Citation
Individual and familial correlates of children's fantasy play

Material Information

Title:
Individual and familial correlates of children's fantasy play
Creator:
Dennis, Lorraine Bradt, 1921-
Publication Date:
Language:
English
Physical Description:
vii, 82 leaves : ; 28 cm.

Subjects

Subjects / Keywords:
Child psychology ( jstor )
Children ( jstor )
Childrens games ( jstor )
Fantasy ( jstor )
Fathers ( jstor )
Mothers ( jstor )
Parents ( jstor )
Pretend play ( jstor )
Questionnaires ( jstor )
Toys ( jstor )
Child development ( lcsh )
Child psychology ( lcsh )
Parent and child ( lcsh )
Play ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Includes bibliographical references (leaves 78-81).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Lorraine Bradt Dennis.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Lorraine Bradt. Dennis. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
025273871 ( ALEPH )
02759851 ( OCLC )

Downloads

This item has the following downloads:


Full Text












INDIVIDUAL AND FAMILIAL CORRELATES OF CHILDREN'S FANTASY PLAY










By

LORRAINE BRADT DENNIS












A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY






UNIVERSITY OF FLORIDA

1976
I II '" I,~ 4.














ACKNOWLEDGMENTS


With humility and deepest gratitude, my indebtedness to a large number of individuals must be acknowledged. Without their efforts and good will, this dissertation truly could not have been accomplished.

I need, especially, to thank Dr. Ira J. Gordon, the chairman of my committee, for his incisive guidance, his wise instruction and for his unfailing calm and cheerful support during the dark moments which are part of the life of every graduate student.

Dr. Robert Jester and Dr. Barry Lester, members of my committee, are also to be thanked for their contributions to the design of this project, and their constructive criticisms which enhanced its execution.

Dr. Jerome L. Singer's interest, his loan of the Barron Inkblots and his help in obtaining the services of one of his graduate students, Claudia Morella, for a reliability study places me in his debt.

Next, I must cite, fondly, the six students who assisted in the data gathering and scoring of protocols from beginning to end, giving tirelessly of their time and good thinking: Marilyn Packard-Luther, Elise Carlson, Ann Leaver, Mary Jane Scott, Don Whitworth and Scott Hicks.

Beverly Malin, Philip Goodwin, Claudia Morella, Tullio Pitassi and Elizabeth Hallenbeck all assisted with reliability studies and, in the process, offered many ideas for improving our techniques.

A special note of thanks goes to my son, Brian Dennis, whose

assistance in coding and computerizing the data was invaluable. Stephen Sledjeski took time from his busy life to program the multiple regression equations.
ii








Mr. Henry V. Diodati, superintendent of the Portsmouth, R.I., schools and the four principals of the elementary schools, Mr. Lawrence P. Mello, Mr. Lawrence Jones, Mr. Robert L. Crudup and Mr. Russell L. Carter must be thanked for their generous help in making rooms available for our testing and facilitating our efforts. Teachers of the kindergarten and first grades of these schools, too, were magnificently tolerant of the many class interruptions as we called for and returned the children during

days of testing.

I wish also to warmly acknowledge my gratitude to the wonderful

children and the splendid families who gave so generously of their time and even more importantly, so promoted this project with their interest.

Finally, it is impossible to express how much this entire undertaking owes to the steady, optimistic encouragement of Larry Dennis and our children, Patrick, Brian, Deborah and Thomas, each of whom in uncounted ways enabled my study and research.























iii














TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS. .................. ..... . ii

ABSTRACT . . . . . . . . . . .. .. . . v

CHAPTER I INTRODUCTION AND STATEMENT OF PROBLEM ....... 1 CHAPTER II REVIEW OF THE LITERATURE. ............ 5

CHAPTER III DESIGN OF STUDY .................. 15

CHAPTER IV RESULTS ........ ..... ........... 36

CHAPTER V DISCUSSION. .................. . 53

APPENDICES ..... ............. . . . . .64
APPENDIX A Directions for Testing. ..... .. ........ 65
APPENDIX B Singer's Structural Fantasy Pre-disposition
Interview (SFPI) . . . . . . . . 66
APPENDIX C 0. J. Harvey Questionnaire . . . . . 67
APPENDIX D Home-Play Questionnaire ............. 68
APPENDIX E Interaction .................. . 70
APPENDIX F Demographic Information ........... 72
APPENDIX G Variables for Correlation Matrix........... 73
APPENDIX H Letter to Parents Requesting Consent . . . 76 APPENDIX I Consent Form . . . . . . . . . 77

REFERENCE LIST . . . . . . . . . . . .. 78

REFERENCE NOTES. .................. ...... 81

BIOGRAPHICAL SKETCH ................... ...... 82















iv













Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy


INDIVIDUAL AND FAMILIAL CORRELATES OF CHILDREN'S FANTASY PLAY

By

Lorraine Bradt Dennis

March, 1976

Chairman: Ira J. Gordon
Major Department: Foundations of Education

The purpose of this study was an attempt to discover individual and familial correlates of children's imaginative play.

Two hundred eighty-one children, representing all those for whom

informed consent had been obtained (61.62% of the total number of children in the kindergarten and first grades of the Portsmouth, Rhode Island school system) were tested with the Peabody Picture Vocabulary Test and three measures of fantasy pre-disposition: The Barron Movement Threshold Inkblots, the Singer Structured Fantasy Pre-disposition Interview (SFPI) and the Tell-a-Story Transcendence Index. Children with PPVT scores below 100 were omitted from the fantasy pre-disposition sample. A median split for each fantasy pre-disposition measure with scoring criterion of rank above or below the median on each of the three tests enabled identification of 25 high fantasy pre-disposition (HF) and 55 low fantasy pre-disposition (LF) children from the remaining 221.



v








Seventeen intact families of HF children and 21 intact families of LF children consented to participate. These were middle-class families with at least one parent in every family having at least one year of college.

Parents, individually, were administered the three fantasy predisposition measures, a test for abstract versus concrete conceptual style and a questionnaire regarding home-play environment. In addition there was a short demographic questionnaire and the mother and child were engaged in a play activity termed "Interaction." Siblings over the age of four were tested with the three fantasy pre-disposition measures.

The mother's Barron and SFPI scores were significantly correlated with the child's SFPI score. The mother's SFPI score was also significantly correlated with the child's Tell-a-Story score. Neither fathers' nor siblings' scores on the fantasy pre-disposition measures correlated with those of the children.

The number of restrictive responses of the mother with regard to her attitudes toward children's play in the home was significantly and negatively correlated with the fantasy pre-disposition of the child. Fathers' restrictive responses were not significant.

The score on the Interaction play activity between mother and child was significantly correlated with the fantasy pre-disposition of the child.

The number of years of school of the mother, but not the father, correlated significantly with the child's fantasy pre-disposition.

The child's PPVT IQ was significantly correlated with the fantasy pre-disposition, despite the fact that all IQs below 100 had been eliminated from the selection sample.


vi








The abstract versus concrete conceptual style of the parent, mother or father, did not differentiate between LF and HF children.

No significant differences between LF and HF groups were obtained in terms of sex, ordinal position, grade in school, age of mother, age of father, outside employment of mother, number of rooms in the home or number of persons in the family.

Individual and familial correlates of children's fantasy play suggested by this study are the child's verbal IQ, mother's scores on two of the fantasy pre-disposition measures, her lack of restrictiveness with regard to play in the home, the number of her years in school and her inclination to engage the child in play which contains fantasy elements. The small number of subjects in relation to the number of variables considered necessitates caution in interpreting findings.



























vii













CHAPTER I

INTRODUCTION AND STATEMENT OF PROBLEM


The systematic study of play is an astonishingly neglected research

area. It is difficult to account for such a blind spot on the part of those interested in child development, especially since there exists a common consensus regarding the prevalence of play in a child's life -- a general notion that given half a chance, the young child would probably play most of the time.

However, although it is impossible to pick up a text directed to teachers of young children without encountering a section praising the virtues of play, as a research topic it is indexed with only four references in the third edition of Carmichael's Manual of Child Psychology, five more if cross-referenced as "games." All except two of these occur in Piaget's chapter on his own theory. The 1974 Annual Review of Psychology indexes "play" not at all.

One deterrent to the study of play seems to have been the sociocultural milieu from which scientists select their problems. Sutton-Smith (1971) and Greg Stone (1971) suggest that the influence of a heavy achievement orientation tended to promote a view of play as the polar opposite of useful work and hence, inconsequential, the occupation of idle children.

A second and major deterrent has undoubtedly been the problem of

formulating an appropriate operational definition. A certain circularity of reasoning occurs in the attempt to differentiate play from not-play, since it is necessary to separate out already-identified "playful" elements.


1





2

Ellis (Note 1) points up a further paradox which hampers experimental rigor

-- the more the investigator gains control over a particular behavior, the less the behavior can be playful. Klinger (1969), attending to the problem of definition, asserts, ". . the relevant behavioral domain is simply insufficiently mapped to support the setting of rigorous positive boundaries on the concept of play"(p. 279).

The task of mapping the "relative domain" of play is, however, beginning to attract investigative attention. A major impetus has come from the proposition that play might have important consequences for later development. Reminiscent of those who, early in the century, saw play as practice for later proficiency, present-day ethologists are suggesting that play is neither antic nor idle, but rather, essential for acquiring the skills which will be necessary for survival (DeVore, 1963; Goodall, 1962). The use of sticks for the retrieval of termites, for example, does not seem to be learned by the isolated ape who has no opportunity to play with sticks, even when exposed to competent models (Mason, 1970). Kempler (Note 2) has recently extended the notion by suggesting that the playing child, creating a repertoire of divergent responses, is building a kind of "future shock" absorber and preparing for the unpredictable.

The idea that play is useful was a major theme for the second annual symposium on play, convened in Atlanta in April, 1974, for "people who share the conviction that play is a rich and natural source of creativity, learning and therapy." The title of the symposium was The Player and the Environment and the program invited participants to consider "the role an individual's environment has in encouraging or suppressing play in a child or adult."

Among the featured speakers, Dr. Jerome Singer and Dr. Dorothy Singer




3


(Note 3) addressed the conference on the importance of an adult model for fostering children's play. They were primarily concerned with the problem of encouraging nursery school teachers to interact imaginatively with the children, suggesting that without such interaction the amount of imaginative play among children decreases, and that the adult interaction seems to have an eliciting effect, providing the child license to play imaginatively.

Such adult influence on children's imaginative play should also be important in the setting of the home. Though studies of parental contributions to child behavior have been criticized for methodological reasons (Yarrow, 1963) for dearth of outcomes (Caldwell and Hersler, 1964) and for conceptual flaws (Bell, 1968, 1971), there is continuing interest in parent attitudes, practices and general home conditions as major factors in several aspects of child development (Rebelsky, 1974; Baumrind, 1971).

Of all of the developmental characteristics, that of symbolic functions seems most securely tied, empirically, to beginning influences, with both language and concept formation tending to thrive or falter in terms of early environmental variables (Gordon, 1969; Hess and Shipman, 1965). Since imaginative play is characterized by the use of symbols (indeed, is termed symbolic play by Piaget) it seems reasonable to suggest that it, too, bears the stamp of the especial confluence of family circumstances, attitudes and practices. However, little is known about the home background factors which might relate to children's imaginative play.

The present investigation was designed to study parent attitudes, practices and certain home conditions which might be correlated with children's imaginative play. Since the study of play can be said to have scarcely begun, it seems important to attempt to obtain data which relates to the child's primary environment, the site of first play experiences.





4


In addition, because the field is young and tools and instruments

have neither been extensively developed nor tested, it was determined to employ several existing measurement devices in an effort to assist in establishing a further gauge of usefulness. Cross-study comparisons are also facilitated when investigators continue with techniques and in directions previously delineated by others.

The purpose of this study, accordingly, is to investigate possible

relationships between level of children's imaginative play, measured with three instruments developed by Singer (1973) and his students, and several familial variables which borrow from suggestions of Bishop and Chace (1971) regarding requisites for a playful environment.













CHAPTER II

REVIEW OF THE LITERATURE

A search of the literature reveals that most contemporary investigations have proceeded on the basis of assumptions that play (especially imaginative play) is important for future development and that children learn (or fail to learn) certain modes of play in terms of certain kinds of environmental influences. However, it would be remiss to omit mention of the fact that the most systematic theory of play to date contradicts such contentions.

Piaget's Theory of Play

For Piaget (1951) play is a kind of inconsequential by-product of the really important organismic business of adaptation. Play occurs when schemas have become so familiar that the adaptive processes of assimilation and accommodation are not invoked; the activity is "merely for pleasure, accompanied by smiles and even laughter, and without the expectations of results characteristic of the circular reactions through which a child learns" (p. 90). Play, therefore, is not accorded any part in the construction of intelligence. As a matter of fact, play proceeds only when efforts toward adaptation are relaxed.

Nor, for Piaget, does imaginative play achieve any special significance. Rather, it is considered an invariant, the intermediate stage between practice games of infancy and the games-with-rules of later life. Furthermore, since Piaget views symbolic play as occurring precisely

5





6

because of the child's inability to think rationally and logically, he would probably be mystified as to why anyone would even try to teach it. Play, in the eyes of Piaget, is the occupation of the immature child, reflecting his as yet inefficient modes of thought and is naturally abandoned along with his childish views of reality as concrete and formal operations are accomplished.
Piaget's theory has been criticized for his apparent insistence that symbolic play is a symptom of immaturity, a kind of "mental digestive system" (Millar, 1966, p. 50) whereby reality can be assimilated wholly in terms of the child's own egocentric preoccupations of the moment. This would seem to relegate adult fantasy and divergent thinking to the same realm, implying that creative adults have simply failed to grow up.

However, no theory as detailed and cohesive has emerged as an alternative. Learning theorists have not attended to play, though it can be suggested that no new paradigm would have to be proposed to account for its appearance and persistence in the life of a child. As an instrumental response it produces satisfying changes in the environment -- baby shakes the rattle and the small child rolls the bright ball. Parental approval of activities would provide differential reinforcement. Imitation of models in the environment as well as instruction might direct the child toward acquiring the various skills and interest in sports or, contrariwise, toward pursuits requiring imagination. These notions explicitly or implicitly underly most current empirical research on play. Several different kinds of environmental "input," seen as influencing play, have been investigated.

Play is Learned -- Environmental Influences




7


Socio-economic Influences

The possibility that socio-economic factors influence the kind of play in which the child typically engages rests heavily on a study conducted by Smilansky (1969) in Israel. Concerned with the school problems of children from certain immigrant groups, her observations led to the conclusion that socio-dramatic play (defined as involving at least one other child in make-believe roles, actions and verbalizations which substitute for real objects and situations, and sustained for at least ten minutes) is virtually absent among these children. The conviction was echoed in this country in terms of our own ghetto children by Lois Murphy (1972), Sutton-Smith (1971) and Weikart, Rogers and Adcock (1970) among others. Descriptions of the play of these children include notes of the preponderance of impulsive and aimless activity, lack of sustained attention and absence of signs of pleasure.

However, the contention that disadvantaged Israeli children do not develop ability to engage in symbolic social play was disconfirmed by Eifermann (1971) who sampled the play behavior of an enormous number of Israeli children, carefully selected in terms of socio-economic class, geographic location, community structure (town, kibbutz, Arab settlement) and culture. She found that "culturally deprived" children play pretend games both at later ages and to a far greater extent than advantaged children. Furthermore, she reports, their pretend play is as elaborate and as well sustained. An explanation for the contradictory findings might lie in the setting in which play is observed. Labov (1972) suggests, with regard to observations of language deficiencies of disadvantaged children, that the usual setting for these judgments is simply too inhibiting for certain children and that only by joining them "on their





8

own turf" can their behavior be truly sampled. Play as Influenced by Instruction

Though reports of socio-economic differences in play may be open to question, the possibility of teaching disadvantaged children to play pretend games is better substantiated. Smilansky's lead (1969) in documenting the results of teaching socio-dramatic play to certain children was followed by similar efforts by Saltz and Johnson (1973) in Detroit and Freyberg (1973) in New York City. Each study reports significant gains in the amount of symbolic play subsequently initiated by the trained children, together with increases in language production, vocabulary range, verbal fluency and ability to sustain attention as well as, for Saltz and Johnson, significant increases in several retention abilities. Each of these studies maintained appropriate controls for pleasant interaction with adults so that the gain can be attributed to the imaginative play rather than to simply the attention and interest of an adult. Persistence of the gains over a long period of time has not been assessed. Specific definition of the make-believe play differs somewhat from study to study, but each definition includes the presence of a theme, roles, objects and elements which are not present in reality.

The Russians, too, seem to be concerned with the role of the adult teacher in fostering imaginative play (Elkonin, 1971), but there is no hint of worry that some children may be deficient in make-believe productions. Rather, it is believed that without adults to organize and guide this activity, the imaginative play of the child might become a kind of idle and self-indulgent dallying (Repina, 1971). Adults are to serve as models for role playing, first mediating the use of playthings, then leading the child into pretending with objects which are used as substitutes




9

for things which are missing (e.g., a spoonful of pine needles as imaginary food). Finally the adult must help the child to imitate the activities of admired grown-ups with appropriate suggestions, such as handing the doll to the youngster and saying, "The baby is sick. Will you examine her, doctor?"

The Russian psychologists see make-believe and imagination as contributing to the child's social identity when he tries out different social roles and also as the bridge to abstract thought.

Play as Influenced by Toys

The question of the kinds of playthings which might encourage play has attracted the attention of several investigators.

Gilmore (1966) found novelty to be an important variable, with his sample of children responding more readily to new rather than familiar objects in every play situation. Hutt (1971) refined the attraction of new objects in attempting to distinguish between play activity and "specific exploration." Play activity would not show a response decrement while exploration would be characterized by a response decrease. Using a specially constructed box with a lever attached to a counter, the events produced by the lever were varied systematically. Hutt found that after the

3 to 5-year old child had explored the features of the box and lever fully, lever pushing decreased. Then the box and lever were incorporated into "play" with little focus on the lever and its effects. She suggests that novelty provokes exploration while play occurs as the object becomes familiar. It is as if the question shifted from "what does this object do?" (exploration) to "what can I do with this object?" in play.

Pulaski (1970) hypothesized that unstructured toys would promote more imaginative play than toys which are very realistic. She tested





10

middle-class kindergarten, first and second graders, selecting a group of "high fantasypre-disposition" and a group of "low fantasypre-disposition" on the basis of several measures developed by Singer (1973). The results indicated that for children of five, six and seven years, the type of toy seems to have little effect. High fantasizers played with both types of toys imaginatively (according to a fantasy rating scale) while low fantasizers simply "fooled around," whether the toy was structured or unrealistic. Pulaski speculated that children's "playing style" may be fairly well formed by age five, Branch (Note 4) is testing this by conducting a similar study, using much younger children, at Yale University. A group of boys and girls who are 20 months and 26 months old are being observed at home during play with high-realistic versus low-realistic toys. Only the preliminary results are available, but these suggest an age effect, with all children engaging in more make-believe at the later age, a sex effect with girls playing more imaginary games at both age levels, and a toy by sex by age interaction with boys decreasing the amount of their pretend play with high realistic toys at the older age level while girls increase the amount of pretend play with high realistic toys at the later age level. Branch notes also that the substantial differences exist in amount of pretend play across subjects and suggests that even at this early age, children may evidence pretend play as a distinctive play mode. Family Influences on Play

Few studies have been located which attend to familial correlates of

play. Several studies can be considered pertinent if a link can be conceptualized between the imaginative productions of symbolic play and the

imagination necessary for creativity.

One study which is directly concerned with family variables and





11

children's make-believe play (Freyberg, 1973) is based on a very small sample. Another (Singer, 1973) reports on some parental practices which were found to be important as a result of interviewing children. The other studies were undertaken with a different focus -- imaginary companions for Manosevitz, Prentice and Wilson (1973), and creativity for MacKinnon (1962), Bishop and Chace (1971), and Dewing and Taft (1973). Nevertheless, findings from these studies are provocative and, taken together, seem to provide some fruitful leads for further investigation.

There is a persistent finding of the importance of birth order with first born and only children showing greater tendencies toward fantasy which has turned up in early studies of daydreaming (Singer, 1973) and appears for imaginary companions (Manosevitz et al., 1973) and children who play most imaginatively (Freyberg, 1973), but not for creative twelveyear-olds (Dewing and Taft, 1973).

Level of parental education seemed to be important as correlated

with more make-believe play (Freyberg, 1973), but only the mother's level of education was significant for creative twelve-year-old children (Dewing and Taft, 1973).

Singer (1973) and Freyberg (1973) both report the possibility that

opportunity for privacy may have some relationship to the child's tendency to play imaginatively. A possible allied finding is that of Dewing andTaft (1973) and Freyberg (1973) that more mothers of creative and imaginative children in their respective samples worked outside the home.

MacKinnon (1962), describing the childhood of creative architects, noted that:

There was often a lack of intense closeness with one or both parents. Most often this appears in relation
to the father rather than the mother, but often it





12


characterized the relationship with both parents.
Thus, if there was a certain distance in the
relationship between parent and child, it had a
liberating effect so far as the child was concerned.
If he lacked something of the emotional closeness which some children experience with their parents, he was also spared the psychological exploitation
that is so frequently seen in the life histories of
clinical patients (pp. 491-492).

Singer (1973), commenting along the same lines, suggests:

. an optimal balance of benign parental contact
and opportunity to be alone seems therefore essential to the development of a rich imaginative life. This
optimal situation occurs most readily where the child's
mother is relatively warm, willing and capable of
spending time with the child, but not so emotionally involved that she cannot at times leave the child to
his own devices (p. 62).

According to Singer (1973) another family variable consists of the kinds of activities in which the child and parent engage together. He suggests that parents who read or tell stories to their children are providing a different kind of medium than those who spend time with their children playing games of skill or chance.

In a study which provided some important conceptual leads for the present investigation, Bishop and Chace (1971) argue that parental personality qualities most likely to encourage children's playfulness at home do not exist as isolated traits. Rather, they suggest, personality theory and research points to a coherence in personality attributes. Drawing upon the long series of publications by Harvey and associates (1961, 1966, 1970), they invoke the proposition that certain personality qualities are manifested in a basic "conceptual system," with differing levels of organization which can be ordered along a dimension of concreteness-abstractness. Four principle modes of conceptual functioning are described in terms of the characteristic beliefs and response patterns.




13


The concrete end of the continuum (System 1), with its simplistic view of the world, stereotyped and polarized in terms of absolutes, its small capacity to "decenter" or take the view of another, together with an inordinate need for structure and rules, is contrasted with the abstract end (System 4) which accepts the world as complex and varied, is open to new ideas and tolerant of ambiguity, views rules as utilitarian, but has no need for reliance on authority or rule for rule's sake.

Reasoning that abstract parents would better enhance children's play

by refraining from imposing conventional standards, by willingness to accept novel (even bizarre) ideas and willingness to live with a certain amount of disorder, Bishop and Chace (1971) measured parents' conceptual systems on Harvey's concrete-abstract continuum. The category of the parent was then related to parental responses on a home-play environment questionnaire which inquired into attitudes and the kinds of conditions maintained for play, especially in terms of the restrictions imposed on play activities. Although no significant differences were found for fathers in the study, the outcome for mothers revealed that abstract mothers provided home play environments which were most free of constraints and that the children of these mothers were the most potentially creative. Potential creativity was measured by indicators of complexity and variety of performance on a laboratory play task.

The investigations of possible home influences on imaginative play are exceedingly few in number. Only slightly more numerous are studies of familial correlates of children's creativity. The latter were of interest because of several studies which have indicated a link between play, fantasy and creativity. Lieberman (1965) established significant correlations between "playfulness" and three measures of divergent thinking




I.





14

in an often-cited study of play and creativity. More recently Schaefer (1969) surveyed 800 academically talented adolescents and discovered significant differences favoring the presence of imaginary companions in childhood for those rated as "creative" in both art and literary expression. Sutton-Smith (1967), also, tested the possibility of a relationship between play and creativity, suggesting that favorite, well-used toys enable the child to increase associations and therefore obtain more novel (i.e., creative) responses, a hypothesis upheld by his study for both boys and girls. Klinger (1969), suggests that fantasy and play are essentially undifferentiated until the third year, thereafter differentiated gradually as play becomes rule-governed. He, too, indicates that the imaginative productions of fantasy may be implicated in creativity.

Taken together, the studies of home and family variables, whether

directed toward children's imaginative play or creativity, suggest possible influences which warrant further investigation. If an adult model is important for imaginative play, as the Singers suggest (Note 3), perhaps that adult would have to be comfortable with fantasy and make-believe. Adults who engaged in imaginative play themselves when young, and who enjoy fantasy in the present, may encourage more imaginative play for their children. Perhaps these are the same adults who, as Bishop and Chace (1971) propose, can refrain from too heavy-handed a management of the child's activity and insistence upon conventional forms of play, thereby providing a home environment which is conducive to fantasy and make-believe. These are the chief considerations which guided the present study.













CHAPTER III

DESIGN OF STUDY

The Problems and Hypotheses

The problem which the present study was designed to address is an

almost total lack of information regarding family background factors which might encourage or discourage children's imaginative play.

Hypotheses were derived from the very few investigations to date,

some of which were focused on creative children rather than children who engage in high levels of fantasy play. However, it seemed reasonable to assume, in the absence of data, that the imagination necessary for creativity and the imagination necessary for fantasy play might require similar facilitating environments, as Bishop and Chace (1971) have suggested.

Two groups of children, pre-selected for high or low fantasy predisposition, and their families were used to test the following hypotheses:

1. There will be a positive relationship between children's scores on the fantasy pre-disposition measures and their parents' scores on the same measures.

This hypothesis was suggested by the findings of Freyberg (1973) and Manosevitz et al. (1973) of more encouragement for imagination and fantasy among parents of children who engage in make-believe play and who have imaginary companions.

2. There will be a positive relationship between children's tendency

15





16


toward fantasy play and parents' tendency toward an abstract rather than concrete conceptual system as measured by Harvey et al. (1961) scale.

This hypothesis was suggested by the findings of Bishop and Chace (1971) with regard to parents of creative children who are characterized by these investigators as encouraging "truly playful play" (p. 321).

3. There will be a positive relationship between children's
imaginative play and a less restrictive play environment in the home.

This hypothesis, also, borrowed from the findings of Bishop and Chace (1971) concerning the type of home play environment provided for more creative children.

4. There will be a positive relationship between scores of the children on the fantasy pre-disposition measures and those of their siblings who are age four or older on the same measures.

The rationale for this hypothesis suggested that if parents who are characterized by abstract conceptual systems and who provide a home play environment which is relatively free of adult constraints do, indeed, enable more imaginative play, the environment should operate similarly for other children in the family.

5. The homes of children with high fantasy pre-disposition will
have significantly more room for privacy. Furthermore, more mothers of these children will be working mothers.

This two-part hypothesis was designed to follow up the findings of Freyberg (1973), Dewing and Taft (1973) and the suggestions of Singer (1973) regarding the need for privacy and for a certain amount of distance between parent and child, presumably obtained when the mother is occupied with employment outside the home.

6. There will be a positive relationship between children's scores





17


on measures of fantasy pre-disposition and an "Interaction" score obtained when mother and child are playing together with toys which can be used either imaginatively or realistically.

This final hypothesis was obtained from the suggestions of Lytton

(1972) and Singer (1973) regarding the importance of the kinds of activities in which parent and child engage together.


Method

Subjects

Informed parental consent was obtained to include 281 kindergarten and first grade children in Portsmouth, Rhode Island for fantasy predisposition screening. This represents 61.62% of all kindergarten and first grade children enrolled in the Portsmouth school district at the time of testing. Consent was not obtained or was refused for the remaining 38.38%.

Sixty of the children were eliminated from the sample because they scored below 100 on the Peabody Picture Vocabulary Test. This was an attempt to avoid confounding fantasy pre-disposition scores with intelligence and verbal production. The remaining 221 children constituted the sample from which the subjects of the investigation were selected.

From their 221 protocols, according to a procedure which will be described, a group of 25 high fantasy pre-disposition (HF) and 55 low fantasy pre-disposition (LF) children were identified.

Families of these children were then contacted by telephone and

invited to participate in the second phase of the study. Table 1 details the attrition of a number of potential families, some because of refusal and some because the family had moved before the contact could be made.





18

Single parent families and those with father absent because of sea-duty were omitted in order to adhere to the criterion of "intact families." Two black families were also omitted in order to maintain a homogeneous sample in terms of race. Two families turned out to be Portuguese immigrants, with some parental English-language deficiency; both were eliminated lest the child's LF score chiefly reflect the bilingual background. Finally, several families either had no phone or an unlisted number which could not be obtained, so that it was impossible to establish contact.

TABLE 1

ATTRITION RECORD OF POTENTIAL SUBJECT FAMILIES


Potential HF Potential LF
Families Families
(N=25) (N=55)

Reason for Elimination Number Eliminated Number Eliminated

Refused 3 12 Moved 3 7 Single Parent 1 5 No Phone or Unlisted Number 1 4 Minority Family 2 Portuguese Language 2 Sib of Other Child Already in Study 2 Total Eliminated 8 34 Remaining for Sample 17 21


Data, accordingly, were collected from 17 HF and 21 LF children, their parents and their siblings over the age of four.

The subjects consisted of 16 girls and 22 boys, 25 of whom were

kindergarteners and 13 in the first grade, and their families. Twenty of the children were first born, two were only children and two were





19

members of sets of twins.

The average IQ was 121.16 (SD = 10.12).

Mothers of these children had a mean age of 31.55 years (SD = 4.80) while the average age of the fathers was 33.87 years (SD = 4.90).

The educational level of the parents was high. Mothers averaged 14.32 years of schooling, a little more than two post-secondary years (SD = 1.76) while fathers had a mean of 15.97 years, only slightly less than four years beyond high school (SD = 1.94).

Two families in the sample lived in apartments. The remainder lived in single family homes. The average number of rooms per household was

7.21 (SD = 1.56).

No family in this sample consisted of other than immediate family members. Average family size was 4.34 persons (SD = .71).

Occupations of fathers are listed in Table 2 and the occupational training of mothers is reported in Table 3.





20







TABLE 2

FATHERS' OCCUPATIONS


Group Frequency Occupation


Low Fantasy 6 Engineers
3 Navy Officers
1 Retired Navy Officer
1 Lawyer
1 Insurance Executive 1 High School Teacher
3 Financial Analysts
1 Contractor
1 Pastor
1 Truck Driver (college student)
1 Barber (college student)
1 Steam Fitter (college student)
High Fantasy 6 Engineers
2 Navy Officers
1 Retired Navy Officer
1 Lawyer
1 Insurance Executive 1 High-School Teacher
1 Electronic Technician (no college work)
1 Anesthetist
1 Technical Buyer
1 Computer Programmer
1 Brick Layer (no college work)





21

TABLE 3

MOTHERS' OCCUPATIONAL TRAINING AND COLLEGE WORK


Groups Frequency Occupational Training


Low Fantasy 2 Registered Nurse
4 High School Teacher
1 B. A. Degree
1 B. S. Degree
5 At least one year college Total 13

High Fantasy 8 Registered Nurse
2 Elementary Teacher
1 Speech Therapy
1 B. S. Degree
5 At least one year college Total 17


Procedure

The 281 children of the initial sample were tested with the Peabody Picture Vocabulary Test and with three fantasy pre-disposition measures by the principal investigator and six assistants, four specially-selected, outstanding undergraduate psychology students and two graduates of the psychology program at Roger Williams College, one of whom holds a master's degree in Child development from the University of Rhode Island. All had received training in the use of the testing devices, had administered the tests to a number of children outside the study and were individually supervised by the principal investigator during the first three weeks of testing. Reliability data appear in a subsequent section.

The three measures of fantasy pre-disposition were:

I. The Barron (1955) Movement Threshold Inkblots. This is a series of 28 inkblots, arranged in a series so that it becomes progressively easier to see human movement (M), which has been found effective in use





22


with young children (Pulaski, 1970; Singer, 1973; Freyberg, 1973). The score is the number of the inkblot for which the child first responds with human movement association. It should be noted that the lower the score on this test, the earlier the M response has occurred, and hence the higher the fantasy pre-disposition.

2. The Singer Structured Fantasy Pre-disposition Interview (SFPI). A series of four questions developed by Singer (1973) which inquire into the child's favorite play activities (Appendix B). Described by Singer as a "clinical interview," the examiner is urged to use all skill in a attempting to ascertain whether there are "pretend" or fantasy elements in the child's responses. Because of the large number of examiners, however, it was determined to hold procedures as standard as possible, even at the risk of missing fantasy which might actually be present, in an effort to guard against the possibility of suggesting responses to the child. Therefore examiners were permitted to probe only with a question which requested "Tell me more about that."

Each interview question is scored with one point if fantasy or pretend is judged to be present in the response. The highest possible score for the SFPI would be 4, while the lowest possible is 0.

3. Tell-a-Story: Stimulus Situation, Card #6, Human CAT test. The stimulus situation for the Tell-a-Story, the sixth card from the Children's Appreception Test (CAT), Human version, depicts two covered, sleeping figures on the ground and a smaller person, head upraised and awake, in the foreground. The child is asked to make up a story about the picture. The story is recorded verbatim and scored according to the Transcendence Index developed by Weisskopf (1950) which awards one point for every element which is supplied by the story-teller, as opposed to those which





23


are obviously depicted in the stimulus situation. Thus, the child would be given three points for describing the larger figures as "the mother and father" . two for ascription of sex to each character and one for identification of a family relationship. Elements once scored were not credited upon repetition.

The validity of indices of fantasy pre-disposition as measures of imaginative play rests on a number of studies conducted by Singer (1973) and his colleagues. All findings point with statistical significance in the same direction. Pulaski (1970) found correlations between amount of imaginative play and ratings for fantasy pre-disposition significant at the .001 level. Freyberg (1973) also found a significant correlation (p=.01l) between fantasy pre-disposition scores and observations of make believe play for the children in her study.

Scoring of the fantasy pre-disposition protocols was accomplished by the investigator and the assistants. Each protocol was scored twice, independently. The individual who had tested the child did not score the protocol. The two independent scorings were compared and any discrepancy resolved by means of consensus between the two scorers. The check for reliability of scoring is detailed in a subsequent section.

After all the protocols had been scored, a median score was obtained for each of the three measures. In order to be assigned to the high fantasy pre-disposition group, a child had to score below the median (because of the reverse scoring) on the Barron and above the median on the other two measures, while the child who was rated low fantasy predisposition would have scored above the median on the Barron and below the median on the SFPI and Tell-a-Story. The scoring criteria are listed in Table 4.





24

TABLE 4

SCORING CRITERIA FOR FANTASY GROUP ASSIGNMENT


Fantasy Pre-disposition Tests
Group Assignment Barron SFPI Tell-a-Story Low Fantasy
Pre-disposition 19-28 0, 1 or 2 0, 1, 2, 3 or 4 High Fantasy
Pre-disposition 1-18 3 or 4 5 and above (maximum = 10)


Consent was obtained from the 17 families of HF children and the 21

families of LF children for home interviews and testing at a time of family convenience. Except in one instance, all visits were scheduled and completed when the father could be home. All home interviews and testing were accomplished by the principal investigator and one or two of the student assistants. Mothers and fathers were separated in order to obtain the various measures, with examiners taking turns interviewing a mother one time and a father the next. The principal investigator did all of the mother-child Interaction observations. The second student assistant, on occasion, tested siblings and accomplished reliability checks for the Interaction observations (see reliability section).

Because screening of the children had taken place in November,

December and January, because so many children had been tested and so many protocols scored and because interviewing of families did not begin until April, it was possible to interview families "blind." By the time the children had been assigned to the HF or LF category, they were just "names." The names were transcribed on to index cards from the category lists by a secretary, together with address and telephone number. Contacts





25


with the family were made using the card; examiners did not inform themselves of the category of the children until interviews were completed.

By means of the home visits, the following tests, Interaction activity, questionnaires and interviews were accomplished:

1. Fantasy Pre-disposition measures. Mothers, fathers and siblings of age four and over were tested with the three fantasy pre-disposition measures previously employed to screen the original children: Barron, Singer Structured Fantasy Pre-disposition Interview (SFPI), and the Tella-Story., Parents were asked to respond to the SFPI in terms of the kind of play they recalled from their own childhood. These measures were scored by two individuals associated with the study, independently, with a subsequent consensus if disagreement existed, exactly as were the children's, previously. The method of scoring had achieved a high degree of reliability (see section on reliability).

2. Abstract-Concrete conceptual system. 0. J. Harvey's "This I

Believe" test (1970) of conceptual level (Appendix C) was used to determine the conceptual system of the parent. Parents responded independently, writing responses to the open-ended statements with ten referents, six of which had previously been used by Harvey and for which detailed scoring criteria are described (Note 5).The four referents added for this study were: work, women's rights, law and education. Each referent is preceded by the statement "This I believe about ... . ." Responses were timed, two minutes allowed for each response.

Each parent was categorized into one of four systems by two judges, scoring the protocols independently and arriving at a consensus with a third judge in case of disagreement. The principal investigator and student assistants served as judges. Because the reliability check on





26


scoring, using an outside observer, was unsatisfactory (detailed in the section on reliability) and because the interest for this investigation lay in terms of the concrete versus abstract direction on the continuum conceptualized by Harvey (et al., 1961), System 1 and System 2 scores were pooled in order to assign a "concrete" category to parents with those scores, while System 3 and System 4 scores were designated as belonging to the "abstract" category. Justification for this procedure obtains both from the fact that emphasis for this study was upon the global differences between concrete and abstract orientations of parents, rather than the qualitative differences among the four systems, and that the contrast between the two categories proved to be much more reliable (again, see the section on reliability studies) than assignment of a System score.
3. Home Play Environment: Parent Permissives and Restrictives. Questions concerned with parent attitudes toward children's play were selected from the original questionnaire developed by Bishop and Chace (1971), utilizing those questions which had successfully discriminated between restrictive and permissive parents in their investigation (Appendix D). The questionnaire is designed with multiple alternative answers plus a space for write-in answer for use if the parent decides that none of the available answers are suitable. One of the alternatives serves as a criterion response indicating restrictiveness and one is a criterion response for permissiveness. The remaining answers and the write-in response represent varying degrees of restriction or permission
-- that is, each contained some contingencies or disclaimers. This enabled a count of the number of wholly restrictive or wholly permissive responses of each parent. These questionnaires contained eight questions





27


and were also completed by each parent independently.

4. Interaction: Mother-Child play activity (Appendix E). Having been previously informed that the principal investigator would like to observe them playing together, the mother and child were requested to sit together at a table or on the floor. The examiner emptied a cannister of previously selected toys, saying, "Here are lots of little toys. Both of you, playing together, build an exciting scene . like for a movie . and then I will ask you to tell me what is happening." (Adapted from Erikson, 1972)

The toys were divided into two categories, "realistic" and "fantasy," with each realistic toy having its fantasy counterpart.

Realistic Toys Fantasy Toys

1. Mother 1. Similar-sized, milk bottle shaped wooden figure.
2. Father 2. Similar-sized, milk bottle shaped, wooden figure.
3. Baby 3. Similar-sized, milk bottle shaped, wooden figure.
4. Rhinocerus 4. Green monster thing.

5. Lion 5. Red thing.

6. Dog 6. "Dumbo" elephant, sitting and smiling.
7. Toy furniture 7. Blocks of assorted shapes.
Identical number of items as
furniture.
8. Toy car 8. Thing on wheels 9. Policeman figure 9. Superman figure

In addition there was a set of colored pipe cleaners and a set of plain popsickle sticks. These were scored as "fantasy" items.





28


A check-list of the toys with separate space for recording mother or child use enabled the observer to record the particular toy and the user. Mother and child were scored separately for the number of toys of either category which each contributed to the scene. Only toys which were actually used in the scene were counted, yielding a "realistic toy" and a "fantasy toy" score for each.

The description of "what is happening" in every instance reported by the child, was recorded verbatim and then read back to the pair for additions or corrections. The description was scored according to the criteria listed below by three judges, independently, with agreement of two out of three or all three constituting the reliability criterion, to yield an Interaction score of 1 to 5 for the mother and child playing together. The higher score reflects more make-believe or fantastic elements.


Score Criteria (Adapted from Pulaski, 1970)


(1) Familiar everyday experience Concrete
Ordinary
Unelaborated
Use of realistic toy in realistic, unelaborated role or function -- for example, the car is used as a car or the policeman as a policeman.
(2) Indirect experience . TV characters, movie characters, story book characters and/or events. Relatively unelaborated narrative.
(3) Content and characters farther removed from experience or more elaborated. Might include "silly" or aggressive or emotional fantasy.
(4) More imaginative, unusual or original than #3 above.
Fantasy elements more detailed.





29


Score Criteria (continued)

(5) Fantastic
Farthest removed from present time, space or reality. Most original, richest imagination. Elaborate details.


5. Demographic data (Interview) (Appendix F). The interview, conducted with either mother or father, ascertained the age of each parent, the number of years of schooling for each, the occupation of the father, the occupation of the mother, the number of persons living in the household and the number of rooms in the house.

Reliability studies

Because of the large group of individuals engaged in both testing and scoring, because certain aspects of scoring involved judgments and because the reliability of certain of the fantasy pre-disposition measures does not seem to have been well-documented, several different reliability studies were conducted during the course of the investigation.

1. Reliability of the testing (fantasy pre-disposition screening) of the children. Testing of the children took place over a period of three months with Christmas vacation intervening. Two reliability studies were accomplished, the first early in the testing period and another during the last three weeks of testing in an effort to guard against "slippage" of examiner technique. Two individuals, both advanced psychology students unconnected with the main investigation, were involved, one completing the early study and the second performing the study at the end of the testing period. The procedure in both instances was identical. The observer sat in on the testing two times for each examiner, completing a protocol each time which was later scored according to the standard




30


scoring procedure. The scores obtained on the reliability-observer's protocol were compared to the scores obtained on the examiner's protocol. Percent agreement was established by the following formula:

100 No. of disagreements % agreement No. of agreements

The first observer, in November, 1974, obtained nine protocols from two observations of four student-examiners and one observation of the principal investigator. The protocols yielded a 100% agreement with those of the examiners.
The second observer conducted his observations in January, 1975. He observed five examiners twice for a total of ten protocols. His protocols obtained an agreement rate of 93.3%.

2. Reliability of scoring for fantasy pre-disposition measures. Although the Barron Movement Threshold Inkblots yield a score which is essentially unequivocal, both of the other measures (SFPI and Tell-a-Story) entail scorer judgments.

The over-all purpose of the three measures was to enable differentiation of a high and a low fantasy pre-disposition group. Accordingly, the reliability of scoring of these measures was tested by determining whether an independent observer, familiar with the use of the measures, could select HF and LF groups by simply inspecting all of the protocols. The degree of correspondence with the groups which had been discriminated by the actual scoring procedure could then be ascertained.

Dr. Jerome Singer cooperated in obtaining the services of one of

his graduate students for this endeavor. The graduate student, informed only of the number which had been selected by the scoring procedures employed for the investigation (25 HF and 55 LF), inspected the total of 221 protocols. She sorted out 24 which she identified as HF and 60 which





31


seemed to her to be LF protocols. Twenty-two of her HF group corresponded to those selected by the investigation procedures, three had been "missed" and two of hers did not correspond to ours. She had selected 60 LF protocols; these included all 55 of those previously selected in the investigation and five which did not correspond (Table 5).

This means that there were ten instances of non-correspondence out of the total of 221 protocols, an agreement rate of 95.5%.

TABLE 5

RELIABILITY OF SCORING FOR FANTASY PRE-DISPOSITION MEASURES


Number of Number of Number
HF LF Remaining Total


Identified by
Scoring Procedures 25 55 141 221 Identified correctly
for Reliability 22 55 134 211 Mis-identified for
Reliability 3 7 10 Agreement Rate = 95.47%


3. Reliability of scoring for 0. J. Harvey "This I Believe" protocols: concrete-abstract conceptual systems. The final sample of 38 sets of parents yielded 76 "This I Believe about . ." protocols which were scored according to the Harvey (Note 5)coring criteria as previously described. An independent judge, a Ph.D. candidate in child development at Purdue University, familiar with the Harvey scoring, scored 74 of the protocols, assigning each into one of 4 systems. Agreement with the investigation scoring was disappointing; 40 protocols corresponded, but 34 did not, a 54.05% agreement.





32

Therefore, as previously noted, System 1 and System 2 scores were

pooled to form a "concrete" category, while System 3 and System 4 scores, grouped together, formed the "abstract" category.

A second inspection of the judge's scores revealed that the broader categorization resulted in much better agreement with 61 of the 74 protocols corresponding to the newly assigned category of "abstract" or "concrete." This, therefore, yields 82.43% agreement.
4. Reliability of the fantasy pre-disposition measures. During the process of visiting the families, it was possible to re-test 16 of the children with the fantasy pre-disposition measures. Testing of all the children had been completed in January; family visits began in late April so at least three months had elapsed between the original testing and the re-test.

Because examiners went into the household "blind," selection of children for re-testing was circumstantial. If it proved possible to separate the child from the parents' testing, so that parents' responses would not influence the child's responses, a re-test might be accomplished. Sometimes children were playing outside while parents were interviewed and could be called in later. In most instances, however, the child wanted to listen to the parent's interview and had to be considered contaminated for a possible re-test. Subsequent inspection of the re-tests revealed that six of the children had previously been categorized as HF and 10 had been assigned to the LF group.

Scores on each of the three measures were correlated with the

original scores obtained by the child on that measure to obtain a correlation coefficient for each of the fantasy pre-disposition tests. Results are tabulated in Table 6.





33

TABLE 6

TEST RE-TEST CORRELATION COEFFICIENTS: FANTASY PRE-DISPOSITION MEASURES


Test r

Barron .274
SFPI .750** Tell-a-Story .616*

*p<.Ol d.f. = N-2 = 14
**p<.OOl

Thus it is evident that the reliability of the Barron is quite

doubtful, while that of the SFPI and Tell-a-Story proved to be much more substantial.

5. Reliability of the Interaction observations. The Interaction

between mother and child was recorded on two kinds of protocols (Appendix E). One check-list was used to record the toy used in the scene and by whom it was used. The second obtained a verbatim account of "what is happening" in the scene. In every instance observations were conducted by the principal investigator. On five occasions a second observer maintained a separate check-list of toy utilization and recorded the description of the scene. The toy lists and Interaction scene descriptions were counted and scored in the prescribed manner and compared with those of the principal investigator.

The check-list of toys and who used them proved less reliable than the account of "what is happening" which furnished the Interaction score. The criterion for checking a toy as used by mother or child was that the toy should be placed into and used in the scene. However, on a number of occasions a toy first used by one of the players might be picked up and used further or in another fashion by the other player. Discrepancies





34


in recording subsequent use of toys resulted in a low reliability for the toy check-lists. None of the observer's corresponded exactly with those of the investigator. The disagreement in actual toy-count was 38 out of 60, yielding an agreement of only 36.67%.

However, the account of "what is happening" provided a reliability of agreement on scores of 100% (i.e., the five observer verbatim records obtained the same score as those of the investigator). Statistical Analysis of Data
Nominal data which included the sex of the child, ordinal position, whether or not the mother was employed outside the home, the concrete or abstract category of the mother and of the father and the encouragement or discouragement of make-believe play by each parent were analyzed by the use of Chi square.

Fantasy pre-disposition scores for children of LF and HF groups, as well as the scores of their parents, the Interaction scores for LF and HF groups and the permissives and restrictives of each parent were subjected to t tests for significant differences.

A correlation matrix was obtained which utilized all of the scores on fantasy pre-disposition measures, IQ scores, age of parents, their years of schooling and similar variables in order to affirm or disaffirm predicted relationships between variables.

Finally, on the basis of the analysis of the correlation matrix, two multiple regression analyses were performed in order to assess the relative contribution of selected predictor variables to a predicted variable, the child's SFPI score.

This study must be considered a pilot investigation into a field
which is in its infancy. Subjects for the study are a restricted sample,





35


limited in terms of numbers, region of the country, socio-economic status and race. No claims are made for representativeness; therefore, extensive inferential statistical procedures were not employed.














CHAPTER IV

RESULTS


Differences Between LF and HF Groups


The tests for significance of differences between LF and HF groups are reported in Table 7.

Neither sex differences, nor that of ordinal position nor the grade level of the child discriminated significantly between the LF and HF groups.

The difference in the Peabody Picture Vocabulary IQ was found to be highly significant, despite an effort to control for IQ effects. All children with an IQ of less than 100 had been eliminated from the selection process. Nevertheless, the LF group had a mean vocabulary-IQ of 117.27 (SD = 8.87) while that of the HF group was 125.94 (SD = 9.73). The t test with 36 df is significant at p = .0005.

The age of the parents did not differentiate between LF and HF groups. The number of persons in the household, the number of rooms in the home, the number of mothers employed outside the home were also not relevant.

Although fathers' years of schooling were not significantly different from group to group, the mothers' years of schooling are significantly higher for HF children. Mothers of HF children average 14.88 years of



36





37


TABLE 7

DESCRIPTION OF L.F. AND H.F. GROUPS: TESTS OF DIFFERENCES


Variable Low Fantasy High Fantasy Test of Significance

Sex 7 girls 9 girls Corrected Chi square 14 boys 8 boys (1) = .782 NS Ordinal Position
First Born 10 10 Corrected Chi square
"only" 0 2 (1) = 1.20 NS
twin 1 1 Grade
Kindergarten 14 11 Corrected Chi square
First Grade 7 6 (1) = .005 NS Peabody Picture
Vocabulary IQ
Range 104-140 106-145 t Test (36) = -3.86**
Mean 117.27 125.94
S.D. 8.87 9.73 Mother's Age
Range 25-40 years 27-45 years t Test (36) = -.620 NS
Mean 31.19 32
S.D. 4.05 5.71 Father's Age
Range 27-43 years 26-46 years t Test (36) = -1.53 NS
Mean 33 34.94
S.D. 4.27 5.54 Mother's Years
of School
Range 10-17 years 13-17 years t Test (36) = -3.48*
Mean 13.86 14.88
S.D. 2.01 1.22 Father's Years
of School
Range 13-19 years 12-21 years t Test (36) = -.909 NS
Mean 15.76 16.24 S.D. 1.64 2.28 Mother's outside
employment
Not employed
outside home 13 11 Corrected Chi square
Part-time 6 6 (1) = .0026 NS
Full-time 2 0 Number of persons
in family
Range 4-6 persons 3-6 persons t Test (36) = 1.79 NS Mean 4.48 4.18 S.D. .68 .73





38


TABLE 7 Continued


Variable Low Fantasy High Fantasy Test of Significance

Number of rooms in home
Range 4-11 rooms 4-10 rooms t Test (36) = -.44 NS Mean 7.14 7.29 S.D. 1.65 1.49

*p<.001
**p<.0005





39


school (SD = 1.22) as contrasted with the mean of 13.86 years (SD = 2.01) for mothers of LF children. The difference is significant at p = .001.

Findings for the Hypotheses

Tables 8, 9, and 10 report results of analysis of scores pertinent to the hypotheses. The correlation matrix for all scores can be found in Appendix G.

1. The first hypothesis stated the prediction of a positive relationship between children's scores on the measures of fantasy play and their parent's scores on the same measures.

The hypothesis was only partially supported and only for certain of the mothers' scores.

It can be seen (Table 8) that the mothers' Barron scores and the SFPI are significantly different from LF to HF group. Mothers' Tell-a-Story scores, however, do not discriminate.

Fathers' scores on the three fantasy pre-disposition measures do not differentiate between LF and HF groups.

Correlation coefficients between parents' and children's scores are presented in Table 9 as abstracted from the complete correlation matrix (Appendix G).

There are no significant correlations between fathers' fantasy predisposition scores and those of the children.

Significant correlations occur between the child's SFPI and two of the mothers' tests, the Barron and the SFPI. Mothers' SFPI is also significantly correlated with children's Tell-a-Story.

2. The second hypothesis stated the prediction that parents of HF children would tend to be rated as "abstract" with regard to conceptual





40








TABLE 8

MEAN SCORES FOR FANTASY PRE-DISPOSITION MEASURES


Test Child Score Mother Score Father Score
L.F. H.F. L.F. H.F. L.F. H.F.

Barron 24.571 12.706 17.952 14.529 14.524 15.824 S.D. 3.22 5.57 5.71 6.28 5.98 4.99 t Test 9.375** 2.367* N.S.

SFPI 1.429 3.353 1.762 2.412 1.667 1.412 S.D. .746 .493 1.04 1.23 1.35 1.12 t Test -16.327** -2.303* N.S.

Tell-a-Story 1.430 7.765 4.290 5.294 4.714 5.00 S.D. 1.33 2.02 2.47 2.47 1.74 2.60 t Test -13.780** N.S. N.S.

*p<.02
**p<. 0001







TABLE 9

CORRELATIONS BETWEEN CHILD'S SCORES ON FANTASY PRE-DISPOSITION MEASURES AND PARENTS' SCORES ON FANTASY PRE-DISPOSITION, RESTRICTIVES AND PERMISSIVES


Mother's Mother's Mother's Father's Father's Father's Mother's Mother's Father's Father's Barron Tell-a- SFPI Barron Tell-a- SFPI Permis- Restric- Permis- RestricStory Story sives tives sives tives

Child's
Barron .133 -.147 -.304 .005 .093 .210 -.020 .241 -.019 -.071
Tell-a-Story
-.224 .214 .327* .057 .132 .197 .048 .217 -.029 .052

SFPI -.389* .139 .369* .104 .160 -.076 .111 -.420* .224 -.080

*p<.05 r = .320 two tail
df = 36





42







TABLE 10

PARENT'S CONCRETE OR ABSTRACT CATEGORY


Category Mother frequencya Father frequencyb L.F. H.F. L.F. H.F.

Concrete 8 4 9 9 Abstract 13 13 12 8
acorrected Chi square (1) = .372 N.S.
bcorrected Chi square (1) = .085 N.S.





43

system on 0. J. Harvey et al. (1961) scale.
The hypothesis was not supported, either for mothers or for fathers. Table 10 shows that the abstract versus concrete ratings of parents did not distinguish between LF and HF groups of children. Mothers' ratings obtained a Chi square of .372, while Chi square for fathers' ratings was .085, both well below the level of significance.

It can be noted that the total number of parents with abstract conceptual systems is greater for both groups than the total number of parents with concrete conceptual systems.

3. The third hypothesis stated the prediction of a positive relationship between the child's HF pre-disposition and a non-restrictive play environment in the home.

This hypothesis received partial support.

Mother and father, each completing a modified version of the BishopChace (1971) questionnaire on home-play environment, received scores indicating the number of unequivocal permissive responses and the number of wholly restrictive responses. Other possible responses to the questions, all of which contained some qualifiers, were not counted for purposes of this scoring. There are eight questions on the questionnaire (Appendix D) so that each parent might have received a maximum of eight completely permissive or eight extremely restrictive scores. The scoring results are tabulated in Table 11.

It can be noted that the number of permissive responses given by

both mothers and fathers is substantially the same for each group. The differences between groups were not significant. Inspection of the correlation matrix (Appendix G) reveals that the permissive scores of both parents obtain few significant correlations; mothers' permissives





44


TABLE 11

MEAN NUMBER OF PERMISSIVE AND RESTRICTIVE RESPONSES


Response Mother Father L.F. H.F. L.F. H.F.

Permissive 2.571 3.058 1.857 2.058 S.D. 1.78 1.71 1.62 1.14 t Test (36) N.S. N.S. Restrictive 1.048 .352 .857 .823 S.D. 1.07 .606 .964 .951 t Test (36) 4.694* N.S.

*p<.0001

are correlated negatively and significantly with mothers' and fathers' restrictives. Fathers' permissives are correlated significantly only with fathers' Tell-a-Story score.

Although the number of restrictive responses from fathers does not differentiate between LF and HF groups, the mothers of LF children give significantly more restrictive responses than mothers of HF children.

Support for the hypothesis, therefore, derives from the indication that although LF and HF children seem to come from homes which are similar in terms of permissive responses of both fathers and mothers, the mothers of these children differ significantly with regard to the number of very restrictive attitudes toward play which they express. Mothers of HF children appear to be less restrictive in attitudes about home-play than mothers of LF children.

The correlation matrix (Appendix G) shows that the mothers' restrictives are significantly related to the child's SFPI and also the several other variables. The significant correlations are abstracted for Table 12. Fathers' restrictives obtain only two significant





45

correlations, a positive relationship with the mothers' restrictives and negative correlation with mothers' permissives.

TABLE 12

VARIABLES SIGNIFICANTLY CORRELATED WITH MOTHERS' RESTRICTIVES


Mothers' Restrictives


Child's SFPI -0.420 Mother's Barron 0.391 (scored in reverse direction) Mother's SFPI -0.379 Mother's Tell-a-Story -0.373 Mother's years of school -0.564 Father's years of school -0.371 Number of Rooms in home -0.417 Father's restrictives 0.404 Room/Person ratio -0.346


The relationship between mothers' restrictives and the child's

fantasy pre-disposition was further explored by means of multiple regression analysis, described in a subsequent section.

4. The hypothesis that the children in this study would resemble

their brothers and sisters on the measures of fantasy play was not upheld.

Thirty-two siblings, with an average age of 7.70 years (SD = 2.85), were administered the fantasy pre-disposition measures and the scores correlated with those of their siblings in the sample. The correlation matrix is tabulated in Table 13.

Not one of the child's fantasy pre-disposition measures correlated with any of those of siblings.

The correlations among the child's own fantasy pre-disposition scores would be expected because of the selection process for the high and low fantasy groups. Only children with tested siblings were entered for this matrix.





46

TABLE 13

CORRELATION MATRIX OF CHILD'S AND SIBLINGS' SCORES ON FANTASY PRE-DISPOSITION MEASURES


Col Col Col Col Col Col
1 2 3 4 5 6
Child's Sib's Child's Sib's Child's Sib's
Barron Barron SFPI SFPI Tell-a-Story Tell-a-Story

1 1.000 .327 -.608* -.237 -.833* -.113 2 .327 1.000 -.240 -.218 -.180 -.164 3 -.608* -.240 1.000 .076 .768* .147
4 -.237 -.218 .076 1.000 .019 .365* 5 -.833* .180 .768* .019 1.000 .038 6 -.113 -.164 .147 .365* .038 1.000

*p<.05 r = .349 two tail
df = 30

No such selection process obtained for the siblings, however. While the siblings do not seem to systematically resemble their sib in the investigation, their own scores on the SFPI and their scores on the Tella-Story are significantly correlated, suggesting that the score on one will tend to predict the score on the other.

5. The fifth hypothesis stated the suggestion that some degree of

privacy would be associated positively with higher fantasy pre-disposition.

Specifically, it was predicted that HF children would have 1) more room for privacy as measured by the Room/Person ratio and 2) more mothers employed outside the home and thereby maintaining some degree of distance from the child.

The prediction was not confirmed. Neither condition distinguished between the groups.

The Room/Person ratio did not correlate with any of the children's scores. Its one significant correlation with any measure was, as





47

recorded on Table 12, a negative association with the number of mothers' restrictive responses (r = -.346).

Chi square analysis of the number of mothers working either part time or full time (grouped together for the analysis since the number of full time working mothers in this sample is so small) revealed no basis for assuming a difference for LF or HF groups (Table 7).

6. The sixth hypothesis stated the proposition that mothers and children who play together imaginatively enable a prediction of high fantasy pre-disposition for the children. Operationally, it anticipated a positive relationship between the Interaction score on the mother-child play activity and the level of the child's fantasy pre-disposition.

The hypothesis was substantially supported.

Table 14 records the mean Interaction scores of LF and HF groups

together with the t test for significance of the difference between means. It can be seen that the Interaction score distinguished between LF and HF groups with a t test which is significant at the .0003 level.

TABLE 14

MEAN SCORES FOR INTERACTION (MOTHER-CHILD ACTIVITY)


Low Fantasy High Fantasy Mean Score 2.095 3.118 S.D. 1.04 1.11 t test (36) -3.986*

*p<.0003

The Interaction score obtained significant correlation coefficients with all three of the child's fantasy pre-disposition measures as reported in Table 15 (abstracted from the complete correlation matrix).





48


TABLE 15

SIGNIFICANT CORRELATIONS WITH INTERACTION SCORE


Interaction


Child's Barron -.380 Child's SFPI .414 Child's Tell-a-Story .520 Father Encourage .374 (obtained from multiple regression correlation matrix)


To summarize the findings in regard to the hypotheses for this study: The prediction that children and parents would resemble each other with regard to fantasy pre-disposition scores was only partially substantiated for only two of the mothers' scores.

The second prediction which suggested that parents of HF children would tend to be categorized as having an abstract conceptual system was not upheld.

The third prediction of a less restrictive home-play environment for HF children was supported in terms of mothers' but not fathers' responses.

The fourth hypothesis stated the suggestion that brothers and sisters of the children in the sample would resemble their LF or HF sib. This hypothesis did not receive any support at all.

The fifth hypothesis stated the prediction of more privacy in terms of a higher room/person ratio and more mothers employed outside the home for HF children. Neither condition obtained and the hypothesis was not confirmed.

The sixth prediction of more imaginative play-interaction between mother and child for HF children was substantially supported.





49

Multiple Regression Analysis

The child measure which achieved the largest number of significant correlations with other variables is the SFPI (Table 16). It is also the most stable measure with a high degree of reliability on re-test (r = .750, p = .001). It was selected as the dependent variable for multiple regression analysis.

TABLE 16

SIGNIFICANT CORRELATIONS WITH CHILD'S BARRON, SFPI AND TELL-A-STORY
(ABSTRACTED FROM COMPLETE CORRELATION MATRIX, APPENDIX G)


Child's Child's Child's Barron SFPI Tell-a-Story

Barron (child) 1.000 -.6167 -.8346 SFPI (child) -.6167 1.000 .7464 Tell-a-Story (child) -.8346 .7464 1.000 Vocabulary-IQ (child) -.4888 .4930 .5474 Interaction (mother-child) -.3798 .4135 .5474 Mother's years of school -.3538 .3695 Mother's SFPI .3695 .3270 Mother's Barron -.3898 Mother's Restrictives -.4195


Variables relating significantly to the child's SFPI are: the child's IQ, the Interaction score, the mother's years of school, the mother's SFPI, and the mother's number of restrictive responses for home-play environment. The father's views on whether make-believe play should be encouraged distinguished significantly between LF and HF groups (Table 17); this variable, termed "father encourage" was also entered into the analysis.

A stepwise multiple regression analysis employing these variables as predictors was performed. The summary table is reproduced for Table 18.





50

TABLE 17

PARENT'S RESPONSES REGARDING ENCOURAGEMENT OF MAKE BELIEVE


Response Mother frequencya Father frequencyb L.F. H.F. L.F. H.F.


Yes 10 8 2 7 No 11 9 19 10

*corrected Chi square (1) = .085 N.S.
bcorrected Chi square (1) = 3.60 Fisher Exact Probability = .056


TABLE 18

SUMMARY TABLE MULTIPLE REGRESSION ANALYSIS I


Step Variable Entered Multiple Rsq Increase F Value no. R Rsq

1 IQ .4830 .2431 .2431 11.5619 2 Interaction .5988 .3586 .1155 6.3043 3 M's Yrs. School .6664 .4441 .0855 5.2318 4 F. encourage .6852 .4695 .0254 1.5788 5 M's restrictives .6952 .4833 .0138 .8518 6 M's SFPI .6995 .4892 .0060 .3618


The analysis indicates that for this equation significant contributors to the variance of the child's SFPI score are IQ (24.31%), the Interaction score (11.55%) and the mother's years of school (8.55%). These three variables, taken together, account for a multiple R of .6664 or 44.41% of the variance of the child's SFPI.

Three variables, father encourage, mother's restrictives and mother's SFPI fail to emerge as significant predictors. Each is significantly correlated with a variable which entered the equation earlier.

A second multiple regression was performed with the five most likely





51

variables of the first equation, three of which had achieved a significant F ratio (IQ, Interaction and mother's years in school) and two which might possibly emerge if the order of entrance were changed. In this instance it was determined that IQ should enter last.

The summary table is reproduced for Table 19.


TABLE 19

SUMMARY TABLE
MULTIPLE REGRESSION ANALYSIS II


Step Variable Entered Multiple Rsq Increase F Value no. R Rsq


1 M's Restrictives .4195 .1760 .1760 7.6903 2 Interaction .5596 .3131 .1371 6.9878 3 M's years school .5941 .3530 .0398 2.0093 4 M's SFPI .6234 .3887 .0357 1.9258 5 IQ .6867 .4716 .0829 5.0202


The five variables account for 47.16% (R = .6867) of the variance of the child's SFPI. Mothers' restrictives and the Interaction score become significant predictors with restrictives accounting for 17.60% and Interaction accounting for 13.71% of the variance of the dependent variable.

IQ, also, obtains significance, accounting for 8.29% of the variance in this equation.

On this occasion, mothers' years of school does not reach significance. It is significantly correlated with restrictives (r = -.564) so that employing one as a predictor seems to relegate the influence of the other to the background.

Mothers' SFPI does not appear to be an important contributor for either equation.





52


In the light of the two equations, it is possible to suggest the influence of four variables on the amount of child's make-believe play as measured by the SFPI. Two of these, the child's IQ and the motherchild Interaction, were significant predictors for both equations. Each of the two remaining, mothers' restrictives and mothers' years of schooling, emerged as significant for one equation. It would appear that use of one as a predictor might preclude use of the other; the substantial intercorrelation suggests shared variance which would necessitate testing beyond the scope of this study.

Summary of Findings

To summarize the individual and familial correlates of children's make-believe play which have been suggested by this investigation: the child's vocabulary-IQ seems to be the most important individual characteristic associated with level of fantasy pre-disposition for children in this study.

A relationship was found for two measures of mothers' fantasy predisposition, the Barron and SFPI, and children's SFPI. The mothers' years of schooling is correlated with children's scores on the Barron and SFPI. The number of restrictive responses made by the mother with regard to home-play environment is negatively associated with the amount of make-believe of the child's SFPI.

Finally, the amount of fantasy in the Interaction score, for which

the mother and child engaged in play together, is significantly correlated with all three of the children's fantasy pre-disposition measures.













CHAPTER V

DISCUSSION


This study was undertaken in an effort to discover possible relationships between individual and family-background factors and children's make-believe play. Because there have been so few investigations of this area, it must be considered exploratory in design and concept. Because the number of subjects is so small in relation to the number of variables, the data must be interpreted with extreme caution.

The sample is limited not only in size but also in composition; it is white, middle class, small family and lives on a large island off the coast of Rhode Island. It contains an unusual number of Navy families because of the large defense installations at Portsmouth and Newport. Inferences from this sample to a general population can be taken only as suggestions for further investigation.

However, the homogeneity of the sample has an advantage of controlling for socio-economic class differences which could otherwise confound findings. These children come from homes with ample space for play and with easy access to beaches, woods and meadows for a wide variety of summer and winter play activities.

Notable was the deep interest of both LF and HF parents in their children and in the study. On the basis of pilot trials approximately one hour had been allotted for the family interviews. It quickly became apparent that the time was greatly underestimated. Parents wanted to


53





54

talk after the interview sessions were completed, comparing notes about their responses, plying interviewers with questions and engaging in discussion about childrearing, often until after midnight.

The children, whether LF or HF, appeared to come from similar homes in terms of parental concern and caring, general socio-economic circumstances and physical environment.

Yet, by the measures used for this investigation, the children

themselves seem to be very different. They represent the two extremes of a larger group of 221 children with regard to scores for the three fantasy pre-disposition tests. A question must be raised with regard to the validity of the measures. Do these children really differ in the amount of make-believe play?

The findings of other investigators concerning the relationship of the measures to make-believe play, observed under several different circumstances, have been cited. The SFPI, the dependent variable for multiple regression analysis in this study, is scored specifically in terms of the pretend play which can be discerned in the responses and seems to yield reliable scores on retest after a passing of several months. Qualitative examination of the protocols was not within the scope of this study, but non-systematic review indicates that the LF children tended to respond to the question "What do you like to play best?" with physical activity such as "play baseball" or "ride my bike," while HF children typically spoke of playing "house" or "school" or "cowboys." However, the .problem of validity is troublesome. Only further research can provide definitive answers to the question. This study was designed to begin with already-existing measures as a step in assaying their utility.




55


The expected correlations between parents' scores on these measures and the scores of their children did not wholly materialize. It had been anticipated that parents who reported a good deal of make-believe play in their own childhood and whose Barron and Tell-a-Story contained substantial fantasy elements would tend to encourage children's fantasy pre-disposition. Although this was suggested for two of the mothers' scores (SFPI and Barron), fathers' scores did not seem related to those of their children at all. It may be that a number of the fathers in this sample are too busy for extensive contact with their children. Ten of the fathers are college students in addition to being heads of households with full-time jobs; half of these are in graduate work. Five are in the Navy with regular sea duty built into their lives.

However, it does appear that an attitude of unqualified approval of children's make-believe distinguishes between fathers of LF and HF children. The variable did not appear to contribute significantly to the child's SFPI on multiple regression analysis. However, father encourage is significantly related (r = .374) to Interaction, a significant predictor in both analyses. It is interesting to conjecture that perhaps fathers' attitudes exert an indirect influence, perhaps in maintaining a supportive rather than a negative atmosphere for fantasy, which better enables the mother and child to engage in make-believe freely when playing together.

A similar speculation might be offered with regard to the mothers' and fathers' restrictives. The Bishop and Chace (1971) study had indicated the importance of lack of restrictiveness in home-play environment provided by the mother. This finding was corroborated by the present investigation. Bishop and Chace were not able to establish any





56


significant differences for fathers' responses to the home-play questionnaire; however, the restrictiveness of fathers in our sample was significantly related to that of their wives (r = .404). Could it be that fathers' restrictive attitudes provide the background for mothers' practices in constraining certain play activities? Clearly, fathers' attitudes, perhaps as they may influence mothers' attitudes and practices, warrant more investigation.

The mothers' restrictiveness was significantly correlated with the child's SFPI, with each of her own fantasy pre-disposition measures, with her years of schooling, the number of rooms in the house and the room/person ratio. The relations are negative (except for the Barron which, it should be recalled, is scored in a reverse direction). It suggests that the higher the fantasy pre-disposition scores, the fewer the restrictives. Also, the more years of schooling, the fewer the restrictives. And the larger the home or the fewer persons per room, the fewer the restrictives. The correlates with restrictiveness seem to make a kind of "common sense" pattern. It is particularly interesting because all of the mothers gave more wholly permissive responses than wholly restrictive responses (Table 11). The entire questionnaire as used in the present investigation (a modified version of the original by Bishop and Chace, 1971) consists of only eight questions. The most restrictive mothers gave an average of 1.04 restrictive responses, so the proportion of restrictive responses to the permissives and to the remainder of the questions does not yield a picture of extreme repression of childhood fun and games. Further use of the home-play environment questionnaire ought to include an item analysis of responses to see if a pattern emerges

-- i.e., is it one kind of play which is being ruled out, or are the





57


restrictives scattered?

In contrast to the Bishop and Chace (1971) investigation, the present study was unable to obtain evidence of differences in abstract versus concrete conceptual systems for LF and HF parents. This might be explained by the relatively high level of education for parents in the sample, with more parents categorized as abstract (46) than were rated as concrete (30). It must also be suggested that in spite of earnest efforts to adhere to the scoring criteria, the scoring of the This I Believe protocols may have been faulty. Such a possibility would also account for the low reliability found for the score procedure.

The level of parental education has been cited as important for

children's make-believe play by Freyberg (1973), while mothers' education but not fathers' was significant for Dewing and Taft's (1973) study of creative twelve-year-olds. This investigation confirmed the importance of mothers' education, which correlated significantly with the child's Barron and SFPI. Her level of education is also found to relate to a less restrictive home-play environment. A question for further research might inquire as to what aspects of higher education promote a less restrictive orientation of the mother toward home play.

Additionally, fathers' education is found, in this sample, to relate to the number of rooms in the household which, in turn, seems to promote a lessened number of constraints on play imposed by the mother. This could, perhaps, be another indication of fathers' indirect contribution, with his level of education enabling the material advantage of a larger house and thereby securing a more relaxed atmosphere for mothers' attitudes toward play. This kind of linking of correlations is, admittedly, highly speculative. Nevertheless, it could provide indications of





58

direction for further research.

In the same vein: Singer (1973) and Freyberg (1973) both report the importance of opportunity for privacy, operationalized as the room/ person ratio, in the child's tendency to play imaginatively. For this study the number of rooms in the home, the number of persons in the household, and the room/person ratio all failed to obtain significant correlations with the children's fantasy pre-disposition measures, or to distinguish between LF and HF groups. However, with the number of rooms in the house (and also, the room/person ratio) significantly related to the mothers' restrictives, one may conjucture that perhaps the amount of space available is more important for the mother's sense of what kinds of play must be constrained with strict rules, than for obtaining privacy. Again, a question for further research.

Also, because of the income level, the number of rooms in these houses was high in proportion to the number of people. This might not obtain for a city apartment-house population. It may be that "room for privacy" becomes important as a factor for imaginative play under circumstances which are typically over-crowded, but does not emerge as important when the amount of space is generally ample.

A final comment -- a caveat: the design for this study implied a notion of unidirectional effect, from parent to child. However, the current literature in child development, beginning with Bell (1968) emphasizes the transactional nature of parent-child interaction. It must be acknowledged that the tendency of the child to play imaginatively might well encourage enjoyment of fantasy by the mother. Perhaps makebelieve play does not necessitate the kinds of restrictive arrangements that high motoric play might demand. Such considerations could help to





59


account for the failure to obtain resemblance between fantasy pre-disposition scores of children and their siblings. One might even suggest, with Smilansky (1969) and others, that make-believe play promotes verbal intelligence.

These are certainly considerations which must guide further research.

Instruments

This study employed several instruments which have been used in

previous investigations: the Barron Inkblot Movement Threshold and the Singer Structured Fantasy Pre-disposition Interview (SFPI) (Singer, 1973; Freyberg, 1973; Pulaski, 1973), and a modified version of Tell-a-Story (Pulaski, 1970) in order to assess the child's, parent's and siblings' fantasy pre-disposition. In addition, the 0. J. Harvey (1970) test of conceptual style and a modified version of the Home-Play Environment attitude questionnaire (Bishop and Chace, 1971) were employed to measure parent variables. It is appropriate to report on our experience with these instruments.

The test-retest reliability of the Barron is so low that serious

doubt must be raised regarding its utility without further effort toward standardization. Such an effort ought also to include an assessment of the present order of the inkblots. Our persistent impression held that certain of the later ones obtained fewer M responses than early ones, though the arrangement supposedly yields a series in which M responses become progressively easier to obtain. The final inkblot (#28) especially seems misplaced.

The SFPI was demonstrated to be a useful and highly reliable measure for our sample. The scoring can apparently be accomplished to discriminate





60


a LF and HF group with an extraordinary degree of agreement between actual scoring of protocols and a more "global" scanning and sorting procedure. It is further possible to administer according to standardized procedures. The Tell-a-Story was only slightly less reliable. It is hoped that other investigators will be interested in further use of these instruments, enabling cross-study comparisons as well as the possibility of obtaining some fully standardized tools for research in children's make-believe play.

The 0. J. Harvey (1970) This I Believe test proved very difficult to score reliably into four Systems according to the scoring criteria.

Better results were obtained if the respondents were simply categorized as "abstract" or "concrete." It would seem that widely-used instruments should provide scoring criteria which enable ready and reliable assessment, again in the interest of cross-study comparisons.

The Home-Play Environment attitude scale, as used in this investigation, was not checked for reliability. If subsequent study can document this, it too can be recommended for further research as an attractive and easily administered instrument. Parents seemed to enjoy responding and, later, discussing their responses with each other and with the examiners.

The Interaction mother-child play activity, contrived for this study, seems to have potential as a predictor of children's tendency toward high or low fantasy play. It was not checked for reliability and the method of recording for use of toys must be refined. It requires only about twenty minutes for administration. Scoring is relatively simple and, though this needs more extensive checking, seems quite reliable. Children and mothers alike were readily engaged in the activity, and children





61

and sibs enjoyed playing with the toys later, while examiners were busy

with parents.

Finally, itshoudbe noted that the research into children's play had great fascination for parents. Parents were charmed with being invited to discuss their own preferred childhood activities (for the SFPI) and were eager to compare notes after separate interviews. It appeared that few parents had described their childhood fun to each other, and the occasion seemed to open new areas for communication. Often, one or the other would exclaim, "You never told me that!" This might offer a fruitful direction for further research, especially if extended into present activities. What do adults do now for "play?" What kinds of childhood activities are carried forward into adult life? Neededed also is an assessment of the role of television in both children's play and adult leisure-time pleasure.

Conclusions

Play, the ubiquitous childhood activity, has until recently been

little researched. One of the difficulties is that an operational definition must almost certainly invoke the notion of spontaneous, freely chosen engagement; therefore when the scientist attempts to quantify it or remove it to his laboratory or circumscribe it for closer scrutiny, he tends to violate its most salient characteristic. It is, of course, the old problem of "observer effect."

This investigation has attempted to check out some previously used

instruments and suggestive findings of other researchers and to test some hunches about background influences on fantasy play. The results were modest. Supported was an indication of mother's influence in terms of





62


her own fantasy pre-disposition, her years of school, her interaction with the child in a play activity and her attitudes about constraints on play in the home. Unsupported was the possibility of a difference in cognitive style of parents as influencing fantasy play or substantial influence from fathers' attitudes or similarity of siblings with regard to fantasy predisposition. Although the IQ of the child seems to be associated with level of make-believe, the sex of the child, ordinal position or number of rooms in the household does not.
Obscured in scores and tests of significance, however, are the many delightful incidents which, unscientifically, have served as reminders of the elusive qualities of play.
There was the tiny kindergartener who, upon being escorted back to her room after testing, stopped in the hall and exclaimed "Oh, I have to go back. I forgot to say goodbye." The student examiner was puzzled. "Yes," responded the small lass solemnly. "The clown and the monkey. They were watching us all the time." The student examiner reporting this was still a little shaken!

There was the little boy with an imaginary companion named "Jethro" who "wakes me up at night to get him a drink of water." And another whose imaginary friend is named "abdominal snow man." The spelling is correct.

There was a young father with eyes alight as he described making fireworks in his basement. He and his own father (the childrens' grandpa) have "wars with castles" whenever the mother is not around to object. They make castles out of blocks at opposite ends of the living room and catapult little rockets at each other's construction. The three small girls wear hard hats and become "knights," retrieving and





63

passing ammunition.

And a set of sisters who captured one of the student examiners, taking her out to their play house and refusing to let her go; each time she suggested returning to the house, they insisted on having another pretend dinner.

Many of the protocols include drawings the children made for us; often we found additions or exchanges in the cannister of toys -- we also discovered early that we would need several duplicate sets of toys because the attrition was so high, attesting, perhaps, to their child-appeal.

Lately it was learned that the study appeared to have introduced a new pretend activity in one neighborhood. The mother of an interviewfamily, meeting the investigator in a store, reported that her children had invented a game called "psychologist." The procedure consists of dumping a cannister of toys out on the table and then making copious "notes" on papers pinned to a clip board.

With the study of play still in its infancy, with its boundaries,

as Klinger (1969) states, "ambiguous," with rigorous operational definitions still to be fashioned, the present study must be regarded as an excursion into unventured territory.

Such investigations can help to discover the new questions to be

asked and assist in charting for further research. It is hoped that in this vein the present study will make a contribution.




























APPENDICES





65


APPENDIX A


Directions for Testing

Barron's Movement Threshold Inkblots

To the child: I am going to show you some cards with designs on them. They aren't made to look like anything -- they are just designs that were made by dropping ink on paper and folding it. I want you to look at each of the designs and tell me what it reminds you of -- what it looks like to you. Every child can see different things and there aren't any right or wrong answers. Sometimes one card may look like more than one thing but I only want you to tell me one thing that you see on each card.
(To the tester: If a child gives more than one response to a card repeat that you only want one answer per card and ask him to tell you which of his answers that design reminds him of most.)

Peabody Picture Vocabulary Test (PPVT)

To the child: "I have some pictures to show you." (Turn to example A and say:) "See there are four pictures on this page. Each of them is numbered. (point to each one) I will say a word, then I want you to point to the picture which best tells me the meaning of the word. Let us try one. Point to the picture that best tells the meaning of 'crib.'" (After the child responds correctly, then turn to example B saying:) "That's fine. Which picture is fin?" (Then to C.) "What picture is 'butterfly?" Fine! Now I'm going to show you some other pictures. Each time I say a word you point to the picture which best tells the meaning of a word. As we go through the book you may not be sure you know the meaning of some of the words, but I want you to look carefully at all of the pictures anyway and choose the one you think is right. What picture is ?"

To the examiner: Be sure that each child looks at each of the alternatives because sometimes younger children will start looking in only one location. Also, always be sure to get a response for every card. If the child doesn't know the word, have him give it a try anyway. You may encourage the child after some responses by saying "good," "fine," etc., but don't do it just after correct responses. You are praising the child's effort not its correctness. You may repeat a word if necessary. Do not precede any word with an article (a, an, the) and do not make any singular words plural.

Scoring: Start with item #25 for all children. Continue forward until the child makes his first error. If this is before he gets 8 correct, go back to a lower level (below 25). The ceiling score is determined by the point at which the child has missed 6 out of 8 words.





66

APPENDIX B

Singer's Structural Fantasy Pre-disposition Interview (SFPI)

To the tester: These questions are designed to discover the nature of each child's play and fantasy life. They should be presented slowly and matter-of-factly to the child with as little comment on your part as possible except to elicit details when an answer is unclear. Because scoring is based on whether or not the child's answer indicates that he is introducing make-believe elements into his play, it will be necessary to follow-up answers which are not clearly make-believe or not makebelieve. However, do not go beyond the probe in questioning, though it can be used more than once.

To the child: I am going to ask you some questions about things you like to do.
(1) What do you like to play best? What do you like to play the most?
(Probe: Tell me more about it.) (Probe: Tell me how you play it.)
(2) What do you like to play best when you are all alone? What do you
like to do best when you are all alone?
(Probe: Tell me more about it.) (Probe: Tell me how you play it.)

(3) Do you ever see make-believe things with pictures in your mind or
think about them? What sort of things?
(Probe: Tell me more about it.)

(4) Do you have a make-believe friend? Do you have an animal or toy or
make-believe person you talk to or take along with you?
(Probe: Tell me more about this.)

Tell-a-Story (Stimulus situation; card #6. Human CAT.

To the child: Now I am going to show you a picture. Tell me a story about what you see.

(Probe: Make up a story about it.)

(Probe: Anymore?)

(Probe: Anymore to your story?)





67


APPENDIX C


0. J. Harvey questionnaire


Now we would like to know your opinions and beliefs about several
other topics. Please write at least two (2) sentences about each topic. You will be timed on each topic at a pace that will make it necessary for you to work rapidly. Be sure to write what you genuinely believe. Take the topics in order, beginning with number 1.

1. This I believe about marriage . .


2. This I believe about religion .


3. This I believe about the American Way of Life .


4. This I believe about friendship .


5. This I believe about people .


6. This I believe about guilt .


7. This I believe about women's rights .


8. This I believe about work .


9. This I believe about law . 10. This I believe about education .




68


APPENDIX D


Home-Play Questionnaire

This is a short questionnaire about some of your feelings about play. Notice that there are possible answers showing about how some people feel about play. Circle the letter of the one which is closest to how you feel. If none of them apply write your own answer in the bottom space.

1. When should children play?

a. When chores and work are all completed.
b. A regularly scheduled play time is best.
c. Whenever they want to.
d. Only with parent's permission.
e.

2. Boys should be discouraged from playing with girls' toys and games.

a. Only when the child seems to play with girls' toys to excess, or
more than he plays with boys' toys.
b. Never.
c. Always.
d. Only when there is no adult male or father in the household.
e.

3. A child should share his toys with other children.

a. Always.
b. With visitors and brothers and sisters.
c. If the parent thinks it should be shared.
d. When ever he wants to.
e.

4. Wrestling or rough-housing should be done:

a. Only outdoors or in designated areas.
b. Only when closely supervised by an adult.
c. No restrictions ought to be placed on it.
d. Only by boys.
e.

5. Children's play should be mainly things that teach them useful ideas
and skills.

a. Completely agree.
b. M'ldly agree.
c. Completely disagree.
d. Mildly Disagree.
e.





69


6. Make-believe play ought to be encouraged.

a. Mainly when children are younger.
b. As much as possible for all children.
c. Unless the child is doing too much of it.
d. No, it rather should be discouraged.
e.

7. Girls should be discouraged from playing with boys' games and toys.

a. Never.
b. Always.
c. Only when child seems to play with boys' toys to excess or more
than she plays with girls' toys.
d. Only as she grows older if she continues to prefer boys' toys,
things.
e.

8. How should a parent react when the child is using a toy in a wrong
way enjoyably?

a. The child should be stopped and taught the correct way to use
the toy.
b. The incident does not call for any special reaction.
c. Probably ought to warn the child and if misuse of toy persists,
punish.
d.





70

APPENDIX E

Interaction

Here are lots of little toys. Both of you, playing together, use any of them to build an exciting scene -- as if for a movie -- and then tell me what is happening. (Permit the building of the scene to occupy five minutes. Then say, "Now can you tell me what is happening?")





71


Interaction

Child M Mother figure Father figure

baby toy furniture car

rhino lion

horse policeman


"mother"

"father" "baby" blocks "car" green thing red thing Dumbo

Superman pipe cleaners popsickle sticks





72

APPENDIX F

Demographic Information

Father's name Age Occupation Last year completed schoolingMother's age Occupation Last year completed schooling Number of children in family____ Name Age Name Age Name Age Name Age Name Age Others living with family Age_ Years of Schooling Number of rooms in house





73

APPENDIX G


Variables for Correlation Matrix

Col 1 Child's IQ Col 2 Child's Barron Col 3 Child's Tell-a-Story Col 4 Child's Fantasy Toys (used for Interaction) Col 5 Mother's Age Col 6 Mother's Years of School Col 7 Mother's Barron Col 8 Mother's Tell-a-Story Col 9 Mother's Fantasy Toys (used for Interaction) Col 10 Father's Age Col 11 Father's Years of School Col 12 Father's Barron Col 13 Father's Tell-a-Story Col 14 Number of Rooms in Home Col 15 Child's SFPI Col 16 Child's Realistic Toys (used for Interaction) Col 17 Interaction Col 18 Mother's SFPI Col 19 Mother's Numbers of Permissives Col 20 Mother's Number of Restrictives Col 21 Mother's Realistic Toys (used for Interaction) Col 22 Father's SFPI Col 23 Father's Number of Permissives Col 24 Father's Number of Restrictives Col 25 Number of Persons in the Household Col 26 Child's Total Number of Toys Col 27 Mother's Total Number of Toys Col 28 Room/Person Ratio



















CORRELATION MATRIX


COL. COL. COL. COL. CCL. COLt.. t.
1 2 3 4 5 6 7 8


1 1.0000 -0.4888 + 0.5474 4 -0.0839 -0.0718 0.2084 0.1104 0.0864
2 -0.4808 -* 1.0000 -0.8346 f -0.0776 -0.15R4 -0.3538 t n.1333 -0.1473
3 0.5474 -0.8346 1.0000 0.0777 0.0649 0.1963 -0.7234 0.2194
4 -0.0639 -0.0776 0.0777 1.0000 0.1754 0.1845 -n.1724 0.177R
5 -0.0718 -0.1584 0.0649 0.1754 1.0000 0.2955 -0.179Q -0.0101
6 0.2064 -0.3538 A 0.1963 01844 0.7954 1.0000 -0.4793 n.n0.76 "
7 0.1104 0.1333 -0.2236 -0.1234 -0.1795 -0.47q93 1.000 -n.A147
8 0.0864 -0.1473 0.2135 0.1778 -0.0101 0.076b 1 -0.1147 1.0000
9 0.0388 0.0901 -0.1907 -0.40634. -0.2178 -0.1247 0.1728 -n0.052
10 0.0440 -0.2276 0.1278 0.1845 n.8s75* 0.7433 -0.1814 O.n096 11 0.1173 -0.1635 0.0127 0.2311 0.0016 0.24R6 -0.1811 -0.0799
12 0.0847 0.0053 0.0571 0.2914 0.058O 0.24qP -0.1409 n.7494 13 0.2273 -0.0930 0.1323 0.3873r -0.1517 0.2008 -n.3185 0.418R6
14 0.1363 -0.1686 -0.0487 0.0169 -0.0087 0.301 -n.104R -n.0271
15 0.49304- -0.6167 it 0.7464" 0.0903 0.0529 0.364 C -0.3899 n.1393
16 -0.1161 -0.3015 0.1859 0.1838 0.41275 0.119? -0.1179 0.0644
IT 0.1579 -0.3798 0.5702* 0.1444* 0.0925 -0.0141 -0.75n 0.1,44
18 0.3603 A -0.3036 0.3270 a- 0.1611 -0.1699 0.1506 -0.1677 .lOS1P s
19 0.0847 -0.0201 0.0480 -0.0780 -0.1987 0.0135 -n.1817 .1177
20 -0.2852 0.2409 -0.2168 -0.1788 n0.623 -0.4444 o.9q ea -0.0710
21 -0.2197 0.0877 -0.1885 0.1173 -0.2496 -0.110A n.?348 -m.1p44
22 -0.2301 0.2096 -0.1965 -0.1396 -0.4134* -0.0819 -n."879 0.)312
23 0.1763 -0.0193 -0.0291 0.1849 -0.0712 0.1702 -0.1t6 0.l 8
24 -0.2968 -0.0712 0.0S24 0.1699 0.2041 -0.1807 0.n164 -0.1560
25 0.0526 0.1010 -0.1001 -0.1130 0.2049 -0.07240 -M.1&48 -n.1009
26 -0.1275 -0.2311 0.1638 0.8194- 0.1645f 0.1176 -n.1641 ".1447
27 -0.1010 0.11sb -0.2463 -0.2711 -0.3015 -0.1539 0.?607 -0.1699
26 0.0919 -0.2317 0.0129 0.0341 -n.1948 0.3447 A- -0-1? 0.0334

COL. COL. COLL. COL.L COL .. CL.
9 10 11 12 13 14 15 16 ROW

1 0.0188 0.0440 0.1173 Q.0J47 0.22713 11363 0.4q'1 -0.1161 2 0.0901 -0.2276 -0.1635 0.0053 -0.090 -0.148@ -0 617 -n.30l
3 -0.1907 0.1278 0.0127 0.0571 0.1373 -0.0487 0.74A4 k .18019 4 -0.4063 & 0.1845 0.2311 0.?914 0.1873 0.016Q 0.ng93 n.183R
5 -0.2178 0.8752r 0.0016 0.0588 -0.1517 -0.0087 n.mn9 n.4t179i
6 -0.1247 0.2433 0.2486 0.2498 0.2008 0.n000 0.36494qM 0.419 7 0.1728 -0.1814 -0.1833 -0.1409 -0.3195 -.108R2 -0.4 898 -0.1329
8 -0.0852 0.0656 -0.0799 0.2594 0.41 A+ -0.0771 0.1193 0.0644
9 1.0000 -0.1021 0.0125 -0.4612* -0.7285 0.0441 -0.1679 -0.OP47
10 -0.1021 1.0000 0.0224 0.0903 -0.0742 0.0176 0.151 0.16nMA 11 0.0125 0.0224 1.0000 -0.0628 0.3qo4-4 0.3411 V- 9.2 9 0.3'47 14
12 -0.4612 0.0903 -0.0628 1.0000 M.1891 -0.0214 0.1047 -0.0415
13 -0.2285 -0.0742 0.3904 0.1891 1.0000 0.1640 O.tbn6 9.3075 14 0.0443 0.0178 0.3411 Ir -0.0214 0.1640 1.0000 0.149A -0.0684


1 -0.1679 0.1541 0.2559 0.1047 A.lj-n 0.0690 1.n000 0.7244 16 -0.0857 0.16064 0.1342* -0.0435 0.047 -0.0684 0.X'S 1.0a000
17 -0.0766 0.0456 0.0420 0.1485 0.7609 -0.2557 0.4139* 0.1747 18 0.0304 -0.0604 0.1447 -0.0093 0.1777 -0.05os 0. 49 -0.0460
19 0.36586 -0.0824 0.1664 0.1315 0.0489 0.1756 0.11i -0.2911 20 -0.0093 0.0272 -0.3710t -0.1852 -0.3139 -0.4170* -0.4194 -0.19A9
21 0.1776 -0.2487 -0.3348 f -0.0864 -0.1775 0.072, -0.1627 -0.1561 +
22 0.1099 -0.3549 0.0062 0.1521 0.1758 0.105 -.076tb -0~ 94q 21 0.1020 -0.0956 0.2264 -0.0720 0.3819 0.1520 0.1138 0.1040 24 -0.2511 0.2577 -0.1794 -0.1446 -0.1331 -0.199 -0.0804 0.0707
25 -0.0227 0.0600 0.0658 -0.1703*- 0.2331 0.,973- -0.1567 -0.12
26 -0.4393 4 0.3410 A 0.3595 + 0.1822 0.3268 -0.0278 0.142 0.71441 27 0.8172 f -0.2185 -0.1872 -0.3?79A -0.2f40 0.0448 -0,7147 -n.7698
28 0.0919 0.005 0.2901 0.1904 -0.046A 0.71T13 L 6.0424




















cnL. COL. COL. COL. (C. COL. CL. COL.
17 is 19 20 21 77 73 ?4


1 0.1579 0.3603 0.0847 -0.2852 -0.2197 -0.73nl 0.1763 -0.7968
2 -0.3798 -0.3036 -0.0207 0.2409 0.0877 0.7096 -5.0193 -0.0717
3 0.5202 r 0.3270-t 0.0480 -0.2168 -0.1885 -0.1965 -.07) 1 O.0~24
4 0.3444 0.1411 -0.0780 -0.178R 0.1173 -0.1316 n.1849 n.1609
5 0.0925 -0.1699 -0.1987 0.0621 -0.2496 -0.4114 j -n.07l7 0.Pn41
6 -0.0343 0.1506 0.0135 -0.56444+ -0.110A -0.0810 n.170? -n.IPn?
7 -0.2500 -0.0677 -0.1812 0.390814- 0.734A -0.0879 -n.165 .n.0144
8 0.1246 0.3508 0.1177 -0.o3730I -0.1044 0.2311? 0).19 -0.1560
9 -0.0766 0.0104 0.38 58F -0.0093 0.1776 0.lO9 A.l02n -M.2511
10 0.0456 -0.0604 -0.0824 0.0272 -0.2487 -0.1140 A -A.0556 4.2 77
11 0.0470 0.1447 0.1664 -0.37104' -.f3748 0.0062 0.7?64 -0.1794
12 0.148S -0.0091 0.1115 -0.1857 -0.0864 0.17? -0.07 -A.1468
13 0.2609 0.1777 0.0489 -0.3139 -0.1729 0.1758 0.9819 t -n.1311
14 -0.2557 -0.0510 0.1756 -0.4170 4 0.0276 0.105 j.1470 -n.15q9
15 0.41354- 0.365* 0.1112 -0.4195 -0.167 -(0.076i ".7748 -0.0804
16 0.1757 -0.0460 -0.2913 -0.1969 -n.3563 4 -0.2949 0.1040 n.A077
17 1.0000 0.1756 0.2819 -0.1080 -0.1901 -0.0111 0.100i -0.18A4
18 0.1756 1.0000 0.0591 -0.9.179 -0.0o 8 -0.0954 0.1500 -0,0661
19 0.2819 0.0591 1.0000 -0.1284 0.0717 0.416k 0.1161 -0.5987 20 -0.1080 -0.3792- -0.3284* 1.0000 0.1100 -0.7672 -0.2171 0.4041
21 -0.1903 -0.0238 0.0717 0.1300 1.0000 0.0889 -0.3114 0.1995
22 -0.0111 -0.0954 0.4163' -0.2622 0.nPQ9 1.0000 0.16p6 -0.1764
23 0.3100 0.1500 0.1161 -0.2321 -0.0114 0.2166 I.0no0 -n.2694
24 -0.1864 -0.0661 -0.19824 0.40414 0.1955 -0.1764 -0.1694 I.nno
25 -0.1031 -0.0882 -0.0496 -0.1037 0.0038 0.0250 0.0455 -0.1190
26 0.3477 0.0879 -0.2254 -0.7422 -0.1741 -0.7714 n.1973 0.16??
27 -0.1661 0.0077 0.3172 0.0695 0.7123 0.114 9 0.0661 -0.0646 28 -0.2377 -0.0567 0.2245 -0.34644- -0.0125 0.0847 0.170 -0.082

COL. COL. COL. COL. (nL. CnL. rLt. CLt.
25 26 77 28
1 0.0526 -0.1275 -0.1010 0.0919 2 0.1010 -0.2311 0.1156 -0.7317 3 -0.1001 0.1638 -0.2463 0.0129 4 -0.1130 0.8194* -0.2211 0.0141 5 0.2049 0.,3655 -0.3015 -0.1548
6 -0.0240 0.3176 -0.1938 0.1447
7 -0.1648 -0.1653 0.2607 -0.0307 8 -0.1009 0.1642 -0.1688 0.0334 9 -0.0227 -0.3393v4 0.8172 0.0919

10 0.0600 0.34184# -0.7185 0.0055 11 0.0658 0.35954* -0.1872 0.2901 12 -0.3203, 0.1822 -0.3796+* 0.1904 11 0.2331 0.3268*- -0.7640 -0.0466 14 0.39734 -0.0276 0.0448 0.71394? 19 -0.1567 0.1952 -0.2147 0.1842 16 -0.1226 0.7141*" -3.2698 0.0525 17 -0.1031 0.3477?1 -0.1661 -0.2377 18 -0.0862 0.087 0.0077 -0.0567 19 -0.0496 -0.2254 0.3172 0.2245 20 -0.1037 -0.2422 0.0695 -0.34644i 21 0.0038 -0.1243 0.7121 n4' -.0125 22 0.0250 -0.2714 0.1305 0.0847 21 0.0455 0.1923 0.0661 n.0970 24 -0.1190 0.1622 -0.0646 -0.0820 25 1.0000 -0.1720 -0.0119 -0.91754" 26 -0.1s5n 1.0000 -0.3148 0.0549 27 -0.0139 -0.3140 1.0000 0.0o83 28 -0.33751 0.0549 0.049 .0 1,0000





76


APPENDIX H

Letter to Parents Requesting Consent

Dear Parent:

This letter is to introduce me and a research project I am conducting regarding children's play. I am Lorraine Dennis, an instructor at Roger Williams College. The research will become a part of my dissertation for the Ph.D. degree at the University of Florida.

For the project, my research assistants (advanced students of
psychology from Roger Williams College) and I will be studying the play of kindergarten and first-grade children. We will be asking the child to do the following: 1) make up a story about a picture of some people who are camping; 2) describe what is seen in a set of ink blots, and 3) answer four questions about favorite games and play activities. Total interview time is about twenty minutes. The purpose of the interview is to obtain information regarding children's preferred ways of playing.

Later, we intend also to study a small number of families to see whether brothers and sisters are generally alike in their style of play and what kinds of play the parents most enjoyed when they were young. These families will be contacted individually and, of course, will be included only if they are willing to grant permission.

It should be emphasized that the identity and name of each child
(and of any participating family) become lost when combined in the group measures and statistics. Findings are expressed in percentages, agelevels, correlations, boys versus girls, trends, etc. No names will ever appear!

We believe that this research will help provide a better understanding of the place of play in the life of a child -- and that the results will be of interest to all of us who live and work with children. Therefore, we plan to offer a report of the general results of this study to any interested parents.

If you have any questions concerning any aspect of this study, please contact me at 255-2105 (my office) or 245-4145 (my home).

We will be deeply grateful for permission to include your child in this first phase of the study. Please sign either form #1 or form #2 indicating your approval or rejection. An envelope is enclosed for your convenience; however, the form may be returned to your child's school. Thank you very much!


Lorraine Bradt Dennis
LBD/mgd Psychology Area Enclosures





77

APPENDIX I


Consent Form form #1

I GIVE PERMISSION FOR MY CHILD
(child's name) TO BE INTERVIEWED IN CONNECTION WITH THE STUDY OF CHILDREN'S PLAY.







(Parent's or Guardian's Signature)






form #2


I DO NOT WISH MY CHILDREN TO BE INTERVIEWED FOR THE STUDY OF

CHILDREN'S PLAY.
(child's name)





(Parent's or Guardian's Signature)













REFERENCE LIST


Barron, F. Threshold for the Perception of Human Movement in Inkblots.
Journal of Consulting Psychology, 1955, 19, 33-39.

Baumrind, D. Current Patterns of Parental Authority. Developmental
Psychology Monographs, 1971, 4, 1-103.

Bell, R. Q. A Reinterpretation of the Direction of Effects in Studies of
Socialization. Psychology Review, 1968, 75, 81-95.

Bell, R. Q. Stimulus Control of Parent or Caretaker Behavior by Offspring.
Developmental Psychology, 1971, 4, 63-72.

Bishop, D., & Chace, C. Parental Conceptual Systems, Home Play Environment and Potential Creativity in Children. Journal of Experimental
Child Psychology, 1971, 12, 318-338.

Caldwell, B., & Hersler, L. Mother-Child Interaction During the First
Year of Life. Merrill-Palmer Quarterly, 1964, 10, 119-128.

DeVore, I. Mother-Infant Relations in Free-ranging Baboons. In H. L.
Rheingold (Ed.), Maternal behavior in mammals. New York: Wiley,
1963.

Dewing, K. & Taft, Some Characteristics of Parents of Creative Twelve
Year Olds. Journal of Personality, 1973, 41, 71-84.

Eifermann, R. Social Play in Childhood. In R. E. Herron & B. Sutton-Smith
(Eds.), Child's Play. New York: Wiley, 1971.

Elkonin, D. Symbolic Representation and its Functions in the Play of
Children. In R. E. Herron & B. Sutton-Smith (Eds.), Child's Play.
New York: Wiley, 1971.

Erikson, E. Play and Actuality in Development. In M. Piers (Ed.), Play
and development. New York: W. W. Norton, 1972.

Freyberg, J. T. Increasing the Imaginative Play of Urban Disadvantaged
Kindergarten Children Through Systematic Training. In J. L. Singer,
The child's world of make believe. New York: Academic Press, 1973.

Gilmore, J. B. Play, a Social Behavior. In R. Haber (Ed.), Current
research in motivation. New York: Holt, Rinehart and Winston, 1966.



78





79


Goodall, J. Nest Building Behavior in the Free Ranging Chimpanzee. Annuals
N. Y. Academy of Science, 1962, 102, 455-467.

Gordon, I. J. Stimulation Via Parent Education. Children, 1969, 16, 57-59.

Harvey, 0. J. Beliefs and Behavior: Some Implications for Education. In
The science teacher. Washington, D.C.: National Science Teacher's
Association, 1970.

Harvey, 0. J., Hunt, D. E., & Schroder, H. M. Conceptual systems and
personality organization. New York: Wiley, 1961.

Harvey, 0. J., White, J., Prather, M., & Alter, R. Teachers' Belief
Systems and Preschool Atmospheres. Journal of Educational Psychology,
1966, 57, 373-391.

Hess, R. D., & Shipman, V. C. Early Experience and Socialization of
Cognitive Modes in Children. Child Development, 1965, 36, 869-886.

Hutt, C. Exploration and Play in Children. In R. E. Herron and B. SuttonSmith (Eds.), Child's Play. New York: Wiley, 1971.

Isaacs, S. Social development in young children. New York: Schocken
Books, 1972. (originally published, 1933)

Klinger, E. Development of Imaginative Behavior: Implications of Play
for a Theory of Fantasy. Psychological Bulletin, 1969, 72, 271290.

Labov, W. Academic Ignorance and Black Intelligence. Atlantic Monthly,
1972, 229, 59-67.

Lieberman, N. J. Playfulness and Divergent Thinking: An Investigation
of Their Relationship at the Kindergarten Level. Journal of Genetic
Psychology. 1965, 107, 219-224.

Lytton, H. Creativity and education. New York: Schocken Books, 1972.

MacKinnon, D. W. The Nature and Nurture of Creative Talent. American
Psychologist, 1962, 17, 484-495.
Manosevitz, M., Prentice, N., & Wilson, F. Individual and Family Correlates of Imaginary Companions in Preschool Children. Developmental
Psychology, 1973, 8, 325-329.

Mason, W. Early Deprivation in Biological Perspective. In V. Denenberg
(Ed.), Education of the infant and young child. New York: Academic
Press, 1970.

Millar, S. The psychology of play. Baltimore: Penquin Books, 1966.

Murphy, L. Infant's Play and Cognitive Development. In M. Piers (Ed.),
Play and development. New York: W. W. Norton, 1972.





80


Piaget, J. Play, dreams and imitation in childhood. New York: W. W.
Norton, 1972. (originally published, 1951)

Pulaski, M. A. Play as a Function of Toy Structure and Fantasy Predisposition. Child Development, 1970, 41, 531-537.

Pulaski, M. A. Toys and Imaginative Play. In J. L. Singer, Child's
world of make-believe. New York: Wiley, 1973.

Pulaski, M. A. The Rich Rewards of Make-Believe. Psychology Today, 1974,
7, 68-76.

Rebelsky, F. Mother-Infant Interaction. In F. Rebelsky & L. Dorman (Eds.)
Child Development & Behavior. New York: Knopf, 1974.

Repina, T. A. Development of Imagination. In A. V. Zaporoahets & D.
Elkonin (Eds.), The psychology of preschool children. Cambridge,
Mass.: The MIT Press, 1971.

Saltz, E. & Johnson, J. Training for Thematic-Fantasy Play in Culturally
Disadvantaged Children: Preliminary Results. In Studies in intellectual development. Detroit: Wayne State University, 1973.

Schaefer, C. E. Imaginative Companions and Creative Adolescents.
Developmental Psychology, 1969, 6, 747-749.

Singer, J. L. Imagination and Waiting Ability in Young Children. Journal
of Personality, 1961, 29, 396-413.

Singer, J. L. The child's world of make-believe. New York: Academic
Press, 1973.

Smilansky, S. The effects of socio-dramatic play on disadvantaged preschool children. New York: Wiley, 1969.

Stone, G. P. The Play of Little Children. In R. E. Herron and B. SuttonSmith (Eds.), Child's play. New York: Wiley, 1971.

Sutton-Smith, B. The Role of Play in Cognitive Development. Young
Children, 1967, 6, 202-214.

Sutton-Smith, B. Introduction. In E. Avedon and B. Sutton-Smith (Eds.),
The study of games. New York: Wiley, 1971.

Weikart, D., Rogers, L., & Adcock, C. Cognitively Oriented Curriculum:
a Framework for Preschool Teachers. Urbana, Ill.: University of
Illinois, 1970.

Weisskopf, E. A Transcendence Index as a Proposed Measure on the TAT.
Journal of Personality, 1950, 29, 379-390.

Yarrow, M. R. Problems of Methods in Parent-Child Research. Child
Development, 1963, 34, 215-226.














REFERENCE NOTES

1. Ellis, M. Play, a paradox for teacher and scientist. Paper presented
at the Second Annual Symposium on Play, Atlanta, Georgia, April,
1974.

2. Kempler, B. Welcome and introductory remarks. Paper presented at
the Second Annual Symposium on Play, Atlanta, Georgia, April,
1974.

3. Singer, D., & Singer, J. L. Television, adult modeling and physical
structure of the setting as factors enhancing play in pre-school
children. Paper presented at the Second Annual Symposium on
Play, Atlanta, Georgia, April, 1974.

4. Branch, A. Personal communication. December, 1973.

5. Harvey, 0. J. "This I Believe" Scoring Instructions, 6/71 and 9/71.
Mimeographed handout at address by 0. J. Harvey, University of
Florida, January, 1972.

























81













BIOGRAPHICAL SKETCH

Lorraine Bradt Dennis was born May 11, 1921, in Norway, Michigan,

attended public schools in Hibbing, Minnesota and graduated in 1938. She earned a combination degree, B.S., R.N., at the University of Minnesota in 1943 and a Master of Science in Psychology at Kansas State College in 1951.

Her professional career has included teaching assignments at Kansas State College, Drake University, the Pennsylvania State University and Marymount College of Virginia. In addition, she is the author of a text book, Psychology of Human Behavior for Nurses, published by W. B. Saunders Co. and was one of the founders of the Child Study Center in Caracas, Venezuela, designed to offer psychological and psychiatric services to English language schools in that city.

She is a member of the American Psychological Association, the Society for Research in Child Development, the American Educational Research Association, Sigma Theta Tau, an honorary nursing society, and Phi Kappa Phi, an honorary scholastic society.

She is, at present, a member of the faculty of Roger Williams College, Bristol, Rhode Island and serving as coordinator of the Psychology Area.










82







I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Ira J. GordJ Chairman Graduate Research Professor Foundations of Education

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the gree of Doctor of Philosophy.



/Emile des
A sociate, ofessor of Education Foundations f Education


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Barry". Lest
Assistant Pr fessor of Psychology

This dissertation was submitted to the Graduate Faculty of the College of Education and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy.

March, 1976


Dean, College of Eduction



Deanf Graduat School




Full Text
23
are obviously depicted in the stimulus situation. Thus, the child would
be given three points for describing the larger figures as "the mother
and father" two for ascription of sex to each character and one for
identification of a family relationship. Elements once scored were not
credited upon repetition.
The validity of indices of fantasy pre-disposition as measures of
imaginative play rests on a number of studies conducted by Singer (1973)
and his colleagues. All findings point with statistical significance in
the same direction. Pulaski (1970) found correlations between amount of
imaginative play and ratings for fantasy pre-disposition significant at
the .001 level. Freyberg (1973) also found a significant correlation
(p=.01) between fantasy pre-disposition scores and observations of make
believe play for the children in her study.
Scoring of the fantasy pre-disposition protocols was accomplished by
the investigator and the assistants. Each protocol was scored twice, in
dependently. The individual who had tested the child did not score the
protocol. The two independent scorings were compared and any discrepancy
resolved by means of consensus between the two scorers. The check for
reliability of scoring is detailed in a subsequent section.
After all the protocols had been scored, a median score was obtained
for each of the three measures. In order to be assigned to the high
fantasy pre-disposition group, a child had to score below the median
(because of the reverse scoring) on the Barron and above the median on
the other two measures, while the child who was rated low fantasy pre
disposition would have scored above the median on the Barron and below
the median on the SFPI and Tell-a-Story. The scoring criteria are
listed in Table 4.


4
In addition, because the field is young and tools and instruments
have neither been extensively developed nor tested, it was determined to
employ several existing measurement devices in an effort to assist in
establishing a further gauge of usefulness. Cross-study comparisons are
also facilitated when investigators continue with techniques and in
directions previously delineated by others.
The purpose of this study, accordingly, is to investigate possible
relationships between level of children's imaginative play, measured with
three instruments developed by Singer (1973) and his students, and several
familial variables which borrow from suggestions of Bishop and Chace (1971)
regarding requisites for a playful environment.


10
middle-class kindergarten, first and second graders, selecting a group of
"high fantasy pre-disposition" and a group of "low fantasy pre-disposition"
on the basis of several measures developed by Singer (1973). The results
indicated that for children of five, six and seven years, the type of toy
seems to have little effect. High fantasizers played with both types of
toys imaginatively (according to a fantasy rating scale) while low fanta
sizers simply "fooled around," whether the toy was structured or unrealis
tic. Pulaski speculated that children's "playing style" may be fairly
well formed by age five. Branch (Note 4) is testing this by conducting a
similar study, using much younger children, at Yale University. A group
of boys and girls who are 20 months and 26 months old are being observed
at home during play with high-realistic versus low-realistic toys. Only
the preliminary results are available, but these suggest an age effect,
with all children engaging in more make-believe at the later age, a sex
effect with girls playing more imaginary games at both age levels, and a
toy by sex by age interaction with boys decreasing the amount of their
pretend play with high realistic toys at the older age level while girls
increase the amount of pretend play with high realistic toys at the later
age level. Branch notes also that the substantial differences exist in
amount of pretend play across subjects and suggests that even at this
early age, children may evidence pretend play as a distinctive play mode.
Family Influences on Play
Few studies have been located which attend to familial correlates of
play. Several studies can be considered pertinent if a link can be con
ceptualized between the imaginative productions of symbolic play and the
imagination necessary for creativity.
One study which is directly concerned with family variables and


APPENDICES


55
The expected correlations between parents' scores on these measures
and the scores of their children did not wholly materialize. It had been
anticipated that parents who reported a good deal of make-believe play in
their own childhood and whose Barron and Tell-a-Story contained substantial
fantasy elements would tend to encourage children's fantasy pre-disposition.
Although this was suggested for two of the mothers' scores (SFPI and
Barron), fathers' scores did not seem related to those of their children
at all. It may be that a number of the fathers in this sample are too
busy for extensive contact with their children. Ten of the fathers are
college students in addition to being heads of households with full-time
jobs; half of these are in graduate work. Five are in the Navy with
regular sea duty built into their lives.
However, it does appear that an attitude of unqualified approval of
children's make-believe distinguishes between fathers of LF and HF chil
dren. The variable did not appear to contribute significantly to the
child's SFPI on multiple regression analysis. However, father encourage
is significantly related (r = .374) to Interaction, a significant pre
dictor in both analyses. It is interesting to conjecture that perhaps
fathers' attitudes exert an indirect influence, perhaps in maintaining
a supportive rather than a negative atmosphere for fantasy, which better
enables the mother and child to engage in make-believe freely when playing
together.
A similar speculation might be offered with regard to the mothers'
and fathers' restrictives. The Bishop and Chace (1971) study had in
dicated the importance of lack of restrictiveness in home-play environ
ment provided by the mother. This finding was corroborated by the
present investigation. Bishop and Chace were not able to establish any


TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ii
ABSTRACT v
CHAPTER I INTRODUCTION AND STATEMENT OF PROBLEM 1
CHAPTER II REVIEW OF THE LITERATURE 5
CHAPTER III DESIGN OF STUDY 15
CHAPTER IV RESULTS 36
CHAPTER V DISCUSSION 53
APPENDICES 64
APPENDIX A Directions for Testing 65
APPENDIX B Singer's Structural Fantasy Pre-disposition
Interview (SFPI) 66
APPENDIX C 0. J. Harvey Questionnaire 67
APPENDIX D Home-Play Questionnaire 68
APPENDIX E Interaction 70
APPENDIX F Demographic Information 72
APPENDIX G Variables for Correlation Matrix 73
APPENDIX H Letter to Parents Requesting Consent 76
APPENDIX I Consent Form 77
REFERENCE LIST 78
REFERENCE NOTES 81
BIOGRAPHICAL SKETCH 82
TV


7
Socio-economic Influences
The possibility that socio-economic factors influence the kind of
play in which the child typically engages rests heavily on a study con
ducted by Smilansky (1969) in Israel. Concerned with the school problems
of children from certain immigrant groups, her observations led to the
conclusion that socio-dramatic play (defined as involving at least one
other child in make-believe roles, actions and verbalizations which sub
stitute for real objects and situations, and sustained for at least ten
minutes) is virtually absent among these children. The conviction was
echoed in this country in terms of our own ghetto children by Lois Murphy
(1972), Sutton-Smith (1971) and Weikart, Rogers and Adcock (1970) among
others. Descriptions of the play of these children include notes of the
preponderance of impulsive and aimless activity, lack of sustained attention
and absence of signs of pleasure.
However, the contention that disadvantaged Israeli children do not
develop ability to engage in symbolic social play was discontinued by
Eifermann (1971) who sampled the play behavior of an enormous number of
Israeli children, carefully selected in terms of socio-economic class,
geographic location, community structure (town, kibbutz, Arab settlement)
and culture. She found that "culturally deprived" children play pretend
games both at later ages and to a far greater extent than advantaged
children. Furthermore, she reports, their pretend play is as elaborate
and as well sustained. An explanation for the contradictory findings
might lie in the setting in which play is observed. Labov (1972) suggests,
with regard to observations of language deficiencies of disadvantaged
children, that the usual setting for these judgments is simply too in
hibiting for certain children and that only by joining them "on their


54
talk after the interview sessions were completed, comparing notes about
their responses, plying interviewers with questions and engaging in dis
cussion about childrearing, often until after midnight.
The children, whether LF or HF, appeared to come from similar homes
in terms of parental concern and caring, general socio-economic circum
stances and physical environment.
Yet, by the measures used for this investigation, the children
themselves seem to be very different. They represent the two extremes
of a larger group of 221 children with regard to scores for the three
fantasy pre-disposition tests. A question must be raised with regard
to the validity of the measures. Do these children really differ in the
amount of make-believe play?
The findings of other investigators concerning the relationship of
the measures to make-believe play, observed under several different cir
cumstances, have been cited. The SFPI, the dependent variable for
multiple regression analysis in this study, is scored specifically in
terms of the pretend play which can be discerned in the responses and
seems to yield reliable scores on retest after a passing of several
months. Qualitative examination of the protocols was not within the
scope of this study, but non-systematic review indicates that the LF
children tended to respond to the question "What do you like to play
best?" with physical activity such as "play baseball" or "ride my bike,"
while HF children typically spoke of playing "house" or "school" or
"cowboys." However, the problem of validity is troublesome. Only
further research can provide definitive answers to the question. This
study was designed to begin with already-existing measures as a step
in assaying their utility.


60
a LF and HF group with an extraordinary degree of agreement between actual
scoring of protocols and a more "global" scanning and sorting procedure.
It is further possible to administer according to standardized procedures.
The Tel1-a-Story was only slightly less reliable. It is hoped that other
investigators will be interested in further use of these instruments,
enabling cross-study comparisons as well as the possibility of obtaining
some fully standardized tools for research in children's make-believe
play.
The 0. J. Harvey (1970) This I Believe test proved very difficult
to score reliably into four Systems according to the scoring criteria.
Better results were obtained if the respondents were simply categorized
as "abstract" or "concrete." It would seem that widely-used instruments
should provide scoring criteria which enable ready and reliable assess
ment, again in the interest of cross-study comparisons.
The Home-Play Environment attitude scale, as used in this investi
gation, was not checked for reliability. If subsequent study can docu
ment this, it too can be recommended for further research as an attractive
and easily administered instrument. Parents seemed to enjoy responding
and, later, discussing their responses with each other and with the
examiners.
The Interaction mother-child play activity, contrived for this study,
seems to have potential as a predictor of children's tendency toward high
or low fantasy play. It was not checked for reliability and the method
of recording for use of toys must be refined. It requires only about
twenty minutes for administration. Scoring is relatively simple and,
though this needs more extensive checking, seems quite reliable. Chil
dren and mothers alike were readily engaged in the activity, and children






42
TABLE 10
PARENT'S CONCRETE OR ABSTRACT CATEGORY
Category
Mother frequency3
Father frequency'3
L.F.
H.F.
L.F.
H.F.
Concrete
8
4
9
9
Abstract
13
13
12
8
Corrected Chi square (1 ) = .372 N.S.
^corrected Chi square (1) = .085 N.S.


1579 0.3603 0.0867 -0.285? -0.2197 -0.2301 0.1763 -0.2968
fefe
fv. 4 r- 40^4 ct 4 r- 4 ir 4 4 c o r. c
o-(v o >t c < >c^Ka Kica8*ifU'iN>r. n'0'^3CC*- CC *\Cnj cococ cecc ecr-c ccoccccecccoc o
i i i i i i i i i iii ii i ii
(*- (7> r\ >, ir O- C >C 4 c (T C I C CC<--*4^C4iT(i-8C
cr o- 4 ~ c xa.cgiAX.i\. <>.* 4co>cfe cc a-rcN.^r-
o ctr-r'-cC^rv r-acu rv. c w f C'£4Cr.c7
C C c-*- CrCi*"fv. r* \jC*\.crC CC* _j
c
c. ccccceccocooceoc.cooc~ec.ccc. i_
itii ii ii i
4 *
IT c.C4C 4 If h
V O O n a4XMflfo4 if iMCO*'" C 4
r 4eO*fe cr-CO^^CON. CvC _i
c
oococc. ccooococooceeo ccceec _
I I I I I I I I I I 1 II
fe fe *
K tn I** X <4 X! ^ cc -4 IT < M* f KCOff4irir.4'
racr-0>C44r^ 4 x> cv f\. r. X o C O r> f3 4 *v< *>
oa.> 4*ria.f-4r a. r- rv, x r cr^r-**'*!-cc*'-
C ^ fe fef^O O OCOCO Cf^C _J

eooc OOCOCC OCOC.CCCCC CCCCCCC o
I II I lili lili I I
vw v fe* fefe v *
ooflC<'4ccc*'fefe^ir^v 4CON<^l'Mr4
ox)-o^.,oo*'i\.4iiva>4)
.4 p-x>>o*r'oc4X4
NN^0if'(*'(7'00l,'^,(^44'< ocooooococooooooocoocooocc
II I II I I I I I I I I I II II I
fe
HHW4440W-*
ai" p8 ir r 4 i" m f
-i ovco *ooo
cc*******
ur^OOC OCOOCC
I I I
V fe fe
4 4X>**''*r*-r-'44*^00'V*>
4* O C 4)n4f'.iX4N4'rnr4*C
oaff4'-rirv4--(r:4 r o
Oim^Cw*-C'iCP.COCO^'OOC
ooccccoocooccc c o o o
I lili II
V
!* OOMMNM 4 4 If ir X
occec**'r-g-rvxJoor'
o.'4S(T--8X4)^4N
OOO O- O -o-
0000000000000
lili I
M V if
M>(r-o4N(i-4MX4ivir
9^- -crirorr--^a -4 cm tm
hM^00i*104-'(,'0n(on
0000 eoooooooo
1 I III
oxrnroNx^
ir o -j n o 7-
OH4NOIPX-
_J rvjfVf^fN.^tn
o o
OVOOOOOOOOO
I lili I
fe v
i'M4)OeK-r-fv'ir*3^.xancfr
'X>4 X-OO
rv ^Ocg\. Cf-. Or-oOOf-C O
oooooo*-*oooooooooo o
lili III III
^ fe
4)0-Cr'CN44MrN0
*rr cror-ooo4 CfVJXXifXlfrX4CMf
rr rn_. On-CO <~
0000000000000
I II III
V v
IT040-N40-N>^
4NOi'NNP'irx if-O-OO^OO-COOC
000-000000000
I III II
*r
r r^o4ir4)(*'N*'
o NHmO'r^trxo-
ir
C _j ni >'*** n
o
O ofMCOCOOCOOO
I II II
rNcN-Mr4Nr4i"^oocr
CTN4INr4r'r'rN4-. cw rv, r. o 4 4
4rM49^4CN4Kfe4XrOH
rfiH'wff c-8 r-nOf9N-*N-H
000000000000000000
I lili II
C r-
u
*fcfe fe %
crr\i44'ox>^x>04'ar'43r-o^<***' o* r- r-
0'04M4Q4irNoirmiroir 00 oi*r^x>r-
SN40i'r^N444XirH f-0r-0^ O 4 X>
nfrooN-ooo-NN4-o-iM-orH-m-N
0000000000000000-00000000000
I III I I I I I I I I
o^ocroatrs
Cfe O 4 4 4 O fM
roo o m x> o fe
-i O (MO O
or
0*^000000000
II lili
r -* fe
or-i"M)HNifeorccc am
OinO*'r-x>fMfn>cmir>ar'*'r-
00*rri**l 00 C C C O*'
000000000000000000
I lilil I III
-M*i4lT4^B90-M*l4irXK90-M^4irXM
iMfgr^fMfsjrgfSiMfM
- 0-N(r4'4fe0-Mf 4r 4KC
---- NNNWigNrMNN
*
o


77
APPENDIX I
Consent Form
form #1
I GIVE PERMISSION FOR MY CHILD
(child's name)
TO BE INTERVIEWED IN CONNECTION WITH THE STUDY OF CHILDREN'S PLAY.
(Parent's or Guardian's Signature)
*************************************************************************
form #2
I DO NOT WISH MY CHILDREN TO BE INTERVIEWED FOR THE STUDY OF
CHILDREN'S PLAY.
(child's name)
(Parent's or Guardian's Signature)


51
variables of the first equation, three of which had achieved a significant
F ratio (IQ, Interaction and mother's years in school) and two which might
possibly emerge if the order of entrance were changed. In this instance
it was determined that IQ should enter last.
The summary table is reproduced for Table 19.
TABLE 19
SUMMARY TABLE
MULTIPLE REGRESSION ANALYSIS II
Step
no.
Variable Entered
Multiple
R
Rsq
Increase
Rsq
F Value
1
M's Restrictives
.4195
.1760
.1760
7.6903
2
Interaction
.5596
.3131
.1371
6.9878
3
M's years school
.5941
.3530
.0398
2.0093
4
M's SFPI
.6234
.3887
.0357
1.9258
5
IQ
.6867
.4716
.0829
5.0202
The five variables account for 47.16%
(R =
.6867) of the variance
of the child's SFPI. Mothers' restrictives and the Interaction score
become significant predictors with restrictives accounting for 17.60%
and Interaction accounting for 13.71% of the variance of the dependent
variable.
IQ, also, obtains significance, accounting for 8.29% of the variance
in this equation.
On this occasion, mothers' years of school does not reach significance.
It is significantly correlated with restrictives (r = -.564) so that
employing one as a predictor seems to relegate the influence of the other
to the background.
Mothers' SFPI does not appear to be an important contributor for
either equation.


66
APPENDIX B
Singer's Structural Fantasy Pre-disposition Interview (SFPI)
To the tester: These questions are designed to discover the nature
of each child's play and fantasy life. They should be presented slowly
and matter-of-factly to the child with as little comment on your part as
possible except to elicit details when an answer is unclear. Because
scoring is based on whether or not the child's answer indicates that he
is introducing make-believe elements into his play, it will be necessary
to follow-up answers which are not clearly make-believe or not make-
believe. However, do not go beyond the probe in questioning, though it
can be used more than once.
To the child: I am going to ask you some questions about things you
like to do.
(1) What do you like to play best? What do you like to play the most?
(Probe: Tell me more about it.) (Probe: Tell me how you play it.)
(2) What do you like to play best when you are all alone? What do you
like to do best when you are all alone?
(Probe: Tell me more about it.) (Probe: Tell me how you play it.)
(3) Do you ever see make-believe things with pictures in your mind or
think about them? What sort of things?
(Probe: Tell me more about it.)
(4) Do you have a make-believe friend? Do you have an animal or toy or
make-believe person you talk to or take along with you?
(Probe: Tell me more about this.)
Tell-a-Story (Stimulus situation; card #6. Human CAT.
To the child: Now I am going to show you a picture. Tell me a story
about what you see.
(Probe: Make up a story about it.)
(Probe: Anymore?)
(Probe: Anymore to your story?)




34
in recording subsequent use of toys resulted in a low reliability for the
toy check-lists. None of the observer's corresponded exactly with those
of the investigator. The disagreement in actual toy-count was 38 out of 60,
yielding an agreement of only 36.67%.
However, the account of "what is happening" provided a reliability
of agreement on scores of 100% (i.e., the five observer verbatim records
obtained the same score as those of the investigator).
Statistical Analysis of Data
Nominal data which included the sex of the child, ordinal position,
whether or not the mother was employed outside the home, the concrete or
abstract category of the mother and of the father and the encouragement
or discouragement of make-believe play by each parent were analyzed by
the use of Chi square.
Fantasy pre-disposition scores for children of LF and HF groups, as
well as the scores of their parents, the Interaction scores for LF and
HF groups and the permissives and restrictives of each parent were
subjected to_t_tests for significant differences.
A correlation matrix was obtained which utilized all of the scores
on fantasy pre-disposition measures, IQ scores, age of parents, their
years of schooling and similar variables in order to affirm or disaffirm
predicted relationships between variables.
Finally, on the basis of the analysis of the correlation matrix, two
multiple regression analyses were performed in order to assess the relative
contribution of selected predictor variables to a predicted variable,
the child's SFPI score.
This study must be considered a pilot investigation into a field
which is in its infancy. Subjects for the study are a restricted sample,


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E9XYW9NHS_8U97WL INGEST_TIME 2015-03-27T18:46:14Z PACKAGE AA00029748_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


59
account for the failure to obtain resemblance between fantasy pre-dis
position scores of children and their siblings. One might even suggest,
with Smilansky (1969) and others, that make-believe play promotes verbal
intelligence.
These are certainly considerations which must guide further research.
Instruments
This study employed several instruments which have been used in
previous investigations: the Barron Inkblot Movement Threshold and the
Singer Structured Fantasy Pre-disposition Interview (SFPI) (Singer, 1973;
Freyberg, 1973; Pulaski, 1973), and a modified version of Tell-a-Story
(Pulaski, 1970) in order to assess the child's, parent's and siblings'
fantasy pre-disposition. In addition, the 0. J. Harvey (1970) test of
conceptual style and a modified version of the Home-Play Environment
attitude questionnaire (Bishop and Chace, 1971) were employed to measure
parent variables. It is appropriate to report on our experience with
these instruments.
The test-retest reliability of the Barron is so low that serious
doubt must be raised regarding its utility without further effort toward
standardization. Such an effort ought also to include an assessment of
the present order of the inkblots. Our persistent impression held that
certain of the later ones obtained fewer M responses than early ones,
though the arrangement supposedly yields a series in which M responses
become progressively easier to obtain. The final inkblot (#28) especially
seems misplaced.
The SFPI was demonstrated to be a useful and highly reliable measure
for our sample. The scoring can apparently be accomplished to discriminate


43
system on 0. J. Harvey et a]_. (1961) scale.
The hypothesis was not supported, either for mothers or for fathers.
Table 10 shows that the abstract versus concrete ratings of parents did
not distinguish between LF and HF groups of children. Mothers' ratings
obtained a Chi square of .372, while Chi square for fathers' ratings was
.085, both well below the level of significance.
It can be noted that the total number of parents with abstract con
ceptual systems is greater for both groups than the total number of
parents with concrete conceptual systems.
3. The third hypothesis stated the prediction of a positive relation
ship between the child's HF pre-disposition and a non-restrictive play
environment in the home.
This hypothesis received partial support.
Mother and father, each completing a modified version of the Bishop-
Chace (1971) questionnaire on home-play environment, received scores
indicating the number of unequivocal permissive responses and the number
of wholly restrictive responses. Other possible responses to the
questions, all of which contained some qualifiers, were not counted for
purposes of this scoring. There are eight questions on the questionnaire
(Appendix D) so that each parent might have received a maximum of eight
completely permissive or eight extremely restrictive scores. The scoring
results are tabulated in Table 11.
It can be noted that the number of permissive responses given by
both mothers and fathers is substantially the same for each group. The
differences between groups were not significant. Inspection of the
correlation matrix (Appendix G) reveals that the permissive scores of
both parents obtain few significant correlations; mothers' permissives


CHAPTER I
INTRODUCTION AND STATEMENT OF PROBLEM
The systematic study of play is an astonishingly neglected research
area. It is difficult to account for such a blind spot on the part of those
interested in child development, especially since there exists a common
consensus regarding the prevalence of play in a child's life -- a general
notion that given half a chance, the young child would probably play most
of the time.
However, although it is impossible to pick up a text directed to
teachers of young children without encountering a section praising the
virtues of play, as a research topic it is indexed with only four references
in the third edition of Carmichael's Manual of Child Psychology, five more
if cross-referenced as "games." All except two of these occur in Piaget's
chapter on his own theory. The 1974 Annual Review of Psychology indexes
"play" not at all.
One deterrent to the study of play seems to have been the socio
cultural milieu from which scientists select their problems. Sutton-Smith
(1971) and Greg Stone (1971) suggest that the influence of a heavy achieve
ment orientation tended to promote a view of play as the polar opposite of
useful work and hence, inconsequential, the occupation of idle children.
A second and major deterrent has undoubtedly been the problem of
formulating an appropriate operational definition. A certain circularity
of reasoning occurs in the attempt to differentiate play from not-play,
since it is necessary to separate out already-identified "playful" elements.
1


9
for things which are missing (e.g., a spoonful of pine needles as imaginary
food). Finally the adult must help the child to imitate the activities of
admired grown-ups with appropriate suggestions, such as handing the doll
to the youngster and saying, "The baby is sick. Will you examine her,
doctor?"
The Russian psychologists see make-believe and imagination as con
tributing to the child's social identity when he tries out different social
roles and also as the bridge to abstract thought.
Play as Influenced by Toys
The question of the kinds of playthings which might encourage play
has attracted the attention of several investigators.
Gilmore (1966) found novelty to be an important variable, with his
sample of children responding more readily to new rather than familiar
objects in every play situation. Hutt (1971) refined the attraction of
new objects in attempting to distinguish between play activity and
"specific exploration." Play activity would not show a response decrement
while exploration would be characterized by a response decrease. Using
a specially constructed box with a lever attached to a counter, the events
produced by the lever were varied systematically. Hutt found that after the
3 to 5-year old child had explored the features of the box and lever fully,
lever pushing decreased. Then the box and lever were incorporated into
"play" with little focus on the lever and its effects. She suggests
that novelty provokes exploration while play occurs as the object becomes
familiar. It is as if the question shifted from "what does this object
do?" (exploration) to "what can I do with this object?" in play.
Pulaski (1970) hypothesized that unstructured toys would promote
more imaginative play than toys which are very realistic. She tested


50
TABLE 17
PARENT'S RESPONSES REGARDING ENCOURAGEMENT OF MAKE BELIEVE
Response
Mother frequency9
Father frequency^
L.F.
H.F.
L.F.
H.F.
Yes
10
8
2
7
No
11
9
19
10
dcorrected Chi square (1) = .085 N.S.
^corrected Chi square (1) = 3.60 Fisher Exact Probability = .056
TABLE 18
SUMMARY TABLE MULTIPLE REGRESSION ANALYSIS I
Step Variable Entered
Multiple Rsq
Increase F Value
no.
R
Rsq
1
IQ
.4830
.2431
.2431
11.5619
2
Interaction
.5988
.3586
.1155
6.3043
3
M's Yrs. School
.6664
.4441
.0855
5.2318
4
F. encourage
.6852
.4695
.0254
1.5788
5
M's restrictives
.6952
.4833
.0138
.8518
6
M's SFPI
.6995
.4892
.0060
.3618
The analysis indicates that for this equation significant contri
butors to the variance of the child's SFPI score are IQ (24.31%), the
Interaction score (11.55%) and the mother's years of school (8.55%).
These three variables, taken together, account for a multiple R of .6664
or 44.41% of the variance of the child's SFPI.
Three variables, father encourage, mother's restrictives and mother's
SFPI fail to emerge as significant predictors. Each is significantly
correlated with a variable which entered the equation earlier.
A second multiple regression was performed with the five most likely


14
in an often-cited study of play and creativity. More recently Schaefer
(1969) surveyed 800 academically talented adolescents and discovered sig
nificant differences favoring the presence of imaginary companions in
childhood for those rated as "creative" in both art and literary expression.
Sutton-Smith (1967), also, tested the possibility of a relationship between
play and creativity, suggesting that favorite, well-used toys enable the
child to increase associations and therefore obtain more novel (i.e.,
creative) responses, a hypothesis upheld by his study for both boys and
girls. Klinger (1969), suggests that fantasy and play are essentially
undifferentiated until the third year, thereafter differentiated gradually
as play becomes rule-governed. He, too, indicates that the imaginative
productions of fantasy may be implicated in creativity.
Taken together, the studies of home and family variables, whether
directed toward children's imaginative play or creativity, suggest possible
influences which warrant further investigation. If an adult model is
important for imaginative play, as the Singers suggest (Note 3), perhaps
that adult would have to be comfortable with fantasy and make-believe.
Adults who engaged in imaginative play themselves when young, and who
enjoy fantasy in the present, may encourage more imaginative play for
their children. Perhaps these are the same adults who, as Bishop and
Chace (1971) propose, can refrain from too heavy-handed a management of
the child's activity and insistence upon conventional forms of play,
thereby providing a home environment which is conducive to fantasy and
make-believe. These are the chief considerations which guided the
present study.


'ty opinion it
r.m.y and is fully
* 'V tr :ooree of
V./t/of the College
Kt! as cartial
/y of Philosophy.
m¡


TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS
ABSTRACT v
CHAPTER I INTRODUCTION AND STATEMENT OF PROBLEM 1
CHAPTER II REVIEW OF THE LITERATURE 5
CHAPTER III DESIGN OF STUDY 15
CHAPTER IV RESULTS 36
CHAPTER V DISCUSSION 53
APPENDICES 64
APPENDIX A Directions for Testing 65
APPENDIX B Singer's Structural Fantasy Pre-disposition
Interview (SFPI) 66
APPENDIX C 0. J. Harvey Questionnaire 67
APPENDIX D Home-Play Questionnaire 68
APPENDIX E Interaction 70
APPENDIX F Demographic Information 72
APPENDIX G Variables for Correlation Matrix 73
APPENDIX H Letter to Parents Requesting Consent 76
APPENDIX I Consent Form 77
REFERENCE LIST 78
REFERENCE NOTES 81
BIOGRAPHICAL SKETCH 82
TV


24
TABLE 4
SCORING CRITERIA FOR FANTASY GROUP ASSIGNMENT
Fantasy Pre-disposition Tests
Group Assignment
Barron
SFPI
Tel1-a-Story
Low Fantasy
Pre-disposition
19-28
0, 1 or 2
0, 1, 2, 3 or 4
High Fantasy
Pre-disposition
1-18
3 or 4
5 and above
(maximum = 10)
Consent was obtained from the 17 families of HF children and the 21
families of LF children for home interviews and testing at a time of family
convenience. Except in one instance, all visits were scheduled and com
pleted when the father could be home. All home interviews and testing were
accomplished by the principal investigator and one or two of the student
assistants. Mothers and fathers were separated in order to obtain the
various measures, with examiners taking turns interviewing a mother one
time and a father the next. The principal investigator did all of the
mother-child Interaction observations. The second student assistant, on
occasion, tested siblings and accomplished reliability checks for the
Interaction observations (see reliability section).
Because screening of the children had taken place in November,
December and January, because so many children had been tested and so many
protocols scored and because interviewing of families did not begin until
April, it was possible to interview families "blind." By the time the
children had been assigned to the HF or LF category, they were just
"names." The names were transcribed on to index cards from the category
lists by a secretary, together with address and telephone number. Contacts




27
and were also completed by each parent independently.
4. Interaction: Mother-Child play activity (Appendix E). Having
been previously informed that the principal investigator would like to
observe them playing together, the mother and child were requested to sit
together at a table or on the floor. The examiner emptied a cannister
of previously selected toys, saying, "Here are lots of little toys. Both
of you, playing together, build an exciting scene like for a movie
. and then I will ask you to tell me what is happening." (Adapted
from Erikson, 1972)
The toys were divided into two categories, "realistic" and "fantasy,"
with each realistic toy having its fantasy counterpart.
Realistic Toys
Fantasy Toys
1.
Mother
1.
Similar-sized, milk bottle
shaped wooden figure.
2.
Father
2.
Similar-si zed, milk bottle
shaped, wooden figure.
3.
Baby
3.
Similar-sized, milk bottle
shaped, wooden figure.
4.
Rhinocerus
4.
Green monster thing.
5.
Lion
5.
Red thing.
6.
Dog
6.
"Dumbo" elephant, sitting and
smiling.
7.
Toy furniture
7.
Blocks of assorted shapes.
Identical number of items as
furniture.
8.
Toy car
8.
Thing on wheels
9.
Policeman figure
9.
Superman figure
In addition there was a set of colored pipe cleaners and a set of
plain popsickle sticks. These were scored as "fantasy" items.


Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
INDIVIDUAL AND FAMILIAL CORRELATES OF
CHILDREN'S FANTASY PLAY
By
Lorraine Bradt Dennis
March, 1976
Chairman: Ira J. Gordon
Major Department: Foundations of Education
The purpose of this study was an attempt to discover individual and
familial correlates of children's imaginative play.
Two hundred eighty-one children, representing all those for whom
informed consent had been obtained (61.62% of the total number of children
in the kindergarten and first grades of the Portsmouth, Rhode Island
school system) were tested with the Peabody Picture Vocabulary Test and
three measures of fantasy pre-disposition: The Barron Movement Threshold
Inkblots, the Singer Structured Fantasy Pre-disposition Interview (SFPI)
and the Tell-a-Story Transcendence Index. Children with PPVT scores be
low 100 were omitted from the fantasy pre-disposition sample. A median
split for each fantasy pre-disposition measure with scoring criterion of
rank above or below the median on each of the three tests enabled
identification of 25 high fantasy pre-disposition (HF) and 55 low fantasy
pre-disposition (LF) children from the remaining 221.
v


8
own turf" can their behavior be truly sampled.
Play as Influenced by Instruction
Though reports of socio-economic differences in play may be open to
question, the possibility of teaching disadvantaged children to play pre
tend games is better substantiated. Smilansky's lead (1969) in documenting
the results of teaching socio-dramatic play to certain children was fol
lowed by similar efforts by Saltz and Johnson (1973) in Detroit and Frey-
berg (1973) in New York City. Each study reports significant gains in the
amount of symbolic play subsequently initiated by the trained children,
together with increases in language production, vocabulary range, verbal
fluency and ability to sustain attention as well as, for Saltz and Johnson,
significant increases in several retention abilities. Each of these
studies maintained appropriate controls for pleasant interaction with
adults so that the gain can be attributed to the imaginative play rather
than to simply the attention and interest of an adult. Persistence of
the gains over a long period of time has not been assessed. Specific
definition of the make-believe play differs somewhat from study to study,
but each definition includes the presence of a theme, roles, objects and
elements which are not present in reality.
The Russians, too, seem to be concerned with the role of the adult
teacher in fostering imaginative play (Elkonin, 1971), but there is no
hint of worry that some children may be deficient in make-believe produc
tions. Rather, it is believed that without adults to organize and guide
this activity, the imaginative play of the child might become a kind of
idle and self-indulgent dallying (Repina, 1971). Adults are to serve as
models for role playing, first mediating the use of playthings, then
leading the child into pretending with objects which are used as substitutes


19
members of sets of twins.
The average IQ was 121.16 (SD = 10.12).
Mothers of these children had a mean age of 31.55 years (SD = 4.80)
while the average age of the fathers was 33.87 years (SD = 4.90).
The educational level of the parents was high. Mothers averaged
14.32 years of schooling, a little more than two post-secondary years
(SD = 1.76) while fathers had a mean of 15.97 years, only slightly less
than four years beyond high school (SD = 1.94).
Two families in the sample lived in apartments. The remainder lived
in single family homes. The average number of rooms per household was
7.21 (SD = 1.56).
No family in this sample consisted of other than immediate family
members. Average family size was 4.34 persons (SD = .71).
Occupations of fathers are listed in Table 2 and the occupational
training of mothers is reported in Table 3.


31
seemed to her to be LF protocols. Twenty-two of her HF group corresponded
to those selected by the investigation procedures, three had been "missed"
and two of hers did not correspond to ours. She had selected 60 LF pro
tocols; these included all 55 of those previously selected in the investi
gation and five which did not correspond (Table 5).
This means that there were ten instances of non-correspondence out of
the total of 221 protocols, an agreement rate of 95.5%.
TABLE 5
RELIABILITY OF SCORING FOR FANTASY PRE-DISPOSITION MEASURES
Number of
HF
Number of
LF
Number
Remaining
Total
Identified by
Scoring Procedures
25
55
141
221
Identified correctly
for Reliability
22
55
134
211
Mis-identified for
Reliability
3
7
10
Agreement Rate = 95.47%
3. Reliability of
scoring for
0. J. Harvey
"This I Believe"
protocols
concrete-abstract conceptual systems. The final sample of 38 sets of
parents yielded 76 "This I Believe about ..." protocols which were
scored according to the Harvey (Note 5)scoring criteria as previously
described. An independent judge, a Ph.D. candidate in child development
at Purdue University, familiar with the Harvey scoring, scored 74 of the
protocols, assigning each into one of 4 systems. Agreement with the in
vestigation scoring was disappointing; 40 protocols corresponded, but
34 did not, a 54.05% agreement.


32
Therefore, as previously noted, System 1 and System 2 scores were
pooled to form a "concrete" category, while System 3 and System 4 scores,
grouped together, formed the "abstract" category.
A second inspection of the judge's scores revealed that the broader
categorization resulted in much better agreement with 61 of the 74 pro
tocols corresponding to the newly assigned category of "abstract" or
"concrete." This, therefore, yields 82.43% agreement.
4. Reliability of the fantasy pre-disposition measures. During the
process of visiting the families, it was possible to re-test 16 of the
children with the fantasy pre-disposition measures. Testing of all the
children had been completed in January, family visits began in late April
so at least three months had elapsed between the original testing and the
re-test.
Because examiners went into the household "blind," selection of
children for re-testing was circumstantial. If it proved possible to
separate the child from the parents' testing, so that parents' responses
would not influence the child's responses, a re-test might be accomplished.
Sometimes children were playing outside while parents were interviewed
and could be called in later. In most instances, however, the child
wanted to listen to the parent's interview and had to be considered con
taminated for a possible re-test. Subsequent inspection of the re-tests
revealed that six of the children had previously been categorized as HF
and 10 had been assigned to the LF group.
Scores on each of the three measures were correlated with the
original scores obtained by the child on that measure to obtain a cor
relation coefficient for each of the fantasy pre-disposition tests.
Results are tabulated in Table 6.


35
limited in terms of numbers, region of the country, socio-economic status
and race. No claims are made for representativeness; therefore, extensive
inferential statistical procedures were not employed.


71
Interaction
Child M
Mother figure
Father figure
baby
toy furniture
car
rhi no
lion
horse
policeman
"mother"
"father"
"baby"
blocks
"car"
green thing
red thing
Dumbo
Superman
pipe cleaners
popsickle sticks


Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
INDIVIDUAL AND FAMILIAL CORRELATES OF
CHILDREN'S FANTASY PLAY
By
Lorraine Bradt Dennis
March, 1976
Chairman: Ira J. Gordon
Major Department: Foundations of Education
The purpose of this study was an attempt to discover individual and
familial correlates of children's imaginative play.
Two hundred eighty-one children, representing all those for whom
informed consent had been obtained (61.62% of the total number of children
in the kindergarten and first grades of the Portsmouth, Rhode Island
school system) were tested with the Peabody Picture Vocabulary Test and
three measures of fantasy pre-disposition: The Barron Movement Threshold
Inkblots, the Singer Structured Fantasy Pre-disposition Interview (SFPI)
and the Tell-a-Story Transcendence Index. Children with PPVT scores be
low 100 were omitted from the fantasy pre-disposition sample. A median
split for each fantasy pre-disposition measure with scoring criterion of
rank above or below the median on each of the three tests enabled
identification of 25 high fantasy pre-disposition (HF) and 55 low fantasy
pre-disposition (LF) children from the remaining 221.
v


Seventeen intact families of HF children and 21 intact families of
LF children consented to participate. These were middle-class families
with at least one parent in every family having at least one year of
college.
Parents, individually, were administered the three fantasy pre
disposition measures, a test for abstract versus concrete conceptual
style and a questionnaire regarding home-play environment. In addition
there was a short demographic questionnaire and the mother and child were
engaged in a play activity termed "Interaction." Siblings over the age
of four were tested with the three fantasy pre-disposition measures.
The mother's Barron and SFPI scores were significantly correlated
with the child's SFPI score. The mother's SFPI score was also signifi
cantly correlated with the child's Tell-a-Story score. Neither fathers'
nor siblings' scores on the fantasy pre-disposition measures correlated
with those of the children.
The number of restrictive responses of the mother with regard to her
attitudes toward children's play in the home was significantly and nega
tively correlated with the fantasy pre-disposition of the child. Fathers'
restrictive responses were not significant.
The score on the Interaction play activity between mother and child
was significantly correlated with the fantasy pre-disposition of the
child.
The number of years of school of the mother, but not the father,
correlated significantly with the child's fantasy pre-disposition.
The child's PPVT IQ was significantly correlated with the fantasy
pre-disposition, despite the fact that all IQs below 100 had been
eliminated from the selection sample.
vi


61
and sibs enjoyed playing with the toys later, while examiners were busy
with parents.
Finally, it should be noted that the research into children's play
had great fascination for parents. Parents were charmed with being in
vited to discuss their own preferred childhood activities (for the SFPI)
and were eager to compare notes after separate interviews. It appeared
that few parents had described their childhood fun to each other, and the
occasion seemed to open new areas for communication. Often, one or the
other would exclaim, "You never told me that!" This might offer a fruit
ful direction for further research, especially if extended into present
activities. What do adults do now for "play?" What kinds of childhood
activities are carried forward into adult life? Needed also is an
assessment of the role of television in both children's play and adult
leisure-time pleasure.
Conclusions
Play, the ubiquitous childhood activity, has until recently been
little researched. One of the difficulties is that an operational defin
ition must almost certainly invoke the notion of spontaneous, freely
chosen engagement; therefore when the scientist attempts to quantify it
or remove it to his laboratory or circumscribe it for closer scrutiny,
he tends to violate its most salient characteristic. It is, of course,
the old problem of "observer effect."
This investigation has attempted to check out some previously used
instruments and suggestive findings of other researchers and to test some
hunches about background influences on fantasy play. The results were
modest. Supported was an indication of mother's influence in terms of


17
on measures of fantasy pre-disposition and an "Interaction" score obtained
when mother and child are playing together with toys which can be used
either imaginatively or realistically.
This final hypothesis was obtained from the suggestions of Lytton
(1972) and Singer (1973) regarding the importance of the kinds of activities
in which parent and child engage together.
Method
Subjects
Informed parental consent was obtained to include 281 kindergarten
and first grade children in Portsmouth, Rhode Island for fantasy pre
disposition screening. This represents 61.62% of all kindergarten and
first grade children enrolled in the Portsmouth school district at the
time of testing. Consent was not obtained or was refused for the remaining
38.38%.
Sixty of the children were eliminated from the sample because they
scored below 100 on the Peabody Picture Vocabulary Test. This was an
attempt to avoid confounding fantasy pre-disposition scores with intelli
gence and verbal production. The remaining 221 children constituted the
sample from which the subjects of the investigation were selected.
From their 221 protocols, according to a procedure which will be
described, a group of 25 high fantasy pre-disposition (HF) and 55 low
fantasy pre-disposition (LF) children were identified.
Families of these children were then contacted by telephone and
invited to participate in the second phase of the study. Table 1 details
the attrition of a number of potential families, some because of refusal
and some because the family had moved before the contact could be made.


16
toward fantasy play and parents' tendency toward an abstract rather than
concrete conceptual system as measured by Harvey et_ aj_. (1961) scale.
This hypothesis was suggested by the findings of Bishop and Chace
(1971) with regard to parents of creative children who are characterized
by these investigators as encouraging "truly playful play" (p. 321).
3. There will be a positive relationship between children's
imaginative play and a less restrictive play environment in the home.
This hypothesis, also, borrowed from the findings of Bishop and Chace
(1971) concerning the type of home play environment provided for more
creative children.
4. There will be a positive relationship between scores of the
children on the fantasy pre-disposition measures and those of their
siblings who are age four or older on the same measures.
The rationale for this hypothesis suggested that if parents who are
characterized by abstract conceptual systems and who provide a home play
environment which is relatively free of adult constraints do, indeed,
enable more imaginative play, the environment should operate similarly
for other children in the family.
5. The homes of children with high fantasy pre-disposition will
have significantly more room for privacy. Furthermore, more mothers of
these children will be working mothers.
This two-part hypothesis was designed to follow up the findings of
Freyberg (1973), Dewing and Taft (1973) and the suggestions of Singer
(1973) regarding the need for privacy and for a certain amount of distance
between parent and child, presumably obtained when the mother is occupied
with employment outside the home.
6. There will be a positive relationship between children's scores


3
(Note 3) addressed the conference on the importance of an adult model for
fostering children's play. They were primarily concerned with the problem
of encouraging nursery school teachers to interact imaginatively with the
children, suggesting that without such interaction the amount of imaginative
play among children decreases, and that the adult interaction seems to have
an eliciting effect, providing the child license to play imaginatively.
Such adult influence on children's imaginative play should also be
important in the setting of the home. Though studies of parental contri
butions to child behavior have been criticized for methodological reasons
(Yarrow, 1963) for dearth of outcomes (Caldwell and Hersler, 1964) and for
conceptual flaws (Bell, 1968, 1971), there is continuing interest in parent
attitudes, practices and general home conditions as major factors in
several aspects of child development (Rebelsky, 1974; Baumrind, 1971).
Of all of the developmental characteristics, that of symbolic functions
seems most securely tied, empirically, to beginning influences, with both
language and concept formation tending to thrive or falter in terms of
early environmental variables (Gordon, 1969; Hess and Shipman, 1965).
Since imaginative play is characterized by the use of symbols (indeed, is
termed symbolic play by Piaget) it seems reasonable to suggest that it,
too, bears the stamp of the especial confluence of family circumstances,
attitudes and practices. However, little is known about the home back
ground factors which might relate to children's imaginative play.
The present investigation was designed to study parent attitudes,
practices and certain home conditions which might be correlated with
children's imaginative play. Since the study of play can be said to have
scarcely begun, it seems important to attempt to obtain data which relates
to the child's primary environment, the site of first play experiences.


ACKNOWLEDGMENTS
With humility and deepest gratitude, my indebtedness to a large numbe
of individuals must be acknowledged. Without their efforts and good will,
this dissertation truly could not have been accomplished.
I need, especially, to thank Dr. Ira J. Gordon, the chairman of my
committee, for his incisive guidance, his wise instruction and for his
unfailing calm and cheerful support during the dark moments which are
part of the life of every graduate student.
Dr. Robert Jester and Dr. Barry Lester, members of my committee, are
also to be thanked for their contributions to the design of this project,
and their constructive criticisms which enhanced its execution.
Dr. Jerome L. Singer's interest, his loan of the Barron Inkblots and
his help in obtaining the services of one of his graduate students,
Claudia Morelia, for a reliability study places me in his debt.
Next, I must cite, fondly, the six students who assisted in the
data gathering and scoring of protocols from beginning to end, giving
tirelessly of their time and good thinking: Marilyn Packard-Luther,
Elise Carlson, Ann Leaver, Mary Jane Scott, Don Whitworth and Scott Hicks.
Beverly Malin, Philip Goodwin, Claudia Morelia, Tullio Pitassi and
Elizabeth Hallenbeck all assisted with reliability studies and, in the
process, offered many ideas for improving our techniques.
A special note of thanks goes to my son, Brian Dennis, whose
assistance in coding and computerizing the data was invaluable. Stephen
Sledjeski took time from his busy life to program the multiple regression
equations.


21
TABLE 3
MOTHERS' OCCUPATIONAL TRAINING AND COLLEGE WORK
Groups
Frequency
Occupational Training
Low Fantasy
2
Registered Nurse
4
High School Teacher
1
B. A. Degree
1
B. S. Degree
5
At least one year college
Total
13
High Fantasy
8
Registered Nurse
2
Elementary Teacher
1
Speech Therapy
1
B. S. Degree
5
At least one year college
Total
17
Procedure
The 281 children of the initial sample were tested with the Peabody
Picture Vocabulary Test and with three fantasy pre-disposition measures
by the principal investigator and six assistants, four specially-selected,
outstanding undergraduate psychology students and two graduates of the
psychology program at Roger Williams College, one of whom holds a master's
degree in child development from the University of Rhode Island. All had
received training in the use of the testing devices, had administered the
tests to a number of children outside the study and were individually
supervised by the principal investigator during the first three weeks of
testing. Reliability data appear in a subsequent section.
The three measures of fantasy pre-disposition were:
1. The Barron (1955) Movement Threshold Inkblots. This is a series
of 28 inkblots, arranged in a series so that it becomes progressively
easier to see human movement (M), which has been found effective in use


REFERENCE LIST
Barron, F. Threshold for the Perception of Human Movement in Inkblots.
Journal of Consulting Psychology, 1955, J9_, 33-39.
Baumrind, D. Current Patterns of Parental Authority. Developmental
Psychology Monographs, 1971, 4_, 1-103.
Bell, R. Q. A Reinterpretation of the Direction of Effects in Studies of
Socialization. Psychology Review, 1968, 7J5, 81-95.
Bell, R. Q. Stimulus Control of Parent or Caretaker Behavior by Offspring.
Developmental Psychology, 1971, 4_, 63-72.
Bishop, D., & Chace, C. Parental Conceptual Systems, Home Play Environ
ment and Potential Creativity in Children. Journal of Experimental
Child Psychology, 1971 12., 318-338.
Caldwell, B., & Hersler, L. Mother-Child Interaction During the First
Year of Life. Merri 11-Palmer Quarterly 1964, JO^, 119-128.
DeVore, I. Mother-Infant Relations in Free-ranging Baboons. In H. L.
Rheingold (Ed.), Maternal behavior in mammals. New York: Wiley,
1963.
Dewing, K. & Taft, Some Characteristics of Parents of Creative Twelve
Year Olds. Journal of Personality, 1973, 41_, 71-84.
Eifermann, R. Social Play in Childhood. In R. E. Herron & B. Sutton-Smith
(Eds.), Child's Play. New York: Wiley, 1971.
Elkonin, D. Symbolic Representation and its Functions in the Play of
Children. In R. E. Herron & B. Sutton-Smith (Eds.), Child's Play.
New York: Wiley, 1971.
Erikson, E. Play and Actuality in Development. In M. Piers (Ed.), Play
and development. New York: W. W. Norton, 1972.
Freyberg, J. T. Increasing the Imaginative Play of Urban Disadvantaged
Kindergarten Children Through Systematic Training. In J. L. Singer,
The child's world of make believe. New York: Academic Press, 1973.
Gilmore, J. B. Play, a Social Behavior. In R. Haber (Ed.), Current
research in motivation. New York: Holt, Rinehart and Winston, 1966.
78


62
her own fantasy pre-disposition, her years of school, her interaction with
the child in a play activity and her attitudes about constraints on play
in the home. Unsupported was the possibility of a difference in cognitive
style of parents as influencing fantasy play or substantial influence from
fathers' attitudes or similarity of siblings with regard to fantasy pre
disposition. Although the IQ of the child seems to be associated with
level of make-believe, the sex of the child, ordinal position or number
of rooms in the household does not.
Obscured in scores and tests of significance, however, are the many
delightful incidents which, unscientifically, have served as reminders
of the elusive qualities of play.
There was the tiny kindergartener who, upon being escorted back to
her room after testing, stopped in the hall and exclaimed "Oh, I have to
go back. I forgot to say goodbye." The student examiner was puzzled.
"Yes," responded the small lass solemnly. "The clown and the monkey.
They were watching us all the time." The student examiner reporting
this was still a little shaken!
There was the little boy with an imaginary companion named "Jethro"
who "wakes me up at night to get him a drink of water." And another
whose imaginary friend is named "abdominal snow man." The spelling is
correct.
There was a young father with eyes alight as he described making
fireworks in his basement. He and his own father (the childrens'
grandpa) have "wars with castles" whenever the mother is not around to
object. They make castles out of blocks at opposite ends of the living
room and catapult little rockets at each other's construction. The
three small girls wear hard hats and become "knights," retrieving and


39
school (SD = 1.22) as contrasted with the mean of 13.86 years (SD = 2.01)
for mothers of LF children. The difference is significant at p = .001.
Findings for the Hypotheses
Tables 8, 9, and 10 report results of analysis of scores pertinent to
the hypotheses. The correlation matrix for all scores can be found in
Appendix G.
1. The first hypothesis stated the prediction of a positive relation
ship between children's scores on the measures of fantasy play and their
parent's scores on the same measures.
The hypothesis was only partially supported and only for certain of
the mothers' scores.
It can be seen (Table 8) that the mothers' Barron scores and the SFPI
are significantly different from LF to HF group. Mothers' Tell-a-Story
scores, however, do not discriminate.
Fathers' scores on the three fantasy pre-disposition measures do not
differentiate between LF and HF groups.
Correlation coefficients between parents' and children's scores are
presented in Table 9 as abstracted from the complete correlation matrix
(Appendix G).
There are no significant correlations between fathers' fantasy pre
disposition scores and those of the children.
Significant correlations occur between the child's SFPI and two of
the mothers' tests, the Barron and the SFPI. Mothers' SFPI is also
significantly correlated with children's Tel1-a-Story.
2. The second hypothesis stated the prediction that parents of HF
children would tend to be rated as "abstract" with regard to conceptual




40
TABLE 8
MEAN SCORES FOR FANTASY PRE-DISPOSITION MEASURES
Test
Child Score Mother Score Father Score
L.F. H.F. L.F. H.F. L.F. H.F.
Barron
S.D.
t Test
24.571
3.22
9.375**
12.706
5.57
17.952
5.71
2.367*
14.529
6.28
14.524
5.98
N.S.
15.824
4.99
SFPI
S.D.
t Test
1.429
.746
-16.327**
3.353
.493
1.762
1.04
-2.303*
2.412
1.23
1.667
1.35
N.S.
1.412
1.12
Tell-a-Story
S.D.
t Test
1.430
1.33
-13.780**
7.765
2.02
4.290
2.47
N.S.
5.294
2.47
4.714
1.74
N.S.
5.00
2.60
*p<.02
**p<.0001


52
In the light of the two equations, it is possible to suggest the
influence of four variables on the amount of child's make-believe play
as measured by the SFPI. Two of these, the child's IQ and the mother-
child Interaction, were significant predictors for both equations. Each
of the two remaining, mothers' restrictives and mothers' years of schooling,
emerged as significant for one equation. It would appear that use of one
as a predictor might preclude use of the other; the substantial inter
correlation suggests shared variance which would necessitate testing
beyond the scope of this study.
Summary of Findings
To summarize the individual and familial correlates of children's
make-believe play which have been suggested by this investigation: the
child's vocabulary-IQ seems to be the most important individual charac
teristic associated with level of fantasy pre-disposition for children
in this study.
A relationship was found for two measures of mothers' fantasy pre
disposition, the Barron and SFPI, and children's SFPI. The mothers'
years of schooling is correlated with children's scores on the Barron
and SFPI. The number of restrictive responses made by the mother with
regard to home-play environment is negatively associated with the amount
of make-believe of the child's SFPI.
Finally, the amount of fantasy in the Interaction score, for which
the mother and child engaged in play together, is significantly correlated
with all three of the children's fantasy pre-disposition measures.


49
Multiple Regression Analysis
The child measure which achieved the largest number of significant
correlations with other variables is the SFPI (Table 16). It is also
the most stable measure with a high degree of reliability on re-test
(r = .750, p = .001). It was selected as the dependent variable for
multiple regression analysis.
TABLE 16
SIGNIFICANT CORRELATIONS WITH CHILD'S BARRON, SFPI AND TELL-A-STORY
(ABSTRACTED FROM COMPLETE CORRELATION MATRIX, APPENDIX G)
Child's
Barron
Child's
SFPI
Child's
Tell-a-Story
Barron (child)
1.000
-.6167
-.8346
SFPI (child)
-.6167
1.000
.7464
Tell-a-Story (child)
-.8346
.7464
1.000
Vocabulary-IQ (child)
-.4888
.4930
.5474
Interaction (mother-child)
-.3798
.4135
.5474
Mother's years of school
-.3538
.3695
Mother's SFPI
.3695
.3270
Mother's Barron
-.3898
Mother's Restrictives
-.4195
Variables relating significantly to the child's SFPI are: the child's
IQ, the Interaction score, the mother's years of school, the mother's
SFPI, and the mother's number of restrictive responses for home-play
environment. The father's views on whether make-believe play should be
encouraged distinguished significantly between LF and HF groups (Table
17); this variable, termed "father encourage" was also entered into the
analysis.
A stepwise multiple regression analysis employing these variables
as predictors was performed. The summary table is reproduced for Table
18.


26
scoring, using an outside observer, was unsatisfactory (detailed in the
section on reliability) and because the interest for this investigation
lay in terms of the concrete versus abstract direction on the continuum
conceptualized by Harvey (et al_., 1961), System 1 and System 2 scores
were pooled in order to assign a "concrete" category to parents with
those scores, while System 3 and System 4 scores were designated as be
longing to the "abstract" category. Justification for this procedure
obtains both from the fact that emphasis for this study was upon the
global differences between concrete and abstract orientations of parents,
rather than the qualitative differences among the four systems, and that
the contrast between the two categories proved to be much more reliable
(again, see the section on reliability studies) than assignment of a
System score.
3. Home Play Environment: Parent Permissives and Restrictives.
Questions concerned with parent attitudes toward children's play were
selected from the original questionnaire developed by Bishop and Chace
(1971), utilizing those questions which had successfully discriminated
between restrictive and permissive parents in their investigation
(Appendix D). The questionnaire is designed with multiple alternative
answers plus a space for write-in answer for use if the parent decides
that none of the available answers are suitable. One of the alternatives
serves as a criterion response indicating restrictiveness and one is a
criterion response for permissiveness. The remaining answers and the
write-in response represent varying degrees of restriction or permission
-- that is, each contained some contingencies or disclaimers. This en
abled a count of the number of wholly restrictive or wholly permissive
responses of each parent. These questionnaires contained eight questions


The abstract versus concrete conceptual style of the parent, mother
or father, did not differentiate between LF and HF children.
No significant differences between LF and HF groups were obtained in
terms of sex, ordinal position, grade in school, age of mother, age of
father, outside employment of mother, number of rooms in the home or
number of persons in the family.
Individual and familial correlates of children's fantasy play sug
gested by this study are the child's verbal IQ, mother's scores on two of
the fantasy pre-disposition measures, her lack of restrictiveness with
regard to play in the home, the number of her years in school and her
inclination to engage the child in play which contains fantasy elements.
j
The small number of subjects in relation to the number of variables con
sidered necessitates caution in interpreting findings.
vn


37
TABLE 7
DESCRIPTION OF L.F. AND H.F. GROUPS: TESTS OF DIFFERENCES
Variable
Low Fantasy
High Fantasy
Test of Significance
Sex
7 girls
9 girls
Corrected Chi
square
14 boys
8 boys
(1) = .782 NS
Ordinal Position
First Born
10
10
Corrected Chi
square
"only"
0
2
(1) = 1.20 NS
twi n
1
1
Grade
Kindergarten
14
11
Corrected Chi
square
First Grade
7
6
(1) = .005 NS
Peabody Picture
Vocabulary IQ
Range
104-140
106-145
t Test (36) =
-3.86**
Mean
117.27
125.94
S.D.
Mother's Age
8.87
9.73
Range
25-40 years
27-45 years
t Test (36) =
-.620 NS
Mean
31.19
32
S.D.
Father's Age
4.05
5.71
Range
27-43 years
26-46 years
t Test (36) =
-1.53 NS
Mean
33
34.94
S.D.
Mother's Years
of School
4.27
5.54
Range
10-17 years
13-17 years
t Test (36) =
-3.48*
Mean
13.86
14.88
S.D.
Father's Years
of School
2.01
1.22
Range
13-19 years
12-21 years
t Test (36) =
-.909 NS
Mean
15.76
16.24
S.D.
Mother's outside
employment
1.64
2.28
Not employed
outside home
13
11
Corrected Chi
square
Part-time
6
6
(1) = .0026 NS
Ful 1-time
2
0
Number of persons
in family
Range
4-6 persons
3-6 persons
t Test (36) =
1.79 NS
Mean
4.48
4.18
S.D.
.68
.73


25
with the family were made using the card; examiners did not inform them
selves of the category of the children until interviews were completed.
By means of the home visits, the following tests, Interaction
activity, questionnaires and interviews were accomplished:
1. Fantasy Pre-disposition measures. Mothers, fathers and siblings
of age four and over were tested with the three fantasy pre-disposition
measures previously employed to screen the original children: Barron,
Singer Structured Fantasy Pre-disposition Interview (SFPI), and the Tell -
a-Story. Parents were asked to respond to the SFPI in terms of the
kind of play they recalled from their own childhood. These measures were
scored by two individuals associated with the study, independently, with
a subsequent consensus if disagreement existed, exactly as were the
children's, previously. The method of scoring had achieved a high degree
of reliability (see section on reliability).
2. Abstract-Concrete conceptual system. 0. J. Harvey's "This I
Believe" test (1970) of conceptual level (Appendix C) was used to determine
the conceptual system of the parent. Parents responded independently,
writing responses to the open-ended statements with ten referents, six
of which had previously been used by Harvey and for which detailed scoring
criteria are described (Note 5). The four referents added for this study
were: work, women's rights, law and education. Each referent is pre
ceded by the statement "This I believe about Responses
were timed, two minutes allowed for each response.
Each parent was categorized into one of four systems by two judges,
scoring the protocols independently and arriving at a consensus with a
third judge in case of disagreement. The principal investigator and
student assistants served as judges. Because the reliability check on


18
Single parent families and those with father absent because of sea-duty
were omitted in order to adhere to the criterion of "intact families."
Two black families were also omitted in order to maintain a homogeneous
sample in terms of race. Two families turned out to be Portuguese im
migrants, with some parental English-language deficiency; both were
eliminated lest the child's LF score chiefly reflect the bilingual back
ground. Finally, several families either had no phone or an unlisted
number which could not be obtained, so that it was impossible to establish
contact.
TABLE 1
ATTRITION RECORD OF POTENTIAL SUBJECT FAMILIES
Reason for Elimination
Potential HF
Families
(N=25)
Number Eliminated
Potential LF
Fami1 ies
(N = 5 5)
Number Eliminated
Refused
3
12
Moved
3
7
Single Parent
1
5
No Phone or Unlisted Number
1
4
Minority Family
2
Portuguese Language
2
Sib of Other Child Already in Study
2
Total Eliminated
8
34
Remaining for Sample
17
21
Data, accordingly, were collected from 17 HF and 21 LF children,
their parents and their siblings over the age of four.
The subjects consisted of 16 girls and 22 boys, 25 of whom were
kindergarteners and 13 in the first grade, and their families. Twenty
of the children were first born, two were only children and two were


CHAPTER III
DESIGN OF STUDY
The Problems and Hypotheses
The problem which the present study was designed to address is an
almost total lack of information regarding family background factors which
might encourage or discourage children's imaginative play.
Hypotheses were derived from the very few investigations to date,
some of which were focused on creative children rather than children who
engage in high levels of fantasy play. However, it seemed reasonable to
assume, in the absence of data, that the imagination necessary for
creativity and the imagination necessary for fantasy play might require
similar facilitating environments, as Bishop and Chace (1971) have
suggested.
Two groups of children, pre-selected for high or low fantasy pre
disposition, and their families were used to test the following hypotheses
1. There will be a positive relationship between children's scores
on the fantasy pre-disposition measures and their parents' scores on the
same measures.
This hypothesis was suggested by the findings of Freyberg (1973) and
Manosevitz et aj_. (1973) of more encouragement for imagination and fantasy
among parents of children who engage in make-believe play and who have
imaginary companions.
2. There will be a positive relationship between children's tendency
15


REFERENCE NOTES
1. Ellis, M. Play, a paradox for teacher and scientist. Paper presented
at the Second Annual Symposium on Play, Atlanta, Georgia, April,
1974.
2. Kempler, B. Welcome and introductory remarks. Paper presented at
the Second Annual Symposium on Play, Atlanta, Georgia, April,
1974.
3. Singer, D., & Singer, J. L. Television, adult modeling and physical
structure of the setting as factors enhancing play in pre-school
children. Paper presented at the Second Annual Symposium on
Play, Atlanta, Georgia, April, 1974.
4. Branch, A. Personal communication. December, 1973.
5. Harvey, 0. J. "This I Believe" Scoring Instructions, 6/71 and 9/71.
Mimeographed handout at address by 0. J. Harvey, University of
Florida, January, 1972.
81


13
The concrete end of the continuum (System 1), with its simplistic view
of the world, stereotyped and polarized in terms of absolutes, its small
capacity to "decenter" or take the view of another, together with an in
ordinate need for structure and rules, is contrasted with the abstract
end (System 4) which accepts the world as complex and varied, is open to
new ideas and tolerant of ambiguity, views rules as utilitarian, but has
no need for reliance on authority or rule for rule's sake.
Reasoning that abstract parents would better enhance children's play
by refraining from imposing conventional standards, by willingness to accept
novel (even bizarre) ideas and willingness to live with a certain amount of
disorder, Bishop and Chace (1971) measured parents' conceptual systems on
Harvey's concrete-abstract continuum. The category of the parent was then
related to parental responses on a home-play environment questionnaire
which inquired into attitudes and the kinds of conditions maintained for
play, especially in terms of the restrictions imposed on play activities.
Although no significant differences were found for fathers in the study,
the outcome for mothers revealed that abstract mothers provided home play
environments which were most free of constraints and that the children of
these mothers were the most potentially creative. Potential creativity
was measured by indicators of complexity and variety of performance on a
laboratory play task.
The investigations of possible home influences on imaginative play
are exceedingly few in number. Only slightly more numerous are studies
of familial correlates of children's creativity. The latter were of
interest because of several studies which have indicated a link between
play, fantasy and creativity. Lieberman (1965) established significant
correlations between "playfulness" and three measures of divergent thinking


48
TABLE 15
SIGNIFICANT CORRELATIONS WITH INTERACTION SCORE
Interaction
Child's Barron -.380
Child's SFPI .414
Child's Tel1-a-Story .520
Father Encourage .374 (obtained from multiple regression
correlation matrix)
To summarize the findings in regard to the hypotheses for this study:
The prediction that children and parents would resemble each other with
regard to fantasy pre-disposition scores was only partially substantiated
for only two of the mothers' scores.
The second prediction which suggested that parents of HF children
would tend to be categorized as having an abstract conceptual system was
not upheld.
The third prediction of a less restrictive home-play environment for
HF children was supported in terms of mothers' but not fathers' responses.
The fourth hypothesis stated the suggestion that brothers and sisters
of the children in the sample would resemble their LF or HF sib. This
hypothesis did not receive any support at all.
The fifth hypothesis stated the prediction of more privacy in terms
of a higher room/person ratio and more mothers employed outside the home
for HF children. Neither condition obtained and the hypothesis was not
confirmed.
The sixth prediction of more imaginative play-interaction between
mother and child for HF children was substantially supported.


k 4 -W V
4 m iT CC l^l>'a^C>INaCa^ fKfTNCs-CiT'rtTcrocf'cr' -4 e p- r
cc.4p-CCcrar^ifrs.n>£f\ir r^*' nC^'Ci*
C rv C *r c C C C r- <40 C rr* rv C
cecccccccococccecccccoccc cc
I lili I II III
Ac
C O fn c ff o- C 4 <\
OPgiSih-P-C t-
oc o c o c c
I I I I I
V
%r r c if- p- x cr c p-
ccaiaafN-Cf'
ar >f c ^ it 4)
p-r~ n r
coccc occc
I I I I I I I I I
At
p- p-
C -4 P- C f cc
oc (7 fr a C ^ < 0 r
p fv r c r c
ccccccccoc
I II II
* -M. fc 4, 4r
j I m if ir C nr ^5 if *t x cr fV/ n o c a p-
r. r" r. c >rr* tvfN.r\jrv,r-(,nfro--C' O O P" P~
OOOOO OCOOOOOCCCOOOOOOOOCOOC
I II I I I I I I I
M v 4 +
a 4 (T 4 o r if' x> cv fru^(^if vr>
ac 4tf,'Oif'^Cr~ir x X)fVf\ f\,c?xif\;(7'Pi 444 4
^-tr^p-ccp* r^cf'fC'r' C C*'-PgOC COCC <--CM'tO'MMrr^
OOOC OOOOCOOCOCOCCCCOOCCOOCC
II III II I I I I I II
* A* Ac
p- c -r r o- o 4
r>r-p*"4Op-.04 ^.^<*>4
INh-CNIPgNOf,'ffN-l00^i^fV.r
-J 0CCC-*-H^-.MMH'00-r-0'-'<^'*' C 4
*-> oooooooooooooooooocoooooooo
II II I I I I I
*+ 4> n
4,0p-^<*'^r>p'p> ir-4afv.ortr r(p
^40r44>oiPiOP'ngp-rg-ir'or~x3>c.cg
4 p- or- 00pg o-rs. 4P-4 /
ifCCO-rM\^-00-'C''-^0^--00'-- OO OOOOOOOOOOOOOOOOOOOOOOOOO
I II I I I I I I I
+ W V if
Ofit*r-ifrooMfMNiniMO-N
Tor'^-irirn'-rcr^'CCCT-^) cn-o~-o r'
J 00p'inr40M-00-Cir<*'i'0N0N00o Oc*
O 0-00000000000000000000000000
lili I II I I I I I I I II I I
X
o
u
4 4
o 4 a o 4 4
O Mfi o
OC^CSOx
O "t If o O M X
e o o n m w
4 p- 4 r- o
ir ac Of r
OOO-OM-
O-9'nt^rv.t-
n^
^ i*" O M fV(
mx-otTOO-
O .0 ^ pg P-
w c* (? if n o a-
pu pg O
o o o
oooooooooooooooooooooooo
I lilil
Of
o
u
x pg|p4ir.0p^ M*4if C pgpgf'jrgpgpgpgrgrg n
a
COL. COL. COL. COI. CPL COL. mi. CPI
OlfiA 0.0440 0.1171 O.O47 0.??71 ^.4<5?0-r -0.11M
v *v
it O' c k r a-raMfif4
-r r r ff m 4 i' C >1 c P
Cc r c a. -O r 4x4
C X 4 I OOIP>CCC
.............
CCCCC CCOCCCCC
I I I I I
W* vv
K4fnO'ar(t-(tr'4i
44cr'4 - 4 (M i a r If c £> 4
4> C C
CCCCCOOOOCCOC
I I I
4.
< p- o p- c p* r g c c
4cp-4p- 4C
4 1 C C C rv 4 4(N O
CC c r 000*^0 o
.............
ococc cocoooc
I I I I I I
"V 4
ropcraroop-eir
4Cr 434 <^C(\40(\
*'(Cf'4eo'if(rct-px rvcOr*r-rv C. r- rv c
c ccccccocccc c
I I I I I I I
** *
C f c jciff-o'iMirciif'ig
C cv 4 o k r c ir o x
r o -t r c r C p
-coc ccc c ecocoe
ill ii i
V +4.
a IC-C*C* (rif^irrf\.irN(rrK4if
4Mllf.N \ ClP(MMtH
occvo >ro pocp-
coooocoooooooc
i i i ii
M +> k
O f*1 N I if 4 tr rv. 4 O O
(f(vN-oaa40(fC4
a c 4 r* p- tr 4
C n- rv r- 4 r* o f* O
ocoocccococ c
I till
*f Ii
< ir a p o o x-*o<
c POPnNr
4 x 4: r- 4-r-r.ir. rfM4,4
Cr*- ori f'
C COOOOOOCOOOOO
ir r-
O ir tr
O O fN.
CO'CO'-OMCffH
irggififfooffiN
CMN4COO-C
OOCOCCO OC
III I
^ k
4 m tr -m MNiCMlfO
044Ci*cirr O O O O fp in
OOOOOOOOCOOOOO
I I II III I
ifs-44r(Mf4C>f-
inr erffMMOMO-
4)-(fi04^NO OfMOPg OOCOOl*- (*
000000000-000
I II I
* kk *
CP PJ o P- 4 o 4 4amN^
lf'4M4'C-4<4(Mf'ffNO
T r- 4 4 if'(ror\jt-43^o
pg rr O onoPgOO pj
OOOOOOOOOOOOOO
II I I
4 IT PU (* 4 4 o 4 pg
Ph4lT4i IN0N04P
NNP4#OCNO'P-
pg pg O OOCOC
00000000-0000
I II I
Ai Ac M
-44Nh-r>oirif>
4CPONP 4Pt'OH
IT4l4f)N4mP<4-0
-|f OOOONPONOPNO
..............
OOOOOOOOOOOOOO
I III I
4
r- m r--
O O 4 *4 fv.
O rg
O-4P0
a
o o o o o o o
till
pg O tr rg r r
if o pg pm oo *
BOO-4N4
O o O 4 M O
0-00000
I I II
* +*
0'P4)4mio-r'WN0'
- Pg a P -
4PmOPoorpg(riH
OOOOO NOP O
OOOOOOOOOOOOOO
III I III
f\io^ir'^p-(y*o rgo4
if 4) P O' O M|f 4if 4S
Mpgpgpgpgpgpgpgpg


68
APPENDIX D
Home-Play Questionnaire
This is a short questionnaire about some of your feelings about play.
Notice that there are possible answers showing about how some people feel
about play. Circle the letter of the one which is closest to how you feel.
If none of them apply write your own answer in the bottom space.
1. When should children play?
a. When chores and work are all completed.
b. A regularly scheduled play time is best.
c. Whenever they want to.
d. Only with parent's permission.
e.
2. Boys should be discouraged from playing with girls' toys and games.
a. Only when the child seems to play with girls' toys to excess, or
more than he plays with boys' toys.
b. Never.
c. Always.
d. Only when there is no adult male or father in the household.
e.
3. A child should share his toys with other children.
a. Always.
b. With visitors and brothers and sisters.
c. If the parent thinks it should be shared.
d. When ever he wants to.
e.
4. Wrestling or rough-housing should be done:
a. Only outdoors or in designated areas.
b. Only when closely supervised by an adult.
c. No restrictions ought to be placed on it.
d. Only by boys.
e.
5. Children's play should be mainly things that teach them useful ideas
and skills.
a. Completely agree.
b. Mildly agree.
c. Completely disagree.
d. Mildly Disagree.
e.


72
APPENDIX F
Demographic Information
Father's name
Age
Occupation
Last year
completed schooling
Mother's age
Occupation
Last year completed schooling
Number of children in family
Name
Age
Name
Age
Name
Age
Name
Age
Name
Age
Others living with family
Age
Years of Schooling
Number of rooms in house


67
APPENDIX C
0. J. Harvey Questionnaire
Now we would like to know your opinions and beliefs about several
other topics. Please write at least two (2) sentences about each topic.
You will be timed on each topic at a pace that will make it necessary for
you to work rapidly. Be sure to write what you genuinely believe. Take
the topics in order, beginning with number 1.
1. This
I believe about marriage .
2. This
I believe about religion .
3. This
I believe about the American Way of Life .
4. This
I believe about friendship .
5. This
I believe about people .
6. This
I believe about guilt .
7. This I believe about women's rights .
8. This
I believe about work .
9. This
I believe about law .
10. This
I believe about education .


73
APPENDIX G
Variables for Correlation Matrix
Col
1
Child's IQ
Col
2
Child's Barron
Col
3
Child's Tell-a-Story
Col
4
Child's Fantasy Toys (used for Interaction)
Col
5
Mother's Age
Col
6
Mother's Years of School
Col
7
Mother's Barron
Col
8
Mother's Tell-a-Story
Col
9
Mother's Fantasy Toys (used for
Interaction)
Col
10
Father's Age
Col
11
Father's Years of School
Col
12
Father's Barron
Col
13
Father's Tell-a-Story
Col
14
Number of Rooms in Home
Col
15
Child's SFPI
Col
16
Child's Realistic Toys (used for
Interaction)
Col
17
Interaction
Col
18
Mother's SFPI
Col
19
Mother's Numbers of Permissives
Col
20
Mother's Number of Restrictives
Col
21
Mother's Realistic Toys (used fo
r Interaction)
Col
22
Father's SFPI
Col
23
Father's Number of Permissives
Col
24
Father's Number of Restrictives
Col
25
Number of Persons in the Household
Col
26
Child's Total Number of Toys
Col
27
Mother's Total Number of Toys
Col
28
Room/Person Ratio


30
scoring procedure. The scores obtained on the reliability-observer's
protocol were compared to the scores obtained on the examiner's protocol.
Percent agreement was established by the following formula:
No. of disagreements
100 No. of agreements = k ^reement
The first observer, in November, 1974, obtained nine protocols from
two observations of four student-examiners and one observation of the
principal investigator. The protocols yielded a 100% agreement with
those of the examiners.
The second observer conducted his observations in January, 1975. He
observed five examiners twice for a total of ten protocols. His protocols
obtained an agreement rate of 93.3%.
2. Reliability of scoring for fantasy pre-disposition measures.
Although the Barron Movement Threshold Inkblots yield a score which is
essentially unequivocal, both of the other measures (SFPI and Tel1-a-Story)
entail scorer judgments.
The over-all purpose of the three measures was to enable differentia
tion of a high and a low fantasy pre-disposition group. Accordingly, the
reliability of scoring of these measures was tested by determining whether
an independent observer, familiar with the use of the measures, could
select HF and LF groups by simply inspecting all of the protocols. The
degree of correspondence with the groups which had been discriminated
by the actual scoring procedure could then be ascertained.
Dr. Jerome Singer cooperated in obtaining the services of one of
his graduate students for this endeavor. The graduate student, informed
only of the number which had been selected by the scoring procedures
employed for the investigation (25 HF and 55 LF), inspected the total of
221 protocols. She sorted out 24 which she identified as HF and 60 which


56
significant differences for fathers' responses to the home-play question
naire; however, the restrictiveness of fathers in our sample was signifi
cantly related to that of their wives (r = .404). Could it be that
fathers' restrictive attitudes provide the background for mothers'
practices in constraining certain play activities? Clearly, fathers'
attitudes, perhaps as they may influence mothers' attitudes and practices,
warrant more investigation.
The mothers' restrictiveness was significantly correlated with the
child's SFPI, with each of her own fantasy pre-disposition measures, with
her years of schooling, the number of rooms in the house and the room/person
ratio. The relations are negative (except for the Barron which, it should
be recalled, is scored in a reverse direction). It suggests that the
higher the fantasy pre-disposition scores, the fewer the restrictives.
Also, the more years of schooling, the fewer the restrictives. And
the larger the home or the fewer persons per room, the fewer the re
strictives. The correlates with restrictiveness seem to make a kind of
"common sense" pattern. It is particularly interesting because all of
the mothers gave more wholly permissive responses than wholly restrictive
responses (Table 11). The entire questionnaire as used in the present
investigation (a modified version of the original by Bishop and Chace,
1971) consists of only eight questions. The most restrictive mothers
gave an average of 1.04 restrictive responses, so the proportion of
restrictive responses to the permissives and to the remainder of the
questions does not yield a picture of extreme repression of childhood
fun and games. Further use of the home-play environment questionnaire
ought to include an item analysis of responses to see if a pattern emerges
-- i.e., is it one kind of play which is being ruled out, or are the


63
passing ammunition.
And a set of sisters who captured one of the student examiners, taking
her out to their play house and refusing to let her go; each time she
suggested returning to the house, they insisted on having another pretend
dinner.
Many of the protocols include drawings the children made for us; often
we found additions or exchanges in the cannister of toys -- we also dis
covered early that we would need several duplicate sets of toys because
the attrition was so high, attesting, perhaps, to their child-appeal.
Lately it was learned that the study appeared to have introduced a
new pretend activity in one neighborhood. The mother of an interview-
family, meeting the investigator in a store, reported that her children
had invented a game called "psychologist." The procedure consists of
dumping a cannister of toys out on the table and then making copious
"notes" on papers pinned to a clip board.
With the study of play still in its infancy, with its boundaries,
as Klinger (1969) states, "ambiguous," with rigorous operational defini
tions still to be fashioned, the present study must be regarded as an
excursion into unventured territory.
Such investigations can help to discover the new questions to be
asked and assist in charting for further research. It is hoped that in
this vein the present study will make a contribution.


58
direction for further research.
In the same vein: Singer (1973) and Freyberg (1973) both report
the importance of opportunity for privacy, operationalized as the room/
person ratio, in the child's tendency to play imaginatively. For this
study the number of rooms in the home, the number of persons in the house
hold, and the room/person ratio all failed to obtain significant correla
tions with the children's fantasy pre-disposition measures, or to dis
tinguish between LF and HF groups. However, with the number of rooms
in the house (and also, the room/person ratio) significantly related to
the mothers' restrictives, one may conjucture that perhaps the amount of
space available is more important for the mother's sense of what kinds
of play must be constrained with strict rules, than for obtaining privacy
Again, a question for further research.
Also, because of the income level, the number of rooms in these
houses was high in proportion to the number of people. This might not
obtain for a city apartment-house population. It may be that "room for
privacy" becomes important as a factor for imaginative play under cir
cumstances which are typically over-crowded, but does not emerge as im
portant when the amount of space is generally ample.
A final comment -- a caveat: the design for this study implied a
notion of unidirectional effect, from parent to child. However, the
current literature in child development, beginning with Bell (1968)
emphasizes the transactional nature of parent-child interaction. It
must be acknowledged that the tendency of the child to play imaginatively
might well encourage enjoyment of fantasy by the mother. Perhaps make-
believe play does not necessitate the kinds of restrictive arrangements
that high motoric play might demand. Such considerations could help to


46
TABLE 13
CORRELATION MATRIX OF CHILD'S AND SIBLINGS' SCORES
ON FANTASY PRE-DISPOSITION MEASURES
Col
Col
Col
Col
Col
Col
1
2
3
4
5
6
Child's
Sib's
Child's
Sib's
Child's
Sib's
Barron
Barron
SFPI
SFPI
Tel1-a-Story
Tell-a-Story
1
1.000
.327
-.608*
-.237
-.833*
-.113
2
.327
1.000
-.240
-.218
-.180
-.164
3
-.608*
-.240
1.000
.076
.768*
.147
4
-.237
-.218
.076
1.000
.019
.365*
5
-.833*
.180
.768*
.019
1.000
.038
6
-.113
-.164
.147
.365*
.038
1.000
*p<.05 r = .349 two tail
df = 30
No such selection process obtained for the siblings, however. While
the siblings do not seem to systematically resemble their sib in the
investigation, their own scores on the SFPI and their scores on the Tel 1-
a-Story are significantly correlated, suggesting that the score on one
will tend to predict the score on the other.
5. The fifth hypothesis stated the suggestion that some degree of
privacy would be associated positively with higher fantasy pre-disposition.
Specifically, it was predicted that HF children would have 1) more
room for privacy as measured by the Room/Person ratio and 2) more mothers
employed outside the home and thereby maintaining some degree of distance
from the child.
The prediction was not confirmed. Neither condition distinguished
between the groups.
The Room/Person ratio did not correlate with any of the children's
scores. Its one significant correlation with any measure was, as


38
TABLE 7 Continued
Variable
Low Fantasy
High
Fantasy
Test of Significance
Number of rooms
in home
Range
Mean
S.D.
nm
4-11 rooms
7.14
1.65
4-10
7.29
1.49
rooms
_t Test (36) = -.44 NS
**p<.0005



INDIVIDUAL AND FAMILIAL CORRELATES OF
CHILDREN'S FANTASY PLAY
By
LORRAINE BRADT DENNIS
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1976

ACKNOWLEDGMENTS
With humility and deepest gratitude, my indebtedness to a large numbe
of individuals must be acknowledged. Without their efforts and good will,
this dissertation truly could not have been accomplished.
I need, especially, to thank Dr. Ira J. Gordon, the chairman of my
committee, for his incisive guidance, his wise instruction and for his
unfailing calm and cheerful support during the dark moments which are
part of the life of every graduate student.
Dr. Robert Jester and Dr. Barry Lester, members of my committee, are
also to be thanked for their contributions to the design of this project,
and their constructive criticisms which enhanced its execution.
Dr. Jerome L. Singer's interest, his loan of the Barron Inkblots and
his help in obtaining the services of one of his graduate students,
Claudia Morelia, for a reliability study places me in his debt.
Next, I must cite, fondly, the six students who assisted in the
data gathering and scoring of protocols from beginning to end, giving
tirelessly of their time and good thinking: Marilyn Packard-Luther,
Elise Carlson, Ann Leaver, Mary Jane Scott, Don Whitworth and Scott Hicks.
Beverly Malin, Philip Goodwin, Claudia Morelia, Tullio Pitassi and
Elizabeth Hallenbeck all assisted with reliability studies and, in the
process, offered many ideas for improving our techniques.
A special note of thanks goes to my son, Brian Dennis, whose
assistance in coding and computerizing the data was invaluable. Stephen
Sledjeski took time from his busy life to program the multiple regression
equations.

Mr. Henry V. Diodati, superintendent of the Portsmouth, R.I., schools
and the four principals of the elementary schools, Mr. Lawrence P. Mello,
Mr. Lawrence Jones, Mr. Robert L. Crudup and Mr. Russell L. Carter must
be thanked for their generous help in making rooms available for our
testing and facilitating our efforts. Teachers of the kindergarten and
first grades of these schools, too, were magnificently tolerant of the
many class interruptions as we called for and returned the children during
days of testing.
I wish also to warmly acknowledge my gratitude to the wonderful
children and the splendid families who gave so generously of their time
and even more importantly, so promoted this project with their interest.
Finally, it is impossible to express how much this entire undertaking
owes to the steady, optimistic encouragement of Larry Dennis and our
children, Patrick, Brian, Deborah and Thomas, each of whom in uncounted
ways enabled my study and research.
i i i

TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS
ABSTRACT v
CHAPTER I INTRODUCTION AND STATEMENT OF PROBLEM 1
CHAPTER II REVIEW OF THE LITERATURE 5
CHAPTER III DESIGN OF STUDY 15
CHAPTER IV RESULTS 36
CHAPTER V DISCUSSION 53
APPENDICES 64
APPENDIX A Directions for Testing 65
APPENDIX B Singer's Structural Fantasy Pre-disposition
Interview (SFPI) 66
APPENDIX C 0. J. Harvey Questionnaire 67
APPENDIX D Home-Play Questionnaire 68
APPENDIX E Interaction 70
APPENDIX F Demographic Information 72
APPENDIX G Variables for Correlation Matrix 73
APPENDIX H Letter to Parents Requesting Consent 76
APPENDIX I Consent Form 77
REFERENCE LIST 78
REFERENCE NOTES 81
BIOGRAPHICAL SKETCH 82
TV

Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
INDIVIDUAL AND FAMILIAL CORRELATES OF
CHILDREN'S FANTASY PLAY
By
Lorraine Bradt Dennis
March, 1976
Chairman: Ira J. Gordon
Major Department: Foundations of Education
The purpose of this study was an attempt to discover individual and
familial correlates of children's imaginative play.
Two hundred eighty-one children, representing all those for whom
informed consent had been obtained (61.62% of the total number of children
in the kindergarten and first grades of the Portsmouth, Rhode Island
school system) were tested with the Peabody Picture Vocabulary Test and
three measures of fantasy pre-disposition: The Barron Movement Threshold
Inkblots, the Singer Structured Fantasy Pre-disposition Interview (SFPI)
and the Tell-a-Story Transcendence Index. Children with PPVT scores be
low 100 were omitted from the fantasy pre-disposition sample. A median
split for each fantasy pre-disposition measure with scoring criterion of
rank above or below the median on each of the three tests enabled
identification of 25 high fantasy pre-disposition (HF) and 55 low fantasy
pre-disposition (LF) children from the remaining 221.
v

TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ii
ABSTRACT v
CHAPTER I INTRODUCTION AND STATEMENT OF PROBLEM 1
CHAPTER II REVIEW OF THE LITERATURE 5
CHAPTER III DESIGN OF STUDY 15
CHAPTER IV RESULTS 36
CHAPTER V DISCUSSION 53
APPENDICES 64
APPENDIX A Directions for Testing 65
APPENDIX B Singer's Structural Fantasy Pre-disposition
Interview (SFPI) 66
APPENDIX C 0. J. Harvey Questionnaire 67
APPENDIX D Home-Play Questionnaire 68
APPENDIX E Interaction 70
APPENDIX F Demographic Information 72
APPENDIX G Variables for Correlation Matrix 73
APPENDIX H Letter to Parents Requesting Consent 76
APPENDIX I Consent Form 77
REFERENCE LIST 78
REFERENCE NOTES 81
BIOGRAPHICAL SKETCH 82
TV

Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
INDIVIDUAL AND FAMILIAL CORRELATES OF
CHILDREN'S FANTASY PLAY
By
Lorraine Bradt Dennis
March, 1976
Chairman: Ira J. Gordon
Major Department: Foundations of Education
The purpose of this study was an attempt to discover individual and
familial correlates of children's imaginative play.
Two hundred eighty-one children, representing all those for whom
informed consent had been obtained (61.62% of the total number of children
in the kindergarten and first grades of the Portsmouth, Rhode Island
school system) were tested with the Peabody Picture Vocabulary Test and
three measures of fantasy pre-disposition: The Barron Movement Threshold
Inkblots, the Singer Structured Fantasy Pre-disposition Interview (SFPI)
and the Tell-a-Story Transcendence Index. Children with PPVT scores be
low 100 were omitted from the fantasy pre-disposition sample. A median
split for each fantasy pre-disposition measure with scoring criterion of
rank above or below the median on each of the three tests enabled
identification of 25 high fantasy pre-disposition (HF) and 55 low fantasy
pre-disposition (LF) children from the remaining 221.
v

Seventeen intact families of HF children and 21 intact families of
LF children consented to participate. These were middle-class families
with at least one parent in every family having at least one year of
college.
Parents, individually, were administered the three fantasy pre
disposition measures, a test for abstract versus concrete conceptual
style and a questionnaire regarding home-play environment. In addition
there was a short demographic questionnaire and the mother and child were
engaged in a play activity termed "Interaction." Siblings over the age
of four were tested with the three fantasy pre-disposition measures.
The mother's Barron and SFPI scores were significantly correlated
with the child's SFPI score. The mother's SFPI score was also signifi
cantly correlated with the child's Tell-a-Story score. Neither fathers'
nor siblings' scores on the fantasy pre-disposition measures correlated
with those of the children.
The number of restrictive responses of the mother with regard to her
attitudes toward children's play in the home was significantly and nega
tively correlated with the fantasy pre-disposition of the child. Fathers'
restrictive responses were not significant.
The score on the Interaction play activity between mother and child
was significantly correlated with the fantasy pre-disposition of the
child.
The number of years of school of the mother, but not the father,
correlated significantly with the child's fantasy pre-disposition.
The child's PPVT IQ was significantly correlated with the fantasy
pre-disposition, despite the fact that all IQs below 100 had been
eliminated from the selection sample.
vi

The abstract versus concrete conceptual style of the parent, mother
or father, did not differentiate between LF and HF children.
No significant differences between LF and HF groups were obtained in
terms of sex, ordinal position, grade in school, age of mother, age of
father, outside employment of mother, number of rooms in the home or
number of persons in the family.
Individual and familial correlates of children's fantasy play sug
gested by this study are the child's verbal IQ, mother's scores on two of
the fantasy pre-disposition measures, her lack of restrictiveness with
regard to play in the home, the number of her years in school and her
inclination to engage the child in play which contains fantasy elements.
j
The small number of subjects in relation to the number of variables con
sidered necessitates caution in interpreting findings.
vn

CHAPTER I
INTRODUCTION AND STATEMENT OF PROBLEM
The systematic study of play is an astonishingly neglected research
area. It is difficult to account for such a blind spot on the part of those
interested in child development, especially since there exists a common
consensus regarding the prevalence of play in a child's life -- a general
notion that given half a chance, the young child would probably play most
of the time.
However, although it is impossible to pick up a text directed to
teachers of young children without encountering a section praising the
virtues of play, as a research topic it is indexed with only four references
in the third edition of Carmichael's Manual of Child Psychology, five more
if cross-referenced as "games." All except two of these occur in Piaget's
chapter on his own theory. The 1974 Annual Review of Psychology indexes
"play" not at all.
One deterrent to the study of play seems to have been the socio
cultural milieu from which scientists select their problems. Sutton-Smith
(1971) and Greg Stone (1971) suggest that the influence of a heavy achieve
ment orientation tended to promote a view of play as the polar opposite of
useful work and hence, inconsequential, the occupation of idle children.
A second and major deterrent has undoubtedly been the problem of
formulating an appropriate operational definition. A certain circularity
of reasoning occurs in the attempt to differentiate play from not-play,
since it is necessary to separate out already-identified "playful" elements.
1

2
Ellis (Note 1) points up a further paradox which hampers experimental rigor
-- the more the investigator gains control over a particular behavior, the
less the behavior can be playful. Klinger (1969), attending to the problem
of definition, asserts, "... the relevant behavioral domain is simply
insufficiently mapped to support the setting of rigorous positive bound
aries on the concept of play"(p. 279).
The task of mapping the "relative domain" of play is, however, be
ginning to attract investigative attention. A major impetus has come from
the proposition that play might have important consequences for later
development. Reminiscent of those who, early in the century, saw play as
practice for later proficiency, present-day ethologists are suggesting
that play is neither antic nor idle, but rather, essential for acquiring
the skills which will be necessary for survival (DeVore, 1963; Goodall,
1962). The use of sticks for the retrieval of termites, for example, does
not seem to be learned by the isolated ape who has no opportunity to play
with sticks, even when exposed to competent models (Mason, 1970). Kempler
(Note 2) has recently extended the notion by suggesting that the playing
child, creating a repertoire of divergent responses, is building a kind of
"future shock" absorber and preparing for the unpredictable.
The idea that play is useful was a major theme for the second annual
symposium on play, convened in Atlanta in April, 1974, for "people who
share the conviction that play is a rich and natural source of creativity,
learning and therapy." The title of the symposium was The Player and the
Environment and the program invited participants to consider "the role an
individual's environment has in encouraging or suppressing play in a child
or adult."
Among the featured speakers, Dr. Jerome Singer and Dr. Dorothy Singer

3
(Note 3) addressed the conference on the importance of an adult model for
fostering children's play. They were primarily concerned with the problem
of encouraging nursery school teachers to interact imaginatively with the
children, suggesting that without such interaction the amount of imaginative
play among children decreases, and that the adult interaction seems to have
an eliciting effect, providing the child license to play imaginatively.
Such adult influence on children's imaginative play should also be
important in the setting of the home. Though studies of parental contri
butions to child behavior have been criticized for methodological reasons
(Yarrow, 1963) for dearth of outcomes (Caldwell and Hersler, 1964) and for
conceptual flaws (Bell, 1968, 1971), there is continuing interest in parent
attitudes, practices and general home conditions as major factors in
several aspects of child development (Rebelsky, 1974; Baumrind, 1971).
Of all of the developmental characteristics, that of symbolic functions
seems most securely tied, empirically, to beginning influences, with both
language and concept formation tending to thrive or falter in terms of
early environmental variables (Gordon, 1969; Hess and Shipman, 1965).
Since imaginative play is characterized by the use of symbols (indeed, is
termed symbolic play by Piaget) it seems reasonable to suggest that it,
too, bears the stamp of the especial confluence of family circumstances,
attitudes and practices. However, little is known about the home back
ground factors which might relate to children's imaginative play.
The present investigation was designed to study parent attitudes,
practices and certain home conditions which might be correlated with
children's imaginative play. Since the study of play can be said to have
scarcely begun, it seems important to attempt to obtain data which relates
to the child's primary environment, the site of first play experiences.

4
In addition, because the field is young and tools and instruments
have neither been extensively developed nor tested, it was determined to
employ several existing measurement devices in an effort to assist in
establishing a further gauge of usefulness. Cross-study comparisons are
also facilitated when investigators continue with techniques and in
directions previously delineated by others.
The purpose of this study, accordingly, is to investigate possible
relationships between level of children's imaginative play, measured with
three instruments developed by Singer (1973) and his students, and several
familial variables which borrow from suggestions of Bishop and Chace (1971)
regarding requisites for a playful environment.

CHAPTER II
REVIEW OF THE LITERATURE
A search of the literature reveals that most contemporary investiga
tions have proceeded on the basis of assumptions that play (especially
imaginative play) is important for future development and that children
learn (or fail to learn) certain modes of play in terms of certain kinds
of environmental influences. However, it would be remiss to omit mention
of the fact that the most systematic theory of play to date contradicts
such contentions.
Piaget's Theory of Play
For Piaget (1951) play is a kind of inconsequential by-product of the
really important organismic business of adaptation. Play occurs when
schemas have become so familiar that the adaptive processes of assimila
tion and accommodation are not invoked; the activity is "merely for
pleasure, accompanied by smiles and even laughter, and without the expec
tations of results characteristic of the circular reactions through which
a child learns" (p. 90). Play, therefore, is not accorded any part in
the construction of intelligence. As a matter of fact, play proceeds
only when efforts toward adaptation are relaxed.
Nor, for Piaget, does imaginative play achieve any special signifi
cance. Rather, it is considered an invariant, the intermediate stage
between practice games of infancy and the games-with-rules of later life.
Furthermore, since Piaget views symbolic play as occurring precisely
5

6
because of the child's inability to think rationally and logically, he
would probably be mystified as to why anyone would even try to teach it.
Play, in the eyes of Piaget, is the occupation of the immature child, re
flecting his as yet inefficient modes of thought and is naturally abandoned
along with his childish views of reality as concrete and formal operations
are accomplished.
Piaget's theory has been criticized for his apparent insistence that
symbolic play is a symptom of immaturity, a kind of "mental digestive
system" (Millar, 1966, p. 50) whereby reality can be assimilated wholly
in terms of the child's own egocentric preoccupations of the moment. This
would seem to relegate adult fantasy and divergent thinking to the same
realm, implying that creative adults have simply failed to grow up.
However, no theory as detailed and cohesive has emerged as an alter
native. Learning theorists have not attended to play, though it can be
suggested that no new paradigm would have to be proposed to account for
its appearance and persistence in the life of a child. As an instrumental
response it produces satisfying changes in the environment -- baby shakes
the rattle and the small child rolls the bright ball. Parental approval
of activities would provide differential reinforcement. Imitation of
models in the environment as well as instruction might direct the child
toward acquiring the various skills and interest in sports or, contrari
wise, toward pursuits requiring imagination. These notions explicitly or
implicitly underly most current empirical research on play. Several dif
ferent kinds of environmental "input," seen as influencing play, have
been investigated.
Play is Learned -- Environmental Influences

7
Socio-economic Influences
The possibility that socio-economic factors influence the kind of
play in which the child typically engages rests heavily on a study con
ducted by Smilansky (1969) in Israel. Concerned with the school problems
of children from certain immigrant groups, her observations led to the
conclusion that socio-dramatic play (defined as involving at least one
other child in make-believe roles, actions and verbalizations which sub
stitute for real objects and situations, and sustained for at least ten
minutes) is virtually absent among these children. The conviction was
echoed in this country in terms of our own ghetto children by Lois Murphy
(1972), Sutton-Smith (1971) and Weikart, Rogers and Adcock (1970) among
others. Descriptions of the play of these children include notes of the
preponderance of impulsive and aimless activity, lack of sustained attention
and absence of signs of pleasure.
However, the contention that disadvantaged Israeli children do not
develop ability to engage in symbolic social play was discontinued by
Eifermann (1971) who sampled the play behavior of an enormous number of
Israeli children, carefully selected in terms of socio-economic class,
geographic location, community structure (town, kibbutz, Arab settlement)
and culture. She found that "culturally deprived" children play pretend
games both at later ages and to a far greater extent than advantaged
children. Furthermore, she reports, their pretend play is as elaborate
and as well sustained. An explanation for the contradictory findings
might lie in the setting in which play is observed. Labov (1972) suggests,
with regard to observations of language deficiencies of disadvantaged
children, that the usual setting for these judgments is simply too in
hibiting for certain children and that only by joining them "on their

8
own turf" can their behavior be truly sampled.
Play as Influenced by Instruction
Though reports of socio-economic differences in play may be open to
question, the possibility of teaching disadvantaged children to play pre
tend games is better substantiated. Smilansky's lead (1969) in documenting
the results of teaching socio-dramatic play to certain children was fol
lowed by similar efforts by Saltz and Johnson (1973) in Detroit and Frey-
berg (1973) in New York City. Each study reports significant gains in the
amount of symbolic play subsequently initiated by the trained children,
together with increases in language production, vocabulary range, verbal
fluency and ability to sustain attention as well as, for Saltz and Johnson,
significant increases in several retention abilities. Each of these
studies maintained appropriate controls for pleasant interaction with
adults so that the gain can be attributed to the imaginative play rather
than to simply the attention and interest of an adult. Persistence of
the gains over a long period of time has not been assessed. Specific
definition of the make-believe play differs somewhat from study to study,
but each definition includes the presence of a theme, roles, objects and
elements which are not present in reality.
The Russians, too, seem to be concerned with the role of the adult
teacher in fostering imaginative play (Elkonin, 1971), but there is no
hint of worry that some children may be deficient in make-believe produc
tions. Rather, it is believed that without adults to organize and guide
this activity, the imaginative play of the child might become a kind of
idle and self-indulgent dallying (Repina, 1971). Adults are to serve as
models for role playing, first mediating the use of playthings, then
leading the child into pretending with objects which are used as substitutes

9
for things which are missing (e.g., a spoonful of pine needles as imaginary
food). Finally the adult must help the child to imitate the activities of
admired grown-ups with appropriate suggestions, such as handing the doll
to the youngster and saying, "The baby is sick. Will you examine her,
doctor?"
The Russian psychologists see make-believe and imagination as con
tributing to the child's social identity when he tries out different social
roles and also as the bridge to abstract thought.
Play as Influenced by Toys
The question of the kinds of playthings which might encourage play
has attracted the attention of several investigators.
Gilmore (1966) found novelty to be an important variable, with his
sample of children responding more readily to new rather than familiar
objects in every play situation. Hutt (1971) refined the attraction of
new objects in attempting to distinguish between play activity and
"specific exploration." Play activity would not show a response decrement
while exploration would be characterized by a response decrease. Using
a specially constructed box with a lever attached to a counter, the events
produced by the lever were varied systematically. Hutt found that after the
3 to 5-year old child had explored the features of the box and lever fully,
lever pushing decreased. Then the box and lever were incorporated into
"play" with little focus on the lever and its effects. She suggests
that novelty provokes exploration while play occurs as the object becomes
familiar. It is as if the question shifted from "what does this object
do?" (exploration) to "what can I do with this object?" in play.
Pulaski (1970) hypothesized that unstructured toys would promote
more imaginative play than toys which are very realistic. She tested

10
middle-class kindergarten, first and second graders, selecting a group of
"high fantasy pre-disposition" and a group of "low fantasy pre-disposition"
on the basis of several measures developed by Singer (1973). The results
indicated that for children of five, six and seven years, the type of toy
seems to have little effect. High fantasizers played with both types of
toys imaginatively (according to a fantasy rating scale) while low fanta
sizers simply "fooled around," whether the toy was structured or unrealis
tic. Pulaski speculated that children's "playing style" may be fairly
well formed by age five. Branch (Note 4) is testing this by conducting a
similar study, using much younger children, at Yale University. A group
of boys and girls who are 20 months and 26 months old are being observed
at home during play with high-realistic versus low-realistic toys. Only
the preliminary results are available, but these suggest an age effect,
with all children engaging in more make-believe at the later age, a sex
effect with girls playing more imaginary games at both age levels, and a
toy by sex by age interaction with boys decreasing the amount of their
pretend play with high realistic toys at the older age level while girls
increase the amount of pretend play with high realistic toys at the later
age level. Branch notes also that the substantial differences exist in
amount of pretend play across subjects and suggests that even at this
early age, children may evidence pretend play as a distinctive play mode.
Family Influences on Play
Few studies have been located which attend to familial correlates of
play. Several studies can be considered pertinent if a link can be con
ceptualized between the imaginative productions of symbolic play and the
imagination necessary for creativity.
One study which is directly concerned with family variables and

11
children's make-believe play (Freyberg, 1973) is based on a very small
sample. Another (Singer, 1973) reports on some parental practices which
were found to be important as a result of interviewing children. The
other studies were undertaken with a different focus -- imaginary com
panions for Manosevitz, Prentice and Wilson (1973), and creativity for
MacKinnon (1962), Bishop and Chace (1971), and Dewing and Taft (1973).
Nevertheless, findings from these studies are provocative and, taken to
gether, seem to provide some fruitful leads for further investigation.
There is a persistent finding of the importance of birth order with
first born and only children showing greater tendencies toward fantasy
which has turned up in early studies of daydreaming (Singer, 1973) and
appears for imaginary companions (Manosevitz et^ aj_., 1973) and children
who play most imaginatively (Freyberg, 1973), but not for creative twelve-
year-olds (Dewing and Taft, 1973).
Level of parental education seemed to be important as correlated
with more make-believe play (Freyberg, 1973), but only the mother's level
of education was significant for creative twelve-year-old children (Dewing
and Taft, 1973).
Singer (1973) and Freyberg (1973) both report the possibility that
opportunity for privacy may have some relationship to the child's tendency
to play imaginatively. A possible allied finding is that of Dewing and Taft
(1973) and Freyberg (1973) that more mothers of creative and imaginative
children in their respective samples worked outside the home.
MacKinnon (1962), describing the childhood of creative architects,
noted that:
There was often a lack of intense closeness with one
or both parents. Most often this appears in relation
to the father rather than the mother, but often it

12
characterized the relationship with both parents.
.... Thus, if there was a certain distance in the
relationship between parent and child, it had a
liberating effect so far as the child was concerned.
If he lacked something of the emotional closeness
which some children experience with their parents,
he was also spared the psychological exploitation
that is so frequently seen in the life histories of
clinical patients (pp. 491-492).
Singer (1973), commenting along the same lines, suggests:
... an optimal balance of benign parental contact
and opportunity to be alone seems therefore essential
to the development of a rich imaginative life. This
optimal situation occurs most readily where the child's
mother is relatively warm, willing and capable of
spending time with the child, but not so emotionally
involved that she cannot at times leave the child to
his own devices (p. 62).
According to Singer (1973) another family variable consists of the
kinds of activities in which the child and parent engage together. He
suggests that parents who read or tell stories to their children are
providing a different kind of medium than those who spend time with their
children playing games of skill or chance.
In a study which provided some important conceptual leads for the
present investigation, Bishop and Chace (1971) argue that parental
personality qualities most likely to encourage children's playfulness at
home do not exist as isolated traits. Rather, they suggest, personality
theory and research points to a coherence in personality attributes.
Drawing upon the long series of publications by Harvey and associates
(1961, 1966, 1970), they invoke the proposition that certain personality
qualities are manifested in a basic "conceptual system," with differing
levels of organization which can be ordered along a dimension of con
creteness-abstractness. Four principle modes of conceptual functioning
are described in terms of the characteristic beliefs and response
patterns.

13
The concrete end of the continuum (System 1), with its simplistic view
of the world, stereotyped and polarized in terms of absolutes, its small
capacity to "decenter" or take the view of another, together with an in
ordinate need for structure and rules, is contrasted with the abstract
end (System 4) which accepts the world as complex and varied, is open to
new ideas and tolerant of ambiguity, views rules as utilitarian, but has
no need for reliance on authority or rule for rule's sake.
Reasoning that abstract parents would better enhance children's play
by refraining from imposing conventional standards, by willingness to accept
novel (even bizarre) ideas and willingness to live with a certain amount of
disorder, Bishop and Chace (1971) measured parents' conceptual systems on
Harvey's concrete-abstract continuum. The category of the parent was then
related to parental responses on a home-play environment questionnaire
which inquired into attitudes and the kinds of conditions maintained for
play, especially in terms of the restrictions imposed on play activities.
Although no significant differences were found for fathers in the study,
the outcome for mothers revealed that abstract mothers provided home play
environments which were most free of constraints and that the children of
these mothers were the most potentially creative. Potential creativity
was measured by indicators of complexity and variety of performance on a
laboratory play task.
The investigations of possible home influences on imaginative play
are exceedingly few in number. Only slightly more numerous are studies
of familial correlates of children's creativity. The latter were of
interest because of several studies which have indicated a link between
play, fantasy and creativity. Lieberman (1965) established significant
correlations between "playfulness" and three measures of divergent thinking

14
in an often-cited study of play and creativity. More recently Schaefer
(1969) surveyed 800 academically talented adolescents and discovered sig
nificant differences favoring the presence of imaginary companions in
childhood for those rated as "creative" in both art and literary expression.
Sutton-Smith (1967), also, tested the possibility of a relationship between
play and creativity, suggesting that favorite, well-used toys enable the
child to increase associations and therefore obtain more novel (i.e.,
creative) responses, a hypothesis upheld by his study for both boys and
girls. Klinger (1969), suggests that fantasy and play are essentially
undifferentiated until the third year, thereafter differentiated gradually
as play becomes rule-governed. He, too, indicates that the imaginative
productions of fantasy may be implicated in creativity.
Taken together, the studies of home and family variables, whether
directed toward children's imaginative play or creativity, suggest possible
influences which warrant further investigation. If an adult model is
important for imaginative play, as the Singers suggest (Note 3), perhaps
that adult would have to be comfortable with fantasy and make-believe.
Adults who engaged in imaginative play themselves when young, and who
enjoy fantasy in the present, may encourage more imaginative play for
their children. Perhaps these are the same adults who, as Bishop and
Chace (1971) propose, can refrain from too heavy-handed a management of
the child's activity and insistence upon conventional forms of play,
thereby providing a home environment which is conducive to fantasy and
make-believe. These are the chief considerations which guided the
present study.

CHAPTER III
DESIGN OF STUDY
The Problems and Hypotheses
The problem which the present study was designed to address is an
almost total lack of information regarding family background factors which
might encourage or discourage children's imaginative play.
Hypotheses were derived from the very few investigations to date,
some of which were focused on creative children rather than children who
engage in high levels of fantasy play. However, it seemed reasonable to
assume, in the absence of data, that the imagination necessary for
creativity and the imagination necessary for fantasy play might require
similar facilitating environments, as Bishop and Chace (1971) have
suggested.
Two groups of children, pre-selected for high or low fantasy pre
disposition, and their families were used to test the following hypotheses
1. There will be a positive relationship between children's scores
on the fantasy pre-disposition measures and their parents' scores on the
same measures.
This hypothesis was suggested by the findings of Freyberg (1973) and
Manosevitz et aj_. (1973) of more encouragement for imagination and fantasy
among parents of children who engage in make-believe play and who have
imaginary companions.
2. There will be a positive relationship between children's tendency
15

16
toward fantasy play and parents' tendency toward an abstract rather than
concrete conceptual system as measured by Harvey et_ aj_. (1961) scale.
This hypothesis was suggested by the findings of Bishop and Chace
(1971) with regard to parents of creative children who are characterized
by these investigators as encouraging "truly playful play" (p. 321).
3. There will be a positive relationship between children's
imaginative play and a less restrictive play environment in the home.
This hypothesis, also, borrowed from the findings of Bishop and Chace
(1971) concerning the type of home play environment provided for more
creative children.
4. There will be a positive relationship between scores of the
children on the fantasy pre-disposition measures and those of their
siblings who are age four or older on the same measures.
The rationale for this hypothesis suggested that if parents who are
characterized by abstract conceptual systems and who provide a home play
environment which is relatively free of adult constraints do, indeed,
enable more imaginative play, the environment should operate similarly
for other children in the family.
5. The homes of children with high fantasy pre-disposition will
have significantly more room for privacy. Furthermore, more mothers of
these children will be working mothers.
This two-part hypothesis was designed to follow up the findings of
Freyberg (1973), Dewing and Taft (1973) and the suggestions of Singer
(1973) regarding the need for privacy and for a certain amount of distance
between parent and child, presumably obtained when the mother is occupied
with employment outside the home.
6. There will be a positive relationship between children's scores

17
on measures of fantasy pre-disposition and an "Interaction" score obtained
when mother and child are playing together with toys which can be used
either imaginatively or realistically.
This final hypothesis was obtained from the suggestions of Lytton
(1972) and Singer (1973) regarding the importance of the kinds of activities
in which parent and child engage together.
Method
Subjects
Informed parental consent was obtained to include 281 kindergarten
and first grade children in Portsmouth, Rhode Island for fantasy pre
disposition screening. This represents 61.62% of all kindergarten and
first grade children enrolled in the Portsmouth school district at the
time of testing. Consent was not obtained or was refused for the remaining
38.38%.
Sixty of the children were eliminated from the sample because they
scored below 100 on the Peabody Picture Vocabulary Test. This was an
attempt to avoid confounding fantasy pre-disposition scores with intelli
gence and verbal production. The remaining 221 children constituted the
sample from which the subjects of the investigation were selected.
From their 221 protocols, according to a procedure which will be
described, a group of 25 high fantasy pre-disposition (HF) and 55 low
fantasy pre-disposition (LF) children were identified.
Families of these children were then contacted by telephone and
invited to participate in the second phase of the study. Table 1 details
the attrition of a number of potential families, some because of refusal
and some because the family had moved before the contact could be made.

18
Single parent families and those with father absent because of sea-duty
were omitted in order to adhere to the criterion of "intact families."
Two black families were also omitted in order to maintain a homogeneous
sample in terms of race. Two families turned out to be Portuguese im
migrants, with some parental English-language deficiency; both were
eliminated lest the child's LF score chiefly reflect the bilingual back
ground. Finally, several families either had no phone or an unlisted
number which could not be obtained, so that it was impossible to establish
contact.
TABLE 1
ATTRITION RECORD OF POTENTIAL SUBJECT FAMILIES
Reason for Elimination
Potential HF
Families
(N=25)
Number Eliminated
Potential LF
Fami1 ies
(N = 5 5)
Number Eliminated
Refused
3
12
Moved
3
7
Single Parent
1
5
No Phone or Unlisted Number
1
4
Minority Family
2
Portuguese Language
2
Sib of Other Child Already in Study
2
Total Eliminated
8
34
Remaining for Sample
17
21
Data, accordingly, were collected from 17 HF and 21 LF children,
their parents and their siblings over the age of four.
The subjects consisted of 16 girls and 22 boys, 25 of whom were
kindergarteners and 13 in the first grade, and their families. Twenty
of the children were first born, two were only children and two were

19
members of sets of twins.
The average IQ was 121.16 (SD = 10.12).
Mothers of these children had a mean age of 31.55 years (SD = 4.80)
while the average age of the fathers was 33.87 years (SD = 4.90).
The educational level of the parents was high. Mothers averaged
14.32 years of schooling, a little more than two post-secondary years
(SD = 1.76) while fathers had a mean of 15.97 years, only slightly less
than four years beyond high school (SD = 1.94).
Two families in the sample lived in apartments. The remainder lived
in single family homes. The average number of rooms per household was
7.21 (SD = 1.56).
No family in this sample consisted of other than immediate family
members. Average family size was 4.34 persons (SD = .71).
Occupations of fathers are listed in Table 2 and the occupational
training of mothers is reported in Table 3.

20
TABLE 2
FATHERS' OCCUPATIONS
Group
Frequency
Occupation
Low Fantasy
6
Engineers
3
Navy Officers
1
Retired Navy Officer
1
Lawyer
1
Insurance Executive
1
High School Teacher
3
Financial Analysts
1
Contractor
1
Pastor
1
Truck Driver (college
student)
1
Barber (college student)
1
Steam Fitter (college
student)
High Fantasy
6
Engineers
2
Navy Officers
1
Retired Navy Officer
1
Lawyer
I
Insurance Executive
1
High School Teacher
1
Electronic Technician
(no college work)
1
Anesthetist
I
Technical Buyer
1
Computer Programmer
1
Brick Layer (no college
work)

21
TABLE 3
MOTHERS' OCCUPATIONAL TRAINING AND COLLEGE WORK
Groups
Frequency
Occupational Training
Low Fantasy
2
Registered Nurse
4
High School Teacher
1
B. A. Degree
1
B. S. Degree
5
At least one year college
Total
13
High Fantasy
8
Registered Nurse
2
Elementary Teacher
1
Speech Therapy
1
B. S. Degree
5
At least one year college
Total
17
Procedure
The 281 children of the initial sample were tested with the Peabody
Picture Vocabulary Test and with three fantasy pre-disposition measures
by the principal investigator and six assistants, four specially-selected,
outstanding undergraduate psychology students and two graduates of the
psychology program at Roger Williams College, one of whom holds a master's
degree in child development from the University of Rhode Island. All had
received training in the use of the testing devices, had administered the
tests to a number of children outside the study and were individually
supervised by the principal investigator during the first three weeks of
testing. Reliability data appear in a subsequent section.
The three measures of fantasy pre-disposition were:
1. The Barron (1955) Movement Threshold Inkblots. This is a series
of 28 inkblots, arranged in a series so that it becomes progressively
easier to see human movement (M), which has been found effective in use

22
with young children (Pulaski, 1970; Singer, 1973; Freyberg, 1973). The
score is the number of the inkblot for which the child first responds with
human movement association. It should be noted that the lower the score
on this test, the earlier the M response has occurred, and hence the higher
the fantasy pre-disposition.
2. The Singer Structured Fantasy Pre-disposition Interview (SFPI).
A series of four questions developed by Singer (1973) which inquire into
the child's favorite play activities (Appendix B). Described by Singer
as a "clinical interview," the examiner is urged to use all skill in a
attempting to ascertain whether there are "pretend" or fantasy elements
in the child's responses. Because of the large number of examiners,
however, it was determined to hold procedures as standard as possible,
even at the risk of missing fantasy which might actually be present, in
an effort to guard against the possibility of suggesting responses to the
child. Therefore examiners were permitted to probe only with a question
which requested "Tell me more about that."
Each interview question is scored with one point if fantasy or pre
tend is judged to be present in the response. The highest possible score
for the SFPI would be 4, while the lowest possible is 0.
3. Tell-a-Story: Stimulus Situation, Card #6, Human CAT test. The
stimulus situation for the Tel1-a-Story, the sixth card from the Child
ren's Appreception Test (CAT), Human version, depicts two covered, sleeping
figures on the ground and a smaller person, head upraised and awake, in
the foreground. The child is asked to make up a story about the picture.
The story is recorded verbatim and scored according to the Transcendence
Index developed by Weisskopf (1950) which awards one point for every
element which is supplied by the story-teller, as opposed to those which

23
are obviously depicted in the stimulus situation. Thus, the child would
be given three points for describing the larger figures as "the mother
and father" two for ascription of sex to each character and one for
identification of a family relationship. Elements once scored were not
credited upon repetition.
The validity of indices of fantasy pre-disposition as measures of
imaginative play rests on a number of studies conducted by Singer (1973)
and his colleagues. All findings point with statistical significance in
the same direction. Pulaski (1970) found correlations between amount of
imaginative play and ratings for fantasy pre-disposition significant at
the .001 level. Freyberg (1973) also found a significant correlation
(p=.01) between fantasy pre-disposition scores and observations of make
believe play for the children in her study.
Scoring of the fantasy pre-disposition protocols was accomplished by
the investigator and the assistants. Each protocol was scored twice, in
dependently. The individual who had tested the child did not score the
protocol. The two independent scorings were compared and any discrepancy
resolved by means of consensus between the two scorers. The check for
reliability of scoring is detailed in a subsequent section.
After all the protocols had been scored, a median score was obtained
for each of the three measures. In order to be assigned to the high
fantasy pre-disposition group, a child had to score below the median
(because of the reverse scoring) on the Barron and above the median on
the other two measures, while the child who was rated low fantasy pre
disposition would have scored above the median on the Barron and below
the median on the SFPI and Tell-a-Story. The scoring criteria are
listed in Table 4.

24
TABLE 4
SCORING CRITERIA FOR FANTASY GROUP ASSIGNMENT
Fantasy Pre-disposition Tests
Group Assignment
Barron
SFPI
Tel1-a-Story
Low Fantasy
Pre-disposition
19-28
0, 1 or 2
0, 1, 2, 3 or 4
High Fantasy
Pre-disposition
1-18
3 or 4
5 and above
(maximum = 10)
Consent was obtained from the 17 families of HF children and the 21
families of LF children for home interviews and testing at a time of family
convenience. Except in one instance, all visits were scheduled and com
pleted when the father could be home. All home interviews and testing were
accomplished by the principal investigator and one or two of the student
assistants. Mothers and fathers were separated in order to obtain the
various measures, with examiners taking turns interviewing a mother one
time and a father the next. The principal investigator did all of the
mother-child Interaction observations. The second student assistant, on
occasion, tested siblings and accomplished reliability checks for the
Interaction observations (see reliability section).
Because screening of the children had taken place in November,
December and January, because so many children had been tested and so many
protocols scored and because interviewing of families did not begin until
April, it was possible to interview families "blind." By the time the
children had been assigned to the HF or LF category, they were just
"names." The names were transcribed on to index cards from the category
lists by a secretary, together with address and telephone number. Contacts

25
with the family were made using the card; examiners did not inform them
selves of the category of the children until interviews were completed.
By means of the home visits, the following tests, Interaction
activity, questionnaires and interviews were accomplished:
1. Fantasy Pre-disposition measures. Mothers, fathers and siblings
of age four and over were tested with the three fantasy pre-disposition
measures previously employed to screen the original children: Barron,
Singer Structured Fantasy Pre-disposition Interview (SFPI), and the Tell -
a-Story. Parents were asked to respond to the SFPI in terms of the
kind of play they recalled from their own childhood. These measures were
scored by two individuals associated with the study, independently, with
a subsequent consensus if disagreement existed, exactly as were the
children's, previously. The method of scoring had achieved a high degree
of reliability (see section on reliability).
2. Abstract-Concrete conceptual system. 0. J. Harvey's "This I
Believe" test (1970) of conceptual level (Appendix C) was used to determine
the conceptual system of the parent. Parents responded independently,
writing responses to the open-ended statements with ten referents, six
of which had previously been used by Harvey and for which detailed scoring
criteria are described (Note 5). The four referents added for this study
were: work, women's rights, law and education. Each referent is pre
ceded by the statement "This I believe about Responses
were timed, two minutes allowed for each response.
Each parent was categorized into one of four systems by two judges,
scoring the protocols independently and arriving at a consensus with a
third judge in case of disagreement. The principal investigator and
student assistants served as judges. Because the reliability check on

26
scoring, using an outside observer, was unsatisfactory (detailed in the
section on reliability) and because the interest for this investigation
lay in terms of the concrete versus abstract direction on the continuum
conceptualized by Harvey (et al_., 1961), System 1 and System 2 scores
were pooled in order to assign a "concrete" category to parents with
those scores, while System 3 and System 4 scores were designated as be
longing to the "abstract" category. Justification for this procedure
obtains both from the fact that emphasis for this study was upon the
global differences between concrete and abstract orientations of parents,
rather than the qualitative differences among the four systems, and that
the contrast between the two categories proved to be much more reliable
(again, see the section on reliability studies) than assignment of a
System score.
3. Home Play Environment: Parent Permissives and Restrictives.
Questions concerned with parent attitudes toward children's play were
selected from the original questionnaire developed by Bishop and Chace
(1971), utilizing those questions which had successfully discriminated
between restrictive and permissive parents in their investigation
(Appendix D). The questionnaire is designed with multiple alternative
answers plus a space for write-in answer for use if the parent decides
that none of the available answers are suitable. One of the alternatives
serves as a criterion response indicating restrictiveness and one is a
criterion response for permissiveness. The remaining answers and the
write-in response represent varying degrees of restriction or permission
-- that is, each contained some contingencies or disclaimers. This en
abled a count of the number of wholly restrictive or wholly permissive
responses of each parent. These questionnaires contained eight questions

27
and were also completed by each parent independently.
4. Interaction: Mother-Child play activity (Appendix E). Having
been previously informed that the principal investigator would like to
observe them playing together, the mother and child were requested to sit
together at a table or on the floor. The examiner emptied a cannister
of previously selected toys, saying, "Here are lots of little toys. Both
of you, playing together, build an exciting scene like for a movie
. and then I will ask you to tell me what is happening." (Adapted
from Erikson, 1972)
The toys were divided into two categories, "realistic" and "fantasy,"
with each realistic toy having its fantasy counterpart.
Realistic Toys
Fantasy Toys
1.
Mother
1.
Similar-sized, milk bottle
shaped wooden figure.
2.
Father
2.
Similar-si zed, milk bottle
shaped, wooden figure.
3.
Baby
3.
Similar-sized, milk bottle
shaped, wooden figure.
4.
Rhinocerus
4.
Green monster thing.
5.
Lion
5.
Red thing.
6.
Dog
6.
"Dumbo" elephant, sitting and
smiling.
7.
Toy furniture
7.
Blocks of assorted shapes.
Identical number of items as
furniture.
8.
Toy car
8.
Thing on wheels
9.
Policeman figure
9.
Superman figure
In addition there was a set of colored pipe cleaners and a set of
plain popsickle sticks. These were scored as "fantasy" items.

28
A check-list of the toys with separate space for recording mother or
child use enabled the observer to record the particular toy and the user.
Mother and child were scored separately for the number of toys of either
category which each contributed to the scene. Only toys which were actually
used in the scene were counted, yielding a "realistic toy" and a "fantasy
toy" score for each.
The description of "what is happening" in every instance reported by
the child, was recorded verbatim and then read back to the pair for
additions or corrections. The description was scored according to the
criteria listed below by three judges, independently, with agreement of
two out of three or all three constituting the reliability criterion, to
yield an Interaction score of 1 to 5 for the mother and child playing
together. The higher score reflects more make-believe or fantastic
elements.
Score
Criteria (Adapted from Pulaski, 1970)
(1) Familiar everyday experience
Concrete
Ordinary
Unelaborated
Use of realistic toy in realistic, unelaborated role
or function -- for example, the car is used as a
car or the policeman as a policeman.
(2) Indirect experience ... TV characters, movie
characters, story book characters and/or events.
Relatively unelaborated narrative.
(3) Content and characters farther removed from
experience or. more elaborated.
Might include "silly" or aggressive or emotional
fantasy.
(4)
More imaginative, unusual or original than #3 above.
Fantasy elements more detailed.

29
Score Criteria (continued)
(5) Fantastic
Farthest removed from present time, space or reality.
Most original, richest imagination. Elaborate details.
5. Demographic data (Interview) (Appendix F). The interview, con
ducted with either mother or father, ascertained the age of each parent,
the number of years of schooling for each, the occupation of the father,
the occupation of the mother, the number of persons living in the house
hold and the number of rooms in the house.
Reliability studies
Because of the large group of individuals engaged in both testing
and scoring, because certain aspects of scoring involved judgments and
because the reliability of certain of the fantasy pre-disposition measures
does not seem to have been well-documented, several different reliability
studies were conducted during the course of the investigation.
1. Reliability of the testing (fantasy pre-disposition screening)
of the children. Testing of the children took place over a period of
three months with Christmas vacation intervening. Two reliability studies
were accomplished, the first early in the testing period and another
during the last three weeks of testing in an effort to guard against
"slippage" of examiner technique. Two individuals, both advanced
psychology students unconnected with the main investigation, were involved,
one completing the early study and the second performing the study at the
end of the testing period. The procedure in both instances was identical.
The observer sat in on the testing two times for each examiner, completing
a protocol each time which was later scored according to the standard

30
scoring procedure. The scores obtained on the reliability-observer's
protocol were compared to the scores obtained on the examiner's protocol.
Percent agreement was established by the following formula:
No. of disagreements
100 No. of agreements = k ^reement
The first observer, in November, 1974, obtained nine protocols from
two observations of four student-examiners and one observation of the
principal investigator. The protocols yielded a 100% agreement with
those of the examiners.
The second observer conducted his observations in January, 1975. He
observed five examiners twice for a total of ten protocols. His protocols
obtained an agreement rate of 93.3%.
2. Reliability of scoring for fantasy pre-disposition measures.
Although the Barron Movement Threshold Inkblots yield a score which is
essentially unequivocal, both of the other measures (SFPI and Tel1-a-Story)
entail scorer judgments.
The over-all purpose of the three measures was to enable differentia
tion of a high and a low fantasy pre-disposition group. Accordingly, the
reliability of scoring of these measures was tested by determining whether
an independent observer, familiar with the use of the measures, could
select HF and LF groups by simply inspecting all of the protocols. The
degree of correspondence with the groups which had been discriminated
by the actual scoring procedure could then be ascertained.
Dr. Jerome Singer cooperated in obtaining the services of one of
his graduate students for this endeavor. The graduate student, informed
only of the number which had been selected by the scoring procedures
employed for the investigation (25 HF and 55 LF), inspected the total of
221 protocols. She sorted out 24 which she identified as HF and 60 which

31
seemed to her to be LF protocols. Twenty-two of her HF group corresponded
to those selected by the investigation procedures, three had been "missed"
and two of hers did not correspond to ours. She had selected 60 LF pro
tocols; these included all 55 of those previously selected in the investi
gation and five which did not correspond (Table 5).
This means that there were ten instances of non-correspondence out of
the total of 221 protocols, an agreement rate of 95.5%.
TABLE 5
RELIABILITY OF SCORING FOR FANTASY PRE-DISPOSITION MEASURES
Number of
HF
Number of
LF
Number
Remaining
Total
Identified by
Scoring Procedures
25
55
141
221
Identified correctly
for Reliability
22
55
134
211
Mis-identified for
Reliability
3
7
10
Agreement Rate = 95.47%
3. Reliability of
scoring for
0. J. Harvey
"This I Believe"
protocols
concrete-abstract conceptual systems. The final sample of 38 sets of
parents yielded 76 "This I Believe about ..." protocols which were
scored according to the Harvey (Note 5)scoring criteria as previously
described. An independent judge, a Ph.D. candidate in child development
at Purdue University, familiar with the Harvey scoring, scored 74 of the
protocols, assigning each into one of 4 systems. Agreement with the in
vestigation scoring was disappointing; 40 protocols corresponded, but
34 did not, a 54.05% agreement.

32
Therefore, as previously noted, System 1 and System 2 scores were
pooled to form a "concrete" category, while System 3 and System 4 scores,
grouped together, formed the "abstract" category.
A second inspection of the judge's scores revealed that the broader
categorization resulted in much better agreement with 61 of the 74 pro
tocols corresponding to the newly assigned category of "abstract" or
"concrete." This, therefore, yields 82.43% agreement.
4. Reliability of the fantasy pre-disposition measures. During the
process of visiting the families, it was possible to re-test 16 of the
children with the fantasy pre-disposition measures. Testing of all the
children had been completed in January, family visits began in late April
so at least three months had elapsed between the original testing and the
re-test.
Because examiners went into the household "blind," selection of
children for re-testing was circumstantial. If it proved possible to
separate the child from the parents' testing, so that parents' responses
would not influence the child's responses, a re-test might be accomplished.
Sometimes children were playing outside while parents were interviewed
and could be called in later. In most instances, however, the child
wanted to listen to the parent's interview and had to be considered con
taminated for a possible re-test. Subsequent inspection of the re-tests
revealed that six of the children had previously been categorized as HF
and 10 had been assigned to the LF group.
Scores on each of the three measures were correlated with the
original scores obtained by the child on that measure to obtain a cor
relation coefficient for each of the fantasy pre-disposition tests.
Results are tabulated in Table 6.

33
TABLE 6
TEST RE-TEST CORRELATION COEFFICIENTS:
FANTASY PRE-DISPOSITION MEASURES
Test r
Barron .274
SFPI .750**
Tell-a-Story .616*
*p<.01 d.f. = N-2 = 14
**p<.001
Thus it is evident that the reliability of the Barron is quite
doubtful, while that of the SFPI and Tell-a-Story proved to be much more
substantial.
5. Reliability of the Interaction observations. The Interaction
between mother and child was recorded on two kinds of protocols (Appendix
E). One check-list was used to record the toy used in the scene and by
whom it was used. The second obtained a verbatim account of "what is
happening" in the scene. In every instance observations were conducted
by the principal investigator. On five occasions a second observer
maintained a separate check-list of toy utilization and recorded the
description of the scene. The toy lists and Interaction scene descriptions
were counted and scored in the prescribed manner and compared with those
of the principal investigator.
The check-list of toys and who used them proved less reliable than
the account of "what is happening" which furnished the Interaction score.
The criterion for checking a toy as used by mother or child was that the
toy should be placed into and used in the scene. However, on a number of
occasions a toy first used by one of the players might be picked up and
used further or in another fashion by the other player. Discrepancies

34
in recording subsequent use of toys resulted in a low reliability for the
toy check-lists. None of the observer's corresponded exactly with those
of the investigator. The disagreement in actual toy-count was 38 out of 60,
yielding an agreement of only 36.67%.
However, the account of "what is happening" provided a reliability
of agreement on scores of 100% (i.e., the five observer verbatim records
obtained the same score as those of the investigator).
Statistical Analysis of Data
Nominal data which included the sex of the child, ordinal position,
whether or not the mother was employed outside the home, the concrete or
abstract category of the mother and of the father and the encouragement
or discouragement of make-believe play by each parent were analyzed by
the use of Chi square.
Fantasy pre-disposition scores for children of LF and HF groups, as
well as the scores of their parents, the Interaction scores for LF and
HF groups and the permissives and restrictives of each parent were
subjected to_t_tests for significant differences.
A correlation matrix was obtained which utilized all of the scores
on fantasy pre-disposition measures, IQ scores, age of parents, their
years of schooling and similar variables in order to affirm or disaffirm
predicted relationships between variables.
Finally, on the basis of the analysis of the correlation matrix, two
multiple regression analyses were performed in order to assess the relative
contribution of selected predictor variables to a predicted variable,
the child's SFPI score.
This study must be considered a pilot investigation into a field
which is in its infancy. Subjects for the study are a restricted sample,

35
limited in terms of numbers, region of the country, socio-economic status
and race. No claims are made for representativeness; therefore, extensive
inferential statistical procedures were not employed.

CHAPTER IV
RESULTS
Differences Between LF and HF Groups
The tests for significance of differences between LF and HF groups
are reported in Table 7.
Neither sex differences, nor that of ordinal position nor the grade
level of the child discriminated significantly between the LF and HF
groups.
The difference in the Peabody Picture Vocabulary IQ was found to be
highly significant, despite an effort to control for IQ effects. All
children with an IQ of less than 100 had been eliminated from the selection
process. Nevertheless, the LF group had a mean vocabulary-IQ of 117.27
(SD = 8.87) while that of the HF group was 125.94 (SD = 9.73). The t
test with 36 df is significant at p = .0005.
The age of the parents did not differentiate between LF and HF groups.
The number of persons in the household, the number of rooms in the home,
the number of mothers employed outside the home were also not relevant.
Although fathers' years of schooling were not significantly different
from group to group, the mothers' years of schooling are significantly
higher for HF children. Mothers of HF children average 14.88 years of
36

37
TABLE 7
DESCRIPTION OF L.F. AND H.F. GROUPS: TESTS OF DIFFERENCES
Variable
Low Fantasy
High Fantasy
Test of Significance
Sex
7 girls
9 girls
Corrected Chi
square
14 boys
8 boys
(1) = .782 NS
Ordinal Position
First Born
10
10
Corrected Chi
square
"only"
0
2
(1) = 1.20 NS
twi n
1
1
Grade
Kindergarten
14
11
Corrected Chi
square
First Grade
7
6
(1) = .005 NS
Peabody Picture
Vocabulary IQ
Range
104-140
106-145
t Test (36) =
-3.86**
Mean
117.27
125.94
S.D.
Mother's Age
8.87
9.73
Range
25-40 years
27-45 years
t Test (36) =
-.620 NS
Mean
31.19
32
S.D.
Father's Age
4.05
5.71
Range
27-43 years
26-46 years
t Test (36) =
-1.53 NS
Mean
33
34.94
S.D.
Mother's Years
of School
4.27
5.54
Range
10-17 years
13-17 years
t Test (36) =
-3.48*
Mean
13.86
14.88
S.D.
Father's Years
of School
2.01
1.22
Range
13-19 years
12-21 years
t Test (36) =
-.909 NS
Mean
15.76
16.24
S.D.
Mother's outside
employment
1.64
2.28
Not employed
outside home
13
11
Corrected Chi
square
Part-time
6
6
(1) = .0026 NS
Ful 1-time
2
0
Number of persons
in family
Range
4-6 persons
3-6 persons
t Test (36) =
1.79 NS
Mean
4.48
4.18
S.D.
.68
.73

38
TABLE 7 Continued
Variable
Low Fantasy
High
Fantasy
Test of Significance
Number of rooms
in home
Range
Mean
S.D.
nm
4-11 rooms
7.14
1.65
4-10
7.29
1.49
rooms
_t Test (36) = -.44 NS
**p<.0005

39
school (SD = 1.22) as contrasted with the mean of 13.86 years (SD = 2.01)
for mothers of LF children. The difference is significant at p = .001.
Findings for the Hypotheses
Tables 8, 9, and 10 report results of analysis of scores pertinent to
the hypotheses. The correlation matrix for all scores can be found in
Appendix G.
1. The first hypothesis stated the prediction of a positive relation
ship between children's scores on the measures of fantasy play and their
parent's scores on the same measures.
The hypothesis was only partially supported and only for certain of
the mothers' scores.
It can be seen (Table 8) that the mothers' Barron scores and the SFPI
are significantly different from LF to HF group. Mothers' Tell-a-Story
scores, however, do not discriminate.
Fathers' scores on the three fantasy pre-disposition measures do not
differentiate between LF and HF groups.
Correlation coefficients between parents' and children's scores are
presented in Table 9 as abstracted from the complete correlation matrix
(Appendix G).
There are no significant correlations between fathers' fantasy pre
disposition scores and those of the children.
Significant correlations occur between the child's SFPI and two of
the mothers' tests, the Barron and the SFPI. Mothers' SFPI is also
significantly correlated with children's Tel1-a-Story.
2. The second hypothesis stated the prediction that parents of HF
children would tend to be rated as "abstract" with regard to conceptual

40
TABLE 8
MEAN SCORES FOR FANTASY PRE-DISPOSITION MEASURES
Test
Child Score Mother Score Father Score
L.F. H.F. L.F. H.F. L.F. H.F.
Barron
S.D.
t Test
24.571
3.22
9.375**
12.706
5.57
17.952
5.71
2.367*
14.529
6.28
14.524
5.98
N.S.
15.824
4.99
SFPI
S.D.
t Test
1.429
.746
-16.327**
3.353
.493
1.762
1.04
-2.303*
2.412
1.23
1.667
1.35
N.S.
1.412
1.12
Tell-a-Story
S.D.
t Test
1.430
1.33
-13.780**
7.765
2.02
4.290
2.47
N.S.
5.294
2.47
4.714
1.74
N.S.
5.00
2.60
*p<.02
**p<.0001

TABLE 9
CORRELATIONS BETWEEN CHILD'S SCORES ON FANTASY PRE-DISPOSITION MEASURES
AND PARENTS' SCORES ON FANTASY PRE-DISPOSITION, RESTRICTIVES AND PERMISSIVES
Mother's
Barron
Mother's
Tell-a-
S to ry
Mother's
SFPI
Father's
Barron
Father's
Tell-a-
S to ry
Father'
SFPI
s Mother's
Permis-
sives
Mother's
Restric-
tives
Father's
Permis-
sives
Father's
Restric-
tives
Child's
Barron -133
-.147
-.304
.005
.093
.210
-.020
.241
-.019
-.071
Tel1-a-Story
-.224
.214
.327*
.057
.132
.197
.048
.217
-.029
.052
SFPI -.389*
.139
.369*
.104
.160
-.076
.111
-.420*
.224
-.080 -
*p<.05 r = .320 two tail
df = 36

42
TABLE 10
PARENT'S CONCRETE OR ABSTRACT CATEGORY
Category
Mother frequency3
Father frequency'3
L.F.
H.F.
L.F.
H.F.
Concrete
8
4
9
9
Abstract
13
13
12
8
Corrected Chi square (1 ) = .372 N.S.
^corrected Chi square (1) = .085 N.S.

43
system on 0. J. Harvey et a]_. (1961) scale.
The hypothesis was not supported, either for mothers or for fathers.
Table 10 shows that the abstract versus concrete ratings of parents did
not distinguish between LF and HF groups of children. Mothers' ratings
obtained a Chi square of .372, while Chi square for fathers' ratings was
.085, both well below the level of significance.
It can be noted that the total number of parents with abstract con
ceptual systems is greater for both groups than the total number of
parents with concrete conceptual systems.
3. The third hypothesis stated the prediction of a positive relation
ship between the child's HF pre-disposition and a non-restrictive play
environment in the home.
This hypothesis received partial support.
Mother and father, each completing a modified version of the Bishop-
Chace (1971) questionnaire on home-play environment, received scores
indicating the number of unequivocal permissive responses and the number
of wholly restrictive responses. Other possible responses to the
questions, all of which contained some qualifiers, were not counted for
purposes of this scoring. There are eight questions on the questionnaire
(Appendix D) so that each parent might have received a maximum of eight
completely permissive or eight extremely restrictive scores. The scoring
results are tabulated in Table 11.
It can be noted that the number of permissive responses given by
both mothers and fathers is substantially the same for each group. The
differences between groups were not significant. Inspection of the
correlation matrix (Appendix G) reveals that the permissive scores of
both parents obtain few significant correlations; mothers' permissives

44
TABLE 11
MEAN NUMBER OF PERMISSIVE AND RESTRICTIVE RESPONSES
Response
Mother
L.F.
H.F.
Father
L.F.
H.F.
Permissive
2.571
3.058
1.857
2.058
S.D.
1.78
1.71
1.62
1.14
t Test (36)
N.S.
N.S.
Restrictive
1.048
.352
.857
.823
S.D.
1.07
.606
.964
.951
t Test (36)
4.694*
N.S.
*p<.0001
are correlated negatively and significantly with mothers' and fathers'
restrictives. Fathers' permissives are correlated significantly only
with fathers' Tell-a-Story score.
Although the number of restrictive responses from fathers does not
differentiate between LF and HF groups, the mothers of LF children give
significantly more restrictive responses than mothers of HF children.
Support for the hypothesis, therefore, derives from the indication
that although LF and HF children seem to come from homes which are
similar in terms of permissive responses of both fathers and mothers,
the mothers of these children differ significantly with regard to the
number of very restrictive attitudes toward play which they express.
Mothers of HF children appear to be less restrictive in attitudes about
home-play than mothers of LF children.
The correlation matrix (Appendix G) shows that the mothers' re-
strictives are significantly related to the child's SFPI and also the
several other variables. The significant correlations are abstracted
for Table 12. Fathers' restrictives obtain only two significant

45
correlations, a positive relationship with the mothers' restrictives
and negative correlation with mothers' permissives.
TABLE 12
VARIABLES SIGNIFICANTLY CORRELATED WITH MOTHERS' RESTRICTIVES
Mothers' Restrictives
Child's SFPI
-0.420
Mother's Barron
0.391 (scored in reverse direction)
Mother's SFPI
-0.379
Mother's Tell-a-Story
-0.373
Mother's years of school
-0.564
Father's years of school
-0.371
Number of Rooms in home
-0.417
Father's restrictives
0.404
Room/Person ratio
-0.346
The relationship between mothers' restrictives and the child's
fantasy pre-disposition was further explored by means of multiple regres
sion analysis, described in a subsequent section.
4. The hypothesis that the children in this study would resemble
their brothers and sisters on the measures of fantasy play was not upheld.
Thirty-two siblings, with an average age of 7.70 years (SD = 2.85),
were administered the fantasy pre-disposition measures and the scores
correlated with those of their siblings in the sample. The correlation
matrix is tabulated in Table 13.
Not one of the child's fantasy pre-disposition measures correlated
with any of those of siblings.
The correlations among the child's own fantasy pre-disposition scores
would be expected because of the selection process for the high and low
fantasy groups. Only children with tested siblings were entered for
this matrix.

46
TABLE 13
CORRELATION MATRIX OF CHILD'S AND SIBLINGS' SCORES
ON FANTASY PRE-DISPOSITION MEASURES
Col
Col
Col
Col
Col
Col
1
2
3
4
5
6
Child's
Sib's
Child's
Sib's
Child's
Sib's
Barron
Barron
SFPI
SFPI
Tel1-a-Story
Tell-a-Story
1
1.000
.327
-.608*
-.237
-.833*
-.113
2
.327
1.000
-.240
-.218
-.180
-.164
3
-.608*
-.240
1.000
.076
.768*
.147
4
-.237
-.218
.076
1.000
.019
.365*
5
-.833*
.180
.768*
.019
1.000
.038
6
-.113
-.164
.147
.365*
.038
1.000
*p<.05 r = .349 two tail
df = 30
No such selection process obtained for the siblings, however. While
the siblings do not seem to systematically resemble their sib in the
investigation, their own scores on the SFPI and their scores on the Tel 1-
a-Story are significantly correlated, suggesting that the score on one
will tend to predict the score on the other.
5. The fifth hypothesis stated the suggestion that some degree of
privacy would be associated positively with higher fantasy pre-disposition.
Specifically, it was predicted that HF children would have 1) more
room for privacy as measured by the Room/Person ratio and 2) more mothers
employed outside the home and thereby maintaining some degree of distance
from the child.
The prediction was not confirmed. Neither condition distinguished
between the groups.
The Room/Person ratio did not correlate with any of the children's
scores. Its one significant correlation with any measure was, as

47
recorded on Table 12, a negative association with the number of mothers'
restrictive responses (r = -.346).
Chi square analysis of the number of mothers working either part time
or full time (grouped together for the analysis since the number of full
time working mothers in this sample is so small) revealed no basis for
assuming a difference for LF or HF groups (Table 7).
6. The sixth hypothesis stated the proposition that mothers and
children who play together imaginatively enable a prediction of high
fantasy pre-disposition for the children. Operationally, it anticipated
a positive relationship between the Interaction score on the mother-child
play activity and the level of the child's fantasy pre-disposition.
The hypothesis was substantially supported.
Table 14 records the mean Interaction scores of LF and HF groups
together with the _t_test for significance of the difference between means.
It can be seen that the Interaction score distinguished between LF and
HF groups with a t test which is significant at the .0003 level.
TABLE 14
MEAN SCORES FOR INTERACTION (MOTHER-CHILD ACTIVITY)
Low Fantasy
High Fantasy
Mean Score
2.095
3.118
S.D.
t test (36)
-3.986*
1.04
1.11
*p<.0003
The Interaction score obtained significant correlation coefficients
with all three of the child's fantasy pre-disposition measures as re
ported in Table 15 (abstracted from the complete correlation matrix).

48
TABLE 15
SIGNIFICANT CORRELATIONS WITH INTERACTION SCORE
Interaction
Child's Barron -.380
Child's SFPI .414
Child's Tel1-a-Story .520
Father Encourage .374 (obtained from multiple regression
correlation matrix)
To summarize the findings in regard to the hypotheses for this study:
The prediction that children and parents would resemble each other with
regard to fantasy pre-disposition scores was only partially substantiated
for only two of the mothers' scores.
The second prediction which suggested that parents of HF children
would tend to be categorized as having an abstract conceptual system was
not upheld.
The third prediction of a less restrictive home-play environment for
HF children was supported in terms of mothers' but not fathers' responses.
The fourth hypothesis stated the suggestion that brothers and sisters
of the children in the sample would resemble their LF or HF sib. This
hypothesis did not receive any support at all.
The fifth hypothesis stated the prediction of more privacy in terms
of a higher room/person ratio and more mothers employed outside the home
for HF children. Neither condition obtained and the hypothesis was not
confirmed.
The sixth prediction of more imaginative play-interaction between
mother and child for HF children was substantially supported.

49
Multiple Regression Analysis
The child measure which achieved the largest number of significant
correlations with other variables is the SFPI (Table 16). It is also
the most stable measure with a high degree of reliability on re-test
(r = .750, p = .001). It was selected as the dependent variable for
multiple regression analysis.
TABLE 16
SIGNIFICANT CORRELATIONS WITH CHILD'S BARRON, SFPI AND TELL-A-STORY
(ABSTRACTED FROM COMPLETE CORRELATION MATRIX, APPENDIX G)
Child's
Barron
Child's
SFPI
Child's
Tell-a-Story
Barron (child)
1.000
-.6167
-.8346
SFPI (child)
-.6167
1.000
.7464
Tell-a-Story (child)
-.8346
.7464
1.000
Vocabulary-IQ (child)
-.4888
.4930
.5474
Interaction (mother-child)
-.3798
.4135
.5474
Mother's years of school
-.3538
.3695
Mother's SFPI
.3695
.3270
Mother's Barron
-.3898
Mother's Restrictives
-.4195
Variables relating significantly to the child's SFPI are: the child's
IQ, the Interaction score, the mother's years of school, the mother's
SFPI, and the mother's number of restrictive responses for home-play
environment. The father's views on whether make-believe play should be
encouraged distinguished significantly between LF and HF groups (Table
17); this variable, termed "father encourage" was also entered into the
analysis.
A stepwise multiple regression analysis employing these variables
as predictors was performed. The summary table is reproduced for Table
18.

50
TABLE 17
PARENT'S RESPONSES REGARDING ENCOURAGEMENT OF MAKE BELIEVE
Response
Mother frequency9
Father frequency^
L.F.
H.F.
L.F.
H.F.
Yes
10
8
2
7
No
11
9
19
10
dcorrected Chi square (1) = .085 N.S.
^corrected Chi square (1) = 3.60 Fisher Exact Probability = .056
TABLE 18
SUMMARY TABLE MULTIPLE REGRESSION ANALYSIS I
Step Variable Entered
Multiple Rsq
Increase F Value
no.
R
Rsq
1
IQ
.4830
.2431
.2431
11.5619
2
Interaction
.5988
.3586
.1155
6.3043
3
M's Yrs. School
.6664
.4441
.0855
5.2318
4
F. encourage
.6852
.4695
.0254
1.5788
5
M's restrictives
.6952
.4833
.0138
.8518
6
M's SFPI
.6995
.4892
.0060
.3618
The analysis indicates that for this equation significant contri
butors to the variance of the child's SFPI score are IQ (24.31%), the
Interaction score (11.55%) and the mother's years of school (8.55%).
These three variables, taken together, account for a multiple R of .6664
or 44.41% of the variance of the child's SFPI.
Three variables, father encourage, mother's restrictives and mother's
SFPI fail to emerge as significant predictors. Each is significantly
correlated with a variable which entered the equation earlier.
A second multiple regression was performed with the five most likely

51
variables of the first equation, three of which had achieved a significant
F ratio (IQ, Interaction and mother's years in school) and two which might
possibly emerge if the order of entrance were changed. In this instance
it was determined that IQ should enter last.
The summary table is reproduced for Table 19.
TABLE 19
SUMMARY TABLE
MULTIPLE REGRESSION ANALYSIS II
Step
no.
Variable Entered
Multiple
R
Rsq
Increase
Rsq
F Value
1
M's Restrictives
.4195
.1760
.1760
7.6903
2
Interaction
.5596
.3131
.1371
6.9878
3
M's years school
.5941
.3530
.0398
2.0093
4
M's SFPI
.6234
.3887
.0357
1.9258
5
IQ
.6867
.4716
.0829
5.0202
The five variables account for 47.16%
(R =
.6867) of the variance
of the child's SFPI. Mothers' restrictives and the Interaction score
become significant predictors with restrictives accounting for 17.60%
and Interaction accounting for 13.71% of the variance of the dependent
variable.
IQ, also, obtains significance, accounting for 8.29% of the variance
in this equation.
On this occasion, mothers' years of school does not reach significance.
It is significantly correlated with restrictives (r = -.564) so that
employing one as a predictor seems to relegate the influence of the other
to the background.
Mothers' SFPI does not appear to be an important contributor for
either equation.

52
In the light of the two equations, it is possible to suggest the
influence of four variables on the amount of child's make-believe play
as measured by the SFPI. Two of these, the child's IQ and the mother-
child Interaction, were significant predictors for both equations. Each
of the two remaining, mothers' restrictives and mothers' years of schooling,
emerged as significant for one equation. It would appear that use of one
as a predictor might preclude use of the other; the substantial inter
correlation suggests shared variance which would necessitate testing
beyond the scope of this study.
Summary of Findings
To summarize the individual and familial correlates of children's
make-believe play which have been suggested by this investigation: the
child's vocabulary-IQ seems to be the most important individual charac
teristic associated with level of fantasy pre-disposition for children
in this study.
A relationship was found for two measures of mothers' fantasy pre
disposition, the Barron and SFPI, and children's SFPI. The mothers'
years of schooling is correlated with children's scores on the Barron
and SFPI. The number of restrictive responses made by the mother with
regard to home-play environment is negatively associated with the amount
of make-believe of the child's SFPI.
Finally, the amount of fantasy in the Interaction score, for which
the mother and child engaged in play together, is significantly correlated
with all three of the children's fantasy pre-disposition measures.

CHAPTER V
DISCUSSION
This study was undertaken in an effort to discover possible relation
ships between individual and family-background factors and children's
make-believe play. Because there have been so few investigations of this
area, it must be considered exploratory in design and concept. Because
the number of subjects is so small in relation to the number of variables
the data must be interpreted with extreme caution.
The sample is limited not only in size but also in composition; it
is white, middle class, small family and lives on a large island off the
coast of Rhode Island. It contains an unusual number of Navy families
because of the large defense installations at Portsmouth and Newport.
Inferences from this sample to a general population can be taken only
as suggestions for further investigation.
However, the homogeneity of the sample has an advantage of con
trolling for socio-economic class differences which could otherwise con
found findings. These children come from homes with ample space for play
and with easy access to beaches, woods and meadows for a wide variety
of summer and winter play activities.
Notable was the deep interest of both LF and HF parents in their
children and in the study. On the basis of pilot trials approximately
one hour had been allotted for the family interviews. It quickly became
apparent that the time was greatly underestimated. Parents wanted to
53

54
talk after the interview sessions were completed, comparing notes about
their responses, plying interviewers with questions and engaging in dis
cussion about childrearing, often until after midnight.
The children, whether LF or HF, appeared to come from similar homes
in terms of parental concern and caring, general socio-economic circum
stances and physical environment.
Yet, by the measures used for this investigation, the children
themselves seem to be very different. They represent the two extremes
of a larger group of 221 children with regard to scores for the three
fantasy pre-disposition tests. A question must be raised with regard
to the validity of the measures. Do these children really differ in the
amount of make-believe play?
The findings of other investigators concerning the relationship of
the measures to make-believe play, observed under several different cir
cumstances, have been cited. The SFPI, the dependent variable for
multiple regression analysis in this study, is scored specifically in
terms of the pretend play which can be discerned in the responses and
seems to yield reliable scores on retest after a passing of several
months. Qualitative examination of the protocols was not within the
scope of this study, but non-systematic review indicates that the LF
children tended to respond to the question "What do you like to play
best?" with physical activity such as "play baseball" or "ride my bike,"
while HF children typically spoke of playing "house" or "school" or
"cowboys." However, the problem of validity is troublesome. Only
further research can provide definitive answers to the question. This
study was designed to begin with already-existing measures as a step
in assaying their utility.

55
The expected correlations between parents' scores on these measures
and the scores of their children did not wholly materialize. It had been
anticipated that parents who reported a good deal of make-believe play in
their own childhood and whose Barron and Tell-a-Story contained substantial
fantasy elements would tend to encourage children's fantasy pre-disposition.
Although this was suggested for two of the mothers' scores (SFPI and
Barron), fathers' scores did not seem related to those of their children
at all. It may be that a number of the fathers in this sample are too
busy for extensive contact with their children. Ten of the fathers are
college students in addition to being heads of households with full-time
jobs; half of these are in graduate work. Five are in the Navy with
regular sea duty built into their lives.
However, it does appear that an attitude of unqualified approval of
children's make-believe distinguishes between fathers of LF and HF chil
dren. The variable did not appear to contribute significantly to the
child's SFPI on multiple regression analysis. However, father encourage
is significantly related (r = .374) to Interaction, a significant pre
dictor in both analyses. It is interesting to conjecture that perhaps
fathers' attitudes exert an indirect influence, perhaps in maintaining
a supportive rather than a negative atmosphere for fantasy, which better
enables the mother and child to engage in make-believe freely when playing
together.
A similar speculation might be offered with regard to the mothers'
and fathers' restrictives. The Bishop and Chace (1971) study had in
dicated the importance of lack of restrictiveness in home-play environ
ment provided by the mother. This finding was corroborated by the
present investigation. Bishop and Chace were not able to establish any

56
significant differences for fathers' responses to the home-play question
naire; however, the restrictiveness of fathers in our sample was signifi
cantly related to that of their wives (r = .404). Could it be that
fathers' restrictive attitudes provide the background for mothers'
practices in constraining certain play activities? Clearly, fathers'
attitudes, perhaps as they may influence mothers' attitudes and practices,
warrant more investigation.
The mothers' restrictiveness was significantly correlated with the
child's SFPI, with each of her own fantasy pre-disposition measures, with
her years of schooling, the number of rooms in the house and the room/person
ratio. The relations are negative (except for the Barron which, it should
be recalled, is scored in a reverse direction). It suggests that the
higher the fantasy pre-disposition scores, the fewer the restrictives.
Also, the more years of schooling, the fewer the restrictives. And
the larger the home or the fewer persons per room, the fewer the re
strictives. The correlates with restrictiveness seem to make a kind of
"common sense" pattern. It is particularly interesting because all of
the mothers gave more wholly permissive responses than wholly restrictive
responses (Table 11). The entire questionnaire as used in the present
investigation (a modified version of the original by Bishop and Chace,
1971) consists of only eight questions. The most restrictive mothers
gave an average of 1.04 restrictive responses, so the proportion of
restrictive responses to the permissives and to the remainder of the
questions does not yield a picture of extreme repression of childhood
fun and games. Further use of the home-play environment questionnaire
ought to include an item analysis of responses to see if a pattern emerges
-- i.e., is it one kind of play which is being ruled out, or are the

57
restrictives scattered?
In contrast to the Bishop and Chace (1971) investigation, the present
study was unable to obtain evidence of differences in abstract versus con
crete conceptual systems for LF and HF parents. This might be explained
by the relatively high level of education for parents in the sample, with
more parents categorized as abstract (46) than were rated as concrete (30).
It must also be suggested that in spite of earnest efforts to adhere to the
scoring criteria, the scoring of the This I Believe protocols may have
been faulty. Such a possibility would also account for the low reliability
found for the score procedure.
The level of parental education has been cited as important for
children's make-believe play by Freyberg (1973), while mothers' education
but not fathers' was significant for Dewing and Taft's (1973) study of
creative twelve-year-olds. This investigation confirmed the importance
of mothers' education, which correlated significantly with the child's
Barron and SFPI. Her level of education is also found to relate to a
less restrictive home-play environment. A question for further research
might inquire as to what aspects of higher education promote a less
restrictive orientation of the mother toward home play.
Additionally, fathers' education is found, in this sample, to re
late to the number of rooms in the household which, in turn, seems to
promote a lessened number of constraints on play imposed by the mother.
This could, perhaps, be another indication of fathers' indirect contri
bution, with his level of education enabling the material advantage of
a larger house and thereby securing a more relaxed atmosphere for mothers'
attitudes toward play. This kind of linking of correlations is, admittedly,
highly speculative. Nevertheless, it could provide indications of

58
direction for further research.
In the same vein: Singer (1973) and Freyberg (1973) both report
the importance of opportunity for privacy, operationalized as the room/
person ratio, in the child's tendency to play imaginatively. For this
study the number of rooms in the home, the number of persons in the house
hold, and the room/person ratio all failed to obtain significant correla
tions with the children's fantasy pre-disposition measures, or to dis
tinguish between LF and HF groups. However, with the number of rooms
in the house (and also, the room/person ratio) significantly related to
the mothers' restrictives, one may conjucture that perhaps the amount of
space available is more important for the mother's sense of what kinds
of play must be constrained with strict rules, than for obtaining privacy
Again, a question for further research.
Also, because of the income level, the number of rooms in these
houses was high in proportion to the number of people. This might not
obtain for a city apartment-house population. It may be that "room for
privacy" becomes important as a factor for imaginative play under cir
cumstances which are typically over-crowded, but does not emerge as im
portant when the amount of space is generally ample.
A final comment -- a caveat: the design for this study implied a
notion of unidirectional effect, from parent to child. However, the
current literature in child development, beginning with Bell (1968)
emphasizes the transactional nature of parent-child interaction. It
must be acknowledged that the tendency of the child to play imaginatively
might well encourage enjoyment of fantasy by the mother. Perhaps make-
believe play does not necessitate the kinds of restrictive arrangements
that high motoric play might demand. Such considerations could help to

59
account for the failure to obtain resemblance between fantasy pre-dis
position scores of children and their siblings. One might even suggest,
with Smilansky (1969) and others, that make-believe play promotes verbal
intelligence.
These are certainly considerations which must guide further research.
Instruments
This study employed several instruments which have been used in
previous investigations: the Barron Inkblot Movement Threshold and the
Singer Structured Fantasy Pre-disposition Interview (SFPI) (Singer, 1973;
Freyberg, 1973; Pulaski, 1973), and a modified version of Tell-a-Story
(Pulaski, 1970) in order to assess the child's, parent's and siblings'
fantasy pre-disposition. In addition, the 0. J. Harvey (1970) test of
conceptual style and a modified version of the Home-Play Environment
attitude questionnaire (Bishop and Chace, 1971) were employed to measure
parent variables. It is appropriate to report on our experience with
these instruments.
The test-retest reliability of the Barron is so low that serious
doubt must be raised regarding its utility without further effort toward
standardization. Such an effort ought also to include an assessment of
the present order of the inkblots. Our persistent impression held that
certain of the later ones obtained fewer M responses than early ones,
though the arrangement supposedly yields a series in which M responses
become progressively easier to obtain. The final inkblot (#28) especially
seems misplaced.
The SFPI was demonstrated to be a useful and highly reliable measure
for our sample. The scoring can apparently be accomplished to discriminate

60
a LF and HF group with an extraordinary degree of agreement between actual
scoring of protocols and a more "global" scanning and sorting procedure.
It is further possible to administer according to standardized procedures.
The Tel1-a-Story was only slightly less reliable. It is hoped that other
investigators will be interested in further use of these instruments,
enabling cross-study comparisons as well as the possibility of obtaining
some fully standardized tools for research in children's make-believe
play.
The 0. J. Harvey (1970) This I Believe test proved very difficult
to score reliably into four Systems according to the scoring criteria.
Better results were obtained if the respondents were simply categorized
as "abstract" or "concrete." It would seem that widely-used instruments
should provide scoring criteria which enable ready and reliable assess
ment, again in the interest of cross-study comparisons.
The Home-Play Environment attitude scale, as used in this investi
gation, was not checked for reliability. If subsequent study can docu
ment this, it too can be recommended for further research as an attractive
and easily administered instrument. Parents seemed to enjoy responding
and, later, discussing their responses with each other and with the
examiners.
The Interaction mother-child play activity, contrived for this study,
seems to have potential as a predictor of children's tendency toward high
or low fantasy play. It was not checked for reliability and the method
of recording for use of toys must be refined. It requires only about
twenty minutes for administration. Scoring is relatively simple and,
though this needs more extensive checking, seems quite reliable. Chil
dren and mothers alike were readily engaged in the activity, and children

61
and sibs enjoyed playing with the toys later, while examiners were busy
with parents.
Finally, it should be noted that the research into children's play
had great fascination for parents. Parents were charmed with being in
vited to discuss their own preferred childhood activities (for the SFPI)
and were eager to compare notes after separate interviews. It appeared
that few parents had described their childhood fun to each other, and the
occasion seemed to open new areas for communication. Often, one or the
other would exclaim, "You never told me that!" This might offer a fruit
ful direction for further research, especially if extended into present
activities. What do adults do now for "play?" What kinds of childhood
activities are carried forward into adult life? Needed also is an
assessment of the role of television in both children's play and adult
leisure-time pleasure.
Conclusions
Play, the ubiquitous childhood activity, has until recently been
little researched. One of the difficulties is that an operational defin
ition must almost certainly invoke the notion of spontaneous, freely
chosen engagement; therefore when the scientist attempts to quantify it
or remove it to his laboratory or circumscribe it for closer scrutiny,
he tends to violate its most salient characteristic. It is, of course,
the old problem of "observer effect."
This investigation has attempted to check out some previously used
instruments and suggestive findings of other researchers and to test some
hunches about background influences on fantasy play. The results were
modest. Supported was an indication of mother's influence in terms of

62
her own fantasy pre-disposition, her years of school, her interaction with
the child in a play activity and her attitudes about constraints on play
in the home. Unsupported was the possibility of a difference in cognitive
style of parents as influencing fantasy play or substantial influence from
fathers' attitudes or similarity of siblings with regard to fantasy pre
disposition. Although the IQ of the child seems to be associated with
level of make-believe, the sex of the child, ordinal position or number
of rooms in the household does not.
Obscured in scores and tests of significance, however, are the many
delightful incidents which, unscientifically, have served as reminders
of the elusive qualities of play.
There was the tiny kindergartener who, upon being escorted back to
her room after testing, stopped in the hall and exclaimed "Oh, I have to
go back. I forgot to say goodbye." The student examiner was puzzled.
"Yes," responded the small lass solemnly. "The clown and the monkey.
They were watching us all the time." The student examiner reporting
this was still a little shaken!
There was the little boy with an imaginary companion named "Jethro"
who "wakes me up at night to get him a drink of water." And another
whose imaginary friend is named "abdominal snow man." The spelling is
correct.
There was a young father with eyes alight as he described making
fireworks in his basement. He and his own father (the childrens'
grandpa) have "wars with castles" whenever the mother is not around to
object. They make castles out of blocks at opposite ends of the living
room and catapult little rockets at each other's construction. The
three small girls wear hard hats and become "knights," retrieving and

63
passing ammunition.
And a set of sisters who captured one of the student examiners, taking
her out to their play house and refusing to let her go; each time she
suggested returning to the house, they insisted on having another pretend
dinner.
Many of the protocols include drawings the children made for us; often
we found additions or exchanges in the cannister of toys -- we also dis
covered early that we would need several duplicate sets of toys because
the attrition was so high, attesting, perhaps, to their child-appeal.
Lately it was learned that the study appeared to have introduced a
new pretend activity in one neighborhood. The mother of an interview-
family, meeting the investigator in a store, reported that her children
had invented a game called "psychologist." The procedure consists of
dumping a cannister of toys out on the table and then making copious
"notes" on papers pinned to a clip board.
With the study of play still in its infancy, with its boundaries,
as Klinger (1969) states, "ambiguous," with rigorous operational defini
tions still to be fashioned, the present study must be regarded as an
excursion into unventured territory.
Such investigations can help to discover the new questions to be
asked and assist in charting for further research. It is hoped that in
this vein the present study will make a contribution.

APPENDICES

65
APPENDIX A
Directions for Testing
Barron's Movement Threshold Inkblots
To the child: I am going to show you some cards with designs on them.
They aren't made to look like anything -- they are just designs that were
made by dropping ink on paper and folding it. I want you to look at each
of the designs and tell me what it reminds you of -- what it looks like
to you. Every child can see different things and there aren't any right
or wrong answers. Sometimes one card may look like more than one thing
but I only want you to tell me one thing that you see on each card.
(To the tester: If a child gives more than one response to a card
repeat that you only want one answer per card and ask him to tell you
which of his answers that design reminds him of most.)
Peabody Picture Vocabulary Test (PPVT)
To the child: "I have some pictures to show you." (Turn to example
A and say:) "See there are four pictures on this page. Each of them is
numbered, (point to each one) I will say a word, then I want you to point
to the picture which best tells me the meaning of the word. Let us try
one. Point to the picture that best tells the meaning of 'crib.'" (After
the child responds correctly, then turn to example B saying:) "That's
fine. Which picture is fin?" (Then to C.) "What picture is 'butterfly?'
Fine! Now I'm going to show you some other pictures. Each time I say
a word you point to the picture which best tells the meaning of a word.
As we go through the book you may not be sure you know the meaning of
some of the words, but I want you to look carefully at all of the pictures
anyway and choose the one you think is right. What picture is ?"
To the examiner: Be sure that each child looks at each of the
alternatives because sometimes younger children will start looking in
only one location. Also, always be sure to get a response for every
card. If the child doesn't know the word, have him give it a try anyway.
You may encourage the child after some responses by saying "good," "fine,"
etc., but don't do it just after correct responses. You are praising
the child's effort not its correctness. You may repeat a word if necessary.
Do not precede any word with an article (a, an, the) and do not make any
singular words plural.
Scoring: Start with item #25 for all children. Continue forward
until the child makes his first error. If this is before he gets 8
correct, go back to a lower level (below 25). The ceiling score is
determined by the point at which the child has missed 6 out of 8 words.

66
APPENDIX B
Singer's Structural Fantasy Pre-disposition Interview (SFPI)
To the tester: These questions are designed to discover the nature
of each child's play and fantasy life. They should be presented slowly
and matter-of-factly to the child with as little comment on your part as
possible except to elicit details when an answer is unclear. Because
scoring is based on whether or not the child's answer indicates that he
is introducing make-believe elements into his play, it will be necessary
to follow-up answers which are not clearly make-believe or not make-
believe. However, do not go beyond the probe in questioning, though it
can be used more than once.
To the child: I am going to ask you some questions about things you
like to do.
(1) What do you like to play best? What do you like to play the most?
(Probe: Tell me more about it.) (Probe: Tell me how you play it.)
(2) What do you like to play best when you are all alone? What do you
like to do best when you are all alone?
(Probe: Tell me more about it.) (Probe: Tell me how you play it.)
(3) Do you ever see make-believe things with pictures in your mind or
think about them? What sort of things?
(Probe: Tell me more about it.)
(4) Do you have a make-believe friend? Do you have an animal or toy or
make-believe person you talk to or take along with you?
(Probe: Tell me more about this.)
Tell-a-Story (Stimulus situation; card #6. Human CAT.
To the child: Now I am going to show you a picture. Tell me a story
about what you see.
(Probe: Make up a story about it.)
(Probe: Anymore?)
(Probe: Anymore to your story?)

67
APPENDIX C
0. J. Harvey Questionnaire
Now we would like to know your opinions and beliefs about several
other topics. Please write at least two (2) sentences about each topic.
You will be timed on each topic at a pace that will make it necessary for
you to work rapidly. Be sure to write what you genuinely believe. Take
the topics in order, beginning with number 1.
1. This
I believe about marriage .
2. This
I believe about religion .
3. This
I believe about the American Way of Life .
4. This
I believe about friendship .
5. This
I believe about people .
6. This
I believe about guilt .
7. This I believe about women's rights .
8. This
I believe about work .
9. This
I believe about law .
10. This
I believe about education .

68
APPENDIX D
Home-Play Questionnaire
This is a short questionnaire about some of your feelings about play.
Notice that there are possible answers showing about how some people feel
about play. Circle the letter of the one which is closest to how you feel.
If none of them apply write your own answer in the bottom space.
1. When should children play?
a. When chores and work are all completed.
b. A regularly scheduled play time is best.
c. Whenever they want to.
d. Only with parent's permission.
e.
2. Boys should be discouraged from playing with girls' toys and games.
a. Only when the child seems to play with girls' toys to excess, or
more than he plays with boys' toys.
b. Never.
c. Always.
d. Only when there is no adult male or father in the household.
e.
3. A child should share his toys with other children.
a. Always.
b. With visitors and brothers and sisters.
c. If the parent thinks it should be shared.
d. When ever he wants to.
e.
4. Wrestling or rough-housing should be done:
a. Only outdoors or in designated areas.
b. Only when closely supervised by an adult.
c. No restrictions ought to be placed on it.
d. Only by boys.
e.
5. Children's play should be mainly things that teach them useful ideas
and skills.
a. Completely agree.
b. Mildly agree.
c. Completely disagree.
d. Mildly Disagree.
e.

69
6. Make-believe play ought to be encouraged.
a. Mainly when children are younger.
b. As much as possible for all children.
c. Unless the child is doing too much of it.
d. No, it rather should be discouraged.
e.
7. Girls should be discouraged from playing with boys' games and toys.
a. Never.
b. Always.
c. Only when child seems to play with boys' toys to excess or more
than she plays with girls' toys.
d. Only as she grows older if she continues to prefer boys' toys,
things.
e.
8. How should a parent react when the child is using a toy in a wrong
way enjoyably?
a. The child should be stopped and taught the correct way to use
the toy.
b. The incident does not call for any special reaction.
c. Probably ought to warn the child and if misuse of toy persists,
punish.
d.

70
APPENDIX E
Interaction
Here are lots of little toys. Both of you, playing together, use any
of them to build an exciting scene -- as if for a movie -- and then tell
me what is happening. (Permit the building of the scene to occupy five
minutes. Then say, "Now can you tell me what is happening?")

71
Interaction
Child M
Mother figure
Father figure
baby
toy furniture
car
rhi no
lion
horse
policeman
"mother"
"father"
"baby"
blocks
"car"
green thing
red thing
Dumbo
Superman
pipe cleaners
popsickle sticks

72
APPENDIX F
Demographic Information
Father's name
Age
Occupation
Last year
completed schooling
Mother's age
Occupation
Last year completed schooling
Number of children in family
Name
Age
Name
Age
Name
Age
Name
Age
Name
Age
Others living with family
Age
Years of Schooling
Number of rooms in house

73
APPENDIX G
Variables for Correlation Matrix
Col
1
Child's IQ
Col
2
Child's Barron
Col
3
Child's Tell-a-Story
Col
4
Child's Fantasy Toys (used for Interaction)
Col
5
Mother's Age
Col
6
Mother's Years of School
Col
7
Mother's Barron
Col
8
Mother's Tell-a-Story
Col
9
Mother's Fantasy Toys (used for
Interaction)
Col
10
Father's Age
Col
11
Father's Years of School
Col
12
Father's Barron
Col
13
Father's Tell-a-Story
Col
14
Number of Rooms in Home
Col
15
Child's SFPI
Col
16
Child's Realistic Toys (used for
Interaction)
Col
17
Interaction
Col
18
Mother's SFPI
Col
19
Mother's Numbers of Permissives
Col
20
Mother's Number of Restrictives
Col
21
Mother's Realistic Toys (used fo
r Interaction)
Col
22
Father's SFPI
Col
23
Father's Number of Permissives
Col
24
Father's Number of Restrictives
Col
25
Number of Persons in the Household
Col
26
Child's Total Number of Toys
Col
27
Mother's Total Number of Toys
Col
28
Room/Person Ratio

k 4 -W V
4 m iT CC l^l>'a^C>INaCa^ fKfTNCs-CiT'rtTcrocf'cr' -4 e p- r
cc.4p-CCcrar^ifrs.n>£f\ir r^*' nC^'Ci*
C rv C *r c C C C r- <40 C rr* rv C
cecccccccococccecccccoccc cc
I lili I II III
Ac
C O fn c ff o- C 4 <\
OPgiSih-P-C t-
oc o c o c c
I I I I I
V
%r r c if- p- x cr c p-
ccaiaafN-Cf'
ar >f c ^ it 4)
p-r~ n r
coccc occc
I I I I I I I I I
At
p- p-
C -4 P- C f cc
oc (7 fr a C ^ < 0 r
p fv r c r c
ccccccccoc
I II II
* -M. fc 4, 4r
j I m if ir C nr ^5 if *t x cr fV/ n o c a p-
r. r" r. c >rr* tvfN.r\jrv,r-(,nfro--C' O O P" P~
OOOOO OCOOOOOCCCOOOOOOOOCOOC
I II I I I I I I I
M v 4 +
a 4 (T 4 o r if' x> cv fru^(^if vr>
ac 4tf,'Oif'^Cr~ir x X)fVf\ f\,c?xif\;(7'Pi 444 4
^-tr^p-ccp* r^cf'fC'r' C C*'-PgOC COCC <--CM'tO'MMrr^
OOOC OOOOCOOCOCOCCCCOOCCOOCC
II III II I I I I I II
* A* Ac
p- c -r r o- o 4
r>r-p*"4Op-.04 ^.^<*>4
INh-CNIPgNOf,'ffN-l00^i^fV.r
-J 0CCC-*-H^-.MMH'00-r-0'-'<^'*' C 4
*-> oooooooooooooooooocoooooooo
II II I I I I I
*+ 4> n
4,0p-^<*'^r>p'p> ir-4afv.ortr r(p
^40r44>oiPiOP'ngp-rg-ir'or~x3>c.cg
4 p- or- 00pg o-rs. 4P-4 /
ifCCO-rM\^-00-'C''-^0^--00'-- OO OOOOOOOOOOOOOOOOOOOOOOOOO
I II I I I I I I I
+ W V if
Ofit*r-ifrooMfMNiniMO-N
Tor'^-irirn'-rcr^'CCCT-^) cn-o~-o r'
J 00p'inr40M-00-Cir<*'i'0N0N00o Oc*
O 0-00000000000000000000000000
lili I II I I I I I I I II I I
X
o
u
4 4
o 4 a o 4 4
O Mfi o
OC^CSOx
O "t If o O M X
e o o n m w
4 p- 4 r- o
ir ac Of r
OOO-OM-
O-9'nt^rv.t-
n^
^ i*" O M fV(
mx-otTOO-
O .0 ^ pg P-
w c* (? if n o a-
pu pg O
o o o
oooooooooooooooooooooooo
I lilil
Of
o
u
x pg|p4ir.0p^ M*4if C pgpgf'jrgpgpgpgrgrg n
a
COL. COL. COL. COI. CPL COL. mi. CPI
OlfiA 0.0440 0.1171 O.O47 0.??71 ^.4<5?0-r -0.11M
v *v
it O' c k r a-raMfif4
-r r r ff m 4 i' C >1 c P
Cc r c a. -O r 4x4
C X 4 I OOIP>CCC
.............
CCCCC CCOCCCCC
I I I I I
W* vv
K4fnO'ar(t-(tr'4i
44cr'4 - 4 (M i a r If c £> 4
4> C C
CCCCCOOOOCCOC
I I I
4.
< p- o p- c p* r g c c
4cp-4p- 4C
4 1 C C C rv 4 4(N O
CC c r 000*^0 o
.............
ococc cocoooc
I I I I I I
"V 4
ropcraroop-eir
4Cr 434 <^C(\40(\
*'(Cf'4eo'if(rct-px rvcOr*r-rv C. r- rv c
c ccccccocccc c
I I I I I I I
** *
C f c jciff-o'iMirciif'ig
C cv 4 o k r c ir o x
r o -t r c r C p
-coc ccc c ecocoe
ill ii i
V +4.
a IC-C*C* (rif^irrf\.irN(rrK4if
4Mllf.N \ ClP(MMtH
occvo >ro pocp-
coooocoooooooc
i i i ii
M +> k
O f*1 N I if 4 tr rv. 4 O O
(f(vN-oaa40(fC4
a c 4 r* p- tr 4
C n- rv r- 4 r* o f* O
ocoocccococ c
I till
*f Ii
< ir a p o o x-*o<
c POPnNr
4 x 4: r- 4-r-r.ir. rfM4,4
Cr*- ori f'
C COOOOOOCOOOOO
ir r-
O ir tr
O O fN.
CO'CO'-OMCffH
irggififfooffiN
CMN4COO-C
OOCOCCO OC
III I
^ k
4 m tr -m MNiCMlfO
044Ci*cirr O O O O fp in
OOOOOOOOCOOOOO
I I II III I
ifs-44r(Mf4C>f-
inr erffMMOMO-
4)-(fi04^NO OfMOPg OOCOOl*- (*
000000000-000
I II I
* kk *
CP PJ o P- 4 o 4 4amN^
lf'4M4'C-4<4(Mf'ffNO
T r- 4 4 if'(ror\jt-43^o
pg rr O onoPgOO pj
OOOOOOOOOOOOOO
II I I
4 IT PU (* 4 4 o 4 pg
Ph4lT4i IN0N04P
NNP4#OCNO'P-
pg pg O OOCOC
00000000-0000
I II I
Ai Ac M
-44Nh-r>oirif>
4CPONP 4Pt'OH
IT4l4f)N4mP<4-0
-|f OOOONPONOPNO
..............
OOOOOOOOOOOOOO
I III I
4
r- m r--
O O 4 *4 fv.
O rg
O-4P0
a
o o o o o o o
till
pg O tr rg r r
if o pg pm oo *
BOO-4N4
O o O 4 M O
0-00000
I I II
* +*
0'P4)4mio-r'WN0'
- Pg a P -
4PmOPoorpg(riH
OOOOO NOP O
OOOOOOOOOOOOOO
III I III
f\io^ir'^p-(y*o rgo4
if 4) P O' O M|f 4if 4S
Mpgpgpgpgpgpgpgpg

1579 0.3603 0.0867 -0.285? -0.2197 -0.2301 0.1763 -0.2968
fefe
fv. 4 r- 40^4 ct 4 r- 4 ir 4 4 c o r. c
o-(v o >t c < >c^Ka Kica8*ifU'iN>r. n'0'^3CC*- CC *\Cnj cococ cecc ecr-c ccoccccecccoc o
i i i i i i i i i iii ii i ii
(*- (7> r\ >, ir O- C >C 4 c (T C I C CC<--*4^C4iT(i-8C
cr o- 4 ~ c xa.cgiAX.i\. <>.* 4co>cfe cc a-rcN.^r-
o ctr-r'-cC^rv r-acu rv. c w f C'£4Cr.c7
C C c-*- CrCi*"fv. r* \jC*\.crC CC* _j
c
c. ccccceccocooceoc.cooc~ec.ccc. i_
itii ii ii i
4 *
IT c.C4C 4 If h
V O O n a4XMflfo4 if iMCO*'" C 4
r 4eO*fe cr-CO^^CON. CvC _i
c
oococc. ccooococooceeo ccceec _
I I I I I I I I I I 1 II
fe fe *
K tn I** X <4 X! ^ cc -4 IT < M* f KCOff4irir.4'
racr-0>C44r^ 4 x> cv f\. r. X o C O r> f3 4 *v< *>
oa.> 4*ria.f-4r a. r- rv, x r cr^r-**'*!-cc*'-
C ^ fe fef^O O OCOCO Cf^C _J

eooc OOCOCC OCOC.CCCCC CCCCCCC o
I II I lili lili I I
vw v fe* fefe v *
ooflC<'4ccc*'fefe^ir^v 4CON<^l'Mr4
ox)-o^.,oo*'i\.4iiva>4)
.4 p-x>>o*r'oc4X4
NN^0if'(*'(7'00l,'^,(^44'< ocooooococooooooocoocooocc
II I II I I I I I I I I I II II I
fe
HHW4440W-*
ai" p8 ir r 4 i" m f
-i ovco *ooo
cc*******
ur^OOC OCOOCC
I I I
V fe fe
4 4X>**''*r*-r-'44*^00'V*>
4* O C 4)n4f'.iX4N4'rnr4*C
oaff4'-rirv4--(r:4 r o
Oim^Cw*-C'iCP.COCO^'OOC
ooccccoocooccc c o o o
I lili II
V
!* OOMMNM 4 4 If ir X
occec**'r-g-rvxJoor'
o.'4S(T--8X4)^4N
OOO O- O -o-
0000000000000
lili I
M V if
M>(r-o4N(i-4MX4ivir
9^- -crirorr--^a -4 cm tm
hM^00i*104-'(,'0n(on
0000 eoooooooo
1 I III
oxrnroNx^
ir o -j n o 7-
OH4NOIPX-
_J rvjfVf^fN.^tn
o o
OVOOOOOOOOO
I lili I
fe v
i'M4)OeK-r-fv'ir*3^.xancfr
'X>4 X-OO
rv ^Ocg\. Cf-. Or-oOOf-C O
oooooo*-*oooooooooo o
lili III III
^ fe
4)0-Cr'CN44MrN0
*rr cror-ooo4 CfVJXXifXlfrX4CMf
rr rn_. On-CO <~
0000000000000
I II III
V v
IT040-N40-N>^
4NOi'NNP'irx if-O-OO^OO-COOC
000-000000000
I III II
*r
r r^o4ir4)(*'N*'
o NHmO'r^trxo-
ir
C _j ni >'*** n
o
O ofMCOCOOCOOO
I II II
rNcN-Mr4Nr4i"^oocr
CTN4INr4r'r'rN4-. cw rv, r. o 4 4
4rM49^4CN4Kfe4XrOH
rfiH'wff c-8 r-nOf9N-*N-H
000000000000000000
I lili II
C r-
u
*fcfe fe %
crr\i44'ox>^x>04'ar'43r-o^<***' o* r- r-
0'04M4Q4irNoirmiroir 00 oi*r^x>r-
SN40i'r^N444XirH f-0r-0^ O 4 X>
nfrooN-ooo-NN4-o-iM-orH-m-N
0000000000000000-00000000000
I III I I I I I I I I
o^ocroatrs
Cfe O 4 4 4 O fM
roo o m x> o fe
-i O (MO O
or
0*^000000000
II lili
r -* fe
or-i"M)HNifeorccc am
OinO*'r-x>fMfn>cmir>ar'*'r-
00*rri**l 00 C C C O*'
000000000000000000
I lilil I III
-M*i4lT4^B90-M*l4irXK90-M^4irXM
iMfgr^fMfsjrgfSiMfM
- 0-N(r4'4fe0-Mf 4r 4KC
---- NNNWigNrMNN
*
o

76
APPENDIX H
Letter to Parents Requesting Consent
Dear Parent:
This letter is to introduce me and a research project I am conducting
regarding children's play. I am Lorraine Dennis, an instructor at Roger
Williams College. The research will become a part of my dissertation for
the Ph.D. degree at the University of Florida.
For the project, my research assistants (advanced students of
psychology from Roger Williams College) and I will be studying the play
of kindergarten and first-grade children. We will be asking the child
to do the following: 1) make up a story about a picture of some people
who are camping; 2) describe what is seen in a set of ink blots, and 3)
answer four questions about favorite games and play activities. Total
interview time is about twenty minutes. The purpose of the interview is
to obtain information regarding children's preferred ways of playing.
Later, we intend also to study a small number of families to see
whether brothers and sisters are generally alike in their style of play
and what kinds of play the parents most enjoyed when they were young.
These families will be contacted individually and, of course, will be
included only if they are willing to grant permission.
It should be emphasized that the identity and name of each child
(and of any participating family) become lost when combined in the group
measures and statistics. Findings are expressed in percentages, age-
levels, correlations, boys versus girls, trends, etc. No names will
ever appear!
We believe that this research will help provide a better understanding
of the place of play in the life of a child -- and that the results will
be of interest to all of us who live and work with children. Therefore,
we plan to offer a report of the general results of this study to any
interested parents.
If you have any questions concerning any aspect of this study, please
contact me at 255-2105 (my office) or 245-4145 (my home).
We will be deeply grateful for permission to include your child in
this first phase of the study. Please sign either form #1 or form #2
indicating your approval or rejection. An envelope is enclosed for
your convenience; however, the form may be returned to your child's
school. Thank you very much!
LBD/mgd
Enclosures
Lorraine Bradt Dennis
Psychology Area

77
APPENDIX I
Consent Form
form #1
I GIVE PERMISSION FOR MY CHILD
(child's name)
TO BE INTERVIEWED IN CONNECTION WITH THE STUDY OF CHILDREN'S PLAY.
(Parent's or Guardian's Signature)
*************************************************************************
form #2
I DO NOT WISH MY CHILDREN TO BE INTERVIEWED FOR THE STUDY OF
CHILDREN'S PLAY.
(child's name)
(Parent's or Guardian's Signature)

REFERENCE LIST
Barron, F. Threshold for the Perception of Human Movement in Inkblots.
Journal of Consulting Psychology, 1955, J9_, 33-39.
Baumrind, D. Current Patterns of Parental Authority. Developmental
Psychology Monographs, 1971, 4_, 1-103.
Bell, R. Q. A Reinterpretation of the Direction of Effects in Studies of
Socialization. Psychology Review, 1968, 7J5, 81-95.
Bell, R. Q. Stimulus Control of Parent or Caretaker Behavior by Offspring.
Developmental Psychology, 1971, 4_, 63-72.
Bishop, D., & Chace, C. Parental Conceptual Systems, Home Play Environ
ment and Potential Creativity in Children. Journal of Experimental
Child Psychology, 1971 12., 318-338.
Caldwell, B., & Hersler, L. Mother-Child Interaction During the First
Year of Life. Merri 11-Palmer Quarterly 1964, JO^, 119-128.
DeVore, I. Mother-Infant Relations in Free-ranging Baboons. In H. L.
Rheingold (Ed.), Maternal behavior in mammals. New York: Wiley,
1963.
Dewing, K. & Taft, Some Characteristics of Parents of Creative Twelve
Year Olds. Journal of Personality, 1973, 41_, 71-84.
Eifermann, R. Social Play in Childhood. In R. E. Herron & B. Sutton-Smith
(Eds.), Child's Play. New York: Wiley, 1971.
Elkonin, D. Symbolic Representation and its Functions in the Play of
Children. In R. E. Herron & B. Sutton-Smith (Eds.), Child's Play.
New York: Wiley, 1971.
Erikson, E. Play and Actuality in Development. In M. Piers (Ed.), Play
and development. New York: W. W. Norton, 1972.
Freyberg, J. T. Increasing the Imaginative Play of Urban Disadvantaged
Kindergarten Children Through Systematic Training. In J. L. Singer,
The child's world of make believe. New York: Academic Press, 1973.
Gilmore, J. B. Play, a Social Behavior. In R. Haber (Ed.), Current
research in motivation. New York: Holt, Rinehart and Winston, 1966.
78

79
Goodall, J. Nest Building Behavior in the Free Ranging Chimpanzee. Annuals
N. Y, Academy of Science, 1962, 102, 455-467.
Gordon, I. J. Stimulation Via Parent Education. Children, 1969, 1_6, 57-59.
Harvey, 0. J. Beliefs and Behavior: Some Implications for Education. In
The science teacher. Washington, D.C.: National Science Teacher's
Association, 1970.
Harvey, 0. J., Hunt, D. E., & Schroder, H. M. Conceptual systems and
personality organization. New York: Wiley, 1961.
Harvey, 0. J., White, J., Prather, M., & Alter, R. Teachers' Belief
Systems and Preschool Atmospheres. Journal of Educational Psychology,
1966, 57, 373-391.
Hess, R. D., & Shipman, V. C. Early Experience and Socialization of
Cognitive Modes in Children. Child Development, 1965, 3£, 869-886.
Hutt, C. Exploration and Play in Children. In R. E. Herron and B. Sutton-
Smith (Eds.), Child's Play. New York: Wiley, 1971.
Isaacs, S. Social development in young children. New York: Schocken
Books, 1972. (originally published, 1933)
Klinger, E. Development of Imaginative Behavior: Implications of Play
for a Theory of Fantasy. Psycholoaical Bulletin, 1969, 72, 271 -
290.
Labov, W. Academic Ignorance and Black Intelligence. Atlantic Monthly,
1972, 229, 59-67.
Lieberman, N. J. Playfulness and Divergent Thinking: An Investigation
of Their Relationship at the Kindergarten Level. Journal of Genetic
Psychology. 1965, 107, 219-224.
Lytton, H. Creativity and education. New York: Schocken Books, 1972.
MacKinnon, D. W. The Nature and Nurture of Creative Talent. American
Psychologist, 1962, 1_7, 484-495.
Manosevitz, M., Prentice, N., & Wilson, F. Individual and Family Corre
lates of Imaginary Companions in Preschool Children. Developmental
Psychology, 1973, 8, 325-329.
Mason, W. Early Deprivation in Biological Perspective. In V. Denenberg
(Ed.), Education of the infant and young child. New York: Academic
Press, 1970.
Millar, S. The psychology of play. Baltimore: Penquin Books, 1966.
Murphy, L. Infant's Play and Cognitive Development. In M. Piers (Ed.),
Play and development. New York: W. W. Norton, 1972.

80
Piaget, J. Play, dreams and imitation in childhood. New York: W. W.
Norton, 1972. (originally published, 1951)
Pulaski, M. A. Play as a Function of Toy Structure and Fantasy Pre
disposition. Child Development, 1970, 4]_, 531-537.
Pulaski, M. A. Toys and Imaginative Play. In J. L. Singer, Child's
world of make-believe. New York: Wiley, 1973.
Pulaski, M. A. The Rich Rewards of Make-Believe. Psychology Today, 1974,
7, 68-76.
Rebelsky, F. Mother-Infant Interaction. In F. Rebelsky & L. Dorman (Eds.)
Child Development & Behavior. New York: Knopf, 1974.
Repina, T. A. Development of Imagination. In A. V. Zaporoahets & D.
Elkonin (Eds.), The psychology of preschool children. Cambridge,
Mass.: The MIT Press, 1971.
Saltz, E. & Johnson, J. Training for Thematic-Fantasy Play in Culturally
Disadvantaged Children: Preliminary Results. In Studies in intel
lectual development. Detroit: Wayne State University, 1973.
Schaefer, C. E. Imaginative Companions and Creative Adolescents.
Developmental Psychology, 1969, (5, 747-749.
Singer, J. L. Imagination and Waiting Ability in Young Children. Journal
of Personality, 1961, 29, 396-413.
Singer, J. L. The child's world of make-believe. New York: Academic
Press, 1973.
Smilansky, S. The effects of socio-dramatic play on disadvantaged pre
school children. New York: Wiley, 1969.
Stone, G. P. The Play of Little Children. In R. E. Herron and B. Sutton-
Smith (Eds.), Child's play. New York: Wiley, 1971.
Sutton-Smith, B. The Role of Play in Cognitive Development. Young
Children, 1967, 6, 202-214.
Sutton-Smith, B. Introduction. In E. Avedon and B. Sutton-Smith (Eds.),
The study of games. New York: Wiley, 1971.
Weikart, D., Rogers, L., & Adcock, C. Cognitively Oriented Curriculum:
a Framework for Preschool Teachers. Urbana, Ill.: University of
Illinois, 1970.
Weisskopf, E. A Transcendence Index as a Proposed Measure on the TAT.
Journal of Personality, 1950, 29_, 379-390.
Yarrow, M. R. Problems of Methods in Parent-Child Research. Child
Development, 1963, 34, 215-226.

REFERENCE NOTES
1. Ellis, M. Play, a paradox for teacher and scientist. Paper presented
at the Second Annual Symposium on Play, Atlanta, Georgia, April,
1974.
2. Kempler, B. Welcome and introductory remarks. Paper presented at
the Second Annual Symposium on Play, Atlanta, Georgia, April,
1974.
3. Singer, D., & Singer, J. L. Television, adult modeling and physical
structure of the setting as factors enhancing play in pre-school
children. Paper presented at the Second Annual Symposium on
Play, Atlanta, Georgia, April, 1974.
4. Branch, A. Personal communication. December, 1973.
5. Harvey, 0. J. "This I Believe" Scoring Instructions, 6/71 and 9/71.
Mimeographed handout at address by 0. J. Harvey, University of
Florida, January, 1972.
81

BIOGRAPHICAL SKETCH
Lorraine Bradt Dennis was born May II, 1921, in Norway, Michigan,
attended public schools in Hibbing, Minnesota and graduated in 1938. She
earned a combination degree, B.S., R.N., at the University of Minnesota
in 1943 and a Master of Science in Psychology at Kansas State College in
1951.
Her professional career has included teaching assignments at Kansas
State College, Drake University, the Pennsylvania State University and
Marymount College of Virginia. In addition, she is the author of a
text book, Psychology of Human Behavior for Nurses, published by W. B.
Saunders Co. and was one of the founders of the Child Study Center in
Caracas, Venezuela, designed to offer psychological and psychiatric
services to English language schools in that city.
She is a member of the American Psychological Association, the
Society for Research in Child Development, the American Educational
Research Association, Sigma Theta Tau, an honorary nursing society, and
Phi Kappa Phi, an honorary scholastic society.
She is, at present, a member of the faculty of Roger Williams College,
Bristol, Rhode Island and serving as coordinator of the Psychology Area.
82

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Ira J. Gorqgln, Chairman
Graduate Research Professor
Foundations of Education
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the^gree of
Doctor of Philosophy. / /
R. Emile Jest
Associate Professor
Foundations
of Education
Education
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
y-fe.
Barrynf. LesU
Assistant Professor of Psychology
This dissertation was submitted to the Graduate Faculty of the College
of Education and to the Graduate Council, and was accepted as partial
fulfillment of the requirements for the degree of Doctor of Philosophy.
March, 1976

'ty opinion it
r.m.y and is fully
* 'V tr :ooree of
V./t/of the College
Kt! as cartial
/y of Philosophy.
m¡







44
TABLE 11
MEAN NUMBER OF PERMISSIVE AND RESTRICTIVE RESPONSES
Response
Mother
L.F.
H.F.
Father
L.F.
H.F.
Permissive
2.571
3.058
1.857
2.058
S.D.
1.78
1.71
1.62
1.14
t Test (36)
N.S.
N.S.
Restrictive
1.048
.352
.857
.823
S.D.
1.07
.606
.964
.951
t Test (36)
4.694*
N.S.
*p<.0001
are correlated negatively and significantly with mothers' and fathers'
restrictives. Fathers' permissives are correlated significantly only
with fathers' Tell-a-Story score.
Although the number of restrictive responses from fathers does not
differentiate between LF and HF groups, the mothers of LF children give
significantly more restrictive responses than mothers of HF children.
Support for the hypothesis, therefore, derives from the indication
that although LF and HF children seem to come from homes which are
similar in terms of permissive responses of both fathers and mothers,
the mothers of these children differ significantly with regard to the
number of very restrictive attitudes toward play which they express.
Mothers of HF children appear to be less restrictive in attitudes about
home-play than mothers of LF children.
The correlation matrix (Appendix G) shows that the mothers' re-
strictives are significantly related to the child's SFPI and also the
several other variables. The significant correlations are abstracted
for Table 12. Fathers' restrictives obtain only two significant


Mr. Henry V. Diodati, superintendent of the Portsmouth, R.I., schools
and the four principals of the elementary schools, Mr. Lawrence P. Mello,
Mr. Lawrence Jones, Mr. Robert L. Crudup and Mr. Russell L. Carter must
be thanked for their generous help in making rooms available for our
testing and facilitating our efforts. Teachers of the kindergarten and
first grades of these schools, too, were magnificently tolerant of the
many class interruptions as we called for and returned the children during
days of testing.
I wish also to warmly acknowledge my gratitude to the wonderful
children and the splendid families who gave so generously of their time
and even more importantly, so promoted this project with their interest.
Finally, it is impossible to express how much this entire undertaking
owes to the steady, optimistic encouragement of Larry Dennis and our
children, Patrick, Brian, Deborah and Thomas, each of whom in uncounted
ways enabled my study and research.
i i i


22
with young children (Pulaski, 1970; Singer, 1973; Freyberg, 1973). The
score is the number of the inkblot for which the child first responds with
human movement association. It should be noted that the lower the score
on this test, the earlier the M response has occurred, and hence the higher
the fantasy pre-disposition.
2. The Singer Structured Fantasy Pre-disposition Interview (SFPI).
A series of four questions developed by Singer (1973) which inquire into
the child's favorite play activities (Appendix B). Described by Singer
as a "clinical interview," the examiner is urged to use all skill in a
attempting to ascertain whether there are "pretend" or fantasy elements
in the child's responses. Because of the large number of examiners,
however, it was determined to hold procedures as standard as possible,
even at the risk of missing fantasy which might actually be present, in
an effort to guard against the possibility of suggesting responses to the
child. Therefore examiners were permitted to probe only with a question
which requested "Tell me more about that."
Each interview question is scored with one point if fantasy or pre
tend is judged to be present in the response. The highest possible score
for the SFPI would be 4, while the lowest possible is 0.
3. Tell-a-Story: Stimulus Situation, Card #6, Human CAT test. The
stimulus situation for the Tel1-a-Story, the sixth card from the Child
ren's Appreception Test (CAT), Human version, depicts two covered, sleeping
figures on the ground and a smaller person, head upraised and awake, in
the foreground. The child is asked to make up a story about the picture.
The story is recorded verbatim and scored according to the Transcendence
Index developed by Weisskopf (1950) which awards one point for every
element which is supplied by the story-teller, as opposed to those which


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Ira J. Gorqgln, Chairman
Graduate Research Professor
Foundations of Education
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the^gree of
Doctor of Philosophy. / /
R. Emile Jest
Associate Professor
Foundations
of Education
Education
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
y-fe.
Barrynf. LesU
Assistant Professor of Psychology
This dissertation was submitted to the Graduate Faculty of the College
of Education and to the Graduate Council, and was accepted as partial
fulfillment of the requirements for the degree of Doctor of Philosophy.
March, 1976


76
APPENDIX H
Letter to Parents Requesting Consent
Dear Parent:
This letter is to introduce me and a research project I am conducting
regarding children's play. I am Lorraine Dennis, an instructor at Roger
Williams College. The research will become a part of my dissertation for
the Ph.D. degree at the University of Florida.
For the project, my research assistants (advanced students of
psychology from Roger Williams College) and I will be studying the play
of kindergarten and first-grade children. We will be asking the child
to do the following: 1) make up a story about a picture of some people
who are camping; 2) describe what is seen in a set of ink blots, and 3)
answer four questions about favorite games and play activities. Total
interview time is about twenty minutes. The purpose of the interview is
to obtain information regarding children's preferred ways of playing.
Later, we intend also to study a small number of families to see
whether brothers and sisters are generally alike in their style of play
and what kinds of play the parents most enjoyed when they were young.
These families will be contacted individually and, of course, will be
included only if they are willing to grant permission.
It should be emphasized that the identity and name of each child
(and of any participating family) become lost when combined in the group
measures and statistics. Findings are expressed in percentages, age-
levels, correlations, boys versus girls, trends, etc. No names will
ever appear!
We believe that this research will help provide a better understanding
of the place of play in the life of a child -- and that the results will
be of interest to all of us who live and work with children. Therefore,
we plan to offer a report of the general results of this study to any
interested parents.
If you have any questions concerning any aspect of this study, please
contact me at 255-2105 (my office) or 245-4145 (my home).
We will be deeply grateful for permission to include your child in
this first phase of the study. Please sign either form #1 or form #2
indicating your approval or rejection. An envelope is enclosed for
your convenience; however, the form may be returned to your child's
school. Thank you very much!
LBD/mgd
Enclosures
Lorraine Bradt Dennis
Psychology Area


70
APPENDIX E
Interaction
Here are lots of little toys. Both of you, playing together, use any
of them to build an exciting scene -- as if for a movie -- and then tell
me what is happening. (Permit the building of the scene to occupy five
minutes. Then say, "Now can you tell me what is happening?")


12
characterized the relationship with both parents.
.... Thus, if there was a certain distance in the
relationship between parent and child, it had a
liberating effect so far as the child was concerned.
If he lacked something of the emotional closeness
which some children experience with their parents,
he was also spared the psychological exploitation
that is so frequently seen in the life histories of
clinical patients (pp. 491-492).
Singer (1973), commenting along the same lines, suggests:
... an optimal balance of benign parental contact
and opportunity to be alone seems therefore essential
to the development of a rich imaginative life. This
optimal situation occurs most readily where the child's
mother is relatively warm, willing and capable of
spending time with the child, but not so emotionally
involved that she cannot at times leave the child to
his own devices (p. 62).
According to Singer (1973) another family variable consists of the
kinds of activities in which the child and parent engage together. He
suggests that parents who read or tell stories to their children are
providing a different kind of medium than those who spend time with their
children playing games of skill or chance.
In a study which provided some important conceptual leads for the
present investigation, Bishop and Chace (1971) argue that parental
personality qualities most likely to encourage children's playfulness at
home do not exist as isolated traits. Rather, they suggest, personality
theory and research points to a coherence in personality attributes.
Drawing upon the long series of publications by Harvey and associates
(1961, 1966, 1970), they invoke the proposition that certain personality
qualities are manifested in a basic "conceptual system," with differing
levels of organization which can be ordered along a dimension of con
creteness-abstractness. Four principle modes of conceptual functioning
are described in terms of the characteristic beliefs and response
patterns.


69
6. Make-believe play ought to be encouraged.
a. Mainly when children are younger.
b. As much as possible for all children.
c. Unless the child is doing too much of it.
d. No, it rather should be discouraged.
e.
7. Girls should be discouraged from playing with boys' games and toys.
a. Never.
b. Always.
c. Only when child seems to play with boys' toys to excess or more
than she plays with girls' toys.
d. Only as she grows older if she continues to prefer boys' toys,
things.
e.
8. How should a parent react when the child is using a toy in a wrong
way enjoyably?
a. The child should be stopped and taught the correct way to use
the toy.
b. The incident does not call for any special reaction.
c. Probably ought to warn the child and if misuse of toy persists,
punish.
d.


80
Piaget, J. Play, dreams and imitation in childhood. New York: W. W.
Norton, 1972. (originally published, 1951)
Pulaski, M. A. Play as a Function of Toy Structure and Fantasy Pre
disposition. Child Development, 1970, 4]_, 531-537.
Pulaski, M. A. Toys and Imaginative Play. In J. L. Singer, Child's
world of make-believe. New York: Wiley, 1973.
Pulaski, M. A. The Rich Rewards of Make-Believe. Psychology Today, 1974,
7, 68-76.
Rebelsky, F. Mother-Infant Interaction. In F. Rebelsky & L. Dorman (Eds.)
Child Development & Behavior. New York: Knopf, 1974.
Repina, T. A. Development of Imagination. In A. V. Zaporoahets & D.
Elkonin (Eds.), The psychology of preschool children. Cambridge,
Mass.: The MIT Press, 1971.
Saltz, E. & Johnson, J. Training for Thematic-Fantasy Play in Culturally
Disadvantaged Children: Preliminary Results. In Studies in intel
lectual development. Detroit: Wayne State University, 1973.
Schaefer, C. E. Imaginative Companions and Creative Adolescents.
Developmental Psychology, 1969, (5, 747-749.
Singer, J. L. Imagination and Waiting Ability in Young Children. Journal
of Personality, 1961, 29, 396-413.
Singer, J. L. The child's world of make-believe. New York: Academic
Press, 1973.
Smilansky, S. The effects of socio-dramatic play on disadvantaged pre
school children. New York: Wiley, 1969.
Stone, G. P. The Play of Little Children. In R. E. Herron and B. Sutton-
Smith (Eds.), Child's play. New York: Wiley, 1971.
Sutton-Smith, B. The Role of Play in Cognitive Development. Young
Children, 1967, 6, 202-214.
Sutton-Smith, B. Introduction. In E. Avedon and B. Sutton-Smith (Eds.),
The study of games. New York: Wiley, 1971.
Weikart, D., Rogers, L., & Adcock, C. Cognitively Oriented Curriculum:
a Framework for Preschool Teachers. Urbana, Ill.: University of
Illinois, 1970.
Weisskopf, E. A Transcendence Index as a Proposed Measure on the TAT.
Journal of Personality, 1950, 29_, 379-390.
Yarrow, M. R. Problems of Methods in Parent-Child Research. Child
Development, 1963, 34, 215-226.


INDIVIDUAL AND FAMILIAL CORRELATES OF
CHILDREN'S FANTASY PLAY
By
LORRAINE BRADT DENNIS
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1976


20
TABLE 2
FATHERS' OCCUPATIONS
Group
Frequency
Occupation
Low Fantasy
6
Engineers
3
Navy Officers
1
Retired Navy Officer
1
Lawyer
1
Insurance Executive
1
High School Teacher
3
Financial Analysts
1
Contractor
1
Pastor
1
Truck Driver (college
student)
1
Barber (college student)
1
Steam Fitter (college
student)
High Fantasy
6
Engineers
2
Navy Officers
1
Retired Navy Officer
1
Lawyer
I
Insurance Executive
1
High School Teacher
1
Electronic Technician
(no college work)
1
Anesthetist
I
Technical Buyer
1
Computer Programmer
1
Brick Layer (no college
work)


BIOGRAPHICAL SKETCH
Lorraine Bradt Dennis was born May II, 1921, in Norway, Michigan,
attended public schools in Hibbing, Minnesota and graduated in 1938. She
earned a combination degree, B.S., R.N., at the University of Minnesota
in 1943 and a Master of Science in Psychology at Kansas State College in
1951.
Her professional career has included teaching assignments at Kansas
State College, Drake University, the Pennsylvania State University and
Marymount College of Virginia. In addition, she is the author of a
text book, Psychology of Human Behavior for Nurses, published by W. B.
Saunders Co. and was one of the founders of the Child Study Center in
Caracas, Venezuela, designed to offer psychological and psychiatric
services to English language schools in that city.
She is a member of the American Psychological Association, the
Society for Research in Child Development, the American Educational
Research Association, Sigma Theta Tau, an honorary nursing society, and
Phi Kappa Phi, an honorary scholastic society.
She is, at present, a member of the faculty of Roger Williams College,
Bristol, Rhode Island and serving as coordinator of the Psychology Area.
82


TABLE 9
CORRELATIONS BETWEEN CHILD'S SCORES ON FANTASY PRE-DISPOSITION MEASURES
AND PARENTS' SCORES ON FANTASY PRE-DISPOSITION, RESTRICTIVES AND PERMISSIVES
Mother's
Barron
Mother's
Tell-a-
S to ry
Mother's
SFPI
Father's
Barron
Father's
Tell-a-
S to ry
Father'
SFPI
s Mother's
Permis-
sives
Mother's
Restric-
tives
Father's
Permis-
sives
Father's
Restric-
tives
Child's
Barron -133
-.147
-.304
.005
.093
.210
-.020
.241
-.019
-.071
Tel1-a-Story
-.224
.214
.327*
.057
.132
.197
.048
.217
-.029
.052
SFPI -.389*
.139
.369*
.104
.160
-.076
.111
-.420*
.224
-.080 -
*p<.05 r = .320 two tail
df = 36


11
children's make-believe play (Freyberg, 1973) is based on a very small
sample. Another (Singer, 1973) reports on some parental practices which
were found to be important as a result of interviewing children. The
other studies were undertaken with a different focus -- imaginary com
panions for Manosevitz, Prentice and Wilson (1973), and creativity for
MacKinnon (1962), Bishop and Chace (1971), and Dewing and Taft (1973).
Nevertheless, findings from these studies are provocative and, taken to
gether, seem to provide some fruitful leads for further investigation.
There is a persistent finding of the importance of birth order with
first born and only children showing greater tendencies toward fantasy
which has turned up in early studies of daydreaming (Singer, 1973) and
appears for imaginary companions (Manosevitz et^ aj_., 1973) and children
who play most imaginatively (Freyberg, 1973), but not for creative twelve-
year-olds (Dewing and Taft, 1973).
Level of parental education seemed to be important as correlated
with more make-believe play (Freyberg, 1973), but only the mother's level
of education was significant for creative twelve-year-old children (Dewing
and Taft, 1973).
Singer (1973) and Freyberg (1973) both report the possibility that
opportunity for privacy may have some relationship to the child's tendency
to play imaginatively. A possible allied finding is that of Dewing and Taft
(1973) and Freyberg (1973) that more mothers of creative and imaginative
children in their respective samples worked outside the home.
MacKinnon (1962), describing the childhood of creative architects,
noted that:
There was often a lack of intense closeness with one
or both parents. Most often this appears in relation
to the father rather than the mother, but often it


65
APPENDIX A
Directions for Testing
Barron's Movement Threshold Inkblots
To the child: I am going to show you some cards with designs on them.
They aren't made to look like anything -- they are just designs that were
made by dropping ink on paper and folding it. I want you to look at each
of the designs and tell me what it reminds you of -- what it looks like
to you. Every child can see different things and there aren't any right
or wrong answers. Sometimes one card may look like more than one thing
but I only want you to tell me one thing that you see on each card.
(To the tester: If a child gives more than one response to a card
repeat that you only want one answer per card and ask him to tell you
which of his answers that design reminds him of most.)
Peabody Picture Vocabulary Test (PPVT)
To the child: "I have some pictures to show you." (Turn to example
A and say:) "See there are four pictures on this page. Each of them is
numbered, (point to each one) I will say a word, then I want you to point
to the picture which best tells me the meaning of the word. Let us try
one. Point to the picture that best tells the meaning of 'crib.'" (After
the child responds correctly, then turn to example B saying:) "That's
fine. Which picture is fin?" (Then to C.) "What picture is 'butterfly?'
Fine! Now I'm going to show you some other pictures. Each time I say
a word you point to the picture which best tells the meaning of a word.
As we go through the book you may not be sure you know the meaning of
some of the words, but I want you to look carefully at all of the pictures
anyway and choose the one you think is right. What picture is ?"
To the examiner: Be sure that each child looks at each of the
alternatives because sometimes younger children will start looking in
only one location. Also, always be sure to get a response for every
card. If the child doesn't know the word, have him give it a try anyway.
You may encourage the child after some responses by saying "good," "fine,"
etc., but don't do it just after correct responses. You are praising
the child's effort not its correctness. You may repeat a word if necessary.
Do not precede any word with an article (a, an, the) and do not make any
singular words plural.
Scoring: Start with item #25 for all children. Continue forward
until the child makes his first error. If this is before he gets 8
correct, go back to a lower level (below 25). The ceiling score is
determined by the point at which the child has missed 6 out of 8 words.


33
TABLE 6
TEST RE-TEST CORRELATION COEFFICIENTS:
FANTASY PRE-DISPOSITION MEASURES
Test r
Barron .274
SFPI .750**
Tell-a-Story .616*
*p<.01 d.f. = N-2 = 14
**p<.001
Thus it is evident that the reliability of the Barron is quite
doubtful, while that of the SFPI and Tell-a-Story proved to be much more
substantial.
5. Reliability of the Interaction observations. The Interaction
between mother and child was recorded on two kinds of protocols (Appendix
E). One check-list was used to record the toy used in the scene and by
whom it was used. The second obtained a verbatim account of "what is
happening" in the scene. In every instance observations were conducted
by the principal investigator. On five occasions a second observer
maintained a separate check-list of toy utilization and recorded the
description of the scene. The toy lists and Interaction scene descriptions
were counted and scored in the prescribed manner and compared with those
of the principal investigator.
The check-list of toys and who used them proved less reliable than
the account of "what is happening" which furnished the Interaction score.
The criterion for checking a toy as used by mother or child was that the
toy should be placed into and used in the scene. However, on a number of
occasions a toy first used by one of the players might be picked up and
used further or in another fashion by the other player. Discrepancies


CHAPTER V
DISCUSSION
This study was undertaken in an effort to discover possible relation
ships between individual and family-background factors and children's
make-believe play. Because there have been so few investigations of this
area, it must be considered exploratory in design and concept. Because
the number of subjects is so small in relation to the number of variables
the data must be interpreted with extreme caution.
The sample is limited not only in size but also in composition; it
is white, middle class, small family and lives on a large island off the
coast of Rhode Island. It contains an unusual number of Navy families
because of the large defense installations at Portsmouth and Newport.
Inferences from this sample to a general population can be taken only
as suggestions for further investigation.
However, the homogeneity of the sample has an advantage of con
trolling for socio-economic class differences which could otherwise con
found findings. These children come from homes with ample space for play
and with easy access to beaches, woods and meadows for a wide variety
of summer and winter play activities.
Notable was the deep interest of both LF and HF parents in their
children and in the study. On the basis of pilot trials approximately
one hour had been allotted for the family interviews. It quickly became
apparent that the time was greatly underestimated. Parents wanted to
53


28
A check-list of the toys with separate space for recording mother or
child use enabled the observer to record the particular toy and the user.
Mother and child were scored separately for the number of toys of either
category which each contributed to the scene. Only toys which were actually
used in the scene were counted, yielding a "realistic toy" and a "fantasy
toy" score for each.
The description of "what is happening" in every instance reported by
the child, was recorded verbatim and then read back to the pair for
additions or corrections. The description was scored according to the
criteria listed below by three judges, independently, with agreement of
two out of three or all three constituting the reliability criterion, to
yield an Interaction score of 1 to 5 for the mother and child playing
together. The higher score reflects more make-believe or fantastic
elements.
Score
Criteria (Adapted from Pulaski, 1970)
(1) Familiar everyday experience
Concrete
Ordinary
Unelaborated
Use of realistic toy in realistic, unelaborated role
or function -- for example, the car is used as a
car or the policeman as a policeman.
(2) Indirect experience ... TV characters, movie
characters, story book characters and/or events.
Relatively unelaborated narrative.
(3) Content and characters farther removed from
experience or. more elaborated.
Might include "silly" or aggressive or emotional
fantasy.
(4)
More imaginative, unusual or original than #3 above.
Fantasy elements more detailed.


57
restrictives scattered?
In contrast to the Bishop and Chace (1971) investigation, the present
study was unable to obtain evidence of differences in abstract versus con
crete conceptual systems for LF and HF parents. This might be explained
by the relatively high level of education for parents in the sample, with
more parents categorized as abstract (46) than were rated as concrete (30).
It must also be suggested that in spite of earnest efforts to adhere to the
scoring criteria, the scoring of the This I Believe protocols may have
been faulty. Such a possibility would also account for the low reliability
found for the score procedure.
The level of parental education has been cited as important for
children's make-believe play by Freyberg (1973), while mothers' education
but not fathers' was significant for Dewing and Taft's (1973) study of
creative twelve-year-olds. This investigation confirmed the importance
of mothers' education, which correlated significantly with the child's
Barron and SFPI. Her level of education is also found to relate to a
less restrictive home-play environment. A question for further research
might inquire as to what aspects of higher education promote a less
restrictive orientation of the mother toward home play.
Additionally, fathers' education is found, in this sample, to re
late to the number of rooms in the household which, in turn, seems to
promote a lessened number of constraints on play imposed by the mother.
This could, perhaps, be another indication of fathers' indirect contri
bution, with his level of education enabling the material advantage of
a larger house and thereby securing a more relaxed atmosphere for mothers'
attitudes toward play. This kind of linking of correlations is, admittedly,
highly speculative. Nevertheless, it could provide indications of


47
recorded on Table 12, a negative association with the number of mothers'
restrictive responses (r = -.346).
Chi square analysis of the number of mothers working either part time
or full time (grouped together for the analysis since the number of full
time working mothers in this sample is so small) revealed no basis for
assuming a difference for LF or HF groups (Table 7).
6. The sixth hypothesis stated the proposition that mothers and
children who play together imaginatively enable a prediction of high
fantasy pre-disposition for the children. Operationally, it anticipated
a positive relationship between the Interaction score on the mother-child
play activity and the level of the child's fantasy pre-disposition.
The hypothesis was substantially supported.
Table 14 records the mean Interaction scores of LF and HF groups
together with the _t_test for significance of the difference between means.
It can be seen that the Interaction score distinguished between LF and
HF groups with a t test which is significant at the .0003 level.
TABLE 14
MEAN SCORES FOR INTERACTION (MOTHER-CHILD ACTIVITY)
Low Fantasy
High Fantasy
Mean Score
2.095
3.118
S.D.
t test (36)
-3.986*
1.04
1.11
*p<.0003
The Interaction score obtained significant correlation coefficients
with all three of the child's fantasy pre-disposition measures as re
ported in Table 15 (abstracted from the complete correlation matrix).


45
correlations, a positive relationship with the mothers' restrictives
and negative correlation with mothers' permissives.
TABLE 12
VARIABLES SIGNIFICANTLY CORRELATED WITH MOTHERS' RESTRICTIVES
Mothers' Restrictives
Child's SFPI
-0.420
Mother's Barron
0.391 (scored in reverse direction)
Mother's SFPI
-0.379
Mother's Tell-a-Story
-0.373
Mother's years of school
-0.564
Father's years of school
-0.371
Number of Rooms in home
-0.417
Father's restrictives
0.404
Room/Person ratio
-0.346
The relationship between mothers' restrictives and the child's
fantasy pre-disposition was further explored by means of multiple regres
sion analysis, described in a subsequent section.
4. The hypothesis that the children in this study would resemble
their brothers and sisters on the measures of fantasy play was not upheld.
Thirty-two siblings, with an average age of 7.70 years (SD = 2.85),
were administered the fantasy pre-disposition measures and the scores
correlated with those of their siblings in the sample. The correlation
matrix is tabulated in Table 13.
Not one of the child's fantasy pre-disposition measures correlated
with any of those of siblings.
The correlations among the child's own fantasy pre-disposition scores
would be expected because of the selection process for the high and low
fantasy groups. Only children with tested siblings were entered for
this matrix.


CHAPTER II
REVIEW OF THE LITERATURE
A search of the literature reveals that most contemporary investiga
tions have proceeded on the basis of assumptions that play (especially
imaginative play) is important for future development and that children
learn (or fail to learn) certain modes of play in terms of certain kinds
of environmental influences. However, it would be remiss to omit mention
of the fact that the most systematic theory of play to date contradicts
such contentions.
Piaget's Theory of Play
For Piaget (1951) play is a kind of inconsequential by-product of the
really important organismic business of adaptation. Play occurs when
schemas have become so familiar that the adaptive processes of assimila
tion and accommodation are not invoked; the activity is "merely for
pleasure, accompanied by smiles and even laughter, and without the expec
tations of results characteristic of the circular reactions through which
a child learns" (p. 90). Play, therefore, is not accorded any part in
the construction of intelligence. As a matter of fact, play proceeds
only when efforts toward adaptation are relaxed.
Nor, for Piaget, does imaginative play achieve any special signifi
cance. Rather, it is considered an invariant, the intermediate stage
between practice games of infancy and the games-with-rules of later life.
Furthermore, since Piaget views symbolic play as occurring precisely
5


2
Ellis (Note 1) points up a further paradox which hampers experimental rigor
-- the more the investigator gains control over a particular behavior, the
less the behavior can be playful. Klinger (1969), attending to the problem
of definition, asserts, "... the relevant behavioral domain is simply
insufficiently mapped to support the setting of rigorous positive bound
aries on the concept of play"(p. 279).
The task of mapping the "relative domain" of play is, however, be
ginning to attract investigative attention. A major impetus has come from
the proposition that play might have important consequences for later
development. Reminiscent of those who, early in the century, saw play as
practice for later proficiency, present-day ethologists are suggesting
that play is neither antic nor idle, but rather, essential for acquiring
the skills which will be necessary for survival (DeVore, 1963; Goodall,
1962). The use of sticks for the retrieval of termites, for example, does
not seem to be learned by the isolated ape who has no opportunity to play
with sticks, even when exposed to competent models (Mason, 1970). Kempler
(Note 2) has recently extended the notion by suggesting that the playing
child, creating a repertoire of divergent responses, is building a kind of
"future shock" absorber and preparing for the unpredictable.
The idea that play is useful was a major theme for the second annual
symposium on play, convened in Atlanta in April, 1974, for "people who
share the conviction that play is a rich and natural source of creativity,
learning and therapy." The title of the symposium was The Player and the
Environment and the program invited participants to consider "the role an
individual's environment has in encouraging or suppressing play in a child
or adult."
Among the featured speakers, Dr. Jerome Singer and Dr. Dorothy Singer


6
because of the child's inability to think rationally and logically, he
would probably be mystified as to why anyone would even try to teach it.
Play, in the eyes of Piaget, is the occupation of the immature child, re
flecting his as yet inefficient modes of thought and is naturally abandoned
along with his childish views of reality as concrete and formal operations
are accomplished.
Piaget's theory has been criticized for his apparent insistence that
symbolic play is a symptom of immaturity, a kind of "mental digestive
system" (Millar, 1966, p. 50) whereby reality can be assimilated wholly
in terms of the child's own egocentric preoccupations of the moment. This
would seem to relegate adult fantasy and divergent thinking to the same
realm, implying that creative adults have simply failed to grow up.
However, no theory as detailed and cohesive has emerged as an alter
native. Learning theorists have not attended to play, though it can be
suggested that no new paradigm would have to be proposed to account for
its appearance and persistence in the life of a child. As an instrumental
response it produces satisfying changes in the environment -- baby shakes
the rattle and the small child rolls the bright ball. Parental approval
of activities would provide differential reinforcement. Imitation of
models in the environment as well as instruction might direct the child
toward acquiring the various skills and interest in sports or, contrari
wise, toward pursuits requiring imagination. These notions explicitly or
implicitly underly most current empirical research on play. Several dif
ferent kinds of environmental "input," seen as influencing play, have
been investigated.
Play is Learned -- Environmental Influences


79
Goodall, J. Nest Building Behavior in the Free Ranging Chimpanzee. Annuals
N. Y, Academy of Science, 1962, 102, 455-467.
Gordon, I. J. Stimulation Via Parent Education. Children, 1969, 1_6, 57-59.
Harvey, 0. J. Beliefs and Behavior: Some Implications for Education. In
The science teacher. Washington, D.C.: National Science Teacher's
Association, 1970.
Harvey, 0. J., Hunt, D. E., & Schroder, H. M. Conceptual systems and
personality organization. New York: Wiley, 1961.
Harvey, 0. J., White, J., Prather, M., & Alter, R. Teachers' Belief
Systems and Preschool Atmospheres. Journal of Educational Psychology,
1966, 57, 373-391.
Hess, R. D., & Shipman, V. C. Early Experience and Socialization of
Cognitive Modes in Children. Child Development, 1965, 3£, 869-886.
Hutt, C. Exploration and Play in Children. In R. E. Herron and B. Sutton-
Smith (Eds.), Child's Play. New York: Wiley, 1971.
Isaacs, S. Social development in young children. New York: Schocken
Books, 1972. (originally published, 1933)
Klinger, E. Development of Imaginative Behavior: Implications of Play
for a Theory of Fantasy. Psycholoaical Bulletin, 1969, 72, 271 -
290.
Labov, W. Academic Ignorance and Black Intelligence. Atlantic Monthly,
1972, 229, 59-67.
Lieberman, N. J. Playfulness and Divergent Thinking: An Investigation
of Their Relationship at the Kindergarten Level. Journal of Genetic
Psychology. 1965, 107, 219-224.
Lytton, H. Creativity and education. New York: Schocken Books, 1972.
MacKinnon, D. W. The Nature and Nurture of Creative Talent. American
Psychologist, 1962, 1_7, 484-495.
Manosevitz, M., Prentice, N., & Wilson, F. Individual and Family Corre
lates of Imaginary Companions in Preschool Children. Developmental
Psychology, 1973, 8, 325-329.
Mason, W. Early Deprivation in Biological Perspective. In V. Denenberg
(Ed.), Education of the infant and young child. New York: Academic
Press, 1970.
Millar, S. The psychology of play. Baltimore: Penquin Books, 1966.
Murphy, L. Infant's Play and Cognitive Development. In M. Piers (Ed.),
Play and development. New York: W. W. Norton, 1972.


CHAPTER IV
RESULTS
Differences Between LF and HF Groups
The tests for significance of differences between LF and HF groups
are reported in Table 7.
Neither sex differences, nor that of ordinal position nor the grade
level of the child discriminated significantly between the LF and HF
groups.
The difference in the Peabody Picture Vocabulary IQ was found to be
highly significant, despite an effort to control for IQ effects. All
children with an IQ of less than 100 had been eliminated from the selection
process. Nevertheless, the LF group had a mean vocabulary-IQ of 117.27
(SD = 8.87) while that of the HF group was 125.94 (SD = 9.73). The t
test with 36 df is significant at p = .0005.
The age of the parents did not differentiate between LF and HF groups.
The number of persons in the household, the number of rooms in the home,
the number of mothers employed outside the home were also not relevant.
Although fathers' years of schooling were not significantly different
from group to group, the mothers' years of schooling are significantly
higher for HF children. Mothers of HF children average 14.88 years of
36


29
Score Criteria (continued)
(5) Fantastic
Farthest removed from present time, space or reality.
Most original, richest imagination. Elaborate details.
5. Demographic data (Interview) (Appendix F). The interview, con
ducted with either mother or father, ascertained the age of each parent,
the number of years of schooling for each, the occupation of the father,
the occupation of the mother, the number of persons living in the house
hold and the number of rooms in the house.
Reliability studies
Because of the large group of individuals engaged in both testing
and scoring, because certain aspects of scoring involved judgments and
because the reliability of certain of the fantasy pre-disposition measures
does not seem to have been well-documented, several different reliability
studies were conducted during the course of the investigation.
1. Reliability of the testing (fantasy pre-disposition screening)
of the children. Testing of the children took place over a period of
three months with Christmas vacation intervening. Two reliability studies
were accomplished, the first early in the testing period and another
during the last three weeks of testing in an effort to guard against
"slippage" of examiner technique. Two individuals, both advanced
psychology students unconnected with the main investigation, were involved,
one completing the early study and the second performing the study at the
end of the testing period. The procedure in both instances was identical.
The observer sat in on the testing two times for each examiner, completing
a protocol each time which was later scored according to the standard