WWW/HIWWWI L

3 1293 01004 9157

OVERDUE FINES ARE 25¢ PER DAY
PER ITEM

Return to book drop-to remove
this checkout from your record.

 

 

 

A STUDY OF THE RELATIONSHIP
BETWEEN TELEVISION EXPOSURE AND LANGUAGE
ACQUISITION OF PRE-SCHOOL CHILDREN

BY

Gary W. Selnow

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

College of Communication Arts and Sciences

Mass Media

1978

‘4 r
J x.

j)“ I i/_\I fl

ABSTRACT

A STUDY OF THE RELATIONSHIP
BETWEEN TELEVISION EXPOSURE AND LANGUAGE
ACQUISITION OF PRE-SCHOOL CHILDREN

BY

Gary W. Selnow

This research focused on the relationship between
television viewing and language development of pre-school
children. It also considered relationships between language
development and maternal factors, then language development
and sibling variables.

In the first series of analyses, child language ability
was correlated with total television viewing time and then
with five subcategories of program types. The next series
considered language ability in terms of a composite exposure-
language level index which represented program exposure
weighted by language SOphistication modeled by television
programs. Language ability, total television viewing and
viewing of programs by category were then considered in terms
of the mother's education level. Finally, there was an
analysis of two elements of Zajonc's Confluence Model which
predicted relationships between (1) sibling number and
language ability and (2) birth order and language ability.

Language samples were collected according to clinical
procedures outlined by the Developmental Sentence Scoring
Analysis for 93 children ages three years and five months

to five years and eleven months. In addition to language

samples, data on television viewing patterns of subjects
were collected from logbooks maintained for a one week
period by parents. Demographic information was also
obtained. Scripts of frequently viewed programs were then
analyzed according to Developmental Sentence Scoring
procedures.

The analysis revealed a relationship between television
viewing and language development for this sample. Viewing
time across all program types was negatively related with
language ability and this negative relationship was strong—
est for cartoons and family drama programs. Using the
exposure-language level index, it was discovered that child—
ren who viewed a greater number of hours of language-
sophisticated shows tended to score higher on the language
assessment instrument. Those who viewed a greater number
of language-poor shows tended to score lower on language
measures.

Mother's education level was positively related to the
child's language ability, as sociolinguistic theory predicts.
Mother’s education level was also positively related to the
language sophistication of programs viewed by children.
Partialling out mother's education reduces, but does not
eliminate, the correlation between language ability and
television viewing.

Finally, language ability was substituted for standard
intelligence measures used in previous research on the
Confluence Model, and the analysis supported two key proposi—

tions: as the number of children in a family increased,

the child's language ability tended to decline. Similarly,
language ability scores tended to decline with birth order

of a child.

ACKNOWLEDGMENTS

The work involved in this dissertation could not be
completed by one person in the three month period that it
took to arrange for subjects and collect the data. Apart
from a time investment, the collection of language samples
from 93 pre-schoolers required great skill and patience and
I am indebted to my wife, Francie, for her work in this
project as an interviewer and assistant, and for her encour-
agement throughout the course of this project.

Four men, my mentors and committee members, contributed
not only to the production of this piece of research, but to
my graduate studies and other research activities during the
recent few years. Committee Chairman, Dr. Tom Muth, teacher,
barrister, spiritual advisor and occasionally drinking part-
ner not only provided encouragement for these scholarly
pursuits, but had the unique talent of breathing life into
some potentially moribund academic subjects with which
graduate students must deal. In his role as advisor and
confidant he displayed an understanding of human nature
with an insight befitting Merlin. Several years ago Dean
Erwin Bettinghaus provided the inspiration for my first look
at the influence of television on language development and
it was largely because of his expertise in this area that

I returned to MSU to study under this great man; the value

ii

of his encouragement and guidance cannot be overstated.
Professor Bill Crano, one of the truly gifted men in social
psychology research left me spellbound in a course six years
ago where I first explored the relationships between
communications principles and elements of human thought.
This man is a scholar who inspires scholarship in his
students; his input to my personal development has been
substantial. More than ten years ago, at Rutgers Univer-
sity, I took an undergraduate journalism course with Dr.
Maurice Cullen, and largely because of his enthusiasm for
understanding the processes of human communication, I
abandoned a previous major and turned to a life of study in
this field. Professor Cullen, paradoxically, projects the
spirit which drove traditional American journalists while
he no less ranks among the contemporary scholars of
communications. His guidance has been invaluable; his
example has been inspirational.

I am grateful to each of these people for their

participation in my studies and my life.

iii

TABLE OF CONTENTS

LIST OF TABLES . . . . . . . . . .

LIST OF APPENDICES . . . . . . . .

Chapter

I

II

RATIONALE AND LITERATURE REVIEW .

Introduction . . . . . . . . .
Scope of the Relevant Issues .
Reasons for the Present Study .
Theories of Language Acquisition
Methodology Problems . . . .
Nativist Positions . . . . .
Biolinguistic Position . .
Linguistic Position . . . .
Environmentalist Positions .
Sociolinguistic Position . .

Television -- Its Power to Influence

and Instruct . . . . . . . .
Learning and Television . . .

Television and Language Learning

Maternal Influences . . . . .

Sibling Influence on Intellectual

Development . . . . . . . .
Summary . . . . . . . . . . .
Focus of the Present Study .

Statement of Hypotheses . . . .

DESIGN AND METHODS . . . . . . . .

Overview of the Study . . . . .
Terminology . . . . . . . . . .
The Child Sample . . . . . . .
Television Programs . . . . . .
Instrumentation . . . . . . . .
Data Collection . . . . . . . .
Data Preparation . . . . . . .
Special Indices . . . . . . .
Language Ability Index . .

Exposure-Language Level Index

iv

vi
vii

Page

downpour-4H

ll
16

20
23
28
35

37
39
40
42

48

48
49
50
53
54
59
61
62
62
63

Table of Contents (cont'd.)

Chapter
Operationalization of Variables
and Specification of Variable
Relationships . . . . . . . . . . .
Data Analysis . . . . . . . . . . . .
Summary . . . . . . . . . . . . . . .
III ANALYSIS OF RESULTS . . . . . . . . .
Introduction . . . . . . . . . . . .
Total Television . . . . . . . . . .
Television Viewing Time by Category
and Child Language . . . . . . . .
Language Ability and Exposure -
Language Level Index . . . . .
Relationships with Mother's Education
Level . . . . . . . . . . . . . .
Number of Siblings and Birth Order
Variables . . . . . . . . . . . . .
Summary . . . . . . . . . . . . . . .
IV DISCUSSION . . . . . . . . . . . . . . .
Structure of the Investigation . . .
General Review of Results . . . . . .
Limitations . . . . . . . . . . . . .
Television Viewing and Child Language
Mother's Education, Television and
Child Language . . . . . . . . .
Number of Siblings and Birth Order
Research Extension . . . . . . . . .
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . .
APPENDIX I
Details of the Developmental Sentence
Scoring Analysis . . . . . . . . .

APPENDIX II
Correspondence and Forms for Parents
of Subjects . . . . . . . . . . . .

Page

64
65
65
67

67
67

70
74
76

83
85

91
91
92
95
98
102
106
110

114

123

128

Table

10

LIST OF TABLES

Children's Ages .

Number of Children in Family

Position of Subject among Siblings

Mother's Education

Father's Education

Parents with whom Child is Living .

Correlations of Language Ability and
Five Categories of Television Programs

Correlations of Mother's Education and
Five Categories of Television Programs

Partial Correlations of Television Viewing
Variables and Language Ability Control-
ing for Mother's Education

Summary of Results

vi

Page
56
56
57
57
58

58

71

79

82

86

LIST OF APPENDICES

Appendix Page
I Details of the Developmental Sentence
Scoring Analysis . . . . . . . . . . . 123
II Correspondence and Forms for Parents

of Subjects . . . . . . . . . . . . . 128

Vii

CHAPTER I

RATIONALE AND LITERATURE REVIEW

Introduction

 

The principal purpose of this research is to explore

the relationship between television viewing and child lan-

guage development. A number of related variables are drawn

into this investigation and contribute to the following

objectives:

1.

To examine the relationship between (a) language
ability and television viewing overall, then (b)
to explore the connection between language ability
and specific subcategories of programs.

To consider the relationship between (a) mother's
education and child language development, and (b)
mother's education with television viewing pat-
terns of children.

To investigate the relationship between (a) lan-
guage ability and sibling numbers, and (b)

language ability and birth order.

Scope of the Relevant Issues

 

To the extent that most language theorists will grant

environmental influences on language development processes,

there is general agreement that parental and especially ma-

ternal language displays provide the single most dominant

input to a child's language learning process. Such an

1

assertion is axiomatic to most environmentalist theories
which argue that constant daily contact between the parent
and child assure such a direct influence; more succinctly,
a child speaks as his parents speak.

For almost three decades with the presence of tele-
vision, there has been a notable change in daily activities
and interaction patterns of children in the United States
and other developed countries. With the proliferation of
television, refinement of targeted programming and social
reinforcement of the medium, television has begun to share
a greater proportion of the young child's time, to an
extent that recent surveys suggest children spend more time
watching television than they do interacting with parents
(Winn, 1977). Given that this is the case, an intriguing
question begs attention: If, as some theories hold, the
child speaks as his parents speak because he spends more
time with them, will it follow that greater exposure to
television will lead to the development of language
heavily influenced by television?

There are a sufficient number of distinctions between
parent-child and television-child language encounters to
invalidate any simple argument based upon exposure alone.
For one, parents can provide immediate feedback, tele—
vision cannot; parents can direct their messages to a
single child, television, being a mass medium, does not
have this Capacity. But television does offer messages

pitched with a keen appeal to young viewers which engages

3
them in the content, and television can embark on a variety
of language trips potentially exceeding the capability
of a single set of parents. Beyond all this, television
is always there, always available to a child with the touch
of a button when parents may be preoccupied with other
activities; television is for the young child, "language

on call."

Reasons for the Present Study

 

The principal objective of this research is to examine
the relationship between the language of young children and
the nature of their exposure to television programs. The
literature suggests that language development can be in-
fluenced by external factors and that television has the
capacity to influence its viewers in a variety of ways. It
appears reasonable to propose the conjunctive position
which predicts a relationship between a child's language
development and his television viewing behavior. The
present study probes this relationship. A variety of other
factors which influence a child's language development have
been recorded in the literature, and two of the most promi-
nent -- mother's influence and sibling influence -- are

explored in this research.

This chapter provides a summary review of literature
from several disciplines which have been brought together in
the present investigation. The following categories have

been given particular attention: (a) theories of language

acquisition, with an emphasis on environmental influence
and intervention, (b) issues related to the instructional
and influential potentials of television, (c) considera-
tion for maternal and sibling influence on child language

development.

Theories of Language Acquisition

 

Methodology Problems

 

Language acquisition theories span a wide range of
philosophical positions reminiscent on the one hand of
empiricism and on the other of rationalism. These two
viewpoints, respectively characterized by the environmen-
talist and nativist positions on language acquisition, split
on the fundamental issue of whether language is generated
within the individual as a creative psychological process,
or whether it is acquired by the child through lessons
provided by other language users. Disciples of both posi-
tions offer supporting evidence, but disputes arise at
every phase of data collection and interpretation.

Data collection during the very early and, thus, per-
haps the most important years of language acquisition, is a
particularly limited activity given that the only source of
data is the child himself who may or may not produce a lan-
guage sample upon request or provocation. Most researchers
believe the basic process of grammatical speech development

is completed by three-and—a-half years of age (McNeill, 1966)

and so research prior to this age must be accomplished if
there is to be understanding of these earliest developmental
stages. Difficulties arise, however, in that children
typically do not speak in lucid phrases, but in enigmatic
fragments and single-word utterances, and it is precisely
here, in the subjective interpretation of this behavior
that language theorists begin to draw sides. Even as
Slobin outlines four levels of evidence for discovering
rules in child and adult grammars, problems inherent to
the subjective nature of analysis become obvious:

l. Analyze speech for regularities.

2. Search for the extension of regularities to new

instances.
3. Look for a normative sense of rules.
4. Analyze responses to direct questions.

(Slobin, 1971:10)

Language during the first few years is restricted to
single word utterances, and because of the child's ego-
centric orientation during this period (Piaget, 1955),
even these generally lack communicative purpose (Piaget,
1955; Vygotsky, 1962). Without an intent to communicate,
the utterances may appear to be random, so the discovery
of regularities becomes a frustrating and often fruitless
task. If regularities cannot be established, then it
follows that an extension cannot be expected nor can

rules be considered a relevant issue. Finally, Slobin

considers the matter of a child's response to direct
questions, a task which requires competence on three
levels: first to interpret the question, then to generate
an answer, and finally to express it verbally. A defi-
ciency at any level would lead to the production of an
uninterpretable response, hence, such attempts can result
in futility:

Interviewer: Adam, which is right, "two shoes" or

"two shoe?"
Adam: Pop goes the weasel (Brown and Bellugi,

1964:135).

So, it is important to establish at the outset that
methodologies available to examine the earliest stages of
language formation are imprecise and thus invite the cre-
ativity of subjective interpretations.

It appears that the tentative data base for early
language production contributes substantially to the
diversity of positions held for language acquisition.
This point becomes obvious in reviewing the dominant

theories.

Nativist Positions

 

The essential tenet of deterministic nativist theories
postulates that the development of language originates from
within an individual. It is an innate process which remains
relatively unaffected by external forces and so can neither

be substantially hastened nor retarded by environmental

intervention. Two broad subcategories of the nativist
position include (1) the biolinguistic theories which View
language development as a biological process and disallow
nearly all environmental impact, and (2) the linguistic
theories which assume a more modified position and provide
for some environmental intervention on language-learning
processes. A review of these nativist positions is impor-
tant to this study which seeks to determine if television,
as an environmental factor, has any effects on children's
language development. Generally, these theories would not
predict a substantial influence by television on language

development processes.

Biolinguistic Position. According to biolinguistic

 

propositions, language development closely parallels
biological development of a child, and as physiological
growth continues so the development of language proceeds
step-by-step. Similar reasoning leads to the biolinguist's
explanation for the species-specific nature of human lan—
guage where biological deficiencies in all animal forms
below man render them incapable of speech (Osser, 1975).
From birth, however, man develops special coordination
centers for motor speech, perception of specialized tem-
poral patterns, cerebral growth, tolerance for prolonged
speech activities and other biological traits which equip
him as a language-capable being. Language development is
a function of these maturational factors and language

acquisition can be explained by the "synchrony between the

attainment of each language milestone and the development

of particular motor skills." (Osser, 1975:290)

This position, also known as the "Critical Age Model"

has been summarized by Golub:

1. Language learning is determined by
physiological and psychological development
which in turn is determined by chronological
age.

2. An individual must pass through each
preceding critical age to get to the next
in a particular linear order.

3. Language cannot be taught, it can only
be learned in its prescribed biological and
psychological sequence.

4. Passage through critical ages requires
the placement in social environments which
stimulate, energize and provide the condi-
tions for an individual to pass from one
critical age to the next (Golub, 1975:1099).

Lenneberg and followers offer three defenses for the

biolinguistic position, citing (l) cross-cultural similari-

ties, which they interpret as support for the pan-human

nature of this theory, (2) clinical research data comparing

deaf

and hearing infants during prelinguistic developmental

stages, and (3) analogue studies relating the progress of

normal children to retarded development of abnormal child-

ren.

lead

First, Lenneberg's studies of cross-cultural data
to the following generalizations:

--Language is a form of behavior present in
all cultures of the world.

--The age of onset is the same for all child-
ren in all world cultures.

9

--The acquisition strategy is the same in all
cultures.

--The formal operating characteristics have

remained the same in recorded history
(Lenneberg, 1969:645).

With these observations, Lenneberg proposes to demon-
strate that language development is not affected by en-
vironmental factors, but is exclusively a function of
biological progressions which exist uniformly among all
humans.

As a second defense for the biolinguistic position,
Lenneberg places heavy emphasis upon his research of
prelinguistic sounds by comparing deaf and hearing infants
(Lenneberg, Rebelsky, and Nichols, 1965). In this series
of studies his arguments are based on environmental
differences between hearing infants who are exposed to the
sounds of language from birth, and deaf infants whose
environment is devoid of language models. Findings of
these studies confirm that deaf infants proceeded through
the same sequence of vocalizations as hearing infants, and
they also produced as much noise (Osser, 1975). It would
seem reasonable to conclude from these observations that
prelinguistic utterances are based on maturational factors
and not environmental stimulation.

Finally, Lenneberg submits studies with mongoloid
children who display the usual stages of development
experienced by normal children, only at a greatly reduced

rate of progression. In these studies, he found the same

10
synchronous patterns of language achievement and motor
development. Since language kept pace with biological
development even in these pathological instances, Lenneberg
holds this as one of the principal demonstrations of proof
for the biolinguistic theory.

It is important to note, however, that in each exhibit
of defense for this theory, Lenneberg limits his observa-
tions to the prelinguistic phase of language development,
that is, during the first 12 months when a child's vocali-
zations are generally limited to cooing and babbling, but
have not yet entered the linguistic stage of actual lan-
guage production. In light of Piaget's notions concerning
the non-communicative function of egocentric speech, or in
this case merely of sounds and babbles, it appears diffi-
cult to extend beyond this formative period into more
advanced stages of language.

When Lenneberg does attempt to extrapolate his data
beyond the conditions under which they were collected, he
faces some difficult problems. While he claims that the
same maturational processes are at play when the child
progresses into word production stages, he relies on more
tenuous data and cumbersome explanations to account for
obvious discrepancies between findings and theory. His
principal defense rests on research by Morley (1957) who
discovered, in an extensive longitudinal language study,
that irrespective of subjects' environmental backgrounds

and linguistic milieus, the onset of language was invariant.

11

However, Morley was unable to substantiate the nature of
parent-child interactions in her sample, a deficiency which
has become the major flaw in this research. Contrary to
Morley's findings, other researchers have documented
differences among varied environmental conditions in
actual language production (McCarthy,l954; Nelson, 1973).
It is important to note that there is little to
recommend biolinguistic theory as valuable in guiding
therapy or designing remedial programs for the language
deficient child. The deterministic orientation disallows
environmental intervention as a productive influence on
the child, and, except for recommending adequate nutri-

tional care, can prescribe little else.

Linguistic Position. As a subset of nativist posi-

 

tions on language learning, linguistic theories (subsuming
psycholinguistic theories for purposes of this review),
too, propose that language is an innate capacity of the
child. They differ, however, from biolinguistic theories
on the fundamental explanation of how language integrates
into growth and development processes of an individual.
Rather than synchronous gearing between biological and lan-
guage maturation processes, linguistic models explain
language acquisition as a function of innate language con-
cepts available to the child which integrate incoming
language information into meaningful sets of rules. The

concept of a sentence is the nucleus around which all

12

environmental language data forms; this existing structure
enables the child to organize and interpret language input
(McNeill, 1970). So, this theory proposes that linguistic
principles are not a learned phenomenon, but are inherent,
conceptual abilities available for language learning tasks
(Chomsky, 1965).

Brown describes the focus of rules in language acqui-
sition:

The emphasis is, to put it simply, that in acquiring

a first language, one cannot possibly be said to ac-

quire a repertoire of sentences, however large that

repertoire may be, but must instead be said to ac-

quire a rule system that makes it possible to gener-

ate a literally infinite variety of sentences, most

of them never heard from anyone else (Brown, 1973:102).

While the role of environmental factors holds only
limited importance in linguistic theories, exposure to lan-
guage models of fluent speakers enables the child to sort
through his existing preknowledge of language universals,
which are rules and constituents underlying all languages,
and discover the rules appropriate for his own language
(Chomsky, 1965). So, by birthright, the child possesses a
notion for correct grammars and the ability to evaluate and
cull out appropriate grammatical elements for the language
to which he is exposed (Osser, 1975). Hence, language en-
vironments serve as models and so are passive in the
language development process. The child plays an active
role since he listens and compares, then eliminates false

rules and adopts appropriate elements for the language he

will use.

13

The admission of environmental influences to this
nativist position is an important trait of linguistic
theory and serves well in the explanation of many observed
language phenomenon. For instance, Osser, Wang and Zaid
(1969) reviewed language samples and environmental data of
black lower-class and white middle-class pre-schoolers and
discovered that the two groups displayed different syntac—
tic structures of standard English. Linguistic theory
accounts for this observation by proposing that the pre-
sence of different environmental language data would lead
the child to choose a different set of rules appropriate to
his own language.

Several elements of the nativist theories are impor-
tant for purposes of this review. First, they offer to
explain an underlying structure which allows human language
learning to take place in a remarkably constant manner for
all individuals irrespective of greatly varying environ-
mental factors. With only minor differences, people
speaking the hundreds of existing languages appear to
progress through similar stages of development as do
individuals within language groups who demonstrate little
variance in their progression through these developmental
stages. Second, nativist theories provide an explanation
for individual creativity of language where novel sentences,
never precisely modeled, can be generated by speakers.
Nativist rule-learning explanations seem to account well

for this occurrence where environmentalists' reinforcement

14

schedules meet with problems.

An obvious shortcoming of nativist theories, however,
is recognized in failure to provide for adequate environ-
mental intervention in the remedy of speech deficiencies.
As a deterministic philosophy it offers little encourage-
ment for remedial programs and therapy for the disadvan-

taged.

Environmentalist Positions

 

Environmentalist positions remain distinct from na-
tivist viewpoints on fundamental philosophical issues,
holding that learning is primarily a function of environ-
mental interactions, rather than the application of an
innate set of rules. Here the environment becomes an ac-
tive agent and imposes on the individual who remains
passive in the learning process.

Principles of learning theory are most dominant among
environmentalist positions and are particularly relevant
to the present study. Advanced chiefly by Skinner (1957)
and Mowrer (1960), learning theory attends to the more
encompassing issue of learning in general, and has been
applied extensively in clinical therapy situations. But,
in dealing with language learning, similar principles of
reinforcement and generalization apply.

In most general terms, elements of reinforcement pro-
vide that when an individual receives a positive reward

(something pleasant) for a behavior, he will tend to repeat

15
that behavior. Conversely, if he receives a negative
reward (something unpleasant) he will tend not to repeat
that behavior.

The basic tenet of this position, applied to language
acquisition, has been summarily described by Mowrer:

The infant begins to learn language by

associating the sounds of the human voice,

particularly his mother's, with need-satis-

fying circumstances (e.g. milk drinking).

The result of this is that when he hears his

own random babbling, he is more likely to

repeat those sounds that are similar to the

pleasurable sounds made by his mother. Thus

the pleasure associated with the mother's

voice now becomes transferred to the child's

own vocalizations (Mowrer, 1960; summarized

by Osser, 1975:299).

Language develoPment continues following the rein-
forcement schedule of rewards by the mother for sounds
which approximate adult language models. These rewards
are pleasurable and so to elicit continued positive rein-
forcement, the child proceeds to voice adult-like speech;
he talks like mother to be rewarded by mother. This
continues throughout the language development process.
Therefore, language is modeled, then through reinforcement
schedules, the child builds his own language repertoire.

Learning theory specifies just how environmental
factors influence the language development process, and
while other researchers support environmentalist positions,
they do not become involved in the dynamics of this pro-

cess. For instance, a variety of recent investigations

have uncovered a strong relationship between parent and

16
child language patterns on several dimensions without
adopting a learning theory explanation (Cazden, 1972;
Medinnus and Bohannon, 1975; Fox, 1976). Fox noted that
the issue is one of environmental intervention and in-
fluence on language development, and her studies clearly
indicate that "the adult enriches the child's linguistic
environment by extending the child's ideas and introducing
different grammatical elements, words, meanings and idea
relationships." (Fox, 1976:668) Similarly, Cazden (1972)
concluded from his studies that adult interaction does
improve language in children.

The environmentalist position has important implica-
tions for this research which investigates the influence
of television on children's language. If, as the theory
holds, the environment displays an impact on language, then
there is room for television, as part of that environment,

to contribute to language development.

'Sociolinguistic Position

The sociolinguistic position represents a subset of
the environmentalist category, and remains neutral with
respect to the controversy between nativists and behav-
iorists over the origin of language behavior (Osser,
1975). and deals with derivatives of language in terms of
social structural factors (Hess and Shipman, 1965;
Bernstein, 1966; Hymes, 1967; Labov, 1970).

Bernstein, whose name is most associated with

l7

contemporary sociolinguistic theories, distinguishes two
major codes as elaborated and restricted, stating that
they generally can be discriminated by the ease with which
linguistic alternatives can be predicted. The restricted
code system selects from a smaller set of syntactical and
lexical options than the elaborated system, and is conse-
quently more predictable in nature. It is generally less
specific and informative in that the repertoire of speech
structures are limited or considerably constrained,
restricting the individual's ability to make fine dis-
cernments. Furthermore, it impedes a continuous learning
process by discouraging verbal inquiries through frequent
role-directed displays of social power. By contrast, the
elaborated code system selects from a more extensive range
of syntactic and lexical alternatives available in the
language, and enables users to differentiate a greater
spectrum of objects and events in the environment. This
suggests that although elaborated and restricted code
users may experience identical events, these events become
more linguistically differentiated for the elaborated code
user, and effectively create more experiences for him.
Acquisition of these linguistic codes takes place apart
from the individual's intellectual capabilities and depends
almost entirely on his family's social class orientation.

Hess and Shipman (1965) support Bernstein's position
with research suggesting that lower-class mothers are not

as explicit as middle-class mothers, and convey only

18
limited information in messages to their children. Their
data was collected on four groups of mothers differing
in education level, various language variables and problem—
solving approaches. They found that better educated
mothers used more complex syntax, longer sentences, more
qualifying modifiers and more total verbal output. Studies
involving similar variables reviewed in a number of other
research contexts tended to confirm these results (Bee, et.
a1. 1969; Brophy, 1970; Steward and Steward, 1973).

Other research lends support to Bernstein's notions of
social influence on language including work by Petersen
(1970) who demonstrated the dominant maternal influence on
language development and Lawton (1968) who found a greater
peer influence.

There is also evidence, however, which suggests that
language style is situationally determined and the child
can modify his use of language appropriately to meet
demands imposed by given conditions. So, he does not
possess abilities for only a single language style, but
can alter his styles to meet demands of a given situation
(Lawton, 1968; Williams and Naremore, 1969). In these
studies, when the child was encouraged to provide elabora-
tion in answer to a question, lower-class children, too,
could offer more complex descriptions. Williams and
Naremore did find in their study, however, that lower—class
children provided answers from a limited, first-person

perspective (self-focused mode of discourse), but that

l9
middle-class children generally described events from the
third person which allowed them to assume a variety of per-
spectives in their communications.

Pozner and Saltz (1974) found that both middle and
lower—class children would learn rules of communication
with equal skill, but middle-class subjects were much
more effective as communicators. These researchers
suggest that distinctions may not necessarily be found in
an ability to communicate, but rather in realizing the
need for expression; they further submit that such diffi-
culty might be viewed as an egocentric orientation in the
lower SES child.

Tulkin (1972) strongly objects to conventional stan-
dards for language assessment which, he claims, largely
employ middle-class orientations, so that many differences
which may have been recorded reveal little valid informa-
tion about the speaker. When observed in a naturally
occurring group interaction, and not in the constrained
laboratory environment, Tulkin argues that language.
deficiences are not apparent (Chandler and Ericson, 1968).
Birren and Hess offer a powerful conclusion for their
research.

Studies of peer groups in spontaneous inter-

actions in Northern ghetto areas show that there

is a rich verbal culture in constant use. Negro

children in the vernacular culture cannot be con-

sidered 'verbally deprived' if one considers them

in a favorable environment - on the contrary their

daily life is a pattern of continual verbal stimula-

tion, contest and imitation (Birren and Hess, 1968:
42).

20

Throughout all of the sociolinguistic literature there
appears to be no contention with the proposition that lan-
guage differences exist among members of different social
classes, and that environmental differences inherent to
these classes account, at least indirectly, for observed
language use differences.

This issue of language quality, as measured in a
variety of terms centering around the capacity to accomo-
date abstractions or complex conceptual notions, or even
minimally to provide an accurate description of physical
occurrences is, it seems, the nuclear issue for all lan-
guage research, at least for environmentalists who believe
the course of learning can be modified. If language used
by any class of individuals is deficient and provides its
users with an impoverished system for communicating ideas,
then the issue of language quality assumes a renewed sig-

nificance surpassing class sensitivities.

Television — - Its Power to Influence_and Instruct

 

According to recent surveys, young children typically
watch television for an average of five hours per day
(Nielson, 1977). This substantial exposure rate has
prompted many investigations of the impact of television
on viewers who stand to learn from the medium, be aroused
by the medium or in other ways be influenced by the sights
and sounds of television.

When television was introduced into the American mass

21

communication system, it assumed the information function
of traditional media, but also displayed a fidelity,
immediacy and social significance unique to itself. In-
dustry accurately envisioned flourishing financial oppor-
tunities in television, audiences were immediately attract-
ed to its format of sound and picture, and social watchdogs
stirred restlessly in the presence of this innovation.

Penetration statistics of television reflect the
enthusiasm with which American people accepted television.
In 1949, each of 4 percent of the homes had a set, and
only five years later television was in over 60 percent of
American homes. By the late sixties, television reached its
saturation level with 97 percent penetration (Defleur, 1970).

Largely because television grew so quickly there was
no available information on which to predict the breadth
and depth of the effects this medium would have, and so
immediate responses tended to be extreme. "Attitudes were
on the one hand fraught with great hope for realization of
the potential of this new medium for education and enter-
tainment." (Christenson, 1969) In review of early studies
on television, Coffin (1955) noted that the medium was
found, in some research, to increase family opportunity
for common experiences and shared interests, but at the
same time interfered with family conversation and face-to-
face interaction. While some proponents of the medium

offered television as a format to present new information

22
and thus broaden children's experiences, critics denounced
the disruptive influence on family life, home study and
further contended the medium would yield harmful effects
because of violence displayed so routinely.

Even though debate over these same issues continues
today the spirit of such urgency appears to have abated and
research has assumed a more dispassionate and analytical
tenor (Christensen, 1969). Even so, information about
the effects of television and its power to influence
viewers is no less critical than before. On the contrary,
as television viewing occupies a greater proportion of
daily schedules, the matter assumes a renewed significance.
Recent Nielson survey findings (Spring, 1977) confirm the
persistence of a trend toward greater television use ge-
nerally among all viewers. More significantly, pre-school
children are reported in this survey to use more television
than previously reported for this group and more than any
other age category as well.

Schramm (1961) noted in 1960 that young children
watched an average of 2 hours per day compared to the two-
fold increase of 4.5 to 5 hours recorded in the 1977
Nielson survey. While a direct relationship between the
extent of television exposure and potential to influence a
viewer may not be firmly established, Siegal (1969) and
Winn (1977) argue that if only because alternative activi-
ties become displaced by viewing, television has a greater

opportunity to influence an individual, particularly a

23

child during his formative years.

The issue at present concerns the power of television
to instruct and influence its audience. As this dis-
cussion turns to the possible effects of television view-
ing on language development, it should be noted that the
relationships considered between these two variables have
implications for environmentalist theories. Recall that
environmentalist positions provide a major role for exter-
nal elements in the development of child language. If it
can be demonstrated that television viewing may effect the
patterns of language growth, such evidence would underscore
the validity of these theories.

Learning and Television

 

Nielson statistics demonstrate the extraordinary tele-
vision exposure time for young children, a fact which in-
vites speculation about the potential for learning to take
place during these viewing hours. Can television serve as
a format of learning for children? Data in hand suggests
that television viewing can provide a learning experience.

In a study involving elementary school children,
DeFleur and DeFleur (1967) assessed the amount and quality
of learning about unfamiliar occupational roles when
information was observed both directly and in televised
film sequences. In general they found that television
portrayals of various occupations imparted information
which was learned by the children as successfully as

through direct face-to-face exposure. They suggested,

24

however, that television typically presented a stereo-
typical profile of occupational roles and this distorted
view was learned and assessed as accurate by subjects.

In a similar study involving television as a source
of information about role relationships, Holloran (1967)
noted that television has the capacity to reach beyond the
direct experiential repertoire of children and present
them with role models and reference groups which would
otherwise not be available. Children integrate this
information into their general growth and learning of

socialization processes.

A number of studies have been conducted on the
effectiveness of Sesame Street as an educator of pre-
school children (Ball and Bogatz, 1970; Frye, 1972; Lesser,
1974). In this series learning has been identified as a
principal goal of programming. Although some researchers
are skeptical of the show's proclaimed success (Cook, et.
a1, 1975) the weight of evidence suggests that Sesame
Street has captured young viewers' attention and facilita-
ted learning of vocabulary, numbers, the alphabet and other
fundamental academic skills. Lesser (1974) concludes from
his review that the program has reached large ethnically
and economically diverse audiences and achieved its in-
structional goals. Children, he says, can and will learn
cognitive material from television, at least when it is

presented in a format which is attractive to these

25

audiences.

A second dimension of television learning, reviewed by
Krugman and Hartley (1970), concerns viewing pacificity and
psychological receptivity to incoming information. These
authors provide an explanation of the neurophysiological
operatives in an individual's relaxed state noting that
frequencies of both Alpha and Beta waves slow during re-
laxation, and while individual differences exist, recepti-
vity to learning in this condition increases for most
people. Furthermore, specific to environmental influences
on an individual's relaxed state, these authors note the
amenability of environmental stimulation which plays a
principal role. The logical conclusion, then, suggests
that when environmental conditions (in this case television
shows) are conducive to a relaxed state, then passive
learning is likely to take place. Implications of this
are described here:

The most special quality of passive learning

is, by definition, an absence of aroused re-

sistance to what is learned; resistance is

exciting and a corollary, therefore, of active

learning. This means that passively learned

material has an important 'advantage' which some

have also associated with so called subliminal

perception, extrasensory perception, or hypno-

tism. This advantage, however, is not a property

of the stimulus but of the respondent; i.e. he

can learn passively so long as the material is

acceptable to him, without conflict (Krugman

and Hartley, 1970:188).

Roberts and Schramm (1971) note a similar phenomenon

which they call "incidental learning" where children were

26
discovered to have learned elements of a television program
which were both peripheral to the story line, and totally
irrelevant to the child's daily repertoire of experiences.
For instance, pre-adolescent children have recalled infor-
mation pertaining to college life, dating activities and
behavior appropriate to various occupational skills.

One subject area where incidental learning may be
particularly relevant involves social learning and
acculturation where an individual is transfused with social
mores and expectations for his behavior in a social con-
text. Bandura (1962) concluded from some of his early
studies that television made available to the individual
visual and symbolic models which he could assess and from
which he could learn. Moreover, Bandura (1969) found
additional evidence from his research during the 1960's
to suggest that television's increasing pervasiveness
both in society and in lives of individual viewers was
leading to this medium's usurpation of parental and teach—
er influence as models for behavior; influence of television
in social learning processes has thus become stronger.

In her workvdJJIthe effects of television violence,
Siegal (1969) takes a particularly pessimistic view of
television as a source of social learning, and while she
underscores the impact of the medium, she concludes that it

produces negative outcomes.

27

The evidence that we do have indicates that

films and television are profoundly educative

for their viewers, teaching them that the

world is a violent and untrustworthy place,

and demonstrating for them a variety of

violent techniques for coping with this

hostile environment. Whether this message

is deemed as fact or fiction, it is accepted

by young children. They incorporate in their

own behavior patterns all the sequences of

adult behavior they observe on television

(Siegal, 1969:282).

Siegal's observations on the conduction of negative
social factors may, indeed, be accurate, but if the strength
of television as a source of social learning can be so
demonstrated, someone must dare ask the complimentary
question about television's potential to perpetuate the
virtues of prosocial actions. Only a handful of studies
have dealt with this question, but so far it appears that
television can also serve as an instrument of positive
influence. Friedrich and Stein (1975) submit evidence
about kindergarten children which supports the notion that
helping behavior can be learned through television viewing.
And Bryan (1971) and Bryan and Walbeck (1970) found that
sharing behavior increased among subjects who viewed a
charitable model on television.

Perhaps the most startling demonstration of television
as an instrument of socialization is found in the longi-
tudinal studies by Gerbner (1967, 1969). Although
Gerbner's methodology has often been challenged, his

observations have received considerable attention in

recent years, and warrant at least a brief review. In

28
the course of his extended "Cultural Indicators" project,
Gerbner and his staff at the Annenberg School of Communica—
tions have attempted to plot a profile of social events as
portrayed on television against viewers' perceptions of the
occurrence of these events in reality. For instance, in
most of his work which involves television's portrayal of
violence, Gerbner has attempted to learn if viewers who
watch an abundance of television perceive the real world as
a place which contains as much violence as displayed on
television. Again, methodological problems notwithstanding,
there is some evidence to suggest a positive correlation
between the amount of viewing and an individual's percep-
tion of reality in accordance with acts displayed on
television programs. It has been stated that heavy viewers
interpret reality through their television experiences;
their television is not simply a screen projecting scripted
programs, but it is seen as a window to the world. If such
a finding could be confirmed, implications for this society
would be most compelling.

Television and Language Learning

 

Recalling that the specific intent of this literature
review is to relate factors of language acquisition and
television as a possible source of language instruction,
any conclusions would, naturally, be conjectural in the
absence of research dealing directly with this issue. To
date there appear to be only several studies which relate,

even peripherally, to language learning and television.

29

In 1961, Wilbur Schramm published the first notable
review of the effects of television on children and
included among his various studies a section relating
television to language development processes. His subjects
were children in school at least two years who were drawn
from either of two towns, one which did receive tele-
vision signals, and the other which did not; this
offered a dichotomous situation no longer available to
researchers simply because U.S. towns without television
no longer exist. Even if some did, their situations would
represent such an aberration that their deviation would
disqualify them as a comparison group. Although Schramm's
measures lacked sophistication by assessing only vocabulary
levels, Schramm found, with some exceptions, that children
reared in the television culture enjoyed a more advanced
vocabulary than their deprived counterparts. Cassirer
(1966) cites studies in Caracas, Venezuela which indicate
that programs produced specifically for pre-schoolers had
the effect of building in viewers richer and more correct
vocabularies.

Vocabulary tests as single measures of language devel-
opment, however, have recently met with considerable
hostility by linguists. Dale notes:

The implicit assumption...is that vocabulary

growth alone reflects language development;

there is little or no supporting evidence for

such an assumption, no correlation between

vocabulary growth and other aspects of lan-
guage development (Dale, 1976:306).

30

Indeed, the direction for language assessment has been
toward more sophisticated measures which attend to syn-
tactical, organizational and process structures of language
(Zigler and Butterfield, 1968; Paraskevopoulos and Kirk,
1969; Zimmerman, 1969; Brown, 1970; Mecham, 1971).

Milkovich, Miller, Bettinghaus and Atkin (1975) under-
took a task to review television and language development
and utilized a word association paradigm to assess
linguistic maturity of their subjects. This test appears
valid and these authors cite research by Entwisle (1966) to
support strength of the measurement tool. As in Schramm's
study, subjects ranged from 7 to 12 years old, so have
already been through, at least, two years of school, but
unlike Schramm, who used dichotomous categories, Milkovich,
et a1. classified subjects according to a range of tele—
vision viewing habits. In general, findings of this study
showed that television viewing had a "consistent negative
effect on the rate of language development." (Milkovich
et al., 1975:12) While this study does deal with language
development among older children, it does not look at
language in the pre-school years where language patterns
are formed. In fact, Milkovich notes that

...heavy viewers may enter kindergarten with

language abilities greater than their light

viewing counterparts; even if television then

had strong negative effects, we would not be

able to determine the point in time at which

they would fall below the norms (Milkovich,
et al., 1975:11).

31

One additional issue requires attention at this point
concerning the specific characteristics of television which
allow it to draw such large and faithful audiences and then,
with viewers in its grasp, perform its voodoo of influence
and instruction. Simply put, what about television makes

it attractive, credible and potent?

In the relatively short period of time that it took
television to become integrated into the information/enter-
tainment schedule of people in this country, it grew in
prominence over other media, specifically newspapers, as
the most appealing and credible medium. The last word of
print gave way to the gospel of television.

One survey (Selnow, 1977) found that the majority of
people sampled first learned about most national and local
news stories through television, would tend to believe
the television over newspaper version of a discrepant news
story, would choose television over newspapers as a sole
source of information, and believed television news
accounts had fewer errors than newspaper stories.

The potency of television may be locked in three
principal elements: its high degree of credibility, ease
of use and social sanctions.

Studies conducted in the late 1960's demonstrated that
credibility of a source can generally be understood in terms
of its constituent parts of trustworthiness, expertise and

dynamism. In television news particularly, where viewers

32

are able to see and hear the reporter, a sense of identi-
fication with that individual contributes to the
believability of the story: Walter Cronkite wouldn't lie!
An even more compelling reason for viewers to place a sense
of trust in television can be found in this medium's capa-
city to bring the audience to the source of a story. Even
though it may not accurately represent the action taking
place (Lang and Lang, 1971), television footage of an event
allows the audience to witness that event and so confirm the
verbal description of the action. Seeing is believing: the
television presentation is a "replica of reality." In one
study cited earlier (Selnow, 1977) a number of respondents
indicated that on television they could see what was happen-
ing so they knew the truth.

There is little question about the dynamic qualities
of television. With the ability to manipulate variables
of sight and sound this medium can massage the two most
keen human senses. Siegal (1969) describes television as
a high-fidelity medium employing two simultaneous modes of
communication which come as close as possible to reality.
Furthermore, unlike newspapers which impose a processing
time lag between event and report, television can cover a
story live and instantly transmit this message around the
globe.

Siegal (1969) contends that the sense of expertise or
authenticity attributed to television is a function of its

vividness and fidelity. She notes that,

33

...both print and film appeal solely to vision,

but film has more intrinsic authenticity because

its fidelity is higher...if a communication

technique is both vivid and of high fidelity,

as is the color film or the color television

image, its intrinsic authenticity is especially

high (Siegal, 1969:268).

It would also be reasonable to conclude that if over a
period of time television has repeatedly demonstrated its
accuracy this too would contribute to its perceived ex-
pertise.

Two situational factors have contributed to the
potency of television. First, it is a medium which re-
quires almost no effort on the part of a viewer. He neither
has to fetch his morning edition nor does he have to ex-
pend a great deal of energy while viewing a program.
Instead, television is always available, costs almost
nothing to use, offers a continuing display of new material,
and is relaxing. Perhaps these characteristics have lead
to what many critics assess as an overuse of the medium.

Finally, as a function of its pervasiveness and
popularity, television has earned a very prominent place
in society, not only as a transmitter of culture as
suggested earlier, but as an element of the culture.
Television is the topic of daily discussions and the focus
of critical debate: Should television be allowed in the
courtroom? Should television air programs depicting
scenes which might inspire antisocial acts? People watch

television because other people watch television, and to

keep in social tempo people must listen to the beat of

34

television. Social sanction and reinforcement underscore

the perceived need for participation in this medium.

In summary, there is evidence to support the proposi-
tion that television can serve as an active source in the
learning of academic material, and as a carrier of infor-
mation for passive learning. It has been found to impart
social learning and so contribute to the socialization
process of young children, both as a negative influence by
its grim portrayals of the world and as a source of pro-
social information. Then, it was suggested that television
may play an even more vital role serving as a distorted
mirror of reality for growing numbers of television
viewers; a confirmation of this would undeniably establish
television as a most powerful influence. Finally, with
specific reference to the primary issue at hand, television
as a tutor of language, there is some evidence that this
medium has a positive influence and some that it has a
negative influence on language learning of older children.
Given all this, there appears to be adequate ground upon
which to conclude that television can both influence its

viewers and contribute to their learning processes.

Although television viewing may influence various
learning processes of children including language devel-
opment, a variety of other factors influence the child's

intellectual and language growth as well. For

35

purposes of this research, two principal elements have been
identified and will be reviewed briefly in the literature
then in light of data collected in this study. The role of
the mother is, perhaps, most salient in primary intellec-
tual growth and will be considered here in terms of socio-
linguistic theories. The effects of birth order and

number of siblings have also been identified as important
variables affecting intellectual growth, and these too will
be reviewed in this section.

Maternal Influences

 

If for no other reason than the mother spends more time
with a young child than anyone else spends with him, it
appears reasonable to infer that frequent interactions
between the child and mother serve most to shape intellec-
tual and language development. This is clearly recognized
by language theories which admit environmental influence,
and, perhaps, most strongly supported by sociolinguistic
positions. Bernstein describes the core of the mother-
child interaction as "a virtuous circle which is set up
and continually reinforced, for the mother will elaborate
and expand the embryo personal qualificatory statements
that the child makes." (Bernstein, 1958) Dynamics of this
mother's influence on child language and intellectual
development are set forth by Dally:

...In the earliest years a mother unconsciously

restricts or elaborates her child's linguistic

development and capacity and therefore his capacity
for educational development according to her own

social and educational background. She may
transmit a 'restricted' form of language

36

from which once adapted, it is difficult

to expand. Or she may transmit an 'ela-
borated' form, which leads to other possi-
bilities, both linguistic and imaginative.
For instance, one mother might say 'Pick

up the red ball and put it on the small
table,‘ whereas another less articulate
would simply say 'Put it there' and supply
the missing details with gestures. If a

child asks 'Why do things fall when I drop
them?‘ there is an immense difference

in attitude and influence between a mother
who explains the laws of gravity in a
simplified form suitable to the child's

level of intellectual development and the
mother who answers 'They just do', or

'Well, why shouldn't they?’ Such differences
repeated in hundreds of ways many times a

day as the child grows up, will have a pro-
found influence on his development and
personality. They affect not only his vocab-
ulary and his linguistic development but also
his feelings towards whatever is discussed
and his attitudes towards them. If he is
always encouraged or always snubbed when he
asks a question, his curiosity and responses
will develop accordingly (Dally, 1976:31).

According to this account, the mother has a substantial
influence on language development, intellectual growth,
curiosity and a host of other psychological processes. She
directly influences the child through her interactions with
him, and indirectly by controlling the events in his envi-
ronment which also affect his development. To a large ex-
tent, she controls the child's daily schedule including his
interactions with other people, his television viewing
patterns and other communications activities. It is impor-
tant to explore the nature of mother-child relationships in

this sample.

37

Sibling Influence on Intellectual Development

 

The issue of sibling influence on intellectual
development has received considerable attention during the
past decade. In one recent investigation, Breland examined
National Merit Scholarship Qualification Test scores in
terms of family size and birth order of examinees. From
this study, Zajonc observed that several findings were
particularly significant:

1. NMSQT scores generally decline with increasing

family size.

2. Within each family size they decline with birth
order.

3. The rate of decline decreases with successive
birth order.

4. There is a discontinuity for the only child, who
scores below a level that would be expected had
intelligence declined monotonically with increas-
ing family size.

5. Twins have comparatively low scores (Zajonc,
1976:227)

In each of these observations Breland found that birth
order and number of children contributed to significant
differences in performance on Merit Examinations, and
while he is quick to note that a variety of diverse condi-
tions contribute to the patterns of test scores, correlation
with sibling variables cannot go unnoticed.

Zajonc and Markus (1975) prOposed to describe the

38

dynamics of this observation with their explication of the
Confluence Model Theory:

The basic idea of the Confluence Model is that
within the family the intellectual growth of
every member is dependent on that of all the
other members, and that the rate of this growth
depends on the family configuration. Different
family configurations constitute different in-
tellectual environments. 'Intellectual environ-
ment' can be thought of in this context as being
some function of the average of the absolute
intellectual levels of its members. Note that
we are not considering IQ, which is a quantity
not related to age, but rather an absolute quantity
such as mental age. If the intellectual environ-
ment is conceived as an average of all the members'
'absolute' contribution, then it changes continually
as the children develop and it manifests the most
dramatic changes when there is an addition to or
departure from the family (Zajonc and Markus, 1975:87).

Zajonc notes that spacing between siblings plays an
important role in this relationship since greater spacing
allows more time for maturation of the previous sibling.
This provides a family community which has a higher average
intelligence level than one where children are spaced close
together. In the extreme case, for instance, if a child
were born only after preceding children reached maturity,
he would enter a more advanced community than if he Were
the only child. This can, perhaps, be better explained
with the following example:

...Consider the absolute intellectual levels

of the parents to be 30 arbitrary units each,

and of the newborn child to be zero. Thus,

the intellectual environment at the birth of

the first child has an average value of 20.

Suppose the second child is born when the

intelligence level of the firstborn reaches

4. The secondborn then enters into a en-

vironment of B30 + 30 + 4 + 0)/4 = 1?]. (Note

that since the intellectual environment is an
average of the absolute intellectual levels

39

of all family members, the individual is

included as a part of his own environment.)

If a third child is born when the intellec-

tual level of the firstborn has reached, say,

7, and that of the secondborn is at 3, the

family intellectual environment will then be

reduced to 14. (Zajonc, 1976:227)

Zajonc cautions against the apparent simplicity of
this example, noting that dynamics of the Confluence Model
are considerably more complex than portrayed here; the
essential elements, however, are represented in this example.

Although this literature makes no specific reference
to the effects of sibling variables on language development,
one might infer from the basic argument that because lan-
guage ability is strongly related to intelligence, a
similar influence might be anticipated.

Summary

Foregoing sections of this chapter have sought to
outline some of the prominent theories of language devel-
Opment spanning the ideological range of positions from
environmentalist to nativist orientations. Following this,
television, a relatively new variable in education and
acculturation processes, was considered in light of con-
temporary research dealing with impact of this medium.
Undeniably, the evidence suggests that television has the
potential to serve an instructional role for its audience.
The final portion of this literature review dealt with the
maternal influence on child development and considerations

of birth order and sibling number on the rate of intellec-

tual and, perhaps, language development.

40

Focus of the Present Study

 

While some research has drawn a link between tele-
vision and its impact on the language patterns of viewers,
none has dealt directly with the issue of television's
influence on language development of pre-school children
whose language is still undergoing rapid development.
Milkovich (1975) stated that his research did not attend
to the language of children entering schools, but assessed
language only after the influence of formal training has
already begun. Milkovich noted this gap in the body of
research and so the present study is concerned with this
group of young viewers.

This study embarks on an exploration of the relation-
ship between the language ability of young children and
their television viewing patterns with specific attention
to the kinds of programs children view. It further
explores maternal influence on the child's language abili-
ties and on the amount and kinds of television programs the
child views. First, by collecting this original data, and
then by establishing links among these variables, this
study represents a crucial step in expanding the under-
standing of the potential for television to influence its
viewers, and particularly in those young children during
the formative years of intellectual and language develop-
ment.

Based upon information from the literature reviewed

in this chapter, four outcomes might be predicted for this

41

analysis:

1. It can be predicted that a relationship will be
discovered between television viewing and language ability
of a child. Long hours of exposure to television language
models can be expected to stimulate some relationship
between viewing and language ability. Since the literature
does not deal adequately with variables of television
language, the expected nature of this relationship is not
particularly clear. There is some evidence, however, by
Milkovich, et al. which suggests that a negative relation-
ship between language ability and television viewing might
be anticipated.

2. It can be predicted from sociolinguistic theories
that a mother's education level (which serves as an index of
language and socioeconomic status) will relate positively
to a child's language ability measures. Furthermore, it
might be expected that a relationship exists between a
mother's education and the kinds of programs viewed by
a child.

Several probes in this investigation will test rela—
tionships outlined in the Confluence Model. In this study,
however, language ability measures are substituted for
standard intelligence measures used in previous research.
This will provide an added dimension to the theory on (a)
intellectual performance and birth order, and (b) intellec-
tual performance and number of children in a family. Based

upon provisions of the Confluence Model the following

42
predictions can be advanced:
3. It can be predicted that the number of siblings
will relate negatively to child language ability.
4. It can be predicted that birth order will relate

negatively to child language ability.

Statement of Hypotheses

 

There is a clear indication from the literature that
television has the capacity to influence its viewers.
This was demonstrated in the review of studies on learning
and television viewing summarized in this chapter. As a
complimentary issue from the literature on language devel-
opment, there is evidence that environmental factors can
intervene in the language development of young children.
By drawing both sets of theories together it appears
reasonable to propose that a relationship can exist between
television viewing and language ability of young viewers.

Evidence as to the nature of this relationship
(whether it is positive or negative) is very tenuous, and
predictions of directionality cannot reasonably be advanced
from the literature. This hypothesis explores a very
untested relationship and is thus stated in the null form:

Hypgthesis 1: There will be no relationship between

the total number of hours a child

spends watching television and his
score on the language ability index.

 

The preceding hypothesis pursues a very general issue

and so addresses the broad relationship between television

43
viewing and language ability. But it is evident from even
a casual observation that not all television programs offer
identical language models; the sophistication of language
varies among programs. In order to unfold the relationship
between total television viewing and language ability, the
following set of subhypotheses examine the relationship
between child language ability and individual program
categories. It is anticipated that this portion of the
analysis will unpack the different relationships which may
exist between child language and five categories of child-
ren's television programs which remained aggregated for the
"total viewing" analysis.

There is no information in the literature on which to
base a prediction concerning the relationships between each
of these five program categories and language ability, so
here, too, the hypotheses are stated in the null form:

Hypothesis la: There will be no relationship

between the number of hours a
child spends watching cartoon

programs and his score on the
language ability index.

 

Hypothesis 1b: There will be no relationship
between the number of hours a
child spends watching family
dramas and his score on the
language ability index.

 

Hypothesis 1c: There will be no relationship
between the number of hours a
child spends watching situation
comedies and his score on the
language ability index.

 

44

Hypothesis 1d: There will be no relationship
between the number of hours a
child spends watching educational
programs and his score on the
language ability index.

 

Hypgthesis 1e: There will be no relationship
between the number of hours a
child spends watching action
drama programs and his score
on the language ability index.

 

If, in fact, there was a demonstrated difference in
the relationships between child language and various pro—
gram categories considered by the preceding set of hypo-
theses, a logical follow-up investigation should consider
the actual language level of each program category in terms
of subject exposure hours. An index which was developed
to accomodate this analysis involved arithmetic operations
among the elements of (1) viewing time for each category,
(2) total viewing time and (3) an average language index of
programs within each category. The formula for this index
is presented in Chapter II.

It seems logical to propose that if a relationship
were discovered, programs with more sophisticated language
would lead to an advancement of child language, and pro-
grams displaying less sophisticated language would contri-
bute less to the advancement of child language. Consistent
with this argument, the following hypothesis was advanced:

Hypothesis 2: The greater the child's language

ability index score, the greater

his score will be on the exposure-
language level index.

 

Sociolinguistic literature posits a strong relationship

45

between the mother's language and child language develop-
ment, noting that language growth is largely dependent on
mother-child interactions. This portion of the analysis
deals specifically with this relationship. Although
language variables for the mother could not be collected,
information about the mother's education level was obtained
and serves in this review as an index of the mother's
education ability and socio-economic status. Bernstein
notes close interrelationships among these three variables,
and in the context of this analysis it is assumed that the
more advanced the mother's education level, the more
developed her language. In terms of the sociolinguistic
propositions, then, the following hypothesis was advanced:

Hypothesis 3: The higher the mother's education

level, the higher the child will
score on the language ability index.

 

Naturally, the mother might be expected to influence
more than just the language of a child. During the first
few years nearly all of a child's activities are controlled,
to some degree by parents, and particularly the mother.
Since television viewing is such an integral element of
this research, it becomes important to investigate the
relationship between the mother's education level (as an
index of language and socio-economic status) and the
television viewing patterns of children. The literature
does not suggest the nature of this relationship and so
directionality cannot be predicted. This hypothesis is

stated in the null form, as follows:

46

Hypothesis 4: There will be no relationship
between the mother's education
level and the total number of
hours of television viewed by
the child.

 

Beyond the general relationship between mother's
education and total language viewing, it is necessary to
sort among individual categories of programs to investigate
the relationships betweena mother's education level and
specific kinds of programs viewed by a child. This
unfolding process will provide specific information about
the influence which mother's education has on a child's
television repertoire. There is no basis on which to
predict directionality of these relationships, so the
hypotheses are stated in a null form:

Hypothesis 4a: There will be no relationship

between the mother's education

level and the number of hours
a child spends watching cartoons.

 

Hypothesis 4b: There will be no relationship
between the mother's education
level and the number of hours
a child spends watching family
drama shows.

 

Hypothesis 4c: There will be no relationship
between the mother's education
level and the number of hours
a child spends watching educa-
tional programs.

 

Hypothesis 4d: There will be no relationship
between the mother's education
level and the number of hours
a child spends watching situa-
tion comedies.

 

Hypothesis 4e: There will be no relationship
between the mother's education

 

47

level and the number of hours
a child spends watching action
drama programs.

Two principal elements of Zajonc's Confluence Model
state relationships between intelligence measures of a
child and (l) the number of siblings, and (2) birth order.
Details are presented earlier in this chapter. In accord-
ance with statements of the model, the final portion of
this research investigates two relationships outlined by
Zajonc, but substitutes a language ability index in place
of conventional intelligence measures. Application of the
language ability variable will provide valuable information
about the extensiveness of the theory. Based on Zajonc's
predictions, the following two hypotheses are proposed:

Hypothesis 5: The greater the number of children

in the family, the lower the child

will score on the language ability
index.

 

Hypothesis 6: Language ability index scores will
decline with birth order.

 

CHAPTER II

DESIGN AND METHODS

The principle question underlying this research con-
cerned the influence of television on language development
of pre-school children. First this study considered
whether or not total television viewing time had any effect
on the language of children, then pursued this issue with
an investigation of the relationships between particular
categories of programs and language ability of pre-school
viewers. Guided by literature which describes an influence
of the mother on child language development, this study
then considered the relationship between mother's education
level and variables of child language and television view-
ing patterns.

Finally, as an adjunct issue of language development,
the focus turned toward the effects of sibling variables
on language ability. Previous research on sibling effects
incorporated standard intelligence measures as dependent
variables; language ability measures are implicated in this

study.

Overview of the Study

 

Language samples were collected from 93 pre-school
children between the ages of three-and-a-half to five-and-
a-half years. Parents provided information about the child's

television viewing habits and also various demographic

48

49
statistics. Language samples were analyzed according to
procedures of the Developmental Sentence Scoring technique
which provided details about the rate of language develop—
ment for each child. This information was then analyzed
in terms of the child's television viewing habits. Popular
television programs were also analyzed according to the same
analytic procedure and this information was incorporated
into the analysis. Finally, this study considered the
relationships between a child's language and mother's edu-
cation level, number of siblings and birth order of the
subject.

Several terms used throughout this discussion will first
be reviewed in this chapter. Then, the sample of television
programs will be described, followed by a review of the child
sample used in this study. The next section will review
instruments used here, then data collection methods will be

considered. Finally, hypotheses will be delineated.

Terminology

 

Formal definitions are required for two terms used
throughout this dissertation in order to avoid any mis-
understanding either in the context of this study or for
extension of these findings to other circumstances:
Language: There is some ambiguity in the literature
concerning a precise definition of language, specifically
relating to the taxonomy of verbal samples. For instance,
what one researcher might term a "language sample," another

might identify as a "prelinguistic utterance." The use of

50

"language" in this research is specific. Here, language
is the verbal utterance illicited in the context of an
informal interview employing procedures outlined in the
Developmental Sentence Scoring procedure by Lee (1974).
(See appendix for details) The words and sentences pro-
duced by subjects during these interviews are the com-
ponents of language considered in this study.

Television: There are legal, technical, behavior research

 

oriented and other definitions ready to describe the sub-
stance, operation and boundaries of television. For
purposes of this research, television is defined as the
audio visual content of those programs which parents have
identified in viewing logs for their children participating
as subjects of this research. For instance, in the state-
ment, "what are the effects of television viewing..." this
specifically refers to effects of the audio visual content

of programs reported to be viewed by individual subjects.

The Child Sample

 

The age range for children participating in this research
(3 years and 5 months to 5 years and 11 months) was selected
for two reasons. First, language of children in this age
range typically represents an emergence from telegraphic
speech and marks the beginning period of complete adult
language. AlSo, even the youngest children in this age
category are able to respond in an interview situation. Toward

the upper age limit, it is desirable to avoid subjects who

51

have been too immersed in formal educational processes
which may impose confounding variables of curriculum
differences, teaching styles, emphasis on language, read-
ing, etc. This research departs from previous studies
which dealt only with older children already well into the
educational process (Schramm, 1961; Milkovich et al.,l975).

So, ideally, the present sample includes that narrow
age range where children have some degree of language
sophistication yet remain relatively uncontaminated by
formal educational systems.

All children in this sample were recruited through
neighboring day care centers and nursery schools which
offered a subject pool of sufficient number to meet the
needs of this research. Also, these institutions main-
tained an established communications channel to parents
whose input was essential in a key portion of the data
collection process. All arrangements with parents and
children were transacted through the schools (or day care
centers). Letters of introduction and appropriate forms
were sent to 190 parents, 102 of whom responded. Three of
these returns were incomplete and could not be used for the
research, 2 children discontinued pre-school and could not
be interviewed, and 5 children chose not to complete the
interview and so were eliminated from the study. In all,
93 subjects were finally included in this analysis. The
following institutions provided assistance in this

research:

52

Eastminster Day Care Center, East Lansing

Edgewood Cooperative Nursery, East Lansing

Edgewood Village Children's Center, East Lansing

Haslett Child Development Center, Haslett

Little People Day Care Center, East Lansing

Okemos Nursery School, Okemos

People's Church Pre-School, East Lansing

At the outset,severalcmaracteristics must be men-
tioned about the sample selected from these sources.
First, there is good reason to believe that the socio—
economic break from this sample drawn from day care centers
and nursery schools will not be representative of the
general population. Children from these schools can be
expected to represent a higher SES, and their parents
would, therefore, be likely to have a higher than average
education level. In fact, the data bears this out. From
this it follows that family activities might not be
comparable for a more general population, i.e., parent-
child interactions may vary on several dimensions including
the nature of these interactions, duration, level of dis-
course, etc. Also, media activities for this sample could
reasonably be expected to deviate from the norm for two
reasons. First, these children may experience a different
media schedule, e.g., more books or more creative activi-
ties and less television. Also, because these children
attend pre-school, television time for them has often
become displaced with this alternative school activity.

They might generally be expected to watch fewer hours of

television than the child who does not attend pre-school.

53

This, too, was substantiated by the data.

Another consideration which initially may appear to
present a problem concerns the educational activity of pre-
schools. This is particularly important in light of the
previous discussion outlining a rationale for the selection
of three to five year olds who are expected to be rela-
tively uncontaminated by formal educational processes.

Upon close review of this sample, however, it appears that
none of the schools listed here engage in formal language

training for the children on a regular basis.

Television Programs

 

This research also considers children's programs which
were identified by viewing popularity among (a) children
nationally and (b) child subjects of this research.

Seven programs with high national ratings were first
selected to provide language samples for this study (iden-
tified by Nielson, 1977); they include:

--Bugs Bunny

--The Waltons

--Little House on the Prairie

--s1x Million Dollar Man
--Land of the Lost

—-Happy Days
--Shazam

 

 

 

 

 

A second sample of television programs, specific to
children participating in this research, was compiled from
the television viewing logs maintained by parents. With
this listing of programs it became possible to focus the

analysis on shows most popular with this child sample.

54

These programs are listed below:

--The Electric Company
--Captain Kangaroo
--Sesame Street

--Mr. Rogers

--Shake, Rattle and Roll
--Superfriends Cartoons
--Wide World of Animals
--Walt Disney

 

 

 

 

 

 

 

 

Subsequent portions of this analysis deal with these
programs categorized according to program type; the following
groupings were used for this purpose:

Cartoons: Bugs Bunny, Shake, Rattle and Roll,
Superfriends Cartoons.

 

 

 

Family Dramas: The Waltons, Little House on the
Prairie, Walt Disney.

 

 

 

Action Dramas: Six Million Dollar Man, Land of the
Lost, Shazam.

 

 

 

Educational: Electric Company, Sesame Street,
Mr. Rogers, Captain Kangaroo.

 

 

 

Situation Commedy: Happy Days.

 

Wild World of Animals was not categorized into any of

 

these five groups.

Instrumentation

 

Several data collection instruments provided the
information necessary for this research.

1. Language Samples. Since the essence of this re-

 

search involved a comparative analysis between television
and child language samples, it was necessary to gather
language information from both these sources. This was
accomplished according to procedures specified by the

Developmental Sentence Scoring technique outlined in a

55
subsequent portion of this chapter.

2. Television Logs. Parents of each subject main-

 

tained a seven-day television log which reported all
viewing activities of their children. The form on which
they recorded this information included half-hour intervals
from 6:00 a.m. to 9:00 p.m. with provisions for any shows
which might be viewed at other times. This information was
then coded according to the name and category of each
program. An example of the television log is included in
the appendix.

3. Demographics. Demographic considerations are

 

standard in most social science research, and information
deemed to be pertinent for the present study was gathered
from parents in a "Confidential Background Information
Form." This questionnaire included the following elements:

--Child's age

--Number of children in family

--Position of this child in family (oldest, youngest,

etc.
--Parents' education
--Parents with whom the child is presently living

Frequencies within each of these demographic catego-

ries has been included in Tables 1 through 6.

Table 1. Children's Ages

56

 

 

 

 

 

 

 

Years/Months Frequency
3/5 3/12 12
4/0 4/4 36
4/5 4/8 18
4/9 4/12 13
5/0 5/4 7
5/5 5/11 7
Table 2. Number of Children in Family
Number of Children Frequency
Only Child 21
2 Children 55
3 Children 11
4 Children 6

 

57

 

 

 

 

 

 

 

Table 3. Position of Subject among Siblings
Position Frequency
Oldest 32
Between Youngest and

Oldest 8

Youngest 32
Only Child 21

Table 4. Mother's Education
Highest Grade Level Frequency
Some High School 2
Graduated High School 8
Some College 16
Graduated College 26
Some Graduate Work 17
Master's Degree 1?

Doctorate

 

58

Table 5. Father's Education

 

 

 

Highest Grade Level Frequency
Some High School 2
Graduated High School 4
Some College 14
Graduated College 8
Some Graduate Work 12
Master's Degree 19
Doctorate 32
Not Reported 2

 

Table 6. Parents with whom Child is Living

 

 

 

Condition Frequency
Mother Only 12
Father Only 2

Both Parents 79

 

59

Data Collection

 

1. Language Samples: Language used by children and

 

language displayed by the television programs children
watch rested at the core of this research. In order to
develop a sound data base for language analysis, it was
necessary to identify an assessment procedure which could
meet several criteria. First, this tool had to employ a
content analysis format to accomodate the language tran-
scriptions from television programs. Then, it had to be
sufficiently sensitive to detect subtle differences in
language patterns of young subjects participating in this
study. Finally, it was necessary to substantiate the
accuracy and reliability of this instrument from its use
record.

An exhaustive review of available instruments identi-
fied the Developmental Sentence Scoring Procedure as most
suitable for analysis of language samples collected in this
study (Lee, 1974). In brief, the Developmental Sentence
Scoring procedure provides a method for making a detailed,
readily quantified and scored evaluation of a child's use
of standard English grammatical rules. In a clinical
setting, where it continues to be widely and successfully
used, this method provides a tool for measuring a child's
growth and progress throughout the period of clinical
therapy. Furthermore, based on an extensive sampling of
children, researchers have charted abstract scores in terms

of normal developmental ages (year and month), which allow,

60

at least with some accuracy, statements concerning the
translation of television and child language scores into
terms of normal chronological age.

Child language samples were gathered according to
prescribed techniques outlined by Lee for the Developmental
Sentence Scoring assessment procedure. Essentially, each
child participated in a thirty minute discussion with a
trained interviewer who provided several props to facili-
tate the illicitation of language in the form of story
telling and role playing. These sessions were audio
recorded and later 50 sentences from this corpus were
transcribed for content analysis according to DSS specifi-
cations.

Television language samples were collected from
selected programs aired during Fall/Winter, 1977, and from
each of these, 50 sentences were transcribed for analysis.

2. Television Logs and Demographic Information: A letter

 

was sent to the parents of each child between the ages of

3 years, 4 months and 5 years, 10 months, and who attended
one of the seven schools identified for this study. The
letter provided a description of this study and explained
how children would be asked to participate. Also, parents
were requested to complete a questionnaire which included
specific demographic data, and a television viewing log in
which they were asked to record the programs watched by
their child. Directions for these forms were provided, and

a stamped, addressed envelope was included for return of

61

the completed material. Parents were invited to complete
an address card if they wished to receive an abstract of
this completed research. A copy of each of these forms,
the directions and letter of description are exhibited in

the appendix.

Data Preparation

 

The data base for this research was comprised of
language samples collected from 93 child subjects and 15
children's television programs. Each of these samples was
analyzed according to the Developmental Sentence Scoring
procedure designed by Lee (1974). Since the specific
procedures are quite extensive, interested readers are
encouraged to refer to Lee for details. In general terms,
though, this process involves the extraction of 50 senten-
ces from a larger corpus obtained in the full 30 minute
interview with child subjects (or from transcriptions of
television program samples). These sentences were then
analyzed according to grammatical content as outlined here:

Eight categories of grammatical forms have been
selected as showing the most significant devel-
opmental progression in children's language:
(1) indefinite pronoun or noun modifier, (2)
personal pronoun, (3) main verb, (4) secondary
verb, (5) negative, (6) conjunction, (7) in-
terrogative reversal in questions, and (8)
why-question...Credit is given only when a
structure meets all the requirements of adult
standard English, and this includes syntactic,
morphological and semantic conventions...a
score of l is added for every sentence which
meets all adult standard rules (Lee, 1974:
136-137).

62

A "total point score" was computed, then, through
simple arithmetic calculations, each language sample was
assigned an average sentence score. Finally, based on
data from previous research by Lee, this score was con-
verted to an age rating at which an average child might
be expected to achieve such a score. (See appendix for more
DSS information)

In summary, this analysis yielded information about
the language age of a child (and television language), the
score for each grammatical category in his language sample,
and a score for the average complexity of words used with-
in each grammatical category.

Special Indices

 

In order to facilitate data analysis, it was necessary
to develop two additional indices:

Language Ability Index: Following all of the data collec-

 

tion and proCessing procedures, two essential statistics
were recorded for each subject: the chronological age and
language age. In order to pursue the analysis in terms of
language progress for each subject, it became necessary to
relate these two elements to determine if the child's lan-
guage was advanced or delayed for his particular chronolo-
gical age. This could reasonably be accomplished in two
ways. First, the child's chronological age could be sub-
tracted from his language age, and this would provide an
index representing language acceleration or retardation

with respect to chronological age. Second, as an analogue

63

to conventional intelligence quota measures, the child's
performance could be divided by his chronological age and
the resulting index would also deal with the development of
the child's language with respect to his actual age.

In fact, both procedures were accomplished and since
the correlation between them was nearly perfect either one
could have been used. While both of them were employed in
all analyses, it is necessary to report only one because of
similarity of results. In this report the second index will
be presented since arithmetically it corresponds most
closely to the intelligence quota measures used throughout
the literature.

Exposure-Language Level Index: Language analysis of tele-

 

vision programs revealed that different categories of
programs offered different levels of language sophistica—
tion. For instance, cartoon shows displayed a more
simplistic language than educational programs. From this
it would follow that a child who viewed an abundance of
cartoons would be exposed to a different kind of language
model than a child who concentrated his television viewing
on educational programs. In order to account for both the
kinds of programs viewed and the number of viewing hours
within each category, an index was developed which multi-
plied the number of viewing hours for a program category
times the average language index of that particular
category. All product results for each category were

added together then divided by the total number of viewing

64
hours. This provided a weighted index of viewing hours

according to category type.

hcar --hours of cartoons viewed by a subject
(Category weight: 441)

 

hsc --hours of situation comedies viewed by a subject
(Category weight: 575)

heduc --hours of educational programs viewed by a
subject (Category weight: 561)

hfd --hours of family drama programs viewed by a
subject (Category weight: 529)

had --hours of action drama programs viewed by a
subject (Category weight: 653)

ELL Index =
(hcar)(44l)+(hfd)(529)+(heduc)(56l)+(hsc)(575)+(had)(653)
(hcar+hfd+heduc+hsc+had)
Operationalization of Variables

and Specification of Variable Relationships

A variety of contemporary research has dealt with the
effects of television on viewing populations. In an expan—
sion of this body of literature, the present study under-
took to examine the effects of television viewing on the
languagecflfpre-school children, an age group which has been
reported nationally to be among the heaviest viewers of
television programs. Because these children are exposed to
such large amounts of television and because their cogni—
tive and language abilities are undergoing such rapid
development, it is particularly important to explore any
effects which television may have on these children.

In this study, the nature of television viewing

65

included the kinds of programs watched and the hours of
exposure for each subject. A complimentary set of data was
generated for television program language samples which
served as an indication of the kinds of language models
available to children. The analysis then reviewed both
sets of data together.

In addition to this principal concern for the effects
of television on the language of young children, this
research sought to review subject language development
according to theories which consider the influence of
maternal and sibling variables. Specifically, according
to socio-linguistic propositions this research considered
the maternal influence on child language and in light of
Zajonc's Confluence Model predictions, reviewed the
influence of birth order and sibling number on child

language.

Data Analysis

 

All of the data gathered in this research was prepared
for computer analysis. Each of the relationships outlined
in these hypothesis statements were reviewed with correla-

tion and partial correlation statistics.

Summary

The principal task of this study was to investigate
relationships which exist between television exposure and
language ability of young children. In pursuit of this

objective a set of hypotheses were proposed which dealt

66
with the amount of viewing and categories of programs
viewed by the subjects of this study. In addition, this
research explored the influence of mother's education and
sibling variables on language development.

The data base for this research consisted of (l) lan-
guage samples from 93 pre-school children, (2) records of
their television viewing patterns, and (3) language samples
of popular children's television programs. Language
samples from both child subjects and television programs
were analyzed according to Developmental Sentence Scoring
procedures. Information from the television logs was
catalogued and coded for analysis.

Demographic data collected from parents provided the
necessary information to explore predictions of the Con-
fluence Model concerning the influence of sibling number

and birth order on language ability.

CHAPTER III

ANALYSIS OF RESULTS

Introduction

 

Results of this analysis will be presented in the
following order:

--The relationship between language ability and
total hours of television viewing.

--Relationships between the five subcategories
of programs and language ability.

--The relationship between mother's education
level and child language ability.

--Relationships between mother's education level
and television viewing patterns of children,
including total viewing time and five subcategories
of programs.

--The relationship between number of children and
language ability.

--The relationship between birth order and language
ability.

Total Television Viewing Time and Child Language

 

The first major hypothesis of this study proposed a
relationship which is fundamental to the central issue in
this research. In the null form, Hypothesis 1 stated that
there would be no relationship between the number of hours
a child spends watching television and his score on the
language ability index. While total viewing time may ig—
nore any differences which exist among program types in the
subject's viewing repertoire, it was anticipated that this

total figure would provide some indication of an overall

67

68

relationship with child language. And, since the litera-
ture offered little basis on which to predict directionality,
investigation of this relationship proceeded without a clear
expectation of how these variables would relate to each
other, thus the hypothesis was stated in a null form.

"Language ability" in this analysis was represented by
an index which divided the child's language age (as compiled
with the Developmental Sentence Scoring procedure) by his
chronological age. "Total hours of television viewing"
represents the complete viewing schedule for one week.

Results of the correlation analysis revealed that
between total viewing hours and child language level there
was a correlation coefficient of -.1669, significant at the
.055 level.

While this correlation is not impressively strong,
given that it accounts for only a small portion of the
variance between these two variables, several points under-
score the importance of this finding.

First, as described in a previous chapter, the
Developmental Sentence Scoring procedure was extracted from
its original setting in clinical speech therapy situations
where it serves to help identify speech deficient children.
With expanded resources for this research it might have
been possible to develop an instrument which was even more
suited to identify subtle distinctions among subjects and,
thereby, expand the variance in language scores. As it

stands, the Developmental Sentence Scoring procedure was

69

successful in identifying substantial differences in lan-
guage performance.

Second, because of conditions established for subject
selection, there is a very narrow age range for subjects
participating in this research. This age limitation leads
to a compression in the variance among language scores,
which, in turn, contributes to less dramatic correlations
among these variables.

Also, the subject pool was limited to children in day
care centers and nursery schools, and this restriction
tends to reduce variance because of the greater concentra-
tion of higher SES individuals. Most significant, however,
is the imposition of a uniform daily schedule on subjects;
the school setting restricts variation in television view-
ing patterns and other activities. If subjects had been
drawn randomly from the general population, a much greater
variance in viewing schedules might be anticipated.

In spite of these elements which were operating against
dramatic results, a substantial correlation was obtained,
significant at the .055 level. Directionality of this
statistic is of principal importance and serves to portend
other findings of this research. It appears from this
observation that television viewing has a significant
negative relationship with language development of young
children in this sample. The null hypothesis is, therefore,
rejected. In light of the evidence disclosed in this

analysis, the hypothesis could be restated as follows:

70

There will be a negative relationship between the number
of hours a child spends watching television and his score
on the language ability index.

Television Viewing Time by Category and Child Lapguage

 

Hypotheses la through 1e consider subcategories
within the general framework of all television programming.
Development of this portion of the research reflects a
recognition that there are a variety of language models
among program types, all programs are not viewed by chil-
dren with equal intensity and attention levels, nor is
child viewing time distributed equally among all program
types. The goals of this series of analyses is to unfold
the total television viewing variable into component parts
and thereby investigate specific relationships between
program types and language development.

A summary of results for the correlation analysis is
presented in Table 7.

Hypothesis la, in the null form, proposes there will
be no relationship between the number of hours a child
spends watching cartoon programs and his score on the
language ability index. As in the case of total television
viewing there was no suggestion in the literature as to the
directionality of the relationship, if any, between cartoon
viewing and child language ability.

As the correlation of -.2391 indicates, there was a
negative relationship between cartoon viewing and child

language, significant at .011. Such a strong relationship,

71

 

 

 

 

mmo. Ahvm.v Aamo.v Amma.v Aamo.v AHHo.V mafiaflbd
mmwa.| «awe. mmma.: whoa. mona.l Hmmm.| ommsmcmq
mmfiuommumo memumoum mEmumonm mfimumoum mfimnmoum mfimnmoum
Emnmoum macho mooaoo accoflu manna coouumu
Haa coHuoa cofiumsufim nmonmm Assess
.mEmnmoum cowmw>maoe
mo mmfluommumo m>flm cam muflaflnm momsmcmq mo mcowumHmHHou .b manme

72

particularly in light of the foregoing explanation of
restraints on the variance in this sample, is a noteworthy
finding which suggests that as cartoon viewing among sub-
jects increases, the language ability displayed by children
tends to decrease.

Given this, the null hypothesis is rejected, and a
restatement of the relationship between these variables may
proceed as follows: "There will be a negative relation-
ship between the number of hours a child spends watching
cartoon programs and his score on the language ability

index."

Hypothesis 1b states there will be no correlation
between the number of hours a child spends watching family
dramas and his score on the language ability index. Here,
too, with a correlation of -.l708 there appears to be a
negative relationship between viewing hours of programs in
the family drama category and language ability. Like
cartoons, family dramas, designed with specific appeal to
young viewers, appear to relate negatively to child lan-
guage development. The null hypothesis here is also
rejected in favor of the following proposition: "There
will be a negative relationship between the number of
hours a child spends watching family dramas and his score

on the language ability index."

Hypothesis 1c predicted there will be no correlation

between the number of hours a child spends watching

73

situation comedies and his score on the language ability
index. With a correlation of —.1399, significant at .091,
the relationship appears to be too weak to reject the null
hypothesis, and with this set of data there can be no
support for a relationship between viewing of situation
comedies and language ability.

Hypothesis 1d deals with the relationship between
educational television and child language, and so has
particular interest given the stated goals of this pro—
gramming type. Even though one might expect educational
television to provide positive language models there is
no previous research to substantiate this position, there-
fore, the hypothesis was stated in null form: There will
be no correlation between the number of hours a child
spends watching educational programs and his score on the
language ability index.

Support for a relationship between hours of educational
television viewing and a child's language is not very
strong with a correlation of .1077, significant at the
.152 level. With this sample there is no substantial
indication that children who viewed more hours of educa-
tional television programs tended to score higher on the
language ability index. With this evidence, the null

hypothesis cannot be rejected.

The final subhypothesis in this series predicts that
there will be no correlation between the number of hours

a child spends watching action drama programs and his score

74

on the language ability index. In this case the correla-
tion is so weak (.0414; significance level, .347), that the
null hypothesis cannot be rejected, and with this set of
data there can be no support for the case of a relationship
between language ability and action drama viewing.

Language Ability and Exposure-Language Level Index

The first series of questions asked by this research
concerned the fundamental issue of television viewing and
language ability, reviewing first the relationship with
television viewing overall and then viewing among individual
program categories. The issue considered in Hypothesis 2
involves a related issue which accounts for (l) the
differences of language level within each program category,
and (2) the amount of time a child spends viewing programs
within each category. Essentially, this weights viewing
time by language sophistication within categories and,
therefore, provides a more direct assessment of the rela-
tionship between television viewing and language ability.
Readers interested in the specific calculations for the
"Exposure Language Level Index" are referred to the dis—
cussion in Chapter II.

Results of the correlational analysis reveal that
between the language ability index and exposure-language
level index there is a correlation coefficient of .1489,
significant at the .077 level. While the significance of
this correlation is not particularly high, the positive

directionality is a very important finding given other

75

evidence already reviewed in this chapter.

From the previous set of hypotheses a trend of rela-
tionships was demonstrated between language ability and
television viewing within categories of programs. For
instance, cartoon programs and family drama shows dis-
played a negative relationship with language ability. In
the present analysis, programs within these categories,
selected on the basis of their popularity among this
sample of children, demonstrated a relative language
sophistication which might be anticipated given previous
findings. The underlying issue in all these instances
concerns the level of language modeled by programs within
each category.

By operating within this rational structure, findings
of the present analysis become more meaningful. Essen-
tially, the positive correlation suggests that as children
viewed a greater proportional number of programs lower on
the language level scale, their own measured language
ability rating tended to be lower. Naturally, the
complimentary position suggests that the greater propor-
tional number of hours a child spent viewing programs
higher on the language level scale, the higher his own
measured language level score tended to be. Several
explanations stand to interpret this observation and
provide some alternative views on the dynamics among
variables at play here. These are reviewed more thorough-

ly in the next chapter. Hypothesis 2 is confirmed by the

76

data.

Relationships with Mother's Education Level

 

This portion of the analysis considers the mother's
influence on several variables relevant to the central
issue of child language ability and television viewing.
The mother plays an integral role in nearly every phase of
a young child's life including fundamental language train-
ing as noted by many sociolinguistic theories. She can
also be expected to influence variables of television
viewing, where she is likely to control, to some degree,
the amount of television watched by the child, and the
kinds of programs included in the child's television
viewing repertoire.

In this series of hypotheses the mother's education
level was used as an index of the mother's language ability
and socioeconomic status, as described in Chapter 2.

The first hypothesis in this series deals with a
fundamental issue in sociolinguistic theories which pre-
dict a positive relationship between the family socio-
economic status (with particular attention to the mother)
on a child's language development. Hypothesis 3 states
that the higher the mother's education level, the higher
the child will score on the language ability index.

Results of this correlation analysis reveal that,
indeed, there is a positive correlation of .3031, which is
significant at the .002 level. This is a clear correlation

between the mother's education level and the child's

77

language ability rating and while it confirms basic

tenets of sociolinguistic theory, it further substantiates
the comparability of this set of data with other language
research. This adds confidence that data involved in this
analysis, collected even under constraints imposed here,
relates positively to data collected under more general

conditions. Hypothesis 3 is confirmed by the data.

While sociolinguistic theories predict directionality
for the influence of mother's education on child language
ability, there is no expected direction for the relation-
ship between mother's education and television viewing by
children. The following set of hypotheses deals with this
issue, first by considering television viewing overall,
then by looking at specific categories of programs.

Because directionality could not be predicted, all hypo-
theses are stated in the null form.

Hypothesis 4 proposes that there will be no correla-
tion between the mother's education level and the total
number of hours of television viewed by the child. A
correlation of -.l717, significant at the .05 level was
found to describe the relationship between these variables,
whereas the mother's education level increases, the child's
overall television viewing decreases. Influence of the
mother on a child's viewing time is evident here, the
remaining question now concerns the mother's influence on

program selection within specific categories.

78

Findings for the next series of hypotheses are
reported in Table 8.

Hypothesis 4a states there will be no relationship
between the mother's education level and the number of
hours a child spends watching cartoons. Results of this
analysis demonstrate a correlation of -.2473, significant
at the .008 level. This finding adds to previous infor-
mation gathered about cartoon programs: cartoons display
the lowest level of language, they relate negatively to a
child's language ability rating and in this analysis they
are found to relate negatively to the mother's education
level. The null hypothesis is, thereby, rejected and the
relationship restated as follows: "There will be a nega-
tive relationship between the mother's education level and
the number of hours a child spends watching cartoons."

Hypothesis 4b proposes.that there will be no rela—
tionship between the mother's education level and the num-
ber of hours a child spends watching family drama shows.
The correlation analysis reveals that like cartoons,
family dramas are negatively correlated with mother's
education level, with a coefficient of -.3217, significant
at the .001 level. It should be indicated that family
dramas did not score high on the language ability rating.

Hypothesis 4b is rejected, and in light of the
results, the following statement is submitted as a descrip-
tion of the relationship between these variables: "There

will be a negative relationship between the mother's

79

 

Amo.v AHv¢.v Ammm.v Ammo.v AHoo.v Amoo.v GOHDMODUM
hHhH.I mmHo.I owho.| mmma. hHNm.I mnvm.l m.Hm£uoz

 

mofluommumu mEmumonm mamumoum mEmumonm mamumoum mfimumoum
Emumoum manna womsou Hmcowu mEmuo coouumu
Had cowpoé coaumsuflm Imosom maflEmm

 

 

.mamnmonm coflmw>mama
mo mmfluommumu m>Hm cam coaumosom m.nwcpoz mo mcoflumamuuou .m magma

80

education level and the number of hours a child spends
watching family drama shows."

Hypothesis 4c dealt with educational shows and stated
that there will be no relationship between the mother's
education level and the number of hours a child spends
watching educational programs. The analysis revealed a
positive correlation of .1756, significant at the .056
level. From this it follows that as the education
level of the mother increases, so do the number of
viewing hours for educational programs. The null hypo-
thesis is rejected in favor of the following restatement
of the relationship: "There will be a positive relation-
ship between the mother's education level and the number of
hours a child spends watching educational programs."

Two final hypotheses in this series state:

(Hypothesis 4d): There will be no relationship

between the mother's education level and the

number of hours a child spends watching situation

comedies.

(Hypothesis 4e): There will be no relationship

between the mother's education level and the

number of hours a child spends watching action

drama programs.

In the first instance, between mother's education and
situation comedies, this analysis revealed a correlation of
-.0780 (not significant) and in the second, between
mother's education and action dramas, a correlation of
-.0155 (not significant). With such weak correlations it

can be concluded that the null hypothesis in each case

cannot be rejected.

81

At this point in the analysis it became evident that
a series of relationships was developing among three
principal variables involved in this research: mother's
education level, child's language ability, and television
viewing variables. First, there was a demonstrated
relationship between child language ability and television
viewing on several dimensions. Then it became evident
that mother's education was related to child language
ability and television exposure patterns of children.
From this an obvious question arises concerning the
independent covariation of child language and television
viewing, each related only to mother's education level.
Specifically, if mother's education was controlled in a
partial correlation analysis, would the relationship
between television viewing and child language ability
diminish toward zero? If this were to occur, there would
be little case for a relationship between television
viewing and child language ability. Table 9 displays the

results of the zero order and first order partial analysis.

82

 

 

AHNH.V mmma.u Aamo.v mmma.| mowomosoo cOAumsufim

1Hk~.c memo. immfl.v sacs. mamnmoum Hmcoaumosem

floom.v nemo.| Aamo.v mona.| mmsmuo maesmm

“avo.v mHmH.I “HHo.v Hmmm.u mcoounmu

AAHH.V HmNH.- immo.1 mmms.: muses meazmh> sauce
mocmoflmacmfim cofipmosom oocmofimecmflm soap
: xmocH m.nonpoz mom I xmcsH [maonuoo
muflaflbm mcaaaonusoo sofiu mufiaflbm noouo oumN

ommsmcmq Imamnuoo umouo umnfim momsmccq

 

coaumoscm m.uocuoz now mcflaaonucoo
mpflaflbm ommsmsmq can moanmanc> mcH3ofl> conH>mee mo mcoflpwaonuou Hafiuumm .m magma

83

In this partial correlation analysis, as one would
expect, there was a slight reduction in the correlation
between television viewing variables and the language
ability index. Such a reduction is to be expected given
that a portion of the variance is restricted when mother's
education is partialled out. However, the significance
of these findings rests largely in the maintenance of
directionality between variables. Specifically, between
the language ability index and total viewing hours,
cartoons, family dramas and situation comedies, there
continues to be a negative correlation even after the
effects of mother's education are removed. Similarly, even
though the correlation between language ability and educa-
tional programs is reduced a positive correlation remains.
It is this observation which underscores the relationship
between television viewing and language ability of chil-
dren in this sample. Irrespective of the influence of
mother's education, which has been demonstrated to have
such a strong influence on the child and his television
viewing patterns, the relationship between television
viewing and language ability is maintained.

Number of Siblings and Birth Order Variables

 

Hypothesis 5 does not concern television viewing
directly, but rather language development as influenced by
the total number of siblings with which a child is raised.
An explanation by Zajonc was provided earlier (Confluence

Model), and the reader will recall that all previous

84

research has dealt with standard measures of intelligence,
and not specifically with language variables. Since the
literature has demonstrated that language shares a strong
relationship with other intellectual activities, it
becomes of interest here to review this language data
in terms of propositions from the Confluence Model.
According to predictions of this theory the hypothesis
stated that the greater the number of children in the
family, the lower a child will score on his language
ability index.

Results of the analysis confirm this hypothesis with
a correlation coefficient of -.l84l, significant at the
.039 level. Based on this finding, the Confluence Model
appears to apply to language as well as standard intelli-
gence measures employed in the mainstream of research on
sibling influence. There is support here for the proposi-
tion that the number of children in a family not only
effects the child's performance on standard intelligence
measures, but also on the way he uses language. Hypothesis
5 is confirmed. V

Hypothesis 6 deals with the position of a child in
relation to his siblings and proposes that language ability
index scores will decline with birth order. This hypothe-
sis, also derived from the Confluence Model, proposes a
negative relationship between numbers of children and
language ability. According to provisions of the theory,

children with no siblings were omitted from this analysis.

85

For all children with siblings, this analysis revealed a
correlation coefficient of -.2005, significant at the .04
level. This second prediction of the Confluence Model was
confirmed by the data, again suggesting that this theory,
which initially concerned standard intelligence measures,
has been accurate in predicting relationships with lan-
guage ability as well. Hypothesis 6 is confirmed by the

data.

Summary

This research considered three principal relationships
concerning the development of child language. First, it
looked at television viewing and language development,
then influence of the mother on child language and finally
it considered the relationship between birth order and
number of siblings on language development. A summary
of findings for the principal hypotheses and sub-hypotheses
is presented in Table 10.

The fundamental concern here was with the effects of
television viewing on language development. It was found
that for television viewing overall, there was a negative
correlation with language ability. Similarly, for
sub-categories of program types, cartoons and family
dramas had a negative correlation with language ability.

In order to assess an overall effect of television
language models on children, an exposure-language level
index was created which accounted for language complexity

of programs within individual categories, and number of

86

BBC.

hem.

mmH.

Hmo.

Hmo.

Hao.

mmo.

awed.

vawo.

whoa.

mmMH.I

mona.l

Hmmm.l

mmmH.l

.xoecfi Hm>oH
mmsmmcmarmHSmomxo on» so on Haws ouoom we: umumoum on“ .muoom
xoccfi huflafinm ommsmcma m.oHHco map noummnm one um mfimwcuommm

 

xoocfi hufiafinm ommsmsma map so ouoom mfin can
mmamuo cowuom acacoumz mocmmm caflno m mason mo amass: on»
coozuon maanOHumaou on on HHH3 muons ”UH mfimonuomwm

 

.xmocfl muflaflnm mmmsmcmH msu so mnoom was cam msmumoum
accoflumosom mafinoums mocmmm oaflno m mason mo Hones: mam
comsumn maanOHumamu on on HHH3 muons "UH mwmocuomwm

.meCH muflaflnm momsmsma may no muoom man can mmfiooaoo
coaumsuam mcwgoum3 mocmmm oaflno m muson mo Hogans may
coo3uon mHanOADMHoH on on HHH3 muons ”0H mflmwcuommm

 

.xoocw muflaflnm omsmmcma may no onoom man can mmEmuo
SHHEmm mcwcoum3 meadow oawco m mason mo Hones: omw
cmm3umn mfinmsofiumamn on on HHH3 whose “DH mflmocuommm

 

.xoocH muHHfinm momsmcma map so muoom me: can msmumoum
coounmo mcacoum3 monomm oaflco a mason mo woman: may
cwozuon awanOHumHmu 0: on Has? ouwna "ma mammnuomwm

 

.xoocfl huflawnm ommsmcma man so ouoom man
can cofimfl>oaou mcwnoum3 monomm pawno m mason mo amass:
on“ com3uon mflcmcowumamu on on HHfl3 muona “H mwmocuomwm

 

 

OOGMU
nauacmam

cofluma
lounou

mommsuomhm

 

 

muasmmm mo mumEEsm .oH wanna

87

 

 

 

 

 

 

 

 

Hos. mmuo.u .ouuno on» an ooaou> mmsmuo
n0uuoo mo uonesn onu coo Ho>oa couuoosoo m.uonuoe onu
noo3uon munmnouuoaou on on HHHB ouona "ow mumonuommm

mom. oooo.- .ouuno on» an cosmu> mouooeoo couumsuum

mo uonEsn onu ono Ho>oa scuuoosoo m.uonuoa onu
noozuon munmnouuoaou on on Haas ouonB "we mumonuommm

omo. omou. .ouuao was an oo3ou> msmuooua Honour
loosco mo uonEsn onu ono Ho>oH nouuoosoo m.uonuoe onu
noo3uon mwnmnouuoaou on on HHH3 ouone Hoe mwmonuomwm

Hoo. namm.| .oauno onu xn oo3ow> moEouo
wausow mo uonEsn onu coo Ho>oa GOuuoosoo m.uonuoE onu
nooBuon muanOuuoHou on on HHH3 ouona “no mumonuomwm

woo. mnem.| .oauno on» an oo3ou> mEoumoum

noouuoo mo uonEsn onu ono Ho>oa couuoosoo m.uonuoE on»
noozuon munmnowuoaou on on Hau3 ouonB uov mumonuommm

omo. ouou.u .ouuno on» an oosou> acumu>muou

mo muson mo uonfidn Houou on» can Ho>oH nouuoosoo m.uonuos
onu noo3uon munmnouuoaou on on Haas ouonB no mumonuommm
moo. umom. .xoocu spuuunm oomsmcmu may no muoom uuus ouuno was nonoun
onu .Ho>oa noHuoosoo m.uonu05 onu uonmun one um mumonuomwm
oonoo nouuoa
lumunmum Iouuov momonuommm

 

 

.A.U.pcoov 0H manna

88

 

 

 

ovo. moom.l .uoouo nuan numB
oneeooo Haws mouoom xoonu eueauno omosmnoq no mumonuommm
omo. Hoou.u .xoocu wouuunm
omsomnoe onu no ouoom Hae3 oeeno on» uozoe onu .SHMEoM
on» nu couoeeno mo uonEsn on» uouooum one um memonuomwm
oonoo coauoe
lemunmem Iouuou momonuomem

 

 

.A.©.ucoov 0H oHQoe

89

viewing hours for each program type. In this analysis

the hypothesis was confirmed with the observation that the
greater the child's language ability index score, the
greater his score was on the exposure—level index. This
suggested that as children viewed a greater proportional
number of programs lower on the language level scale, their
own measured language ability rating tended to be lower.
Similarly, the greater proportional number of hours a
child spent viewing programs higher on the language level
scale, the higher his own measured language level score
tended to be.

A second set of investigations dealt with the relation-
ships between mother's education level and (1) child lan—
guage ability and (2) television viewing patterns of the
child. In the first instance a positive correlation was
found between the mother's education level and the child's
language ability index. Then a negative correlation
described the relationship between mother's education
level and the total number of television hours viewed by
a child. Within separate program categories, negative
correlations were also found between mother's education
level and cartoons, and family dramas. A positive
relationship was noted between mother's education and
educational programs.

Finally this research considered language ability in
terms of theories advanced by the Confluence Model con-

cerning the total number of siblings with whom the child

90
is raised, and the birth order of a child. In the first
instance, the hypothesis was confirmed noting that the
greater the number of children in the family, the lower
the child tended to score on the language ability index.
Then, the second hypothesis was also confirmed with the
finding that language ability index scores declined with

birth order.

CHAPTER IV

DISCUSSION

Structure of the Investigation

 

The primary investigation of this study examined the
relationship between television viewing and language devel-
opment of young children. In this portion of the analysis
a child language ability index was correlated first with
total television viewing time and then with five subcate-
gories of program types. The next analysis considered
language ability in terms of a composite exposure-language
level index which represented program exposure weighted by
language sophistication modeled in different categories.
Language ability, total television viewing and viewing of
programs by category were then considered in terms of the
mother's education level. Finally, there was an analysis
of two elements of Zajonc's Confluence Model which pre-
dicted relationships between (1) sibling number and
language ability and (2) birth order and language ability.

Language samples were collected according to clinical
procedures outlined by the Developmental Sentence Scoring
Analysis for 93 children aged three years and five months to
five years and eleven months. In addition to language samples,
data on television viewing patterns of subjects was collected
from logbooks maintained for a one week period by parents.
Scripts of frequently viewed programs were then analyzed

according to Developmental Sentence Scoring procedures.

92

General Review of Results

 

The present research involves a variety of investi-
gations, some of which confirm existing theories or expand
the scope of these theories, and others which identify new
sets of relationships. These findings offer a small
contribution to the literature, but they also have practi-
cal implications since the issues deal with very common
elements of television viewing and language development
of young children.

The purpose of this section is to describe these
findings in general terms and outline the relationships
which have been identified in this analysis.

The underlying issue of this investigation is to
determine if television viewing accelerates or retards
language growth. But the present research is designed to
deal only with potential relationships between these two
variables and cannot attend to the causal issue. Nonethe-
less, the findings here represent an essential first step
and lay the groundwork for additional research.

First, this data provides evidence that there is a
relationship between television viewing and language
development of young children. Across all program
categories it was found that children who watched more
than average television tended to achieve lower scores on
the language ability rating. It was then speculated that
the aggregation of all television programs may have masked

any differential influence which exists among various types

93
of programs. In pursuit of this possibility, the analysis
then considered each program category typically viewed by
children including cartoons, educational programs, family
dramas, action dramas and situation comedies. This analy-
sis confirmed the suspicion. It appears that cartoons and
family drama programs had a negative relationship with
children's language ability; as viewing time of cartoons
and family dramas increased, scores on the language ability
rating declined. Readers should note that this does not
imply a causal effect of television on child language
since other interpretations exist to explain this obser-
vation. For instance, perhaps children who already have
less developed language skills can only understand cartoons
and family dramas, so they confine their viewing schedule
to these kinds of programs. Children with more advanced
language abilities may include more sophisticated programs
which they can understand. These findings note a relation-
ship between the various television categories and child
language and while this is certainly an important observa-
tion, it cannot deal with causal statements.

Following this set of investigations the study then
explored the relationship between the mother's education
level (which served as an index of language level and
socio-economic status) and child's language ability.

Guided by sociolinguistic theory, a positive relationship
was predicted. This was, in fact, borne out by the data.

The higher a mother's education level, the higher a child

94
tended to score on the language ability index.

Next, a series of analyses considered the relationship
between variables of television viewing and education level
of the mother. The specific purpose here was to discover
whether or not a mother's education level had any relation-
ship with the kinds of television programs viewed by a
child. Here, too, a pattern became evident. Generally, as
the level of mother's education increased, the child's
total television viewing declined, and among specific
categories of programs, cartoons and family drama programs
also declined. Viewing of educational programs increased
with an increase in mother's education level. It appears
from these results that the mother does influence tele-
vision viewing patterns of a child, and mothers with more
education tend to select for their children programs with
more sophisticated language models.

At this point it became important to determine if
television viewing and language ability of a child were
each related only to mother's education level, but not
related to each other. If this were the case there would
be no strength in the argument that television viewing and
language development are related at all. It was determined
that the relationship between language and television
viewing was not spurious, and, indeed, the relationship
appears to exist.

Finally, as an adjunct portion of this research, this

data was applied to predictions of the Confluence Model

95
which proposes a relationship between (1) the number of
siblings and achievement on intelligence tests, and (2)
birth order and intelligence. In this research, language
ability was substituted for standard intelligence measures
to test the extensiveness of these predictions. Complete
confirmation was obtained. As the number of children in a
child's family increased, his language ability index tended
to decline. Similarly, language ability scores tended to

decline with birth order of a child.

Limitations

 

Nearly all of the conditions established for this
research were conservative in nature and worked against the
discovery of significant findings. Subjects chosen for the
study, data collection methods, analysis of language sam-
ples all contributed to a reduction in the variance among
elements explored in this study, and this leads to a dimi-
nution of impressive statistical results. In other words,
the setup for this research stacked the deck against signi-
ficant findings. Still, a number of significant findings
were obtained. A discussion of the results will be
forthcoming, however, it is first necessary to elaborate on
some of the specific factors which contributed to a con-
strained variance among elements in this study.

One very important consideration is the limitation on
subject selection. All of the children represented in this
sample were recruited from nursery schools and day care

centers located near a major university. Such a sample

96

cannot reasonably be expected to demonstrate characteris-
tics of a general population. Furthermore, to the extent
these children spend their days in a controlled environ-
ment, their schedules do not permit large variations in
their interactions with other children, play activities,
and particularly television viewing. During their stay

at pre-school they all generally watch an equal amount of
television; the only variation in viewing which can occur
among subjects must be at home in the evenings, on weekends
and during days absent from school.

Another unusual characteristic of this sample is the
exceptionally high education level of parents which could
naturally be expected given the proximity of these pre-
schools to a major university. Consequently, differences
which might be anticipated in some variables affected by
SES may not be evident because of such a skewed education
distribution. If subjects were selected from the general
population, a more normal distribution of parent's educa-
tion level would likely be obtained and so more dramatic
findings might be expected.

Another important consideration in the interpretation
of these results is the age of children included in this
sample. This study specifically limited the age range of
subjects (roughly between three-and-a-half and five-and-a-
half years old):haorder to consider children whose language
was sufficiently advanced to respond in an interview

situation, yet not old enough to be heavily influenced by

97

formal language training in school. There is a reduction
in the variance among elements of language here simply
because such a restriction was imposed on the ages of sub—
jects.

Finally, there is concern with the language assess-
ment method chosen for this research. The Developmental
Sentence Scoring procedure is an accurate and reliable
tool in its original setting of clinical speech therapy
where it is used to help identify speech deficient children.
This instrument correlates highly with other testing
procedures and has the confidence of researchers and
speech therapists in situations for which it was designed.
But, the present research extracted the DSS from its
original setting and applied it to the task of analyzing
the language ability of normal children and television
language samples. The DSS was chosen because it most
closely matched the specific needs of this research: it is
a content analysis procedure, it can accomodate a fairly
wide age range and it provides specific information about
grammatical use categories.

Even though the DSS proved to be a very successful
tool for this study, it seems likely that more sensitive
language assessment instruments might be developed for
future research, or a combination of assessment procedures
might be utilized for even finer distinctions among lan-
guage samples.

In summary, a variety of factors have led to the

98

compression of variance in this research and so tended

to work against the display of more dramatic correlation
coefficients among variables investigated here. Conse-
quently, even small correlations with modest statistical
significance must be carefully reviewed for information
about the nature of the relationship between two variables.
Directionality is particularly important and so is the
emergence of patterns for variables considered in a variety
of contexts.

Television Viewing and Child Language

 

The initial portion of this research considered the
general question of whether or not total television viewing
shared any relationship with a child's language ability.
The results demonstrated a negative correlation between
total viewing time and performance on language measures,
leading to the general conclusion that as children in
this sample watched more television their language scores
tended to be lower. While the correlation describing
this relationship is evident and clear, dynamics with
respect to a causal inference are not. It is difficult
to determine from this single investigation whether or
not this negative correlation is a result of opportunity
costs of viewing television, where alternative and perhaps
language building activities became displaced, or whether
the language models in television programs contributed
directly to the lower language performance scores. If the

first explanation accurately describes the relationship

99
between these variables, then, perhaps, the child who
views an abundance of television is deprived of inter—
actions with parents or siblings, or of playing games and
reading books, activities which might engender a better
language system in the child. If the second description
were in operation then television itself might be respon-
sible for contributing to the lower language scores of
heavy viewers by providing deficient language models for
the child. While none of the data collected here can
confirm either position, further analysis in this study
contributes to an explanation.

In order to examine this relationship more closely.
and unfurl the "total television" variable, this study
then turned to separate investigations which considered
five categories of television programs. Here language
ability was reviewed in terms of the numbers of hours
children spent watching cartoons, family dramas, situation
comedies, educational programs and action drama shows.
Significant negative correlations were found between
language ability and viewing of cartoons and family
dramas. There were no significant correlations obtained
for language ability and situation comedies, educational
programs, and action dramas.

In general terms these results contribute to the
observation that as children in this sample viewed a
greater number of hours of cartoon and family drama

programs, their language ability index tended to be lower.

100

Given the sophistication of language models which one
might expect to find in cartoons, for instance, as compared
to educational programs, an immediate interpretation of
these findings might note that television language modeling
contributed to the difference in language ability among
children in this sample. This position holds that heavy
viewers of cartoon programs achieve lower language ability
index scores because language models in cartoons are defi-
cient. Indeed, this may be a reasonable explanation of
interactions among variables, but the data here cannot
confirm such a causal relationship. Alternatively, it is
quite reasonable to propose that children with slower
language development select cartoons more often because
they can understand little more than cartoon language.

At this stage of the analysis, it is only clear which
program categories are negatively correlated and which
ones are positively correlated with the language ability
index.

Finally, in the series of investigations on television
and language development, this study combined information
obtained on viewing patterns of subjects with language
levels for five program categories. This was accomplished
with an arithmetic procedure described in Chapter II which
yielded an index accounting for the exposure times by
category, weighted by the average language sophistication
of programs within that category. This index was then

correlated with language ability indices for each child.

101
It was anticipated that findings from this correlation
analysis would yield information relating to the language
ability of the child and his exposure times to relatively
simplistic or advanced language models.

Results of this analysis demonstrated a positive
correlation which suggested that children with lower lan-
guage indices tended to view more hours of programs lower
on the language scale, and, conversely, children with
higher language indices tended to View more hours of
programs higher on the language scale. Such a finding
reinforces interpretations submitted in previous investi-
gations by adding objective support to the observation
that program categories lower on the language scale have
a negative correlation with child language ability. While
this finding provides added confidence in the previous ex-
planations of the relationship between television and child
language, unfortunately, it does not contribute to the
defense of either causal explanation. This positive
correlation between the exposure-language level index
and child language ability index could be interpreted as
support for the proposition that children with poorer
language skills watch programs with less language
sophistication because they are only capable of under-
standing language of such simplistic shows. Meanwhile,
their language sophisticated counterparts view more
language-sophisticated shows because they understand and

are attracted to these kinds of programs. But, this

102

finding also could be applied to the argument that children
who View more language-sophisticated shows,through exposure,
deve10p a more sophisticated language. And, similarly,
those who concentrate their television viewing on less
language developed programs, are stagnated in language
growth because of exposure to relatively simplistic tele-
vision language models.

While the causal relationship between these variables
of television viewing and language ability remain unclear,
one of the most valuable findings of this research is the
disclosure of such a relationship. This in itself, is an
essential discovery. An interaction exists between these
variables: the language of television and the language of
child viewers are related.

Mother's Education, Television and Child Language

 

In its most basic form, the theory advanced by socio-
linguist Basil Bernstein proposes that children learn
language from parents, and primarily from the mother, and
so a child will tend to speak as his mother speaks. This
position was considered by this research using mother's
education level as an index of mother's language. This
may be a rough measure, but a positive correlation can
reasonably be expected between language sophistication and
education level, and given restrictions on data gathering
in this research, mother's education level served as a
representation of language ability.

In the first analysis of this series, the correlation

103
between mother's education level and the child language
ability index demonstrated a significant positive rela-
tionship. Indeed, children whose mother had higher
education tended to score higher on the language ability
index than those whose mother had less education. This
relationship was expected and adds support to some of the
principal sociolinguistic theories. It also tends to
confirm the comparability of this data with other data
collected under varied circumstances.

The second portion of this series concerned the
relationship between mother's education level and tele-
vision exposure for children in this sample. The first
analysis correlated mother's education with the total
number of hours of television viewing, then subsequent
analysis unfolded this total viewing time and looked at
the mother's education level in terms of program category
type.

In the first instance, a significant negative
correlation was found which suggested that as the mother's
education level increases, the total number of viewing
hours for the child decreases. Although this data does
not attend to the dynamics of this relationship, it
appears reasonable to propose that more educated mothers
encourage their children to engage in alternative activi-
ties in place of a heavy television viewing schedule.
Possibly these activities include more interaction between

parents and child, and this may account for the more

104
developed language of children with higher educated
mothers as demonstrated in the previous analyses. Again,
it is the discovery of the relationship here that bears
attention: mother's education and child exposure to tele-
vision are negatively correlated.

As for the relationship of mother's education level
with individual categories of programs, a recognizable
pattern emerges. There is a clear negative correlation
between mother's education and both cartoon programs and
family drama programs. In these instances, as the mother's
education level increases the child's viewing of cartoons
and family dramas decreases. The other significant re-
lationship discovered in this analysis is a positive
correlation between mother's education level and education-
al programs, suggesting that children of higher educated
mothers tend to watch a greater number of hours of educa—
tional shows.

Here, too, the data does not bear information which
would support a description of the dynamics at play in the
relationships between the mother's education level and the
amount and kinds of programs viewed by children. It would
be necessary to design an entirely different study to deal
with this issue. This data does, however, establish
several relationships not previously reported in the
literature, and, thereby, uncovers information which may
have substantial heuristic value in describing viewership

patterns for young children.

105

The series of analysis so far portrays relationships
between (a) child language and television viewing, (b)
mother's education level and child language, and (c)
mother's education level and child television viewing pat-
terns. Since mother's education level is related both to
child language and television viewing, an obvious question
arises concerning the covariation of child language and
television viewing independent of each other, but related
only to mother's education level. Specifically, it is
important to determine if a spurious relationship exists
between television viewing and child language. This was
investigated through a partial correlation analysis con—
trolling for mother's education, which revealed that, in
general, the relationships between language ability and
the various elements of television viewing were maintained.

Given all the relationships presented so far it would
be helpful to offer a composite sketch and draw all the
variables into a single frame.

It appears that the mother's education level has two
'principal effects. First, it relates positively with the
child's language abilities, and so as sociolinguistic
theories predict, the mother has a direct influence on the
language of her children. It also seems that the mother's
education level has a negative relationship with total
television viewing and certain categories of programs
viewed by the child. Here it appears that the mother has

an influence on the kinds of shows the child views; this

106

might be expected on logical grounds, but is substantiated
here with empirical data. Finally, those television shows
which the child does view have a relationship with the
development of his language. Programs lower on the lan-
guage ability scale tend to be viewed by children who
perform lower on the language index, and programs higher
on the scale tend to be viewed by children who perform
higher on the language index. In summary, of the rela-
tionships among these three variables: a mother influences
the child's language directly, she influences the amount
and kinds of programs he watches, and then these shows,
too, have an effect on the child's language abilities.

Number of Siblings and Birth Order

 

The final portion of this research investigated two
propositions of the Confluence Model developed by Zajonc.
Where all previous research on the Confluence Model
involved standard measures of intelligence, for instance,
the National Merit Scholarship Qualification Tests or
Scholastic Aptitude Tests, this study considered sibling
number and birth order in terms of language ability
measures. The language ability index used here is analog-
ous to standard intelligence quota measures since it
involves test performance adjusted for age.

According to provisions of the Confluence Model, a
child is born into an intellectual milieu described by an
average of the intellectual capacities of all family mem-

bers. As the number of children increases, by the

107
arithmetic of averaging, the intellectual level of that
milieu is reduced (assuming children have lower intelli-
gence levels). From this, Zajonc predicts a negative
relationship between the number of children in a family
and a child's performance on intelligence measures. In
this present study, with the substitution of language
ability for standard intelligence measures, the prediction
was confirmed: language ability shares a negative rela-
tionship with the number of children.

In light of this finding, it appears reasonable to
propose a similar description of the linguistic-intellec-
tual milieu. With a greater number of children, the
average level of language is reduced. Parents must
apportion their attention to a greater number of children
and so each child is the recipient of fewer hours of
direct exposure/interaction with a sophisticated language
model.

A complimentary portion of the Confluence Model
proposes that birth order, too has a negative relationship
with child performance on intelligence measures. In the
present research a similar relationship was discovered with
the measure of language performance. For each child with
siblings ("only" children were excluded from this analysis
according to provisions of the Confluence Model) each
successive child tended to achieve a lower language ability
score. The reasoning follows from a portion of the model

previously reviewed, where each successive child is

.108
introduced into a linguistic-intellectual milieu which
has already been lowered by his preceeding siblings.
Since the quality of the milieu is determined through an
averaging process, it would hold that unless all previous
siblings have reached maturity, later children arrive into
a reduced linguistic-intellectual environment.

In summary, this portion of the analysis supported
predictions of Zajonc's Confluence Model demonstrating that
the number of children and birth order of a child relate
negatively to performance on language measures employed
here. Significance of these findings is recognized in an
extension oftfldntmodel from basic intelligence measures

to language performance measures.

***

Three fundamental issues became the concern of this
research. First, the principal matter and focus of this
study involved the relationship between television viewing
and language ability of young children. Several findings
from this portion of the analysis demonstrate interesting
and noteworthy connections. First, it was discovered that
the total amount of television viewing relates negatively
to language ability. In other words, the more television
a child views, the lower his language ability scores tend
to be. Then, in terms of individual categories of
programs, cartoons and family dramas have a negative

relationship with language ability. A final look at this

109
question involved an index which considered the language
level of program categories and viewing times of each
category. This index was compared to language ability
where, again, there was a signficant positive correlation.
This suggests that as children viewed a greater propor-
tional number of programs lower on the language level
scale, their own measured language ability rating tended
to be lower. The converse holds.

The second issue of concern involved the relationship
between mother's education level and both language ability
and television viewing patterns. In support of predictions
from sociolinguistic theories, there was a positive rela-
tionship between mother's education and child language
ability. As for the correlation between mother's education
and television viewing patterns, total television viewing,
cartoons and family dramas were found to have a negative
relationship and educational programs had a positive
relationship.

In order to assess the relationship between television
viewing and child language with the effects of mother's
education level removed, a partial correlation analysis
revealed that the fundamental relationship remained between
these variables.

Finally, this data was reviewed in terms of the
Confluence Model which suggests relationships between
sibling number and birth order on intellectual performance

measures. In this series of analyses, language ability was

110

substituted for the usual intelligence measures used in
other research. The predictions, nonetheless, were borne
out. It was confirmed that the greater the number of
siblings, the lower the performance on language ability
measures. Also, it was found that as birth order increas-
ed, performance on language ability measures decreased.
Predictions of the Confluence Model were confirmed.

Research Extensions

 

This study represents an initial attempt at uncover-
ing the relationships which exist between language develop—
ment processes and television viewing patterns of young
children. Although the scope was narrow, the findings
were significant and have considerable heuristic value for
expanded research on this very important issue.

As this study has indicated in several discussions,
most of the elements employed here contributed to a con-
strained variance, and so led to conservative results in
most cases. Future research should work to expand these
findings by broadening the scope of study along several
dimensions.

First, it is very important to consider a more
expansive population in the selection of subjects. Where
this study tapped nursery schools and day care centers,
and, thereby, recruited children with similar SES back-
grounds, continued research must draw from a much more
general population and include particularly those children

from lower SES families. Studies have demonstrated that

111
these children typically watch more television, and so
effects of the media might be most dramatic among this
sample. Furthermore, these children who spend their
days at home, and not in pre-schools, will have more
opportunity to develop individual activity schedules
unlike subjects in this study who were regimented by
schedules of the institution. This kind of sample might
be drawn at random from the population (using hospital
birth records to identify families with young children),
or a stratified sample might be considered to assure
sufficient representation of lower SES subjects.

The selection of a language assessment instrument
should be another concern for continued research. Without
question, the Developmental Sentence Scoring procedure met
the needs of the present research and it offered a valuable
tool for identifying language distinctions among subjects.
Particularly as the sample selection expands, and children
from a wider SES spectrum are included, it may be necessary
to utilize a combination of instruments to accurately
assess language samples. Since the age range (three to
five years) is so restricted only a precision tool will be
successful at distinguishing among the variables of lan-
guage and provide the necessary information about language
development.

While there were significant correlations between
language ability and many of the variables considered by

this research, the size of these correlations were not

112

dramatic which suggests the influence of other factors on
language development. Naturally, this is not a startling
observation, the identification of television as an
influential variable is the noteworthy contribution of
this research, the influence of other factors was always
understood. It would be valuable for future research to
consider in a single study a wide variety of factors,
including television viewing, which may have an impact
on language development. This would provide information
about the relative strength of the television influence
in this hierarchy of variables. The literature clearly
lacks a comprehensive review of the variables which impact
on language development. Many studies point to many
factors, but it will require the statistical aggregation
of these potential influences to indicate which are
primary and which are tangential. Furthermore, as media
habits vary among age groups and change over time for the
entire population, it would be valuable to repeat the
study over the course of several years to plot the pro-
files of change. If television does play an important
role in shaping child language, does this change as the
media habits change both for an individual and for an
entire population which shifts its mass communication
activities? This kind of research would provide a major
contribution to the literature.

Finally, this study confirmed the importance of

parental influence, not only in the language development

113
process as sociolinguistic theories predicted, but also
in the selection of television programs viewed by a child.
A variety of politically active groups have called for
increased parental participation in children's television
viewing activities, and it seems reasonable to offer this
kind of research in evidence of the value of such input.
Continued research on the influence of parents in the
television viewing processes of young children would make
a valuable contribution to the literature and provide

pragmatic information for immediate use.

BIBLIOGRAPHY

BIBLIOGRAPHY

 

Arons, L., and May, M. (Eds.). Television and Human
Behavior.New York: Appleton-Century-Crofts
Division of Meredily Publishing Co., 1961.

 

Baker, R., and Ball, S. (Eds.). Mass Media and Violence.
Staff report to the National Commission on the
Causes and Prevention of Violence. Washington,
D.C.: U.S. Government Printing Office, 1969.

 

Ball, 8., and Bogatz, G.A. The First Year of Sesame
Street: An Evaluation. Princeton, N.J.: Educa-
tional Testing Service, 1970.

 

 

Bandura, A. "A Social-learning theory of identification
processes," in D.A. Goslin (Ed.), Handbook of
Socialization Theory and Research. Chicago:
Rand McNally, 1969. Pp.213-262.

 

Bandura, A. "Influence of model's reinforcement contin-
gencies on the acquisition of imitative processes."
Journal of Personality and Social Psychology,
1965(A), 1, 589-595.

Bandura, A. "Vicarious processes: a case of no-trial
learning," in L. Berkowitz (Ed.), Advances in
Experimental Social Psychology. New York: Academic
Press, 1965(B), 2. Pp.l-55.

 

Bandura, A. "Imitation of film mediated aggressive models."
Journal of Abnormal and Social Psychology, 1963,
66, 3-11.

Bandura, A. "Social learning through imitation," in M.R.
Jones (Ed.), Nebraska Symposium on Motivation.
Lincoln: University of Nebraska Press, 1962.
Pp.211-274.

Bandura, A., and Menlove, F. "Factors determining vicarious
extinction of avoidance behavior through symbolic
modeling." Journal of Personality and Social‘Psye
chology, 1965, 8, 99-108.

114

115

Bibliography (cont'd.)

Bee, H., Van Egern, L., Streissguth, H., Numan, B., and
Leckie, M. "Social class differences in maternal
teaching strategies and speech patterns."
Developmental PsyChology, 1969, 1, 726-734.

 

Berkowitz, L. (Ed.) AdvanCeS'in Experimental Social
Psychology. New York: Academic Press, 1965, 2.
Pp.l-55.

 

 

Bernstein, B. "A Sociolingusitic approach to socialization
with some reference to educability," in J. Gumperz
(Ed.), Directions in Socio-linguistics. New York:
Holt, Rinehart and Winston, Inc., 1972.

 

Bernstein, B. "Elaborated and restricted codes: an outline."
Sociological Inquiry, 1966, 2.

 

Bernstein, B. "Some sociological determinants of perception."
British Journal of Sociology, 1958.

 

Birren, J. and R. Hess. "Influences of biological, psycho-
logical and social deprivations on learning and
performance," in Perspectives on Human Deprivation.
Department of Health, Education and Welfare.
Washington, D.C.: Government Printing Office, 1968.

 

Braine, M. "On learning the grammatical order of words."
Psychological Review, 1963, 70, 323-348.

 

Brophy, J. "Mothers as teacher of their own pre-school
children: the influence of socioeconomic status and
task structure on teaching specificity." Child
Development, 1970, 41, 79-94.

 

Brown, R. A First Language: The Early Stages. Cambridge,
Mass: Harvard University Press, 1973(A).

Brown, R. ngcholinguistics. New York: Free Press, 1970.

 

Brown, R. and Bellugi, U. "Three processes in the child's
acquisition of syntax." Harvard Educational Review,
1964, 34, 133-151.

 

Brown, R., Cazden, C. and Bellugi, U. "The child's grammar
from I to III," in J.P. Hill (Ed.), Minnesota
Symposium on Child Psychology. MinneapoIis: Uni-
versity of Minnesota Press, 1969, 28-73.

 

Bryan, J.H. "Model effect and children's imitative altruism."
Child Development, 1971, 42, 2061-2065.

116

Bibliography (cont'd.)

Bryan, J.H., and Welbek, N.H. "The impact of words and deeds
concerning altruism upon children." Child Develop—
ment, 1970, 41, 747-757.

 

Carmichael, L. (Ed.). Manual of Child Psychology, 1954.

 

Cassirer, H. "TV for the pre-school child." The UNESCO
Courier, 1966, November, 4-11.

 

Catton, W., Jr. "Value modification by mass media," in R.
Baker and S. Ball (Eds.), Mass Media and Violence,
Staff Report to the National Commission on the
Causes and Prevention of Violence, Washington, D.C.:
U.S. Government Printing Office, 1969.

 

Cazden, C. Child Language and Education. New York: Holt,
Rinehart and Winsotn, 1972.

 

Cazden, C. "The neglected situation in child language
research and education," in F. Williams (Ed.),
Language and Poverty. Chicago: Markham, 1970.
Pp.81-101.

 

Chandler, B. and Erickson, F. "Sound of society: a demon-
stration in group inquiry." Final Report #6-2044.
Washington, D.C.: U.S. Government Printing Office,
1968.

 

Chomsky, N. Aspects of the Theory of Syntax. Cambridge,
Mass: M.I.T. Press, 1965.

 

Christenson, E. "Television: tiger by the tail," in S.
Sunderlin (Ed.), Children and T.V.: Television's
Impact on the Child. Washington, D.C., 1969.

 

 

Coffin, T.E. "Television's impact on society." American
Psychologist, 1955, 10, 630-641.

 

Comstock, G.A. and Rubinstein, E.A. (Eds.), Televisionyand
Social Behavior. Washington, D.C.: U.S. Government
Printing Office, 1972, 2. Pp.l36-l40.

 

Cook, T.D., Appleton, H., Conner, R., Shaffer, A., Tamkin,
G., and Weber, S.J. Sesame Street Revisited: A
Study in Evaluation Research. New York: Russel
Sage Foundation, 1975.

 

 

Davies, A. (Ed.). Problems of Language and Learning. London:
Heinemann EducatIonal Books, 1975.

 

Dale, P. Language DevelOpment: Structure and Function. New
York: Holt, Rinehart and Winston, 1976.

 

117

Bibliography (cont'd.)

Daler A- G. Mothers: Their Power and InfluenCe. London:-
Weidenfield and Nicholson, 1976. -.1

 

DeFleur, M.L. Theories of Mass Communication, 2nd ed. New
York: David McKay, 1970.

 

DeFleur, M.L., and DeFleur, L.B., "The relative contribution
of television as a learning source for children's
occupational knoweldge." American Sociological
Review, 1967, 32, 777—789.

 

Dunn, L. Peabody Picture Vocabularlerest. Minneapolis:
American Guidance ServIce, 1965.

 

Entwisle, D. "Form class and children's work associations."
‘Journal of Verbal and Learning and Verbal Behavior,
1966, 5, 558-565.

Ferguson, C.A. and Slobin, D.C. (Eds.). Studies of Child
Language Develppment. New York: Holt, Rinehart and
Winston, I973.

 

 

Feshback, S. "The effects of aggressive content in televi-
sion programs upon the aggressive behavior of the
audience," in L. Arons and M. May (Eds.), Televi-
sion and Human Behavior. New York: Appleton-Cen—
tury-Crofts Division of Meredily Pub.Co., 1961.

 

Fisher, H. and Logemann, J. The Fisher-Logemann Test of
Articulation Competence. Boston: Houghton-M Ifflin,
1971.

 

 

Fox, S.E. "Assisting children's development." The Reading
Teacher, 1976, 29, No.7, 666-670.

 

Francis, H. Language in Childhood. New York: St. Martin's
Press, 1973.

 

Frederich, L.K., and Stein, A.H. "Prosocial television and
young children: the effects of verbal labeling and
role playing on learning and behavior." Child
Development, 1975.

 

Frye, Karen H. Television's Sesamegtreet: An Experiment in
Early Education. Los Angeles: UCLA Center for Afro-
American Studies. CAA Monograph No.3, 1972.

 

 

Gerbner, G. "Toward cultural indicators: the analysis of
mass mediated public message systems," in Gerbner,
et al. (Eds.), The Analysis of Communication Con-
tent: Developments In Scientific Theories and Compu-
ter Techniques. New York: Wiley, 1969, Pp. 123-232.

 

 

 

118
Bibliography (cont'd.)

Gerbner, G. "An institutional approach to mass communication
research, " in Thayer, L. (Ed.), Communication:
Theory_and Research. Springfield, 111: Charles C.
Thomas, 1967. Pp. 429-451.

 

 

Gerbner, G., Holsti, O. Krippendorff, K., Paisley, W., and
Stone, P. (Eds.). The Analysis of Communication
Content: Developments in Scientific Theories and
Computer Techniques. New York: Wiley, 1969.

 

 

 

Golub, L.S. "A critical age model of language learning."
Learning Arts, 1975, 52, No.6, 1097-1103.

 

Gumperz, J. (Eda).Directions in Sociolinguistics. New York:
Holt, Rinehart and Winston, Inc., 1972.

Halliday, M. "Talking one's way in: a sociolinguistic per-
spective on language and learning, " in A. Davies
(Ed.), Problems of Language and Learning. London:
Heinemann Educational Books, 1975. Pp.8-28.

 

Hess, R. and Shipman, V. "Early experience and the sociali-
zation of cognitive modes in children." Child
Develgpment, 1965, 36, 869-886.

 

Hill, J.P. (Ed.). Minnesota Symposium on Child Psychology.
Minneapolis: University of Minnesota Press, 1969, 2.

 

Holloran, J.D. "Television and Violence," in O.N. Larsen
(Ed.), Violence and the mass media. New York:
Harper and Row, 1967. Pp.l39-151.

 

Hymes, D. "Models of the interaction of languages and social
setting." Journal of Social Issues, 1967, 23, 8—30.

 

Jenkins, J. and Palermo, D. "Mediation processes and the
acquisition of linguistic structures," in U. Bellugi
and R. Brown (Eds.), The Acquisition of Language.
Monograph of the Society for Research in Child
Development, 1964, 29.

 

Johnson, D. "The influences of social class and race on
language test performance and spontaneous speech of
pre-school children." Child Development, 1974, 45,
517-521.

 

Kaplan, B.J. "Malnutrition and mental deficiency." Psycho-
‘logical Bulletin, 1972, 78, 321-334.

 

119
Bibliography (cont'd)

Kenny, D.C. "Threats to the internal validity of cross-
lagged panel inference, as related to television
violence and child aggression: a follow-up study,"
in G.A. Comstock and E.A. Rubinstein (Eds.),
Television and Social Behavior. Washington, D.C.:
U.S. Government Printing Office, 1972, 3, 136-140.

Krugman, H.E., and Harley, E.L. "Passive learning from
television." Public Opinion Quarterly, 1970, 34,
184-190.

 

Labov, W. The Study of Nonstandard English, published by the
NatIonal Council of Teachers of English, 1970.

Lang, G.E. and Lang, K. "The unique perspective of tele-
vision: a pilot study." American Sociological
Review. 18, 1953, 3-12.

 

Lawton, D. Social Class, Language and Education. London:
Rutledge and Keen Paul, 1968.

Lenneberg, E.H.‘Whlexplaining language". Science, 1969,
64, No.3880, 635-643.

Lenneberg, E.H., Rebelsky, F., and Nichols, I., "The
vocalization of infants born to deaf and hearing
parents." Vita Humana, 1965, 8, 23-37.

 

Lesser, G.S. Children and Television: Lessons from Sesame
Street. New York: Random House, 1974.

Lewis, M. and Rosenbaum, L. (Eds.). Interaction! Conversa-
tion, and the Develgpment of Language. New York:
John Wiley and Sons, 1977.

 

McCarthy, D. "Language development in children," in L.
Carmichael (Ed.), Manual of Child Psychology, 1954.

McCarthy, J., and Kirk, S. Examiner's Manual: Illinois Test
of Psycholinguistic Abilities. Champaign, Ill:
UniverSity of Illinois Press, 1961.

McNeill, D. The Acquisition of Languagey A Study of Develop-
mental PsychOlipguistic Abilities. Champaign, Ill:
University of Illinois Press, 1961.

Mecham, M. Verbal Language Development Scale, Manual. Circle

Pinces, Minnesota: American Guidance Service, Inc.,
1971.

120
Bibliography (cont'd)

Medinnus, G.R. and Rohannon, K. "An exploratory study of
some psycholingusitic correlates of maternal
language." Journal of'Genetic Psycholggy, 1975,
126, 2, 263-268.

 

Milkovich, m., Miller, M. with Bettinghaus, E., and Atkin,
C. Exploring the Relationship between Television
Viewing and Language Develgpment. Paper submitted
to the Office of Child Development, Department of
Health, Education and Welfare, 1975.

 

 

Morley, M. The Development and Disorders of Speech in
Childhood. London: Livingstone, 1957.

 

 

Mowrer, 0. Learning Theory and the Symbolic Processes.
New York: Wiley, 1960.

 

Nelson, K. "Structure and strategy in learning to talk."
Monographs of the Society for Research in Child
Development, 1973, 38, No.149.

 

 

Osser, H. "Biological and social factors in language
development," in S. Rogers (Ed.), Children and
Language. London: Oxford University Press, 1975.
Pp. 289-308.

 

Osser, H., Wang, M.D., and Zaid, F. "The young child's
ability to imitate and comprehend speech: a com-
parison of two subcultural groups." Child
Development, 1969, 40, 1063-1075.

 

Paraskvopoulos, J., and Kirk, S. The Development and
Psychometric CharacteristiCs of the Revised Illinois
Test of Psycholinguistic Abilities. Urbana: Univer-
sity of Illinois Press, 1969.

 

 

Peterson, D. "A sociolinguistic study of elaborated and
restricted code systems. ” Unpublished Ph.D. Thesis,
Michigan State University, 1970.

Piaget, Jean. The Language and Thought of the Child. London:
Routledge and Kegan Paul, Ltd., latest publishing,
1955.

Pozner, J. and Saltz, S. "Social class, conditional
communication and egocentric Speech." Develgpmental
Psychology, 1974, 10, 764—771.

 

 

121
Bibliography (cont'd.)

Roberts, D.F., and Schramm, W. "Children's learning from the
mass media," in W. Schramm and D.F. Roberts (Eds.),
The Process and Effects of Mass Communications.
Urbana, University of Illinois Press, 1971.

 

Rogers, 8. Children and Language. London: Oxford University
Press, 1975.

 

Schramm, W., and Roberts, D.C. (Eds.). The Process and
Effects of Mass Communication. Urbana: University
of Illinois Press, 1971.

 

 

Schramm, W., Lyle, J., and Parker, E. Television and the
Lives of Our Children. Stanford, Californis:
Stanford University Press, 1961.

 

 

Selnow, G.W. "Survey of media use patterns in Saginaw County,
conducted for 'International Export Project'," U.S.
Department of Agriculture, 1977.

Siegal, A. "The effects of media violence on social learn-
ing," in R. Baker and S. Ball (Eds.), Mass Media and
Violence, a Staff Report to the National Commission
on the Causes and Prevention of Violence. Washington,
D.C.: U.S. Government Printing Offic, 1969.

 

Singer, Jerome L. The Child's World of Make-Believe. New
York: Academic Press, 1973.

Skinner, B. Verbal Behavior. New York: Appleton-Century-
Crofts, 1957.

 

Slobin, D.C. (Ed.). The Ontogenesis of Grammar. New York:
Academic Press, 1971.

 

Smith, A. Communication and Culture. New York: Holt, Rine-
hart and Winston, 1966.

 

Steward, M. and Steward, D. "The observation of Anglo-
Mexican and Chinese-American mothers teaching their
young sons." Child Development, 1973, 44, 329-337.

Sunderlin, S. (Ed.). Children and T.V.: Television's Impact
on the Child. Washington, D.C. 1969.

 

Thayer, L. (Ed.) Communication: Theory and Research.
Springfield, Illinois: Charles C. Thomas, 1967.

 

Tulkin, S.R. "An analysis of the concept of cultural
deprivation." Developmental Psychology, 6, 1972,
326-339.

 

122
Bibliography (cont'd.)
Vygotsky, L.S. Thought and Language. Translated by E.

Hanfmann and G. Vakar. Cambridge, Mass.: M.I.T.
Press, 1962.

 

Williams, F. (Ed.). Language and Poverty. Chicago: Markham,
1970.

 

Williams, F., and Narmore, R.C. "On the functional analysis
of social class differences in modes of speech."
Speech Monographs, 36, 77-102, 1969.

Winn, M. The Plug-In Drug. New York: The Viking Press, 1977.

 

Zajonc, R.B. "Family configurationanuiintelligence." Science,
April, 1976, 192, 227-236.

Zajonc, R.B., and Markus, G.B. "Birth order and intellectual
development." Psychological ReviewaBZ (1), 1975,
74-88.

Zigler, E., and Butterfield, E. "Motivational aspects of
change in IQ test performance of culturally deprived
nursery school children." Child Development, 1968,
39, 1-14.

 

Zimmerman, R., Steinger, V., and Evatt, R. Preschool Lan-
guage Scale. Columbus, Ohio: Merrill PubliShifig
Company, 1969.

 

 

APPENDICES

APPENDIX I

DETAILS OF THE DEVELOPMENTAL
SENTENCE SCORING ANALYSIS

APPENDIX I

DETAILS OF THE DEVELOPMENTAL
SENTENCE SCORING ANALYSIS

The Developmental Sentence Scoring procedure is a
language analysis instrument born of many developmental
language theories and is an eclectic formulation derived
from the combined elements of many existing language
assessment techniques. In the development of a rationale
fortjޣ;procedure, Lee has drawn liberally from many
sources including Chomsky, McNeill, Fillmore, Brown, Fries
and a host of others. The technique rests firmly on the
prepared ground of language theory from prior research.

Reliability and validity of DSS procedures were
reviewed by Koenigsknecht in the basic text outlining pro-
cedures for this instrument. A portion of the summary is
presented here.

(From: Koenigsknecht, R. "Statistical Information

on Developmental Sentence Analysis," in Lee, L.

Developmental Sentence Analysis, Evanston: North—
western University Press, I974. Pp. 222-268).

 

The results of this series of investigations of the
validity and reliability of the DSS procedures and its
reweighted scoring system strongly support its usefulness
as an objective measure of syntax development in children.

The validity of the DSS measure was indicated by the
123

124

significant differences produced between age groups by

the over—all scoring procedure and by each of its component
grammatical categories. As age levels increased, the
scores which purported to measure spontaneous syntax and
morphology usage increased systematically, indicating that
grammatical development was being measured. All nine of
the component categories of the DSS, eight involving
specific grammatical structures and one the over-all sen-
tence point, were useful in discriminating between adjacent
age groupings. The discriminating power of the categories
varied. The rank order from the most to the least discri-
minating was (1) Main Verbs, (2) Conjunctions, (3) Indefi-
nite Pronouns-Noun Modifiers, (4) Personal Pronouns, (5)
Secondary Verbs, (6) Negatives, (7) Sentence Points, (b)
Wh- questions, and (9) Interrogative Reversals. Evidence
was presented which showed that the DSS successfully
measures the impact of not only the developmental level or
maturity of syntax usage but also the typical length of a
child's utterances.

The validity of the DSS construct was further indicated
by the confirmation of the grammatical hierarchies in the
reciprocal averaging procedure itself. The stability and
accuracy of the DSS scoring system was enhanced with this
analysis through the reweighting of items within grammatical
categories.

Major support for the usefulness of the scoring system

was supplied by the high reliability coefficients obtained

125

by this measure. The over-all internal consistency of the
instrument as measured by Coefficient Alpha was 0.71. The
within-subject internal consistency of the scoring proce—
dure was assessed in a split—half procedure in which odd
items and even items were combined and then correlated.
This resulted in an over-all estimate of reliability of
0.73 which indicated good stability of the scoring proce-
dure within subjects on a 50-utterance sample. Each of the
individual grammatical categories was positively correlated
with the over-all DSS scores, indicating that each of the
grammatical categories is measuring increases in oral
language development and contributing to the total DSS
score. Item analyses for each of the grammatical categories
revealed the usefulness of each category at specific age
levels.

The discriminating power of individual categories
within the overall DSS instrument for sorting among the
actual ages of children was presented in the summary above.
A similar analysis was conducted with the present data to
assess the discriminating power of the nine categories in
terms of language ability index scores described in
Chapter II. The following order was observed: (1) Main
Verbs, (2) Conjunctions, (3) Personal Pronouns, (4) Indefi-
nite Pronouns, (5) Sentence Points, (6) Negatives, (7)
Secondary Verbs, (8) Interrogatives, (9) Wh- Questions.
This rank order was determined through the Strength of
Pearson Correlations between individual categories and

the Language Ability Index.

126

Developmental Sentence Scoring

Weighted Scores

vub:

I:
ing: He 1:

past: 3!! (VIM In u:
no (pun. n.

Auxiliary cm. are. me
ms. were

0'! [on 19]

my
be. any
4 verb:
* 2n:

I‘ve (a! it.‘

 

(From: Lee, L.L. Developmental Sentence Analysis.
Evanston: Northwestern ”niv. Press. 1974.)

127

Developmental Sentence Scorin
Weighted Scores

 

*—

 

 

(con t ' u‘Y
. IN'I‘ERROGATIVE I
NECATIVES (“ONJUNPTIONS REVERSALS . VIN-QUESTIONS
it. this. that + copula or Reversal of co ula:
auxiliary is, ’s. + not: Isn‘t it red? ere they
We not mine. there?

is is not a dog.
at is not MOVII'I‘.

 

A. who. what. what f noun:
0 am I? What is be
eating? What book are
you reading? _ '
B. where. how many. how
am .wiiai . . .do.
at... or
Where did it so?
aw much do you want?
What is he doing?
Who! is a hammer [or.’

 

 

 

 

 

 

 

can‘t. don‘t Reversal ofi auxiliary be:
I: he commg? Isn't_he
coming! War he a30mg?
lilan’ he going.
ian’t. won't A. but when. how how + adjective
3. to. find so.» that When I come?
C. or, How do you do it?
flow by is it?
heeauaa A. Obligtt’o do. does.
did; I W run? Does
it into? Didn’t it hurt?
I. :Beversal oi modal: _
1.0! you PIJ '1’. Won't it
hurt. Shall sit down?
' . i8 filmmfli.
t' fun on r it?
t an't fund's“! ,
loiher negatives: why.wlii if. how come
UnconiraCied negatives: how about + gerund
lean not go. Why are you crying?
He has nor zone. What if I won‘t do it?
Pronoun-auxrliary or How come he i:_crying?
pronounycopula flowchart coining with me?
KPHUICUOI'IZ

In; not coming.
lie snot here.
Auxiliary-negative or
«pub-negative
contraction: .

ing.

lie hasn't een seen.
It couldn't be mine.

aren't big.

 

A. where. when. how.
while.whethcr (or not).
till until unlexs. Since.
before. after. for. as as
+ adjective + as. as i .
like. that. than
I know where on are.
Don't come ri (cull. .

I. Obligatory deletions:

run faster than you

n .
abiguaman [is
' like d
‘03:}? a o;
C. Elliptical deletions
sco l'

[c .
al's why ll took it].
I know how [I can do

D. '«wordi + infinitive:
I know how to do it.

 

 

know where to no.

A. Reversal of auxiliary
have'

Ila: he seen ou?

B. Reversal W" i two or
three auxiliaries;
Has he been eating?
Could n't he have
waited?

Could he have been

cryin .’
Woulfin‘i he have been
‘OIIII?

whose. which. which 9 noun
Whose car is that?
Which book do you want?

 

APPENDIX II

CORRESPONDENCE AND FORMS FOR
PARENTS OF SUBJECTS

l"

128

Letter of Introduction to Parents

 

October 7, 1977

Dear Parents,

During the next two months I will be working with children through
various nursery schools and day care centers in the East Lansing
area on a project which will look at the effects of television
viewing on young children. As you probably know, the impact of
television on children has received considerable attention by-
educators, parent's groups and the U.S. government in recent days,
and I feel this study will be a valuable addition to this general
body of information. I have been in contact with Karen Ritts,
Director of the Eastminster Day Care Center and she has reviewed'
our project and approved the procedures we have proposed.

Specifically, we would like to ask two things from the children

and their families. First we would request that parents complete

a ”Television Viewing Log" for their 4/5 year old child enrolled at
Eastminster Day Care Center. Essentially, this will provide infor-
mation about television viewing by your child during a typical
7—day week. We have enclosed a Television Viewing Log form and

set of instructions.

Second, we would ask that each child be interviewed by graduate
students from MSU for 15 minutes during one class day. These students
will record the child's conversation and later review it for various
language patterns. There will be only one talk session with each
child and it will be conducted at the school.

I would greatly appreciate your cooperation in this project and ask
that you please complete the Television Viewing Log which we.have
enclosed and mail it back in the envelope provided. -Naturally, I
will be glad to answer any questions which you may have. '

Sincerely,

Gary w. Selnow

Graduate student in Communication
Michigan State University
355-9578

1J29

INSTRUCTIONS FOR COMPLETION OF THE TELEVISION VIEWING LOG

Your entries on this Television Viewing Log will provide basic information
about the kinds of television shows your child watches during a typical seven
day week. He would ask that you select a convenient week in September or early
October and list programs viewed by your child next to the appropriate time
slot for each day of that week. If, for some reason, you are unable to determine
the name of the program, would you please indicate which channel your child is
watching or, as a minimum indication, please mark the appropriate time slot
with a check. If at all possible, would you make entries in this log throughout
the day as your child watches television. If this is not possible, would you
record this information at the end of each day while your memory is still fresh.

Here is an example of how your entries might look.

SATURDAY (Date: 829* 4’4 )

 

Morning Program .
_ V This indicates that the child watched
6:00-6:30 tv, but parent is not sure of the show

 

or channel.
5:30-7:00 CAR-room l-loen

7:00-7:30 LITTLE RQSCQLS

7:30-8:00 HR “.3515qu

. . __.. A dash or blank indicates child did not
8'00'8'30 "' watch tv during this time.

 

How to Return Forms _ .
When you have completed the "Television Viewing Log" for a seven day period,

please send us the form along with the Background Information form in the enclosed

self addressed envelOpe. If this envel0pe should become misplaced, please send

‘ these forms to: Gary Selnow, Ext. Ag. & Mktg., 29 Chittenden Hall, Michigan State

University, East Lansing, MI 48824.

13I)

[Example of Single day from Television Viewing Lug

 

8 UN DAY “)6 te 3 )

MORNING PROGRAM AFTERNOON CONTINUED PROGRAM

 

6:00-6:30

 

6:30-7:00

 

7:00-7:30

 

7:30-8:00

 

8:00-8:30

 

8:30-9:00

 

9:00-9:30

9:30-10:00
10:00-10:30
10:30-11:00
11:00-11:30
11:30-12:00

W
12:00-12:30

12:30-1:00

 

 

 

 

 

1:00-1:30

 

1:30-2:00

3:00-3:30

3:30-4:00'

4:00-4:30

4:30-5:00

5:00-5:30

5:30-6:00

EVENING‘

.6:00-6:30

6:30-7:00

7:00-7:30

7:30-8:00

8:00-8:30

 

 

 

 

 

 

 

 

 

 

8:30-9:00

 

AFTER 0:00 p.m.

 

 

 

 

2:00-2:30

 

2:30-3:00

 

 

 

131

Instructions for Completion of
the Background Information Form

On this form we are requesting some information which
is important to the basic design of this study. As you will
note, due to the nature of some information requested, we
will treat your responses with complete confidentiality.
This information is important and we would appreciate your

responses to all seven questions.

1332

 

 

C 0 N F I D E N'T I A L
Background Information Form

Information on this form will remain confidential and will be used only to

provide background statistics for purposes of this study. When all the data has
been collected. identification of the reSpondent will be eliminated from the
records for all analysis. '

. _Name of child

 

Child's age: Years . Months

Number of children in family

 

Position of this child to brothers and sisters.

Oldest -
Between oldest and youngest
Youngest

No brothers or sisters

 

How would you rate the ability level of your child compared to other children
of the same age?

Hell above average
Above average
About average
Below average
Nell below average

Grammer school Grammer school

Some high school , Some high school
Graduated high school , Graduated high school-
- Some college . Some college

Graduated college Graduated college

Some graduate work Some graduate work
Masters degree Masters degree
Doctorate Doctorate

 

 

would you please indicate with which parent(s) the child is presently living.

Mother only
Father only
Both parents '
Other

 

C O N F I D E N T I A L

 

 

133

Form: Request for Research Abstract

 

Your help is greatly appreciated! If you would
like to receive the results of this study, please fill
out your name and address below, and send this along with

the other forms.

 

name

 

street

 

city state zip

"lllllfljllllili'fllllfi