EFFECTS OF STORY ENACTMENT AND TEACHER-LED DISCUSSION
ON PRESCHOOL CHILDREN'S STORY COMPREHENSION
BY
TERESA C. BENNETT
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
1983
Copyright 1983
by
Teresa C. Bennett
This work is dedicated to
my son and my father.
"The lowest ebb is the turn of the tide"
Henry Wadsworth Longfellow
ACKNOWLEDGEMENTS
I wish to extend my thanks to the following people:
Dr. Linda Lamme, who was demanding, insightful, help-
ful, interested, and enthusiastic;
Dr. Steve Olejnik, my teacher, who helped me with the
analysis and spent time critiqueing my writing;
Dr. Dorene Ross, who critiqued my writing, helped me
revise and provided support;
Dr. Pat Ashton, who was a good listener and counselor;
Dr. Bob Algozinne, who provided feedback on my writing
and humor at important moments.
My thanks extend to other members of my family and
friends, my mother, Lolly and Chip, Gussie and Bob Mautz,
Sharen Halsall.
My thanks also go to Lois Rudloff, my typist, who did
such a great job.
I want to express my appreciation to an unofficial
member of my committee, Dr. Anthony Pelligrini, of the Uni-
versity of Georgia, who is an authority on play and young
children. He helped me formulate the framework of this
study.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS . . . . . . . .
ABSTRACT . . . . . . . . . .
CHAPTER
ONE INTRODUCTION . . . . . . .
Introduction . . . .
Need for the Study . . .
Statement of the Problem .
Significance of the Study
Limitations of the Study .
Definition of Terms . .
Formal Story Elements . .
Summary . . . . .
TWO REVIEW OF LITERATURE . . . . . .
Introduction . . . . . . . .
Reading Readiness . . . . . .
Story Comprehension . . . . . .
Formal Elements of a Story . . . .
Low Socioeconomic Status Black Children:
Language Development and Reading
Readiness . . . . . . . .
Play and Young Children . . . . .
Story Enactment and Story Comprehension
Play Training . . . . . . .
Teacher-Led Discussion . . . . .
Fairy Tales . . . . . . . .
Summary . . . . . . . . .
THREE RESEARCH DESIGN AND METHODOLOGY . . .
Subjects . . . . . . . . .
Sample Selection . . . . . . .
Variables . . . . . . . .
Instrumentation . . . . . . .
Design . . . . . . . . .
Data Collection . . . . . . .
Hypotheses . . . . . . . .
Assumptions . . . . . . . .
Page
iv
viii
1
1
4
6
8
9
10
12
13
14
14
17
18
20
22
25
29
31
39
41
43
Page
Procedure . . . . . . . . 55
Teacher Training Workshops . . . .. 57
Curriculum Implementation . . . .. 59
FOUR RESULTS . . . . . . ... 62
Little Red Riding Hood: Formal Elements
of a Story . . . . . . 62
The Gingerbread Man: Formal Elements of
a Story . . . . . . . . 77
Post Hoc Analyses . . . . ... 87
Little Red Riding Hood . . . ... 91
The Gingerbread Man . . . . .. 97
FIVE SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 106
Factors Affecting Results of the Study 106
Criterion-Referenced Test: Little Red
Riding Hood . . . . . . .. 113
Story Retelling . . . . . .. 114
Broad Implications for Future Research 122
Practical Implications for Future
Research . . . . . . .. 123
Enactment Treatment Observations ... . 124
Implications for Day Care Teachers . 127
Conclusions . . . . . ... 134
APPENDIX
A CRT LITTLE RED RIDING HOOD, 4th FAMILIAR
STORY . . . . . . . ... 137
B READING ALOUD TO CHILDREN SCALE (REVISED). 138
C BOOKS USED IN THIS STUDY . . . .. 141
D INFORMATION FOR TEACHERS ON THE STUDY
ABOUT HOW CHILDREN UNDERSTAND STORIES . 142
E SCHEDULE FOR STORIES . . . . .. 145
F QUESTIONS FOR THE THREE BEARS BY PAUL
GALDONE . . . . . . . ... 146
G QUESTIONS FOR THE THREE LITTLE PIGS ILLUS-
TRATED BY AURELIUS BATTAGLIA. . . . 149
H QUESTIONS FOR THE THREE BILLY GOATS
GRUFF . . . . . . . ... 152
Page
I TEACHER-LED DISCUSSION QUESTIONS TO BE
USED FOR THE STORY LITTLE RED RIDING
HOOD BY JACOB AND WILHELM GRIMM ... . 155
J SAMPLE STORIES . . . . . .. 158
K PROCEDURES . . . . . . . .. 168
L SUMMARY FOR TREATMENT EFFECTS WITH
TEACHERS NESTED WITHIN TREATMENTS USED
AS THE ERROR TERM . . . . . .. 169
REFERENCE NOTES . . . . . . . .. 170
REFERENCES . . . . . . . . .. 171
BIOGRAPHICAL SKETCH . . . . . . .. 175
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
EFFECTS OF STORY ENACTMENT AND TEACHER-LED DISCUSSION
ON PRESCHOOL CHILDREN'S STORY COMPREHENSION
By
Teresa C. Bennett
April 1983
Chairperson: Linda Leonard Lamme
Major Department: Curriculum and Instruction
The purpose of this study was to investigate the effect
of story enactment and teacher-led discussion on low income
black preschool children's story comprehension as measured
by a criterion-referenced test and two retelling tasks which
were analyzed for formal elements of a story.
The study involved 45 preschoolers (3 1/2 to 4 1/2
years old) in six Title XX (federally funded) daycare cen-
ters in Gainesville, Florida. The pretest administered by
graduate students in the Speech Department was the Test of
Early Language. The posttest data, the criterion-referenced
test and two retelling tasks were collected by the teachers
who carried out the four-week curriculum and by the experi-
menter.
viii
Analysis of covariance was used to test for treatment
effects on 11 dependent variables. Chi-square analysis was
done on 6 variables which were dichotomous. The signifi-
cance level was set at a=.05.
The four-week literature curriculum was carried out by
volunteer teachers in six daycare centers. The enactment
group was provided books, filmstrips, cassettes, instruc-
tions and props for each story. The teacher-led discussion
group was provided books, filmstrips, cassettes, instruc-
tions and specific questions and appropriate answers for
each story. The control group was provided books, film-
strips and cassettes for each story.
The enactment and teacher-led discussion treatments
had a significant positive effect on the criterion-refer-
enced test on Little Red Riding Hood, total formal elements
score and unity on The Gingerbread Man. These results sug-
gest that a literature curriculum utilizing enactment or
teacher-led discussion can significantly improve preschool
children's story comprehension, particularly in regard to
recall.
CHAPTER ONE
INTRODUCTION
Introduction
Psycholinguistic theories of how children learn to
read suggest that reading is above all a thinking process,
a "psycholinguistic guessing game" in which children test
hypotheses about how to derive meaning from print (Downing,
1979; Goodman, 1981; Smith, 1978). Rather than learning a
series of hierarchical skills, children learn to read by
interacting with whole texts while reading (Smith, 1978).
If reading is an integrated semantic process in which mean-
ing is paramount, children need to be given experiences in
problem-solving and making inferences to "fill in the gaps"
in reading materials such as stories (Bransford, 1979). A
story is a stable and organized body of knowledge and in or-
der to understand it a child must comprehend the continuity
and connectedness of the story's events and structure (Stein,
1980). Through exposure to stories, children develop an in-
ternal representation, a set of expectations about what com-
prises a story, called a story schema. A story schema
helps children integrate and understand what they read (Dur-
kin, 1981; Stein, 1980).
Language proficiency helps children understand and re-
member stories that are read to them. Story comprehension
is enhanced by good language skills and consistent exposure
to stories. Low income black children come from environ-
ments where they may not be exposed to literature and their
language development may be termed delayed (Deutsch, 1967;
Templin, 1975). Children from this population who are in
daycare centers might benefit from a program which empha-
sizes a holistic literature approach. Reading many stories over
a period of time facilitates the reading readiness of these
children through listening, memorizing, and inventing stor-
ies using book language (Levenstein, 1970). Better compre-
hension is possible when children understand stories and
their plots. The ability to comprehend the relationship of
events in a story helps children be more ready to read
(Stein, 1980).
Research with low income black children and reading
falls into two categories. Research which used isolated
skills as outcome measures found that direct instruction in
these skills most benefitted the children (Bereiter & Engel-
man, 1966). The use of higher order measures, like overall
language development and comprehension with children taught
by direct instruction, indicates that these children are not
as successful in comprehending what they read as children
taught by a meaning approach (Becker, 1977; Bruner, 1968).
The goals of this study are not involved with skill hier-
archies, since comprehension cannot be broken down into
skills (Mason, Osborn, & Rosenshine, 1977). This study has
a much broader goal, the integration and understanding of
story material.
One theory about the learning style of low income
black children terms their learning style, social in-
teractive (Gordon, 1982). Play, which is social interac-
tive, is identified as a powerful means for develop-
ing the language of young children (Piaget, 1962; Vygotsky,
1978). Play, an intellectual process, helps the young child
assimilate new information into present mental structures
(Piaget, 1962). Language and meaning grow through play as
the child becomes aware of symbols and the relationship be-
tween the signifier and the signified (Wolfgang, 1974). A
special type of play treatment, story enactment, may be ideal
for facilitating the language growth and story comprehension
of this population, because it is social interactive. Also
previous research has shown that story enactment has had pos-
itive effects on the story comprehension of low income chil-
dren who are five years of age (Pelligrini & Galda, 1982).
The traditional teaching method used to help children
listen to and understand stories is a questioning or
classroom discussion technique (Pearson & Johnson, 1978).
Teacher-led discussion, one treatment in this study, is more
direct than the other treatment, story enactment. This
study seeks to juxtapose a dynamic, interactive treatment,
story enactment, with the more direct method, teacher-led
discussion, to determine which condition enhances the story
comprehension and language of four year old black children.
Story comprehension and language, in turn, are important
aspects of reading readiness in a holistic approach to lit-
eracy.
Need for the Study
Theories abound as to why low socioeconomic status
black children do poorly in school settings. Whether the
problem is lack of adequate stimulation at home, biased
tests or biased teachers, the fact remains that children
in this population have difficulty academically (Gordon,
1982). It seems important to do research on these children
before school entry to ascertain the most appropriate teach-
ing methods for facilitating their learning.
This research is needed to extend the generalizability
of previous research in the area of story comprehension to
low socioeconomic status black four year olds. Six known
studies utilizing a small group enactment treatment have
found significant results on story comprehension measures
(Milner, 1982, Pelligrini & Galda, 1982; Saltz, Dixon, &
Johnson, 1977; Saltz & Johnson, 1974; Silvern, Williamson,
Taylor, Surbeck, & Kelley, Note 1; Silvern, Williamson, Tay-
lor, & Kelley, Note 2). This research shows that story en-
actment does facilitate story comprehension in older chil-
dren. There has not been enough research on children un-
der five years of age to know how they understand stories.
Nancy Stein (1980), a major researcher in the area of story
comprehension, states that more research is needed with
young children below the age of five to complete our under-
standing of how children think about stories.
The present study is similar to Milner's (1982) study
and a study by Pelligrini and Galda (1982). Milner used a
story enactment treatment with middle income white four
year olds and found significant results on total formal ele-
ments of a story included in a retelling task. Milner's re-
sults can only be generalized to similar populations, pre-
school children of college students. The proposed study
extends the generalizability of Milner's results to low SES
black four year olds. The second treatment in this study,
teacher-led discussion, is similar to a treatment used by
Pelligrini and Galda (1982). The discussion group in the
Pelligrini and Galda study did facilitate story comprehen-
sion, but the primary treatment, story enactment, facil-
itated better story comprehension for low SES black five
year olds.
Another reason this study is needed is to develop a
literature program of fairy tales for daycare staff using
story enactment and teacher-led discussion. This study
will broaden our knowledge of the best teaching methods to
use while involving children with literature. There is
little research data on the impact of children's experiences
with literature in the daycare curriculum. The
effectiveness of the training provided for the teachers is
assessed by determining its impact upon the students.
Statement of the Problem
Research is needed on low socioeconomic status four
year old black children in daycare settings to determine
the best methods for promoting conceptual learning, such as
story comprehension. This study seeks to investigate the
effects of two treatment conditions, story enactment of
fairy tales and teacher-led discussion about fairy tales on
preschoolers' story comprehension as measured by knowledge
of formal elements of a story and a criterion-referenced
test. The experimenter is providing teacher training for
those teachers who will implement the treatments in separate
daycare centers. The primary treatment is enacting a series
of four fairy tale stories with teachers taking an active
role in dramatization. The second treatment consists of a
teacher-led discussion after reading the same stories. The
control condition is a situation in which the teacher will
read the same stories.
The question under investigation in regard to story
comprehension is: Does story enactment, in preselected
groups of four children, or teacher-led discussion, facil-
itate the story comprehension of low socioeconomic status
black four year olds (N=45) as measured by a CRT and formal
elements of a story?
The formal story elements to be used in scoring the
children's story retellings are those used by Applebee
(1978), Isbell (1979), and Milner (1982): formal opening
(i.e., once upon a time), formal closing (i.e., the end),
and number of characters, number of incidents, number of
times conversational quotations are used and story unity.
The adult-led discussion treatment utilizes a questioning
mode developed by Sadow (1982) which is based on Rumelhart's
story grammar. The questions investigate these areas: set-
ting; initiating event, reaction, action, and consequence.
The control group is hearing the same fairy tales read aloud
and participating in their regular curriculum.
All three groups are hearing the same four fairy tales.
They are Little Red Riding Hood, The Three Little Pigs, The
Three Billy Goats Gruff, and The Three Bears. Fairy tales
are chosen as the genre of literature because they have
simple plots and thematic development. Fairy tales possess
special appeal to children who identify with the conflicts
and problems dealt with in the stories as well as the regu-
lated pattern inherent in the tale, i.e., use of repetition,
causal relations, formal opening and closing (Favat, Note 3).
Significance of the Study
This study is significant because it will develop a
curriculum model for teachers at the early childhood level
and validate its effectiveness for children. This study
also validates the theory that one academic skill area,
holistic reading readiness, can be enhanced either by a
thematic fantasy play curriculum or a more traditional dis-
cussion curriculum. The results of this study indicate
whether low socioeconomic status black children learn best
in a social interactive curriculum, i.e., story enactment,
or the more passive, abstract, discussion curriculum.
A carefully developed daily literature curriculum has
been developed for this study. The children hear one fairy
tale every week for four weeks. Each teacher reads the
story two days a week and audiovisual media presents the
story on two other days during the week. The story enact-
ment group enacts the story immediately after hearing it or
seeing it presented. This type of treatment may be espe-
cially appropriate for low SES black children, since it is
a condition in which social interaction is the focus. This
active treatment, termed story enactment, is contrasted with
the more traditional discussion treatment to determine which
condition is best for this population.
Another important facet of this study lies in its con-
tribution to new theories about reading readiness. Theories
about reading readiness have changed in recent years. The
theory of readiness as visual and auditory discrimination,
letter identification, and copying letters is being rede-
fined with research findings by psycholinguistic theorists
(Downing, 1979; Goodman, 1981; Smith, 1978). Clay (1972)
says there are certain language concepts children need to
master before they are ready to learn to read. These lan-
guage concepts involve exposure and interaction with books,
inventing stories, using book talk and memorizing stories
(Clay, 1972; McDonell & Osburn, 1978). The story enactment
treatment in this study is aimed at facilitating these lan-
guage concepts. Inventing stories and talking in book lan-
guage are practiced while enacting the stories and speaking
dialogue. Memorizing stories develops with exposure to the
same stories over a period of time (Schickedanz, 1978).
The significance of this study, then, lies in its potential
for giving us information about which kind of curriculum is
most beneficial with this particular population in regard
to story comprehension, which is the central component of
holistic reading readiness.
Limitations of the Study
One limitation of this study was that the quality of
the literature intervention is dependent on the training,
abilities and interests of the teachers. Attempting to re-
mediate this, the experimenter divided the teachers by edu-
cational level before randomly assigning them to treatments.
Also, teacher training was provided to standardize implemen-
tation of the treatments. Nevertheless, this study was
susceptible to teacher effects. Possibly more teacher
training is needed for successful implementation of the en-
actment treatment.
Two other factors which limited the study were small
sample size and the short length of the treatments. The
total sample size was 45, after 17 subjects were lost because
of attrition and absenteeism. Six centers were used because
that was the maximum the experimenter could site visit per
week. Only the four year olds in the six centers were sub-
jects. This limits the number of students in the study.
The experiment .was scheduled for four weeks. This might
be too short a time span to see changes in the dependent
measures.
Another limitation of the study was that a federal day-
care audit took place during the second week of the experi-
ment. The audit affected the morale of the teachers, di-
rectors and students. Also loss of some subjects was due to
the audit.
Because subjects could not be randomly assigned to
treatments, analysis of covariance was chosen for the analy-
sis to adjust for initial differences between subjects.
The covariate consisted of a language quotient score on the
Test of Early Language (Note 6).
Definition of Terms
Schema. In this study a schema is defined as a mental
construct which includes information about events or happen-
ings which must be met before a situation may qualify as a
particular type of event or action. Schemata guide the as-
sumptions human beings make while comprehending, learning,
and remembering. Schemata may also be called scripts or
frames. A specific script involves a group of concepts or
events. For example, a birthday party schema would include
a mental script of what events will take place at a birth-
day party, who will attend, etc. Schemata help us organize
knowledge to better understand incoming information about
the world (Bransford, 1979).
Story schema. In order to connect the ideas and events
to one another in a story, a story schema develops to help
the reader establish continuity between events in the story
(Rumelhart, 1975).
Story grammar. These have been developed by authors
and consist of a setting and a number of episodes which are
related to one another in a meaningful way.
Story representation. A mental representation in the
mind of the reader concerning the actions or events in a
story and how they are connected to one another. Assessment
of how well a reader comprehends the story is usually done
by asking the reader to retell the story and by analyzing
how many connections are made.
Story enactment. Children are read a story, are as-
signed roles and then enact the story in groups of four
with active participation by the teacher.
Thematic fantasy play. A situation in which children
enact a role and theme not related to their personal expe-
rience.
Self-directive dramatization. The pupil's own orig-
inal, imaginative, spontaneous interpretation of a charac-
ter of his/her own choosing in a story.
Teacher-led discussion. An activity in which teacher,
after reading to the students, utilizes a questioning mode
to go over the important elements of a story.
Reading readiness. Visual and language concepts devel-
oped over time with exposure to literature and print. The
reading readiness area that relates to this study involves
story comprehension specifically memorizing telling and in-
venting stories, and using book talk (Clay, 1972).
CRT. A test to measure recall (Appendix A).
Formal Story Elements
Formal opening. A designated beginning to the story,
i.e., once upon a time.
Formal closing. A designated ending to the story, i.e.,
The end.
Number of incidents. A count of the number of inci-
dents recounted in the retelling, i.e., He jumped on his back.
Number of conversational quotations. A numerical count
of the times characters speak dialogue, i.e., He said, Hello."
Number of characters mentioned. A numerical count of
how many different characters are mentioned in the retelling.
Unity. A measure of the child's skill in retelling the
story with a sense of thematic development, i.e., If the
child brings in incidents not related to the story, unity
was not scored.
Summary
The overall purpose of this study is to investigate
the usefulness of a story enactment treatment or a teacher-
led discussion treatment on the story comprehension of black
low socioeconomic status four year olds as measured by for-
mal elements of a story included in two retelling tasks and
scores on a criterion-referenced task.
CHAPTER TWO
REVIEW OF THE LITERATURE
Introduction
Experience with literature has a significant effect on
a child's literacy development (Teale, 1978). Interaction
with adults and literature in the home environment influ-
ences the interest children have in books and other areas
related to reading, i.e., language development and vocabu-
lary development (Durkin, 1981; Levenstein, 1970). Studies
show that the home environments of early readers included the
following: printed materials were present, reading was'done
in the environment, the environment facilitated contact with
paper and pencil, and adults in the environment responded
to the child's efforts with quality interaction (Teale,
1978). This description of home environments producing
children who will read early and love to read would not
characterize the homes of low socioeconomic black children
(Levenstein, 1970). Studies show that there was a lack of
printed materials in low SES homes, that books were infre-
quently read to children in these homes, and these parents
wereless sophisticated verbally (Deutsch, 1967; Templin,
1957). These factors have a profound effect on the language
development and potential reading readiness of the children
from low SES environments. This studywas designed to en-
hance the literature experiences that low SES children may
lack at home, but which can be provided at a day care center.
It has been accepted that young children need the ex-
perience of hearing stories read. Schickedanz (1978) stated
that therewere specific skills children learn during the
story reading experience, one of which is memorizing the
story. The ability to remember and tell a story serves the
purpose of helping the child develop a story schema, a set
of expectations about what is contained in a story. The
best teaching method to help low SES black children develop
a story schemawas the focus of this study.
There are varying views concerning the best teaching
methods to use with low SES preschoolers. Becker and Engel-
man (ote 4)of the Oregon Direct Instruction model emphasized
individual and group classroom drill on basic skills as the
best teaching method. They pointed out the positive overall
performance of the didactic direct instruction models in the
Follow Through Evaluations. Bereiter and Engelman (1966) as-
serted that direct instructionwas the best teaching tech-
nique for low SES children. Programs like DISTAR did raise
scores on reading readiness skill tests, but these same
children fell below the national average on the reading
comprehension test of the Metropolitan Achievement Test
(Becker, 1977; Bruner, 1968). It seems obvious that direct
instructionwas the best method for teaching skill
hierarchies but probably not for teaching higher order
thought. The focus of the present studywas a higher order
mental process, story comprehension.
Comprehension of stories develops through quality ex-
periences with literature. The development of a young
child's total language, syntax, semantics, phonology, and
vocabulary enables the child to be a better reader
(Livo, 1972). One theorist stated that the best precursor
to reading achievement was a program of total language de-
velopment which was filled with interverbal communication
(Livo, 1972). The two treatments in this study,
story enactment and teacher-led discussion, were aimed at
immersing the child in verbal interaction in order to facil-
itate the development of a story schema. This mental idea
of story elements help the child to better understand
stories when read (Rumelhart, 1975).
This review will begin by discussing reading readiness
and how a story schema develops through exposure and memor-
ization, and this aids story comprehension. Research stud-
ies which have utilized formal elements of a story as a de-
pendent measure will be reviewed next. The teaching meth-
ods used to teach low SES children reading readiness in the
past will be reviewed with particular attention to language
deficits and difficulty in reading. Finally, the two meth-
ods which have been helpful in teaching children story com-
prehension, story enactment and teacher-led discussion
will be discussed.
Reading Readiness
Ideas about what reading is fall into two camps. One
states that reading is a hierarchy of skills which can best
be taught by direct instruction (Becker, 1977; Bruner, 1968).
The other camp states that reading is an interactive process
between reader and whole text in which meaning is the cen-
tral element (Downing, 1979; Goodman, 1981; Smith, 1978).
The holistic concept of reading is more learner-centered
than the skill hierarchy approach. Ideas about reading
readiness differ according to which side of the reading
argument one accepts. The authors who advocate holistic
meaning approach to reading emphasize the gestalt of the
reading episode, for example, knowing that books tell stor-
ies, knowing that book talk is different from conversation,
memorizing and inventing stories (Clay, 1972; Schickedanz,
1978; Smith, 1978). These are broad language concepts which
develop over time with exposure to literature. This study
examined two methods of teaching children how to
approach the reading experience with these broad language
concepts needed for comprehending print. Preparing chil-
dren to be holistically ready to readwas a broad goal of
this study.
The ability to tell a story has been highly correlated
with reading readiness. In one study children were read
"Peter Rabbit" 10 times. Then they were asked to state as
many incidents as they could recall about the story. There
was a .78 correlation with the Lee-Clark Reading Readiness
Test (Livo, 1972). The knowledge children gain by experienc-
ing literature helps them develop the mental scaffolding
needed to comprehend when they read in later years (Livo,
1972).) The two most important resources children bring to
the reading experience are competence in oral language and
the knowledge that reading is the expression of and compre-
hension of meaning (Goodman, 1981). The meaning approach to
reading and reading readiness emphasizes the communicative
nature of the reading act. Reading is getting information
from books and gaining an understanding of what is contained
in books constitutes reading readiness (Downing, 1979).
This orientation is the view adopted by this research study.
Cognitive psychologists state that human beings inter-
pret experience through existing mental structures, schemata,
which aid human comprehension. It is important for children
to develop story schemata to enable them to better compre-
hend the stories they read or are read to them (Durkin,
1981; Rumelhart, 1975).
Story Comprehension
John Bransford (1979) in his book Human Cognition ex-
plores the relationship between schema theory and comprehen-
sion. The concept of schemata is derived from the work of
Piaget and Kant (Bransford, 1979). Schemata are also called
scripts or frames. Schemata characterize the way in which
conceptual structures are built. A story schema helps or-
ganize the information in the story in a logical, coherent
manner. A restaurant schema would include a sequence of
events which occur in a restaurant. Human comprehension de-
pends upon these schemata and subschemata (the sequence in-
herent in the entire script) to make sense of what will be
read and understood (Bransford, 1979).
We all depend on our prior knowledge of the world and
prior experience to help us reason about events or situations.
Bransford (1979) states that
Comprehension consists of: (1) finding a
schema that fits a particular input [i.e.,
at a birthday party, or that sequence of
events which constitute a birthday party.]
(2) discovering those entities that corres-
pond to particular roles required in the
schema, (3) making inferences to fill in
the gaps in the story. (p. 185)
Bransford (1979) discusses the role of inference in compre-
hension. Comprehension depends on one's ability to think
inferentially and make assumptions based on general knowl-
edge. Understanding stories requires one to make assump-
tions concerning relations between events. People make
sense of what is heard or read by connecting the events of
a story in some logical way. Readers or listeners "fill in
the gaps" of a story based on their experience level.
The holistic model of teaching reading,which the the-
ories of story schema fit into, states that reading compre-
hension is an interactive process in which both text and
world knowledge play key roles (Durkin, 1981; Rumelhart,
1975). Low SES children need quality experiences with lit-
erature to develop story schemata for better understanding
and integration of what is read.
Formal Elements of a Story
In an attempt to discover how children think about
stories, some researchers have analyzed stories told by
children. This analysis has revealed there are certain
elements of a story. Three studies will be reviewed in
this section. The first study is by Applebee (1978), who
analyzed the 360 stories collected by Pitcher and Prelinger
(1963). The stories were told by middle class American
two, three and four year olds in response to the question,
"Tell me a story." Applebee scored these stories for for-
mal elements: formal opening, formal closing, the use of
consistent past tense. All three conventions showed a
steady rise from two-five years. Applebee also found that
number of words, number of T-units, number of characters,
number of incidents, and average number of words per T-unit
all showed a consistent and significant rise with age
whether considered individually or as a set.
Table 1.
Use of Formal Elements of a Story
by Applebee (1978) p. 163
Age 2 Age 3 Age 4 Age 5 Chi-Square
Formal 30.0 43.3 76.7 86.7 26.87 (significant)
Beginning
Formal 0.0 13.3 13.3 46.7 23.82 (significant)
Ending
Consistent 63.3 80.0 93.3 86.7 9.63 (significant)
Past Tense
Isbell (1979) used a story reading and storytelling
condition with 12 middle class subjects who were four and
five years old. Contemporary children's stories were used.
Isbell studied the same dependent measures that Applebee de-
veloped. She found that the storytelling group included
more incidents in the retelling, more formal endings, and
conversational quotations. One weakness of Isbell's study
was the small number of subjects. It is difficult to draw
conclusions about a treatment with only six subjects in each
treatment, but her findings do correspond to those of Apple-
bee.
Milner (1982) conducted a study utilizing an eight
week story enactment treatment with four year olds at Baby
Gator Research Center, in Gainesville, Florida. Milner meas-
ured the effect of her fairy tale curriculum with several
outcome measures of oral language, formal story elements
[similar to Applebee (1978) and Isbell (1979)] empathy, and
reading readiness. She found significant effects for condi-
tion on the empathy measure, formal opening, number of char-
acters mentioned and total score for the use of formal story
elements. Milner's subjects completed a retelling task and
tapes were analyzed by the experimenter. The control group
in Milner's study were read contemporary children's stories
instead of fairy tales. Milner's curriculum was a total in-
tegration of the fairy tale story throughout the entire pre-
school curriculum. Since she was a teacher in the school
where the treatment was held, she planned numerous activities
every day which related to that week's story. Milner's re-
sults are limited in that they can only be generalized to a
small population -preschoolrchildren of college students.
The research on formal story elements concludes that
children's concept of a story grows from age two to five in
formal elements of a story included in a retelling task
(Applebee, 1978). Milner's (1982)study demonstrated that
middle income children of student parents can grow in story
comprehension as measured on total formal elements of a
story through exposure to thematic-fantasy play.
Low Socioeconomic Status Black Children:
Language Development and Reading Readiness
The language of a low income child differs from the lan-
guage of a middle income child. These differences have been
researched thoroughly concluding that the crucial differ-
ence between lower-class children and middle-class children
is not in the quality of language but in its use (Deutsch,
1967). Low SES children generally have scores on language
tests which are below their mental ages. Deutsch (1967)
stated that "being lower class, black or white, makes for
lower language scores."
Research has paid specific attention to the language
of low SES children drawing the following conclusions:
(1) they have limited language ability;
(2) they possess syntactic inferiority;
(3) they use more simple sentences;
(4) they use more mispronounced words;
(5) they have deficits in auditory attention and in-
terpretation skills;
(6) they lack some communication skills;
(7) they lack adequate adult models in the environ-
ment;
(8) they have a restricted vocabulary (Dunn, Neville,
Pfost, Pochanart, & Bruininks, 1968, page 8).
Low SES children generally understand more language than
they use. The school setting may be particularly difficult
for them to adjust to because school language is so differ-
ent from the low income child's language.
Language proficiency is highly correlated to reading
achievement (Livo, 1972). Because of the limited language
proficiency lower-class children possess, they are 4 to 10
times more likely to be poor readers in comparison to the
entire school population (Dunn et al., 1968). Low SES chil-
dren enter school less ready to learn to read in comparison
to advantaged children (Dunn et al., 1968). For this reason
the low SES population has been the focus of a great deal of
reading readiness research.
The reading readiness skills identified 10 to 20 years
ago as the best predictors of reading achievement were the
isolated skills of knowing letter sounds and names, auditory
blending and visual discrimination. Facilitating the read-
ing achievement of low SES children in school settings was
the focus of such programs as DISTAR. The Follow Through
Evaluations show that direct instructionwas the best method
for teaching low SES children the hierarchy of reading readi-
ness skills mentioned above (Bereiter & Engelman, 1966).
The DISTAR program did raise achievement to one standard de-
viation above the national norm on the WRAT (word recogni-
tion subtest). However, on the reading comprehension test
of the MAT, these same students fell below the national norm
(Becker, 1977). This research points out the weakness of
the Direct Instruction Model for teaching a complex global
ability like reading comprehension.
New ideas about reading readiness emphasize the impor-
tance of understanding the meaning behind an author's mes-
sage (Smith, 1978). Theorists now assert that low-income
children need teaching strategies which enable them to ver-
balize with peers and use motor abilities (Dunn et al.,
1968). The story enactment treatment in this study seems
ideal, then, for this population to facilitate story compre-
hension, a very important reading readiness skill for pre-
readers. Story enactment was derived from the developmental
theory of play and its value for young children as their
natural mode of learning.
Play and Young Children
What is the value of play? Many theories have stated
that adaptive intelligence involves both differentiation and
integration. Piaget (1962) called this accommodation and
assimilation. The value of play rests in the process of in-
tegrating and consolidating recent learning and conceptual-
ization. Play develops from a self-directed activity in
which children imitate actions to an other-directed activity
which includes social interaction and language. Play can be
defined as voluntary, pleasurable activity which is not goal-
directed or dependent on the restraints of time and space.
In play elements of reality are incorporated into the imag-
ination. Play is active, structured, symbolic activity in-
volving mental processes which develop adaptive intelligence.
The preschool child makes use of a new psychological
process upon arrival at Piaget's preoperational stage,
symbolic play (Piaget, 1962). Symbolic ability allows the
child to call up objects or actions which are not present,
but have been observed. Play is the fusion of reality and
fantasy. Through use of symbols and signs (language), the
child begins to think abstractly.
In play, thought is separated from objects
and action arises from ideas rather than
things: a piece of wood becomes a doll and
a stick becomes a horse. . It is ter-
ribly difficult for a child to sever thought
(the meaning of a word) from object.(Vygot-
sky, 1978, p. 97)
The very young child is bound to every ac-
tion by situational constraints . it is
impossible for very young children to sepa-
rate the field of meaning from the visual
field because there is such intimate fusion
between meaning and what is seen. (Vygot-
sky, 1978, pp. 96-97)
The preschooler can call up mental images of objects,
actions, and situations. Play allows manipulation of reality
through fantasy. Young children who are developing symbolic
ability need specific objects for play. However, as the
child grows older, fewer props are needed and words suffice
as symbols. Abstract thought has arrived. Play provides the
meaningful context in which children can develop competent
language use. Stern, Bragdon, and Gordon (1976) identified
three cognitive areas directly related to symbolic play.
These are the use of symbolic representation, involvement
(focus of attention), and language.
During symbolic play, small groups of children pretend,
verbalize, problem-solve and use their primitive conceptions
of the reality of meaning. Verbalization during play is cru-
cial to the development of linguistic meaning and thought
(Jurkovic, 1978; Vygotsky, 1978).
The relation of thought to word is not a
thing but a process, a continual movement
back and forth from thought to word and
from word to thought. . Thought is not
merely expressed in words; it comes into
existence through them. Every thought
tends to connect something with something
else, to establish a relationship between
things. Every thought moves, grows, and
develops, fulfills a function, solves a
problem. (Vygotsky, 1962, p. 125)
The basis of all cognition is flexibility and fluency with
symbols. This develops in young children through the mean-
ingful experience of play.
The development of symbolic language and thought compe-
tence in children has its roots in late infancy with the
emergence of representational, symbolic ability (Piaget,
1962). Piaget (1962) stated that in the fourth stage of the
six sensori-motor schemes, an infant will search for an ob-
ject that has been placed out of the child's visual field,
i.e., under a blanket. This searching on the part of the in-
fant demonstrates that the child has a mental image of an ab-
sent object. This is the beginning of symbolic activity.
The gradual freeing of the symbol from what it represents
leads to symbolic behavior as an autonomous medium, such as
language.
Gowen (Note 5) has defined and identified the important
elements of symbolic play. She defined symbolic play as
using inanimate objects as animate, performing everyday ac-
tivities in the absence of materials (i.e., drinking out of
an empty cup), substitution of one object for another, role-
playing, and novel or unusual endings to play activities.
Seventy-eight percent of the children in her study did play
symbolically. Gowen identified three structural elements to
symbolic play: the signifier (child or object), the signi-
fied (objects or beings), and the mode of representation.
She found that three-four year olds used objects more often
to pretend and four-five year old children used more verbal
communication without objects and actions. This supports
Vygotsky's assertion that meaning is derived from action and
objects, and that verbal symbols become adequate for convey-
ing meaning with maturity.
The two major theorists who have contributed most to
discussions of playwereVygotsky and Piaget. Their orien-
tations were similar in some ways, yet very different in oth-
ers. Piaget believed playwas intelligent behavior involving
processes of assimilation over accommodation (Piaget, 1962).
By encoding symbols, the egocentric preschool child has a
method of rethinking the reality of past experience and as-
similating these experiences into existing mental structures
(Fein, 1979). Vygotsky, in contrast, saw play as an emo-
tional process in which the child seeks to reduce tension
and understand the social meaning of the world. To Piaget
child's play was an egocentric experience, to Vygotsky it was
a social experience in learning the social code of the cul-
ture, language (Fein, 1979).
The story enactment treatment in this study incorpo-
rated all the benefits of play as a social experience in
learning. The structure of a story provided the conceptual
framework for the play episode. The roles of characters in
the story provided the vehicle for peer-peer interaction with
dialogue, actions and appropriate props. This researcher
hypothesized that the social interactive mode of learning,
story enactment, may be a good way to teach low SES black
children to appreciate and comprehend literature which, in
turn, may make them better readers.
Theories state that play facilitates symbolic develop-
ment. Acting "as if" objects or actions are real helps young
children develop representational ability (Vygotsky, 1978).
Story enactment may be especially appropriate for low SES
children to help them develop their skills with symbol ma-
nipulation. Story enactment may be the mental mediation
necessary for young children who are not verbally precocious
to develop complex mental abilities like story comprehension.
Story Enactment and Story Comprehension
There are six studies which have used a story enactment
treatment as an independent variable to facilitate story
Table 2.
Studies which Relate to the Proposed Study
Pelligrini Silvern Silvern Milner Saltz Carlton Proposed
& Galda et al. et al. (1982) et al. & Moore Study
(1982) (Note 4) (Note 5) (1977) (1968)
Age of
Subjects
Number of
Subjects
Treatment
Length
Levels of the
independent
Variable
Types of
Stories Used
Who imple-
mented
treatments
Results
5-9
505
8 weeks
5-8
340
8 weeks
5-7
108
1 treatment
2 trng ses.
teacher-led
discussion
*TFP and
drawing
fairy tales
researcher
TFP signifi-
cant for story
comprehension
4 years
60
8 weeks
3-4 years 6-10 years 4 years
80
1 year
TFP and TFP, socio-
control dramatic play,
discussion, &
control
fairy tales fairy tales
teachers one
of which was
researcher
TFP significant
for total ele-
ments of story
& empathy
Dependent 1-two factor 1-story recall 1-story recall 1-story recall
Measures criterion ref- Steins 12 Steins 12 task 2-empathy
erenced test propositions propositions measure 3-read-
2-retelling 2-Borke's IPT 2-Borke's IPT ing readiness
4-oral language
SES rural low SES rural low SES rural
*Thematic Fantasy Play
middle income
120 100 approx.
14 weeks 4 weeks
self-directed story enactment,
dramatization teacher-led
discussion
children
self-selected
stories
fairy tales
self-directed dramatization
and control & teacher-led
discussion
control
contemporary 6 unfamiliar
children's stories, and
stories 6 familiar
fairy tales
teachers teachers
were were
trained trained
dramatization significant
for story comprehension
teachers some teachers teachers
of which were were were trained
researchers trained
TFP signifi- self-directed
cant for gains dramatization
in IQ & sequen-treatment sig-
tial memory & nificant for
empathy reading gains
1-PPVT 2-se- Gray-Votaw- formal elements
quential mem- Rogers Achieve-of a story on 2
ory 3-fantasy ment test on stories and CRT
judgment test vocabulary
4-empathy for primary
5-story in- & intermediate
terprctation grades
test 6-impulse
control
low SES low SES low SES
comprehension. These studies all indicate story enactment
has a significant effect on story comprehension even though
story comprehension was measured in different ways. A chart
which summarizes the data of these studies follows.
Play Training
Saltz, Dixon and Johnson (1977) conducted a study in
which low SES preschoolers were trained in one of three
types of fantasy activities during a school year. The 80
subjects were randomly assigned to four groups. The exper-
imental sessions were 15 minutes three days per week for
each group. The treatments were thematic-fantasy play,
(acting out fairy tales), socio-dramatic play, fantasy dis-
cussion, and a control condition which was a regular pre-
school curriculum. Subjects were pretested on the PPVT.
An alternate form of the PPVT was one posttest. Subjects
were given another posttest, the story interpretation test,
which measured their ability to relate events to one another
and to measure causal relations. Sequential memory was mea-
sured by telling the child a story with five simple pictures
After hearing the story the child was given the pictures in
random order and asked to put them in their correct order to
tell the story. This procedure was repeated with a second
story. Also measured in this study were empathy and impulse
control (Saltz, Dixon, & Johnson, 1977).
This study isolated the elements of the socio-dramatic
play and thematic fantasy play which were important for the
development of cognitive processes. One important element
seemed to be the play enactment and the use of symbols. An-
other important element was the motoric component inherent
in the treatment. Children moved when acting out situations
or stories. The role changes required for more complex pat-
terns of behavior in response to changes in situations and
people in the episode were important. Thematic fantasy play,
in particular, seemed to facilitate the ability to perceive
causal relationships. This study has consistently had im-
pressive results: an average gain of 23 IQ points for the
play-training groups versus an average gain of 16 IQ points
for the control groups.
The present study examined the utilization of a play-
training model similar to that used by Saltz, Dixon, and
Johnson (1977). Three groups were utilized, one in which
stories were enacted, one in which the teacher led
a discussion after reading the story, and a control condi-
tion in which the children just beard the same stories
read.
Pelligrini and Galda (1982) did a similar study with 108
low SES black children, 54 boys and 54 girls, in kindergar-
ten, first and second grade. The children were randomly as-
signed to one of three treatment conditions, thematic fantasy
play, adult-led discussion or drawing. The children in the
thematic fantasy play group enacted roles in fairy tales.
The adult-led discussion group did not participate in fan-
tasy reconstruction with peers, but after hearing the fairy
tale read discussed the story with the teacher. Children
in the drawing condition were read the fairy tale and given
blank paper and crayons and asked to draw as much about the
story as they could.
The three treatment conditions were carried out with
nine groups of four children within each grade. Each group
consisted of two boys and two girls. A researcher carried
out all the treatment conditions. One of the dependent mea-
sures in this study was performance on a criterion-referenced
test of 10 questions about the last story which was read.
The total number and sequence of events recalled about the
story in a retelling task was the other dependent measure.
The story retellings were broken down into nine main constit-
uents, setting, seven episodes and conclusion. Retellings
were scored as including the constituent if the "gist" was
included irrespective of order. Retellings were also scored
for sequence.
ANOVA analysis revealed that the play group in this
study (x = 5.888) scored significantly higher for condition
than both the discussion (x = 4.722) and drawing (x = 3.666)
groups. The discussion group scored higher than the drawing
group as evidenced by the means. The significant effect for
grade indicated that older children did better than younger
children. Second graders x = 6.361) scored significantly
higher than first graders (x = 4.611). First graders
scored higher than kindergarteners (x = 3.305).
The results of this study indicate that story compre-
hensionwas a function of two factors, age and training in
verbally reconstructing the story. Second graders outper-
formed the younger children. On the criterion-referenced
test measuring story comprehension, children who were ex-
posed to thematic-fantasy play or adult led discussion an-
swered correctly more story related comprehension ques-
tions. For kindergarteners and first graders, thematic-
fantasy play provided the most effective means of understand-
ing stories.
This study was similar in several ways to the Pelligrini
and Galda study. This study included a thematic fantasy
play treatment and a teacher-led discussion treatment. The
third group was a drawing group, but a condition in
which the children heard the same stories read. The popula-
tion was younger, i.e., four year olds. Therewas a similar-
ity in one dependent measure, the retelling task, although
the tapes were analyzed according to formal elements of
a story originated by Applebee (1978).
Pelligrini and Galda (1982) provide support for two
methods of encoding language to promote story comprehension,
thematic fantasy play and adult-led discussion. More re-
search is needed with younger children, though, in order to
conclude that younger children are incapable of understand-
ing stories as well as older children.
Carlton and Moore (1968) studied children in first
through fifth grade over a period of eight years. They
used a story enactment treatment called self-directive dram-
atization. The authors defined self-directive dramatization
of stories as referring to the pupil's own original, imag-
inative, spontaneous interpretation of a character of his/
her own choosing in the story which is selected and read co-
operatively with other pupils in his group. It wasnot chil-
dren putting on plays or dramatics. Of particular interest
to this research was the chapter on the use of self-directive
dramatization with 120 culturally disadvantaged pupils. The
school population used in this part of the study was 85 per-
cent black. Children in grades one through four were in-
cluded in the treatment. Experimental subjects who received
exposure to self-directive dramatization were matched with
control subjects who did not receive the training but partic-
ipated in a basal reading program.
The children were exposed to the self-directive ap-
proach. They began by learning how to select stories they
wanted to read. They read to each other in groups of two.
They dramatized parts of stories that the teacher read to
them and by acting out a character for the other children
to guess. The classroom teachers carried out the treatment
which lasted for 14 weeks. Usually self-directive drama-
tization took place two or three times a week.
The dependent measure used in this study was the Gray-
Votaw-Rogers Achievement Test (vocabulary and paragraph
meaning section) to ascertain reading gains. Although not
familiar to this researcher, this dependent measure seemed
highly related to story comprehension. The results of this
study show that the experimental group scored significantly
better than the control group in all grades. Self-concept
was also improved through self-directive dramatization.
This study supports other studies in the following conclu-
sion; whether called self-directive dramatization, thematic-
fantasy play or just story enactment, children taking roles
and acting out stories with some teacher participation was
facilitative for promoting conceptual growth such as story
comprehension. Although used with older children, it will
be possible to adapt the self-directed approach outlined by
Carlton and Moore with the following changes. The choice of
stories was decided by the researcher. Since children
are younger the story enactment groups need to be four
to a group. Teachers may take a
more active role, either by taking a role or narrating the
enactment.
The self-directed dramatization method was used by Sil-
vern, Williamson, Taylor, Surbeck andKelley (Note 1), a
group of researchers at Auburn. They have done two studies
utilizing a play treatment to facilitate story comprehen-
sion. The first study involved self-directed dramatization.
Thirteen teachers near Auburn, Alabama, conducted the self-
directive dramatization treatment and control conditions.
There were 505 children, aged five to nine years, 266 boys
and 239 girls, from a rural environment, in 26 intact,
rural public and private school classrooms participating in
the study. The population was low to middle SES with no
racial information given. The same six stories were used
as stimuli in treatment and control conditions. They were
contemporary children's books chosen by the investigators.
Teachers reviewed the books to insure that the books were
unfamiliar to the children. To be sure that the children's
actions were not based on past experience with the stories,
teachers read from typed copies with no pictures as stimuli.
Teachers were trained in how to carry out self-directed
dramatization. The researchers found that the children in
the treatment group had significant increases in story re-
call. Story recall in this study was measured according to
Stein's (1980) 12 propositions formed into 10 multiple
choice questions.
Silvern, Williamson, Taylor, and Kelley (Note 2) con-
ducted another study measuring children's story recall as a
product of play, story familiarity and adult intervention.
There were 340 children in this study in 20 intact classrooms,
in kindergarten through third grade. There was no information
on sex, race or SES given in the article. The same multi-
ple choice test was used as in the previous study. Teach-
ers were trained in the self-dramatization process. Teach-
ers in the treatment group read the class a story and imme-
diately the children acted out the story. In the control
group the teacher read a story to the class and conducted
a discussion. The treatment group did significantly better
on story recall.
Two separate sets of stimuli were used, one set was
unfamiliar, one was familiar. The unfamiliar set included
the contemporary stories used in the first study reviewed
by Silvern et al. (Note 1). The familiar group of stories
were the classic fairy tales, i.e., Three Pigs, Three Bears.
Teachers read from typed copies of the content of the stor-
ies.
Five of the cooperating teachers in the treatment con-
dition volunteered to take a facilitative role in the play,
and five took a directive role in the play. The teachers
were trained in self-directive dramatization techniques.
The play treatment did significantly better on story recall.
These results support the other studies in this section
which have found significant results with story enactment.
The teachers stated that novel stories should be used if
children are not playing them out.
In summary, the studies utilizing a story enactment
treatment have consistently had impressive positive results.
With older children, story enactment facilitated better
reading comprehension. With young children, story enact-
ment helped children recall story elements as measured in
a retelling task.
Teacher-led Discussion
The teacher questioning technique is a widely used
method to teach reading comprehension. When teachers ask
the right questions, experience and research support the
value of this technique (Guszak, 1967). Pure recall ques-
tioning is deemed trivial and not very useful by reading
specialists (Guszak, 1967; Hare, 1982). Reading is a reason-
ing process which involves inferential thinking, evaluation,
explanation, prediction and conjecture (Downing, 1969).
Guszak (1967) did a study of teachers in second, fourth and
sixth grade. They asked 70 percent literal questions con-
cerned with the factual makeup of stories. Questioning
techniques which help children exercise higher order thought
processes stimulated better reading comprehension.
Pearson and Johnson (1978) developed a taxonomy for
evaluating the entire comprehension question-response se-
quence. The three classes of question-response sequences
were: (1) textually explicit--this requires no inferential
thinking, just literal recall; (2) textually implicit--this
requires reading between the lines and making inferences;
(3) scriptally implicit--this requires reading beyond the
lines and relying on story schema--and prior knowledge--to
integrate story material. Reading approaches which empha-
size meaning tend to stress more scriptally implicit ques-
tion information to develop comprehension (Hare, 1982). In
preparing young children in reading readiness, teachers
need to be aware of the value of developing story schemata
to help children comprehend reading material.
Teacher-led discussion designed to develop story sche-
mata in young children was chosen as a treatment in this
study. The questions the teachers used were developed from
elements of a story by Applebee (1978) and Rumelhart's
(1975) story grammar. Rumelhart's (1975) grammar waschosen
as a basis for discussion method in this study because of
its simplicity. The first part of the story grammar is the
setting. Then the initiating event sets the story in mo-
tion. The main character then reacts with some feeling,
thought, desire or goal which motivates action. Finally,
the action is responsible for some consequence, (Sadow,
(1982).
In this study a discussion wasused as an alternate
method to teach children the elements of stories. Discus-
sion may prove as effective or more effective than the story
enactment treatment. If both methods of teaching are effec-
tive, teachers could alternate methods or use discussion
only if it proves as good as story enactment. Fairy tales
are chosen as the most appropriate type of literature to
use because of structure, characters and appeal to chil-
dren.
Fairy Tales
The magic of fairy tales and their appeal to children
have been addressed by several authors. Bettleheim (1977)
states that fairy tales are important for healthy emotional
development of the child. Children's identification with
the conflicts and problems dealt with in these stories pro-
vides motivation for role-playing during play. Fairy tales
are full of fantasy situations. Children identify with the
central hero/heroine of the tale. Interest in the fairy
tale emerges in the pre-reading age and declines at about
10 years of age (Favat, Note 3).
Why do children like fairy tale literature? Piaget
(1962) has described the thought processes of the young
child in great detail. His beliefs about children and the
elements of the fairy tale help form the rationale for us-
ing a fairy tale curriculum with preschool children. The
child's belief in magic, animism, the morality of constraint,
and the transformations the child performs in play correlate
well with the content of the fairy tale (Favat, Note 3).
Animism, giving human characteristics to objects and animals,
is present in fairy tales. Moral justice in fairy tales
usually entails punishment for evil doings. This is
the morality of constraint. The fairy tale is a stable men-
tal construction of magic, animism, and authority. Fairy
tales appeal to the preschooler because of these qualities
(Favat, Note 3).
Children respond to the form of the fairy tale: its
regulated patterns (beginning and ending,) use of repetition,
patterned contrastive repetition, the simple nature of the
plot, cause and effect and short length. Saltz, Dixon
and Johnson (1977) attributed the benefits of thematic-fan-
tasy play training to several components of acting out fairy
tales. First, fairy tale enactment allowed children to deal
with events and themes extremely remote from their personal
experiences. Secondly, the strong plots in fairy tales and
their inherent cause and effect relationship improved chil-
dren's comprehension of causal relations. Third, the mo-
toric aspect of acting out fairy tales facilitated the par-
ticipation of children. Vygotsky (1962) stated that behav-
ing toward an object as if it were something other than what
it actually is is a basic factor in the development of mean-
ing and cognition of the child.
A fairy tale curriculum was chosen for this study
because of its appeal to preschool children for the reasons
discussed (Favat, Note 3). The plots of the storieswere strong
and simple, easy to understand, and there were causal rela-
tions inherent in the stories which facilitate comprehension
skills like cause and effect and sequencing.
Summary
Young children need quality experiences with litera-
ture to develop reading readiness skills, such as story com-
prehension. A method of enhancing story comprehension is
to develop a story schema, a set of expectations about how a
story fits together in a coherent, meaningful whole (Rumel-
hart, 1975). Low SES preschoolers have been the target of
research concerning reading readiness as defined by know-
ing letter sounds and names, visual and auditory discrimina-
tion. The best method for teaching this population these
reading readiness skills was direct instruction (Becker,
1977; Bruner, 1968). However, the same children who devel-
oped isolated reading readiness skills did poorly on compre-
hension tests (Bereiter & Engelman, 1966). This led
to the conclusion that direct instruction in phonics and
isolated skills does not enhance story comprehension. The
best method for teaching young children story schema for
the development of story comprehension is in the process of
being researched. Studies which have utilized a story en-
actment treatment have concluded that story enactment is
helpful for developing story comprehension. Teacher-led
discussion, the traditional method of teaching comprehen-
sion, has proved effective for developing story schema if
inferential questions are used. The treatments in this
study, story enactment of fairy tales and teacher-led dis-
44
cussion of fairy tales, were chosen because they seem most
appropriate for helping young children develop story sche-
mata which will, in turn, help them comprehend stories bet-
ter.
CHAPTER THREE
RESEARCH DESIGN AND METHODOLOGY
This study extended the knowledge about how low income
black low socioeconomic status four year olds recalled
story information. The treatments in this study repre-
sented two teaching modes, a structured yet abstract dis-
cussion led by the teacher and a structured, active treat-
ment in which children were read a story, assigned roles,
and then enacted the story in groups of four with active par-
ticipation by the teacher. The control group in this study
heard the same stories but did not participate in discussion
or enactment.
Subjects
Subjects were 45 four year old children in six Alachua
County Coordinated Child Care centers which are funded by
Title XX federal funds. Title XX is a federal program
which subsidizes the cost of daycare for parents who are
working or in training for employment in some institution.
The children who attend these centers are from low socio-
economic status homes. All of the children who participated
in this study were black and four years old (3.5 4.5
years). Seventeen children were lost from the study.
Sample Selection
Eight teacher volunteers were solicited through the
administrative office of Alachua County Coordinated Child
Care. Treatments were randomly assigned to these volun-
teers. The use of volunteers makes this study susceptible
to external validity threats. The sample of 45 low income
black children was chosen in order to broaden the general-
ization of the results of this study. Research has shown
story enactment facilitates middle income children's cogni-
tion of formal elements of a story (Milner, 1982).
Variables
There were three levels of one independent variable,
the story enactment treatment, the teacher-led discussion
treatment, and the control condition in which the teacher
read to the group and there was no enactment or discussion.
The dependent variables were formal elements in two stories
told by children: formal opening, formal closing, story
unity, number of characters mentioned, number of incidents,
and number of conversational quotations. The recall section
of the story comprehension was a criterion-referenced test.
Instrumentation
The Test of Early Language, by Hresko, Reid and Ham-
mill (Note 6), is a new test of receptive and expressive
language. The test gives information by item
related to the content and form aspects of language. The
TELD was impressive in reference to reliability and valid-
ity. This will be reported in detail.
The TELD was normed on 1184 children in 11 states and
1 Canadian province. Norms are provided for every six month
interval from 3-0 to 7-11 years. Since this test will be
used for a black population, it should be stated that 8 per-
cent of the norming population for the TELD was black. This
corresponds to a comparable 11 percent of the nation which
is black.
Reliability of the TELD has been computed in two areas,
content and time. Internal consistency, or the degree of
homogeneity among the items within the test, yielded a coef-
ficient alpha of .91 for the four year old group. The stan-
dard error of measure spanned 1.75 to 2.0 for ages three to
seven and was 2.10 for four year olds. Coefficient alpha
for all ages averaged to .90. Test retest reliability with
two weeks between tests was computed using 177 children in
Dallas who were three to seven years old. Correlations were:
three year old group = .84, four year old group = .72, five
year old group = .86, six year old group = .85, and seven
year old group = .87. The correlation on the total 177 was
.90.
Validity reported included criterion-related and con-
struct. For criterion-related validity, correlations were
done with other valued measures of performance established
tests. Results are reported for each age level except four
years. Scores on TELD by three year olds were correlated
with the Zimmerman Preschool Language Scale equalling .46.
The correlation of TELD with TOLD for five year olds was
.66. The authors state that all tests run are significant
beyond the .01 level of significance and are large enough
to support the TELD's criterion-related validity. Construct
validity is reported in two ways. Age of child and mean raw
scores ascend developmentally as evidenced by the following
figures:
Mean Raw Score
3 years 8.7
4 years 14.9
5 years 22.4
6 years 26.3
7 years 31.5
Information on construct validity was computed by relating
the TELD scores at different ages to tests of intellectual
development, reading and school readiness. For four year
olds, the correlation with the Test of Early Reading was .54.
For five year olds the correlation with the Slosson Intelli-
gence Test was .75.
The CRT was developed by the experimenter. Teachers
participating in the study were asked to write down the ques-
tions which they thought were important for recall on Little
Red Riding Hood. From these questions the experimenter com-
piled a list of factual and inferential questions. Ques-
tions four and nine were the only inferential questions
asked. The other questions were factual (Appendix A). Cor-
rect answers and a distractor were provided for the ques-
tions. Since different versions of the story were used,
some questions had several answers which were counted cor-
rect. The procedure was that the teacher asked the question
and paused four or five seconds. Then the child was given
the correct answer and the distractor.
Formal elements of a story to be used in this study
were taken from The Child's Concept of a Story by Applebee
(1978). Audiotapes were recorded of children in the two
treatments and control group retelling two stories, after
the curriculum was completed. One story was the last in
the series of four and one was a less familiar story which
was not included in the curriculum. These tapes were num-
bered and coded by the experimenter. Formal elements will
be treated as separate dependent variables. Treatment and
control groups were compared as to whether the element was
included or not. The elements are: formal opening, formal
closing, unity. Number of incidents, number of characters
and number of conversational quotations were counted. There
were a total of eight dependent variables for each story.
Design
A three group pretest/posttest quasiexperimental de-
sign was utilized. Analysis of covariance was used for the
data analysis. The covariate was the pretest TELD score.
0 X O O
0 X O O
O X O O
TELD Retelling CRT
pretest tasks
X1 = story enactment treatment
X2 = teacher-led discussion treatment
X = control group--teacher reads only
C T, T,
I I I I
C = control group
T1 = story enactment group
T2 = teacher-led discussion group
Data Collection
Volunteer teachers were solicited through the Alachua
County Coordinated Child Care administrative office. Treat-
ments were randomly assigned to eight teachers, and workshops
were provided for them. Four year old students of the vol-
unteer teachers were the subjects of this study.
A site visit schedule was formulated. The site visits
were scheduled at least once a week for each participating
teacher. Each teacher was visited a minimum of four times.
Records were kept on the progress of the treatments.
Hypotheses
Little Red Riding Hood
All hypotheses were tested at .05 level of significance.
Hypothesis 1. There will be no difference between four
year old children in the enactment, teacher-led discussion
or control group in frequency of including a formal opening
in the first retelling task.
Hypothesis 2. There will be no difference between four
year old children in the enactment, teacher-led discussion
or control group in frequency of including formal closing in
the first retelling task.
Hypothesis 3. There will be no difference between four
year old children in the enactment, teacher-led discussion
or control group in unity of thematic development in the
first retelling task.
Hypothesis 4. There will be no significant difference
in mean scores on the pretest between four year old children
in the enactment, teacher-led discussion or teacher reads
only group.
Hypothesis 5. There will be no interaction between
treatment level and pretest score on total formal elements
of a story.
Hypothesis 6. Adjusting for differences on the pretest,
there will be no difference in mean scores for total formal
story elements between four year olds in the story enact-
ment, teacher-led discussion or teacher reads only group.
Hypothesis 7. There will be no pretest by treatment
interaction on number of characters.
Hypothesis 8. There will be no difference between four
year old children in the enactment treatment, teacher-led
discussion or control group in number of characters mentioned
in the story retelling task.
Hypothesis 9. There will be no pretest by treatment
interaction on number of incidents mentioned.
Hypothesis 10. There will be no difference between
four year old children in the enactment treatment, the
teacher-led discussion treatment or control group in number
of incidents mentioned in a story retelling task.
Hypothesis 11. There will be no pretest by treatment
interaction on number of conversational quotations.
Hypothesis 12. There will be no difference between
four year old children in the enactment treatment, the
teacher-led discussion treatment or control group in the
number of times conversational quotations are used in a
story retelling task.
Hypothesis 13. There will be no pretest by treatment
interaction on the total score of characters, incidents,
and conversational quotations.
Hypothesis 14. There will be no difference between
four year old children in the enactment treatment, teacher-
led discussion and control group in total score on charac-
ters, incidents and conversational quotations.
Hypothesis 15. There will be no interaction between
treatment level and pretest score on the criterion-refer-
enced test.
Hypothesis 16. Adjusting for differences on the pre-
test, there will be no difference in mean scores on a cri-
terion-referenced test in four year olds in the story enact-
ment, teacher-led discussion or teacher reads only story.
Gingerbread Man
Hypothesis 17. There will be no difference between
four year old children in the enactment group, the teacher-
led-discussion group and the control group in the frequency
of including a formal opening in the story retelling task.
Hypothesis 18. There will be no difference between
four year old children in the enactment, teacher-led discus-
sion and control group in frequency of including a formal
closing in the story retelling task.
Hypothesis 19. There will be no differences between
four year old children in the enactment, teacher-led discus-
sion and control group in unity or thematic development in
the story retelling task.
Hypothesis 20. There will be no interaction between
treatment level and pretest score on formal elements of a
story score.
Hypothesis 21. Adjusting for differences on the pre-
test, there will be no difference in the mean scores for
total formal elements used in the retelling task between
four year old children in the enactment group, the teacher-
led discussion group and the control group.
Hypothesis 22. There will be no treatment by pretest
interaction on number of characters.
Hypothesis 23. There will be no difference between
four year old children in the enactment, teacher-led discus-
sion treatment or control group in number of characters men-
tioned in the story retelling task.
Hypothesis 24. There will be no treatment by pretest
interaction on number of incidents.
Hypothesis 25. There will be no difference between
four year old children in the enactment, teacher-led discus-
sion and control group in number of incidents mentioned in
the retelling task.
Hypothesis 26. There will be no treatment by pretest
interaction on number of conversational quotations.
Hypothesis 27. There will be no difference between
four year old children in the enactment, teacher-led dis-
cussion and control group in number of times conversational
quotations were used in the retelling task.
Hypothesis 28. There will be no pretest by treatment
interaction on total score on characters, incidents and
conversational quotations.
Hypothesis 29. There will be no difference between
four year old children in the enactment treatment, teacher-
led discussion and control group in total score on charac-
ters, incidents and conversational quotations.
Assumptions
The experimenter assumed that the teachers could carry
out the treatments, children could tell stories, and the co-
variate would be highly related to storytelling ability.
The assumptions for analysis of covariance were that scores
were independent, variances were equal and scores were nor-
mally distributed at each level of the covariate, that there
was no interaction between pretest and posttest, and that
there was a linear relationship between the covariate and
the dependent variable.
Procedure
There were two phases to this study, teacher training
and curriculum implementation. Separate workshops were
conducted by the experimenter for teachers in each treat-
ment, teacher-led discussion, story enactment and the con-
trol group. Teachers in the enactment treatment were pro-
vided information on these topics: (1) the importance of
developing a sense of story in the young child, (2) setting
up the environment to facilitate pretend play, i.e., props,
toys, and time during the day for pretend play, (3) how
teachers can take an active role and model appropriate be-
havior for children in pretend play situations, (4) how to
facilitate peer-peer social interaction in groups of four
children, and (5) role playing of these techniques. The
workshops for the enactment treatment were approximately
two hours total. The teachers in the discussion treatment
received training in these areas: (1) the importance of de-
veloping a sense of story in the young child, (2) the use of
questioning to help children better understand formal ele-
ments of a story, and (3) role playing of techniques. The
control teachers were taught the Reading Aloud to Children
Scale (Appendix B).
The second phase of this study was the implementation
of one of three conditions by the trained teachers. The
teacher in each condition was provided a complete kit with
everything he or she needed. The enactment group received
a kit containing props and the book. The teacher-led dis-
cussion group received a book, a script and a list of ques-
tions to use during the discussion time after reading the
story. The control group received the book only.
Teacher Training Workshops
Story Enactment Workshop
Teachers were instructed on why story enactment was a
valuable activity. Play and its importance for facilitat-
ing the young child's intellectual growth, creativity and
social skills and as the child's natural mode of learning
were discussed. Language and its link to the process of
play were also discussed. A brief review of the studies
which have utilized story enactment with positive effects
for children was presented.
The purpose of a story enactment treatment is to ex-
tend children's use of language and concept of a story.
Language competence is basic to school progress. If the
language of these children can grow in vocabulary and mean-
ing, it will be beneficial for reading achievement and
achievement in other areas. Knowledge of story structure
will also help these children with reading achievement and
general comprehension as a mental exercise in encoding in-
formation for retrieval.
The classroom environment, how it is set up and how it
operates, greatly influences the development of language ex-
tension and play. Time during the day must be allocated for
play, either with or without teacher involvement. Props are
essential to embellish play in the preschool years because
these children are just beginning to use symbols. Also,
teachers can stifle or enrich the play environment by model-
ing play behavior or setting up structured situations for
play, i.e., acting out stories. The teacher's attitude
toward play will greatly influence the outcome of the play
experiences.
The procedure to be followed was presented. Teachers
role-played the enactment procedure. Teachers read the
story before any enacting by the children. The teacher
randomly divided the class into groups of four. These same
four children enacted the stories together over the four
week treatment. The props were distributed. Groups were
invited to watch each other enact and chant the dialogue
with the speakers. The teacher might be the narrator or
might take a role. It was all right for a child to be the
leader.
After the story enactment, the teacher gave children
feedback about how they did and prepared for the next enact-
ment group. Teachers worked with two groups per day in
story enactment. One story was to be enacted each week.
Two versions of each story were given to the teacher. Mon-
day through Thursday were official days for enactment.
Friday were utilized if children were absent or for some
reason the treatment was missed one day.
Teacher-Led Discussion Workshop
The teachers in this workshop were introduced to ma-
terial on the importance of story structure and story com-
prehension. The training focused on how to carry out the
questioning technique. The importance of wait-time when
teaching through discussion was explored. Teachers were
instructed about how the questions were developed from
story schema theory. The experimenter emphasized the im-
portance of planning discussion by constructing questions
with goals in mind, i.e., formal elements of a story. An-
swers to the questions were provided in the lesson plan.
These were discussed with the teachers. Teachers were
given a set of questions for each story with specific ques-
tions for each day. The lesson plans are shown in Appen-
dices F, G, H, and I. Teachers were asked for comments and
suggestions.
Control Group Workshop
Teachers in the control group were trained in methods
of reading aloud to groups of children. The following tech-
niques were discussed: reading with expression, pointing
to words and pictures, choosing appropriate books for young
children. The Reading Aloud to Children Scale (Appendix B)
was discussed with the teachers.
Curriculum Implementation
Treatments occurred simultaneously in all centers dur-
ing the 30 minute morning story hour time. The curriculum
was implemented for four weeks. Following the four week
curriculum, children were posttested on three measures, two
retelling tasks (one familiar and one unfamiliar) and a re-
call measure, a criterion-referenced test. Children were
randomly asked to retell the last story of the literature
curriculum, Little Red Riding Hood. These retellings were
coded for formal elements of a story included by the child.
The child was asked 10 questions about Little Red Riding
Hood before retelling it. The child was read a less fa-
miliar story and asked to retell that story also. It was
also coded for formal elements of a story included by the
child. Data were analyzed and recommendations were made
for teachers on the effectiveness of each literature cur-
riculum carried out by the daycare center teachers.
CHAPTER FOUR
RESULTS
This study investigated the effects of two story treat-
ments, enactment and teacher-led discussion, on preschool
children's story comprehension. The control group listened
to the same stories read. There were 17 dependent variables
included in the posttests, two retelling tasks and one cri-
terion-referenced test designed to measure recall. The
children were asked to retell two stories, Little Red Riding
Hood and The Gingerbread Man. Little Red Riding Hood was
familiar because it was the last in the series of four in
the treatment. The other story, The Gingerbread Man, was
a less familiar story which the children in all three groups
heard read only one time. The purpose of a retelling task
with an unfamiliar story was to ascertain if the benefits
of enactment or teacher-led discussion transfer to other
situations. The criterion-referenced test was administered
on Little Red Riding Hood, which was the last in the lit-
erature curriculum for all subjects.
The three treatments did not have an equal number of
subjects. The enactment group had 25, the discussion group
had 11, and the control group had 9 subjects. The existence
of an unequal number of subjects does not seriously affect
the analysis of covariance results as long as the assump-
tions of ANCOVA are not violated.
61
Six dependent variables were analyzed by chi-square
because of the yes/no nature of the responses. The total
sample of 45 individuals was used to test these hypotheses
which concerned formal opening, formal closing and unity on
both stories. The other 11 dependent variables were ana-
lyzed by analysis of covariance. In the enactment group,
the same four children enacted the stories over the four
week period. The individuals in these groups were not in-
dependent of each other, so the means of eight groups were
used as scores for the enactment group only. Individual
scores were used for the discussion and control groups.
The sample size for all of the ANCOVA analyses was 28.
The experimenter did two analyses, one using individual
scores for all three groups (N=45) and one using the eight
group means for the subjects in the enactment group (N=28).
There was no difference in the outcome of the analysis, so
only the analysis using group means for the enactment treat-
ment was reported.
Little Red Riding Hood
Formal Elements of a Story
Formal Opening
Formal opening was defined by whether the child made
it clear a story was beginning.
Hypothesis 1: There will be no difference be-
tween four year old children in the enactment,
teacher-led discussion or control group in
frequency of including a formal opening in
the first retelling task.
The chi-square statistic was 6.229. The p value was
.0444. Hypothesis 1 was rejected at a=.05. There was a
statistically significant difference between children in the
three groups in the use of a formal opening in the retelling
task. Twenty-one out of 25 subjects in the enactment group
did not use a formal opening. Significantly fewer children
in the enactment group did not use formal opening.
Results for formal opening are shown in Table 3.
Table 3.
Chi-Square for Formal Opening
Little Red Riding Hood
Did Use Did Not Use
Group Formal Opening Formal Opening Total
Enactment 4 21 25
Discussion 6 5 11
Control 4 5 9
Total 14 31 45
N = 45
X = 6.229
Formal Closing
Formal closing was defined by whether the child made
it clear in the retelling that the story was ending.
Hypothesis 2: There will be no difference be-
tween four year old children in the enactment,
teacher-led discussion or control group in fre-
quency of including a formal closing in the first
story retelling task.
The chi-square statistic was 3.410. The p value was
.1818. Hypothesis 2 was not rejected at a=.05. There was
no statistically significant difference between the groups
on use of a formal closing. There was no relationship be-
tween group and use of formal closing.
Results for this variable are shown in Table 4.
Table 4.
Chi-Square for Formal Closing
Little Red Riding Hood
Did Use Did Not Use
Group Formal Closing Formal Closing Total
Enactment 11 14 25
Discussion 5 6 11
Control 1 8 9
Total 17 28 45
N = 45
2 = 3.410
x = 3.410
Unity
Unity was defined as a measure of the child's skill in
retelling the story using a sense of thematic development.
Hypothesis 3: There will be no difference
between four year old children in the enact-
ment, teacher-led discussion or control
group in frequency of using unity in the
first retelling task.
The chi-square statistic was .739. The p value was
.6910. Hypothesis 3 was not rejected at a=.05. There was
no statistically significant difference between the groups
on unity. Similar proportions of children in each group did
not use unity.
Results for unity are shown in Table 5.
Table 5.
Chi-Square for Unity
Little Red Riding Hood
Group Did Use Unity Did Not Use Unity Total
Enactment 15 10 25
Discussion 8 3 11
Control 5 4 9
Total 28 17 45
N = 45
2 = .739
x = .739
Pretest
A one-way analysis of variance was done on the TELD
pretest to ascertain if there were differences between the
three groups.
Hypothesis 4: There will be no significant
difference in mean scores on the pretest be-
tween four year old children in the enactment,
teacher-led discussion or control group.
The F statistic was .28. The p value was .76. Hypoth-
esis 4 was not rejected at a=.05. There was no statistically
significant difference between the three groups at a=.05.
The language abilities of the three groups were similar prior
to the experiment.
The results for the ANOVA on the pretest are shown in
Table 6.
Table 6.
TELD Pretest Data
Variable df SS MS F
Pretest 2 29.679 14.83 .28
Error 42 2256.230 52.47
N = 45
Total Score: Use of Formal Story Elements
A total score of use of formal story elements was cal-
culated scoring one point for each, formal opening, formal
closing, and story unity. A high score of three was pos-
sible.
Hypothesis 5: There will be no interaction
between treatment level and pretest score on
formal elements of a story score.
The F statistic was 1.59. The p value was .2273. Hy-
pothesis 5 was not rejected. There was no pretest by treat-
ment interaction at a=.05.
Hypothesis 6: Adjusting for differences on
the pretest, there will be no significant
difference in the mean scores for total for-
mal story elements between four year old
children in the enactment, teacher-led dis-
sion or control group.
The adjusted means for the groups were Enactment, 1.49;
Discussion, 1.82; Control, 1.09. The F statistic was 1.38.
The p value was .2715. There was no statistically signif-
icant difference between the three groups on this variable.
The results for this variable are shown in Table 7.
Table 7.
Total Formal Elements of a Story Score
Little Red Riding Hood
Variable Df SS MS F
Pretest 1 .202 .2020 .21
Treatment 2 2.610 1.3000 1.38
Error 24 22.810 .9504
Adjusted means for groups
Enactment 1.49
Discussion 1.82
Control 1.09
N = 28
Number of Characters
The total number of characters each child mentioned in
the story retelling was counted ou.t of a possible five.
Hypothesis 7: There will be no pretest
by treatment interaction on number of
characters.
The F statistic was .48. The p value was .6233. Hy-
pothesis 7 was not rejected. There was no interaction at
a=.05.
Hypothesis 8: There will be no difference
between four year old children in the enact-
ment, teacher-led discussion or control
group in number of characters mentioned in
the story retelling task.
The adjusted means for the groups were Enactment, 4.18;
Discussion, 4.17; Control, 3.67. The computed F statistic
was 1.48. The p value was .2475. Hypothesis 8 was not re-
jected. There was no statistically significant difference
between the three groups on number of characters mentioned
in the retelling task.
Results for this variable are shown in Table 8.
Table 8.
Number of Characters
Little Red Riding Hood
Variable df SS MS F
Pretest 1 .010 .010 .02
Treatment 2 1.550 .775 1.48
Error 24 12.560 .523
Adjusted means for groups
Enactment 4.18
Discussion 4.17
Control 3.67
N = 28
Number of Incidents
The total number of incidents mentioned by the child
in the retelling was counted.
Hypothesis 9: There will be no pretest by
treatment interaction on number of incidents
mentioned.
The F statistic was .81. The p value was .4586 at
a=.05. Hypothesis 9 was not rejected. There was no inter-
action.
Hypothesis 10: There will be no difference
between four year old children in the enact-
ment, teacher-led discussion or control
group in regard to number of incidents men-
tioned in the retelling.
The adjusted means on this variable were Enactment,
6.56; Discussion, 8.20; Control, 7.17. The F statistic was
1.28. The p value was .2973. There was no statistically
significant difference between the three groups in regard
to number of incidents.
Results for this variable are shown in Table 9.
Number of Conversational Quotations
The number of times conversational quotations were
used by the child in the retelling task was counted.
Table 9.
Number of Incidents
Little Red Riding Hood
Variable df SS MS F
Pretest 1 1.19 1.19 .23
Treatment 2 13.09 6.54 1.28
Error 24 123.10 5.12
Adjusted means for groups
Enactment 6.56
Discussion 8.20
Control 7.17
N = 28
Hypothesis 11: There will be no pretest
by treatment interaction on number of con-
versational quotations.
The computed F statistic was 12.66. The p value was
.0002. There was a pretest by treatment interaction. Hy-
pothesis 11 was rejected.
The slope for the enactment group was -.43. The slope
for the discussion group was -.033. The slope for the con-
trol group was 1.03. The regression equations for this
variable are
Y1 = 45.4 .43x
Y2 = 2.03 .033x
Y3 = -93.42 + 1.03x
The pretest by treatment interaction on this variable
indicates that scores on the pretest by subjects in the en-
actment group were negatively related to the number of con-
versational quotations as a dependent variable.
For subjects in the control group, there was a positive re-
lationship between pretest score and number of conversa-
tional quotations.
Hypothesis 12: There will be no differ-
ence between four year old children in
the enactment, teacher-led discussion or
control group in regard to number of times
conversational quotations were used.
The adjusted means for this variable were Enactment,
3.75; Discussion, 4.42; Control, 7.08.
The F statistic was 1.38. The p value was .2715.
There was no statistically significant difference at a=.05
between the three groups. The ANCOVA analysis results
should not be interpreted, because there was a pretest by
treatment interaction.
Results of the pretest by treatment interaction are
shown in Figure 1.
2
o
In /
0
80 90 100 110
Pretest
Figure 1.
Total Score on Characters, Incidents and
Conversational Quotations
For each category, a percentage was computed in regard
to number of characters, number of incidents and number of
conversational quotations. These percentages were added to
get a total score. For example, the highest score possible
on number of characters was 5. If a child used three in the
retelling, a percentage, 3 out of 5, was calculated for that
variable. The same process was used for number of incidents
and number of conversational quotations. Then the percen-
tages were added.
Hypothesis 13: There will be no pretest by
treatment interaction on total score on char-
acters, incidents and conversational quotations.
The computed F statistic was 2.88. The p value was
.0775. Hypothesis 13 was not rejected at a=.05.
Hypothesis 14: There will be no differ-
ence between four year old children in
the enactment, teacher-led discussion or
control group in regard to a total score
of characters, incidents and conversa-
tional quotations mentioned in the retell-
ing task.
The adjusted means for the group were Enactment, 1.59,
Discussion, 1.76; Control, 1.72. The F statistic was .55.
The p value was .5847. There was no statistically signif-
icant difference between groups in total score on charac-
ters, incidents and conversational quotations.
The results for total score on these variables are
shown in Table 10.
Table 10.
Total Score of Characters, Incidents, and
Conversational Quotations
Little Red Riding Hood
Variable df SS MS F
Pretest 1 .3143 .3143 2.36
Treatment 2 .1461 .0730 .55
Error 24 3.1900 .1330
Adjusted means for groups
Enactment 1.59
Discussion 1.76
Control 1.72
N = 28
Criterion-Referenced Test
A 10 item test was administered orally to each child.
Hypothesis 15: There will be no interaction
between pretest and treatment on a criterion-
referenced test.
The F statistic was .55. The p value was .5872. Hy-
pothesis 15 was not rejected. There was no interaction at
a=.05.
Hypothesis 16: There will be no difference
between four year old children in the enact-
ment, teacher-led discussion or control group
in mean scores on criterion-referenced test.
The adjusted means on the CRT were Enactment, 8.75;
Discussion, 8.13; Control, 5.56.
The F statistic was 43.08. The p value was .0001.
There was a statistically significant difference between the
enactment and discussion groups and the control group on re-
call of story information. Hypothesis 16 was rejected at
t=.05.
Results are shown in Table 1i.
The computed t for group one versus group two was 1.78.
The p value was .0883. This indicated there was not a sta-
tistically significant difference between the enactment and
discussion groups. The computed t for group one versus
group three was 8.59. The p value was .0001. This indicates
a statistically significant difference when comparing the
enactment and control groups. The computed t for group two
versus group three was 7.46. The p value was .0001. This
indicates a statistically significant difference between the
discussion group and the control group. The enactment and
discussion groups did significantly better than the control
group on recall of story information.
Results are shown in Table 12.
Table 11.
Chart for Criterion-Referenced Test
Little Red Riding Hood
Variable df SS MS F
Pretest 1 6.63 6.63 11.43
Treatment 2 49.99 24.90 43.08*
Error 24 13.92 .58
Adjusted means for groups
Enactment 8.76
Discussion 8.13
Control 5.56
N = 28
* Significant at .05
Table 12.
Estimates for the CRT
Estimates
Parameter Estimate T for Ho P
1 vs 2 .628 1.78 .0883
1 vs 3 3.200 8.59 .0001
2 vs 3 2.570 7.46 .0001
The Gingerbread Man
Formal Elements of a Story
The first three dependent variables, formal story ele-
ments, formal opening, formal closing and unity, were ana-
lized by chi-square.
Formal Opening
Hypothesis 17: There will be no difference
between four year old children in the enact-
ment, teacher-led discussion or control group
in frequency of including a formal opening in
the story retelling task.
The computed chi-square statistic was 2.881. The p
value was .2452. Hypothesis 17 was not rejected at a=.05.
There was no statistically significant difference between
the groups or use of a formal opening when retelling The
Gingerbread Man.
The results for formal opening are shown in Table 13.
Table 13.
Formal Opening
The Gingerbread Man
Did Use Did Not Use
Group Formal Opening Formal Opening Total
Enactment 4 21 25
Discussion 3 8 11
Control 0 9 9
Total 7 38 45
N = 45
2
X2= 2.811
Formal Closing
Hypothesis 18: There will be no difference
between four year old children in the enact-
ment, teacher-led discussion or control group
in frequency of including a formal closing in
the story retelling task.
The chi-square statistic was 3.740. The p value was
.1541. Hypothesis 18 was not rejected at a=.05. There was
no statistically significant difference between the groups
in regard to formal closing.
The results for this variable are shown in Table 14.
Table 14.
Formal Closing
The Gingerbread Man
Did Use Did Not Use
Group Formal Closing Formal Closing Total
Enactment 6 19 25
Discussion 6 5 11
Control 2 7 9
Total 14 31 45
N = 45
2
X = 3.740
Unity
Hypothesis 19: There will be no difference
between four year old children in the enact-
ment, teacher-led discussion or control group
in unity or thematic development in the story
retelling task.
The chi-square statistic was 13.073. The p value was
.0014.
Hypothesis 19 was rejected at a=.05. There was a sig-
nificant difference between the enactment treatment, teacher-
led discussion and control group in regard to unity or the-
matic development as a formal story element. One hundred
percent of the teacher-led discussion group, 64 percent of
the enactment and 22 percent of the control group demon-
strated unity in their retelling of The Gingerbread Man.
Results for unity are shown in Table 15.
Table 15.
Unity
The Gingerbread Man
Did Use Did Not Use
Group Unity Unity Total
Enactment 16 9 25
Discussion 11 0 11
Control 2 7 9
Total 29 16 45
N = 45
2
X = 13.073
Total Score for Formal Elements of a Story
Hypothesis 20: There will be no interaction
between treatment level and pretest score on
formal elements of a story score.
The F statistic was .81. The p value was .4591. Hy-
pothesis 20 was not rejected. There was no treatment by pre-
test interaction.
Hypothesis 21: Adjusting for differences on
the pretest, there will be no difference in
the mean score for total formal elements
used in the retelling task between four
year old children in the enactment, teacher-
led discussion or control group.
The adjusted means for the groups were Enactment, 1.10;
Discussion, 1.81; Control, .44.
The F statistic was 11.91. The p value was .0003.
There was a statistically significant difference between
the three groups in regard to total score on formal elements
of a story. Hypothesis 21 was rejected at a=.05.
The computed t for group one versus group two was
-2.47. The p value was .0209. The discussion group did
better than the enactment group on this variable. The com-
puted t for group one versus group three was 2.15. The p
value was .0415. There was a statistically significant dif-
ference favoring the enactment group over the control group.
The computed t for group two versus group three was 4.86.
The p value was .0001. This indicates a statistically sig-
nificant difference between the discussion group and the con-
trol group. The enactment and discussion groups did signif-
icantly better than the control group in total formal ele-
ments of a story score. And the discussion group did better
than the enactment group on this variable. When looking at
formal story elements as a total score, the two treatments
were effective in developing an awareness of formal elements
in four year old children.
Results for total score on formal elements of a story
are given in Tables 16 and 17.
Table 16.
Total Formal Elements
The Gingerbread Man
Variable df SS MS F
Pretest 1 .013 .013 .04
Treatment 2 9.160 4.580 11.91*
Error 24 9.220 .384
Adjusted means for groups
Enactment 1.10
Discussion 1.81
Control .44
N = 28
*Significant at a .05
Table 17.
Estimates for Total Formal Elements
The Gingerbread Man
Estimates
Parameter Estimate T for Ho P
1 vs 2 .71 -2.47 .0209
1 vs 3 .65 2.15 .0415
2 vs 3 1.36 4.86 .0001
Number of Characters
Hypothesis 22: There will be no pretest by treat-
ment interaction on number of characters.
The F statistic was 1.62. The p value was .2208.
This hypothesis was not rejected. There was no interaction
at a=.05.
Hypothesis 23: There will be no difference
between four year old children in the enact-
ment, teacher-led discussion or control group
in the number of characters mentioned in the
story retelling task.
The adjusted means for the groups were Enactment, 4.62;
Discussion, 4.37; Control, 4.85. The computed F statistic
was .41. The p value was .5586. Hypothesis 23 was not re-
jected at a=.05. There was no statistically significant
difference between the three groups.
Results for number of characters are shown in Table 10.
Number of Incidents
Hypothesis 24: There will be no treatment
by pretest interaction for number of incidents.
The F statistic was .49. The p value was .6195. This
hypothesis was not rejected at a=.05. There was no treat-
ment by pretest interaction.
Table 18.
Number of Characters
The Gingerbread Man
Variable df SS MS F
Pretest 1 .8358 .8358 .63
Treatment 2 1.0900 .5450 .41
Error 24 31.9300 1.3300
Adjusted means for groups
Enactment 4.62
Discussion 4.37
Control 4.85
N = 28
Hypothesis 25: There will be no difference
between four year old children in the enact-
ment, teacher-led discussion or control group
in number of incidents mentioned in the re-
telling task.
The adjusted means on this variable were Enactment,
5.84; Discussion, 6.10; Control, 5.41.
The F statistic was .62. The p value was .5461. Hy-
pothesis 25 was not rejected at a=.05. There was no dif-
ference between the three groups in number of incidents.
Results for number of incidents are given in Table 19.
Table 19.
Number of Incidents
The Gingerbread Man
Variable df SS MS F
Pretest 1 .439 .439 .24
Treatment 2 2.29 1.145 .62
Error 24 44.46 1.85
Adjusted means for groups
Enactment 5.84
Discussion 6.10
Control 5.41
N = 28
Number of Conversational Quotations
Hypothesis 26: There will be no treatment by pretest
interaction on number of conversational quotations.
The F statistic was 1.35. The p value was .2789. Hy-
pothesis 26 was not rejected. There was no treatment by
pretest interaction.
Hypothesis 27: There will be no difference
between four year old children in the enact-
ment, teacher-led discussion or control group
in number of times conversational quotations
were used in the retelling task.
The adjusted means on this variable were Enactment,
5.84; Discussion, 6.10; Control, 5.41.
The F statistic was 1.88. The p value was .1750. Hy-
pothesis 27 was not rejected at a=.05. There was no statis-
tically significant difference between the three groups on
number of conversational quotations.
Results for number of conversational quotations are
given in Table 20.
Table 20.
Number of Conversational Quotations
The Gingerbread Man
Variable df SS MS F
Pretest 1 5.45 5.45 .67
Treatment 2 30.35 15.17 1.88
Errov 24 194.20 8.09
Adjusted means for groups
Enactment 4.79
Discussion 6.23
Control 3.76
N = 28
Total Score on Characters, Incidents and
Conversational Quotations
For each category a percentage was computed of number
of characters, number of incidents, and number of conversa-
tional quotations. These percentages were added to get a
total score.
Hypothesis 28: There will be no pretest by
treatment interaction on total score on char-
acters, incidents and conversational quotations.
The F statistic was 2.19. The p value was .1358. Hy-
pothesis 28 was not rejected at a=.05. There was no pre-
test by treatment interaction.
Hypothesis 29: There will be no difference
between four year old children in the enact-
ment, teacher-led discussion or control group
in regard to a total score on characters, in-
cidents and conversational quotations.
The adjusted means for the groups were Enactment, 1.68;
Discussion, 1.80; Control, 1.57.
The F statistic was .71. The p value was .5032. There
was no statistically significant difference between the
groups in total score on characters, incidents, and conver-
sational quotations. Hypothesis 29 was not rejected.
Results for total score of characters, incidents and
conversational quotations are shown in Table 21.
Tables 22 and 23 summarize the results of this study.
Post hoc analysis on teacher effects is presented next.
Post Hoc Analyses
The analysis described in the previous section tested
hypotheses concerning treatment effects. The assumption was
made that teachers were equally effective within groups.
Table 21.
Total Score of Characters, Incidents and
Conversational Quotations
The Gingerbread Man
Variable df SS MS F
Pretest 1 .0002 .0001 .000
Treatment 2 .2503 .1250 .710
Error 24 4.2500 .1770
Adjusted means for groups
Enactment 1.68
Discussion 1.80
Control 1.57
N = 28
Table 22.
Summary of Results for Little Red Riding Hood
Hypothesis
Hypothesis Statistically
Dependent Variable Number Significant
A. Formal Opening 1 *S at a.05
B. Formal Closing 2 NS
C. Unity 3 NS
D. Total Score 5-6 NS
Formal Elements
E. Number of Characters 7-8 NS
F. Number of Incidents 9-10 NS
G. Number of Conversa- 11-12 Pretest by
tional Quotations treatment in-
teraction (Hy-
pothesis 11)
H. Total Score Characters, 13-14 NS
Incidents and Conversa-
tional Quotations
Criterion-Referenced Test 15-16 *S ata.05
Hypothesis 16
NS = Not Significant
*S = Statistically Significant
Table 23.
Summary of Results for The Gingerbread Man
Hypothesis
Hypothesis Statistically
Dependent Variable Number Significant
A. Formal Opening 17 NS
B. Formal Closing 18 NS
C. Unity 19 *S ata.05
D. Total Score 20-21 *S ata.05
Formal Elements Hypothesis 21
E. Number of Characters 22-23 NS
F. Number of Incidents 24-25 NS
G. Number of Conversa- 26-27 NS
tional Quotations
H. Total Score Characters, 28-29 NS
Incidents and Conversa-
tional Quotations
NS = Not Significant
*S = Statistically Significant
After examining group means, the experimenter suspected the
existence of teacher effects. Further analyses were needed
to test the assumption of teacher effects on the pretest and
11 continuous variables of the total 17 dependent variables.
The post hoc analysis first tested the hypothesis that
teachers were equally effective within groups. If the hy-
pothesis was rejected, the effect of the treatment was re-
analyzed using teachers nested within programs as the error
term. In reanalysis only five of the variables indicated a
significant teacher effect. The results of these analyses
and the reanalysis of the treatment effect are reported be-
low. A summary table for these analyses on the dependent
variables follows at the end of the presentation of these
results.
Little Red Riding Hood
A total score on characters, incidents and conversa-
tional quotations was calculated by the method described
earlier in this chapter. The hypothesis tested for this
variable was: teachers are equally effective within each
treatment group on a total score on characters, incidents
and conversational quotations.
The computed F statistic was 5.03. The p value was
.0013. This hypothesis was rejected. There was a teacher
effect on this variable. Teachers were not equally effective
|