agemates (Rubin et al., 1976; cited in Rubin et al., 1983). Middle and high SES children engage in more symbolic play than lower SES peers (Smilansky, 1968; Smith, 1977; Tizard, Philips & Plewis, cited in Rubin et al., 1983; wain & Shmukler, 1981, cited in Johnson et al., 1987). 26 Smilansky (1968) and wain & Shmukler (1981, cited in Johnson et al., 1983) assert that these deficits are due to the differential ways that lower SES parents treat their children. Smilansky says that upper class parents are more likely to play with their children while lower class parents are less likely to do so. There are many confounding variables in studies of social class differences. Most studies cited are done in the school or laboratory settings. Schwartzman’s (1984) study is interesting in that it shows that outside of these settings, even lower SES children’s play demonstrate the maturity of their more advantaged counterparts. They are also highly creative, more verbal than in the academic settings, and they display a variety of social and survival skills. Schwartzman (1984) and Feitelson (1977) point out several basic differences in the lifestyles of these children. Children who work (i.e. who engage in child care and/or other economic responsibilities) cannot or do not have enough time to play. And children of lower class families frequently have to assume such responsibilities. Children must also have sufficient space and toys to engage in symbolic play, both of which are deficient in lower income homes. 27 Personality differences Play activities are a good medium for observing personality differences among children. Stockinger—Forys and McCune-Nicolich (1984) put together dyads of 3-year olds who did not know each other. They observed that the socially dominant child seemed able to wait while the other child warmed up. Pellegrini (1989) found differences in rough-and—tumble play of popular and rejected children. For the former, this type of play was nonaggressive while for the latter, aggression seems to dominate it. Other researches identify personality patterns associated with play. Johnson and his colleagues (1987) identified a "fantasy—making" predisposition. They found this to be related to higher level of imagery, positive affect, social interactions and cooperation during free play. Wolf and Gardner (1979, cited in Johnson et al., 1987) differentiated between "patterners" and "dramatists". The former engage in more object play, while the latter engage in more social play. As most of these studies are observational and correlational, causal relationships between personality and types of play in cannot be properly established. However, since personality or dispositions are often inferred to be relatively enduring, stable traits, then it could be said 28 that choice of play activities is a result of certain dispositions. In this study, self—concept and peer popularity are two relatively stable traits that were related to play. As popular and unpopular children, as well as children of high and low self-concept differ in behaviors, it is hypothesized that they would differ in play as well. Summary of play patterns The differences described above show that both maturation and socialization play a part in differentiating play behaviors. As the literature cited describes children in Western countries, predominantly the U.S., it would be interesting to see if such patterns also exist among play of Filipino children. One would expect similarities in patterns that are due to maturation, while differences between cultures would be evident in patterns that are due to socialization. The Philippine experience Most of the research cited above involved American children in the U.S. and children in other western countries such as England and the Netherlands. Far less is known about Filipino children’s play. The following sections describe what has been written about how children play and parent-child relationships in the Philippines. 29 Filipino children’s play As in the western literature, there are sex differences in the way Filipino children play. Girls play inside the house more frequently than outside, and more frequently than boys (Minoza, Botor & Tablante, 1984). Filipino children’s play is also sex-typed. Sevilla (1982) observed that the favorite play activities of 3-6 year old girls is playing house, while for boys, it is gunfights. Many of the games however are not sex-typed, such as playing in the sand and catching insects. There are also social class and ecological differences. Urban middle- and high-SES children are more likely to play indoors and have commercial toys, while rural and lower SES urban children play outdoors and have few toys (Minoza et al., 1984). Social skills are demonstrated in play. Minoza and her colleagues found that most children (81.6%) could get along with others at play. When they quarrel, group pressures tend to resolve the quarrel (Mendez et al., 1984). WM In child-rearing, Filipino mothers are more involved than the fathers. Ramirez (1974) reports that the father is always out of the house (including non-work activities) to the neglect of the attention due to his wife and children. Minoza et al. (1984) found that only 23.5% of the husbands they surveyed were enthusiastic in helping their wives with 30 child care, although about half (51%) report helping with discipline. And yet, in one rural community, more fathers (84% vs. 13% of mothers) reported companionship and warmth between parent and child as a primary motive for parenthood (Jurilla, 1986). The father also acts as the child’s playmate while the wife attends to her chores. In a nationwide survey done by Sevilla (1982), she reports that 73% of parents get directly involved with their children’s play, either as an active onlooker or as actual participants. In a primarily authoritarian society, this is significant in that in the area of play, parents and children are, for once, on a more equal basis (Atienza, 1982). The apparent relationship between parents and children in play, particularly the role of fathers, is interesting in that it allows the participants to step out of their customary authoritarian relationship. But very little has been written about children’s play in the Philippines. A major aim of the present study is to add to the literature about play patterns of Filipino children. Methodological issues McCune—Nicolich and Fenson (1984) recommend that in doing play research, one must consider the setting, participants and their roles, subjects, and observation procedures. 31 Settings Most studies are done at home (Belsky, Goode & Most, 1980; Dunn & Dale, 1984; Dunn & Wooding, 1977; Miller & Garvey, 1984; Minoza et al., 1984; O’Connell & Bretherton, 1984), in a laboratory (Eckerman & Stein, 1990; McCune— Nicolich & Fenson, 1984; O’Connel & Bretherton, 1984; Phillips & Sellito, 1990; Rosen, 1974; Smith & Connolly, 1980 in Johnson et al., 1987; Sorce & Emde, 1981 cited in Cohen, 1987), or in school play areas such as the playroom or playground (Christie & Johnsen, 1989; Pellegrini, 1989; Pellegrini & Perlmutter, 1990; Rubin, 1982; Van der Kooij, 1989). It appears that the setting varies according to the purpose of the research. When studies are done in the laboratory, there is usually manipulation of some variable such as spatial density, and there is usually adult intervention (Eckerman & Stein, 1990; Smith & Connolly, 1980, in Johnson et al., 1987; Sorce & Emde, 1981, cited in Cohen, 1987). Most purely descriptive studies are done where free play is most readily observed, such as the home or playground. But other free play research is also done in the classroom or the laboratory. In such cases, a bias seems to exist in favor of more socioeconomically advantaged children who show more mature forms of play in the laboratory or classroom (Schwartzman, 1984). 32 As this study is exploratory in nature and has as one focus the question, "What do Filipino children play at home?", a survey was deemed most appropriate to get as many responses as possible from many children. Participants and their roles Most free play observations include other same—age children. This appears to be the most "naturalistic" form of observation in that most play activities occur within the context of peer relations. A second focus of this study was to document Filipino children’s free play activities in an academic setting. For this purpose free play observations were done during recess. This was limited to observations of children identified as popular or unpopular through a sociometric measure. When others are involved in the child’s play activities as documented in research studies, it is usually the mother (Belsky, Grade & Most, 1980, cited in Cohen, 1987; Dunn & Wooding, 1977; Dunn & Dale, 1984; Minoza et al., 1984; O’Connel & Bretherton, 1984; Phillips & Sellito, 1990; Sorce & Emde, 1981 cited in Cohen, 1987). However, most of these studies are experimental, involving effects of some manipulation on children’s play. Another adult is usually requested to manipulate the situation in some way. The observation is then focused on the effects of such manipulations on the child’s play. Examples include level 33 of child’s play when mother ignores/attends to child (Eckerman & Stein, 1990; Phillips & Sellito, 1990; Sorce & Emde, 1981, cited in Cohen, 1987), and effects of play tutoring (Rosen, 1974; Smilansky, 1968). Some mother-child observations are of free play. In these observations, the focus is more on how the mother plays with the child (Dunn & Dale, 1984; Dunn and Wooding, 1977). Parent-child play activities were also explored in this study and as this researcher was concerned mostly with getting a variety of responses from many children, survey was conducted rather than observations at home. Subjects Most studies of play include infants, preschoolers and kindergartners. Older children are rarely studied (Pellegrini, 1989; Van der Kooij, 1989b). As such, certain types of play are overrepresented in the psychological literature. These include functional, constructive, and in particular, symbolic play. Since play of older children is not much studied, games with rules are neglected (Block, 1981). And as we associate play with childhood, play in adolescence and adulthood, except probably in the context of sports, is practically ignored whether in the U.S. or in the Philippines. 34 For this study, two different childhood levels were chosen, first graders to represent early middle childhood, and fourth graders to represent later middle childhood. These were chosen because these groups are young enough for parental interactions to remain active and significant, yet old enough that opportunity for peer interactions are available. The different forms of play, from physical play to the more complex games with rules, are also observed in children of these age levels. Conceptual Framework This research is not guided by any one theory, and all hypotheses are based on the literature previously reviewed. Variations in play activities The topic most studied in play research is age differences. This study is focused on social play of children in middle and late childhood. Other studies describe differences in types of play engaged by these two age groups (Piaget, 1962). Parten (1932, in Johnson, et al., 1987) observed play patterns to change from solitary to cooperative (social). Piaget (1962) and Smilansky (1969) identified early to mid-childhood as the age when sociodramatic play is most common, and mid—late childhood when games—with—rules emerge. Changes in friendship patterns and cognitive development could account, in part, for these differences. 35 As children grow older, their peer group size changes from dyadic to small to bigger groups. However there is a sex difference in that girls tend to participate in small peer groups while boys are in larger groups (Block, 1981). Cognitively, games-with-rules require more maturity as they require knowledge and retention of several rules as well as the use of strategies to win in these usually competitive activities. Sociodramatic play, on the other hand, requires different cognitive skills, primarily the ability to take the perspective of others which requires a knowledge of the roles of others, and also the flexibility to "pretend." In this study, two very different age groups, first and fourth graders, comprise the sample. Based on the literature, the younger group would be expected to engage in more sociodramatic play, while the older is now capable of games with rules. These two age groups were chosen primarily because of this difference. In order to see more diversity in play patterns, children who are capable of these more sophisticated play activities rather than preschool children were selected. Also as children of these age groups are more social, i.e. tend to play in groups compared to preschoolers, important social variables, such as popularity as an indicator of social competence, could be investigated in relation to social play. Furthermore, children of these age groups still spend a considerable amount of time with the 36 family, so that impact of that relationship on play may be investigated as well. Sex differences have been found in types of play activities (Kalverboer, 1977; Smith, 1977), toys (Block, 1981) and roles played in sociodramatic play (Rubin et al., 1983). These sex differences in play seem to reflect sociocultural sex role stereotypes. In this study, observing play of Filipino children would allow one to see sex-role stereotypes which exist in the Philippine culture. Roles Filipino children play would reflect roles that are perceived in the culture as more typical of each sex. Studies on socioeconomic status differences show less mature forms of play among lower SES children (Smilansky, 1968; Rubin et al., 1983). Children of ages 6-7 are expected to engage in a considerable amount of sociodramatic play. These studies show that children of lower SES tend not to do so. Instead they engage in the less mature play forms of solitary, parallel, functional, or constructive play. Smilansky (1968) attributes such differences to parent— child interactions. Schwartzman (1984) says differences in lifestyle account for the appearance of less mature forms of play among lower SES children. MacDonald (1993) describes a parenting style which resembles that of the lower class. He associates this style with low involvement in parent-child play. Mendez et al. (1984) shows that lower class Filipino 37 children frequently take care of siblings and do household chores leaving less time for play. Play varies in different settings. There are differences between schools and homes in terms of types of available play materials (Christie & Johnsen, 1989). Age and sex differences in play seem to reflect dispositions to certain play activities due to maturity and to social learning. Effects of settings is probably more external in that children will adjust their play activities according to what is physically available and possible in various settings. Play and popularity As most peer interactions occur in the context of play, whether a child is liked or not by her or his peers could be observed in play interactions. As most play activities in mid— to late childhood are social, a child with good social skills would fit in well in play interactions. It is also possible that the valuable social skills that popular children have are learned in play interactions. Problem behaviors can also be seen in the play situation. For example, some differences between popular and rejected children have been observed in rough-and—tumble play (Pellegrini, 1989). The latter tend to show more hostility than playfulness compared to the former. Among isolated children, a basic problem is their inability to 38 enter play groups. These general findings would be expected in Filipino children’s play. These early socialization interactions could also affect how children view themselves. They see the self according to how they think others see them. Being liked and accepted by peers is therefore a possible determinant of a positive self-concept. In this study, it is assumed that being liked is in part determined by positive social interactions with peers. For these age groups, these positive social interactions occur mostly during play. Accordingly, children’s self-concept was related to play patterns in this study. Play with parentsI parent—child relationshipI and self-concept As parent—child relations are positively correlated with a child’s self-concept, parent-child play interactions would indirectly affect self-concept due to its relation to parent-child relationships. In this research, parent—child relationship is measured as the child’s perception of parental warmth as opposed to rejection or indifference. Playful interactions are assumed to be indicators of warmth. A parent who cares enough to spend time playing with a child is perceived as one who must "really care". In childhood, play is probably one of the few parent— child interactions where they are on an equal level. It is also usually pleasant. Thus opportunities for parents and 39 their children to engage in pleasant, equalitarian play interactions could contribute to the development of a positive parent-child relationship. Summary of framework This study includes descriptions of play activities at home reported by boys and girls, of lst and 4th grades and from public and private schools representing the lower and middle classes respectively. In addition, play and its relation to certain interpersonal and intrapersonal variables will be analyzed. It is expected that play patterns will differ among children varying in popularity and self—concept. On the other hand, parent-child play activities and frequency of parent—child play will be related to parent-child relations and self- concept. It is believed that parent—child play is related to parent-child relations and that both directly, or indirectly relate to self—concept. Statement of the Problem This study seeks to describe play activities of Filipino children in different settings, and to relate these activities to selected sociodemographic, personality, and social variables. Specifically, the following questions guide this study: 40 What play activities are engaged in by children categorized according to: a. sex? b. age (lst vs. 4th grade)? d. socioeconomic status (low and middle)? What is the relationship between playing with parents, child’s report of parent-child relations, and child’s self-concept? What is the relationship between a child’s peer play activities, peer popularity, and self-concept? Hypotheses Accordingly, the following hypotheses are proposed: There are differences in types of play engaged in by a. boys and girls, and b. children in first and fourth grades, c. children from low and middle socioeconomic status. Children’s reports of experiencing or having experienced playing with their fathers and/or mothers is signifi- cantly related to the child’s perception of his or her relationship with the father/mother, and with the child’s self-concept. Specifically, the more frequent the play interaction between child and parent(s), the more the child will perceive parental warmth, and the higher the child’s score in the Self-concept Scale. 41 3a. There are differences in types of play engaged in by popular and unpopular children. 3b. The more frequently the child engages in social play , the higher the peer rating of popularity of that child, and the higher the child’s score in the Self—concept Scale. Definition of Terms Play- This is any activity that is "enjoyable, flexible, and most typically characterized by pretend" (Smith & Vollstedt, 1985, p. 1049). In this study, it refers to any activity that would have any two of the three characteristics mentioned above that is observed during recess. It also refers to any activity that the children would interpret as play in response to the Play Activities Scale. Types of play- This refers to the kinds of play behaviors engaged in by children. There are different categories described in the literature. The following are included in the observation guide which is to be used in identifying types of play: Rough and tumble Dlav- This type of play involves gross body movements with physical contact between players. It includes the following, though this list is not exhaustive: tease, hit at/kick at, poke, pounce, sneak 42 up, play fight, pick on, chase, hold and push (Pellegrini, 1989). Sensorimotor play— This involves simple repetitive muscle movements with no effort at adaptation. It is done for the sake of the movement itself. It is also sometimes called "practice" play as the child appears to be simply exercising a movement. Examples include running, jumping, bouncing ball. Functional play- This is a manipulatory activity that is not constructive. It involves manipulation of objects but with no attempt to construct or make something out of the manipulation. An example is kneading clay but not making something out of the clay. Games with rules- This type of play activity has prearranged formal rules, and is usually competitive. Outdooerotor games- These are group, motor games which require gross motor skills such as running, jumping or throwing but which are differentiated from ball games. These games have rules and are played in groups. Examples include tag, leap—frog, hide-and-seek. Reciprocal play— In this type of play the actions reverse other chilren’s actions demonstrating awareness of each other’s roles. No conversation or other social exchange takes place. An example is a two—player video game where each player interacts with the action on the screen and not necessarily with the other player. 43 Reciprocal social play— This is similar to reciprocal play with the added dimension of social exchanges such as conversation. Socioeconomic status (SES)— This is the social and financial status of the subjects. Low socioeconomic status is inferred from enrollment in a public school while enrollment in a private school by virtue of the tuition structure is believed to indicate a middle SES (Gonzales, 1986). In this study, the terms "public" and "private" school may be used interchangeably with "low" and "middle" social class respectively. Popularity— This is the degree to which one is liked by other people. In this study, a sociometric rating scale which measures liking and/or acceptance is used to classify children as popular or unpopular. In this study, only same-sex peer ratings were used as same—sex interactions are more common among these age groups. Parent—child relationship— In this study, this is defined as the child’s perception of parental warmth as measured by the Parental Acceptance—Rejection Questionnaire (PARQ, Rohner, 1980). Parental warmth- Rohner (1980) conceptualizes parental warmth as a bipolar dimension with rejection or the absence of parental warmth and affection at one pole, and acceptance at the opposite pole. A parent is perceived as low in parental warmth or high in parental 44 rejection if the PARQ total score is low. "Parental acceptance" or "Parental rejection total" is used in the text to refer to the PARQ composite scale to distinguish it from the subscales Warmth and Rejection. Warmth (subscale) This subscale measures a child’s perceptions of parents as giving love or affection. Examples of parental behaviors include kissing, showing approval, comforting child. Reiection— This includes the child’s perception of the parent's dislike, disapproval and/or resentment of him or her. This is manifested in two ways: as (1) hostility, which includes anger, resentment, and enmity, or (2) indifference, which is the lack of parental concern or interest. A parent is perceived as high in parental rejection if the PARQ total score is high. The rejection dimension has 3 subscales: Hostility, Neglect and Rejection. Hostility Subscale refers to conditions where the child believes his or her parents (a) are angry, bitter, or resentful of him or her, or (b) intend to hurt the child physically or verbally. Examples of hostility behaviors include derogatory remarks, nagging, hitting. Neglect Subscale The child sees parents as unconcerned or uninterested. Sample behaviors include parents 45 spending minimum amount of time with the child or ignoring the child’s call for help. Rejection Subscale The child sees parents as withdrawing warmth but where such rejection does not clearly reflect either aggressive/hostility or neglect/indifference. An example is the child’s report that "my mother does not love me." Self-concept- This refers to a person’s View of oneself in a continuum of low (or negative) to high (or positive). In this study the Pasao Pictorial Self-Concept Scale by Munnariz and Pasao (1988) was used. The total score refers to self-concept or level of adjustment. It has two subscales: 1. Endearing Traits: measures how a child views his or her traits. This subscale answers questions such as "what do you like about yourself?" and "what do you like to do?" It includes items related to school activities (including peers) and the child’s view about the self. In this study, this scale shall be referred to as Self-concept: view of self, and 2. Relationships with Others: measures how a child views the self in relation to significant others, particularly significant adults such as parents, grandparents and teachers. This shall be referred to as Self-concept: Relation to others. 46 In general, the higher the score, the higher the self- concept. Scope and Limitations This study seeks to describe play activities of older children as most play studies include infants, toddlers and preschoolers. The ages 6—7 and 9—10 (lst and 4th graders respectively) were chosen because at these ages, the children are already social and yet they are sufficiently different developmentally to expect that their play patterns should differ as well. Also children of these ages already demonstrate two types of play, sociodramatic play and games with rules, which are investigated in this study. Because this is a cross—sectional study, only age differences rather than developmental progression can be described. First graders were especially chosen because for most schools in the Philippines, this is the beginning of "formal" education. Most of the social interaction takes place during breaks, such as recess, as there are no longer free times during the school day to engage in socialization activities as are found in preschools. This study includes only the low and middle income classes. The upper bracket was excluded because they are a small minority; also children from upper SES families tend to be enrolled in private sex—segregated rather than coeducational schools. 47 Observations of play activities were done only in the school setting. Ideally, different contexts should be investigated: the home, neighborhood, parks, etc. Play activities in the home were not observed but were reported by the children. This serves as a limitation in that the basis for analysis of the two settings are not the same. Clear comparisons could not be done. Instead, descriptions of play in each setting were done separately. Four coeducational elementary schools in Metro Manila were included in this study. Given the urban setting, generalizations of the findings to children in other areas of the Philippines are limited. Significance of the Study This study’s primary significance is its potential contribution to the developmental literature in Filipino psychology. The observations of children’s play in the playground could show important differences in types of play of children of two different age groups, of varying popularity, sex and socioeconomic status. In addition, comparisons with findings of play research in western cultures could be made. Differences and/or similarities in play patterns between Filipino and western children could be identified. Relating peer play with a child’s popularity among peers and his or her self—concept in relation to peers could 48 guide teachers and counselors to possible interventions regarding children with problems in peer relationships. And lastly, in attempting to relate playing with parents and variables such as parent—child relations or self—concept, potential benefits of such an interaction in childhood would hopefully be identified. In a society where dual incomes is becoming more a necessity, and where for the middle— and upper classes, child caregiving is left to a hired caregiver, there may be less opportunities for parent— child interactions. If parents could see why play is useful, then directing these limited opportunities toward play interactions might be the most efficient use of that limited time they do spend with their children. Chapter 2 METHODOLOGY In this study, first and fourth graders in four school systems completed a series of instruments: 1. Sociometric Rating Measure, 2. Home Play Survey, 3. Child Parental Acceptance— Rejection Questionnaire (PARQ), and 4. Pasao Pictorial Self—Concept Scale. Different kinds of information were used for different purposes: 1. Tabulations of responses to the Home Play survey were done to describe play activities of children at home, 2. Data were derived from the Home Play survey, PARQ and Self-Concept scale to establish correlations between frequency of parent-child play, parental acceptance and child’s self—concept, and 3. Based on outcomes of the Sociometric measure, selected children were observed in play activities during recess in school to describe differences between activities of those differing in popularity and self-concept. 49 50 Subjects Two sections (classes) each of Grades One and Four were recruited from 2 public and 2 private coeducational schools in Metro Manila. Data collection was not finished for two sections, one section each of Grades One and Four of the second private School due to lack of time. A total of 14 classes/sections were given the questionnaires (see Table 1 for distribution of classes). Table 1 Distribution of Classes from each Grade Level per School Public Public Private Private School 1 School 2 School 1 School 2 Grade 1 2 2 2 1 Grade 4 2 2 2 1 Two of the schools, one private and one public, were located in Las Pinas, a suburb of Manila, and the other two were located in inner city Manila. Las Pinas is an industrialized community, with factories and other manufacturing companies immediately outside the residential areas. Inner city Manila has less distinct zoning patterns, 51 where residences and commercial establishments are found in the same areas. The private and public school in each city were located in the same general vicinity, within 3 kilometers of each other. The children from the two schools in each area differ in socioeconomic status as evidenced by the tuition structures of the two types of schools. Children in the free public schools are generally from the lower income class while those in the two profit-oriented private schools are from the middle class (Gonzales, 1986). For this particular sample, the class distinction is further supported by the children’s residence wherein housing zones serviced by the private schools selected are predominantly middle class. Only one school, a private school, had a playground; it consisted of a vacant lot with monkey bars. The other schools had corridors and some space immediately outside the classrooms that could be used during recess. The classrooms and these areas are available for the children’s use during recess. The number of subjects varied according to analysis as only those with completed survey forms and observation data could be included. Some were included in the descriptive analysis but not in the correlational analysis due to missing data. Table 2 shows distribution of subjects 52 according to grade level, school and sex for the descriptive/ correlational analyses. Table 2 Number of Subjects in Play Survey Data GRADE ONE GRADE FOUR PUBLIC PRIVATE PUBLIC PRIVATE BOYS 64 (60) 3O (28) 52 (44) 54 (38) GIRLS 68 (65) 38 (34) 69 (63) 62 (36) NOTE: Numbers in parentheses refer to number of subjects in correlational analysis. For the observational study, the Sociometric outcomes were used to select four children from each section (two popular and two unpopular, and one of each sex within each popularity category). However, it was not possible to complete all observations groups for two reasons: 1) some sections did not have children with low popularity ratings and 2) only children with parental consent could be included. When more than two children qualified per section, random selection was done to identify which child was to be included in the observations. A total of 32 children identified through the procedure described above were observed. Table 3 shows the number of 53 children in the observational study. To permit assessment of inter-rater reliability, sixteen children were observed by two independent observers. Tafle3 Distribution of Subjects in Observations GRADE ONE GRADE FOUR PUBLIC PRIVATE PUBLIC PRIVATE BOYS 2 2 2 3 GIRLS 2 3 2 3 UNPOPULAR: (n = 13) BOYS 2 3 1 1 GIRLS 2 2 2 0 Instruments The children were asked to complete (fourth graders) or respond orally (first graders) to four questionnaires: 1. Sociometric Rating Measure (Appendix A) 2. Home Play Survey (Appendix B) 3. Child PARQ (Appendix C) 4. Pasao Pictorial Self-Concept Scale (Appendix D) 54 The first three instruments were translated into Filipino for children who are more comfortable in that language. Each instrument went through 3 translations: the first by a Psychology—Education double major, the second by a Psychology graduate (also the study’s research assistant), and the final one by the researcher herself. The research assistant reviewed the first translation, correcting words and phrases for simplicity and informality. She and the researcher went over the suggested revisions, and a third version was drafted through this collaboration. The research assistant tried out this third version with a class of first graders in a private school who were not part of the study. This was done to check if questions would be raised regarding meaning of words and/or sentences. Sociometric Rating Measure A rating scale proposed by Asher, Singleton, and Tinsky (1979) was used for this study. A sheet with a list of the child’s classmates was given to each child. Each child checked one of three possible choices: I like this classmate a lot, I somewhat like this classmate, and I do not like this classmate. For the first graders, the names were handwritten in large letters, and a researcher read the names out loud to each child. Figure 1 shows a sample of the sociometric questionnaire. I like this I somewhat I do not like classmate a lot like this classmate this classmate (Gustong-gusto (Medyo gusto ko (Hinding-hindi ko ang kaklaseng ang kaklaseng ko gusto ang ito) ito) kaklaseng ito) MARIA MIKO Fi re 1. Sample sociometric questionnaire. Scoring the sociometric scale Each child in a class was given a popularity rating. The choices were coded as follows: I like this classmate a lot : 3 I somewhat like this classmate: 2 I do not like this classmate: 1 Each child’s popularity rating was then computed using the formula: POPULARITY RATING = SUM OF RATINGS NUMBER OF SAME-SEX CHILDREN IN THE CLASS GIVEN THE TEST 56 Possible scores range from 1.0 to 3.0. The higher the rating, the more popular the child. Under the assumption that a popular child will be rated on the average between "somewhat like" and "like very much", with more ratings of the latter, the midpoint between these two ratings plus .01 was used as the cutoff score. On the other end, an unpopular child was assumed to be rated between "do not like" and "somewhat like", with more ratings of the former. The midpoint between these two ratings was used as the cutoff. Thus, the cutoff scores for determining popularity were the following: Categopy Range of Scores Popular 2.51 - 3.00 Average 1.51 — 2.50 Unpopular 1.00 — 1.50 In class sections where no child scored in the unpopular category, a cut—off score of 1.70 was considered a low rating if it also corresponded to a standard Z score of —1.5 (or popularity rating is 1.5 standard deviations below the mean for that class). This criterion was used for three cases . 57 Rationale for choosing peer rating There are different ways of measuring popularity as defined by peer likability. The most common methods of sociometric analysis are peer nominations, peer behavioral descriptions, and peer rating scales, each having advantages and disadvantages. The method of peer behavioral descriptions specifically identifies skills or roles of socially competent children (E.g. children are asked to identify "Someone who everybody listens to", in The Class Play of Bower, 1969, cited in Hughes, 1990). The problem of applicability to the Philippines of behaviors identified by American children as true of socially competent peers rules out the use of this method for this study. The most commonly used method is peer nominations, where each child is asked to select a restricted number of classmates with regard to some criterion (E.g. Name 3 classmates you most like to play with). It is the easiest and quickest to administer but has several disadvantages. Nomination procedures tend to measure friendship patterns rather than general peer acceptance and likability. There are also ethical concerns of negative nomination questions (Hughes, 1990). In response to this issue however, Hayvren and Hymel (1984) observed that preschool children did not discuss their choices after the 58 administration of positive and negative sociometric questions. They also did not interact differently after the sociometric measure. And lastly, nomination procedures appear to produce more variable and lower reliability coefficients ranging from .22 to .74 (Hartup, 1983). Asher (1979), and Oden and Asher (1977) compared the nomination procedure with peer rating. They found peer ratings to have significantly higher reliability coefficients than peer nominations. For this study, the peer ratings method appears to be most suited. Hughes (1990) in his review of studies using sociometric methods suggests that peer ratings in contrast to peer nominations measure acceptance and likability rather than friendship patterns. Because each child is rated by each member of a group, peer ratings indicate a child’s acceptance by all of the group members. It decreases the likelihood of a person not being chosen because he or she was momentarily forgotten. Also because all children are rated, average ratings are more stable over time. Test-retest correlations have been found to be quite high, ranging from .74 to .81 after a four-week interval (Hughes, 1990). The concurrent validity of peer ratings has also been established. Rubin and Daniels—Biernes (1983) looked at concurrent correlates of the sociometric ratings of 59 kindergarten and grade one children using the method of Asher and colleagues. They found sociometric status to correlate significantly and negatively with measures of solitary—dramatic play, and positively with positive peer interactions and prosocial strategies suggested in the Social Problem—Solving Skills Test. In this study, same-sex ratings were used because children’s ratings have been shown to reflect a sex bias (Singleton & Asher, 1977, cited in Rubin & Daniels—Biernes, 1983). Also to address the ethical questions of having children give negative ratings to their classmates, this study followed Hughes’s (1990) suggestion that children be specifically instructed not to talk about their choices, and to administer this particular measure in—between classes, rather than before free time. Home Play Survey This brief survey asked for some demographic information and asked several questions regarding the child’s play activities at home (Appendix B). Fourth graders were given the questionnaire to answer on their own and they were tested as a group. First graders needed individual assistance and their answers were noted down by research assistants. 60 Item #1, which asked about toys at home, was included by the researcher to supplement item #2 which asked about play activities at home. The second item was adapted from Seagoe's Play Report (Johnson, 1976). Seagoe’s observations scale has been used chiefly to compare differences in play styles between cultures and subcultures and also in identifying individual play styles. Reliability coefficients of observers’ ratings range from .84 to .99 for boys and .76 to .99 for girls. Seagoe’s items have also been found to be related to chronological age (r=.63 for boys and .65 for girls). Item #3 was added for prompting additional answers to #2. Items 4-9 were based on MacDonald and Parke’s (1984) findings using home observation measures. For their American sample, they found significant correlations between popularity of boys and girls and (1) parents’ physical play with their children, and (2) parents’ engagement in child’s activities (both measured as observed frequencies of physically active parent-child interactions, and active participation of parent in child’s activities respectively). 61 Child Parental Acceptance—Rejection Questionnaire (Child PARQ) This instrument was designed by Rohner (1980) to elicit respondents’ assessments of their experiences in terms of perceived parental warmth. This instrument is meant for children ages 7—12 to reflect on how their primary caretaker now treats them. For this study, children were asked to respond to the Child PARQ twice, once for the father, and once for the mother. This instrument is ideal for this particular study as it has been validated cross—culturally. Rohner (1980) cites the following studies: Saavedra (1980) found that Puerto Rican adolescents’ perceptions of self—esteem and self-adequacy varied with their perceptions of both maternal and paternal warmth. This finding is relevant to this research as the Philippine culture bears more resemblance to Puerto Rican than to western cultures. Rohner, Hahn, and Rohner (1980) found that 7-12 year old working class Korean immigrants perceived themselves as more rejected than their middle-class counterparts. Also they reported significantly lower feelings of self—esteem and self-adequacy. 62 Scoring the PARQ The PARQ is a 4—point Likert-type scale where respondents are asked to check degree of agreement with each statement. The choices are coded from 1 to 4. Points are added to form a subscale score. The higher the score in a subscale, the more that attribute is reported to occur. For example, the higher the score in the Warmth subscale, the stronger the agreement that the parent is warm. The higher the score in Hostility, the stronger the agreement that the parent is hostile. The Warmth subscale was reverse—scored when computing the composite score (PARQ total). The higher the PARQ total, the more rejecting the parent is perceived by the child. Table 4 shows minimum and maximum scores for the subscales and total scale. Reliability and validity of the Child PARQ Reliability was determined for this sample of 187 first graders and 181 fourth graders. In checking internal consistency, it was found that the reliability improved when items were removed from 3 subscales: Warmth, Hostility and Neglect. These items had negative or near zero correlations with the total subscale score. In the analysis, these items were removed. Furthermore, a different item was removed for the Neglect scale for the two grade groups. Thus 63 Table 4 Minimum and Maximum Scores for the PARQ Scale # of items Minimum Maximum {16111151 """""""" é """"""" é """" é; """""""" Hostility 14 14 56 Neglect 9 9 36 Rejection 3 3 12 PARQ total 35 35 140 reliability coefficients were determined separately for the Grade One and Grade Four samples. Tables 5 and 6 show internal consistency reliability coefficients before and after items were removed from each scale for the two groups. For the first grade sample, internal consistency reliability is satisfactory for the hostility subscale and for the total score (coefficient alpha > .70). The rest of the subscales had moderate reliability (coefficient alpha between .47 to .57). For the fourth grade sample, most of the subscales and the total score had satisfactory reliability (coefficient alpha > .60). The coefficient alpha is quite low for the 64 Rejection subscale. For this study, all results related to this subscale are rendered uninterpretable. Table 5 Internal Consistency Reliability Coefficients (alpha) for PAR Scales and Total Scale Grade One Sam le Coeff. alpha Item # Coeff. alpha Scale All items removed with item removed iléérléfi’Zéééfiééim’mféé """""" ii”””m""féé """ Warmth (Mother) .51 #3 .56 Hostility (Father) .81 #1 .83 Hostility (Mother) .77 #1 .81 Neglect (Father) .48 #7 .57 Neglect (Mother) .43 #7 .55 Rejection (Father) .56 none .56 Rejection (Mother) .47 none .47 Total Score (Father) .79 none .79 Total (Mother) .72 none .72 * Cronbach’s coefficient alpha, p< .001 for all correlations Concurrent validity has also been established for 3 of the 4 scales and the total scale for the Grade Four sample. The PARQ scales were correlated with a subcale of the Pasao 65 Self—Concept measure used in this study. This subscale included items on how the children View themselves in relation to mother, father, grandparents and teacher. As this self—concept scale partially measures self—concept derived from children’s relationship with parents, it is expected that it should be associated with a measure of parent-child relationship such as the PARQ. Table 7 shows that this self—concept scale is significantly related to the PARQ for the fourth grade sample. This finding was not found for the Grade One sample as the PARQ scale behaved differently for this group. This will be evident and will be discussed more thoroughly in the Results and Discussion sections. Pasao Pictorial Self—Concept Rating Scale This instrument was constructed by educational psychologists from the University of the Philippines (Munnariz & Pasao, 1988). It contains norms for Filipino children from urban and rural settings, including grades Kindergarten to Grade Four. Each child was shown a series of 34 plates depicting a child in different situations. The child in the picture is the same sex as the subject. Each Child was then asked to choose from among four pictures or figures the one he or she 66 Table 6 Internal Consistency Reliability Coefficients (alpha) for PAR Scales and Total Scale Grade Four Sam le Coeff. alpha Item # Coeff. alpha Scale All items removed with item removed Warmth (Father) .59 none .59 Warmth (Mother) .62 none .62 Hostility (Father) .75 #1 .77 Hostility (Mother) .78 #1 .80 Neglect (Father) .60 #10 .65 Neglect (Mother) .58 #10 .64 Rejection (Father) .34 none .34 Rejection (Mother) .23 none .23 Total Score (Father) .71 none .71 Total (Mother) .74 none .74 * Cronbach’s coefficient alpha, p< .001 for all correlations 67 Table 7 Correlations between PARQ Scales and Self—concept in Relation to Significant Others (Grade Four only) Self—concept in relation to others GIRLS BOYS Warmth (Father) .24” .29"‘ Warmth (Mother) .20‘ .22‘ Hostility (Father) —.22* -.21* Hostility (Mother) -.21“ —.09 Neglect (Father) -.24“ —.43”‘ Neglect (Mother) -.22" -.38“ Rejection (Father) —.03 —.05 Rejection (Mother) .10 -.04 Total Score (Father) —.30“ —.28“ Total (Mother) —.28" -.36“' 5 ' ' significant- at alpha— : 70—5 __________________________ " significant at alpha = .01 '” significant at alpha = .001 saw as himself or herself. There are separate plates for boys and girls. The placement of the four options is identical, but figures represent the appropriate sex. 68 This pictorial scale has two subscales: "Endearing traits", also referred to in this study as "View of Self", and "Relationships with others". The latter subscale includes situations with significant adults: parents, grandparents and teacher, and one plate shows a situation with peers. Scoring the Pasao scale Each of the response alternatives to a plate has a corresponding score. Scores for each plate range from 1 to 4. Scores are added for the lst 18 plates to form the Subscale "Endearing traits", and for the next 16 plates to form the subscale "Relationship with significant others." The two scales combine to form the total score. The higher the score, the higher the self-concept. Raw scores were used for most of the statistical analysis: mean scores and correlations. Percentile ranks were used to identify children high and low in self-concept in the observational part of the study. Percentile ranks were determined from the norms set by Munarriz and Pasao (1988) for urban children in Grades One and Four respectively. 69 Reliability and Validity of the Pasao Scale Both reliability and validity data were obtained from the test manual developed by Munarriz and Pasao (1988). Table 8 shows internal consistency and test—retest reliability coefficients for each of the subscales and the total scale. Table 8 Reliability Coefficients for the Pasao Pictorial Self— concept Scale Internal Test— Consistency Retest Scale 1 (Endearing Traits) .66 .52 Scale 2 (Relationships with .77 .55 Significant Others) Total Scale .79 .54 *all coefficients are significant at .01 alpha level To test the validity of this instrument, teachers were asked to rate children on such traits as: Confident, Not Confident, Healthy, Not Healthy. Total scores were correlated with the Pictorial Scale. (see Table 9) Although the figures below are not high, comparisons between children identified by teachers as high and low in 7O self-concept are probably most relevant for this study. For these groups of children, differences in scores on the Pictorial Self-concept Scale are reflected in teachers’ ratings. Table 9 Validity Coefficients for the Pictorial Scale Teacher Rating Scale 1 (Endearing Traits) .17* Scale 2 (Relationships with .07(NS) Significant Others) Total Scale .14* *coefficients are significant at .010 alpha level Those rated high in self-concept scored significantly higher in the Pictorial Scale than those rated low in self-concept by teachers (M: 117.08 and M=107.96 respectively, t=3.82, p<.01). The Scale appears to be able to discriminate between those with high and those with low self—concept. Peer Play Scale The original intention was to conduct structured observations using pre-set play categories. However, difficulties were encountered with the original structured 71 observation guide. There were two observers, one with a bachelor’s degree in Psychology, and the other a 3rd-year graduating Psychology major. The use of the structured play scale were discussed with them and they practiced the use of this scale in two visits each to a community playground where they separately observed children at play. They then practiced together in the two schools in Las Pinas, twice for each grade. The researchers found that: 1. Due to a short recess period, many of the children did not play, and when they did, did so only briefly*, 2. There were other apparently significant play and non-play behaviors not included in the checklist, and 3. They had difficulty deciding categories spontaneously and found it easier to describe all behaviors observed leaving categorization to be done later. Due to these observations, this researcher decided that unstructured observations would yield more accurate data, including observations that could be potentially useful. The observers then practiced simultaneously the unstructured observations with children not chosen for this part of the study. 'I' It was also observed that the four schools had limited play space and equipment, with only an empty lot or corridors available for space, and monkey bars in the lot of one school. 72 Inter—observer reliability Reliability was determined by computing the percentage of behavioral units that were similarly recorded by the two observers as part of the data used in this study. A behavior or group of behaviors was considered a unit if it seems to have a common theme, and co-exist to meet that theme. For example, in the observation "he read a textbook and started writing down answers in his notebook", reading and writing were considered part of the same unit as they are work—related and the latter could not be done without the reading aspect. On the other hand, the observation "he copied notes on the board while talking to his classmates" consisted of two units, as there are two independent behaviors, one work-related, the other social. Percent agreement was computed using the formula: Number of behavior units recorded by % Agreement = both observers x 100 Total number of behavior units observed The two researchers made simultaneous observations of 16 children out of a total of 32 children. The other observations were done independently. The first observer agreed with the second observer 71.3% of the time (72 similar observations out of a total of 101 behavioral units 73 observed), while the second observer agreed with the first 74.2% (72/97) of the time. Average percent agreement is 71.7% (144/198) . Procedure Pre-data gathering phase Permission to conduct observations of, and administer surveys to, two sections each of first and fourth graders were requested from principals of two public and two private coeducational elementary schools in Metro Manila. Once granted, the researchers requested the cooperation of the teachers of selected sections. Only sections with a fairly equal number of boys and girls were included. When there were several such sections, two from each grade level were randomly selected. Letters were sent to parents informing them of the study and asking for permission for possible inclusion of their child in the study (Appendices F & G). Children were also informed of the study and their assent to participate obtained (see Appendix H). Data-gathering phase 1) Administering Questionnaires All children with parental consent in the selected sections were given four questionnaires at two or three 74 different times. The Sociometric survey was done first as it was necessary for identifying children for the observations. It was immediately followed by either the PARQ for mother or for father. This was done in order to minimize the possibility of children discussing the peer ratings activity. The PARQ for mother and for father were administered one after the other. In half of the cases, the one for mother was given first, in the other half, the reverse was done. The three other instruments were given at another time in counterbalanced order. For the fourth graders, measures were administered in the classroom. For the first graders, individual administration was necessary. 2) Sampling from the sociometric scale Scores were tallied and the mean computed for each child. The researcher then categorized children as low or high in popularity and randomly selected two per category (one boy and one girl) from each class if there were more than two for each category. In three classes, there were no children with very low sociometry ratings. Thus there were fewer unpopular children included in the observations (see Table 3). The names of the selected children were given to the observers. The observers did not know who were rated as popular or unpopular. 75 3) Observations The selected children were observed unobtrusively during recess (see Appendix I for consent procedure). A total of 32 children were observed during recess. Twenty— five were observed on two separate occasions, two weeks apart; 7 were observed only once due to absenteeism during the weeks of observations. The observers situated themselves in the classroom/corridor or playground where they could see the three areas. They were far enough from each observed child that they would not come in contact with the child unless they were directly approached. But they were near enough that they could easily see what the child did. They wrote down all observations of the target child’s behaviors. They were also instructed not to compare their observations. Data Analysis Different sources of data were used to assess each problem. Analyses of the data were done accordingly: Problem #1 To test the hypothesis "There are differences in types of play engaged in by boys and girls, children in first and fourth grades, and children from the low and middle socioeconomic status" , cross-tabulations of play activities were done with the variables above. For this analysis, data 76 from 437 children were available. Play activities were based on responses to the Home Play Survey. Some activities or toys were grouped together due to similarity of function (e.g. videogames was used as the overall term for very similar games such as computer games, Sega/Nintendo, Gameboy). Differences between the groups designated above were determined by looking at frequencies and percentages of children giving similar responses. Problem.#2 To test the second hypothesis, "Playing with father/mother is significantly related to the child’s perception of his or her relationship with the father/mother, and with the child’s self-concept", a correlation matrix was produced for groups of children categorized according to Grade level, sex and school. The correlation matrix included the children’s scores in the four scales of the Child PARQ (for mother and father), total score of Child PARQ (for mother and father), scores in the Self-Concept subscales and total scale, and children’s reported frequency of their mothers/fathers playing with them. Only children with complete data in all relevant variables were included. Data from 187 first graders and 181 fourth graders were included in the analysis. 77 Regression analysis was also done to test possible models for predicting self—concept, with parental rejection and parent-child play as predictors. Problem #3 To answer the third problem, "What is the relationship between a child’s peer play activities, peer popularity and self-concept?", two hypotheses were presented. The first hypothesis that "there are differences in types of play engaged in by popular and unpopular children" was tested using observational data from 32 children identified as popular and unpopular. Cross—tabulations of frequency of occurrence of various behaviors were done to describe play activities as well as other social behaviors of these two groups. The behaviors were categorized according to the categories included in the original Peer Play Scale. New categories were also derived from the other observed non- play behaviors such as "fighting", "work-related", etc. (Appendices J & K). The second hypothesis was, "The more frequently the child engages in social play, the higher the peer popularity rating of that child, and the higher the child’s score in the self-concept scale". It was not possible to test this hypothesis with the data obtained as the play activities observed were not numerous enough for any useful quantification to be possible. Instead, the groupings were 78 refined such that play and other social behaviors of children varying in both popularity and self-concept could be described. As only twenty-five children had completed the Pasao Pictorial scale, only data from these children were included in the analysis. The 60th percentile was the median split for this group and it was used to categorize children as high or low in self—concept. Frequency counts of the different observed behaviors were cross—tabulated according to both popularity and self— concept. To see if a relationship exists between popularity and self—concept, two analyses were done: 1. Correlation between popularity ratings and self- concept scores, and 2. Comparison of mean self-concept scores of popular and unpopular children. Chapter 3 RESULTS The content of this chapter is arranged according to the order of the problem statements. The first two sections describe play activities in the home, in general, and specifically play with their parents. The third section describes play in school with their peers. Play activities at home This first section answers the question posed in the first problem, "What play activities are engaged in by children categorized according to sex, age, and socio- economic status?" It was hypothesized that there would be differences in play activities between boys and girls, between first and fourth graders and between low and middle SES children. Responses of the children to the question "What play activities do you play at home?" were tabulated, and frequency counts were done according to the groups mentioned above. Sex differences in play at home Responses of boys and girls reflect similarities and also differences. Table 10 summarizes boys’ and girls’ most frequent play activities. Similar data showing most frequently mentioned play activities by sex, age and school/ socioeconomic status are in tables 11 - 14. 79 80 There were three play activities mentioned by both sexes. Videogames, outdoor games, and ball games were among the top five most frequently mentioned games by both boys and girls. Table 10 Play activities of Boys and Girls at Home Play of Play of Rank Girls f % Boys f % n = 237 n = 200 1 dolls 206 87.0 cars 97 48.5 2 house 131 55.3 videogames 82 41.0 3 outdoor/motor 78 32.9 outdoor/motor 81 40.5 games games 4 videogames 70 29.5 ball games 74 37.0 5 ball games 48 20.0 "fighting" 69 34.5 dolls 6 stuffed toys 39 16.4 guns 48 24.0 Videogames include computer games, TV-hook-up games such as Nintendo, Sega, and portable games such as Game Boy. Outdoor games are group, motor games which require gross motor skills such as running, jumping or throwing, but which are differentiated from ball games. These include the more traditional Filipino games such as 81 "langit-lupa" (heaven—earth), "tumbang-preso" (knocking off a can), "patintero", "habulan" (tag), "luksong-tinik" (leap-thornbush), and "luksong-baka" (leap—cow).* The differences in boys’ and girls’ play seem to reflect sex—typed play activities. Playing with dolls and playing house were the top 2 play activities of girls but these were rarely mentioned by the boys. While boys also played with "dolls", these were not the baby dolls or Barbie as mentioned by girls. Boys play with "fighting" dolls which include figures such as the Ninja Turtles, GI Joe, X—men, superheroes and small plastic soldiers. Apparently even doll play is sex-typed. Boys play with cars, trucks and other models of vehicles; these were rarely listed by the girls. Ball games were also mentioned by more boys than girls. Seventy-four of the boys (37%) mentioned ball games compared to forty-eight (20.2%) of the girls (Chi Square = 14.29, p < .001). * loose translations by writer. See Appendix L for brief descriptions of these games. 82 Among the girls, only the older girls from the public schools mentioned ball games with relative frequency (for more detailed listings of play activities according to grade, sex and school, see Tables 11 - 14). Table 11 Play Activities of First Grade Girls at Home Public Private n = 68 n = 38 f % Rank f % Rank ééii; """"" £73313 ”””” i """"" 5.5”?!5 """ i" playing house 53 77.9 2 25 65.8 2 motor games* 23 33.8 3 6 15.8 (hide/seek,etc) videogames 17 25.0 4 10 26.3 3 stuffed toys 9 13.2 8 21.0 4 ball games 8 6 *hide/seek includes other outdoor, motor, group games such as "patintero", "tumbang preso", "langit—lupa", etc. 83 Table 12 Play Activities of Fourth Grade Girls at Home Public Private n = 69 n = 62 f % Rank f % Rank (3511;. """"" éémééié """ i """"""" {35"}??? """ I playing house 39 56.5 2 24 38.7 3 motor games 36 52.2 3 13 21.0 6 videogames 15 21.7 5 28 45.2 2 stuffed toys 7 10.0 15 24.2 4 ballgames 25 36.2 4 9 14.5 boardgames 14 22.6 5 Age differences in play activities at home For both boys and girls, more of the older children reported playing ball games than younger children. Fewer of the older girls played house, although this was still a pOpular activity among the fourth graders. Table 15 compares number of first and fourth graders reporting these play activities. The older boys did not report playing with robots which the younger boys did. There were some age differences which were also a function of socioeconomic status. More of the older 84 Table 13 Play Activities of First Grade Boys at Home Public Private n = 64 n = 30 f % Rank f % Rank a;£;;’§,;a;;""3;mééfi """" i ”””””” i """""""" cars/trucks 28 43.8 2 17 56.7 1 videogames 21 32.8 4 12 40.0 4 ballgames 18 28.1 5 7 23.3 6 guns 22 34.4 3 2 superheroes/ 17 26.6 6 13 43.3 2.5 soldiers robots 12 18.8 10 33.3 5 legos 2 3.1 13 43.3 2.5 *superheroes/soldiers include Ninja turles, wrestlers, "tau-tauhan", etc. children from the middle class played videogames compared to their younger counterparts and to the lower class children (see Table 16). Playing with legos was popular only among the middle class first grade boys (43%). This was not mentioned by the older children, and only by 2 children (3.1%) of the public school first grade boys. Table 14 85 Play Activities of Fourth Grade Bovs at Home motor games cars/trucks videogames ballgames guns superheroes soldiers 29 10 16 55. 19. 30. Private n 54 f % Rank 14 25.9 5 5 26 48.1 2 33 66.1 1 20 37.0 4 14 25.9 5.5 23 42.6 3 Table 15 Grade One and Grade Four Children Reporting Ball Games and Playing House Ball games Playing house (girls only) 12 .23*" 14 . 76*” significant I ,~.!—,:-o.;-.._~__.~_11_.-_; -,-_~-_ - 86 Table 16 Number of Children Playing Videogames by Age and Socioeconomic Status PUBLIC/LOW SES PRIVATE/HIGH SES Grade 1 Grade 4 Grade 1 Grade 4 TOTAL Yes 38 31 22 61 152 (28.8)a (25.6) (32.4) (52.6) (34 8) No 94 90 46 55 285 (71.2) (74.4) (67.6) (47.4) (65 2) TOTAL 132 121 68 116 437 3 numbers in parentheses refer to percentages NOTE: Chi Square = 22.96, p < .001 Socioeconomic/school differences in play activities at home There were some differences in play activities of children from the public and private schools. Stuffed toys and constructions toys ("Legos" were specifically mentioned) were among those mentioned by girls and boys respectively from the private schools but not by children from the public schools. Among the fourth graders, boardgames seem to be common among the private school girls but not the public school girls (see Table 12). While "dolls" were the most frequently mentioned play activity of girls regardless of socioeconomic status, there is an interesting difference in the kind of 87 dolls they have (See Table 17). More of the private school girls specifically mentioned "Barbie" (71.2%), an import from the U.S., while more of the public school girls (68.1%) used the more generic terms "dolls" or "manika" (Filipino word for doll; Chi Square = 12.09, p < .001) . Table 17 Dolls Mentioned by Public and Private School Girls Public Schools Private Schools n = 119 n = 87 f % f % Barbie 51 42.8 62 71.2 Dolls/Manika 81 68.1 37 42.4 *Percentages do not total 100 because a child may give both responses. Table 18 shows that outdoor play activities, both traditional outdoor games and ballgames, are not as common among the private school children as among the public school children. More of the children from the public schools, regardless of age and sex, reported outdoor games as one of their play activities at home. 88 Table 18 Outdoor Games of Public and Private School Children Public Private Chi Square (n = 253) (n = 184) f % f % Ball games 72 28.5 36 19.6 4.06‘ Outdoor/ 112 44.3 37 20.1 26.61"* motor games Further analysis shows that there is an interaction between sex, age and socioeconomic status. The difference between public and private school children reporting ball games was evident only among older girls. There was no difference between younger children and older boys from the public and private schools (see Table 19). 89 Table 19 Number of Children Playing Ball Games by Age. Sex and Socioeconomic Status PUBLIC PRIVATE CHI f 9 f % SQUARE GRADE ONE: girls 8 11.8 6 15.8 .08 boys 18 28.1 7 23.3 .06 GRADE FOUR: girls 25 36.2 9 14.52 6.92" boys 29 55.8 10 37.0 3.02 ” significant at alpha = 01 Parent—child play This section is divided into two parts. The first answers the second problem "What is the relationship between playing with parents, child’s report of parent- child relations, and child’s self-concept?" The second part describes patterns of parent-child play according to sex, age and socioeconomic status. While this was not part of the problem, the researcher decided to include this analysis as differential patterns could give more insight into the relationship between the variables state above. 90 Playing with parents, parental warmth and self-concept It was hypothesized that the more frequent the play interaction between child and parent(s), the more the child will perceive parental warmth (lower scores in the PARQ), and the higher the child’s self-concept (higher scores in the Pasao). Children were asked how often their mothers and fathers played with them. They were asked whether their mothers and fathers play with them almost everyday, 3-4 times a week, 1-2 times per week, or do not play with them at all. Responses to this question were correlated with the subscales of the Parental Acceptance-Rejection Questionnaire and the Pasao Self-Concept Scale. Tables 20 and 21 show descriptive data for each of these scales. Correlations between age, sex, and socioeconomic status and each of the PARQ and self-concept scales are presented in the appendix (see Appendices M - 0). Three- way analyses of variance, with age, sex, and socio- economic status as independent variables, and the PARQ and self-concept scales as dependent variables, are presented in Appendix P. A correlation matrix was derived for each of the subgroups of children classified according to sex, age and socioeconomic status. These subgroups were kept l—— .. _ ’ _ 2“”, 91 separate if the correlations differed in terms of direction and magnitude (1 e. having reached statistical significance or not). These analyses show that patterns emerged differently for Grade One and Grade Four data. Thus for this section, data are presented separately for the two age groups. Correlation matrices show both the uncorrected coefficients and the correlations corrected for attenuation. Plavinq with parents, parental warmth and self—concept: Grade Four data No significant correlations emerged between frequency of parent—child play and either the PARQ and the Self— concept scales from data obtained from the boys (see Table 22). Thus the hypothesized relationship between these variables was not confirmed for boys. Table 23 shows correlations obtained from girls. There was a similar pattern for mothers and fathers in that frequency of parent—child play was positively correlated with warmth, and negatively correlated with neglect and hostility by parents. The more frequently a parent engaged in play with the daughter, the warmer and the less neglecting and hostile the parent was perceived by her. The neglect subscale had the highest correlation with frequency of playing with parents. The rejection subscale 5W.— 92 Table 20 Mean Scores in the Pasao and PARQ (Grade Four Data) PRIVATE PUBLIC GIRLS BOYS GIRLS BOYS n = 36 n = 38 n = 63 n = 44 FATHER: WARMTH 20.19 23.00 23.32 22.52 (2 80) (3 17) (3 43) (3 80) HOSTILITY 19.56 22.00 22.19 22.68 (4 65) (6 66) (6 10) (7 85) NEGLECT 12.58 13.53 14.17 15.77 (3.28) (4.25) (3.43) (4.95) REJECTION 5.47 6.37 5.54 6.25 (1.90) (1.92) (2.19) (2.33) TOTAL 53.83 57.78 58.29 60.77 (9.17) (11.36) (11.20) (16.40) MOTHER: WARMTH 20.97 23.97 23.94 22.36 (3.10) (2.71) (3.16) (4.06) HOSTILITY 20.92 20.91 23.10 23.73 (5.57) (5.92) (7.14) (8.76) NEGLECT 12.89 12.32 14.11 15.98 (3.31) (4.58) (3.17) (5.09) REJECTION 5.81 6.11 5.71 6.23 (1.94) (1.71) (2.19) (2.08) TOTAL 54.69 57.56 58.89 63.36 (11.23) (12.40) (12.24) (16.22) SELF—CONCEPT 62.89 61.24 64.41 64.30 :VIEW OF SELF (3.98) (5.39) (7.06) (4.09) :RELATION TO 54.17 53.16 55.29 53.50 OTHERS (6.75) (6.55) (8.74) (5.40) :TOTAL 117.03 113.03 118.35 117.86 (9.50) (10.05) (9.90) (7.83) NOTE: Standard deviation values within parentheses. Table 21 93 Mean Scores in the Pasao and PARQ (Grade One Data) FATHER: WARMTH HOSTILITY NEGLECT REJECTION TOTAL MOTHER: WARMTH HOSTILITY NEGLECT REJECTION TOTAL SELF-CONCEPT :VIEW OF SELF :RELATION TO OTHERS :TOTAL Standard deviation values within parentheses. PRIVATE GIRLS BOYS n 35 n = 27 20.52 20.56 (2.79) (2.91) 21.64 21.78 (6.70) (7.52) 14.55 15.11 (4.23) (4.19) 5.97 5.74 (2.28) (2.51) 58.00 59.22 (10.07) (12.85) 20.46 20.56 (3.10) (2.99) 21.83 21.85 (6.44) (7.26) 14.31 15.07 (3.96) (4.24) 5.97 5.85 (2.20) (2.18) 59.29 59.44 (9.70) (13.37) 62.51 63.07 (7.09) (6.15) 53.80 47.52 (5.83) (6.95) 117.23 110.70 (8.10) (10.88) PUBLIC GIRLS BOYS n = 65 n = 60 21.11 20.77 (2.83) (2.91) 22.02 21.30 (7.85) (7.05) 15.66 14.85 (4.57) (4.23) 5.57 4.68 (2.42) (2.00) 59.03 56.63 (14.50) (11.81) 21.32 21.45 (2.60) (3.02) 21.83 21.15 (7.11) (6.78) 15.86 14.80 (4.46) (3.73) 5.68 4.78 (2.32) (1.07) 58.69 55.88 (13.00) (11.62) 60.43 61.48 (5.81) (6.82) 54.63 53.85 (6.54) (6.18) 115.14 115.20 (10.15) (11.76) 94 Table 22 Correlations between Freguency of Parent—son Play and Scores in the PARQ: Grade Four Data (n = 82) PLAYMOM PLAYDAD ____________________________________________________________ I Warmth of Mother .14 (.17) ‘ Hostility of Mother .09 (.10) ) Neglect by Mother —.09 (-.11) Rejection subscale (Mother) .01 (.02) Rejection Total (Mother) -.01 (-.01) Warmth of Father .03 (.03) Hostility of Father .01 (.01) Neglect by Father —.15 (—.19) Rejection subscale (Father) -.001 (-.001) Rejection Total (Father) -.07 (- 08) Self-concept -.01 (-.02) -.18 (—.22) (view of self) Self-concept (in relation .02 (.02) -.09 (-.11) to significant others) Total Self—concept -.17 (—.19) — 05 (- 05) NOTE: PLAYMOM is correlated with mother variables only, PLAYDAD with father variables. correlations in parentheses are corrected for attenuation 'k significant at alpha 5 .05, otherwise not significant 95 Table 23 Correlations between Frequency of Parent—daughter Play and Scores in the PARQ: Grade Four Data (n = 99) PLAYMOM PLAYDAD Warmth of Mother .22‘ (.28)‘ Hostility of Mother —.20* (-.22)* Neglect by Mother -.36* (-.45)* Rejection subscale —.10 (—.21) (Mother) Rejection Total (Mother) -.28* (.33)‘ Warmth of Father .27* (.35)* Hostility of Father -.20* (-.23)* Neglect by Father -.35* (-.43)* Rejection subscale (Father) —.03 (-.05) Rejection Total (Father) -.30* (-.36)* Self-concept .15 (.18) .16 (.20) (view of self) Self-concept (in relation .06 (.07) .05 (.06) to significant others) Total Self-concept .11 (.16) .14 (.16) NOTE: PLAYMOM is correlated with mother variables only, PLAYDAD with father variables. correlations in parentheses are corrected for attenuation * significant at alpha 5 .05, otherwise not significant 96 was not significantly correlated with parent-child play. Given the poor reliability of this subscale, this finding is uninterpretable. Frequency of parent—child play was not significantly correlated with the self-concept scales; these results do not support the hypothesis that these variables are correlated. Correlations between parental warmth and self-concept: Grade Four data For this analysis, data from boys and girls were combined as patterns were very similar between these two groups, i.e. correlations between the same sets of variables were both significant and in the same direction for these two groups. There were significant correlations between the Parental Acceptance-Rejection and the self-concept subscales. The correlations were weaker for the "View of self" subscale, and stronger for self-concept in relation to significant others (see Table 24). In general, the less rejecting (or the more accepting) a parent, the higher the self-concept of the child. The Neglect—self—concept relationship appeared to be strongest. Neglect Of parent moderately correlated with self-concept in relation to Others. The less neglecting a 97 parent, the higher the self-concept of the child in relation to significant others. In summary, the hypothesized relationships were partially shown to be true. Among the fourth grade girls frequency of parent-daughter play was correlated with self- concept. For both boys and girls, perceived parental warmth was correlated with self-concept. Predicting self-concept: Grade Four data A regression analysis was done to check how well frequency of parent—child play and parental rejection (PARQ total score) predict self-concept. Separate analyses were done for boys and girls as the patterns of correlations differed between these two groups. Separate analyses were also done for mother and father. For the female fourth graders, parental rejection directly correlated with self—concept. Frequency of parent- daughter play was no longer significantly correlated with self—concept once parental rejection has been taken into account (see Tables 25 and 26 for correlation coefficients and beta weights respectively). For this group, parental rejection served as an intervening variable between parent- child play and child’s self-concept. 98 Table 24 Correlations between PARQ Scales and Self—Concept Scales: Grade Four Data (n = 181) SELF-CONCEPT view of relation to TOTAL self significant others MOTHER: Warmth .14* (.22)* .22* (.32)* .17* (.24)* Hostility - 20' (-.28)* -.22* (-.28)* -.24* (—.30)* Neglect -.09 ( 14) -.29* (-.41)* -.28* (-.39)* Rejection —.02 (-.05) -.09 (—.21) —.05 (—.12) (subscale) Rejection —.18* (-.25)* —.31* (—.41)* -.29* (-.38)' (total) FATHER: Warmth .16' (.26)‘ .26' (.39)‘ .23' (.34)’ Hostility - 18* (—.25)* -.22* (-.29)* -.22* (-.28)* Neglect -.09 (-.14) -.33* (-.47)* -.31* (-.43)* Rejection (subscale) - 02 (-.04) -.06 (-.12) -.01 (-.02) Rejection total -.12 (-.18) -.28* (-.38)' -.25* (-.33)* NOTE: figures in parentheses are correlations corrected for attenuation * significant at alpha 5 .05, otherwise not significant. 99 Table 25 Correlations Between Parent—child Play, Parental Rejection, and Self-concept: Grade Four girls (n=99) Parent-child Parental Self—Concept Play Rejection MOTHER: Parent-child 1.00 -.33 .16 Play Parental 1.00 -.38 Rejection Self-concept 1.00 FATHER: Parent—child 1.00 -.36 .16 Play Parental 1.00 -.33 Rejection Self-concept 1.00 NOTE: correlations presented are corrected for attenuation Figure 2 illustrates a model representing this relationship between the three variables. 100 MOTHER: -.33 -.37 Mother-daughter --------- > Maternal --------- > Self—Concept play Rejection FATHER: —.36 -.31 Father—daughter --------- > Paternal --------- > Self-Concept play Rejection Figure 2. A model predicting self-concept of fourth grade girls. For the male fourth graders, different models are presented for mother and father. Only rejection by the mother was directly related to self—concept. The less rejecting the mother, the higher the self-concept of the child. Frequency of playing with mother was neither related to self-concept nor to parental rejection (see Table 27). Relations with the father predicted self—concept in a different way. Rejection and father-son play independently predicted self-concept of fourth grade boys. The less rejecting the father, and the less frequent father-son play, the higher the self-concept of the child. Parent—child play was not related to parental rejection. Figure 3 shows this relationship. Tables 28 and 29 show correlations and beta weights/partial correlations between the three variables respectively. 101 Table 26 Beta Wei hts Partial Correlations between Parent-child Play, Parental Rejection and Self-Concept: Grade Four girls Parent- Parental Self S.E. R child play Rejection Concept of B MOTHER: Parent-child 1.00 -.33 .04 .12 .38 Play Parental 1.00 -.38 .13 Rejection Self-concept 1.00 FATHER: Parent-child 1.00 -.36 .05 .12 .33 Play Parental 1.00 -.31 .14 Rejection Self-concept 1.00 102 Table 27 Correlations Between Mother-child Play, Maternal Rejection, and Self-concept: Grade Four boys (n=82) Parent—child Parental Self—Concept Play Rejection MOTHER: Parent-child 1.00 —.01 -.05 Play Parental 1.00 -.36 Rejection Self—concept 1.00 NOTE: correlations presented are corrected for attenuation —.21 Father-son play ———————1__ > Self— r——> concept Rejection by Father -.27 Figure 3. A model predicting self—concept of fourth grade boys. 103 Table 28 Correlations Between Father-son Play, Paternal Rejection, and Self-concept: Grade Four boys (n=82) Parent-child Parental Self-Concept Play Rejection FATHER: Parent-child 1.00 -.08 -.19 Play Parental 1.00 -.25 Rejection Self—concept 1.00 NOTE: correlations presented are corrected for attenuation Table 29 Beta Wei hts Partial Correlations between Parent—child Play, Parental Rejection and Self-Concept: Grade Four Boys Parent— Parental Self S.E. R child play Rejection Concept of B FATHER: Parent-child 1.00 -.08 -.21 .12 .33 Play Parental 1.00 -.27 .14 Rejection Self-concept 1.00 104 Playinq.with parents. parent—child relationship and self— concept: Grade One data Grade One data were combined for all subjects. In contrast to the Grade Four data, the relationship between frequency of parent-child play and the PARQ scales were the reverse of what was hypothesized. Table 30 shows that frequency of parent-child play was positively correlated with most of the scales measuring perceptions of negative parent behaviors, and inversely correlated with parental warmth. Generally it can be stated that the more frequently a parent plays with a child, the more rejecting (or less accepting) the parent is perceived by the child. This finding is the opposite of what was expected, and contradicts that Of the fourth grade data. Possible explanations are in the discussion section. When play was correlated with self-concept, the expected outcome was found to be true. The more frequent the parent-child play, the higher the self-concept of the child. 105 Table 30 Correlations between Fregpency of Parent—child Play and Scores in the PARQ: Grade One Data (n = 187) PLAYMOM PLAYDAD Warmth of Mother -.19* (-.25)* Hostility of Mother .29' (.32)* Neglect by Mother .27* (.36)‘ Rejection subscale .14 (.20) (Mother) Rejection Total (Mother) .30* (.35)* Warmth of Father -.24* (—.28)* Hostility of Father .22* (.24)* Neglect by Father .13 (.17) Rejection subscale (Father) .14 (.19) Rejection Total (Father) .22* (.25)* Self-concept (view of self) .23' (.28)* .30‘ (.37)* Self-concept (in relation .24* (.27)‘ .29' (.33)* to significant others) Total Self-concept .32' (.36)* .39‘ (.44)* NOTE: PLAYMOM is correlated with mother variables only, PLAYDAD with father variables. correlations in parentheses are corrected for attenuation * significant at alpha 5 .05, otherwise not significant 106 Correlations between parental warmth and self—concept: Grade One data Unlike the fourth grade data, there were more significant correlations between the Parental acceptance- rejection scales and "view of self" self—concept rather than the self—concept in relation to others subscale (see Table 31). The correlations between the PARQ and self—concept scales were also the Opposite of what was expected. While the correlations were weak, the direction of these coefficients shows that the more rejecting (or less accepting) a parent, the better the child views himself or herself. Among the PARQ subscales, it was the Rejection subscale which has the strongest relationship (although still weak) with self—concept, i.e. the more rejecting a parent is perceived, the higher the self-concept Of the child. Predicting self-concept: Grade One One model is presented for first graders since data from boys and girls, using mother and father variables, fit the same model. Parent—child play and parental rejection both significantly predict self-concept. Figure 4 shows that the more frequent the parent—child play, and the more rejecting the parent, the higher the self—concept of the child. 107 Table 31 Correlations between PARO Scales and Self—Concept Scales: Grade One Data (n = 187) SELF-CONCEPT view of relation to TOTAL self Significant others MOTHER: Warmth -.17* (-.28)* -.04 (- 06) -.18* (-.27)* Hostility .22* (.30)* .17* (.22)* .26' (.33)* Neglect .11 (.18) .12 (.18) .19' (.29)‘ Rejection .26‘ (.47)’ .09 (.15) .22' (.36)* (subscale) Rejection .21’ (.30)* .08 (.11) .21' (.28)* (total) FATHER: Warmth -.15* (—.25)* -.09 (- 14) —.20* (-.30)* Hostility .22‘ (.24)* .15* (.19)* .24* (.30)* Neglect .14‘ (.23)‘ .14' (.21)‘ .21' (.31)’ Rejection .26* (.43)* .07 (.11) .21' (.32)‘ (subscale) Rejection .21* (.29)* .11 (.14) .23* (.29)‘ (total) NOTE: correlations in parentheses are corrected for attenuation * significant at alpha 3 .05, otherwise not significant 108 Parent-child play contributes more to the variation in self- concept than does parental rejection. Tables 32 and 33 show correlations and beta weight matrices respectively for the three variables. MOTHER: ———> Parent—child play .25 -——> Parental Rejection Mother FATHER: F__> Parent-child play .35 -——> Parental Rejection Father Figure 4. A model predicting .40 h .18 .30 E .19 Self— concept Self— concept self—concept of first graders. . 'n'rfir'firr- 109 Table 32 Correlations Between Parent—child Play, Parental Rejection, and Self-concept: Grade One (n=187) Parent—child Parental Self—Concept Play Rejection MOTHER Parent-child 1.00 .25 .44 Play Parental 1.00 .28 Rejection Self—concept 1.00 FATHER: Parent-child 1.00 .35 .36 Play Parental 1.00 .29 Rejection Self-concept 1.00 NOTE: correlations presented are corrected for attenuation 110 Table 33 Beta Wei hts Partial Correlations between Parent-child Play, Parental Rejection and Self-Concept: Grade One Parent— Parental Self S.E. R child play Rejection Concept of B MOTHER: Parent-child 1.00 .25 .40 .07 .47 Play Parental 1.00 .18 .09 Rejection Self-concept 1.00 FATHER: Parent-child 1.00 .35 .30 .08 .40 Play Parental 1.00 .19 .09 Rejection Self—concept 1.00 .-_-~-——-—_-’------—--_-‘--_--——‘~_—--—--—-----—--—---—’---- Other results: parent-child play activities As in the first section, responses of the children were tallied according to sex, grade level and school. Tables 34 to 37 show that there was quite a diversity of parent-child play activities. As the children in this study generally gave only one or two responses to the question "What games do you play with her (mother)/ him (father)?", there were 111 few common responses. A condensed table showing only common responses is shown below (see Table 38). Table 34 Mother-daughter Plav Activities —-———---————-————---——---———-—_-———-———_————-——————-—--—— GRADE ONE GRADE FOUR Public Private Public Private n = 68 n = 38 n = 69 n = 62 f f f f dolls 7 5 1 4 house 8 2 1 7 videogames 10 6 5 15 boardgames 3 2 3 1 jokes/stories 2 5 6 4 NONE 19 7 29 8 Other responses: ball games piano books stuffed toys cards . sungka garter teacher-teacher/doctor-doctor ghosts tickling guns legos outdoor games 112 Table 35 Mother-son Play Activities GRADE ONE GRADE FOUR Public Private Public Private n = 64 n = 30 n = 52 n = 54 f f f f videogames 16 6 4 19 boardgames 1 2 9 tag 7 1 4 6 NONE 21 14 34 12 Other responses: badminton jokes ball games legos bike puzzles cards reading cars sungka ghosts teacher-teacher GI Joe tennis guns tickling house 113 Table 36 Father-daughter Play Activities GRADE ONE GRADE FOUR Public Private Public Private n = 68 n = 38 n = 69 n = 62 f f f f videogames 4 7 6 11 boardgames 1 1 4 12 jokes/stories 4 3 1 7 outdoor games 16 1 8 10 NONE 16 8 32 14 Other responses: ball games piano bike singing cards scissors-paper-stone cars stories dolls stuffed toys ghosts swimming guns teacher—teacher horsey tennis house tickling lambingan television market/store wrestling 114 Table 37 Father-son Plav Activities GRADE ONE GRADE FOUR Public Private Public Private n = 64 n = 30 n = 52 n = 54 f f f f videogames 9 6 4 10 ball games 2 6 5 15 NONE 13 5 33 13 Other responses: bike kite billiards lambingan boardgames legos cards marbles cars outdoor games darts pingpong drawing reading ghosts robot guitar scissors-paper-stone guns sipa horsey tickling house TV jokes wrestling/karate 115 Table 38 Play Activities Of Boys and Girls with Mothers and Fathers condensed Mother Father Girls Boys Girls Boys n = 237 n = 200 n = 237 n = 200 Videogames 15.2 22.5 11.8 14.5 Boardgames 8.4 —-—- ---— ---- Ballgames —-—- —--— -—-— 14.0 Outdoor games —--- 9.0 14.3 -——- (Active like tag) NONE 26.2 40.5 29.1 32.0 *Blanks indicate very small frequencies. Sex comparisons in parent-child play Videogames were the most frequently mentioned parent- child play activity. These were mentioned by both boys and girls with both mothers and fathers. The boys reported more active games with either parent, "tag" with their mothers and ball games with their fathers. 116 Girls, on the other hand, do also engage in active games but only with their fathers. Boardgames were reported to be played with their mothers. In analyzing types Of boardgames, there also existed a sex:difference (see Table 39). As there was a very similar gmattern for boys and girls, data were combined to reflect differences in boardgames played with mothers and fathers “Chi Square = 9.11, p = .003) Table 39 Boardgames Played with Mothers and Fathers Mother Father f f Word games 16 6 Strategy games 5 17 Others 7 5 -———---—-----_—--—--—————---—----————---——--—- More mothers play "word games" or boardgames which require or test verbal skills. Examples are Scrabble (the most frequently mentioned) and Boggle. More fathers play "strategy games" such as chess and checkers. These games require not only strategic planning, they are also "combat" 117 games wherein the objective is to conquer the game pieces of the opponent. In analyzing "None" responses (i.e. to the question "What games do you play with her [mother]/ him [father]?"), more boys than girls reported not playing any games with either mother or father. The difference is largest when one looks at playing with mother. Chi Square analyses of number of boys and girls who play games and who do not play games with their mothers and fathers show that the difference is significant only for the mothers (see Tables 40 & 41). That is, more boys than girls reported that their mothers do not play any games with them, while more girls than boys reported the opposite. There was no significant difference in the percentage of fathers who play with their sons and daughters. Table 40 Number of BO 5 and Girls Whose Mothers Pla do not la with Them Girls Boys Yes 175 (73.84%) 119 (59.50%) Mothers play No 62 (26 16%) 81 (40.50%) 118 Table 41 Number of Boys and Girls Whose Fathers Play/do not play with Them Girls Boys Yes 168 (70.88%) 136 (68.0%) Fathers play No 69 (29.11%) 64 (32.0%) Age and school differences in parent—child play The only significant differences were in the responses of "None" to the question, "What games do you play with your mother/ father?". First, there was an interaction between age and socioeconomic status/school in that the difference between private and public school children was significant for the fourth graders. Table 42 shows that more public school than private school children reported that their parents do not play with them. This was true only for the fourth graders. First grade public and private school children do not differ in the percentage of "None" responses. 119 Table 42 Children by Grade and School Reporting that Parents Do Not Play with Them PUBLIC PRIVATE CHI SQUARE f % f % Grade One: Mother 40 30.3 21 30.9 not significant Father 29 22.0 13 19.1 not significant Grade Four: Mother 63 52.1 20 17.2 30.05”* Father 65 53.7 27 23.3 21.85"* *" significant at alpha = 001 Second, there was also an age difference, although the pattern differed in the different socioeconomic classes. Among the public school/ low SES children, more of the older children reported that their parents do not play any games with them (Chi Square = 11.50 and Chi Square = 25.911 for mother and father respectively, p < .001 in both cases). By fourth grade, more than half of the public school children said that their parents do not play with them. 120 Popularity, play and self-concept For this section, data from observations of the 32 children selected on the basis of their popularity scores were analyzed. Twenty—five children were observed on two occasions, 7 were Observed once. In this analysis, data for popular and unpopular children were summed over sex, grade level and type of school. This was done as there were very few cases per cell, and the patterns among these different categories appeared to be very similar. Table 43 shows number of cases per cell. Table 43 Number of Children Observed per Category GRADE ONE GRADE FOUR boys girls boys girls TOTAL POPULAR 4 5 4 5 18 UNPOPULAR 5 4 3 2 14 TOTAL 9 9 7 7 32 121 All Observed behaviors were listed and categories were then derived (see Table 44). This included play as well as other behaviors. The Peer Play Scale was also used to categorize play activities Observed (see Table 46). For each child, a category is checked if it was Observed to have occurred at least once. For example, if a child played tag more than once, the category "played with others" was checked for that child without taking frequency into account. Number of occurrences was not considered in order to control for number of observations; some children were observed twice, and others (7 of the 32) were Observed only once. This section answers the question "What is the relationship between a child’s peer play activities, peer popularity and self-concept?" It was hypothesized that: 1. There are differences in types of play engaged in by popular and unpopular children, and 2. The more frequently the child engages in social play, the higher the peer popularity rating of that child, and the higher the child’s score in the self— concept scale. The second hypothesis was not tested as originally planned as the data did not allow sufficient quantification of "frequency of occurrence of social play." Instead, a different approach was taken examining differences in types 122 of play engaged in by children with high and low self— concept. In addition, since unstructured general observations were done instead of structured observations of play behaviors only, other social behaviors aside from play were also noted. While this study centers on play activities, it was decided that other social behaviors be included in the analysis in order to get a more comprehensive view of how differently children Of varying popularity and self-concept behave. Popular and unpopular children Play behaviors of popular and unpopular children Because of the Short 20-minute recess that the schools allowed, and because of the lack of play space and facilities, play activities Observed were very limited in number. Nevertheless, some differences in play behaviors emerged between popular and unpopular children. Table 44 shows that ten of the popular children (55.6%) played with others. In contrast, only 4 (28.6%) of the unpopular children did so. Table 39 lists play activities Observed. Of the 18 activities Observed from play of popular children, 12 involved 3 or more participants, while 6 were dyadic interactions; no one played alone. Among the unpopular children, 6 out of the 10 activities were group play, 2 were dyadic, and 2 were isolated. 123 Table 44 Behaviors During Recess of Popular and Unpopular Children POPULAR UNPOPULAR (n = 18) (n = 14) f % f % EATING with others 13 72.2 1 7. alone 0 0.0 9 64. PLAYING with others 10 55.6 4 28. alone 0 0.0 2 14. did not play 8 44.4 8 57. WORK-RELATED 8 44.4 6 42. TALKING with 2 or more 14 77.8 4 28. with one child 2 11.1 2 14. APPROACHED 8 44.4 3 21. OTHERS APPROACHED BY 9 50.0 1 7. OTHERS TAUNTING/TEASING 2 11.1 5 35. FIGHTING physical 1 5.6 2 14. verbal 0 0 0 5 35 AGGRESSIVE/* 1 5.6 5 35. HYPERACTIVE ALONE ENTIRE 0 0.0 8 57. RECESS This includes behaviors such as aimless running around the room, climbing chairs/shelves, throwing things, etc. 124 Table 45 Play Activities of Popular and Unpopular Children # OF PARTICIPANTS MOTOR GAMES alone 2 33 motor aggres- physica sive POPULAR (n = 18) scissor-paper stone x Chanting (market song) x tag (n=2) x tug-of—war hitting hands pinching pulling x chasing/tickling (n=2) rubber bands (n=2) gameboy hide-and-seek chasing (n=2) startling game street fighter x x (karate) > F?" r’r" - fl. 2 '1!" J‘nJJ" rJT r "2:: r5 ~ ,- r. . “’7 m" . r JP __...:u \‘N , r- 3 I “2 " ”'17,.‘Trhlrz t ”'1 l (”Uy,") L a rr.r..‘."""xv .2"