PARENT INVOLVEMENT IN THE
COMPOSING PROCESSES OF KINDERGARTEN CHILDREN
CONSTANCE RUTH GREEN
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
This study is dedicated to my daughters-Christy, Marie, and
Emily-the most important part of my life.
First, I would like to thank my committee: Dr, Linda Lamme,
for her prompt and thorough feedback, her patience, and her unfading
enthusiasm and friendship; Dr. Dorene Ross, for her thorough criticisms,
availability, and friendship; Dr. Suzanne Krogh, for her stimulating
questions and suggestions; Dr. Athol Packer, for his wonderful support
and calmness; and Dr. Linda Crocker, for her insightful, creative ideas
on measurement and data analysis and for being a responsive listener.
Next I would like to thank Noille Silk, Ann Amos, and Cindy
Holmes, for their time, suggestions, and involvement; Betsy Nies, for
collecting writing samples; Karen Kilgore, Rhonda Royston, and Sharen
Halsall for rating the writing and fluency samples; Jean Brown, Dwight
Rogers, Tish Denny, and Tess Bennett for encouragement and support, and
the wonderful parents and children who willingly and enthusiastically
participated in the study.
Finally, I would like to thank Christy, Marie, and Emily, for
helping me keep my life in perspective; Paul, for his gentle understand-
ing; and my parents and sister, Carol, for believing in me.
TABLE OF CONTENTS
ACKNOWLEDGMENTS . . . . . .
LIST OF TABLES . . . . . .
LIST OF FIGURES . . . . . .
ABSTRACT . . . . . . . .
I INTRODUCTION . . . . .
Definition of Terms ..
Statement of the Problem .
Theoretical Rationale . .
Significance of the Problem
Limitations and Assumptions
Summary . . . . .
II REVIEW OF RELATED RESEARCH AND LITERATURE .
Introduction . . . . . . . .
Connections between Drawing, Oral Language,
Reading, and Writing . . . . .
Drawing and Writing . . . .
Oral Language and Writing . . .
Reading and Writing . . . .
Summary . . . .
The Development of Writing . . . .
Writing Stages . . . . . .
Summary . . . . . . .
Invented Spelling . . . . . .
Spelling Development . . . .
Suimmarv . . . .
Fostering the Writing Process in
and at Home . . .
Writing Environments . .
Summary . . . . .
. . .
. . . . . iii
. . . . . . . ix
. . . . . . . x
. . . I9
. . . 11
. . . 11
. . . 13
. . 17
. . . 23
. . . 25
. . 25
. . . 30
. . . 31
. . 31
. . . 36
. . 37
. . . 37
. . . 45
Educational Programs Supplemented by
Parent Involvement . . . . . . . ... .47
Studies of Parent Involvement . . . ... 47
Summary . . . . . . . . .. . 57
Sex Differences in Language Arts . . . . .. 58
Reading and Oral Language . . . . .. 63
Summary . . . . . . . .... . 63
Summary . . . . . . . . .. . . 63
III ORDER OF THE STUDY . . . . . . . ... . 68
Sample . . . . . . . . ... .. .. . .68
Study Design . . . . . . . . ... 69
The Settings . . . . . . . . ... 69
Instrumentation . . . . . . . . . 73
Concepts about Print . . . . . .... .73
Lamme/Green Composing Scale . . . .. 74
Parent Response Sheet . . . . .... .75
Procedure . . . . . . . .... . .. 75
Data Collection . . . . .. . . . 76
Writing Samples . . . . . . ... 77
Fluency . . . . . . . .... . 78
Data Analysis . . . . . . . .... . 78
IV RESULTS . . . . . . . . ... . . . .80
Introduction . . . . . . . . .. ... 80
Treatment Effects on Concepts about Print . . .. 82
Treatment Effects on Writing Fluency . . . .. 87
Treatment Effects on Compositional
Writing Achievement . . . . . . . . 87
Effects of Class, Gender, and Interactions . . .. 90
Summary . . . . . . . . ... . . 97
V SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS . . . .. 100
Introduction . . . . . . . . . . 100
Summary of the Problem . . . . . . . . 100
Findings and Conclusions . . . . . . . 101
Concepts about Print . . . . . . . 102
Writing Fluency . . . . . . ... 102
Writing Achievement . . . . . .... 103
Class Differences . . . . . .... 104
Gender Differences . . . . . .... .106
Two-way Interactions . . . . . .... .106
Assistance during Writing . . . . . 107
Types of Writing Children Did at Home . .. .108
Parent Response Sheets . . . . .... 108
Summary . . . . . . . .. .. . 108
Recommendations . . . . .
Recommendations for Researchers
Recommendations for Teachers . .
Recommendations for Parents . .
Recommendations for Administrators
Summary . . . . . . .
Summary . . . . . . . .
A CONCEPTS ABOUT PRINT . . . . . . .
B LAMME/GREEN COMPOSING SCALE AND SAMPLES . .
C WRITING BOOKLET . . . . . . . .
D RESPONSE SHEET . . . . . . . .
E OUTLINE OF PARENT WRITING WORKSHOP . . .
F TEACHER LETTERS . . . . . . . .
G PARENT INTERVIEWS . . . . . . .
H LEAST SQUARES MEANS FOR CONTROL AND TREATMENT
GROUPS IN WRITING ACHIEVEMENT . . . .
REFERENCES . . . . . . . . . . .
BIOGRAPHICAL SKETCH . . . . . . . . .
. . . 19
. . . 146
. . . 158
. . . 161
. . . 171
. . 172
- . 179
LIST OF TABLES
1. Concepts about Print pretestt) means and
standard deviations . . . . . . .
2. Summary of correlation coefficients of the
pretest with the five dependent variables .
3. Summary of ANCOVA for Concepts about Print
4. Summary of least squares means for
Concepts about Print . . . . . .
5. Summary of ANCOVA for Writing Fluency . .
6. Summary of least squares means for Writing
Fluency . . . . . . . . . .
7. Summary of ANCOVA for Writing Achievement-
first posttest . . . . . . . .
8. Summary of least squares means for Writing
Achievement-first posttest . . . . .
9. Summary of ANCOVA for Writing Achievement-
second posttest . . . . . . . .
10. Summary of least squares means for Writing
Achievement-second posttest . . . .
11. Summary of ANCOVA for Writing Achievement-
third posttest . . . . . . . .
12. Summary of least squares means for Writing
Achievement--third posttest . . . . .
13. Total number of sharing episodes per class
14. People with whom children wrote (all classes)
15. People with whom children wrote (by class)
. . . . 83
. . . . 84
. . . . 85
. . . . 86
. . . . 88
. . . . 89
. . . . 91
. . . . 92
. . . . 93
. . . . 94
. . . . 95
. . . . 96
. . . . 106
. . . . 106
. . . . 106
16. Types of writing children did at home . . . . . 109
17. Number of response sheets returned . . . . .. 109
LIST OF FIGURES
1. Writing Achievement-second posttest, gender
x treatment interaction . . . . . . . . .
2. Writing Achievement-first posttest, gender
x class interaction . . . . . . . . . .
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
PARENT INVOLVEMENT IN THE
COMPOSING PROCESSES OF KINDERGARTEN CHILDREN
Constance Ruth Green
Chairperson: Linda Leonard Lamme
Major Department: Curriculum and Instruction
This study investigated the impact of a home composing curriculum
on kindergarten children's concepts about print (a measure of reading
readiness), writing fluency (the number of words children can write), and
writing achievement. In addition, the study investigated gender differ-
ences on the above measures, types of writing done by children at home,
and family members with whom children wrote.
A sample of 74 children was drawn from four kindergarten classes,
two at a university laboratory school and two at a public school.
Parents of children in the treatment group participated in a parent
workshop and received a booklet which introduced the concept of writing
with young children, the importance of writing, and the interrelationship
between writing and reading. During the ten-week treatment period
parents and children wrote together at home several times each week.
The home writing samples were brought to school for sharing time with
other children in the treatment group.
Data were analyzed by analysis of covariance, using the Concepts
about Print test as the covariate. This procedure was conducted for
each of the five dependent variables: Concepts about Print posttest,
writing fluency, and three writing achievement samples. The treatment
group was found to be significantly higher on the Concepts about Print
test, writing fluency, and the first posttest of writing achievement.
There were no significant gender differences on any of the measures.
Data from parent response sheets indicated that these children
wrote notes, letters, and lists more than other types of writing. They
wrote with their mothers or alone more often than with other family
The findings of this study suggest that a flexible home composing
curriculum, combined with the sharing of writing at school can have an
impact on reading readiness, the number of words children can write,
and possibly writing ability.
Research in many fields of education has demonstrated the
influence of the home environment in children's cognitive development
(Durkin, 1966; Gordon, Greenwood, Ware, and Olmsted, 1974). For
example, Gordon et al. (1974) stated that the home situation provided "a
major source of a student's pattern of achievement, as well as his per-
sonality structure . ." (p. 1). Olmsted (1980) reinforced this idea
in concluding that parental teaching style combined with the home
atmosphere had a decided impact on children's learning.
One area in which the out-of-school environment has been shown to
be particularly influential is that of reading. Certain characteristics
of the home have been found to be extremely important for reading
development (Chomsky, 1971; Durkin, 1966), leading educators to encourage
parental involvement in reading. In studying children who read before
formal instruction, Clark (1976) and Durkin (1966) found that none of
these children learned to read on their own. All had adults and/or older
siblings who modeled reading and writing behaviors and responded to the
children's interest in print by reading to them, providing writing
materials, and answering questions. Thus, both Durkin and Clark con-
cluded that the home environment was a crucial factor in the development
of reading competence.
Recent ethnographic studies of parents and their children as
they acquire literacy (Bissex, 1980b; Taylor. 1983) indicate that
parents' interactions with their children through writing, as well as
through reading, help children become literate. These studies point
out many ways in which writing fits naturally into family routines.
Few parent involvement programs in schools have focused on
Recommendations that parents provide reading materials, that
they read to their children, and that they demonstrate a
model of reading behavior are common. Perhaps corresponding
recommendations about writing should also be given. Early
exposure to and experience with writing as communication
may facilitate interest in both writing and reading and may
contribute to initial success in school. (Hall, Moretz,
and Staton, 1976, p. 585)
It seems logical that the home environment may provide an appropriately
subjective audience for beginning writers as well as beginning readers.
Possible suggestions for parental activities would include interacting
with children as they write (Dyson and Genishi, 1982), providing purposes
for writing, and responding positively to early attempts at written
Because the process of composing/writing is important not only
as a means of communication, but also in relationship to reading
(Cramer, 1978; Durkin, 1966; Hildreth, 1936; Shanahan, 1980), teachers
need to inquire into ways to foster composing at home, as well as at
school. Almost no experimental research exists in this area. The few
qualitative studies that have been done point to a need for controlled
quantitative studies (Graves, 1982).
Teachers have occasionally noted developmental differences and
differences in fine motor control between boys and girls (Maccoby and
Jacklin, 19/4). Specific differences in levels of writing or differen-
tial effects of home writing experiences on boys and girls have not
Definition of Terms
Composing/Writing-Communicating ideas and feelings through the process
of drawing, writing, or dictation
Handwriting-The motor process of putting graphic symbols on paper
Fluency-The number of words a child writes at any given sitting
Invented Spellings-Experimentation with sound/symbol relationships for
the purpose of writing words
Parent Involvement-The participation of parents in the home composing
project; this included workshop attendance, completing and
returning weekly forms, and working with children on at least
one composing activity each week
Home Composing Curriculum-A booklet consisting of eight composing
activities parents used with their children at home; the intro-
duction includes suggested parental behaviors, ideas for moti-
vating young writers, and a materials list; response sheets
are included in the parent booklet (Appendix E)
Statement of the Problem
The impact of a home composing curriculum at the kindergarten
level was investigated in this study. The results provide information
about the effects of the curriculum on children's writing achievement,
concepts about print, and writing fluency. This study contributed to
the present knowledge of parent involvement, the composing processes of
young children, and the interaction of these two areas.
Specifically, the purpose of the study was to examine the
following research questions:
1. Would children who received the home composing
curriculum display a higher mean score for reading
readiness than children in the control group when
reading readiness was measured by the Concepts about
2. Would children who received the home composing cur-
riculum display a higher mean level of writing fluency
than children in the control group when they were
asked to write as many words as they knew how to
3. Would children who received the home composing cur-
riculum display a higher mean level of writing
achievement than children in the control group when
writing achievement was measured by three separate
posttest writing samples, evaluated by the Lamme/Green
4. Would there be differences between the mean scores
of kindergarten girls and kindergarten boys who
were exposed to the home curriculum on any of the
5. Would there be any interactions between gender
and treatment on any of the five measures?
6. Would there be any interactions between class
and treatment on any of the five measures?
7. Would there be any interactions between gender
and class on any of the five measures?
8. Would there be differences among the mean
scores of the four classes on any of the five
9. Would there be any three-way interactions among
class, treatment, and gender on any of the five
According to research by Clark (1976) and Durkin (1966), the
home environment can provide reading and writing experiences that
schools find difficult to offer. Dyson and Genishi (1982) and Teale
(1982) posit the view that natural literacy development depends upon
social interaction and speech. For many children, the home affords a
situation where they can interact with adults and older siblings as
they begin to read and write. A supportive home environment can enhance
the interrelationships between reading and writing (Durkin, 1966;
Hildreth, 1926; Taylor, 1983; Teale, 1978, 1982).
The connection between reading and writing was documented
when Clay (1975) found a high correlation between writing vocabulary
(fluency) 'core', and her word .reading test score for children five
years, six months old. A correlation of 0.79 between Clay's Concepts
about Print and Word Reading tests indicate interrelationships among
these instruments and the constructs they measure.
Some specific items in the Concepts about Print test appear
to be directly related to the home writing curriculum used in this
study. Several items assess left-to-right movement. Adults model
left-to-right movement when they take dictation from children and
follow the text with their fingers as they read back the words. Per-
ception of word boundaries and matching spoken and written words can
also be learned through dictation and rereading experiences (Clay,
1975). Three items in the Concepts about Print test assess knowledge
of punctuation marks. Children can become aware of these symbols as
their parents model appropriate use of punctuation in meaningful
Though writing development does not follow a set sequence
(Vukelich and Golden, 1982, 1984), the Lamme/Green scale developed
for this study provides a hierarchy of categories through which
children typically progress as they learn to compose/write. Since
writing is enhanced by interaction with supportive adults and other
children (Dyson and Genishi, 1982; Teale, 1982), it seems likely that
the level of composing should increase as a result of experiences with
the home writing curriculum.
Significance of the Problem
In light of the academic skills orientation that is currently
flourishing in early childhood programs it is not difficult to
understand parental anxiety over reading and writing readiness.
Most parents are concerned that their children become competent
writers, but are unsure of materials and methods they can use to
facilitate the writing process. The results of this study will pro-
vide parents with information about the effectiveness of a field-
tested composing curriculum.
This study may also influence the ways in which teachers
approach parent involvement. Teachers are a potential source of infor-
mation for parents as they select home reading and writing activities.
With the many commitments and responsibilities of kindergarten
teachers, it is difficult for them to spend time developing home as
well as school curriculum. Teachers will be able to use materials
from this study as a guide when they plan parent workshops and home
writing curricula for their individual settings.
Knowledge about parent involvement should also be expanded by
the results of this study of home writing. Literature in early child-
hood reading aptly describes the home environments of early readers
(Durkin, 1966; Teale, 1978, 1982), but does not address specific activi-
ties and parental teaching strategies that foster writing at home.
Because the home situation is such an important factor in cognitive
development, it seems imperative that research on home learning be
extended to the area of writing. Research in the area of home writing
has seriously neglected quantitative data to justify home writing cur-
ricula, indicating a need for further study.
Limitations and Assumptions
The internal validity of this study may have been threatened
by interaction between the writing/composing curricula (history) that
were used in the classrooms and the treatment. It was difficult to say
whether or not the changes in the dependent variables could be attributed
solely to the treatment. Classroom curricula are discussed in Chapter
The generalizability of the study was limited to parents who
were able to attend workshops and spend a minimal amount of time each
week working on composing activities with their children.
Posttest scores could have been influenced by several factors.
In the researcher's classroom scores may have been higher because the
parents tried to do more to help with the study. Occasionally children
in the experimental group discussed the home writing activities with
the control groups. Parents were told that the other group would have
these activities the second semester and were urged to wait until that
time to participate.
The researcher investigated the effects of a kindergarten home
composing curriculum on three important areas of reading and writing
readiness: writing achievement, writing fluency, and concepts about
print. The study was designed to add to the existing knowledge of
parent involvement, early childhood composing, and the interaction of
these two variables. Results of the study should assist kindergarten
teachers in fostering composing at home and at school.
REVIEW OF RELATED RESEARCH AND LITERATURE
Parents are teachers of their own children, either conscious
or unconsciously, as they model, directly instruct, and respond to th ir
Most parents intuitively do an adequate job of helping their
babies learn to talk, but often do not feel prepared to teach their
children to write. Letter formation and correct spelling are more of on
emphasized than the trial and error approach that these parents used
when their children were learning to speak (Forester, 1980). In
addition, the tenor of language transactions between parents and their
young children is often directive and controlling (Hoffman and McCull; ,
1984). When parents suggest words to be written, spell, and give
directions about spacing, children may learn to feel incompetent about
Most children learn the intricacies of oral language by imita'-
ing what they hear at home and by experimenting with speech patterns id
rules. Sometimes infants play with language as they babble and repea'
sounds. At other times they use language for a purpose, e.g., "dat"
tells an adult what the child wants. Children's awareness of meaning
takes precedence over the specific language that is used (Donaldson,
1978). Parents are usually able to understand their own children's
language. Bissex (1981) states that "Children learn language among
people who respond to their meanings before their forms" (p. 787).
A parallel can be drawn between the development of oral and
written language. Just as speech develops only in an environment rich
in oral language, writing develops when children are regularly exposed
to print. The availability of writing and drawing materials and exposure
to adults who write are potent influences in the lives of young writers
(Bissex, 1981; Calkins, 1980; Durkin, 1966).
Research on comparative language environments demonstrates that
in programs where children's composition is an integral part of language
arts, children become more proficient readers and writers (Birnbaum,
1980; DeFord, 1981). Kindergarten children are just beginning to explore
the rudiments of written language. The home environment can do much
to support the school's writing curriculum if teachers are able to
communicate to parents appropriate ways to facilitate young children's
writing. There is a need for educators to learn the most effective
ways to foster parent involvement in the composing process.
Few research studies directly address this specific area.
However, research in several related areas suggests some potentially
important relationships between parent involvement and composing/
writing curriculum. A survey of the literature related to parent
involvement in writing revealed five areas for investigation: (a)
connections between drawing, oral language, reading, and writing;
(b) the development of writing in early childhood; (c) fostering the
writing process in early childhood; (d) education programs supplemented
by parent involvement; (e) sex differences in language arts.
Writing is not an isolated process. Oral language, drawing,
and sometimes reading precede and/or accompany young children's writing.
Scribbles and designs form a common origin for both drawing and writing
(Kellogg, 1969) and both processes can be enhanced by appropriate con-
versation (Hoffman and McCulley, 1984). Similarly, exposure to printed
materials and literacy experiences enhance both reading and writing
development. The interrelationships between these language and literacy
processes will be examined in the following section.
Drawing and Writing
For most children, writing evolves naturally from drawing (Emig,
1977; Kane, 1982; Zepeda de Kane, 1980). As children gain experience in
drawing, their scribbles become symbolic and often look like shapes or
letters (DeFord, 1981; Kellogg, 1969). Children who observe adults
writing and live in an environment where printed materials are valued
will be more likely to make "mock letters" (Birnbaum, 1980; Hall,
Moretz, and Staton, 1976) and attempt to read their own messages (Dyson,
1981). These children are beginning to understand that ". .the marks
on paper are a written version of speech" (Donaldson, 1978).
In studying over 500,000 children's drawings, Kellogg (1969)
found that 20 types of scribbles emerged. The scribbles involved the
use of six categories of lines: vertical, horizontal, diagonal, cir-
cular, alternating, and dots (no line movement). These strokes, which
appear in the formation of alphabet letters, are practiced spontaneously
I/ ,,iini children as they draw. When a two- or thrpe-year-old draws a1
%cribble thdt reosmbles d shape, this shape is often rcii( nimb( rd and
repeated in subsequent artwork. The shape stage follows patterned
scribbling and emerges in the third or fourth year, depending upon
individual experiences. Shapes evolve into six distinct diagrams:
rectangles (including squares), ovals (including circles), triangles,
Greek crosses, diagonal crosses, and odd shapes. At first diagrams
are combined with scribbles, then later with other diagrams, forming
combines. Aggregates, combinations of three or more diagrams, are
the most predominant art form of three- to five-year olds. Kellogg
(1969) observed that these stages are self-taught through practice,
rather than learned by adult instruction.
The diagrams, combines, and aggregates that Kellogg described
are not always perceived as drawing by young children. Frequently pre-
schoolers classify certain shapes or patterns in their drawing as
writing. The transition between drawing and writing represents a shift
in modes of self-expression. ". .[U]ntil the task of writing has
been mastered, the system of drawing is the only one sufficiently
elaborated to permit expression of inner life" (Gardner, 1980).
Scribbling experiences allow children to refine the motor skills
necessary for writing (Kane, 1982). Children move spontaneously from
scribbling to representational drawing to writing. The drawing/writing
process often allows children to think through their experiences before
expressing their thoughts orally (Kane 1982).
The literature regarding the relationship between drawing and
writing stressed the importance of scribbling and forming mock letters
as precursors to writing. Frequent and spontaneous experiences with
pencil and paper provide children with opportunities to move through
the developmental sequences of graphic representation.
Gardner (1980) also drew an analogy between language play and
art activities of young children. Both types of play featured recurring
patterns, versatility, and a concentration of energy. Oral language
frequently accompanies the drawing activities of young children
(Gardner, 1980; Kane, 1982).
Just as young children's drawing activities are often accompanied
by talk, writing also has an important link to oral language. The
following subsection will expore this connection.
Oral Language and Writing
Ethnographic and observational studies of beginning writers focus
on the importance of oral language during the writing process (Bissex,
1980b; Childers, 1981, Dyson and Genishi, 1982). Graves (1980) and two
other researchers observed sixteen first through third grade children
from a rural New Hampshire School over a two-year period, to gather
direct classroom information on the writing process. Data were col-
lected by audio and video taping during the writing process, interviews,
structured interventions, and analysis of the children's writing.
Though efforts were made to minimize the influence of the researchers
in the classroom, Graves admitted that their presence may have affected
some of the data.
Graves and his associates found that oral language almost always
accompanied the writing of these young children. The following types
of oral language were recorded: sounding to prove for sound-symbol
relationship, sounding to break off a phonetic unit from a word, reread-
ing the composition to reorient, conversation with friends, procedural
talk, advanced statement of text, and conversation before and after com-
posing. When teachers paid careful attention to the talk that sur-
rounded writing, they were better able to help children gain and main-
tain control of the writing process. "In summary, the amount of language
a child must produce before, during, and after the written event is
paramount. Beginning writers show through voice alone that writing is
much more of a speech event than a writing event" (Graves, 1982, p. 22).
Following Graves' observational research methods, Calkins (1983)
found that when third-graders shared their writing verbally with peers
they received constructive responses which helped them grow as writers.
The sharing sessions took several forms; whole class meetings with
several children sharing each time, clusters of three to four children,
quiet sharing where responses were written, sharing which focused on
one specific aspect of writing, sharing about the process of writing,
and giving writing to friends or the class library. The teacher in
the classroom emphasized that everyone in the classroom was a writing
teacher and encouraged verbal communication among the children as
they wrote, as well as during sharing times.
Lamme and Childers (1983) observed the composing processes of
three young children (ages three to five) in a laboratory setting. All
children were from white, middle-class homes. The children were video-
taped as they composed in weekly sessions for sixteen weeks. The
researcher presented a topic for each session, then became an
observer-participant and audience for the young writers. The taped
transcripts were later analyzed in four categories: composing behaviors,
behaviors accompanying composing, nature of oral interaction while com-
posing, and functions of oral interaction while composing.
It was reported that these children ". . talked constantly
while they were composing." They engaged in basically four types of
oral interaction during the composing sessions: questions, answers/
responses, sharing/telling, and taking breaks. By far, the sharing/
telling interactions were most common among this group of children.
Sixteen functions of oral language were also recorded, the most common
being response to questions, sharing products, and explaining (Childers,
1981; Lamme and Childers, 1983).
This study was one of the first to examine the activities of
young children as they composed. The researchers thoroughly categorized
behaviors of the children as they wrote, raising questions to be
answered by future research. The population and setting of the study
limited the generalizability of their results.
The relationship between the oral and written language of
kindergarteners was observed by Dyson (1983) in two separate three-
month studies conducted in a self-contained public school classroom.
Dyson set up a writing center in the classroom that the children could
come to whenever they wished. As an observer-participant, Dyson was
able to eavesdrop on conversations between children and collect writing
Names of friends and family members were the most conmnon words
written by these children. Second in popularity were lists of words that
labeled the environment. Both of these categories were particularly
relevant to the children. Words and topics for writing often emerged
as the children talked together at the writing center. Some children
were more dependent on this talk which surrounded the writing than
others. The four children discussed in this article used oral
language for different purposes as they wrote. One child used oral
language to interpret her writing. Another narrated her writing by
talking as she wrote. A third little girl used the talk around her
for writing stimulus.
Dyson (1981, p. 783) concluded that ". writing does not
necessarily begin with the understanding of the alphabetic principle."
For some children writing began with an idea, then they found a way to
put that idea into print. Other children began with print, then needed
to find a way to make that print meaningful. In either case, oral
language seemed to be the salient factor preceding or extending child-
Dyson's research is a valuable asset to the literature on
beginning writing. It is one of only a few studies which focus on
the role of oral language during composing. Two strengths of this
study were that it was conducted in two segments over a considerable
length of time (six months total) and that the researcher was able to
interact with the children as the study was being conducted.
Dyson and Genishi's (1982) observations of two first graders
confirmed findings by Graves (1981) and Dyson (1981). The social con-
text seemed to be extremely important to these young writers as they
clarified spellings and discussed their "work" back and forth. The
children also directed speech to themselves. In conclusion, the
researchers stated that "these children's interactions had positive
effects on their ability to write" and enhanced their capacity to "take
responsibility for their own learning-to seek out needed information-
and to contribute positively to another's learning" (Dyson and Genishi,
1982, p. 131).
Reading and Writing
The language arts are integrated processes for young children.
Just as writing depends on speech, reading can be learned or enhanced
through writing. The early levels of basal reading series and kits
used in most kindergartens approach reading first as the pronunciation
of a sequence of letters and then the comprehension of meaning (Kita,
1979). Chomsky (1971) proposed that this approach to reading disregards
what children already know about language and distorts children's
understanding of the purpose of reading. When children learn to read
primarily through phonetics they may come to view reading as a decoding
process, rather than a communication process (Kita, 1979).
Chomsky (1971) argues that developmentally children are able
to write before they are able to read. It is easier for young children
to compose words by the sound-symbol relationship than to decode an
unfamiliar word (Chomsky, 1971). Teaching reading before writing
seems to be a backwards approach, according to several researchers
(Chomsky, 1971; Graves, 1982; Hall, Moretz, and Staton, 1976; Hildreth,
Graves (1983) found that through writing, children were moti-
vated to read as well as write. "Children acquire perceptions by
writing. Eye, hand, mouth, and ear work together to aid a child to
understand the process of putting words on paper. Because they write,
children's perceptions expand. Children learn to read their own writing
and the writing of others which is very different from reading a pub-
lished reader or a library book" (Graves, 1983, p. 152).
Hall, Moretz, and Staton (1976) studied 18 children in four
nursery school classes in Maryland to determine the sequence of learning
to read in relation to learning to write. The children, who ranged in
age from three years, four months, to six years,one month, came from homes
where most of the parents were college graduates and the fathers were
professionals. Seventeen of the eighteen children showed an interest
in writing before an interest in reading. The availability of writing
materials, books, magazines, and newspapers, along with parental
responsiveness and the learning of letter names seemed to be salient
factors in the writing development of these children.
Kita (1979) studied the concepts of reading and writing of
twenty kindergarteners in Virginia from a range of socioeconomic back-
grounds. These children ranged in age from five years five months to
six years three months. Interviews were conducted within the classroom
setting. In response to the researcher, the children indicated that the
nature and purpose of reading could vary. Most believed that word
recognition was necessary for reading signs, but not for reading books.
They believed that an important part of reading books was looking at
pictures, and in fact, that was one way to read books. The most fre-
quent purpose the children mentioned for reading books was to learn
how to read. Their purposes for reading signs, labels, and newspapers
were similar to those of mature readers.
The same children's responses to questions about the nature and
purpose of writing showed a more accurate understanding of this process.
All reported that they had purposes when they wrote and that writing
helped in learning to read. Writing samples indicated that most of
the children tried to tell a story through writing. Kita (1979) con-
cluded that language learning should include both books with related
illustrations and an informal writing program where children "can
experiment with print in a meaningful context" (Kita, 1979, p. 8).
Although limited to a small number of children from only one
classroom, Kita's (1979) study has the strength of including lower SES
children in the sample. Using the teacher as an interviewer could
have affected the children's answers. The study is, however, a
beginning for research on children's concepts about reading and writing.
Research by Chomsky (1979) and Graves (1982) indicated that
children can begin composing words as soon as they have learned a few
letter-sound combinations. Graves (1983) reports that 90 percent of
the children entering school believed they could write. Only 15 percent
believed they could read.
Graves (1980) stated that children want to write. His year-
long investigation into the current status of writing in the United
States revealed a trend toward more reading, less writing, and an
emphasis on language arts subskills (e.g., grammar, punctuation) over
writing content. Ethnographic studies by Graves and his colleagues
(Calkins, 1980; Kamler, 1980) support the value of informal writing
experiences in learning to read and in children's choice of writing
topics. When children were encouraged to write independently and
invent their own spellings for words, grammar and punctuation improved
along with reading comprehension and word recognition (Shanahan, 1980;
Sulzby, 1980; Whiteman, 1980).
Birnbaum (1980)conducted a case study of eight children from
the fourth and seventh grades in two school districts. The students were
selected by teachers and administrators as the most proficient writers
and readers in their grades. Video-tapes were made of each student three
times during two semesters as she/he composed and read. Thirty hours
of classroom observation, as well as parents' and teachers' interviews,
were part of the study. Compositions written by the students were rated
by two Educational Testing Service (ETS) readers and the researcher.
Differences were found in several areas between the more pro-
ficient and less proficient writers. More proficient writers were more
reflective as they composed and showed an awareness of writing for an
audience. Less proficient writers viewed composition as an externally
imposed task and were more concerned with conventions of writing than
the meaning conveyed by their texts. The proficient writers saw them-
selves as good writers and readers, whereas less proficient writers said
their teachers thought they were good writers.
The home environment of the less proficient writers lacked two
features that were present in the homes of the more proficient writers.
The first characteristic was the presence of an adult who wrote
extensively and impressed upon the child the importance of writing.
The second was the presence of audiences who responded with interest
to the children's compositions. School environments of more proficient
writers stressed composing for audiences, peer conferences, and
dramatic presentation of student writings. The language arts curricu-
lum at the other school was based on a series of performance objectives
which allowed for individual pacing. The majority of the time students
worked alone on their instructional exercises.
Birnbaum's (1980) study made a thorough investigation into the
attitudes, composing styles, and academic and non-academic backgrounds
of eight young writers. By limiting the sample size, the investigator
was able to collect data on most factors that appeared to be relevant
to successful composing. The sample was limited in size, ages of
children, and geographic setting. However, the strengths of the study
far outweighted the weaknesses and merit consideration for further
research on environmental influences that affect children' writing.
The relationship between reading and writing also was studied
by Durkin (1966). In a longitudinal study of children who learned to
write and spell words, availability of a chalkboard and availability
of reading materials in the home were important influences to these
children. These early readers, identified from 61 schools in Oakland,
California, achieved consistently higher reading scores through the
sixth grade. The strengths of this study lie in both the large sample
size and the longitudinal aspects. Generalizability is limited by the geo-
graphic location and SFS level in the area where the study was conducted.
Zeman (1969) examined the relationship between reading compre-
hension and the basic sentence types and sentence structural patterns
in the compositions of 180 second and third graders in seven Allentown,
Pennsylvania, classrooms. Two of the schools were in a rural area
and one was in a suburban area. Students in each grade level were
separated by sex and reading level (above-average, average, below
average) based on reading comprehension scores. The sample population
was randomly selected from each of these twelve groups. Information
on SES status of the students was not provided. Writing analysis was
done on endings to unfinished stories. Analysis of the data revealed
that above-average readers used the most complex sentence structures in
To explore the impact of different language environments on
developing reading and writing strategies, DeFord (1981) collected
writing samples from three first grade classes over a seven-month
period. The language emphasis in the three classrooms stressed phonics,
skills, or whole language models. The phonics teacher introduced each
letter and sound separately to the class. The skills teacher used
drills, workbooks, and lessons from the Ginn 720 series. The teacher in
the whole language classroom integrated language experience, literature,
and children's writing.
Oral readin-g strategies were analyzed in all three classrooms by
Goodman's miscue analysis. Children in the phonics room exhibited a
high dependence on decoding strategies. They frequently pronounced
nonwords that looked and sounded like the text, but made no sense.
Children from the skills room had a high percentage of omissions and
substitutions of graphically similar words. Readers from the whole
language classroom manipulated the text by substituting words that
were similar in meaning to the author's words. To these children, it
seemed very important that the text make sense.
Reading comprehension was higher in the language/literature
classroom than in the other two classrooms. Children were better
at retelling stories, used more story conventions, and recalled more
story information. The writing of children in the whole language class-
room also "produced a wider variety of literary forms, such as stories,
informational prose, songs, poetry, and newspaper reports" (DeFord,
1981, p. 656).
In conclusion, DeFord stated that interaction between reading
and writing is necessary for children to become.literate. As they
read, children learn about writing, and they learn about reading as
they write. The classrooms that separated reading and writing instruc-
tion were, therefore, self-defeating.
The controversy over methods of reading instruction has con-
tinued for years. In this controversy, writing has been almost ignored.
Yet, the studies cited above point toward a strong positive relationship
between compositional writing and proficient reading for young children.
When children are provided with materials and allowed to talk and write
in an accepting environment, they tend to compose whole texts from the
bt niinninj (Deford, 1980; Gundlach, 1'I'1).
As children write they use letters and sounds to form words,
then phrases, sentences, and longer compositions that have meaning for
them. After the symbols are written they can be read back by the child
and understood. Hildreth (1964) postulates that
In the initial stages of learning to read any experience
with writing benefits reading no matter what methods are used
in reading instruction. With more emphasis on writing
paralleling the reading experience fewer children would
reach an early plateau in reading and be unable to read at a
normal rate (p. 16)
When young children are exposed to print and writing materials,
writing evolves naturally from drawing. Writing almost invariably
enhances the reading process when children master these abilities out-
side the school setting and without formal instruction (Chomsky, 1971;
Durkin, 1966). Chomsky (1971) emphasizes that through writing children
develop awareness of print and purposes for reading and writing. She
states that we should allow children to ". .. get to the point where
they can make their own productions before they are expected to read
other people's productions. This would be true preparation for learning
to read!" (p. 299).
The majority of studies that link the language arts areas
together have been cross-sectional and limited to middle SES children
and populations that were convenient to the researcher, leaving ques-
tions about their implications for more diverse populations. Few
studies have looked at the effect of sex differences on drawing or
oral language during the writing process. These areas need to be
explored by further research.
It is clear from these studies that all areas of language arts
are interconnected in young children's learning. The nature of these
connections over time and across populations, and outside influences on
these connections, need further investigation.
The Development of Writing
Children who have been exposed to print and who have had
experiences with drawing materials may begin to distinguish between
drawing and writing by age three (DeFord, 1980; Gardner, 1980; Hiebert,
1978; Lavine, 1972). During the early stages, a child's writing may
not appear much different from scribbling to an adult (DeFord, 1980).
The difference lies in the pattern of the scribbles and the child's
interpretation. When DeFord (1980) studied the spontaneous writing
processes of fifty children, ages two to seven, she found that graphic
symbols children called writing were usually characterized by linearity,
directionality, uniformity, flow, and rhythm and were interpreted as
messages by the writers. This understanding clearly differentiates
writing from drawing. DeFord observed that some rules of writing, such
as linearity, uniformity, and directionality, may "swing in and out of
conventional use between the ages of two and seven" (p. 159). The
development of writing is enhanced when children play alone with print
or share the writing experience with another child. Although DeFord
cautioned that the following stages she observed were not necessarily
sequential, they demonstrate a movement from global to specific concepts
These stages are
2. Differentiation between drawing and writing
3. Concepts of linearity, uniformity, inner complexity,
symmetry, placement, left-to-right motion, and top to
4. Development of letters and letter-like shapes
5. Combination of letters, possibly with spaces, indicating
understanding of units (letters, words, sentences), but
may not show letter/sound correspondence
6. Writing known isolated words-developing sound/letter
7. Writing simple sentences with use of invented spellings
8. Combining two or more sentences to express complete
9. Control of puncutation-periods, capitalization, use of
upper and lower case letters
10. Form of discourse-stories, information materials,
letters, etc. (DeFord, 1980, p. 162)
Harste, Burke, and Woodward (1981) video-taped three- to six-
year-old children engaged in three literary tasks: reading environmen-
tal print, writing their names and anything else they wished to write,
drawing self-portraits, and signing their names. The study included
children from lower, middle, and upper SES homes. Socioeconomic status
did not appear to be a relevant factor in children's performance on the
literary tasks. Like DeFord, these researchers found that three-year-
olds often distinguish drawing from writing. When writing, children
use different movements which appear to be more purposeful. The tape
recordings of the sessions also revealed that when children write,
they put meaning over object, rather than later labeling their marks.
Clay (1975) looked beyond writing development to children's
awareness of written language. She studied children in school settings
in New Zealand where copying words and tracing were encouraged. There-
fore, her progression may not reflect the natural sequence children
would go through without this form of stimulus. According to Clay,
children develop an awareness of written language in the following
1. Understanding that print talks
2. Forming letters
3. Building up memories of common words they can construct
out of letters
4. Using those words to write messages
5. Increasing the number and range of sentences used
6. Becoming flexible in the use of sentences
7. Disciplining the expression of ideas with the spelling
and punctuation conventions of Fnglish (Clay, 1975,
Clay posits that children do not learn about writing at one level, then
move to higher levels, as many curriculum sequences would imply.
Knowledge of all alphabet letters does not necessarily precede writing
words, sentences, and stories. Language learning is more intermingled
Through her research Clay (1980) revealed a number of principles
commonly found in children's early writing development. The recurring
principle, exemplified by repeated letters or words, gives young writers
the self-selected practice they need to provide them with a sense of
accomplishment. The directional principle refers to children's
developing the habit of starting at the left side of the paper when they
write. Clay reports that this principle usually takes about six months
to become established. Children demonstrate the generating principle
by writing strings of letters or words. The inventory principle is
exhibited when children write lists of letters or words. The contrast
principle is reflected in the writing of opposites-letters which con-
tain contrasting lines, or words which are opposite in meaning. When
using the abbreviation principle, children let one or two letters
represent a word. The flexibility principle refers to children's
tendency to explore the limits of letter and word formation. These
principles are not sequential but are integrated in different ways as
children gain experience with writing.
Clay did not address the transition from drawing to writing or
the impact of oral language on writing. Her subjects were taken from a
school district where writing instruction was already in progress,
therefore she was not able to observe the natural progress of children
who had not had formal writing instruction. The development of grapheme/
phoneme relationships and other spelling concepts were basically not
addressed by Clay.
Observational research by Vukelich and Golden (unpublished
manuscript) confirms DeFord and Clay's stages of writing development.
Vukelich and Golden investigated the nature of children's concept of
writing as a symbol system and the predominant patterns in children's
writing. They collected writing samples from 34 four-year-olds and
39 five-year-olds on two different days in October, January, and April.
The children were invited to a writing table in groups of five or six,
given writing books and pencils, and directed to "Write anything you
wish to write." When children indicated they were finished, the
researcher asked them individually to "Tell me what you wrote."
The writing samples were classified into seven categories, with
several subcategories in each. The main categories were
1. Drawing scribbles
2. Linear scribbles
3. Shapes, lines, symbols
4. Recognizable picture
5. Recognizable picture and unrelated writing
6. Recognizable picture and relating writing
7. Writing only
This study exemplified the wide range of drawing/writing pro-
duced by nursery school and kindergarten children and the progress they
made during the year. Vukelich and Golden observed two distinct stages
in the children's developing concept of message. First was the recog-
nition that symbols are necessary to write; second, an awareness that a
"precise message can be represented using written symbols" (Vukelich
and Golden, p. 19). These children write letters and numbers in
strings, rather than scattering them about the page, thus demonstrating
an early awareness of linearity.
The recent findings of Harste, Burke, and Woodward (1981)
supported Vukelich and Golden's (unpublished manuscript) observation
that age is not necessarily correlated with language development. Their
study of the effects of home environments on the reading and writing
development of three- to six-year-olds indicated that when children
have equal opportunities to write and a wide exposure to environmental
print, their responses to writing tasks do not differ across age levels.
Several important implications emerge from the research on
children's writing development. First, it appears -that children can
differentiate their writing from drawing at an early age (DeFord, 1980;
Gardner, 1980; Harste, Burke, and Woodward, 1981; Hiebert, 1978;
Lavine, 1972). Early writing appears before children produce letters
or words that can be understood by adults (Clay, 1980; Vukelich and
Golden, unpublished manuscript). Secondly, writing development does
not necessarily follow an orderly sequence (Clay, 1980; Ford, 1980;
Newman, 1983; Vukelich and Golden, unpublished manuscript). Thirdly,
children's writing should be interpreted within the context of oral
language in order to understand the meaning of the message (Harste,
Burke, and Woodward, 1981; Newman, 1983; Vuchelich and Golden, unpub-
lished manuscript). Finally, children seem to practice common principles
as they gain experience in early writing (Clay, 1980).
Some of the earliest research on invented spelling was con-
ducted by Read (1971), who found that a few children began to write
their own words at age three-and-a-half, often before learning to read.
The youngest children Read studied used blocks or other alphabet toys
to form words. He found that instead of memorizing, these children
were learning ". .. a complex, but generally systematic phonology
system (Read, 1971, p. 1).
Read presented a complex sequence that he found children
followed in developing spelling skills. The first sounds that children
related to letters were clear consonants (p t k b d f v s z j m n r 1).
The next step was for children to learn long vowel sounds. Read
believed that children distinguished letter names from the sounds they
represent. One child in his study represented this by putting dots
over names of letters. At about the same time children were usually
taught some sight words by their parents. The acquisition of ending
sounds and short vowel sounds was linguistically described in Read's
The parents in Read's study provided writing materials for
their children and accepted their invented spellings. None of the
parents were experts in this area. Some of the children in the study
began to read, as well as write early, though no statistical data were
Read concluded that children who invent spellings have made
abstract inferences about a complex phonetic system. This type of
language development, to some extent, is essential for reading and
Building on the work of Read and Chomsky, Paul (1976) categor-
ized features and stages of invented spelling based on writing samples
she collected in her own kindergarten classroom. The first stage
involved writing the first letter or phoneme of each word or syllable.
In the second stage, children added ending phonemes and sometimes long
vowels. Short vowels appeared in the third stage. The children inter-
changed vowels that were formed similarly in the mouth. A variety of
vowels were used to represent each vowel sound. Children at the fourth
stage were usually reading already. Their writing included memorized
sight words and some diagraphs.
Paul also found that the thinking process involved in invented
spelling took priority over the final product. Children seldom
repeated the same invented spelling for a word. Often different
spellings for the same word appeared several times in a text. Children
frequently could not read back what they had written immediately after
their composing and almost never by the next day.
Paul concluded that children's spontaneously invented
spellings provide them with opportunities for independence in enhancing
their written communication skills. The strategies children employ
reflect their developing language abilities. Unfortunately, her
study was limited to kindergarten children in her own classroom and does
not reflect the continuity of spelling development in first and second
The stages of young children's spelling were investigated
by Forester and a first-grade teacher in Victoria, B.C., Canada, over
a three-year period (Forester, 1980). The data consisted of classroom
observations and anecdotal records. The stages Forester outlined sup-
port funding of other researchers (Bissex,-1980b; Read, 1971).
Forester noted many similarities between stages of oral language
and written language development. In particular, she noted that
children vary widely in the length of time they spend at each stage
and that movement to lower stages is common at first. "Learning is not
a linear process, but one of gradual synthesis and integration" (Forester,
1980, p. 187). From her observations, Forester concluded that spelling
development is usually characterized by the following sequence of
1. Consonants (beginning, final, median)
2. Blends (ch, sh, bl, tr, ek), morphologic markers (-ed,
-ing, -'s, etc.)
3. Words in frequent use (today, we, have, etc.)
Miscues in writing are usually indicative of growth, rather than mis-
takes, Forester cautioned. This can be exemplified by the observational
findings of Gentry (1981) and Bissex (1981)
Gentry (1981) proposed five stages in early spelling development.
Children in the deviant stage have no knowledge of letter/sound cor-
rrfspondence. Letters, and sometimes numbers, are used randomly to repre-
sent words. At this stage children are rarely able to "read back" their
compositions. Pre-phonetic spellers use the most salient sound
features in their representations of words. Pre-phonetic words usually
contain one, two, or there letters, and rarely contain vowels. The
phonetic stage is characterized by an almost exact correspondence
between letters and sounds. At this stage children have enough
command of sound/letter relationships to write fluently.
The transitional stage emerges when children become familiar
with more standard spellings and are able to apply many correct spelling
patterns. The correct spelling stage indicates a readiness for formal
spelling instruction. Gentry cautioned that time spent on creative
writing experiences should exceed spelling instructions. Gentry's
stages, based on Read's (1971) findings, explained spelling development
in a way that could be understood and applied by teachers and parents.
Bissex's (1980a, 1980b) five-year case study of her son Paul's
writing development supports many of the findings of Read (1971) and
Gentry (1981) on invented spellings. Paul's early attempts at writing
were characterized by. nonlinear, letter-like forms that communicated
a message, but did not represent individual sounds and words. Next,
he used what he had learned about letter-sound relationships from the
environment to construct messages. He used letter-name relationships
(R = ARE), predominantly consonants, to represent some words at this
stage. When the lack of spacing was pointed out to Paul, he used dots
to separate words. Messages grew in length and vowels gradually
became more prevalent.
Biss.ex described Paul as being very ego-involved with his
writing and becoming upset when she could not read it. For Paul,
writing with invented spellings preceded both reading and formal
writing instruction. As an awareness of words and correct spellings
evolved from reading, Paul made his own spelling book, complete with
correctly spelled words. At this point he became concerned with
correctness and fluctuated between phonetic spellings, word analysis,
and memorized spellings.
At approximately age six years seven months, Paul began using
punctuation, double consonants, and visually recalled the approximate
spellings of some words he had not memorized. By the end of first
grade he had moved into conventional spacing and correct representations
of short vowel sounds.
The case study approach is valuable in that it can lead to
questions and patterns to investigate in broader studies. It also
provides insight into the uniqueness of children's interests and pat-
terns of development that large-scale studies are not able to do.
Paul's development as an invented speller paralleled the stages
outlined by DeFord (1981), Read (1971), and Gentry (1981). Gentry (1981)
used Paul's spellings as an example of his five stages of developmental
spelling. At the pre-communicative stage (formerly called deviant),
Paul demonstrated some knowledge of alphabet letters, but no awareness
of letter-sound correspondence. The pre-phonetic stage (formerly called
semi-phonetic) lasted only a few weeks for Paul. The letter-name
strategy seemed to be the salient feature of this stage. During the
phonetic stage all of the important sound features of words were repre-
sented in some way. Word segmentation and spacing also appeared. At
l.his stage Paul was aware that words could be spelled phonetically in
more than one way. The transitional stage was characterized by
visual spelling strategies, based on more extensive experiences with
print. Reversals of letters within words were common at this stage.
Paul reached the correct stage of developmental spelling at age eight,
when he had established the basic rules of the English orthographic
Most of the studies described above were conducted in middle
or upper SES homes or schools. This leaves questions about their
applicability for lower SES children. Also, none of the studies com-
pared differences in writing development between boys and girls, and
factors that might influence differences.
Until this time, all of the research on early writing develop-
ment has been observational. There is a strong need for more controlled
studies that can more clearly investigate some of the conclusions of
The studies on early writing and spelling strengthen each other
in their findings. They generally conclude that there is a develop-
mental trend in children's writing, which does not necessarily follow
the same sequence for every child. Spelling generally begins with
scribbles, mock letters, or pretend writing. Next, one letter, usually
a consonant, is used to represent a word. The use of two or more
consonants, and occasionally long vowels, follows. Short vowels, and
close letter-sound correspondence characterize the next stage. As
children begin to read, their spelling becomes more conventional
(Bissex, 1981; Forester, 1980; Gentry, 1981; Paul, 1976; Read, 1971).
Fostering the Writing Process
in School and at Home
Parents usually provide a climate that allows trial-and-error
practice in learning to speak (Forester, 1980). There is also evidence
that many parents have found successful ways to foster reading and
writing development at home. It has been suggested that many positive
factors in home language climates might be applied to classroom settings
(Durkin, 1966; Forester, 1980; Harste, Burke, and Woodward, 1981; Moss
and Stansell, 1983; Vukelich and Golden, unpublished manuscript). This
section will examine ways that both parents and teachers can foster the
writing process in early childhood. Recommendations emerging from
these studies will be synthesized into two lists.
Hall, Moretz, and Staton (1976) studied the home backgrounds of
early writers "to ascertain the sequence of learning to write in rela-
tion to learning to read" (Hall, Moretz, and Staton, p. 585). The study
included ten girls and nine boys ranging in age from three years, four
months to six years, one month. Most of the parents were college
graduates, the fathers holding professional positions. They were all
identified by their teachers as early writers. Classification was
confirmed by a writing sample. Questionnaires were completed by both
parents and teachers, and both children and parents were interviewed.
Children were selected from four different nursery-kindergarten classes.
Pll the children had frequently observed parents and/or sibling.
writing and had access to writing materials. Reading was a frequent
activity in these homes.
Although the population in this study was extremely limited
in number and SES levels of students, the findings were similar to
findings of other studies of early writers and readers.
Durkin (1966) also studied the home environments of early
readers to determine common factors in their development. The popula-
tion consisted of first graders from 61 schools in Oakland, California.
Children were identified by word or sentence reading ability and
scores on Gates Reading Tests. The children were followed until sixth
grade and found to have consistently higher reading scores.
The strengths of Durkin's study were found in the number of
schools and subjects included in the investigation as well as the use of
standardized measures to assess reading achievement. It also showed
the continuing importance of a positive home reading environment for
children from similar populations.
Snow (1983) examined the oral language interactions of an
academically oriented middle-class mother and her preschool son. Like
many middle-class families, this parent/child dyad frequently engaged
in literacy-focused dialogue. In homes where activities with print
(i.e., reading, writing, typing, word games) are common, conversations
between preschoolers and parents often center around letters, numbers,
words, and books. Snow discussed several techniques this parent used
to facilitate language acquisition and literacy. These included
expanding on the child's utterances, adding new information to the
topic, clarifying questions, answering questions, taking steps to help
the child focus on the task at hand, and insisting that a task be
completed. Snow concluded that differences in reading achievement
between middle-class and working-class children may be attributed
partly to availability of literacy materials and partly to the way
parents prepare children for writing and reading though their oral
The quality, or tenor, of oral language transactions between
parents and children was observed by Hoffman and McCully (1984). The
researchers recorded conversations between parents and children when
the children drew pictures and when they wrote about their pictures.
As the children drew, their parents were warm and encouraging, demon-
strating an acceptance of the meaning in their children's art. The
tenor changed abruptly when the children began writing. Parents
became directive and controlling, instructing their children about
spelling, spacing, punctuation, and even the content of their writing.
In essence, the parents demonstrated a totally different set of
expectations for their children as artists and as writers.
These researchers concluded that several conditions in the home
environments of early writers may contribute to writing development.
1. Adults and older siblings in the home wrote and showed
an interest in the children's writing (Birnbaum, 1980;
Durkin, 1966; Haley-James, 1982; Hall, Moretz, and
Staton, 1976; Harste, Burke, and Woodward, 1981).
2. Writing materials were available to the children
(Durkin, 1966; Harste, Burke, and Woodward, 1981;
3. H wide range of printed materials was available to
the children and the children were read to regularly
(Durkin, 1966; Harste, Burke, and Woodward, 1981;
4. Children were included in family outings (e.g., cook-
ing, writing to Grandma, reading the comics) where
language development occurs spontaneously (Harste,
Burke, and Woodward, 1981).
Moss and Stansell (1983) compared the home and school writing
of a five-year-old girl. This case study revealed extreme differences
in the child's perceptions of writing in these two environments. At
school, Heather said she wrote to learn how to make letters and spell
words. At home, she reported writing for enjoyment-"Notes to people
and stuff like that" (Moss and Stansell, 1983, p. 347). At home she
was able to choose her topic and method of writing and make decisions
about language, meaning, and writing conventions.
Although the case study involved only one child, it raised
questions about children's perceptions of writing when taught by two
very different methods.
Gentry (1978) concluded from observational research (not
described) that kindergarten, first, and second grade teachers could
improve the writing and spelling instruction in their classrooms by
encouraging creative writing, de-emphasizing standardized spelling, and
responding to nonstandard spelling appropriately. In order to implement
these steps at home and at school, teachers must be aware of children's
stages of spelling development (Gentry, 1981) and be dble to communi-
cate these to parents. "Teachers should accept children's early
misspellings in the same spirit that parents accept the early mispro-
nunciations in their children's oral language" (Cramer, 1978).
Because writing and oral language are so interrelated for young
children, it is important to allow children to write together and talk
freely as they write (Childers, 1981). Dyson and Genishi (1982) and
Lamme and Childers (1983) found that writing was as much an oral as a
written process for the children they studied. Children use oral
language to ask for assistance, to "sound out" words, to reread what
they have written to each other, to answer and respond to questions, and
to share their writing with others. These case studies provide valuable
information for parents and teachers in planning environment conducive
A comparison of the writing strategies of five kindergarten
children in formal and informal language arts settings was made by Dyson
(1982). The subjects were already participants in one of Dyson's two
other studies, in which they were invited to write at a writing center.
This writing was not observed by their teacher. Parent and child inter-
view supplemented the information obtained at the writing center.
The five children Dyson described were all at different stages
in their writing development. The open-ended activities at the writing
center allowed them to use their current writing strategies and explore
new ones. The language arts curricula in the classrooms focused on
workbooks and worksheets used for practicing names, sounds, and formation
of alphabet letters. For the children who were inventing spellings on
their own and knew many letter sounds, writing strategies were com-
pletely stifled at school. For the child who was not yet aware of the
connection between talk and print, the attempt to teach letter sounds
was not very helpful.
Although the small sample size and number of classrooms
observed limited the generalizability of this study, it does demonstrate
the ineffectiveness of a "typical" pre-reading curriculum for enhancing
the writing strategies of kindergarten children.
In a study of 17 four- and five-year-olds who had no formal
writing instruction, Wiseman and Watson (1980) found that children can
learn a great deal about printed language without being taught. The
researchers collected three writing samples from each child. For the
first sample the children were asked to write everything they could
write. The second sample was a written conversation in which children
responded with writing and reading to statements read and written by an
adult. Family pictures and stories about the pictures comprised the third
sample. The children demonstrated numerous writing skills ranging from
a combination of scribbles and occasional letters to punctuated sentences
with phonetically spelled words that were close to standard spelling.
The writing samples and interviews with the children provided evidence
that young children can become aware of grapheme/phoneme relationships
and experiment with writing without formal instruction. Wiseman and
Watson concluded that perhapsas out-of-school situations in [these
children's] lives should be approximated in the classroom to encourage
a natural learning process" (1980, p. 753).
Once again a small sample size and lack of information about
the children's socioeconomic backgrounds limits the generalizability
of this study's results. No information was given on the home language
environments of these children. Two strengths of Wiseman and Watson's
study were that they collected three writing samples in different ways
and interview data was collected from the children.
Another study of positive classroom writing environments was
conducted by Florio and Clark (1982). Data were collected in two
classrooms: a second/third grade open space classroom and sixth grade
middle school classroom. The open space classroom, on which this
article was based, had 100 students from two distinct populations;
children of students and professors from a nearby university and lower
SES students (Title I). Approximately half the children came from
single parent homes. Data collection for study included ethnographic
field notes, selected video-tapes, weekly journals by teachers on
teaching, writing, teacher interviews, and collections and discussion
of children's written work.
The researchers found that most of the writing in this class-
room was not evaluated by the teachers, writing initiated by children
gave them the most control over their own writing, and the most informal
writing experiences appeared to involve children the most. This ethno-
graphic study has the strengths of a wide range of data and a large,
Lickteig (1981) reviewed research studies on writing to gather
recommendations for teachers of composition at all levels. She asserted
that the "basic skills" speak to little that is basic in education.
"[T]eacher attitude, which is audibly and visibly reflected in teacher
words and actions, is the single most important ingredient in a success-
ful composition program for children" (Lickteig, 1981, p. 45). In a
supportive atmosphere, children can afford to take risks and make
mistakes. This is true for composition, as well as other areas of self-
expression. Another basic for successful writing is real experiences
accompanied by discussion. Ideas, feelings, and attitudes are also
important to share before writing. Because reading, writing, and talking
are such integrated processes, the importance of oral and silent reading
in the classroom cannot be over-emphasized. Finally, frequent writing
experiences are necessary if growth is to occur. Children should have
opportunities to write individually, in pairs, and in small groups with
teacher support, but not constant instruction and guidance.
The value of teachers' responses to children's writing was dis-
cussed by Tway (1980). The researcher observed and worked with ten
children, ages six and one-half to eleven at the McGuffey Laboratory
School of Miami University. These children were selected by their
teachers as talented writers. Tway found that teachers' writing with
children, encouraging discussions of writing, and accepting writing in an
authentic way helped children grow as writers. Another successful tech-
nique for motivating children's writing was to invite children's authors
to visit the classroom, read their books, and interact with children.
Graves and Giacobbe (1982) investigated first grade children's
progress in writing from December to May, through a series of questions
asked before and after the children wrote. Ten of the twenty-three
children in Giacobbe's class, representing a continuum from low to
high writing performance, participated in the study. Xeroxed writing
samples were collected to demonstrate parallels between concepts of
writing and actual performance. Two children, one showing high writing
performance and one showing low writing performance, were used as case
Between December and May the lower child showed differences in
his process of choosing a subject, a clearer ability to differentiate
between drawing and writing, and a sense of continuity about writing.
During the same time span the more advanced child showed a decrease in
the number of oral words that accompanied her writing and the percentage
of written words made up from oral. The length of her writing increased.
She became more aware of the process of writing, composed over a longer
period of time, developed a sense of audience, and evaluated her own
work more extensively.
This study provided insights into ways that effective question-
ing techniques and a classroom atmosphere conducive to composition
writing can enhance certain areas of writing development. The lack of
a control group was a definite handicap in this study. It was the
only study reviewed that focused on the topic of teachers' questions.
The current research on writing development points to certain
implications for parents and early childhood teachers.
1. Introduce alphabet letters informally (Forester, 1980;
2. Allow children to choose writing topics and messages;
provide frequent opportunities for creative writing
(Forester, 1980; Gentry, 1978; Lickteig, 1980; Templeton,
1980; Vukelich and Golden, unpublished manuscript,
1984; Wiseman and Watson, 1980)
3. Encourage oral language during the composing process
(Dyson and Genishi, 1982; Gentry, 1978; Lickteig, 1981)
4. Immerse children in print (DeFord and Harste, 1982;
5. Provide language experience activities (Forester, 1980;
Vukelich and Golden, unpublished manuscript;
Wiseman and Watson, 1980)
6. Make writing materials accessible to children (Vukelich
and Golden, 1984; Wiseman and Watson, 1980)
Include markers, pens, pencils, paper, magnetic letters,
typewriter, wooden letters, and sand (Vukelich and
7. React to the meaning of children's writing, rather than
conventional spelling and punctuation (DeFord and Harste,
1982; Florio and Clark, 1982; Forester, 1980; Gentry,
1978; Hoffman and McCulley, 1984; Moss and Stansell,
1983; Tway, 1980)
8. Provide word-game activities (Moss and Stansell, 1983;
9. Base creative writing on real experiences (Lickteig,
1981; Tway, 1980)
10. Enhance risk-taking and self-expression by providing
a supportive atmosphere for young writers (Lickteig,
11. Encourage children to explore and experiment with
writing (Dyson, 1983; Lamme and Childers, 1983)
12. Provide opportunities for children to share what they
write (Childers, 1981)
13. Collect writing samples throughout the year to allow
children to observe their own progress and evaluate
their own work (Wiseman and Watson, 1980)
14. Become models of good writing by writing with students
(Forester, 1980; Newman, 1983; Tway, 1980)
Supplemented by Parent Involvement
Studies of Parent Involvement
A generation ago parent involvement was usually limited to
parents dropping their children off at school, occasional P.T.A. meetings,
and voting on school bond issues. The inception of the Elementary and
Secondary Education Act (1965), the California Early Childhood Education
Program, and the Florida Primary Education Program, among other state-
wide programs, have emphasized the role of parents in education.
The majority of parent involvement programs in the past twenty
years have focused on early intervention with lower-class families.
The programs typically involved children from birth to three years or
the preschool years from three to five. Research on these programs was
usually correlational in design and did not always control for factors
such as maternal education, family income, pre-test information, or the
voluntary status of the parents (Irvine et al., 1979). Many of these
programs involved home visitations as well as center-based parent
participation. Current research projects have experimentally studied
the impact of home interventions on children in preschool programs and
elementary schools as well. Both types of programs will be examined in
this section to provide an understanding of the importance of parent
involvement in children's cognitive development.
Bronfenbrenner (1974) evaluated data from twenty-seven pre-
school parent involvement projects, some that included home visits and
some that did not. Only those that had follow-up data for two years
following the termination of the program were included in the evalua-
tion. Bronfenbrenner's conclusions emphasized the importance of family
involvement during early childhood (Bronfenbrenner, 1974; Boger et al.,
1. Children made substantial initial IQ gains in early
intervention programs in a group setting. These gains
showed a progressive decline that seemed to be related
to home factors.
2. Programs focusing solely on home-based tutoring had
similar results to center-based, group programs.
3. IQ gains were maintained for three to four years in
programs that focused simultaneously on parent and
child. These gains also had a carryover effect for
siblings of the target child. Gains were most substan-
tial when the parent (usually the mother) felt she was
primarily responsible for the child's learning.
Bronfenbrenner credits the continuity of the family social structure with
the maintenance of IQ gains. Families were able to reinforce and con-
tinue positive teaching behaviors in the years following an intervention
In Parents as Teachers of Young Children: An Evaluative Review
of Some Contemporary Concepts and Programs, Goodson and Hess (1975)
studied the impact of twenty-nine parent involvement programs. Nine
of the programs included follow-up testing from three months to five
years after the program. Significant IQ gains were maintained for
children in seven of the nine programs over these various lengths of
time. The few programs that measured achievement gains had evidence
that children maintained these gains one to two years later. Goodson
and Hess concluded that these twenty-nine parent involvement programs
. . consistently produced significant immediate gains in children's
IQ scores, seemed to show long-term effects on children's IQ and school
performance, and seemed to positively alter the teaching behavior of
parents" (Boger et al., 1978).
Gordon's Parent Education Head Start Planned Variation Program
was implemented in four counties in different areas of the United States.
(Olmsted et al., 1980). Paraprofessionals were employed to spend half
days making home visits and half days in the classroom. On a series of
five roqnitive measures the children in this particular program performed
about as well as children in other Head Start Programs and exceeded
the control group on the Caldwell Preschool Inventory.
The Parent Education Follow Through Program (PEFTP) is currently
serving children in grades kindergarten through three in ten communities
in nine states. The population of the program includes urban, rural,
and multiethnic families. Parent educators have a dual role of working
as home visitors and classroom aides. A major focus of the program is
to include parents in six major areas of parent involvement: teacher
of own child, classroom volunteer, paraprofessional, decision maker,
learner, and audience. Home learning activities were developed by
teachers, parents, and paraprofessionals. The Abt evaluation indicated
long-term achievement gains for children involved in PEFTP. Subsequent
evaluations confirmed positive achievement results for children involved
in this project (Olmsted et al., 1980).
Teacher attitudes and experiences appear to affect methods of
parent involvement. A survey conducted by Becker and Epstein (1982)
revealed techniques used by teachers to involve parents in home learning
activities. Data were collected from 3,700 public elementary school
teachers of grades one, three, and five, in over 600 schools in Maryland.
The survey questioned teachers about 14 possible techniques they could
use to involve parents. The responding teachers (73 percent) reported
a positive view of home learning and varying use of these techniques.
"Only a minority of teachers initiate interactions with parents
that go beyond what is traditionally expected of them" (Becker and
Epstein, 1982, p. 88). For example, only 7 percent of the teachers
reported conducting three or more parent workshops or group meetings
a year. Conventional methods of parent communication, e.g., talking
with parents, open-school night, and notes home were reported by a vast
majority of teachers.
Reading to children or listening to children read aloud was the
most common technique for involving parents of younger children in
home learning. Seven out of eight first grade teachers reported using
this activity. In addition, some teachers supplemented the reading by
suggesting that parents take children to the public library or by loaning
books and teaching materials to parents. Reading activities were used
by teachers with families from all educational levels. Teachers of
younger children also reported more frequent use of classroom observa-
tions by parents to help them learn some teaching strategies. Only
one-third of the teachers surveyed believed that many or most of their
parents would attend a workshop, even if it were held in the evening.
The survey did not include the possibility of a weekend meeting.
The results of this survey provided an overall view of tech-
niques that elementary teachers in one state used to involve parents in
home learning. Although a response rate of 73 percent is very good, it
still leaves questions about methods used by the other 27 percent of
teachers. More specific techniques for involving parents were discussed
in the following article.
Epstein and Becker (1982) also examined the comments of over
1,000 teachers concerning parent involvement in home learning. These
comments were part of the survey described above. They found that
variation in comments was related to number of years of teaching
experience and number and types of contacts with parents.
The issue of time spent on parent activities seemed to be a
crucial consideration for many teachers. They questioned whether or
not the time necessary to plan and implement parent projects was worth
their effort. This is a very valid question considering the lack of
research on the effects of home learning programs in schools. In their
conclusion, Epstein and Becker, 1982, p. 111) noted:
Because of an absence of research on the effects of parent
involvement, it is impossible to assure teachers that
certain practices will lead to improved parent-child
exchanges, or improved parent-teacher relations.
Without research evidence, it would also be difficult to answer
questions teachers have about the most valuable ways for parents to
spend their limited time with their children. Some teachers felt that
parents should spend their time on socialization and development. Others
believed that a short amount of time spent each day on specified skills
was beneficial. In contrast, one teacher remarked parents might become
frustrated and impatient if expected to work on skills at home.
In commenting on children's time and feelings about home learning,
some teachers expressed the belief that academics should be kept to a
minimum and more time spent on play and individual interests. Others
added that parents should instill values, responsibility, and home-
related skills. One reason given for not asking parents to tutor their
children was the added stress it might cause in the family if a parent
took on this role. The authors also questioned whether the benefits of
parent involvement were the same for older and younger children, parents
of all educational levels, and married and single parents.
Based on statistical analysis of the survey and additional
teacher comments, Epstein and Becker identified areas where research is
needed. These included investigating the most educationally significant
type of parent involvement, advantages and disadvantages of various
programs, and skills parents would need. The attitudes of parents,
teachers, administrators, and students toward parent involvement warrants
further research, as does the role of the teacher in organizing parent
activities (Epstein and Becker, 1982).
The following two experimental studies demonstrate that parental
assistance with reading appears to have a positive influence on
children's standardized reading scores.
The impact of a parent involvement reading project in an inner
city school was the focus of a study by Shuck, Ulsh, and Platt (1983).
Their population included 150 third and fifth grade remedial reading
students randomly assigned to experimental and control groups. All
students were given a Slosson Intelligence Test and were pre- and
posttested using the Stanford Achievement Test, reading comprehension
subtest. Both groups also attended the reading laboratory daily for
30 minutes. Students in the experimental group did individualized
activities at home with their parents. Records were kept via calendars
that parents used to record the types of activities and time spent on
them. A behavior modification system, using points and prizes, rein-
forced homework and class participation. Parent-teacher conferences
were held three times during the year, though not all parents were able
The data were analyzed by analysis of covariance, using per-
centage of school attendance and reading achievement scores as co-
variates. There was a significant main effect for parental tutoring.
The experimental group had a mean posttest grade equivalent score of
3.8, compared to the control group's mean grade equivalent of 2.6.
Limitations of this study included the possibility of a teacher
variable, and an inability to generalize from an urban population. Also,
the use of a behavior modification system could have confounded the
results. The authors indicated that future research is needed to deter-
mine whether or not home programs are effective for different age levels,
in curriculum areas other than reading, and whether or not parent
tutorial programs have lasting effects.
A research study from Great Britain (Tizard, Schofield, and
Hewiston, 1982) showed that significant gains in reading achievement
were obtained by children whose parents listened to them read. The
research project was conducted in six schools in a working-class section
of London. Two schools were randomly selected for the experimental
parent involvement group, two were selected for an in-school reading
intervention, aid the remaining two schools served as a control group.
One class was chosen from the top infant class (six- to seven-year-olds)
in the two experimental schools to have the intervention. The other
classes formed within a school control group. Parents were first
informed of the project by letter, second by a school meeting, and
finally by a home visit by the researcher. The project continued for
two full school years. During the first year, children took home their
school readers. Second-year students read supplementary books and
library books in addition to school readers.
All children were tested with three reading tests at the
beginning, middle, and end of the intervention period. Mean scores on
all three instruments showed a highly significant improvement for
children who received the parent intervention, but no comparable improve-
ment for children receiving the in-school reading instruction or the
control group. This study was extremely well planned and well designed.
The length of the study (two years) and the home visits by the researcher
added to its strength. Limitations would include possible teacher effect
because of no in-class control group, and the possibility of between-
Several recent studies have exemplified methods that early
childhood teachers can use to involve parents in their language arts
programs. Teachers at Wayne State University Nursery School in Michigan
used journals as a home activity for parents and children (Elliott,
Norwosad, and Samuels, 1981). Parents were introduced to purposes and
procedures regarding the journals at a Saturday morning workshop.
Mimeographed pages provided specific assignments related to school
topics that parents'could focus on when writing with their children,
e.g., toys at home, friends at home. The topics appeared at the top of
the page and space remained for dictated sentences and copying by those
children who were ready. The journals were sent home on Friday and
expected to be returned on Monday. Journal topics were discussed at
school and home writing was read and shared at the book center.
The teachers believed that this was one way parents could be helped
to ". . understand that family conversation, family outings, looking
at books, being read to, and seeing adults read are as important as
learning the alphabet" (Elliott et al., 1981, p. 691).
Burgess (1982) found that parent participation in a readiness
program can have a positive effect on overall readiness and language
readiness scores on standardized tests. A sample of 90 families was
chosen from a total population of 200 children entering kindergarten in
a small town in Maine. Sixty of the parents volunteered to participate
in the parent training; 30 prior to their children's testing, and 30
after the children's testing. The control group of 30 parents did not
volunteer to participate.
The researcher and an assistant conducted eight, two-hour parent
workshops during February and March preceding the children's entrance
in school. The workshops consisted of sharing and evaluation, presen-
tation of new information, and the preparation of games and activities.
The following eight topics were covered in the workshops: reading to
children; stimulating oral language; color, size, shapes, textures,
and classification; more abstract sequencing, comparison, and classi-
fication; body awareness; listening skills; number recognition, one-
to-one correspondence; and writing stories-taking dictation and modeling
correct letter formation.
The two posttest measures were the Utah Test of Language
Development and four subtests of the Metropolitan Readiness Test, Level 1.
The experimental group scored significantly higher than either of the
control groups on both measures.
The use of volunteers was well controlled in this study by
having an extra group who volunteered, but did not receive the treat-
ment until after testing. However, future researchers might consider
ways of motivating parents who did not volunteer. This study was con-
ducted with a limited rural population. Similar studies should be
attempted with more divergent groups. Other early intervention pro-
grams have been shown to have long-term gains when parents have been
highly involved (Bronfenbrenner, 1974; Olmsted et al., 1980). Follow-
up studies similar to the one described above might support these
findings. Also no mention was made of whether or not any of these
children had attended preschools. This should be considered as a factor
in future research.
It is evident from the research studies reviewed in this section
that various types of parent involvement have positively affected
children's cognitive development. Limitations in design, population,
and curriculum areas indicate a need for closely controlled experiments
that have a wide generalizability. With one exception (Elliott et al.,
1981), parent participation in the language arts area has focused on
reading. Early childhood teachers, in particular, often encourage
parents to read to their children but may overlook other equally
important areas, such as writing.
Future research should explore the strengths and weaknesses
of different types of parent involvement and methods of motivating
parents and children in various situations (Epstein and Becker, 1982).
Methods of training parents and communicating with them during the
period they are working with their children should also be investigated
(Epstein and Becker, 1982). While several of these studies looked at
differences in grade level, socioeconomic level, and educational level
of parents, none examined the possibility of a parent involvement pro-
gram having a differential effect for boys and girls. The possibility
of sex differences in home learning needs to be explored by future
Sex Differences in Language Arts
Reading and Oral Language
Any sex differences that exist in the area of language arts
could be attributed to a number of factors: heredity, brain specializa-
tion, cultural attitudes toward sex roles, prenatal development,
creativity, achievement, motivation, and ability. Because of a lack
of research on sex differences in writing, this review will predomi-
nantly examine research on sex differences in attitudes toward reading,
language achievement, and verbal ability. Though other areas of
research are possibly relevant, it would be difficult to draw any
definite conclusions from them in relation to this study.
The superior reading ability and achievement of young girls has
been widely accepted in the field of education. A study by Downing et al.
(1979) and his associates sought to discover the impact of cultural
expectations and sex role standards on attitudes toward reading using
a cross-national population. The population sample included adults
and children in grades one, four, eight, and twelve in the following
countries: Denmark, England, Finland, Israel, Japan, and the United
States. One hundred subjects were randomly selected from each age
group in each country. Interviewers were trained to administer ques-
tions from two test booklets. One booklet showed objects, which were
described as presents. The subjects were told to circle either the
picture of the girl or boy, depending on whom they thought should receive
the presents. An activity book, marked the same way, showed neuter
stick figures engaged in various actions. Adult subjects were inter-
viewed individually in their homes, whereas school-age subjects were
tested in total classroom groups.
Females of all levels, in all seven countries, gave a majority
of "girl" responses, indicating that they believed reading was a
feminine activity. Younger boys in all countries started out with a
majority of "boy" responses. "Girl" responses increased with age,
particularly in the United States, England, Canada, and Israel. The
results of this study indicate that cultural expectations and sex-role
standards about reading differ between countries. It appeared inmany
countries that viewing reading as a feminine activity increases with age.
Downing and his associates reported that the stick figures may
not have been psychologically neutral in all cases. Also, in Denmark,
where children are rarely tested, it may have been difficult to obtain
accurate responses. The use of intact classrooms may have produced
some teacher bias in the results.
Using methods similar to those of Downing et al. (1979), May
and Ollilla (1981) had very different findings for younger children.
The researchers tested 136 children from day-care centers and public
school kindergartens, ranging in age from three-and-one-half to five-
and-one-half years. The data indicated that preschoolers of both sexes
assess reading objects and actions to be more appropriate for boys. In
addition, one girl out of sixteen reported that her father read to her,
but eight boys out of eighteen said they were read to by their fathers.
Research on differences in achievement and ability of males
and females was recently reviewed by Levine and Ornstein (1983).
Several studies they cited indicated that girls have higher reading
achievement than boys in the early 1970s, but that the gap had narrowed
by 1980. Other studies supported the finding that sex differences are
diminishing in most areas of achievement and attainment (Levine and
Ornstein, 1983, p. 66).
Plomin and Foch (1981) reanalyzed data from several large
studies that found sex differences in verbal ability in favor of girls.
They found that the average differences between sexes were far greater
than differences within the groups. Gender was found to account for
only about 1 percent of the differences in verbal ability and samples
of at least 1,000 were needed to find this difference.
In a lengthy and comprehensive review of sex differences in
many fields, Maccoby and Jaklin (1974) reported that few differences
appeared between sexes on ability tests over most age ranges. However,
in 18 studies of children under the age of seven, girls had slightly
higher scores in eight studies, boys in only one study. These findings
may be attributed to cultural differences, since the studies that
found higher scores for girls were predominantly done among "disad-
vantaged" groups. The type of items on the test may also have had an
effect if the test was not well balanced for verbal and non-verbal
Maccoby and Jaklin (1974) also reported that research up until
the time of the publication of their book gave some support to the
generalization of females having higher verbal ability than males.
These differences were almost always non-significant, and in many popu-
lations did not exist at all. Among preschool children, only one study
in the thirteen reviewed by the authors showed girls to be superior to
boys in verbal abilities. From the age of 10 or 11 on, the research
showed a stronger trend for girls to have higher verbal ability. For
children under seven, no sex differences were found on verbal tests of
creative ability. After age seven, girls usually demonstrated superior
performance to boys on these tests.
Graves (1975) found sex differences when he studied the writing
of 94 second graders from a middle-class community. The length of
girls' writing exceeded boys' writing. However, boys wrote more on
their own than girls. Differences were found on topics chosen, style
of writing, spacing, letter formation, and neatness.
The large number of writing samples examined and the four phases
of data collection added to the thoroughness of this study. Additional
research is n(.drdd to determine whether or not these differences extend
to other aqv groups and populations.
Bodkin (1978) found significant differences in the writing of
boys and girls at third and sixth grade levels. One hundred twenty
children were randomly selected from a population of 301 children.
The subjects were part of four intact classrooms. All children were
given writing journals and instructed to write daily for ten days on
any subject they chose. The analysis revealed that girls wrote more
in primary territory-about family, home, school, and interpersonal
relationships. Boys wrote more in extended territory-about sports,
metropolitan, world, and catastrophic events. Girls also wrote more
often and at greater length, regardless of their socioeconomic level.
The differences between boys' and girls' writing topics appear
to be related to cultural, sex-role attitudes more than ability. Girls'
greater interest in writing could possibly be attributed to their moti-
vation to please a teacher or parents. Further research is needed
before reasons can be determined for these differences.
Even though most intelligence tests have been standardized to
minimize sex biases, some research on sex differences still indicates
that females are higher in verbal ability from preschool through
adulthood (Deno, 1982). In critiquing studies of sex differences in
cognition, Deno pointed out that many weaknesses existed in the research.
First, she noted that sex differences in verbal ability generally
account for only 1 percent to 5 percent of the population variance.
Second, most of the research in this area has been done on white,
middle-class Americans, often of high ability. Third, tests need to be
developed that will measure more specific abilities to aid in the
understanding of cognitive differences between males and females.
Fourth, the higher variability of male scores over female may be
responsible for some measurement error that has been overlooked by
researchers. Fifth, there was a lack of longitudinal studies.
In addition, most of the research has been done with high school
and college students, some with elementary, but very little with pre-
school and kindergarten children. Because of developmental differences
it would be dangerous to generalize between age groups. Even if
differences in verbal ability exist, it would be difficult to say what
aspects of language arts would be affected by these differences.
Evidence of sex differences in intelligence, verbal ability,
and language achievement of young children is weak. Most researchers
in this field would agree that a tendency exists for girls to have
higher verbal ability than boys and that this tendency increases with
age. Cultural attitudes toward sex roles in reading seem to be a
salient factor influencing these differences, possibly in combination
with other factors (Downing et al., 1979; Levine and Ornstein, 1983).
Young children who have been widely exposed to print, who have
seen adults writing, and who have had many drawing experiences, often
start writing with no formal instruction (Gardner, 1980; Kane, 1982).
Oral language seems to play an important role in the drawing/writing
development of successful writers (Bissex, 1980a; Childers, 1981;
Dyson and Genishi, 1982; Graves, 1980). Talking during the writing
process helps children as they learn about sound-symbol relationships
reread compositions, ask questions, answer/respond, and share or tell
about their writing (Calkins, 1983; Graves, 1980; Lamme and Childers,
1983). Thus, writing is often a social experience for young children
(Dyson and Genishi, 1982; Graves, 1981; Lamme and Childers, 1983).
Some research implies that writing should, and often does,
precede reading (Chomsky, 1981; Graves, 1982; Hall, Moretz, and Staton,
1976; Hildreth, 1936). Composing words by sound-symbol relationship
may be a less complex task than decoding unfamiliar words (Chomsky,
1971). Also, more young children believe they can write than believe
they can read (Graves, 1983). Children who read early almost always
have a simultaneous interest in writing and spelling (Durkin, 1966).
Children are able to distinguish their writing from their
drawing before they are able to produce letters or words recognizable
to adults (Clay, 1980; Vukelich and Golden, 1984). After the early
stages of scribbling and mock letters, writing development does not
necessarily follow an orderly sequence (Clay, 1980; DeFord, 1980;
Newman, 1983). Children may move back and forth over a "range" of
writing stages, depending on the topic of writing and their motivation.
When children are encouraged to write independently, their
spelling evolves through a predictable sequence of stages. Invented
spelling usually begins with the child representing words with the
initial letter of each word or syllable (Bissex, 1980a;Forester, 1980;
Gentry, 1981; Paul, 1976; Read, 1971). Ending consonants and
occasionally long vowels follow in the second stage. Short vowels
are used, though not always accurately, in the third stage. A
transition to conventional spelling is characterized by the wide use of
short vowels, blends and diagraphs, and memorized sight words. The
thinking process involved in invented spelling is a higher importance
than the resulting product and plays an important part in reading
and writing development (Paul, 1976; Read, 1981).
Research on early writers indicates that characteristics in
the home environments of these children might be successfully adapted
in school settings. Early writers were exposed to printed materials
and read to regularly, owned a variety of writing materials,
saw adults and siblings writing, and were included in family outings
(Birnbaum, 1980; Durkin, 1966; Haley-James, 1982; Hall, Moretz, and
Staton, 1976; Harste, Burke, and Woodward, 1981). In addition to
these factors teachers should encourage student selection of writing
oral language during writing, exploration, and experimentation with
writing (Dyson, 1981; Forester, 1980; Gentry, 1978; Lamme and Childers,
1983; Lickteig, 1981; Shane, 1980; Vukelich and Golden, 1984; Wiseman
and Watson, 1982).
There is a tendency for girls to perform better than boys on
tests of general intelligence, verbal ability, and language achievement,
and for this tendency to become stronger as age increases. In addition,
cultural attitudes toward sex roles appear to influence attitudes
toward reading (Downing et al., 1979; Levine and Ornstein, 1983).
Although most of the research on parent involvement in
language arts has involved reading activities rather than writing,
the methods and findings are relevant to this study. Holsinger (1979)
found that primary age children improved significantly on one measure of
reading achievement when their parents helped them at home using
individually prescribed activities. A second finding of Holsinger's
study was that parents who participated in the home program had sig-
nificantly more contacts with the school than parents who did not
participate. Shuck, Ulsh, and Platt (1983) found that a combination
of home reading activities, parent record-keeping on calendars, home
visits, and classroom reinforcement resulted in significant differences
between experimental and control groups in an inner-city school.
Research by Chomsky (1971), Graves (1982), Hall, Moretz, and
Staton (1976), and Lamme and Childers (1983) shows that pre-first grade
children can show a keen interest in writing if writing materials are
available, parents are responsive, and there is some familiarity with
letter names. Several research studies (Gentry, 1978; Lickteig, 1981;
Wiseman and Watson, 1980) that positive language/literary environments,
at home or at school, can enhance children's growth as writers.
Writing development also appears to be fostered when children
share the processes and products of writing with others (Calkins,
1983; Dyson and Genishi, 1982; Graves, 1983). The process of sharing
seems to help not only the writer, but those who listened and responded
as well (Calkins, 1983).
Writing fluency, the number of words a child can write at any
given sitting, can logically be tied with children's experiences in
writing. It has been suggested that writing words is a good indicator
of knowledge about print and writing (Clay, 1980; Hildreth, 1964).
A large body of research (Chomsky, 1971; Clay, 1980; DeFord,
1981; Durkin, 1966; Graves, 1983; Gundlach, 1981; Hildreth, 1964) shows
that early informal writing experiences can have a decided impact on
reading readiness. Clay's Concepts about Print test (1975) is a valid
and reliable measure of reading readiness. The twenty-four item test
includes items designed to measure children's knowledge about books,
letters, words, and punctuation marks (see Appendix A) all indicators
of reading readiness.
Several researchers (Clay, 1980; DeFord, 1980; Vukelich and
Golden, unpublished manuscript) have noted that, although children do
not move through a definite sequence in their writing development,
there is a hierarchical nature to writing development. These studies
provided the framework for the development of a valid and reliable
scale (Lamme/Green scale) to measure children's writing achievement
at the kindergarten level.
The research cited above points out the value of parent involve-
ment in language arts. The presentstudy is designed to measure the
effects of parent involvement, using a specified curriculum, on child-
ren's writing achievement, writing fluency, and concepts about print.
The methodology of the study will be presented in Chapter III.
THE DESIGN OF THE STUDY
This study was undertaken to examine the impact of a home
composing curriculum on kindergarten children's concepts about print,
writing fluency, and writing achievement. The lack of research on
parent involvement in the composing/writing process and writing research
pointed to a need for quantitative research on young children's writing.
The sample was drawn from four kindergarten classes: two at
a university laboratory school and two at a public elementary school.
The two classes at the laboratory school were representative of the
population of the state in terms of race, sex, and SES. The public
school drew from two distinctly different populations. Seventy percent
of the students, predominantly white, were from middle SES families.
Thirty percent of the students, predominantly black, came from lower SES
Seventy-four children participated in the study: 38 female and
34 male. These children were drawn from a total population of 90
children. Eight children were not included because they could not be
matched with another child of the same gender on the pretest. Two
other children did not participate because of parent request. Six
children moved during the study or were absent when postest data were
A four-group, randomized-matched-subjects, pretest-posttest
design was used. One covariate was used to control for initial
differences in children's concepts about print.
All children in the four classrooms were administered the
Concepts about Print test during the first month of school. Three
writing samples were gathered from each child during the first four
weeks of school. These were analyzed using a revision of the Vukelich/
Golden scale (Vukelich and Golden, unpublished manuscript). One pre-
test sample of writing fluency was also gathered from each child during
the initial four weeks of school. These variables were later dropped as
covariates because of inconsistent data collection techniques and their
low correlations with posttest data.
Data from the Concepts about Print test were used to rank-order
the boys and girls in each of the four classes. The top-ranked two
girls formed a match pair (block), and so on, until as many girls as
possible had beem matched. The same procedure was followed for boys in
each class. From the matched pairs children were randomly assigned to
either the experimental or control group.
The two kindergarten classes at the public school are described
t 'o other because the classes shared an open-space area and combined
.Ludents for an hour of lanquagq arts instruction each day. The forty-
five children in the two classes were hoioqc nrou.rly grouped into four
ability groups. The lowest level, a readiness group, worked primarily
on listening skills, following directions, print awareness, writing
letters in the air, and name writing. The teachers used the Ginn 720
series, Level 1, Module 1, with this group. The middle two groups
started with Ginn 720, Level 1, Module 3, where letter recognition and
beginning sounds were emphasized. The high group began with Ginn 720,
Level 2. This group was introduced to two D'Nealian letter forms each
day. Throughout the Ginn series there was an emphasis on literature and
print awareness which these teachers stressed in their small group
During total group instruction the teachers wrote experience
stories on chart paper from the children's dictation. They exposed the
children to a wide variety of literature supplemented with flannelboard
stories, puppets, and films. In addition, parent volunteers wrote dic-
tated stories for individual children and children were given key
vocabulary words (words that were particularly meaningful for each child).
The two kindergarten teachers at the university laboratory
school used different approaches to language arts instruction. In one
classroom the Alpha-Time series formed a basis for introducing letter
names and sounds during the first two months of school. Language
experience was also part of the language arts program in this classroom.
Small and large groups of children dictated stories on topics such as
trips, visitors to the classroom, special plays, etc. These charts were
used to introduce basic sight words, punctuation, and comprehension.
Dictated sentences were also written about individual children's pictures.
Writing was sometimes copied from the chalkboard or a chart. During
the learning center time, the teacher divided the class into three
groups. One group played in the courtyard under the supervision of
the aide, a second group worked independently on manipulative activi-
ties, and a third group had direct instruction in math or reading.
These groups rotated during the learning center time.
The other kindergarten class at the laboratory school was taught
by the researcher. This classroom was organized into nine basic learn-
ing centers: math, art, dramatic play, books and listening, manipulative
games, drawing and writing, unit blocks, large hollow blocks, and sand
play. The latter two centers are accommodated in an outside courtyard.
Writing was frequently modeled by the teacher, university
students, and volunteers in the classroom. Total group experience
stories, thank-you notes, and lists were frequently dictated by the
children. The teacher also wrote words to songs and chants on chart
paper for the class to read together. Stories were read to the children,
with or without props, from one to four times daily. The writing and
drawing center was one of four centers all children were required to
attend for fifteen minutes each morning. The center was supplied with
markers, pencils, crayons, and various sizes snd shapes of paper. The
children typically wrote or drew on a topic related to the week's theme
or wrote notes back and forth to each other or the teacher. In
addition, some children also chose to come to the writing and drawing
center during the afternoon activity times.
Beginning November 1, approximately halfway through the inter-
vention period of this study, the laboratory school began implementing
the John Henry Martin Writing to Read program for all kindergarten and
first grade children. During the two weeks of the program children in
the two classes were introduced to all the computers and tape-recorded
lessons on stories. The children were grouped and paired by ability
during this introductory period. Each progressed through the program
at an individual pace and regrouping frequently occurred. Each class
used the Writing to Read lab for one hour daily. The children were
typically divided into four groups and rotated centers. During the time
they were in the lab each day the children used the computer to learn
about the phonetic sounds in one word for a ten to fifteen minute
period. After this exposure, they listened to audio-tapes for rein-
forcement for these sounds and wrote the word they had learned on the
computer, as well as other words containing these sounds. During the
next fifteen minutes the children listened to story tapes. At the last
center they typed letters, words, sentences, or stories. Both the
classroom teacher and a Writing to Read teacher were present in the lab
It is the opinion of the researcher that the Writing to Read
program had very little impact on the children in the study. For two
weeks, the children attended the lab for short periods to be introduced
to equipment and procedures. The average child was introduced to only
three words (nine phonemes) by the end of the study.
Concepts about Print
The Concepts about Print test (Appendix A) was designed to
measure young children's book awareness and understanding of print. The
purpose of this instrument is to provide the teacher with an indication
of each child's level of readiness for various reading experiences. It
is appropriate for use with non-readers because it involves pointing
and brief verbal explanations. The test is administered individually and
the 24 items are dichotomously scored.
Two tables are provided in the manual for converting raw scores
to stanines. The first table, based on an urban population of 320
children ages five years, zero months, to seven years, zero months, were
used in this study.
The reliability of Concepts about Print, calculated on 40 urban
children aged five years, zero months to seven years, zero months, in
1968, was 0.95 (Kuder-Richardson). Test-retest reliability coef-
ficients were reported to range from 0.73 to 0.89 for 56 Texas kinder-
garten children in 1978. Corrected split-half coefficients ranged
from 0.84 to 0.88 for the same population (Clay 1972). The correlation
between Concepts about Print and Clay's Word Reading test was 0.79,
based on a population of 100 children at six years, zero months in 1966.
Concepts about Print was chosen for this study because the
researcher believed thdt writing experiences and interaction with
paretns during writing would enhance children's understanding of many of
the roncupts coveredd on this test. This instrument was used by the
county public schools and the laboratory school kindergartens as part
of their initial screening.
Lamme/Green Composing Scale
Three writing samples were collected from each child in each
of the four classes during August and September. The four teachers
who participated in the study asked small groups of children in their
classes to "Write anything you want to write." From this pilot data,
the authors of the scale developed a set of 22 descriptors. These
descriptors were based on studies conducted by Vukelich and Golden
(unpublished manuscript) and DeFord (1980). For the present study,
the descriptors ranged from scribbles to stories or letters with five
or more sentences. Based on these descriptors, the authors grouped
writing samples that appeared to have similar characteristics. When
these groups of samples were arranged hierarchically, thirteen cate-
gories emerged. Three raters used the thirteen category scale to
evaluate the pilot test writing samples. They found that several cate-
gories either overlapped or contained very few samples. As a result,
the categories were condensed into six hierarchical groupings which
adequately described the data. The Lamme/Green Composing Scale
(Appendix B) is a description of these categories.
Three raters (two graduate students and one professor) used
the Lamme/Green Composing Scale to rate the three posttest samples of
writing achievement collected in December. High interrater correla-
tions were found for the three raters (r = 0.94, 0.90, 0.91). Validity
of the scale is based on correlations with the Concepts about Print
test (r = 0.78), writing fluency (r = 0.75), and total percentiles
for the Metropolitan Readings Test (r = 0.57).
Parent Response Sheets
On Friday of each week during the study, parents were asked
to complete a parent response sheet (Appendix D) and return it to
their children's teachers. The response sheet provided space for
recording the writing activities the child did each day and persons
with whom the child wrote. Additional space allowed parents to comment
on other writing activities of their children.
A workshop introduced the parent in the experimental group
to the concept of writing at home with their children, the importance
of writing, and the interrelationship between writing and reading.
(The workshop outline appears in Appendix E.) The workshop at the
laboratory was conducted by the researcher's chairperson and the
workshop at the public school was conducted by the researcher.
The laboratory workshop was held on a weekday evening and the
public school workshop was held on a Saturday morning. Both sessions
lasted approximately one hour. Fifteen parents attended the workshop
at the laboratory school and five parents attended at the public school.
Those parents who were unable to attend, but still wanted to participate
in the study, met with their children's teachers on an individual
The workshops focused on the importance of informal and spon-
taneous writing activities parents could use with their kindergarteners.
Both workshop leaders emphasized that writing should be pleasurable
activity for both parents and children. Parents were told that writing
need not occur at a specific time each day, but to try to write for and
with their children several times each week. The workshop leaders
showed samples of children's writing and explained typical categories
in writing development. Parents were encouraged to accept invented
spellings as a natural part of young children's writing.
At the end of the workshop a booklet was given to the parents,
presenting ideas for home writing (Appendix C). Three ways parents can
help young children with writing were presented and discussed. Samples
of various writing activities were shown and parents shared other
writing ideas they had tried.
Each parent received ten response sheets to indicate which
activities were used with a child each day and to identify who did the
activity with the child. Space was provided for comments and the
listing of additional activities. Parents were encouraged to send
their children's writing to school where it was shared. Biweekly notes
(Appendix F) were sent home to parents encouraging continued participa-
tion and suggesting additional home writing ideas.
The study began the second week in October and ended the first
week in December.
Data were collected both before the study in August and Septem-
ber, and after the study in December and January. Posttest writing
samples were collected in all four classrooms by an experienced univer-
sity student who was taking an early childhood language art class at the
time. Consistent directions were given to children in all four classes.
Samples were collected in groups of eight to nine children during the
second and third week of December. The time children stayed at the
writing tables ranged from ten minutes to one hour, with an average
time of approximately thirty minutes.
All children were given black felt-tip pens and white 8-1/2x 11"
unlined paper. They were encouraged to write for as long as they
wanted. The groups were taken outside or to the cafeteria when they
wrote. All samples were scored using the Lamme/Green composing scale
During orientation in August and during the first few weeks of
school all children were individually administered the Concepts about
Print test. The teachers administered the test at the public school.
At the laboratory school the Title I teacher and aide administered the
test. In December and January the Concepts about Print test was
administered by the teachers at the public school and by the Title I
dide and a research assistant at the laboratory school.
Levels of writing fluency were established by asking the
children, in large groups, to "Write all the words you know how to
write." If children hesitated, they were asked, "Can you write your
name? Can you write any other words?" One sample was collected from
each child during the first three weeks of school and again during the
second week of December. Three raters counted both the number of
words spelled correctly and the number of words which could be iden-
tified (including misspellings). These ratings were labeled fluency A
and fluency B, respectively. Fluency B was dropped during analysis
because of its high correlation with fluency A (r = 0.93672).
The same test of fluency is currently used as part of the
kindergarten screening for the public schools in the county where the
study was conducted. In this fluency test, only the number of words
spelled correctly is counted.
The first step in the analysis was to determine correlations
between pretest and posttest data. It was found that the Concepts
about Print test given in September had the highest correlations with
a posttest data. Therefore, only the Concepts about Print test was used
as a covariate.
The independent variables in this study were gender, class, and
treatment (the home composing curriculum). The dependent variables
included Concepts about Print, three measures of writing level (rated
using the Lamme/Green scale), and writing fluency.
The significance of observed mean differences between treat-
ment, classes, and gender groups and the effect of group interactions
was tested with a general linear model procedure, described in more
detail in Chapter IV. An overall alpha level of 0.10 was used for
testing effects on the three compositional writing measures, so that
for the compositional writing sample at each of the three occasions,
an alpha level of 0.033 was used. An alpha level of 0.05 was specified
for testing effects on Concepts about Print and writing fluency.
In Chapter IV the results of these analyses are presented and
answers to each of the research questions is discussed.
The purpose of this study was to determine the impact of a
nome composing curriculum on kindergarten children's concept about
print, writing fluency, and writing achievement. Children from four
classrooms were rank-ordered and matched, based on pretest results
from Clay's (1980) Concepts about Print test. Approximately half of
the children in each of the four classrooms participated in the home
composing curriculum (treatment). The treatment involved parents writ-
ing at home with their children, following guidelines described in a
parent workshop (Appendix E) and writing composing booklet and the
sharing of home writing in the classroom. The control group partici-
pated in outside play with a classroom aide during the sharing of home
writing products, then shared science materials while the treatment
group played outside. Following the intervention period, all children
in the four classes were administered Clay's Concepts about Print
test (1975). Three writing achievement samples and one writing fluency
sample were also collected from each child.
To control for initial differences in the student population
data were analyzed by analysis of covariance, using Concepts about
Print as the covariate. This procedure was conducted for each of the
five dependent variables: Concepts about Print, writing fluency, and
the three writing achievement posttests.
The first General Linear Model procedure tested for interac-
tions between the covariate and treatment, using the following model:
Y! = aO + alCAP + a2GEN + a3CL + a4TRT
+ a5(GEN)(TRT) + a6(TRT)(CL) + a7(GEN)(CL)
+ a8(GEN)(TRT)(CL) + ag(CAP)(TRT) + ERROR
Y! = i dependent variable and i ranges from 1 to 5
for the five separate variables: posttest 1, posttest
2, posttest 3, Concepts about Print (posttest), and
CAP = Concepts about Print pretestt)
GEN = gender
CL = class
TRT = treatment
a0 = the intercept and al-ag are the regression coefficients
When no covariate by treatment interaction was found, the follow-
ing reduced model was used:
Y! = a0 + al(CAP) + a2(GEN) + a3(CL) + a4(TRT)
+ a5(GEN)(TRT) + a6(TRT)(CL) + a7(GEN)(CL)
+ a8(GEN)(TRT)(CL) + ERROR
Table 1 summarizes the pretest means and standard deviations
for the three dependent variables: gender, treatment, and class. As
can be seen from the table, girls began at a slightly higher level
than boys. The experimental group was slightly higher than the control
and classes 1 and 2 were significantly (a < 0.01) higher than 3 + 4.
Class 1 also had much wider variance than the rest of the classes.
Table 2 summarizes the correlation coefficients of all pretest
and posttest measures. The Concepts about Print pretest was signifi-
cantly correlated with all five posttest measures. Therefore, it was
considered to be a good choice for the covariate. Each of the writing
achievement posttests was highly correlated to the other two, indicating
that they measured the same construct.
Treatment Effects on Concepts about Print
The first question to be answered from the statistical analysis
was: Were the observed differences in mean levels of the Concepts
about Print test, statistically significant (at the alpha level < 0.05)
for children in the experimental and control groups? Results of the
analysis of covariance (ANCOVA) reported in Table 3 indicate that there
was a significant difference between the two treatment groups on the
Concepts about Print test. There were also significant differences
among the four classes.
Least squares means were used in this analysis to include the
effect of the covariate. Table 4 summarizes the least squares means on
the Concepts about Print test. The treatment section shows that the
mean of the treatment group was higher than the mean of the non-treatment
Table 1. Concepts about Print pretestt) means and standard deviations
4-) 00 C0 IO Cj -,a- "-
- iO- C- 00 (0) in ir
co 4-) Io >n r>_) IL_ IC o
C- ",- t t i i
00 n0 00 O 0
C 1 O0 C:) r-_ Cn C C:)
S0 r-I 0
-O CA C) CO ID r- i o C)
rc o D o- on
4-) 4-) ( C C C
S, 0 0 0 ,
4 4-C' CD al O LO C
0VC C h 0 L
. C0 0 0 00 n 0 0
, O O0
-r- C ) C)
S4- C 4
-3 ( i 4- i - i n c
o C A C ) C) V) ) 0 (C
) a- 0 0 0 o _
0 0 04 (
4-' = >
0 C) C ) C) 0 u 1
- 0 0-- i 0 n 3 -
E) - - ) C -
0 L 0 C C C)
( .- .,- -
4-' 0) 0 m 0) 4-) :K
CA- CA CA CA d- --
a)0 r- ) 0 4c C
1..- a) 4/ 4-) ( CA 4-. C
U 0- 0 0 0 4- E3
4- 4- 2 4 49 4-' 4
Ss- a) a) a) o. u
.. 0- E E E 4 (U
E > 4-1 C)
E : 4 a) a) a) C 0 (
CI *. *= a) C) a) r
U) Sf C C C S U *-
ra U (U () -) O V CC 4
S 4- ) >r > r >U U 4-) 0 4
-J (-Q -= 4- 4-. 4 0U a)
Table 3. Summary of ANCOVA for Concepts
Source DF SS F P
Concepts about Print 1 414.08 53.61 0.0001
Gender 1 17.69 2.29 0.1357
Class 3 69.02 2.98 0.0384*
Treatment 1 35.03 4.54 0.0375*
Gender x Treatment 1 4.90 0.63 0.4291
Treatment x Class 3 45.60 1.97 0.1275
Gender x Class 3 11.54 0.50 0.6891
Gender x Treatment x Class 3 16.31 0.70 0.5569
*p < 0.05.
Table 4. Summary of least squares means for Concepts about Print
Concepts about Print Least Squares Means
group. Children who wrote at home and shared their writing at school
had higher mean scores on this measure of reading readiness. Follow-
up analysis using the Scheff6 formula showed that class 2 was signifi-
cantly lower than class 1 and class 4 was significantly lower than
classes 1 and 3. The mean score for females was slightly higher than
the mean score for males.
Treatment Effects on Writing Fluency
The second question to be answered from the statistical analysis
was: Were the observed differences in mean levels of writing fluency
statistically significant for children (at an alpha level < 0.05) for
children in the experimental and control groups?
The data in Table 5 indicate that there was a significant
difference between the two treatment groups on the writing fluency
Table 6 summarizes the least squares means for writing fluency.
The treatment section shows that the mean of the treatment group was
higher than the mean of the non-treatment group. Classes 2 and 3 were
lower than classes 1 and 4. The mean score for females was slightly
higher than the mean score for males.
Treatment Effects on
Compositiona I Writi_ Achievement
The third question to be answered from the statistical analysis
was: Were the observed differences in mean levels of writing achieve-
ment statistically significant (at an alpha level 4 0.05) for children
Table 5. Summary of ANCOVA for Writing Fluency
Source DF SS F P
Concepts about Print 1 329.4428 16.97 0.0001
Gender 1 23.3577 1.20 0.2775
Class 3 96.4730 1.66 0.1857
Treatment 1 107.0356 5.51 0.0225*
Gender x Treatment 1 45.8289 2.36 0.1302
Treatment x Class 3 12.6176 0.22 0.8835
Gender x Class 3 139.4192 2.39 0.0771
Gender x Treatment x Class 3 42.9609 0.74 0.5373
*p < 0.05.
Summary of least squares means for Writing Fluency
Writing Fluency Least Squares Means