"H 1
in”?
LJ

3'}?
5251:

o 1“! to .- ' .A. _

>5 i— l ”2“! ; “2119’;

'r .i A." "mifi‘:

5 [”5513‘9211'1521135x' 5

'JJI‘Q, . "~23

.3011' h! «11' 3- . 3.6.: h ‘6,

mg 11.15; ii! "'5‘3‘55'214'53't.

i’z' ‘1- “Ti '5 ‘

15‘: “‘3 inf" if “Ju'ugéj .
' ii :1} LIA”

.
'l
'I

“£323.61,
I
WE.

:-
..
. .
"\‘I
. .

37'
'4

T--.
.45..
fig

”05
.¢ w
n

. . 0“
.‘.'.
. .
u .

Ml ‘i I
. 1“ mm; ‘ Mg . . (5.5.1' u,
1;” .511:;- 15"‘gfiqmu1' “Wm '1‘;,:';'91_‘:v§1:;1 ‘
11F,“ ‘1'“ NE. W IRE hhfi $505111 u. ‘ 1"} Eflfl.
53.51'35‘a'fk3'151 '1] f) 22‘1"”)a'1-1fl :E' a”
'5' "'= 5 'mt ... Jug:
[311” '15 fth! 43.;

.l:: . 1:185?"

m--.“
_' 43)., .
- K .

.. u 4"
.2:
r‘.

.
_.- u o
.‘n‘.

.‘W

-
1'...
la

EMA?
[o

a.
a
'7"

4 -_ .
- .4 - u .
“ff.

.4.

:EW.

3 :1“ - 11'-
" 5'5'*.:’F'Jl:1"

Y: 5 .ill‘" 0‘, I
l "“ I "I
1 I

0L . ‘3" N . 4 ‘1‘
:LM'iiu'u fié‘“; '1"1ff;¥:il‘!i t ’
131%,;‘3: “hr" ‘ {“II'HQ
, ‘ ‘Qfifiliflh '1: xgrl'la .

I?! 4"2'.’ ‘ 1.3:}; lzfifii fig‘
5‘1’§1"fi1"'15‘:1‘,"':

Ega.

3:01:

‘ I
‘n! l .,

a
t ..
..
x "
u.
.
.
\

I
. :53
\MR

.
"‘ .. ..
n .>
. .

4-.

p.

' Q'W‘VA .

0. -_e

mréfz': -
....

-—\~

‘- ~ _
b. . .
-‘ V. .
..._4 m". -—-
"4-
.~ ‘_._
u.

v
LI
-9 :1. “-5-,

_...,
5:.
..

 

5. .32,”
'1'15'5‘ Na-
1. .' :1' 11.6"] “Him!“V

~._._4v
- ~ ..¢-—~

 

§_r if . 77v ,,ir..#_ __ . .__.___

llllllllllllllllill!llllllllUlHlllllllllllllllllll

321293 01716 5279

This is to certify that the

thesis entitled

CHILDREN'S STORY RETELLING AS AN ORAL LANGUAGE SAMPLING
TECHNIQUE IN DIFFERENT MODALITY CONDITIONS

presented by

JAMIE CHRISTINE GAZELLA

has been accepted towards fulfillment
of the requirements for

_MASIERS_QF_degree in AUDIOLOGY AND SPEECH
ART SCIENCES

Mfiw

Major professor

 

Date [z/lslly;

0.7539 MS U is an Affirmative Action/Equal Opportunity Institution

 

l

 

LIBRARY

Michigan State
Unlverslty

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

MTE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1/99 woummu

CHILDREN’S STORY RETELLING IN DIFFERENT MODALITY CONDITIONS
By

Jamie Christine Gazella

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Audiology and Speech Sciences

1997

ABSTRACT
CHILDREN’S STORY RETELLING IN DIFFERENT MODALITY CONDITIONS
BY

Jamie Christine Gazella

This study explored story retelling' as a ‘procedure for
standardizing language outcomes. Its purpose was to determine
whether the child's linguistic response to the story retelling
task was differentially influenced by the presentation
modality (audio alone or audiovisual). This independent group
design was conducted on twenty-nine white, English-only
speaking males aged 4;3 to S;6 years. It compared global
measures of language to see which modality caused children to
produce the most complex language. Results indicated that
there were no statistical significance between the two
modality presentations in the areas of language analyzed--
quantity of language (total number of utterances, total number
of words, and Mean Length Utterance), lexical quality of
language (Type Token Ratio and total number of different
. words), and syntactic qmality of language (sentence type,

completeness, and complexity).

This is dedicated to Richard L. Gazella. Grandpa, I “hung in there!”

iii

ACKNOWLEDGMENTS

This thesis could never have been completed without the assistance of many important people. First,
I would like to thank Dr. Ida Stockrnan, my faculty thesis advisor. Words cannot express my appreciation for
all of her guidance these past years. She is such an inspiration and role model for me. I would also like to
thank my thesis committee members, Dr. Paul Cooke and Dr. Alice Whiren. Both of them dedicated so much
time and energy toward the success of my thesis.

I would also like to give credit to my assistants Melissa McGowen and Hien Van Luu. Without their
loyalty and commitment, I would never have completed this project. A special acknowledgment to Lori Henry
who came in and saved the day with her kindness. Thanks also to Gina DiNatale and Don Coon for their artistic
abilities and undying determination to assist me with the technical aspects of this project.

In addition, I would like to thank the faculty, students, parents, and staff at Michigan State
University’s Child Development Laboratory. Working with these individuals was very insightful and
pleasurable. I also would like to thank the Herbert J. Oyer Speech-Language-Hearing Clinic for its use of an
audiometer during the preliminary hearing screenings of the children in the study.

Furthermore, I would like to offer words of gratitude to my friends for their undying faith in me.
They were my cheerleaders throughout this endeavor. From the A-Team to the 410 Grove Girls, they are all
the best friends that a person could ever ask for in their life!

Last, but certainly NOT least, I would like to recognize my family. They are my rock and foundation,
ALWAYS! Thank you for your unconditional love and patience. Timmy and friends came “alive” because of
all of you! I love you always!

This work was made possible by a generous grant and funding from the Kappa Kappa Gamma

Foundation and Dr. Ida Stockman. Thank you also to the Psychological Corporation for their generosity.

iv

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. ix
LIST OF FIGURES ............................................................................................................ xi
CHAPTER 1
BACKGROUND INTRODUCTION .................................................................................. 1
Oral Language Sampling ..................................................................................................... 1
Benefits of Language Sampling ............................................................................... 2
Limitations of Language Sampling .......................................................................... 3
Story Telling ........................................................................................................................ 6
Story Generation ...................................................................................................... 7
Benefits of Story Generation ....................................................................... 8
Limitations of Story Generation .................................................................. 8
Story Retelling ........................................................................................................ 9
Benefits of Story Retelling .......................................................................... 9
Control of language forms ................................................... 9
Familiarity of the task .................................................................... 10
Reduction of transcription time .................................................... 10
Predictor of academic success ........................................................ 10
Technical standardization .......................................................................... 10
Limitations of Story Retelling ................................................................... 11
Naivety of the listener .................................................................... 11
Lack of pragmatic information provided ....................................... 12
Interest of subjects ......................................................................... 12
Lack of standardization .................................................................. 12
Story Generation Versus Story Retelling ............................................................... 13
Statement of the Problem ................................................................................................... 15
Modality of Input ................................................................................................... 15
Visual only ................................................................................................. 15
Audio only ................................................................................................. 16
Audiovisual ................................................................................................ 16
Comparative Effects of Different Modality Conditions ............................ 17
Limitations of existing research ................................................................. 19
Purpose ............................................................................................................................... 21

CHAPTER 2

METHODS ........................................................................................................................ 22
Subjects .................................................................................................................. 22
Description of Subjects .............................................................................. 22

Subject Selection Criteria .......................................................................... 24

Subject Selection Procedures .................................................................... 24

Gaining subject access ................................................................... 24
Speech and hearing screening ........................................................ 26

Assignment of Subjects to Treatment Conditions ..................................... 28

Description of the Experimental Procedures ......................................................... 29
Construction of the Story .......................................................................... 29

Development of the Videotaped Stimuli .................................................. 29

Data Collection and Procedures ............................................................................ 31
Testing of the Modality Conditions ........................................................... 31

Order of testing .............................................................................. 31

Testing site ..................................................................................... 32

Testing Procedures: Story Retelling .......................................................... 32

Testing Procedures: Direct Questioning .................................................... 33

Recording the Data ............................................................................................... 34

Data Distractions ....................................................................................... 35

Defining Utterances ................................................................................... 35

Data Processing ..................................................................................................... 36

Data Analysis ......................................................................................................... 36
Quantitative Data Analysis ........................................................................ 37

Qualitative Lexical Data Analysis ............................................................. 37
Syntactic Qualitative Analysis ................................................................... 38
Reliability ....................................................................................... 39
Data processing .............................................................................. 39
Statistical Analysis of Data .................................................................................... 39

CHAPTER 3

RESULTS .......................................................................................................................... 41
Responses to the Story Retelling Task ............................................................................... 41
Modality Group Comparisons in Quantity of Language ....................................... 42
Modality Group Comparisons in Lexical Quality of Language ............................ 43
Modality Group Comparisons in Syntactic Quality of Language ......................... 44
Responses to the Direct Questioning Task ............................................................ 45
Modality Group Comparisons in Quantity of Language ........................... 45

Modality Group Comparisons in Lexical Quality of Language ................ 46

Modality Group Comparisons in Syntactic Quality of Language ............. 47

vi

Comparative Analysis of the Retelling and Direct Questioning Tasks ................. 49

Quantity of Language ............................................................................................ 49
Lexical Quality of Language .................................................................................. 52
Syntactic Quality of Language ............................................................................... 54
Conclusions ........................................................................................................................ 57
CHAPTER FOUR
DISCUSSION .................................................................................................................... 59
Comparison to Other Studies ................................................................................. 60
Interpretation of Modality Effects .............................................................. 61
Interpretation of No Modality Effects ........................................................ 62
Clinical Implications of Story Retelling for Language Sampling ......................... 65
Research Implications ............................................................................................ 67
Types of Questions to be Addressed in Future Research ........................... 67
Methodological Issues in Future Research ................................................ 68
Conclusion ............................................................................................................. 70
APPENDICES
Appendix A - Teacher Questionnaire .............................................................................. 72
Appendix B - Letter to Parents ....................................................................................... 73
Appendix C - Parent Consent Form ................................................................................ 74
Appendix D - Parent Questionnaire ................................................................................ 75
Appendix E - Criteria Reference F orm ............................................................................ 76
Appendix F - Letter with Results to Parents ................................................................... 77
Appendix G- Letter to Parents When Child Didn’t Participate ........................................ 78
Appendix H - Human Subjects Approval Form ............................................................... 79
Appendix I - Story Sample ............................................................................................. 80
Appendix J - Comprehension Questions ......................................................................... 82
Appendix K - Utterance Form ........................................................................................ 83

vii

Appendix L - Sample SALT Forms ................................................................................ 84

Appendix M - Data Form ............................................................................................... 94
Appendix N - Syntax Rules ............................................................................................ 95
Appendix 0 — Syntax Forms .......................................................................................... 98
BIBLIOGRAPHY ....................................................................................................... 100

viii

LIST OF TABLES

Table 1 - Ages for Subject Pairs in Modality Presentations at the Time of

Testing ................................................................................................................................ 23
Table 2 - Overview of Dependent Variables Compared for Each Story Retelling and
Direct Questioning Modality Condition ............................................................................ 37
Table 3 - The Number of Utterances, Words, and MLU Generated for the Retelling

Task .................................................................................................................................... 42
Table 4 - The Number of Different Words (NDW) and Type Token Ratio (TTR)
Generated for the Retelling of Task ................................................................................... 43
Table 5 - The Syntactic Structure of Language Generated for the Retelling Task ............ 44

Table 6 - The Number of Utterances, Words, and Mean Length Utterance Generated
for the Direct Questioning Task ......................................................................................... 46

Table 7 - The Number of Different Words (NDW) and Type Token Ratio (TTR)
Generated for the Direct Questioning ................................................................................ 47

Table 8 - The Syntactic Structure of Language Generated for the Direct Questioning
Task .................................................................................................................................... 48

Table 9A - Comparative Quantitative Analysis of Number of Utterances, Words, and
Mean Length Utterance Generated for the Retelling and Direct Questioning Tasks in the
Audiovisual Modality ........................................................................................................ 50

Table 9B - Comparative Quantitative Analysis of Number of Utterances, Words, and
Mean Length Utterance Generated for the Retelling and Direct Questioning Tasks in the
Audio OnlyModality .......................................................................................................... 51

Table 10A - Comparative Lexical Qualitative Analysis of Number of Different Words
(NDW) and Type Token Ratio (TTR) Generated for the Retelling and Direct Questioning
Tasks in the Audiovisual Modality .................................................................................... 53

Table 10B - Comparative Lexical Qualitative Analysis of Number of Words and Type
Token Ratio Generated for the Retelling and Direct Questioning Tasks in the Audio
Only Modality .................................................................................................................... 53

ix

Table 11A - Comparative Syntactic Structure Analysis Generated for Retelling and
Direct Questioning Tasks in the AV Modality .................................................................. 55

Table 11B - Comparative Syntactic Structure Analysis Generated for Retelling and
Direct Questioning Tasks in the Audio Only Modality ..................................................... 56

Table 12 - Number of Children in This Study Who Fell Above or Within One Standard
Deviation Below the Mean for MLU and TTR Measures According to Mille and
Chapman (1991) ................................................................................................................. 66

LIST OF FIGURES

Figure 1 - Categories of Syntactic Expressions ............................................................ 38

xi

Chapter One

BACKGROUND INTRODUCTION

Oral Language Sampling

Oral language sampling has become a valuable tool for clinically assessing language
skills (Morris-Friehe & Sanger, 1992). This procedure typically involves individuals
engaged in conversation in a relaxed natural setting with limited structure. Children may talk
about a topic of their choice or respond to cues about a topic given by the clinician. Best
clinical practices require the speech-language pathologist to follow the child’s lead in
conversation.

The speech samples are expected to be representative of a person's habitual language
patterns. The components of language, (i.e., pragrnatics, semantics, morphology, phonology,
and syntax) can all be evaluated with this procedure. According to Miller (1981), "the
evaluation of children's productive language through free-speech sample analysis is one of
the most revealing procedures available to us.” (Page 9) Shipley and McAfee (1992) concur
that an evaluation of language capabilities is not complete without taking language samples.
They help determine whether language impairment exists, and specifically, what the deficit

is. The child’s needs then can be specified, and treatment strategies designed to meet them.

2

It is reasonable to ask why spontaneous oral language sampling is so valued given
that many different standardized tests are available for evaluating a child’s expressive and
receptive language competency. However, many existing language tests analyze a child's
language only at a word or sentence level (Merritt & Liles, 1989). They typically rely on
such elicitation methods as recalling, categorizing, and imitating words and sentences. John
and Bemey (1967) stated that imitative tasks may be dull for younger children. According
to Mandler, Scribner, Cole, and DeForest (1980), these methods make categorical and
associational demands that don't always allow children to express their "true” habitual
language. For example, the act of categorization requires a very specific type of
decontextualized language response that seldom is required for routine communication.
Younger children and those with limited education may not have fully developed this skill
(Mandler, et al., 1980). In contrast, language sampling allows language to be assessed in

a more comprehensive manner.

Willing

Stockrnan (1996) identified four advantages of oral language sampling. First, the
language itself is highly accessible to observation when elicited in natural contexts.
Language sampling could be present virtually anywhere that natural conversation occurs.
Therefore, no specific planning is required to elicit the sample.

Second, language sampling reflects a child’s individual language style, including
his/her cultural distinctions. Ethnic variation in language use may be misinterpreted or

penalized by conventional tests that do not take into account these differences. Such cultural

3
biases can be minimized in language sampling when the child leads the conversation using
his/her own words, experiences, lifestyle, and upbringing as a guide.

Third, oral language sampling also has practical utility. It can be used with all people
who communicate using speech. Language sampling, unlike standardized tests, does not
require a certain response to be provided by the child during the assessment.

Finally, language sampling permits a comprehensive competency assessment. It
offers the opportunity to observe form, content, and use. Pragmatics is particularly
neglected and difficult to assess with standardized tests. But with oral language sampling,
many routine pragmatic features can be observed (e.g., children’s tum-taking pragmatic
skills when communicating with a clinician). Form and content can be analyzed by
observing the number and types of words used, as well as the accuracy of a child’s phonemic
productions. Sentence structure, or syntax, can be analyzed to determine the type,

appropriateness, and complexity of the grammar produced by the child.

I' 'I I' [I S l'

Stockrnan (1996) also identified four limitations of language sampling. The typical
method for obtaining language samples is via conversational speech. It may be difficult to
a elicit adequate language samples from young children though because the child plays a
large role in the success of getting the language sample (Dollaghan, Campbell, & Tomlin,
1990). In clinical contexts, children may be shy or uncomfortable around strangers.

Therefore, they may not talk as much or in the same manner as they do in more familiar

4
circumstances. The child's attention to the task may also have an impact, as well as his/her
motivation.

Second, language which is adapted to a particular context of use need not reveal the
full range of skills in a child’s repertoire. Since there is a reduction of commrmicative
pressure on children to use language other than what is easy for them, the language sample
may not reveal the extent of a person’s language competencies or skills. Wren (1995)
demonstrated the effect of elicitation context on children’s syntactic constructions. She
compared seven different techniques (involving free play with puppets, storytelling,
explanation of a game, stories from pictures, viewmaster, specific tasks-birthday party, and
sentence building tasks) for eliciting language. Using a repeated measures design, she
examined how these language elicitation techniques affected the quantity and quality of
language produced by children. This was done by evaluating the utterance length and
number and types of syntactic structures that were at or above the criteria for Stage V of the

- (LARSP). Each of the seven

 

techniques were examined with the goal of revealing which one yielded the most complete
picture of a child’s language abilities. She concluded that no one task elicited the most
representative example of a child’s language. For example, the birthday task elicited the
most utterances, whereas the story retelling technique yielded the most adjectives. Only a
combination of the techniques revealed the array of skills that a child was capable of
producing. According to Wren (1985),

"most recommended techniques for eliciting language from young children (even such ones

as “tell me about your favorite TV show this week”) yield either minimal language or
language that is not representative of what the child can actually produce" (83).

S

A third limitation of language sampling is that it requires a lot of time to administer
and analyze appropriately. An adequate sample for analysis may require a lengthy
conversation between the examiner and child. The number of utterances needed for good
representation of language skills also is debatable, although 50 utterances seem to be
regarded as appropriate for most research and clinical purposes (Wren, 1985). Transcribing
this many utterances plus those of the examiners can be time intensive. There also are no
set topics in conversational speech. Thus, a clinician may have to make assumptions about
what was said by the children when their speech is unclear. The clinician may be forced
to exclude the utterances, or worse, obtain another language sample.

Finally, considering all the factors that can create variability in an oral language
sample, it is not surprising that the standardization of the language sampling procedure has
surfaced as an important clinical issue. This study responded to the lack of standardized
language sampling analysis procedures. Standardized testing generally is beneficial
because it decreases the potential effect of extraneous situational factors on performance.
A clinician needs to be sure that an individual’s performance on a test is due to linguistic
skill and not to the manner of testing. Presently, standardized procedures for the language
sampling approach do not exist. Consequently, it would be impossible to replicate language
sampling conditions and outcomes. Standardizing the language sampling method would
allow a clinician to obtain language under uniform conditions. Once a standardized language
sampling procedure is established, norms may be generated so that a child’s language sample

can be compared to that of his/her peers.

6
Language sampling is an integral part of language assessment. Developing a
language sampling analysis procedure in which the disadvantages of language sampling can
be reduced and the advantages accentuated is warranted. If language sampling is to remain
a viable procedure for language assessment, one must find a way to overcome these

limitations. Story telling methods offer one way to do this.

Story Telling

Interest in using story telling methods in the field of speech language pathology has
increased in recent years as researchers and clinical practices have come to focus on
pragrnatics and discourse competency. Although the main focus of using story telling
currently has been to test a child's story comprehension and narrative discourse skills, there
is the potential to use this technique for eliciting language samples to grammatical skills.
According to Merritt and Liles (1967),
"Narrative analysis takes into account a wide range of language abilities that are used
regularly in the child's educational environment. It facilitates a broad description
of a child's language use encompassing the interaction of sentence formulation, use of
linguistic devices to conjoin meaning across sentences, and the general organization of
content.” (p. 438)
Story telling could be used effectively to elicit a language sample. Liles (1993) suggested
that stories provide valuable data about children’s pragmatic and linguistic abilities. At the
same time, story telling is a natural cultural experience in many societies. Stories provide

part of the scaffolding for learning language in children’s environments. They stimulate a

child’s imagination (John & Bemey, 1967). Narratives are common forms of stimulation for

7
children because of "their critical role of exposing children to language routines" (John,
Homer, & Bemey, 1970, p. 250).

Language sampling that relies on habitual conversation may not detect impairments
due to the lack of task difficulty. Children may avoid words and respond to the clinician as
simply as possible to avoid using more complex language. If unnoticed by a clinician, these
children could be portrayed as normal speakers. In contrast, story telling may prove to be
a strong indicator of language impairment, even if the child appears normal in habitual
conversation. This method may actually indicate language impairment better than some of
the standardized tests available (Culatta, Page, & Ellis, 1983). The present study was
motivated by the possibility that story telling can also be used as a standardized oral

language sampling procedure. Story generation is one way to elicit language from a child.

W
Story generation requires children to invent a narrative using their own words. It
is not the same task as telling a story that one already knows. If children were familiar with
a story, requesting them to talk about it would be a retelling task. Story generation allows
children to be creative and original in their narrative presentation. In many cases, children
are shown pictures (static or dynamic) and asked to make up a story about what they saw

(Schneider, 1996, Dollaghan, et al., 1990, Shroyer & Hamilton, 1996).

Write

Story generation allows a child to use his/her own words. In addition, there is less
structure and pressure placed upon the child to answer a question or talk about a subject/topic
a clinician proposes. Transcription time is also decreased so long as the story context is
bounded by a picture or theme. The clinician does not have a significant verbal role while
a child tells a story, so obviously, the child takes the lead. The narrative abilities of a child
during the story generation task could also indicate if a child is learning-disabled. Some
children who are learning disabled have the skills needed to produce narratives, but do not
utilize these skills as efficiently as normally developing children (Morris-Friehe & Sanger,

1992)

Wu

The difficulty with a story generation task is the lack of structure surrounding the
topics generated by the children. Even if pictures are presented for children to describe (as
is the case of many research studies), there is no definitive indication that the stories told by
one child will be ones that can be compared to another. Different life experiences may cause
different interpretations of pictures. In addition, the children may discuss dissimilar topics,
which makes reliable comparisons between the two impossible. Therefore, the obstacle of
discovering a way to make a standardized language sampling technique remains. In addition,
pragmatics cannot be analyzed using this technique because of the limited interaction
between the clinician and the narrator. Wren (1985) suggested that comparability among

children, their typical productions, and the representativeness of the language elicited must

9
be kept in mind when selecting language eliciting tasks. Story generation does not allow
comparisons across different children because there is no control over the story theme. Story

retelling may alleviate this problem.

Wing

Story retelling is another story telling method. It requires the child to be presented
a story (i.e. written, verbal, or oral with visual components). The child must then retell the
events that occurred in the story to another person. Story retelling differs from story
generation in some crucial ways. Story retelling gives the child a specific topic to discuss.
The clinician’s familiarity with the story script makes for easier and shorter transcription.
When words are unclear, this familiarity makes interpretation more accurate. If all children
are told the same story under the same conditions, then norm referencing may be possible

for the language sampling task.

We
W. Storytelling has many assets as pointed out above.

Hedberg and Stoel-Gammon (1986) proposed that retellings could expose complex language
organization skills. Culatta, et. al, (1983) asserted that story retelling validly reflects the
child’s use of language. The clinician provides a foundation for language expression in the
form of the story, and the child then is given the flexibility to use his/her own words (within
the vocabulary boundaries supplied by the story) and grammatical style to tell the story.

Sutter and Johnson (1995) observed that with longer stories, memory is taxed enough so that

l 0

children cannot just repeat the story verbatim, but relay it in their own linguistic style. John
and Bemey (1967) showed the importance of story telling and designed their experiment to
prove that story retelling demonstrates "how a child selects from and organizes the
continuous flow of stimuli--a process which is truly representative of his everyday
environment" (p. 266). By providing a child with the stimuli and story that is retold, the
clinician is able to control the topic of the language sample, and alleviate the problem created
by lack of similarity in story themes among children.

MW. Story retelling is a common language experience for many
children beginning at an early age. Applebee (1978) stated that during the preoperational
stage of development (two to six years of age), the typical response to discussing stories is
to retell the story. Applebee (1978) also found that most six and nine year old children
elaborated stories by recalling or retelling them. As age increases, storytelling skills become
more refined and the stories more complex in structure.

WW. Storytelling is also advantageous because there is
less language to transcribe. The child generally is the only person who speaks during the
retelling task.

MW. Narrative tasks require strong metalinguistic skills
and processing demands are also placed on the child with narrative tasks such as story
retelling. Therefore, story retelling could be a good index of future academic success (Paul
& Smith, 1993).

Wm. Advancements in media technology make it possible to

standardize story delivery protocols in ways that were not available years ago. Radio and

l 1
television are the biggest influences in modern society. Thus, a child could be presented a
story under the same visual/oral conditions as any other child participating in the same story
retelling task.

So far, different story presentation conditions have used. For example, a clinician
may tell a child the story (with or without pictures present). Consequently, there is no
control over the way the story is nan'ated. One clinician may not tell the story the same way
as another (e.g., words may be stressed differently). This could give one child an advantage
over another. Even when the story is told by the same clinician, it may not be narrated the
same way for every child. Using media technology this obstacle to story retelling
standardization can be overcome.

Stories can be presented in three different ways using media technology. These are
audio only, video only, and audiovisual presentation possibilities. The video only method
is viable for the story generation task whereas the other two methods are viable for the story

retelling task.

Ii il l' [SI Bllli

WM. There are a few limitations associated with story retelling.
For instance, children may not provide an adequate recapitulation of a story because they
assume that the clinician is familiar with it. Details are omitted, which most likely include
descriptions and elaborations (Liles, 1993). This problem can be solved by having the child
retell the story to a different person who s/he knows is not familiar with the tale (Liles, 1993;

Ripich &

1 2
Griffith, 1988; and Schneider, 1996). A more complete retelling ought to yield a larger
sample of language.

mmmmmmmm. Another problem arises because no tum-
taking interactions are required. Therefore, a broad range of pragmatic skills cannot be
observed. This problem may be minimized again by including a new participant in the
interaction. For example, when the new person enters the room, the manner in which the
child greets the newcomer, if at all, could be examined. This could also be accomplished by
asking questions about the story after the child has retold it. Tum-taking that occurs between
the two individuals during this question and answer period could be analyzed, focusing on
specific areas such as the child’s eye contact, response appropriateness or relevance to the
topic.

W- Cooperation and motivation of the subject remains an important
factor. One way to increase story retelling incentives is to alter the story presentation.
Dynamic audiovisual presentations may be more interesting to a child then looking at static
pictures with an oral accompaniment.

WWW- Finally, there are not standardized procedures for story
retelling. Some investigators have orally presented stories to be retold. (Schneider, 1996),
while others have presented static pictures with oral narration (Schneider, 1996, Shroyer &
Hamilton, 1996). Hayes, Kelly, and Mandel, (1986) utilized dynamic pictures that subjects
watched. Without a standard method to present the stories, story retelling would not be

useful as a norm-referenced for language sampling procedure.

l 3
e ' Ve e in

Both story generation and story retelling are useful methods for eliciting language.
When assessing a child’s narrative skills, the clinician discovers how well an individual has
mastered the skills of correct sequencing of events, memory/recall, stating a narrative that
is cohesive, and utilizing story schema. Also, a clinician could determine if the child is
sympathetic to an unfamiliar listener when telling the story. All of these areas provide
descriptive information about language discourse language information beyond the single
word level or even sentence level so often found in formal standardized language tests
(Ripich & Griffith, 1988).

Merritt and Liles (1989) compared the story generation and story retelling methods.
Their study aimed to show which method was of more clinical value. Forty children (20
normal and 20 language-disordered children, ages 9;0 to ll;4 years), were observed. The
targeted performances measured were story length, and number of story grammar parts,
complete episodes and story components mentioned.

The children were exposed to story stems in the story generation part of the
experiment. These story stems consisted of “a human protagonist and setting information
and was designed to evoke images of an adventure involving a series of goal-based events,”
(p. 439). The children were shown what to do before they generated a story. They generated
a total of three stories.

For the story retelling task, the children were given two stories to retell. The stories

were modified versions of the stories Buried Alive and Shipwrecked. The children each took

1 4
part in a practice retelling, and then were shown a black and white video of the stories. After
exposure to the stories, they were asked to retell them.

For the story component measures, both the normal and language-disordered children
showed a larger number of story components in the story retelling task. Both groups (normal
and impaired) used more complete episodes in the story retelling method. They used more
clauses in the story retelling task as well. Finally, the story retelling task showed more
emphasis on the setting in the story than did the story generation task.

The investigators concluded that both tasks were strong techniques that reflected
children’s narrative abilities. But they felt the story retelling task would be more beneficial
to a clinician. This task provides a longer language sample, and the longer the language
sample, the more representative of the child’s abilities it ought to be. Merritt and Liles
mentioned that the story retelling method took less time to transcribe because the clinician
was familiar with the story and could decipher what the child said from contextual
information. This made the retelling tasks more reliable when scoring. The story retellings
by children also contained more story grammar components in their narrations. Finally,
because the stories were the same for the children, comprehension testing could be done.
This could not occur with story generation. Examining a child’s comprehension increases
the scope of the assessment, giving the clinician more information for making diagnostic
decisions. Consequently, the study’s results suggest that the story retelling procedure may
be the best task to use when using story telling to standardize language sampling. However,
there are still problems to solve when using the retelling method. This study was designed

to address some of these issues, as considered below.

15

Statement of the Problem
Language sampling is a valuable and necessary resource when describing an
individual's language strengths and weaknesses. A major obstacle to its use, however, is that
standard language sampling procedures do not exist. Story retelling is a possible way to
increase the validity and reliability of language sampling across children. However,
problems still remain to be solved. This study was developed to address one of these

difficulties, namely the problem of using consistent story delivery procedures.

WHEEL!!!
The modality of input used for story delivery is an important aspect to address.
Visual only, audio only, or combined audiovisual modalities have been used to elicit stories

from children. Each condition is described below.

XisuaLinx

Many studies utilize the story telling task by asking children to generate a story
about what is happening in a video or series of static pictures. For example, Dollaghan, et
al., (1990), did a study in which children watched an 108 second black and white video and
were expected to tell what happened afterwards. In the visual only condition, the child tells
a story based on what was seen. This visually aided story generation task is not considered
a story retelling task (Liles 1993 as cited in Schneider 1996). Although some information

about the events are provided to the child, the actual verbal rendering of a story is not.

l 6
Therefore,
“visually presented stories not accompanied by narration present the additional requirement
of translating visual into verbal information. This differs from orally presented stories,
which provided prior linguistic structures” (p. 86).
A child has to create the linguistic structure, making the visual presentation a generation task.

Oral/verbal story retelling, strictly interpreted, requires one of two modality inputs, either

audiovisual or audio only.

Asking!!!

Audio only input requires children to listen to the story with no visual input.
Investigators may read a story out loud to the children (Culatta, Page, & Ellis, 1983; Ripich
& Griffith, 1988; Schneider, 1996). Another way to elicit a story retelling with audio only

input has involved children listening to a radio presentation (Hayes, et.al., 1986).

l 1' . I

Audiovisual input allows children to listen to the story at the same time they see
pictures of the story events. Some researchers have exposed children to oral readings with
static visual stimuli (usually pictures) shown at the same time (Schneider, 1996; Paul &
Smith, 1993; Shroyer & Hamilton, 1996; Pomper, Sauer, & Hedberg, 1996; and Pratt &
MacKenzie-Keating, 1985). Sutter and Johnson (1995) had subjects listen with headphones
to a story from a tape with a static picture accompaniment. John and Bemey (1967) had a
librarian read children a story and then the children retold it using pictures to guide them.

Hayes, et., al., (1986) compared television and radio presentations of stories that children

17
were to retell. Finally, Gibbons, Anderson, Smith, Field, and Fischer, (1986) presented
children with either an audiovisual or audio only story that was seen and/or heard on
television. Those children who saw the audiovisual story were then able to use toys that

represented the figures in the story when they were retelling what had happened.

 

Research has hinted that story retelling is a way to obtain information about one’s
grammatical language abilities, although studies have focused mainly on discourse areas of
competency. These studies have differed in their story delivery methods and end results also
have varied. Pratt and MacKenzie-Keating (1985) compared the effects of aural/oral and
videotaped story presentations on language performance. They found that the children who
both watched and listened to a tape produced more semantic referential errors than did those
who just heard the story. In contrast, Gibbons, et al., (1986) found that four-year-old
children produced more elaborations and remembered dialogues better in audiovisually
presented story condition than in an audio only condition. However the two studies
employed different types of language measures. Beagle-R005 and Gat (1983), as cited in
Gibbons, et a1. (1986), compared televised audiovisual stories with radio. They found that
individuals who were exposed to televised audiovisual stories remembered more details
about characters and action events. Radio input helped facilitate the recall of spoken words.
Like Beagle-Roos and Gat (1983), Hayes, et al., (1986) observed that their subjects had a
more difficult time remembering and understanding the story when they listened to a radio

instead of watching television. Those who heard the story were able to recall sound effects

l 8
and more of the conversations between the speakers. Salomon (1979), as cited in Hayes, et
al., (1986) believed that visual input contributed the most to the spatial and figural aspects
of the story depicted. Salomon (1979) also thought that verbal understanding and logical
thinking were best facilitated by verbal presentation.

Schneider’s recent 1996 study seems to have been the most comprehensive
comparative study of the effect of story presentation modalities. She examined the effects
of various story presentations on the language productions 16 language impaired children
ages 5:7 to 9;9 years. None had more than mild articulation difficulties. She presented the
stories to every child in four conditions: (1) The children saw the story events in pictures
only (story generation task) or (2) they heard the story with no pictures (story retelling audio
only task) or (3) they heard the story with picture accompaniment at the same time and (4)
picture accompaniment after the oral story presentation (story retelling audiovisual task).
The purpose of the study was to determine if the different presentations impacted a child’s
story, and “whether each presentation form may be appropriate for gathering information
about different aspects of narrative abilities” (Page 87).

The stories with pictures present were from the book, qus, by Mercer Mayer. For
all stories used, the same main character was present, with the introduction of new
supporting animal characters in each additional story. The stories had the same number of
words, clauses conjunctions, etc.. The children told/retold the stories to an aide who was
introduced as someone not familiar with the story to the children. The unfamiliar listener
only responded to the child’s story with neutral responses.

The results did not show a consistent bias effect of the input modalities explored.

l 9
For example, there was no statistically significant evidence that the children produced longer
stories in one modality condition than another, as measured by either the number or words
or utterances. However, the story generation task (pictures only) yielded a smaller number
of story grammar units than the other modes of stimulus presentation. The pictures only
yielded in the largest amount of both irrelevant and new information. The oral (audio only)
presentation resulted in the largest number of story grammar units and the most complete

story episodes.

The above cited studies do not show consistent agreement on the effect of different
types of story presentation modalities on language performance. A given method may be
better for one aspect of language but not another. Few studies have compared story
presentation modalities modalities using morphosyntactic measures of language
performance. Such measures could include utterance length, number and types of words
used, as well as the complexity of grammatical structure.

Although Schneider (1996) took the initial steps of toward story retelling responses
to obtain a language sample from children, many aspects were not explored. There are four
issues that should be considered that were not addressed in Schneider’s (1996) study. First,
there has not been systematic exploration of morphosyntactic variables. Schneider did
examine the sample in terms of the Mean Length Utterance (MLU) and vocabulary diversity
as assessed by the Type Token Ratios (TTR). Both are important aspects of language to

examine when analyzing language samples, but she did not look at the syntactic structure of

2 O
the utterances spoken by the children in her study. Analysis of syntactic structure would
indicate if a child was speaking at a word, phrase, or sentence level. If the child uses
sentences in speech, then the complexity of the sentences could be examined. For example,
one could monitor if a child used noun and/or verb elaborations, simple, complex, and/or
elaborated sentences. The later suggests more mastery of language.

The second issue to address is the population used in Schneider’s study. The children
were all language impaired. In order to obtain from a standard language sampling procedure,
normally developing children need to be tested. In addition, the children used in Schneider’s
study were older. To standardize language sampling for screening children with potential
language impairments, a clinician needs to know whether a procedure is suitable for a young
child. Early identification is important for intervention success. Therefore, research needs
to show how younger children perform on a story retelling task.

Third the use of media technology in the study may also have a strong effect on the
outcome of the research. Schneider (1996), like most studies, used static still pictures which
need not be presented in a uniform way. Dynamic audiovisual and audio only presentations,
which have already been prerecorded, control for the manner in which each child is orally
told the story. Schneider did not control for this variable in her study. In addition, static
still pictures may be harder for children to understand when they are required to interpret

action scenes.

Finally, story retelling and story generation needs to be clearly differentiated.

Schneider confounded both story retelling and story generation in her study, yet does not

2 I
analyze the strength of one language sampling elicitation method. Instead, many methods
are examined. This study takes only one of those methods, story retelling, and attempts to
determine if the modality input of the story (either audio only of audiovisual) affects the

quantity and quality of language produced by the children.

Purpose
All four factors mentioned above were considered when designing this study.
Therefore, the purpose of this investigation was to determine whether young normal males’
morphosyntactic responses to the story retelling task were differentially influenced by their
exposure to a specific presentation modality (audio only or audiovisual). Specifically, the

following questions were asked:

1. Is there a difference between audio only and audiovisual story presentations in the
quantity or amount of the language provided by the children during story retelling?
2. Is there a difference between audio only and audiovisual story presentations in the
syntactic and lexical quality of the language provided by the children during story

retelling?

Chapter Two

METHODS
In this experiment, story retelling was used as a standard procedure for eliciting oral
language samples from children. Selected grammar measures were evaluated under two
different modalities (audio alone and audiovisual) in an independent group design. The goal
was to determine whether there are differences between the modalities in the quantity and

quality of language elicited.

I! 'I' [Sl'l

The subjects were 29 normally developing children. All were Caucasian males who
were monolingual English speakers. The mean age of the children was 4;10, with a range
of 4;2 to 5,6 (see Table 1). They were all students at Michigan State University’s Child
Developmental Laboratories in East Lansing.

This study followed an independent group design with random assignment of subj ects
to treatment conditions. Because of the subjects’ large age range, the two groups were
carefully matched so that the ages in one group corresponded to those children in the other

group as shown below.

22

Table 1. Ages for Subject Pairs in Modality Presentations at the Time of Testing.

23

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Subject Number Age Modality Condition
(yr;mo;day)

l 5;4;23 Audio

2 5;4;15 Audiovisual
3 5;3;22 Audio

4 5;3;21 Audiovisual
5 4;7;4 Audiovisual
6 4;6;24 Audio

7 5;6;10 Audiovisual
8 5;4;6 Audio

9 4:8;10 Audio

10 4;9;2 Audiovisual
ll 4;2;8 Audio

12 4;2;9 Audiovisual
13 5;3;8 Audio

14 5;2;l l Audiovisual
15 4;2;25 Audiovisual
l6 4;4;23 Audio

17 4;1;20 Audio

18 4;2;15 Audiovisual
19' Audio

20" Audiovisual
21 4;11;10 Audiovisual
22 4;10;O Audio

23 4;5;22 Audiovisual
24 4;6;9 Audio

25 5;0;3 Audio

26 5;1;8 Audiovisual
27 4;6;1 l Audiovisual
28 4;8;4 Audio

29 5;2;26 Audio

30 5;2;27 Audiovisual
31 5;0;6 Audiovisual
32' Audio

 

a - Identifies children who refused to participate in the experimental task.
b - Identifies a child who was out of town during the testing process.

 

24

5!. IS] I' C'I'

Normally developing children who met the following criteria were selected as

subjects if they were:
1. male.
2. Caucasian (non-Hispanic).

3. 4;0 to 5;6 years of age.

4. monolingual English speaking.

5. Preschool-enrolled at Michigan State University’s Child Development Laboratories,
and

6. had passed a speech and hearing screening.

An attempt was made to keep the sample as homogenous as possible because of the small
subject pool. Therefore, the subjects attended the same school and came from similar
socioeconomic backgrounds. The sample also consisted only of young males in order to
control for potential gender biases. Males were chosen because they have been shown to be

at higher risk for language disorders than females.

Wm

W. The children were enrolled in either a morning or afternoon
class at Michigan State University’s Child Developmental Laboratories. Selection of the
children involved several steps. A list of males between the ages of 4:0 and 5;6 years who

met the selection criteria was obtained from the school. Next, these children’s teachers

2 5

responded to a questionnaire that elicited information about each child relative to the
selection criteria (see Appendix A). Only those children who received a good/normal overall
rating as designated on the form from the teacher survey were chosen as potential subjects.
This screening phase generated an initial list of approximately 54 potential subjects. Parental
consent was sought in a letter describing the study (see Appendix B). The correspondence
also contained a parental consent form and a parent questionnaire as shown in Appendices
C & D, respectively.

The returned parent consent forms and questionnaires were examined by the
investigator. Approximately 13 parents did not want their children to participate. These 13
children were removed from the potential subject pool, and their teacher questionnaire forms
discarded.

The remaining subject pool with parent consent forms was reviewed. None of the
parents expressed concern about their son’s language development or hearing. However,
about 27% of them reported that their children had multiple ear infections at younger ages.
Originally, these children were going to be excluded from the subject pool because ear
infections have been linked to language delay. But, removing them would have depleted
the subject pool extensively. Furthermore, none of the parents were concerned about the ear
infections at the time of the study and did not feel that their children had any hearing
difficulties related to the previous infection/s. This opinion also could be corroborated by
the language screening. Once the parent consent forms and questionnaires had been

reviewed, there were 41 potential subjects for the study.

2 6

W. In the final step of the selection process, the 41
children were given the Preschool Language Scale-3 (PLS-3) (Zimmerman, Steiner, &
Pond, 1979) and a hearing screening. These standard procedures were administered by the
investigator and a second year graduate student from Michigan State University’s Audiology
and Speech Sciences program. The graduate student subsequently assisted with data
collection and transcription of the children’s language samples as well. Other than being
alerted to her expected duties, the student was not told about the purpose of the study.

Before the screening and testing occurred, time was set aside for developing rapport
between the test givers and the children. The children’s classrooms were visited at least
twice in order to allow for the children and teaching staff to feel comfortable with the
investigator and assistant.

Screening began after this rapport period. The screening testing was done in two
rooms designated by the Child Development Laboratories. One of the test givers
accompanied the child from the classroom to the testing site. Often, the teaching staff
assisted with the process when the child appeared to be concerned about leaving the
classroom. The child was given the opportunity to refuse to participate. If the child agreed
to join the test giver, he was escorted to the screening site. The children were always
supervised and never left alone.

The PLS-3 (Zimmerman, Steiner, & Pond, 1979) was administered to ensure that the
potential subject's expressive and receptive language abilities fell within the range of normal

performance for the child's age. This test provides norm referenced information about

2 7
expressive and receptive language abilities in the age range studied. Scores were analyzed
using a 90% confidence interval.

Hearing screening was done using standard play audiometry procedures. The child’s
hearing was screened bilaterally at 20 - 25 dbl-IL for 500, 1000, 2000, and 4000Hz using an
audiometer and headphones. The child was instructed to respond to the tone by placing a
block into a bucket or raising his hand. In one testing room, there was a tile floor, that
amplified ambient noise from vents, cars outside, and distant classrooms. Therefore, some
children required the audiometer to be turned up from 20dB to 25dBHL in order to hear
tones at 500Hz and 1000Hz. In these cases, the child’s response to the auditory
comprehension section of the PLS-3 was scrutinized carefully to make certain that he was
within normal limits for his age. If the child passed the auditory comprehension portion of
the PLS-3, then he was judged to have adequate hearing for the story retelling task despite
having passed the hearing screening at 25dB for the lower frequencies. The second room
was carpeted and did not have the ambient noise levels that prevented the children from
passing the screening at 20dBHL for all frequencies. Following the speech and hearing
screening, each child was allowed to select a small piece of colored construction paper and
a sticker as a reward before being escorted back to the classroom.

The PLS-3 and hearing screening test results were analyzed. All the children
performed within normal limits. Only one child failed the hearing screening. This child was
not included as a subject in the study. Letters were sent to parents informing them of the

screening results(see Appendices F and G).

2 8
WWW

Once the initial screening was completed, the investigator identified subjects for the
study. Children were excluded if they did not meet the age or ethnic criteria, refused to
participate, or were too shy to do the screening. A total of 32 male children from the original
list met the screening criteria and were chosen as subjects for the study.

The 32 children were placed in pairs. Subjects closest in age were paired. The
largest age gap between subjects in a pair was three months. Once the matched pairs were
determined, each child was randomly given either an odd or even subject number, ranging
from one to 32. For example, if the first member of the pair was given the number five, his
partner was given the number six. A random numbers table was used to assign the children
in each pair to either the audio or audiovisual experimental condition. The first number in
the pair that occurred in the table was placed in the audio condition, and his partner was
placed in the audiovisual modality condition. For example, if the number six was found first
in the random numbers table, then this subject was placed in the audio condition, and subject
five, his partner, was put in the audiovisual group. The final subject pool consisted of 16
subjects each in the audio and audiovisual conditions (see Table 1).

Note that only 29 of the 32 children who met all the screening criteria were actually
given the experimental tasks. This is because one child from each modality condition
refiised to participate (subjects 19 and 32), and a third child was out of town at the time of
the retelling task (subject 20). Therefore, the audiovisual modality condition ended up

having one more language sample than the audio only modality condition.

2 9
i t e ' al Pr dure

W

The subjects were exposed to a single story in either an audio only or audiovisual
condition. The story was composed by the investigator (see Appendix I). An original story
prevented all the children from knowing the story line before the study began. The story
was designed following Stein and Glenn's (1979) story grammar. This required the story
to have six story grammar components in it--setting, initiating event, internal responses,
attempts, direct consequences, and reactions. Telling the story required approximately seven

minutes.

I! I | n“ | 15'. II

The video input for the story was made using moving puppets as the characters
engaged in the actions. The use of dynamic pictures in the story was critical to the
audiovisual version. Dynamic action presentations were expected to provide a more realistic
representation of story events and experiences in real time. Static pictures offer less realistic
representation of events because objects do not move. The child would have to use more of
his imagination to translate the “action” seen in a non-moving picture to a dynamic picture
(Stockrnan, personal communication, 1996).

The puppets were animated to portray human characteristics. The children were
expected to identify best with actions being performed by human characters. Stories having
animals or inanimate objects performing human-like tasks require children, especially those

who do not see visual images, to imagine something foreign to them. For example, cows

3 0
talking are not a part of their everyday world experiences. Therefore, to make sense of the
story, the children would have to be very imaginative. That is, they need more mental
images to match these fictional and unrealistic characters/actions. Even when watching
videos that provide these visual pictures, having surreal stories may cause some children to
be confused when an object or animal performs atypical actions. As a result, it could be
very difficult for a child to put into words what is viewed.

A graphic arts designer created the puppet characters used in the story. Four
individuals provided the story narration/sound effects. The investigator was the narrator,
and three other individuals irnpersonated the three characters in the story (Timmy, Tirnmy’s
mom, and Timmy’s friend, Amy). However, they were not professional artists. Their voices
had the regional dialect typical of a middle-class individual from lower Michigan. The
narrator attempted to say words either before or after a character spoke or performed an
action. This was done so as not to distract the listener too much from the story events.

The puppets were manually manipulated, to act out the story line (see Appendix I).
The story actions were fihned using a Sony High 8 video camera. The puppets’ actions
were performed using props and fishing line to imitate talking and other movements such as
waving and walking. Sequences were videotaped at different angles to create the action of
the story (i.e. showing the ball rolling in the mud and having the camera man walk in the
mud after the ball, like the character Timmy would have done). The story was videotaped
without sound included. After all of the scenes were completed, they were edited by cutting
and pasting them together for a smooth flow of events. An editing machine which, allowed

for fade outs/ins, was used.

3 1
Once the visual condition was completed, the voices were recorded and synchronized
with the videotaped pictures. The synchronized audiovisual story product was copied from
the High Video 8 machine to a regular VHS for story presentation. The audio alone
videotape was made in a similar way. The audiovisual version of the story was sent to the

VCR but without the video input cable attached.

Wm

Each child received an individual subject number. Once they received the number,
the child was referred to as this subject number in the data collection. Thus, all names were
eliminated in order to ensure confidentiality, and guard against any unwanted bias in
handling data.

A television monitor with a VCR was used to present the children the story. The
children who received the audiovisual version of the story listened to and viewed the story
from the television. The same equipment was used to deliver the audio only version of the
story. The television monitor was covered by a blanket to minimize distraction. As a

technical precaution, every session was audio recorded on separate backup equipment.

I I' [M ll'l C [1'

W. The testing of the modality conditions began approximately two
weeks after all of the children had been screened. The order of testing in each modality was

randomized.

3 2
Mugging. The child was accompanied to the test site by the investigator and/or the
graduate assistant. The room contained a small table on which a television set was placed.
The VCR was on a metal shelf beside the television. The same Sony High 8 video camera
was also set up in the corner of the room to record the child’s words, as well as a tape
recorder on a chair just behind the child to be used for back-up purposes. No reference was
made to the video camera or the tape recorder so that the children would not be distracted by

them.

e 11'
Once the child reached the testing site, the investigator said to the child,
" , this is my friend (pointing to the assistant who stayed in the room with the
children during each story presentation). You may remember her fiom earlier. She brought
a story especially for you that I want you to pay close attention to. I haven 't seen/heard it
before. I have to leave for a minute, but will be right back. When I come back, maybe you
can tell me a little bit about the story."

It was assumed that the children would be less likely to omit parts of the story when
communicating with someone unfamiliar with it (Liles, 1993). The children were directed
to sit on a focal spot on the floor marked by “X.” This location controlled for the distance
that the children sat from the monitor, and also placed the child within the video camera’s
scope of view. As the investigator left the room, she discretely turned on the tape recorder.

The graduate assistant also turned on the video camera without the child’s

knowledge. She then turned on the VCR, and played the story in the specific modality

33

designated for that child according to his random assignment. The assistant did not distract
the child unless he needed to be redirected to pay attention to the story.

At the conclusion of the story presentation, the investigator returned to the room. The
graduate assistant turned off the television and VCR, said goodbye to the child, and left the
room. The investigator then said to the child, “can you tell me about the story you just
saw/heard. I missed it. ” For most children, that was the only prompt needed to initiate a
story retelling. For others, the statement had to be repeated or the following additional
prompt was used, “I heard the story was about a boy named Timmy. What happened to

Timmy?” At this time, most of the children attempted to retell the story to the investigator.

The responses given to the child during the story retelling took the form of head
nods, facial expressions and occasional “oh,” “oh my goodness,” and “I see” affirmations.
No other verbal responses occurred with one exception. If a child mmnbled under his
breath or turned away fi'orn the camera so that his words were not audible for recording, then
the investigator repeated the child’s words. If the child finished speaking but did not
indicate that he had completed telling the story, the investigator asked, “is that al . ” or
“anything else?” After the child completed the story retelling, he was thanked for telling the

story. The story retelling task was immediately followed by a direct questioning task.

Waning
The pilot sessions suggested that the story retelling samples may not provide enough

language to analyze for every child. Therefore, direct questioning about the story was added

34

as a post hoc procedure in order to elicit more language from the children. A standard set
of nine questions was created. The questions were asked of every child in the same sequence
shown in Appendix J. The children’s responses were included in the analysis regardless of
its semantic accuracy relative to the story content. The answers to the last question, which
asked the child if he had ever gotten his own shoes muddy, were not used in the analysis.
This is because the child’s answers need not have been related to the story presented. This

same sequence of events occurred with each child.

W

After the data were collected, each child’s recorded experimental session from the
Sony High 8 video camera was copied on to a VHS tape so that the samples could be viewed
for transcription. Then the language samples were orthographically transcribed by the
investigator and assistant and represented on a standard transcription form (see Appendix K).
Every word that the child said was transcribed. The investigator and the same graduate
assistant who witnessed the experimental sessions transcribed the language sample together
by viewing and listening to the video and audio recorded samples of the children. A
transcribed utterance was included for analysis only if both observers agreed on its surface
form. When a word was not agreed upon by the two observers, they re-listened to the tape.
If the words were still unclear, they were judged unintelligible, and the utterance was

counted for the total number of utterances, but not analyzed for any other measure.

3 5
Two separate language samples resulted from the study. The first sample was
entitled “retelling.” The second sample contained the directed questions and answers and

was called, “directed questions.”

I! | D . I I'

During the testing session, some children introduced other conversational topics that
were not related to the story. For example, they talked about school projects. Utterances that
did not contain information about the story were excluded from analysis. In some cases,
the investigator repeated what the child said in case the child’s words were not clear on the
tape recordings. Often the utterances were repeated in an intonation which suggested that
a question was being asked of the child. Therefore, many children replied with one word
affirmatives. However, because not all subjects were given the opportunity to reply to these
“incidental questions,” these responses were excluded from the analysis. The goal of this
study was to determine how the method of story retelling affected the child’s language

sample using a uniform method of obtaining these samples.

Defininglltteranm

Naturally, the children retold the story using running discourse in which sentence
boundaries are not always obvious. Thus, it was necessary to use an utterance identification
criterion. The investigator segmented the child’s responses into utterances using Nelson’s
(1993) guidelines (p. 346). The faculty advisor also independently segmented the child’s

responses into utterances. Then a discussion and resolution of discrepancies between the

3 6
investigator and faculty advisor was made. Once the utterances were agreed upon within the
samples, they were submitted to a second assistant to enter into the computer for data

processing.

ata o s '

Once the orthographic transcriptions were completed, the data were entered verbatim
into the computer using the Systematic Analysis of Language Transcripts (SALT) (Miller
& Chapman, 1991). A second assistant, a computer science student at Michigan State
University, entered the data for two separate language samples on each child, one for the
retelling task, and one for the direct questioning task. Once completed, the investigator r

reviewed the transcripts, checked for errors, misspellings, and revised them accordingly.

mm

Data were divided into three categories of dependent variables(see Table 2). The
total number of words, Mean Length Utterance, and total number of utterances said by the
child were placed under one heading entitled “quantity of language.” A second heading
titled “lexical quality of language” consisted of the child’s total number of different words
and Type Token Ratio (TTR). The third heading, “syntactic quality of language” was
established. It consisted of the remaining syntactic information (percentage of complete
sentences, elaborated sentences, simple sentences, elliptic simple sentences, complex

sentences, and elliptic complex sentences).

37

Table 2. Overview of Dependent Variables Compared for Each Story Retelling and Direct
Questioning Modality Condition.

 

 

A. Quantity of Language
1. Total number of utterances
2. Total number of words
3. Mean Length Utterance

B. Lexical Quality of Language
1. Type Token Ratio
2. Total number of different words

C. Syntactic Quality of Language
1. Percentage of elaborated sentences
2. Percentage of complete sentences
3. Percentage of simple sentences
4. Percentage of elliptic simple sentences
5. Percentage of conrplex sentences
6. Percentage of elliptic complex sentences

 

W
The SALT program (Miller & Chapman, 1991) separately totaled the number of words and

number of utterances produced and computed the Mean Length Utterance (MLU) for each child in
both language samples. The MLU averaged the number of morphemes per utterance using an

adaptation of Brown’s rules (Brown, 1973).

W
The SALT program (Miller & Chapman, 1991) totaled the number of different words used
in a sample and computed the Type token Ratio (T'I'R)for every child in both language samples.

TTR is the ratio of different words related to the total number of words used in a sample.

 

3 8
S I I'll Iilll ! l'

Syntactic structure was determined by observing the types of sentences produced. See the
classification scheme in Figure 2. Sentences were identified as incomplete, complete, elaborated,
complex, and so on. Children were not penalized for incorrect semantic responses made during the
story retelling or directed questioning about the story. Strict guidelines (see Appendix N) were
developed and followed for the syntactic analysis to insure consistency of category assignment
across different observers.

When scoring the syntactic component, the investigator and her faculty advisor
independently completed the syntax analysis form (see Appendix 0) according to the type of
sentence that the child used. This task was completed without either person being aware of the
modality condition to which the child had been assigned. Sentences were judged as complete or
incomplete. Then sentences in each category were differentiated according to grammatical features
of ellipses, complexity, and elaboration. The investigator and her faculty advisor met together and
compared their syntactic analyses. The independent analyses were reviewed and a 100% agreement
was made on the syntactic qualities of each utterance. The combined analysis was then placed on

the forms formd in Appendix 0.

 

 

Incomplete Complete

1
Elliptic /Non-elliptic

Elliptic Simple Elliptic Complex Simple \\Cmnmcx\
Elaboration None Elaboration None Elaboration None Elaboration None

 

Figure 1. Categories of Syntactic Expressions.

 

3 9

mm. A second year graduate student at Michigan State University’s Audiology and
Speech Sciences department performed the same sentence structure analysis. This was done to
check for the reliability of the investigator and faculty advisor’s syntactic evaluation. A point-to-
point comparison of all data across subjects on the audio only condition yielded 97% and 95%
agreement for the retelling and direct questioning task, respectively. The reliability for the
audiovisual group was 93% for the retelling and 97% for the directed questioning task,
respectively.

W. Once the data forms were completed for each child, they were computer
processed using Excel Software. Scores were computed for each syntactic category based on the
total number of utterances in each sample. These scores were entered on the data form as shown

in Appendix M.

W

Separate analyses were done for the direct questioning and retelling tasks for each child.
Multiple analysis print outs were provided, as shown in Appendix L. The numerical data were

consolidated and represented on a single sheet as shown in Appendix M.

The student Minitab (Schaefer, RL., & Anderson, R.B., (1987)Software was used to do the
statistical analysis. The Minitab program provided basic measures of central tendency and the tests
of differences between the two treatment groups along with the Type One error probability.
Although the data were suited to multi variate statistical analysis, it was not undertaken here because
of the limitations of the Minitab program. Follow-up analysis involving average sine

transformations of the data will be pursued for future dissemination of the work.

4 0
Because the study involved a comparison of two independent groups, comparisons between
the two modality treatment groups could be made using a two sample t-test for each dependent
variable. The two treatment conditions, audio only and audiovisual, were compared on three sets
of dependent quality and quantity variables for the retelling task. The same analysis was applied

separately to the oral language samples elicited from the direct questioning task.

Chapter Three

RESULTS

In this study, subjects were presented a story in either an audiovisual or audio only
condition. Spontaneous nonverbal and verbal responses were observed during the story
presentation. Those who saw and heard the story sat very still, watching the monitor
intently. There was little or no movement or talking. Children who only heard the story
reacted differently. A few subjects wandered around the room. Others played with their
shoes, narnetags, clothes, and so on while listening. Some looked around the room, facing
away from the monitor. These observations suggested that the children exposed to the audio
only condition would have a more difficult time retelling the story. However, the following

results countered this expectation.

Walt
Children were exposed to either an audiovisual or audio only story presentation. Data
were collected for the global language aspects studied. The areas of quantity of language,

lexical quality of language, and syntactic quality of language were measured.

41

4 2
Story retelling responses for the two modality conditions were compared on number
of utterances, number of words, and mean length utterance (MLU). The results are shown
in Table 3.

Table 3. The Number of Utterances, Words, and MLU Generated for the Retelling Task.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NAV= 15 Mean No. of SD Median Range t Score
NA = 14 Utterances
Low High
Audiovisual 8.60 5.28 8.00 l - 21
Audio Only 7.79 3.02 7.50 3 - 13
0.51
p > 0.05
Mean No. of SD Median Range tScore
Words
Audiovisual 54.70 33.60 48 3 - 144
Audio Only $7.50 34.10 44 12 - 124
Median Range -0.23
P > 0.05
Mean MLU SD t Score
Audiovisual 7.57 1.81 7.20 5.00 - 11.00
Audio Only 8.15 2.58 8.23 3.20 - 12.11
-0.69
P > 0.05

 

 

 

An analysis compared the two groups on each of three dependent variables: the
number of utterances, number of words, and utterance length. The means for each dependent
variable were close. For example, the means for MLU were 54.7 and 57.5 for the
audiovisual and audio only modalities, respectively. The means were not consistently higher

for one modality than the other across the measures observed. For example, the mean for the

43

audio only condition was highest for MLU, but lowest for total number of utterances. The
t test of differences between group means revealed no statistically significant difference on

any of the quantity measures (see Table 3).

in xi a i
For this analysis, story retelling responses in the two modality conditions were
compared on type token ratio (TTR) and number of different words (see Table 4).

Table 4. The Number of Different Words (NDW) and Type Token Ratio (TTR) Generated
for the Retelling Task.

 

 

 

 

 

 

 

 

 

NAV= 15 Mean TTR SD Median Range t Score
NA = 14 Low High
Audiovisual 0.65 0.13 0.66 0.47 - 1.00
Audio Only 0.6! 0.13 0.62 0.38 - 0.77
0.83
P > 0.05
Mean NDW SD Median Range t Score
Audiovisual 32.60 17.40 31.00 3 - 81
Audio Only 32.90 15.60 28.50 5 - 63
-0.04
P > 0.05

 

 

 

 

 

 

 

The numerical closeness of the means was undeniable. For example, the TTR means
were 0.65 and 0.61 for the audiovisual and audio only modalities, respectively. The means
were not consistently higher for one modality than another. The modality conditions did not
differ significantly for neither the number of different words nor the type token ratio (see

Table 4).

44
d ' arison ’n S ntactic uali f Lan u
Language used in story retelling was analyzed for its syntactic quality on six

dependent variables as shown in Table 5.

Table 5. The Syntactic Structure of Language Generated for the Retelling Task.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NAV= 15 Mean % of SD Median Range t Score
NA = 14 Complete
Sentences Low High
Audiovisual 87.80 12.00 90.91 62.50 - 100.00
Audio Only 88.70 14.10 91.99 50.00 - 100.00
-0. 17 P> 0.05
Mean % of SD Median Range t Score
Elaborated
Sentences
Audiovisual 86.50 14.90 90.91 62.50 - 100.00
Audio Only 87.10 13.20 88.89 30.00 - 100.00
-0.11 P> 0.05
Mean % of Simple SD Median Range t Score
Sentences
Audiovisual 45.80 22.90 42.86 16.67 - 100.00
Audio Only 47.50 22.40 43.65 0.00 - 77.78
-0.20 P> 0.05
Mean % of SD Median Range t Score
Complex
Sentences
Audiovisual 33.40 19.80 28.57 0.00 - 60.00
Audio Only 37.10 26.60 38.89 0.00 - 100.00
-0.42 P> 0.05
Mean % of Elliptic SD Median Range t Score
Simple Sentences
Audiovisual 4.26 7.46 0.00 0.00 - 25.00
Audio Only 0.79 2.97 0.00 0.00 - 11.11
1.67 P> 0.05
Mean % of Elliptic SD Median Range t Score
Complex
Sentences
Audiovisual 5.29 8.16 0.00 0.00 - 20.00
Audio Only 3.22 5.99 0.00 0.00 - 20.00
0.78 P> 0.05

 

 

 

 

 

 

 

45

A consistent trend for the means to be higher for one modality than another across
all measures was not observed. For example, the mean for the audio only condition was
higher for the percentage of simple sentences, but lower for the percentage of elliptic simple
sentences. The means for the two presentation modalities did not differ significantly on any
of the six syntactic variables. Table 5 shows that both story presentation conditions yielded
language samples that were predominated by complete and elaborated sentences, and more
simple than complex sentences. Elliptic structures accounted for the smallest percentage of

the syntactic characteristics in both conditions (see Table 5).

WW

Linguistic responses to direct questions about the story were analyzed separately.

The analyses paralleled those used for the story retelling.

 

The audiovisual and audio only conditions were compared on three dependent
quantity variables. These comparisons for number of utterances, number of words used, and

mean length utterance are shown in Table 6.

46

Table 6. The Number of Utterances, Words, and Mean Length Utterance Generated for the
Direct Questioning Task.

 

 

 

 

 

 

 

 

 

 

 

 

NAV= 15 Mean No. of SD Median Range t Score
NA - 14 Utterances Low High
Audiovisual 12.00 3.25 11.00 8 - 20
Audio Only 12.00 3.98 11.00 8 - 23
0.00
P > 0.05
Mean No. of SD Median Range t Score
Words
Audiovisual 68.40 28.80 62.00 33 - 122
Audio Only 73.40 41.80 64.50 27 - 182
-0.37
P > 0.05
Mean MLU SD Median Range t Score
Audiovisual 6.60 1.70 6.36 4.00 - 10.17
Audio Only 6.61 1.79 6.35 3.78 - 9.46
-0.01
P > 0.05

 

 

 

 

 

 

 

 

The analysis revealed that the group means for each dependent variable were close.
For example, the MLU means were 6.60 and 6.61, respectively for the audiovisual and audio
only modalities. The t test of differences between the group means revealed no statistically

significant difference on any of the quantity measures (see Table 6).

 

The two story presentations were compared on the lexical quality of language using

number of different words and Type Toke Ratio (TTR), as shown in Table 7.

47

Table 7. The Number of Different Words (NDW) and Type Token Ratio (TTR) Generated
for the Direct Questioning Task.

 

 

 

 

 

 

 

 

 

NW: 15 Mean TTR SD Median Range t Score
NA = 14 Low High
Audiovisual 0.61 0.1 1 0.60 0.44 - 0.79
Audio Only 0.61 0.12 0.62 0.45 - 0.85
-0.13
P > 0.05
Mean NDW SD Median Range t Score
Audiovisual 38.90 10.60 37 24 - 54
Audio Only 41.60 17.70 40 20 - 86
-0.49
P > 0.05

 

 

 

 

 

 

 

The means for TTR were identical and the standard deviations were extremely close.

For example, the means for the audiovisual and audio only conditions were 0.61, with
standard deviations of 0.11 and 0.12, respectively. The means for the number of different
words also were close at 38.9 and 41.6 for the audiovisual and audio only conditions,
respectively. The audio only condition had the largest range of different words, but only one
child used more than 58 different words. However, none of the differences between the

means for the two presentation modalities were statistically significant in the lexical quality

of the language produced (see Table 7).

 

Language used in the direct questioning task was analyzed for its syntactic quality

using six dependent variables as shown in Table 8.

48

Table 8. The Syntactic Structure of Language Generated for the Direct Questioning Task.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NAV= 15 Mean % of SD Median Range t Score
NA -= 14 Complete
Sentences Low High
Audiovisual 81.40 7.52 81.82 70.00 - 91.67
Audio Only 77.50 12.90 81.82 55.56 - 92.31
098
P > 0.05
Mean % of SD Median Range t Score
Elaborated
Sentences
Audiovisual 81.40 7.52 81.82 70.00 - 91.67
Audio Only 78.40 11.10 81.82 50.00 - 92.31
083
P>005
Mean % of SD Median Range t Score
Simple
Sentences
Audiovisual 45.20 19.70 44.44 5.00 - 90.00
Audio Only 44.92 9.83 43.65 35.71 - 57.14
004
P > 0.05
Mean % of SD Median Range t Score
Complex
Sentences
Audiovisual 20.10 17.40 20.00 0.00 - 53.33
Audio Only 20.60 16.70 21.83 0.00 - 43.48
009
P>005
Mean % of SD Median Range t Score
Elliptic Simple
Sentences
Audiovisual 6.24 8.48 5.00 0.00 - 30.00
Audio Only 4.39 7.80 0.00 0.00 - 22.22
0“
P > 0.05
Mean % of SD Median Range t Score
Elliptic
Complex
Sentences
Audiovisual 6.91 6.16 6.67 0.00 - 18.18
Audio Only 8.50 4.70 9.09 0.00 - 15.38
078

 

 

 

 

 

P > 0.05

 

 

4 9
For each syntactic category, none of the differences between the means of the two
presentation modalities were statistically significant for any of the six syntactic structures

compared (see Table 8).

o 'v ' th Ret lin ir t uest' in s
The language produced in the retelling and direct questioning tasks was compared
separately for each presentation condition. The analyses paralleled the focus of the two

preceding analysis on the quantity and quality of language produced.

QuantituLLananage

Data for the quantity variables are shown in Tables 9A and 9B for the audiovisual
and audio only conditions, respectively. The results showed that the direct questioning task
generated consistently more utterances and words than did the retelling task in both

modalities.

50

Table 9A. Comparative Quantitative Analysis of Number of Utterances, Words, and Mean

Length Utterance Generated for the Retelling and Direct Questioning Tasks in the
Audiovisual Modality.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N = 15 Mean No. of SD Median Range t Score
Utterances
Low High
Retelling 8.60 5.28 8.00 1 - 21
Direct 12.00 3.25 11.00 8 - 20
Questioning
-2.12"'
p < 0.05
Mean No. of SD Median Range t Score
Words
Retelling 54.70 33.60 48 3 - 144
Direct 68.40 28.80 62 33 - 122
Questioning
-1.20
P > 0.05
Mean MLU SD Median Range t Score
Retelling 7.57 1.81 7.20 5.00 - 11.00
Direct 6.60 1.70 6.36 4.00 - 10.17
Questioning
1.52
P > 0.05
" Statistically Significant

 

51

Table 9B. Comparative Quantitative Analysis of Number of Utterances, Words, and Mean
Length Utterances Generated for the Retelling and Direct Questioning Tasks in the Audio
Only Modality.

 

 

 

 

 

 

 

 

 

 

 

 

 

N = 14 Mean No. of SD Median Range t Score
Utterances
Low High
Retelling 7.79 3.02 7.50 3 - 13
Direct 12.00 3.98 11.00 8 - 23
Questioning
-3.16"'
p < 0.05
Mean No. of SD Median Range t Score
Words
Retelling 57.50 34.10 44.00 12 - 124
Direct 73.40 41.80 64.50 27 - 182
Questioning
-1.10
P > 0.05
Mean MLU SD Median Range t Score
Retelling 8.15 2.58 8.23 3.20 - 12.11
Direct 6.61 1.79 6.35 3.78 - 9.46
Questioning
1.84
P > 0.05

 

 

 

 

 

 

 

‘ Statistically Significant

However the differences between tasks were statistically significant for only the
number of utterances in each modality condition but not for the number of words or mean
length utterance. This outcome can be explained by the fact that each child responded to a
standard number of questions. The direct questioning task guaranteed that at least eight
utterances would be produced. The story retelling task yielded more variable output among

the children- some retold long stories with as many as 21 utterances while others did not.

52
In each modality condition, some children produced as few as one utterance. Therefore, the
mean number of utterances was reduced for the story retelling samples relative to those
obtained from the direct questioning task.

With respect to mean length utterance (MLU), the story retelling task generated
longer utterances on the average than did the direct questioning task in both conditions.
However, the difference between group means was not significant for neither the audiovisual
nor the audio only conditions. The consistency of shorter MLU’s for both conditions in the
direct questioning task was likely due to the greater use of elliptic utterances during the

direct questioning than story retelling task.

I . I O H [I

Data for the lexical quality variables are shown in Tables 10A and 10B for the
audiovisual and audio only conditions, respectively. The results showed that the direct
questioning task generated consistently more different words than did the retelling task in
each modality conditions. But the difference between the tasks were not statistically

significant.

53

Table 10A. Comparative Lexical Qualitative Analysis of Number of Different Words
(NDW) and Type Token Ratio (TTR) Generated for the Retelling and Direct Questioning
Tasks in the Audiovisual Modality.

 

 

 

 

 

 

 

 

 

 

 

 

 

N = 15 Mean TTR SD Median Range t score
Low High
Retelling 0.65 0.13 0.66 0.47 - 1.00
Direct 0.61 0.11 0.62 0.44 - 0.85 1.00
Questioning P > 0.05
Mean NDW SD Median Range t Score
Retelling 32.60 17.4 37 24 - 54
Direct 38.90 10.6 40 20 - 86
Questioning
-1.19
P > 0.05

 

Table 10B. Comparative Lexical Qualitative Analysis of Number of Words and Type Token
Ratio Generated for the Retelling and Direct Questioning Tasks in the Audio Only Modality.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N = 14 Mean T'I'R SD Median Range t score
Low High
Retelling 0.61 0.13 0.62 0.38 - 0.77
Direct 0.61 0.12 0.62 0.45 - 0.85
Questioning
-0.05
P > 0.05
Mean No. of SD Median Range t Score
Different
Words
Retelling 32.90 15.60 28.50 5 - 63
Direct 41.60 17.70 40.00 20 - 86
Questioning
-1.38
P > 0.05

 

 

 

S 4

With respect to type token ratio (TTR), the means were nearly identical for the
samples generated by the story retelling and direct questioning tasks in both conditions. For
example, the means for the audio only modality were identical (.61) for both the retelling and
direct questioning tasks. The TTR means for the audiovisual modality were .65 and .61 for
the retelling and direct questioning tasks, respectively.

The number of different words spoken by the children was also larger for the direct
questioning than the story retelling task in each modality. This trend was most likely due
to the to the larger number of utterances produced by the children. More words were said,
and a larger variety of vocabulary words were therefore produced. However, none of the
differences between the two means for the tasks were statistically significant on any of the

lexical quality variables in either modality.

WWW:

Data for the syntactic quality variables are shown in Tables 11A and 11B for the
audiovisual and audio only conditions, respectively. The results showed that the story
retelling task generated consistently more complete and elaborated sentences than did the
direct questioning task in both modalities. However, the direct questioning task generated
more elliptic sentence structures than the retelling task did. But differences between the two
tasks in the total number of complete sentences was statistically significant only for the audio

only condition.

55

Table 11A. Comparative Syntactic Structure Analysis Generated for Retelling and Direct
Questioning Tasks in the AV Modality.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N = 15 % of Complete SD Median Range t Score
Sentences Low High
Retelling 87.80 12.00 90.91 62.50 - 100.00
Direct Questioning 81 .40 7.52 81 .82 70.00 - 91.67
1.77 p> 0.05
% of Elaborated SD Median Range t Score
Sentences
Retelling 86.50 14.90 90.91 62.50 - 100.00
Direct Questioning 81 .40 7.52 81 .82 70.00 - 91 .67
1.19 P> 0.05
% of Simple Sentences SD Median Range t score
Retelling 45.80 22.90 42.86 16.67 - 100.00
Direct Questioning 45.20 19.70 44.44 5.00 - 90.00
0.09 P> 0.05
% of Complex Sentences SD Median Range t score
Retelling 33.40 19.80 28.57 0.00 - 60.00
Direct Questioning 20.10 17.40 20.00 0.00 - 53.33
1.96 P> 0.05
% of Elliptic Simple SD Median Range t score
Sentences
Retelling 4.26 7.46 0.00 0.00 - 25.00
Direct Questioning 6.24 8.48 5.00 0.00 - 30.00
.0.68 P> 0.05
% of Elliptic Complex SD Median Range t score
Sentences
Retelling 5.29 8.16 0.00 0.00 - 20.00
Direct Questioning 6.91 6.16 6.67 0.00 - 18.18
-0.62 P> 0.05

 

 

56

Table 11B. Comparative Syntactic Structure Analysis Generated for Retelling and Direct
Questioning Tasks in the Audio Only Modality.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N = 14 % of Complete Sentences SD Median Range t Score
Low High
Retelling 88.70 14.10 91.99 50.00 - 100.00
Direct Questioning 77.50 12.90 81.82 55.56 - 92.31
2.19‘ P < 0.05
% of Elaborated SD Median Range t Score
Sentences
Retelling 87.10 13.20 88.89 30.00 - 100.00
Direct Questioning 78.40 1 1.10 81.82 50.00 - 92.31 1.87 P > 0.05
% of Simple Sentences SD Median Range t score
Retelling 47.50 22.90 43.65 0.00 - 77.78
Direct Questioning 44.92 9.83 43.65 35.71 - 57.14
0.40 P > 0.05
% of Complex Sentences SD Median Range t score
Retelling 37.10 26.60 38.89 0.00 - 100.00
Direct Questioning 20.60 16.70 21.83 0.00 - 43.48
1.96 P > 0.05
% of Elliptic Simple SD Median Range t score
Sentences
Retelling 0.79 2.97 0.00 0.00 - 11.11
Direct Questioning 4.39 7.80 0.00 0.00 - 22.22
-1 .61 P > 0.05
% of Elliptic Complex SD Median Range t score
Sentences
Retelling 3.22 5.99 0.00 0.00 - 20.00
Direct Questioning 8.50 4.70 9.09 0.00 - 15.38
-2.59‘ P< 0.05

 

"' Statistically Significant

 

5 7

For both modality conditions, the retelling task produced a larger percentage of
complete and elaborated sentences than did the direct questioning task. This most likely
occurred because the response to directed questioning did not require the children to answer
them in complete sentences, according to the syntactic rules used in this study (see Appendix
N). In fact, a larger percentage of elliptic responses was generated in the direct questioning
than in the story retelling task, in both modalities. But, the difference in mean percentage
of elliptic responses between the two tasks was significant for the audio only condition but

not the audiovisual condition.

Conclusions
The results of this study showed that there were no statistically significant differences
between the modality conditions for story presentation on any of the linguistic measures
analyzed. This means that both the audiovisual and audio only conditions yielded story
retelling samples equivalent in:
1. Quantity of language (Number of utterances, Number of words, and Mean Length

Utterance (MLU)).

2. Lexical quality of language (Type Token Ratio (TTR) and Number of different words

(NDW))
3. Syntactic quality of language (complete, elaborated, simple complex, elliptic simple,

and elliptic complex sentence structure).

5 8
Since no significant differences were found (except for a larger number of utterances in the
direct questioning task than in the story retelling and direct questioning tasks), the children

also produced language samples that were equivalent across tasks as well.

Chapter Four

DISCUSSION

This study aimed to determine whether one story presentation condition (audiovisual
or audio only) elicited a larger amount and complexity of language in the story retellings of
4 - 5 year old normally developing children than the other. Twenty-nine normally developing
males at Michigan State University’s Child Developmental Laboratories served as subjects
in an independent group design. They were randomly assigned to one of two groups. Each
group was presented the same story in either an audiovisual or audio only modality
condition. The children then retold the same story and answered a standard set of questions
about it. Their story retelling was analyzed for the quantity, lexical quality, and syntactic
quality of language produced. Results revealed no statistically significant differences
between presentation modalities for any of the language areas measured in the story retelling
and direct questioning tasks.

It was concluded that the added visual input did not make a difference in the length
or complexity of language produced by a child during the story retelling or direct questioning
tasks. Only one child made a comment that could have been attributed directly to having
seen the video. He commented on the color of mud, calling it “brown” instead of using the
actual word. No other comments suggested that the addition of dynamic pictures made a

difference in the language samples of the children.

59

6 0
r's i

This study can be compared directly to Schneider’s (1996) research, which examined
language use during a story retelling task. Although she focused mainly on how various
modality inputs (e.g., oral only presentation, oral presentations with picture accompaniment,
etc.) affected a child’s narrative abilities, one of the analyses focused on the amount of
language produced by subjects. The quantity variables investigated (Mean Length of T-unit
in Morphemes, number of words produced by the children, and number of utterances said)
were the same as those used in this study. Schneider’s analysis revealed no statistically
significant differences among modality conditions for story presentation on any of the three
quantity measures. This outcome is consistent with the results of this study, in which no
modality outcome effects were observed for the quantity of language variables.

However, Schneider (1996) did observe modality differences on other language
discourse measures. For example, the audio only presentation produced the largest number
of story grammar units, while the visual only story delivery condition yielded the least
number of mazes. Thus, while other studies have revealed modality effects, there are two
interpretations of an absent modality effect in this study: (1) A modality effect exists but this
study was not able to reveal it or (2) No modality effect exists for young normally

developing children on the variables measured.

II II' [MII'IEEEI
It is possible that this research study did not reveal a modality bias because it focused

on different measures of language performance than did other studies. Global aspects of

6 1
language, (e.g., number and length of utterances and morphosyntactic patterns), may be less
vulnerable to modality effects than are the measures of narrative discourse typically
examined by other studies. For example, the aforementioned outcome of Schneider’s (1996)
study showed that the audio or oral only modality produced the largest number of story
grammar units and most complete stories. But in the same study, no modality bias showed
up on the amount of language production.

Second, the type of subjects used in this study may have also created different data
outcomes relative to other studies. The subjects in this study were young, normally
developing males. In contrast, Schneider’s (1996) subjects were older language impaired
children. In her study, the language impaired children showed an oral only modality bias
for some discourse areas measured. For example, the largest number of story grammar units
were provided in the oral only condition. It is not clear whether the audio only bias was due
to their language impairment status or older age.

Third, the manner in which the story recall was tested in this study may have been
a factor. Children may have preferred the audio only presentation in Schneider’s study
because still pictures were used in the video condition and still pictures might have been
difficult for children to interpret, especially if action events were depicted in the pictures.
In this study, however, dynamic pictures were used. These pictures may have been so clear
and easy for the children to interpret the context of the picture. The children were able to
view the all the actions of the characters, from the initiation to the completion of movements.
They did not see a single still frame (as found in static pictures). However, no prompts were

used to reveal visual imagery.

6 2
W

Although it is reasonable to expect the audiovisual modality to offer more than one
source of input about the story, it is possible that no modality effect exists, at least for the
global type of language measures used in this study. It is also possible that one modality
presentation is actually better than two. When receiving both new audio and visual stimuli,
the children may not be able to process two modality inputs at one time. What this suggests
is that children, when engaged in activities, may have to sacrifice some information input in
order to perform a task. In this study, each child knew he had to recall the story. Therefore,
the task required him to pay close attention to what was happening in the story. It is
possible that the children focused upon just the audio or visual input when presented the
story in both modalities. Focusing on both may have been too much of a challenge for the
subjects at their young age. If so, then the children exposed to the audiovisual condition in
this study may not have taken advantage of the additional information provided to them
because it was too much information for them to process at one time.

Research suggests that multiple sources of information compete for the same limited
cognitive resources. Bloom (1996) proposed the cognitive cost hypothesis. Bloom (1996)
observed children’s affective display of negative, positive, or neutral emotion in relation to
when they spoke and/or manipulated toys. In a longitudinal study, children were observed
when first words were emerging, and at the time of the vocabulary spurt. Results showed
that the children in the early vocabulary stage either talked or demonstrated some form of
emotion, but both did not occur at the same time. At older ages, when the children

developed a larger vocabulary, they talked and demonstrated affect together, but not without

63

some cognitive costs. The words used by the children were familiar and routine. Emotional
expression was either neutral or positive- the valence associated with least cognitive effort.

Bloom (1996) also observed object manipulations and use of language. She reported
that unless the children were accustomed to the object manipulation activity, they talked very
infiequently, if at all. Emotional affect was also practically non-existent when a less familiar
object manipulation task was being performed.

Tomasello (1995) summarized research on verb acquisition in children. In one
study, Tomasello and Kruger (1992) used 48 two year-olds as participants. Each child
played with a parking garage toy. The examiner then introduced the verb “plunk” to the
child, and demonstrated how to “plunk” a doll while using the toy garage. A child heard the
word in one of three conditions. In the impending condition, the child heard the word before
the action occurred. Either the child performed the action after the examiner asked him to,
or the examiner informed him that she was about to “plunk” the doll. In the second
“ongoing” condition either the examiner or child performed the action of “plunking” at the
same time that the word was spoken. In the completed condition, the word “plunk” was
spoken after the clinician or child completed the “planking” action.

Comprehension of the targeted word’s meaning was measured afterwards. The child
was asked to perform the target action “plunk” by manipulating the toys. Distractor actions
were requested as well. The children who were exposed to the impending condition used
the targeted verb more often than those in the ongoing and completed groups. Those in the
ongoing action group used the targeted verb the least. Tomasello and Kruger (1992)

interpreted the results in terms of attentional demands. They suggested that task demands

6 4

were increased because the ongoing condition required the child to process information about
both the words and the actions at the same time. Consequently, the children could have paid
more attention to the action that was occurring than the novel words. Although this study
did not involve an action event, attentional principles could apply to its outcome. Given the
novelty of the story for the subjects, there may have been some cognitive cost to having to
process a story with more than one modality of input.

The contention that added visual input to the auditory condition may not have been
used by the children seems supported by Schneider’s (1996) study. Schneider (1996)
presented stories to children using audio only, visual only, and audio plus visual story
presentations. No statistically significant differences among the modalities were observed
in the quantity of language used. The audio only produced the same quantity of language
as the visual only and audio plus static picture accompaniment during or after the oral
portion presentations. This outcome clearly shows that the audiovisual mode was no better

than either modality used singly.

 

There are advantages to using story retelling as a standardized language elicitation
task for global screening purposes. Children produce complex and elaborated sentences, as

well as a variety of sentence types (e.g., elliptic and morphological markers).

Adding a direct question task at the end of the retelling also expanded the amount of

language spoken in a time efficient manner. In some instances, children did not provide

6 5
enough language in their story retelling. They gave as few as one to three utterances in some
cases. The eight directed questions guaranteed that at least eight utterances would be spoken
by the children. This part of the study also showed that the children could produce syntactic
forms (e.g., elliptic forms) that were almost nonexistent in the retelling task.

Other aspects of language that were not the focus of this study, could be monitored
as well. For example, the responses could have been analyzed for semantic accuracy.
Although an informal observation revealed that the children seemed to understand the story,
their answers were not judged on semantic accuracy. Some pragmatic characteristics such
as monitoring tum-taking skills could be observed as well.

Another advantage is that the story retelling task took a short amount of time relative
to most language sampling tasks. The entire process of story presentation, retelling, and
answering the directed questions afterwards lasted approximately fifteen minutes. Seven of
the minutes consisted of the story presentation. The length of time needed to get a useful

sample is drastically reduced.

Despite the limited number of utterances and words spoken by the children ( no more
than 23 by one child), the data yielded by MLU and TTR performances were comparable to
the norms in the SALT Database (Miller and Chapman, l991)(see Table 20). Their database

included language samples from over 200 children, and were based on 50 utterances.

66

Table 12. Number of Children in This Study Who Fell Above or Within One Standard
Deviation Below the Mean for MLU and TTR Measures According to Miller & Chapman
(1991)

 

MLU- MLU- Direct TTR- TTR-
Retelling Questioning Retelling Direct Questioning

 

No. Children in This
Study Above or Within 26 25 27 27
1 SD Below the Mean

 

No. Children in This
Study Below 1 SD 3 4 2 2
Below the Mean in
This Study

 

 

 

 

 

 

 

From 86% to 95% of children in this study fell above or within one standard deviation below
the mean, even though none produced fifty utterances combined for the story retelling nor
the direct questioning tasks. In addition, two of the children who did not fall within one
standard deviation on the direct questioning task for MLU, performed well above the MLU
mean on the retelling task. Therefore, those children may not have talked as much during
the direct question task because they judged that enough detail had already been provided
during the story retelling task. Twenty-five of the 29 subjects in the retelling task had a
higher TT'R than Miller and Chapman’s (1991) study would predict, and 23 subjects had a
higher TTR than Miller and Chapman’s (1991) norms in the direct questioning task.
When combining the data from the story retelling and direct questioning tasks, results
paralleled those found above. The average number of utterances spoken by the children
when combining both tasks was 20.21, with a range of 11 - 33. When averaging the MLU’s
for both tasks, only four of the 29 subjects were more than one standard deviation below the

mean. When averaging the TTR’s for both tasks, all of the children had TTR’s that were

6 7
better than the mean given by Miller and Chapman (1991). This suggests that it may not

be necessary to elicit 50 utterances to evaluate global language measures.

Research Implications

r s e s

In this study normally developing white males between the ages of 4;0 and 5;6 years
were observed. Although the data outcomes were clear cut for them, more studies need to
be done in order to determine how generalizable the outcomes are to other groups. One goal
of such a study would be to examine the applicability of the story retelling method cross-
culturally. The characters and names in the story could be changed so that they fit the culture
of the children who are presented the story. Discovering if the retelling task produces the
same outcomes as those from this study would be valuable to developing a standardized
procedure.

Another shortcoming associated with the study was the population and school setting
used for observation. The children used as subjects all came from the same educational
background. They were a very homogenous group with socioeconomic status most likely
within the top ten percent of society (Whiren, 1997). The Child Developmental Laboratories
provide children with story retelling tasks in its curriculum. This task may have been easy
to do for the subjects of this study because they were familiar with it. It is not known
whether children who are not from a similar economic background and educational settings

would respond to the task in the same manner as these children did. Children who are not

6 8

familiar with the story retelling task either in their home life or educational settings may not
react to the task like the subjects of this study. In addition, the number of years that the
children have been students in the Child Development Laboratories could have affected how
well the children retold the story. For example, some children may have been students at the
CDL since they were babies. Therefore, they had more story retelling experience than those
who became students at the school at the age of four-years-old. Those children with more
experience may have provided better language samples during the story retelling task.

Another suggestion for future studies would be to determine how younger language
impaired children’s responses to the story retelling task are influenced by different modality
presentations. It is well known in the literature that younger language impaired children
have difficulty with auditory processing. Therefore, these children may actually favor the

visual only input modality.

“1111'” 'El B 1

Several issues surfaced in this study that need to be given attention in future studies
of story retelling. First, some of the children produced utterances that did not pertain to the
story that they were presented in the investigator’s judgment. Difficulties arose when
attempting to analyze these unforeseen distractions to data gathering. In the end, it was
decided to eliminate all words that did not directly pertain to the story. As a result, the
length of the language sample was reduced for some children. It is unknown how much was
lost in the language sample analysis because of this decision.

There also were obstacles to the analysis of the syntactic structure. The manner in

6 9

which the child’s sentence structure was scored needed to be consistent in order to ensure
reliability. Therefore, strict guidelines were used to assign utterances to syntactic categories
(see Appendix N). In turn, it became more difficult for a child to be credited with a complete
sentence, even if his utterance was complete from a semantic point of view. For example,
when some children were asked where Timmy and his mom were going, they answered, “to
a party.” Semantically, that is an accurate response and appropriately complete structurally
in terms of pragmatic rules for grammatical ellipses. However, because noun and verb forms
were missing from the utterance, it was considered syntactically incomplete. Some observers
may score a child’s answer to such questions in a different manner. The plan followed in this
study allowed the examiner to be very reliable when scoring the syntactic portion of the data,
but at the expense of semantic validity. The focus was on sentence structure not the
semantic accuracy of what was said. However, this decision no doubt undermined the
children’s real competence since they were credited with producing fewer complete
sentences than they did.

In addition, another methodological issue relates to the types of directed questions
asked of the children. The questions asked were about the sequence of the events that took
place in the story. Therefore, if a child was familiar with sequencing, these questions would
have been fairly easy for him. Asking questions that require children to draw on their visual
imagery, such as “what do you think that Timmy’s house looked like?” may have provided
the investigator with different results. The answer to this question would have been more

difficult for children who did not see the story, and therefore, a modality bias could have

 

 

7 0
resulted. Future research addressing the types of questions asked of the children would be
beneficial.

Designing and filming the story were expected to be cost efficient. It was assumed
that finding someone to create a dynamic depiction of story events would be a relatively
simple task. Instead, it became extremely complex. Costs limited the number of people
willing to create the story for the study. In the end, the video that was used was not
professionally created. In addition, the characters and narrator were not professional story
tellers. There is a possibility that the manner in which the video was made may have caused
the lack of modality differences found in the study. A professionally prepared video may
have captivated the children’s attention enough to make a modality difference in
presentation. The video could have not been engaging to the children because of the lack of
professionalism (e.g., unskilled character voices, puppets that were not appealing to the eye,
etc.). Future studies that have the funding available to produce a story using professionals
is strongly encouraged.

Cnnslusinn

In conclusion, the results of this study suggest that the addition of the visual input to
the story presentation does not enhance the language used in the story retelling task. No
significant differences were observed on any of the dependent quantity and quality variables
analyzed, nor did visual input enhance the children’s response to the questions asked about
the story. Adding the direct questioning task did significantly increase the number of

utterances observed, but there were no statistical differences between the two tasks in lexical

7 l
or syntactic quality of the language produced. Therefore, combining the data from the two

tasks may yield a more robust sample of language.

Results also imply that it is possible to standardize language sampling using the story
retelling method. The addition of a direct question task provided the possibility of eliciting
even more language from the children in a standard way. Further research needs to be
conducted to determine if these results can be generalized to other groups, but based upon
this study’s outcomes, the future looks promising for using story retelling to standardizing

language sampling.

APPENDIX A

‘ Please answer the following questions for child:

Care Giver/s Education (if lmown):

APPENDIX A

Teacher Questionnaire

 

 

Is English the primary language spoken at home? :

Has he and/or she ever mentioned any concern in language development or hearing for

 

 

? YES NO NOT SURE

1. Hearing
__ Good/Normal Questionable __ Poor
2. Social Interaction
__ Good/Normal Questionable Poor
3. Behavior

Good/Normal _ Questionable __ Poor
4. Problem Solving Skills
__ Good/Normal Questionable Poor
5. Reasoning
_ __ Good/Normal __ Questionable _ Poor
6. Vision

Good/Normal Questionable Poor
7. Communication Skills
__ Good/Normal Questionable Poor
8. Overall Rating (Of the above information)

Good/Normal _ Questionable Poor

 

Thank you for your time and your help on this project. It is very much appreciated.
‘(Note that after the subjects have been chosen, the questionnaires for children not chosen for the experiment will be destroyed, and
those that will be used will have their names blocked out and a random number assignment will be used in its place)

72

APPENDIX B

APPENDIX B

Letter to Parents

February 6, 1997

Dear Parent,

I am requesting for permission for your child to possibly participate in a language research
study. Your child will be given a free hearing screening following the American Speech-Hearing-
Language Association guidelines to see if your child fits the study. Your child will also be given
the Preschool language Scale-3 (PLS-3), a language test, to see if your child fits the study. You will
be notified of the results of these tests. Finally, your child’s teacher will be asked to fill out a
questionnaire about his language and hearing skills. If your child passes the above tests and his
teacher shows no concern about your child’s hearing and language, then he could be used in the
study.

Your child will be told a story and asked to retell it. The research study will take place
dining school hours at school, and last approximately twenty minutes. The time that yom' child will
be participating in the study will be arranged carefully in order to not disturb yoru' child’s normal
schedule as much as possible. Your child will have the choice to participate in the study. The
children will be invited to participate, and they can decline. They then will be invited to participate
one more time. If your child declines repeatedly, then your child will not be forced to participate
in the study. All information about your child’s identity will be kept confidential and will not be
revealed in the results from this project.

There is a questionnaire for you to fill out if you are allowing your child to participate in the
study. Because the study will take place at your child’s school during school hours, aside from
filling out the questionnaire, there will not be any additional requirements of you. If you have any
questions regarding the project, please feel free to call me at 432-2033 (work) or 351-6732 (home),
or Dr. Whiren at 355-1900. Attached is a consent form and survey. Please fill out each form and
return it to me using the self-addressed envelope no later than February 14, 1997.

Respectfully,

Jamie Gazella, B.A.
Graduate Student
Audiology and Speech Sciences

73

APPENDIX C

'I ”'9'

APPENDIX C

Parent Consent Form

Title of the Project: Children’s Story Retelling Under Different Modality Conditions

Investigator:

Jamie Gazella

Michigan State University

Department of Audiology and Speech Sciences
East Lansing, MI. 49924

Consent Form

My child CAN / CANNOT participate in the
research project on language sample analysis the Child Development Laboratory.

If consent is given, I understand that my child may refuse to participate or discontinue
participation at any time without penalty. I understand further that my child’s identity will
be kept confidential. Neither I nor my insurance company will incur any expense as a result
of my child’s participation in this project.

 

 

(Signature of parent/ guardian) Date

74

APPENDIX D

jar

APPENDIX D

Parent Questionnaire
Dear Parent,

Please answer the following questions as accurately as possible if you have consented
to have your child participate in this study. All information will be kept confidential. If your
child is chosen for the study, names will be blocked out and number assignments will be
used in its place. If your child is not a participant, this form will be destroyed. If you have
any questions, please feel free to call me at 351-6732. Thank you for you time and help.
1. Is your child bilingual? YES NO
2. How would you describe your child's language development?

Good/Nomal Questionable Poor

Comments:

 

 

 

 

3. How would you describe your child's hearing?

 

 

 

Good/Normal Questionable Poor
Comments:
4. Has your child had a large number of ear infections? If so, when?

 

 

Thank you for taking the time to fill out this form.

75

 

APPENDIX E

WA:L1§

APPENDIX E

Criteria Reference Form

(+ signifies passed, - signifies fail)

 

 

 

1. Hearing
Teacher assessment
Care giver assessment
Hearing screening
Right ear Left ear
_ 250Hz __ 250Hz
__ 500Hz _ 500Hz
_ 1000Hz _ 1000Hz
__ 2000Hz _ 2000Hz
__ 4000Hz _ 4000Hz
2. Care giver concerns YES NO

3. Level of education:

 

4. Bilingual YES NO

 

5. Teacher questionnaire overall score:

Comments:

 

 

 

6. Care giver consent formed signed YES NO
7. Care givers concerned about language development? YES NO

8. PLS-3 score:
Normal Other

 

9. Ear Infections/Comments:

 

 

Passed all criteria: YES NO

76

t?“

'IV

APPENDIX F

APPENDIX F

Letter with Results to Parents

March 23, 1997
Dear Parent:

Thank you again for your assistance with my research project. It was a pleasure working
with your son. Enclosed are the results of the hearing screening and language screening for:

 

 

 

Wt auburn 'U‘

Hearing screening:

Passed

 

Failed

 

Comments:

 

 

 

 

Language Screening (Using the Preschool Language Scale-3):

 

 

 

Again, I would like to thank you and your son for your help. I couldn’t have done this
project without you. The next part of my study will begin within a week. If your child
participates in the next part of the study, he will be told a story and then asked to retell it
to me.

If you have any questions regarding this information or about my project, please feel free to
call me at 432-2033 (work) or 351-6732 (home).

Sincerely,

Jamie C. Gazella
B.A. Speech-Language Pathology

77

APPENDIX G

Ivan-'2“
I
.A

APPENDIX G

Letter to Parents When Child Didn’t Participate

April 7, 1997

Dear Parent:

Thank you again for your permission for to participate in my
language study. When your son was asked if he wanted to join me, he indicated that he
was not interested. Therefore, hearing and language screenings were not obtained.

 

I would like to thank you though for your support. Without willing parents like yourself,
I could never have been able to do this research study. If you have any questions
regarding this information or about my project, please feel flee to call me at 432-2033
(work) or 351-6732 (home).

Sincerely,

Jamie C. Gazella
B.A. Speech-Language Pathology

78

APPENDIX H

"I 71-2" I
I ._

 

FAX: 517143-1171

'Dflmraeuu-y
wanna-M
my“

”balm
mm

APPENDIX H

HUMAN SUBJECTS APPROVAL FORM

MlCHlCANsm-E
UNIVERSITY

Jm up In”?

“I“
W*m‘”* tit .
' nan: gun’,-’ °
The Maternity hi: ' '
m1 , m as m a
asst: “mugging; mun. FEW WEE!
tor-‘3‘“ annual wwuhmmr
above. ' m w m. war-e: an ally m um

 

MS WW“

. . at ten:- such emanated “is ’
CO ‘ .. mm.e

either of m - .
have. the me et
....-....-...-..‘:.‘*&en:- .. w .1...” mi:-

5 nee-rah m

. the «new
mummmwanngmmwmm

them "
(817) -fill-Ien. do a: ire-teen in Queen: ea

79

APPENDIX I

WWW”

APPENDIX I

Story Sample

Story Example: (narrator in normal script, dialogue in bold)

Once upon a time, in a small town in Michigan, there lived a a young boy named
Timmy. Timmy lived in a big white house with pretty flowers around it.

Today was a special day for Timmy because he and his mom were going to a party.

“Hi, my name is Timmy.”

Timmy was already dressed up in his best party clothes. He was so excited for the party.
So, he skipped into the house hoping his mother was ready to go to the party because he
loved parties!

“Not yet,” she said.

“Can I go outside and play?” Timmy asked her.

“Ok, but whatever you do, don’t get dirty!” she warned.

“Ok, mom,” Timmy said as he closed the door.

Timmy walked outside. He saw his best friend Amy sitting on a rock in her backyard,
playing with her new basketball.

“Hi Tim, do you want to play with my new ball?” Amy asked.

“Sure Amy, throw it to me,” Tim said and walked over to Amy. He was very excited.
Amy slid off the rock she was on to get the ball.

“Throw it to me.” Timmy said

“Ok, here goes,” Amy said.

“Oh wow, this is fun!” Tim said after he caught the ball.

“Your turn,” he said as he threw the ball back to Amy.

Amy and Tim were having a great time playing catch with Amy’s new basketball. (Amy
and Tim are heard laughing).

“Ha, ha, ha,” Amy laughed, “this is fun,” she said, “here catch.” She threw the ball to
T irn. It rolled past him into some mud.

“I am sorry Tim!” Amy yelled as the ball kept rolling in the mud. Tim went to get the
ball.

(Noise of Tim chasing the ball in the mud)

“Oh no,” Tim said as he looked down at his shoes. They were all muddy.

“Don’t worry Timmy. We’ll think of something,” Amy said. “ I know, we have
some white paint at home. You can paint your shoes white and make them look like
new.”

Amy went home to get some white paint so Tim could paint his shoes to make them look
like new.

80

 

74

 

 

81

“Here’s some white paint Tim. Good luck, my mom says I have to go home and
eat.” Amy said and then left.

Tim looked at his muddy shoes and started to apint them. He painted them very
carefully. He didn’t want to miss a spot. He spent a lot of time painting his shoes. He
finally thought they looked just like new.

When he was finished, he skipped into the house to see if his mom was ready to go to the
party. He walked up the stairs but forgot that the paint on his shoes was still wet. He left
a trail of white footprints all over the stairs.

Tim’s mother came to the head of the stairs and looked down. She saw the paint marks.

“Oh,” she cried, “Tim, what did you do?” she asked.

Tim came to the head of the stairs and looked down too.

“Uh oh,” Tim said, seeing the white footprints on the steps.

“I am sorry mom,” he apologized. “ I got my shoes muddy. I painted them like
new so you wouldn’t be mad.”

She went to the door and Tim was scared that his mom was mad at him.

“Tim, I told you to stay clean. Now, I want you to stay right here and clean up your
shoes and the stairs. We won’t have time for the party. I am going next door to tell
them we can’t go,” Tim’s mom said as she walked outside and closed the door.
“Ohhh,” Tim sighed. He was upset that he couldn’t go to the party. He really wanted to

go.

Tim’s mom walked down the street to her neighbor’s house to tell them that Tim and she
couldn’t go to the party. But, just like Tim she did not look where she was going, and,
just like Tim, she stepped in mud too. Tim’s mom got her shoes very very muddy.

Tim’s mom walked back into the house. She opened the door and called to Tim, “Tim,

guess what,” she yelled to him, I got my shoes muddy too. Come on,” she said, “let’s
both change our shoes and clean up this mess. If we work together, then I am sure

we will make it to the party.”

“Yahoo!” Timmy said. He was happy because he was going to the party.

“I love you Tim/mom very much,” they said. They were happy and gave each other a
big hug. Then, they cleaned up the paint, changed their shoes, and went of to the party.
They had a great time together.

THE END

‘W Ax: anfim

 

APPENDIX J

0‘“

W -.-'1

APPENDIX J

Comprehension Questions

Where did Tim and his mom want to go?

What happened to Tim’s shoes?

How did Tim get his shoes muddy?

Who helped Tim fix his shoes?

How did Tim’s best friend Amy help him with his shoes?
Why was Tim’s morn upset?

What happened to Tim’s mom?

When did Tim and his mom go to the party?

Did you ever get your shoes muddy?

82

 

. .s 4..
{UN-fl, J 0 Streets 1’.

APPENDIX K

Transcriptions

Number:

APPENDIX K

Utterance Form

 

Modality:

 

Page:

 

Utterance Number:

Utterance:

 

 

 

 

 

 

 

 

 

 

 

 

83

 

APPENDIX L

nnmnnnmmnmnmmmmnn++++++++++++++++m

APPENDIX L

"SAMPLE SALT FORMS"

Child, Exami

ID:23

NAME:

SEX:

DOE:

DOB:

CA:4;6

E:

LOC:

CONTEXTzNAR

DOT:

T:

[301: Overgeneralization error (word level)
[EP]: Pronoun error (word level)

[EW]: Error at the word level

[EU]: Error at the utterance level

épéo: Prompt that does not interfere with discourse
It about Timmy.

Oh wow.

Can you tell me about it?

I/m sorry.

I miss/ed it.

No.

What happen/ed to Timmy?

He got his shoe/s mud/dy and it got paint all over the stair/s
[laughs] Oh.

What else happen/ed?

paint was all over the stair/s and he got his shoe/s mud/dy.
and his mom got his shoe/s mud/dy too [laugh].
and her dress mud/dy [laughs].

Anything else? ‘4

No.

They went to the party.

84

 

85

TRANSCRIPT SUMMARY

Total Utterances
Statements (.1
Rxclamations (1)
Questions (?)
Intonation Prompts (~)
Abandoned Utterances (>1
Interrupted Utterances (‘1

anverbal Utterances
unintelligible Utterances
Responses to Questions

Yes/No Responses to Questions
Responses to Intonation Prompts
Imitations

One-word Utterances

37.50
12.50

Utterances with Mazes

Utterances with Pauses

Utterances with Omissions
Utterances with Overlapping Speech
Utterances with Word Codes
Utterances with Utterance Codes

Analysis Set (C&I verbal Utts)
Statements (.1
Rxclamations (I)
Questions t?)
Intonation Prompts (~)

Responses to Questions

Yes/No Responses to Questions
Responses to Intonation Prompts
Imitations

One-word Utterances

Utterances with.Mazes

Utterances with Pauses

Utterances with Omissions
Utterances with Overlapping Speech
Utterances with Word Codes
Utterances with Utterance Codes

UOOOOO NOOHU OOOQQ NOOOOO N’OOHWOO OOOOOOO
I
I
1

Other
Between-Utterance Pauses: O

 

86

Analysis Set: C&I Verbal Utts

WORD AND MORPHEME SUMMARY

 

Child
Analysis Total
Set Utterances

MLU in Words 5.88 5.88
MLU in Morphemes 7.00 7.00

Brown’s Stage Post V Post V
Expected Age Range (mos.) 41‘- -- 41 - --
TTR 0.51 0.51
No. Diff. Word Roots 24 24
Total Main Body Words 47 47
Total Maze Words 0 0
No. Omitted Words 0 ‘ 0

MLU note: expected age ranges are the values within one standard deviation

of predicted values for that linguistic stage; based on a linear fit of a
sample of 123 upper-middle class children (17-59 months of age) from.Madison,
Wisconsin (Miller and Chapman, JSHR, 24, 1981, 154-161). They are irrelevant
for adults and for transcripts without bound morpheme coding.

 

87

WORD ROOT TABLE
Table Expanded by Bound Morphemes and Codes
CaI verbal Utts
Main Body
lst Speaker

Total and d
ABOUT Exp e Total Expanded

DRESS

a:
P1
a:
hJHOULJN nwunaptaaaaoawsura

ooowwoo o ooooo oouoooocoow

0'!
E;
IJF‘HIJF'Hln N

"I“ 4 “WIS-tum

£3100!!!an NNONNONDWOOONONOGOWGONDN mmu++++++++++++++++m

88

Child, Exami

ID:23

NAME:

SEX:

DOE:

DOB:

CA:4;6

E:

LDC:

CONTEXT:CON

DOT:

T:

[BO];

[EP]: Pronoun error (word level)

[EW]: Error at the word level

[EU]: Error at the utterance level

ngo: Prompt that does not interfere with discourse
Oh they went to the party.

[name] I just want to ask you a couple quick question/s
you just told me cause you told me a lot.
Where did Tim and his mom go?

to a party.

What happen/ed to Tim/s shoe/s?

they got all mud/dy [layeddown]. '
they got all paint and they were still wet.‘
Oh my goodness.

and they got (all over the stair/s) all over the stair/s.

they got on all of the step/s and mom look/ed down.

and they saw (footprint/s toot track/s) tootprint/s.
I see. '

Then his mom got angry.

Ohh.

and then she stepp/ed.

and she didn/t know where she was go/ing.

and she stepp/ed in mud just like Timmy did [laugh].
She did. Oh my goodness.

and she got her dress all mud/dy [laugh].

How did Tim get his shoe/s mud/dy?

by walk/ing in mud.

by walk/ing in mud.

Who help/ed him.£ix his shoe/s [name]?

I don/t know.

You don/t know.

Hey [name] why don/t you come back here for one second
over there.

I have a couple more quick question/s.

How did Tim/s friend help him fix his shoe/s?

paint.

paint. Ooh.

and they gOt on and I don/t know.
got paint on.

Ok. A couple more quick question/s.

 

*WHI

 

CONNOR]

89

Oh. I see. What happen/ed to Tim/s mom?
She step in mud too.

Oh. That/s right. You told me.
When did they go to the party?

after they clean/ed up the stair/s and Chang/ed their shoe/s.
then they went to the party.

 

 

9O

TRANSCRIPT SUMMARY

Total Utterances 20
Statements (.) ' 20 100.00
Exclamations (I)
Questions (?)
Intonation Prompts (~)
Abandoned Utterances (>)
Interrupted Utterances (‘)

 

 

Nonverbal Utterances .
unintelligible Utterances
Responses to Questions
Yes/No Responses to Questions

Responses to Intonation Prompts
Imitations

One-word Utterances

Utterances with Mazes

Utterances with Pauses

Utterances with Omissions
Utterances with Overlapping Speech
Utterances with Word Codes
Utterances with Utterance Codes

Analysis Set (C&I verbal Utts)
Statements (.)
Exclamations (I)
Questions (?)
Intonation Prompts (~)

NM
01000000 H0000 00000 “000000 P000000 00000
5.:
"I
0
0

Responses to Questions
Yes/No Responses to Questions

Responses to Intonation Prompts
Imitations

One-word Utterances

Utterances with Mazes

Utterances with Pauses

Utterances with Omissions
Utterances with Overlapping Speech
Utterances with Word Codes
Utterances with Utterance Codes

15.00

Other
Between—Utterance Pauses: 0

 

91

Analysis Set: C&I Verbal Utts

WORD AND MORPEEME SUMMARY

Child
Analysis Total
Set Utterances

MLU in Words 5.80 5.80
MLU in Morphemes 6.65 6.65

Brown's Stage . Post V Post V
Expected Age Range (mos.) 41 - -- 41 - --
TTR . 0.46 0.46
No. Diff. Word Roots 53 53
Total Main Body Words 116 116
Total Maze Words 7 7

No. Omitted Words 0 0

MLU note: expected age ranges are the values within one standard deviation
of predicted values for that linguistic stage. based'on a linear fit of a
sample of 123 upperdmiddle class children (17-59 months of age) from Madison,
Wisconsin (Miller and Chapman. JSBR, 24, 1981.'154-161). They are irrelevant
for adults and for transcripts without bound morpheme coding.

 

 

 

W

WORD ROOT TABLE
Table Expanded by Bound Morphemes and Codes
C&I verbal Utts
Main Body
lst Speaker

A Total Expanded Total Expanded

AFTER
ALL
AND 1
ANGRY
BECAUSE
BY
CHANG
CHANG/fl!
CLEAN
CLEAN/ED
DID
DIDN
DIDN/T
DON
DON/T
DOWN
DRESS
FOOTPRINT
POOTPRINT/S

«Mum

mu HHWHbNHPm P HHH N HH H HHHHHmHH
NU HOHUPméOO N 000 N ab 0 opoonooo

w

v

'5';
HmHNUHNH
pwouowoo

SHOE
SHOE/S 1 4

STAIR/ S
STEP
STEP
STEP / S
STEPP
STEPP /
STILL
THE
THEIR
THEN
THEY
TIMMY

N

HHI—‘HH HHHNquput-H

93

NHNN

HOHHH HOOU‘ONHONO 0

000°

 

 

 

APPENDIX M

APPENDIX M
Data Form
Subject number:
Retelling:
I Number of words:
I Number of different words:
I Type Token Ratio (TTR):
I Mean Length Utterance (MLU):
I Number of Utterances:
I Syntactic Complexity: (Ratios)
I Incomplete sentences:
I Complete sentences:
I Simple sentences:
I Simple Elliptic Sentences:
I Complex sentences:
I Complex Elliptic Sentences:
I Elaborated sentences:
Comprehension:
I Number of words:
I Number of different words:
I Type Token Ratio ('I'I'R):
I Mean Length Utterance (MLU):
I Number of Utterances:
I Syntactic Complexity: (Ratios)

Incomplete Sentences:
Complete sentences:

Simple sentences:

Simple Elliptic Sentences:
Complex sentences:
Complex Elliptic Sentences:
Elaborated sentences:

94

W’“ -lia Q's-gm

 

APPENDIX N

APPENDIX N

Syntax Rules
INCOMPLETE COMPLETE (any can be elaborated)
Elliptic Non-Elliptic \
Simple Complex Simple Complex
Rules Regarding Syntax:

All utterances are looked at for structure only. The inappropriate responses that a child may give
to the retelling or comprehension tasks are ignored. Only the syntactic structure will be analyzed.

Complete Sentences:

SimeleScntcnses;

~

Simple sentences contain a subject and a verb. They are considered complete. Some
examples of simple sentences are:

0 “he go to the party,” “I forget,” “You should have stayed.”

A special case occurs when the child produces what appears to be a complex utterance, but
is missing both noun and verb forms in the first part of the utterance. If neither noun or verb
form is present in the first portion of the clause, but a complete sentence is found in the
second part of the utterance, consider this utterance to be a simple complete sentence. This
is true even if the words following the first part of the utterance appear to be a complex
sentence. An example of this would be, Examiner: “Where was Timmy go/ing?” Child:
“party, and they go right now.” The child produces an incomplete utterance by saying party
(both noun and verb forms are absent), but continues with a simple complete sentence
following it. Therefore, score this as a complete simple sentence.

Wm

~

Complex sentences contain at least two verb forms and are considered complete. It can
consist of two independent clauses joined by a conjunction. Examples of conjunctions are:
0 after, and, as, because, but, if, or, since, so, then, until, while

- An example of this is, “he go to the store and he eat vegetables.”

95

96

Infinitive forms of verbs automatically make the utterance a complex one. An example of
this is “He had to go bed.” This is a complex complete sentence.

Ell' . S _

Elliptic sentences contain at least one noun form or one verb form. Therefore, either the
subject is present in the utterance, or the predicate is present. Whichever part of speech that
is missing should be clearly stated in the utterance immediately preceding the elliptic
utterance.

Elliptic sentences are considered complete.

Elliptic sentences have part of the answer or topic feimd in the utterance immediately
preceding it. For example, the examiner may ask, “Where was Tim go/ing?” If the child
responds, “go to a party,” this is considered a simple elliptic sentence. The subject Tim has
been clearly stated in the examiner’s utterance right before the child’s response, and the
child has already stated a verb form. Therefore, the child has at least one part of speech
(noun or verb-subject or predicate) found in his utterance. If the child only stated “to a
party,” this would not be considered a complete sentence because he has not said a verb
form or noun form. Both are missing, causing the sentence to be incomplete. Another
example may be the child saying, “he brush his shoes and comb his hair.” In this sentence,
the child had omitted the subject in the second part of the utterance. The subject is clearly
stated in the first part of the utterance though as he. Therefore, this would be another
example of an elliptic sentence (complex- for rules governing simple and complex, read
below).

For elliptic sentences that are missing the predicate form and rely on the preceding
utterance, the complexity of the elliptic form is based upon whether the implied verb is
simple or complex from the preceding utterance. For example: Examiner: “Who help/ed
Tim to fix his shoe/s?” Child: “Amy” This is a complex complete elliptic sentence. The
child states the noun form only, making his reply an elliptic one. The verb form in the
preceding utterance clearly is complex (because of the infinitive form), making this a
complex elliptic sentence.

Incomplete Utterances:

~

If a sentence is incomplete, do not bother to determine if they are elaborated or not. Just
mark them as incomplete and continue to the next sentence.

Answers to comprehension questions or utterances not containing noun or verb forms are
incomplete. Therefore, a child saying, “mud/dy” or ‘ eah” would not have said a complete
sentence, even if his reply made sense semantically.

Elaborated Sentences:

Elaborated sentences contain bound morphemes, adjectives, lexical markers, and
conjunctions. An example of this is, “and I did,” because the child used a conjunction in
his utterance. Another example would be, “He got shoe/s mud/dy.” In this sentence the
bound morphemc on shoes and muddy elaborate the sentence, as well as the adjective
muddy.

APPENDIX 0

 

APPENDIX 0

SYNTAX FORMS

_mcu_mo~ % nu

 

 

gun-:8

58:. 2o magma—o

mi. m5. 0

 

m... 0 Q... o

0053

mi.

002: ex

@32an U

 

 

 

 

maze—Sm

 

 

 

 

 

.a_a_a

 

 

 

 

 

 

 

A
ololaialmr w M ..

 

 

 

444

 

 

#uN-I

 

d
0"

 

3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

nos as we: m5. o

 

Q6...“ was...

mi? 02: ox

 

 

 

alpaca. 88$

..~.mo$

m...

 

 

$8:

0 Con.
0.8.x.

OX

@3038.—

 

 

leg

 

97

98

i;

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mew—mo. a PM 93328 __:83Ww.o Ooamfifi m... =0 02: e. mica—23
_ a. N\ .
mm§==m im# a\ N \ x
u .< x .\ x .\
A \\ \ V -\\
M ’\ \ —\\ P\\
o \ ..\ ,\ .\
z .\ \ \ \
is p\ .\ .\
b_
3
la—
4»
lm_
4a
3
3 u
a
no
no
u:
we -
nu—
ma
nu
Egan? EEK 93 o m... :0 9:. e me: 003 9. m... .5003 9. $988.8

 

 

 

SECOND PERSON RELIABILITY CHECK FORM

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

99

 

.mhwo.oo;nw c.8338 .sooamoa 003 26 mi. 9:. a m... 0%.: a mi. 003 ax mafia ooamox @3222.

 

 

 

 

 

 

0035:2683

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_a
A—L—L“-A_L_AAAL

 

 

 

44.4

 

Eve-soccer“

 

‘
d
‘

 

.a
A
—L

 

.m

 

 

 

 

 

 

.5

CD
dd—L—A—AAA

a...
4444—...

 

 

 

 

 

 

 

 

 

 

€831.20, 9.312.. was. m_sm...l,m_.§_o 23% 3.30312. mesa coats. mega

 

 

 

 

_ 5.8.x. Immhooww 8.8.x. 98$ ”900$ 900$ mmbofi.

 

 

 

 

 

 

 

 

 

BIBLIOGRAPHY

American Speech-Language—Hearing Association. (1985, May).
Guidelines for identification audiometry. ASHA, 49-52.

Applebee, A. (1978). Ine_thld;s_§gngept_gj_§tgry. Chicago:
University of Chicago Press.

Beagle-Roos, J., & Gat, I. (1983). The specific impact of
radio and television on children’s story comprehension.
Qeurnal_ef_Educatienal_Eexchelesx. 75, 128-137.

Bloom, L. (1996). The integration of expression into the
stream of everyday activity. Lecture at Michigan State
University’s Symposium: Movement and Action: Links to
Intelligent Behavior, East Lansing, Michigan. March,
1996.

Bloom, L. & Beckwith, R. (1989). Talking with feeling:
Integrating affective and linguistic expression in

early language development. Ceenition_and_EmotienJ 3.

313-342.

100

101

Brown, R. (1973). A First Language. Cambridge, MA: Harvard

University.

Culatta, B., Page, J.L. & Ellis, J. (1983, April). Story

retelling as a communicative performance screening

tool. n e e

Schools. 14. 66-74.

Dollaghan, C.A., Campbell, T.F., & Tomlin, R. (1990,

August). Video narration as a language sampling

context. Qcurnal_9f_3eeech_and_8earins_niserders.

55,
582-890.

Gibbons, J., Anderson, D.R., Smith, R., Field, D.E., &

Fischer, C. (1986). Young children's recall and

reconstruction of auditory and audiovisual narratives.

Child_Dexelcnment. 57. 1014-1023.

Glenn, C.G. (1978). The role of episodic structure and

story length in children's recall of simple stories.

Qcnrnal_cf_yerbal_Learnins_and_yerbal_aehaxior. 17.
229-247.

Hamilton, M. and Shroyer, J. (1996). Comparison of
Storytelling and Retelling in Preschoolers. A poster

session presented at the American Speech-Hearing-

102
WI 78! 341-346-

Hedberg, N.L., & Stoel-Gammon, C. (1986). Narrative
analysis: Clinical procedures. Igpigs_in_Language
Disorders, 7, 58-69.

John, V.P. and Berney, T.D. (1967). Analysis of story
retelling as a measure of the effects of ethnic content
in stories. Final Report, Office of Economic
Opportunity, Project # 577.

John, V.P., Horner, V.M., & Berney, T.D. (1970). Story
retelling: A study of sequential speech in young
Children. In H. Levin & J.P. Williams (Eds.), Basic
fitnd1e5_in_3eaging. New York: Basic Books.

Liles, B. (1993, October). Narrative discourse in children
with language disorders and children with normal
language: A critical review of the literature. figurnal
W, 36, 868-882.

Mandler, J.M., Scribner, S., Cole, M., & DeForest, M.
(1980). Cross-cultural invariance in story recall.
Child.D§l§lQDm§nL, 51, 19—26.

Merritt, D.D., & Liles, 8.2. (1989, August). Narrative
analysis: Clinical applications of story generation and
story retelling. flgnrnal_gf_fipeegh_and_H§a£1na
Disorders, 54, 438—447.

Miller, J. (1981). W

Fr; #1.. _
__ f m

 

103

(1980). Cross-cultural invariance in story recall.
Child_Derelonment, 51, 19-26.

Merritt, D.D., & Liles, B.Z. (1989, August). Narrative
analysis: Clinical applications of story generation and
story retelling. Qoaraal_of_§eeech_and_nearing
Qiggggggg, 54, 438-447.

Miller. J. (1981). Assessiag_Language_£rgdngtion_in_
Children1_Exnerimental_2rocedures. Baltimore:
University Park Press.

Miller, J., & Chapman, R. (1991). Systematig_3naly§i§_gf_

Language_Transcriets. Madison: University of

Wisconsin.

Morris-Friehe, M.J., & Sanger, D.D. (1992). Language
sampling using three story elicitation tasks and
maturation effects. Qggrngl_gfi_ggmmunigatign
Diggxdgxfi, 25, 107-124.

Nelson, N.W. (1993). Ch1ldhggd_Langgage_Di§grdgrg_1n_
Contexti_Infang¥_Ihroggh_Adolescence. New York:

MacMillan Publishing.
Paul, R., a Smith, R.L. (1993). Narrative skills in 4—year-

olds with normal, impaired, and late-developing

language. Jogrnal_of_Speech_and_fiearing_flesearch. 36.

 

104
Esyghglggy, 21, 350-356.
Ripich, D.N., & Griffith, P.L. (1988). Narrative abilities
of children with learning disabilities and nondisabled

children: Story structure, cohesion, and propositions.

Journal_of_Learning_Disabilities. 21. 165-173-
Salomon. G. (1979). Interacticn_cfamedial_ccgnitionl_and_

lgarning. San Francisco: Jossey-Bass.

 

Schaefer, R.L., & Anderson, R.B. (1987). fitndent_£fiitign_
Qf_Minitah. Addison Wesley Publishing Co. & Benjamin
Cummings Publishing Co., Raleigh, Massachusetts.

Schneider, P. (1996). Effects of pictures versus orally
presented stories on story telling by children with
language impairment. American_gonrnal_of_Saeeeh:
Language_£atholog¥. 5. 86-96-

Shipley, K. and McAfee, J. (1992). Afifiggsmgnfi_in_§ngegh;
Langnage:2athclog¥i_A_Besonrce_Manual. San Diego=

 

Singular Publishing Company.
Stein, N.L., & Glenn, C.G. (1979). An analysis of story

comprehension in elementary school children. In R.O.

Freedle (Ed.), Nea_Directions_in_Discourse_Rrocessinga

Yol1_Zi_Adrances_in_Disccnrse_£rocessingp Norwood, NJ:
Ablex.

Stockman, I. Personal interview. Feb. 1996.

Stockman, I. (1996). The promises and pitfalls of language

105
W. 5, 86-96.

Shipley, K. and McAfee, J. (1992). Asggsgment_ig_§pe§gh;
WW1. San Diego=
Singular Publishing Company.

Stein, N.L., & Glenn, C.G. (1979). An analysis of story

comprehension in elementary school children. In R.O.

Freedle (Ed.), W
Waging Norwood, NJ:

Ablex.

Stockman, I. Personal interview. Feb. 1996.

Stockman, I. (1996). The promises and pitfalls of language
sampling analysis as an assessment tool for linguistic
minority children. Langgage‘_§peegh_gnd_fle§;1ng
W 27. 355-366.

Sutter, J.C.L., & Johnson, C.J. (1995). Advanced verb form
production in story retelling. Jggrnal_gfi_fipeegh_gnd
hgg;igg_3gggg;gh, 38, 1067-1080.

Tomasello, M. (1995). Pragmatic context for early verb
learning. In M. Tomasello & W. Merriman (Eds.), Bgyggg

Na in - ' ' ' ' ' f

106

yggbg. Hillsdale: Lawrence Erlbaum Associates.

Whiren, A. Personal interview. 7 Aug. 1997.

Wren, C. (1985). Collecting language samples from children
with syntax problems. Lan a
Serrices_in_§chools. 16, 83-102.

Zimmerman, T., Steiner, V., and Pond, R. (1979). Erggghggl
nggggg§_§gglg (Rev. ed.). Columbus, Ohio: Charles E.

Merrill.

 

 

 

Ausmngun
93mg uefituolw

AHVHBI‘I

 

 

MICHIGAN STATE UNIV. LIBRARIES
ll11111111111111111111111111111111
31293017165279