ELSEVIER

NARRATIVE SKILLS FOLLOWING TRAUMATIC BRAIN INJURY IN CHILDREN AND ADULTS KATHLEEN R. BIDDLE Eliot-Pearson Department of Child Study, Tujqs University

ALLYSSA McCABE University of Massachusetts-Lowell and Harvard Graduate School of Education

LYNN S. BLISS Department of Communication Disorders and Sciences, Wayne State University

Personal narratives serve an important function in virtually all societies (Peterson & Mc- Cabe, 1991). Through narratives individuals make sense of their experiences and represent themselves to others (Brunet, 1990). The ability to produce narratives has been linked to ac- ademic success (Feagans, 1982). Persons who have sustained a traumatic brain injury (TBI) are at risk for impaired narrative ability (Dennis, 1991). However, a paucity of information exists on the discourse abilities of persons with TBI. This is partly due to a lack of reliable tools with which to assess narrative discourse. The present study utilized dependency analy- sis (Deese, 1984) to document and describe the narrative discourse impairments of children and adults with TBI. Ten children (mean age 12;0) and 10 adults (mean age 35;2) were compared with matched controls. Dependency analysis reliably differentiated the discourse of the individuals with TBI from their controls. Individuals with TBI were significantly more dysfluent than their matched controls. Furthermore, their performance on the narrative task revealed a striking listener burden. © 1996 Elsevier Science Inc.

Educational Objectives: (1) Readers will acquire knowledge and understanding of the im- portance of assessing spontaneous personal narrative ability of individuals with traumatic brain injury; (2) develop knowledge of the use of dependency analysis to assess personal narrative skill; (3) understand the relationship between traumatic brain injury and functional communication outcomes in children and adults with TBI; and (4) recognize potential for developmental differences in language outcomes for children versus adults with traumatic brain injury.

IN TR OD U CTION Traumatic brain injury (TBI) is a complex impairment that affects linguistic, cognitive and pragmatic functioning. Previous research has shown that tradi-

Address correspondence to Allyssa McCabe, Department of Psychology, University of Massachu- setts-Lowell, 1 University Avenue, Lowell, MA 01854

J. COMMUN. DISORD. 29 (1996), 447-469 © 1996 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010

0021-9924/96/$ 15.(30 SSDI 0021-9924(95)00038-F

448 BIDDLE et al.

tional language measures do not adequately describe the deficits of individu- als with TBI (Chapman et al, 1992; Jordan, Cannon, & Murdoch, 1992; Mc- Donald 1992, 1993). They tend to score within normal limits on traditional tests of aphasia, despite the presence of significant communicative impair- ments (Chapman et al., 1992; Jordan et al., 1992; Jordan & Murdoch, 1990; Levin, Benton, & Grossman, 1982; McDonald, 1993; Mentis & Prutting, 1987; Wyckoff, 1984). Their recovery of linguistic skills, as shown by con- ventional language measures, masks a more complex and functional disorder of language usage (Dennis & Barnes, 1990; Dennis & Lovett, 1990; Ylvi- saker, 1986, 1993).

Language assessment measures of aphasia are generally ineffective as a means of studying the deficits associated with TBI for several reasons. First, the lesions associated with TBI often involve frontal lobe damage (Katz, 1992). The frontal lobes are involved in the planning, execution and control of many behaviors including language functioning (Alexander, Benson, & Stuss, 1989). Tests of aphasia do not assess this aspect of language ability (Ylvi- saker, 1986). In fact, formal language testing may provide cuing and structure from which individuals with TBI may benefit, thus, giving them inflated scores (Ylvisaker, 1986). In addition, language tests, which emphasize decon- textualized skills and ignore complex cognitive processing demands, are un- likely to reveal communicative impairments that individuals with TBI exhibit (Ewing-Cobbs, & Fletcher, 1987; Wyckoff, 1984; Ylvisaker, 1993). As a con- sequence, standard assessment procedures, based on models and tests for aphasia, are limited in their usefulness in planning intervention or predicting functional outcomes for individuals with TBI (Benton, 1987; Erlich, 1988; Wyckoff, 1984; Ylvisaker, 1986, 1993).

A more promising avenue of research lies in the realm of narrative dis- course. Narrative skills tax the linguistic, cognitive and communicative abili- ties of speakers (Peterson & McCabe, 1983). The executive functions of plan- ning and control are required for an individual to construct a coherent narrative. Narrative analysis accounts for the dynamic interaction between language and cognition and is relevant to both children and adults with TBI (Chapman et al., 1990; Hartley & Jensen, 1991, 1992; Jordan et al., 1992; Liles, Coehlo, Duffy, & Zalagens, 1989; McDonald, 1993; Marsh & Knight, 1991; Mentis & Prutting, 1987, 1991). In Table 1, a summary of the research describing the narrative and discourse abilities of adults and children with TBI is presented. This review will highlight some of the major findings pertinent to TBI.

The discourse of adults with TBI has been described as reduced in coher- ence, completeness and fluency (Hartley & Jensen, 1991; 1992; Marsh & Knight, 1991, Liles, et al., 1989; Mentis & Prutting, 1991). Their fictional nar- ratives are generally incomplete, characterized by a reduced number of com- plete idea units, episodes, subjects and verbs (Glosser & Deser, 1990; Hartley

NARRATIVE SKILLS 449

& Jensen, 1991, 1992; Liles, et al., 1989; Mentis & Prutting, 1991). Irrelevant and inaccurate information may also be present (McDonald, 1993; Hartley & Jensen, 1991, 1992). Individuals with TBI often do not supply critical infor- mation for the listener to understand the text (McDonald, 1993). This defi- ciency may be the result of impaired self-monitoring skills (Hartley & Jensen, 1991).

Cohesion ability has been one area in which agreement has not been ob- tained regarding the narrative abilities of adults with TBI. Mentis and Prutting (1987) and Hartley and Jensen (1991, 1992) documented fewer cohesive ties in the discourse of adults with TBI than a control group. In contrast, Glosser and Deser (1990) and McDonald (1993) found no evidence of impairments in cohesion with the adults in their investigation. Liles and her colleagues (1989) reported that cohesive ability varied with discourse task; narrative retelling was not associated with cohesion impairments while spontaneous narrative tasks revealed a reduced number of cohesive ties in comparison to a control group. Different methods of elicitation and analyses may have resulted in these conflicting data. In contrast, manner of production by adults with TBI has been consistently shown to be impaired. Deficiencies in rate, fluency and redundancy have been documented (Hartley & Jensen, 1991, 1992; Marsh & Knight, 1991; McDonald, 1993). In particular, increased hesitations, pauses, and false starts have been reported (Hartley & Jensen, 1991, 1992).

There is a paucity of research describing the discourse and narrative abili- ties of children with TBI. Only two studies have been reported in the litera- ture. Chapman and her colleagues (1990) employed a retelling paradigm in or- der to investigate the narrative abilities of children and adolescents who had sustained mild, moderate or severe TBI. Their stories were analyzed for lan- guage and information structure as well as information flow. Children with mild and moderate TBI did n o t show narrative deficits while subjects with se- vere injury were impaired in their abilities to present critical story information and delineate episodes in narratives. In contrast, Jordan, Murdoch and Buttsworth (1991) did not find differences in the production of story grammar and cohesive elements when their subjects with TBI were asked to tell a story about a doll. However, all the subjects produced a few complete episodes, thereby limiting the data for analysis. The task limitations in this investigation may have contributed to their results. These two studies differ in their method- ology, one a retelling task and one a spontaneous generation. The variations in elicitation may have resulted in the differences in the results. In addition, al- though Chapman and her colleagues (1990) did include a working memory measure in their investigation, it is possible that some of the differences they found may have been due to memory rather than discourse ability, since list recall may place different demands on working memory than the recall of a complex adventure story. Their procedure may not have tapped children's ability to organize and spontaneously produce complex narratives. The task of

Tabl

e 1.

Sum

mar

y of

Stu

dies

of D

isco

urse

and

Nar

rativ

e Sk

ills

of A

dults

and

Chi

ldre

n w

ith T

BP

X A

ge

Stud

y Sa

mpl

e (y

ears

) M

etho

ds

Ana

lysi

s R

esul

ts

&

Adu

lt st

udie

s Er

lich,

198

8 n

= 10

29

.5

Nar

rate

in

resp

onse

to

Con

tent

uni

ts (

CU);

# sy

llabl

es/

Few

er C

U/m

in b

TSI

= 7-

36 m

os.

"Coo

kie

Thef

t m

inut

e; C

U's

/min

ute

Red

uced

eff

icie

ncy

pict

ure

Red

unda

ncy

inde

x D

escr

ibe

fam

ily

Synt

actic

com

plex

ity

Des

crib

e a

wor

k C

oher

ence

(lo

cal a

nd g

loba

l)

expe

rien

ce

Coh

esio

n (H

allid

ay &

Has

an,

1976

Si

ngle

wor

d pr

oduc

tion

Glo

sser

&

n =

9 24

.3

Des

er,

1990

TS

I =

2-14

mos

.

Har

tley

&

n =

11

26.7

N

arra

te c

omic

stri

p Je

nsen

, 19

91

TSI

= 1-

20 m

os.

Ret

ell o

ral s

tory

Te

ll ho

w to

buy

gr

ocer

ies

Prod

uctiv

ity

Qua

ntity

R

ate

Con

tent

Ta

rget

con

tent

uni

ts

Inac

cura

te u

nits

Pr

oble

ms

of re

fere

nce

Coh

esio

n (H

allid

ay &

Has

an,

1976

)

Om

issi

ons

of s

ubje

ct

and

verb

h Im

pair

ed lo

cal a

nd

glob

al c

oher

ence

b M

ore

lexi

cal e

rror

s b

No

diff

eren

ce in

co

hesi

on

No

diff

eren

ce in

sy

ntac

tic c

ompl

exity

TB

I fe

wer

coh

esiv

e tie

s b

TBI

few

er ta

rget

co

nten

t uni

ts b

TBI

mor

e in

accu

- ra

te c

onte

nt u

nits

b TB

I sl

ower

rate

of

spee

ch, l

ess

spee

ch b

TBI

high

er p

erce

ntag

e of

dys

flue

ncie

s b

TBI

mor

e pr

oble

ms

with

ref

eren

ce

t-

Har

tley

&

Jens

en,

1992

Lile

s, Co

elho

, D

uffy

, &

Zala

gens

, 19

89

Mar

sh &

K

nigh

t, 19

91

Mcd

onal

d,

1993

n=ll

TS

I =

1-20

mos

.

n=4

TSI

= 5-

14 m

os.

n =

18

TSI

= 20

-146

m

os,

n=2

TSI

= 36

mos

. an

d 10

8 m

os.

26.7

23.0

30.4

29

42

Nar

rate

com

ic s

trip

Rete

ll or

al st

ory

Rete

ll fil

mst

rip

Gen

erat

e st

ory

for

pict

ure

Enga

ge in

5 p

robl

em

solv

ing

task

s with

kn

own

pers

on

10 m

in. c

onve

rsat

ion

with

stra

nger

D

escr

ibe

proc

edur

e to

blin

dfol

ded

pers

on

Prod

uctiv

ity

Qua

ntity

Ra

te

Cont

ent

Targ

et c

onte

nt u

nits

In

accu

rate

uni

ts

Prob

lem

s of r

efer

ence

Co

hesi

on (H

allid

ay &

Has

an,

1976

)

Sent

ence

pro

duct

ion

Cohe

sion

(Hal

liday

& H

asan

, 19

76)

Stor

y gr

amm

ar (S

tein

&

Gle

nn,

1979

) Be

havi

oral

ly R

efer

ence

d Ra

ting

of In

term

edia

te S

ocia

l Ski

lls

(BRI

SS)-

--6

verb

al s

cale

s

7-po

int r

atin

g sc

ale:

repe

ti-

tiven

ess,

orga

niza

tion,

eff

ec-

tiven

ess,

clar

ity an

d de

tail

Cohe

sion

(Hal

liday

& H

assa

n,

1976

) Pr

opos

ition

al (i

nfor

mat

iona

l co

nten

t) to

tal,

repe

ated

, or

igin

al, e

ssen

tial,

irrel

evan

t

Thre

e pr

ofile

s:

Con

fuse

d--

Hig

h #

clar

- ity

pro

blem

s hig

h #

inac

cura

te co

nten

t uni

ts Im

pove

rish

ed -

- m

arke

d de

nsity

in

prod

uctiv

ity an

d co

nten

t, lo

w c

ohes

ion

Inef

fici

ent-

--ou

tput

>

cont

rols

, hig

h #

dysf

luen

cies

R

etel

l--n

o di

ffer

ence

s G

ener

ate-

-TB

I few

er

cohe

sive

ties

TB

I few

er c

ompl

ete

epis

odes

La

ck fl

uenc

y, c

lari

ty

cont

inui

ty

Less

app

ropr

iate

in u

se

of la

ngua

ge a

nd

deliv

ery

of sp

eech

Re

petit

ive,

dis

orga

nize

d,

inef

fect

ive

Irre

leva

nt p

ropo

sitio

ns ~

Out

side

rang

e fo

r #

prop

ositi

ons

No

diff

eren

ce in

co

hesi

on

Z t"

t"

cont

inue

d 4~

Tabl

e 1.

Con

tinue

d I,

o

X A

ge

Stud

y Sa

mpl

e (y

ears

) M

etho

ds

Ana

lysi

s R

esul

ts

Men

tis &

"n

= 3

32

C

onve

rsat

ion

with

C

ohes

ion

(HaU

iday

& H

asan

, Fe

wer

ties

in th

e Pr

uttin

g,

TSI

= 12

-44

part

ner

1976

); ad

ded

"inc

ompl

ete

narr

ativ

e b

1987

m

os.

Des

crib

e yo

ur w

ork,

co

hesi

ve"

Hig

h %

inc

ompl

ete

how

to p

lay

a sp

ort

ties b

ch

ange

a ti

re, o

r ba

ke a

cak

e M

entis

&

n =

1 24

Te

n sa

mpl

es:

Topi

c m

anag

emen

t ab

ility

N

onco

here

nt c

hang

es

Prut

ting,

TS

I -

58 m

os.

6 C

onve

rsat

ions

, To

pic

and

subt

opic

s; i

ntro

duc-

A

mbi

guou

s un

its

1991

co

ncre

te a

nd

tion,

rei

ntro

duct

ion,

rel

ated

, In

com

plet

e un

its

abst

ract

no

ncoh

eren

t or n

ew in

for-

R

educ

ed c

oher

ence

4

Mon

olog

ues,

m

atio

n an

d co

ntin

uity

co

ncre

te a

nd

abst

ract

(e.

g.,

Hal

low

een,

gre

ed)

Chi

ldre

n St

udie

s n

= 19

12

.6

Ret

ell o

ral s

tory

La

ngua

ge s

truct

ure

Gro

up e

ffec

t for

C

hapm

an, e

t al.

TSI

= 12

-26

Info

rmat

ion

stru

ctur

e (L

abov

, la

ngua

ge a

nd i

nfor

- 19

92

mos

. 19

72;

Van

Dijk

, 19

85)

mat

ion

stru

ctur

e b

Info

rmat

ion

flow

(i.e

., Se

vere

TB

I:

effi

cien

cy)

Less

lan

guag

d'

Few

er e

ssen

tial

stor

y co

mpo

nent

s b of

set

- tin

g an

d co

mpl

icat

ing

actio

n Fa

ilure

to s

igna

l new

ep

isod

e b

Om

it es

sent

ial

actio

n in

form

atio

n b

Den

nis

&

n =

33

9.2

Barn

es,

1990

TS

I =

10-1

11

mos

.

Test

of l

angu

age

com

pete

nce,

ex-

pa

nded

edi

tion

(Wiig

& S

ecor

d,

1989

) A

dditi

onal

non

- di

scou

rse

mea

sure

s

Subt

est s

core

s C

ompo

site

sco

res

Inte

rcor

rela

tion

of su

btes

t sco

res

Corr

elat

ion

of su

btes

t sco

res

to

non-

disc

ours

e m

easu

res

Jord

an,

n =

20

11.6

Te

ll a

stor

y ab

out a

St

ory

gram

mar

(Ro

th &

M

urdo

ch &

TS

I =

14-2

8 do

ll fig

urin

e Sp

ekm

an,

1986

) Bu

ttsw

orth

, m

os.

Coh

esio

n (H

allid

ay &

Has

an,

1991

19

76)

Prob

lem

s w

ith g

loba

l co

nten

t b

No

diff

eren

ce in

in

form

atio

n fl

ow

Littl

e di

ffer

ence

bet

wee

n m

ild/m

oder

ate

79%

at l

east

one

sub

test

sc

ore-

-im

pair

ed

rang

e 67

% c

ompo

site

sco

re--

im

pair

ed r

ange

H

ighe

r in

terc

orre

latio

n be

twee

n m

akin

g in

fere

nces

and

re-

cr

eatin

g se

nten

ces

Wor

king

mem

ory

pred

icts

inf

eren

cing

ab

ility

b V

ocab

ular

y kn

owle

dge

pred

icts

met

apho

r co

mpr

ehen

sion

and

ab

ility

to u

nder

stan

d am

bigu

ity b

No

sign

ific

ant

diff

eren

ces

Z 7>

< rrl r-

"TB

I den

otes

trau

mat

ic b

rain

inju

ry (

this

ter

m is

bei

ng u

sed

inter

chan

geab

ly w

ith C

H/,

the

termi

nolo

gy w

hich

was

pre

vious

ly us

ed I

o de

scri

be in

divid

uals

with

a c

lose

d he

ad i

njur

y);

TSI d

enot

es ti

me

sinc

e in

jury

. b D

enot

es s

igni

fican

t fin

ding

.

454 BIDDLE et al.

Jordan and her colleagues (1991) may have been too abstract for the children since fantasy narratives are rarely used by children in their spontaneous dis- course (Preece, 1987).

The purpose of the present investigation was to expand on the literature con- ceming the narrative discourse abilities of children and adults with TBI. This study differs from previous ones by using spontaneous personal narratives for assessment as well as a unique discourse analysis approach. Personal narratives merit study because they are used so frequently in social (Bruner, 1990; Preece, 1987) and educational (Feagans, 1982) settings. Competent discourse ability is required in daily life. Spontaneous generation tasks provide a more accurate in- dex of functional communication ability than retelling tasks (Liles, 1993). Be- cause of the increased demands for spontaneous organization and production of discourse in this investigation, additional differences in the narrative discourse abilities of individuals with TBI versus control sugjects should be apparent. Dependency analysis (Deese, 1984) has never been used to study the discourse abilities of individuals with TBI. This approach will identify and quantify ele- ments which may be impaired in the discourse of individuals with TBI, such as omission of critical information, redundancy and fluency.

M E T H O D

Subjects There were 40 participants in this study, 20 individuals with TBI and 20 con-

trol group members. The adults and children with TBI were matched on the ba- sis of age, gender, ethnicity, educational level, and region of the United States with a comparison group of individuals who had no history of head injury.

The 10 children with TBI who participated in this study included seven boys and three girls, ranging in age from 7;4 to 16;2 years. Their mean age was 12;0 years (SD = 3;2). These children were selected and recruited from a larger sample of children who are involved in a prospective study of injury impact being conducted by a medical center in the Northeastern United States. One criterion for inclusion was that the children had no premorbid history of learning or emotional difficulties. The time since injury for the children with TBI ranged from 22 to 46 months (M - 33; 1 months, SD = 8;8 months). Glas- gow coma scale (GCS) scores (Teasdale & Jennett, 1974), taken at the time of injury, were available for all but one of the children. Based on the GSC scores, seven of the 10 children were classified as having a mild to moderate injury (GCS > 8). Two of the children were in the severe TBI range (GCS > 8). The mean GCS score for the group was 9;6. Of the 10 children, five were receiving resource room support services. Three of those five children were also receiv- ing speech-language services. With the exception of one child, these children did not have any obvious physical handicaps.

NARRATIVE SKILLS 455

Ten children were in the comparison group. They ranged in age from 7; 1 to 16;2 years (M = 12;1, SD ~ 3;2). These children were of average intellectual abil- ity and were not receiving special educational services for any reason including giftedness. Every child received a small gift for participating in the study.

The five women and five men with TBI who participated in the study ranged in age from 20 to 52 years, with a mean age of 35;2 years (SD = 9.89). They rep- resented a range of educational levels from high school through graduate school. It was not possible to obtain the medical records of these individuals. However, all were within 12 months of the injury at the time of the interview.

The 10 adults in the control group ranged in age from 19 to 52 years (M = 34;4, SD = 9.98): All of the adult control participants were high school gradu- ates. Some of the adults had attended college or trade school.

Procedure The children's narratives were collected by the first author in a home setting. The narratives were elicited using the conversational map technique developed by Peterson and McCabe (1983). A generation task was selected because the goal of the study was to examine the participant's ability to spontaneously orga- nize discourse on demand with minimal contextual support (see Liles, 1993).

Using this procedure, the interviewer related a personal experience and then asked the participant if she/he had had a similar experience. Once the in- dividual began recounting that experience, the interviewer responded to the narrative in a manner designed to encourage elaboration (e.g., "Uh huh"). The interviewer did not evaluate the speaker's experience, nor did she request spe- cific details of the event. Each child received four narrative prompts includ- ing, "A pet does something funny," "A trip to an amusement park," "Getting lost," and "Being stung by a bee."

The adult narratives used in this study were collected and transcribed under the direction of the third author. Clinicians elicited the narratives using the same conversational map technique (Peterson and McCabe, 1983) used with the children. The story prompts were modified in order to make them appro- priate for adults. The adult prompts included "Getting lost," "A broken bone," "A favorite experience with a child," and "Having a car accident." For both the children and the adults, each narrative was recorded and was later tran- scribed for analysis.

Measures The narratives were analyzed according to the principles of dependency

analysis (Deese, 1984). This method has been described in detail with exam- ples by Peterson and McCabe (1983). Unlike High Point analysis (Labor, 1972), which assesses independent clauses and their informational content,

456 BIDDLE et al.

dependency analysis focuses on the story's syntax (Peterson and McCabe, 1983). It also provides quantitative measures of dysfluency and listener bur- den. Propositions were the units of analysis and the relationships between propositions illustrated the basic form of the narrative, that is, its complexity. For example the sentence, "I just got lost last week at the beach" was divided into the main proposition "I got lost," with dependent propositions "just," "last week," and "at the beach" as modifiers of "got lost."

Because length is an excellent index of narrative complexity (McCabe and Peterson, 1990), the longest narrative of each participant which was unre- hearsed and not a retelling was selected for analysis. The goal was to view the upper limits of each participant's performance. Each story was divided into propositions. Absolute counts and ratios were coded in the categories defined by Deese (1984) and described by Peterson and McCabe (1983). There was one additional category added to the analysis. The categories and examples of each are listed below.

Absolute Measures/Frequencies Explicit Propositions. All propositions provided by the speaker. Implicit Propositions. Information which had to be inferred by the listener.

Implicitness was coded when a definite article was used without a previ- ous referent. For example, "So we were walking around the lake," when a lake had never been mentioned. In this case the coder would add "There was a lake" to the narrative in parentheses. Partial implicitness was coded when a portion of a sentence was not supplied by the speaker. For exam- ple, "Went to the beach." The implicit propositions were divided into complete (subject and verb) or partial propositions.

Repeated Propositions. e.g., "It was so hot." "It was so hot." A repeated proposition could occur anywhere in the text and need not be verbatim.

Total Dysfluencies. False starts, internal corrections, and fillers were com- bined for a total dysfluency measure.

False Starts. e.g., "We it was s o . . . " Fillers. This was added to the list of absolute measures and included

pauses, hesitations (such as "um"), fillers, and colloquialisms (e.g., "like," "you know"), and mispronunciations (e.g., "supposes" for sup- posed).

Internal Corrections. e.g., " . . . nights, I mean weekends." Ratio Measures/Proportions

Listener Burden Ratio. The total number of implicit propositions divided by the total number of explicit propositions provided the listener burden ratio. Ratio measures were also calculated for the number of complete implicit propositions and the number of partial implicit propositions rela- tive to the total number of explicit propositions.

Redundancy Ratio. To calculate this ratio, the total number of repeated propositions was divided by the total number of explicit propositions.

NARRATIVE SKILLS 457

Table 2. Results of Dependency Analysis

TBI Controls F Value Outcome Measure X (SD) X (SD) (df 1, 18)

Explicit propositions 33.70 (29.12) 33.40 (18.10) 0.00 Implicit propositions 5.25 (3.71 ) 3.15 (2.20) 7.68*

Partial implicit 3.35 (3.42) 2.00 (1.95) 3,16 Complete implicit 1.90 (1.25) 1.15 (1.14) 3.95

Repeated propositions 1.30 (1.66) 0.15 (0.37) 9.50** Total dysfluencies 6.55 (7.55) 2.25 (2.02) 7.25*

False starts 1.50 (2.20) 0.20 (0.52) 9.66** Fillers 4.75 (5.5) 1.80 (1.60) 6.65* Internal correct 0.45 (0.69) 0.35 (0.64) 0.71

Listener burden radio (implicit/explicit) .230 (. 195) .097 (.072) 8.58**

Partial ratio .129 (. 125) .058 (.061) 4.98* Complete ratio .099 (. 104) .038 (.039) 5.87"

Redundancy ratio .035 (. 104) .010 (.031 ) 3.30 Dysfluency ratio .211 (.232) .068 (.045) 7.77*

False start ratio .036 (.042) .004 (.009) 11.61 ** Filler ratio .160 (2.02) .055 (.039) 6.13*

Intemal correct ratio .016 (.027) .007 (.020) 1.20

*p < .05. **p < .01.

Dysfluency Ratio. For this ratio the total number of false starts, internal corrections, and additional dysfluencies was divided by the total number of explicit propositions.

False Start Ratio. The total number of false starts was divided by the total number of explicit propositions for this measure.

Internal Corrections Ratio. This ratio was the total number of inernal cor- rections divided by the total number of explicit propositions.

Filler Ratio. To calculate this ratio, the total number of fillers was divided by the total number of explicit propositions.

The ratio measures provided an index of the coherence of the narrative in addition to its syntactic complexity. Higher scores than normal in dysfluency, redundancy, and listener burden would have implied a less coherent narrative.

Reliability of Narrative Measures Thirty percent of the data was coded by an independent rater who had re- ceived extensive training in dependency analysis. Reliability was calculated using a Pearson correlation for measures of total explicit propositions and to- tal implicit propositions. The results were an r = .99 (p < .001) for explicit propositions and r = .95 (p = .003) for implicit propositions. Disagreements about the coding were resolved through discussions.

458 BIDDLE et al.

R E S U L T S

On a variety of measures, the individuals with TBI were substantially less ar- ticulate than their peers. Dependency analysis differentiated the discourse of the TBI participants from that of the comparison group in a series of two way anovas using matched groups for the head injury control variable. Table 2 summarizes the results of the analysis and presents means and standard devia- tions for each group. As is standard procedure, a p value of .05 was chosen to determine the significance of the results. There were large standard deviations for the group with TBI on several measures. This was not the case for the comparison group.

Absolute Measures/Frequencies Total Explicit Propositions. On the measure of total number of explicit

propositions, individuals with TBI did not differ from members of the comparison group. In fact, the mean number of propositions generated by each group was nearly equal.

Total Implicit Propositions. There was a significant main effect for injury status on this count [F(1, 18) = 7.68, p < 05]. In their narratives, persons with TBI left out more information than their non-injured peers.

Complete Implicit Propositions. Individuals with TBI omitted complete propositions more often than their controls. However, group differences related to injury injury status were not significant.

Partial Implicit Propositions. Group differences were not significant for the absolute amount of partial information omitted in the narratives of in- dividuals with TBI as compared to the controls.

Repeated Propositions. Significant group differences demonstrated that TBI individuals repeated themselves more often than did individuals in the comparison group [F(1, 18) = 9.50, p < .01 ].

Total Dysfluencies. TBI individuals generated more dysfluencies than the control group. These differences were significant [F(1, 18) = 7.25, p < .05].

False Starts. Significant differences were noted in the number of false starts made by members of each group [F(1, 18) = 9.66, p < .01]. Per- sons with TBI produced more false starts than their matched peers.

Fillers. TBI individuals used more pauses, hesitations, and fillers than their non-impaired counterparts. The differences between the groups were significant [F(1, 18) = 6.65, p < .05].

Internal Corrections. Group differences were not significant for the num- ber of internal corrections made by the participants.

Ratio Measures/Proportions Listener Burden Ratio. There was a significant main effect for injury sta-

tus [F(1, 18) = 8.58, p < .01]. TBI individuals left more information im-

NARRATIVE SKILLS 459

plicit in their narratives relative to their explicit propositions than did members of the control group.

Complete Implicit Proposition Ratios. TBI participants omitted more complete propositions than members of the comparison group. The group differences were significant [F(1, 18) = 5.87, p < .05].

Partial Implicit Proposition Ratios. In addition to omitting complete prop- ositions, TBI individuals left out parts of propositions in their narratives more often than their non-injured matches. Again, the differences were significant [F(1, 18) = 4.98, p < .05)].

Redundancy Ratio. Group differences were not significant on this measure. Dysfluency Ratio. There was a significant main effect for injury status on

this measure [F(1, 18) = 7.77, p > .05]. TBI individuals were more dys- fluent relative to their total output of speech than their control group.

False Start Ratio. This measure of dysfluency also distinguished the TBI participants from their matched counterparts. There was a main effect for injury status which was significant [F(1, 18) = 11.61, p > .01].

Fillers Ratio. In their use of fillers, one type of dysfluency, TBI partici- pants were significantly different than their peers [F(1, 18) = 6.13, p > .05]. Individuals with TBI produced more fillers in their narratives than the comparison group members.

Internal Corrections Ratio. Group differences were not significant on this measure.

Age Effects. Significant main effects for age were found for the following variables, p > .05: explicit propositions [F(1, 18) = 4.70]; fillers [F(1, 18) = 8.57]; total dysfluencies [F(1, 18) = 8.91]. In other words, children produced more explicit propositions than adults. However, the children also used more fillers than the adults and their total number of dysfluen- cies was greater than the total for the adult participants. In addition, the children omitted more partial information from their stories than the adults, [F(1, 18)= 4.37,p = .05].

Interaction Effects. Significant interactions occurred on the absolute mea- sures of fillers [F(1, 18) = 4.59] and total implicit propositions [F(1, 18) = 4.46, p > .05]. Head injury had a more profound impact on chil- dren than on adults with respect to the number of fillers they used and the amount of information omitted. Children with TBI were vastly more dys- fluent than all other groups in the study and omitted substantially more information.

DISCUSSION The results of this investigation provide information concerning the narrative discourse abilities of children and adults with TBI. The findings will be dis- cussed first with reference to differences in the narrative abilities of children

460 BIDDLE et al.

and adults with TBI in comparison to their uninjured peers. Following that, the differences between the children and adults with TBI will be reviewed.

On 11 of the 17 dependent variables, the children and adults with TBI per- formed significantly worse than their matched controls. These findings sup- ported Dennis' (1991) assertion that children and adults are at risk for dis- rupted discourse ability following TBI. Dependency analysis identified those impairments and reliably distinguished the narratives of individuals with TBI from those of the comparison group. These quantitative differences are illus- trated in the examples of narratives in Table 3.

Interestingly, the children and adults with TBI did not differ from their comparison groups in their mean number of explicit propositions. Unlike the adult participants in Hartley and Jensens' (1991) study and contrary to the findings of Chapman et al. (1992) with children with severe TBI, the individu- als with TBI in this study produced almost exactly as much discourse as their matched peers. Despite this fact, however, the participants with TBI were sub- stantially less efficient and less effective in accomplishing their communica- tive goals. This study confirms previous findings of an inefficient style of dis- course among individuals with TBI (Erlich, 1988; Hartley & Jensen, 1992) and illustrates that discourse deficits persist past the initial injury period.

The individuals with TBI in this study did not appear to adequately monitor their narratives. In both the absolute measures and in proportion to their total output, individuals with TBI failed to include critical information in their nar- ratives. Sometimes the children and adults with TBI omitted partial informa- tion. Other times, however, they neglected to include entire units of informa- tion which were necessary in order to make sense of their narratives. Omissions in the narratives of individuals with TBI have been noted in the lit- erature (Mentis & Prutting, 1991; Glosser & Deser, 1990; Hartley & Jensen, 1991, 1992). However, this study extends those findings to demonstrate that individuals with mild to moderate TBI habitually omit information from their stories.

The gaps in the stories of the participants with TBI resulted in a striking lis- tener burden. That is, the onus was on the listener to make sense of the narra- tive. For example listeners may attempt to fill in the missing information. Al- ternatively, listeners may simply be left wondering just exactly what the speaker was attempting to convey. As Hartley and Jensen (1991) have hypoth- esized, it appeared that children and adults with TBI in this study were unable to adequately take the listener's perspective and did not deliver their discourse in an effective manner.

In addition to omissions, the individuals with TBI, like McDonalds' (1993) participants, made numerous repetitions. Although the redundancy ratio was not significant, it is worth noting that the average listener would not correct for length when hearing a story. Consequently, most listeners would likely find the discourse of these individuals overly repetitive. The results of the

NARRATIVE SKILLS 461

Table 3. Examples of Narratives

Female child with TB1, age: 7 years, 4 months

Ummm, I, once, there was a, we went. There was a for. There was this umm fort. A tree fell down. And there was dirt, all kinds of stuff there. It was our fort. And one day, I have a friend named Jude. She 's umm grown up. She has a kid. She has a cat named Gus, a kitten. It 's so cute. But once, when she d idn ' t have that kitten, one day, me, my brother, my cousin Matt, and her, and my dad, and one of his friends, went into the woods to see the fort, to show her. And we went up there. I stepped on a bee 's nest. And they chased us all the way back. And I got stung and my cousin Matt got stung in one of the private parts. And umm I had a bite right here (points), right here (points), right there (points), and umm one on my cheek. And right here. And when I umm went over, when we got back to my friend Jude 's house, in her bathroom she had this clean kind of stuff. And I put it on me. She put it one me right here (points). But umm, I had to go to the bathroom to put it on, you know. It hurt! And my brother Jason he got stung once. He got stung I think three right here (points). I remember where I got stung, but I don ' t remember where Jason got stung. My friend Jude didn ' t even get stung. She ran so fast that she didn ' t even get stung. The bees chased us and I looked back. And there was one right in front of my face. That 's when I got stung here (points). There was like two hanging around my legs. I was running and trying to get them off me. They both went, "Bzzzzz." It hurt! I was crying my head off.

Female child control, age: 7 years, 1 month

I got stung on the same place twice. One time I was running home because I was playing with my neighbor. Her name is Holly. I was in my bare feet. And I got my foot stung. The next time, I got my toe stung. We had like cement steps that were going up to where we park. There was a bee hive. I was stepping on a step. And it stung me again.

Male with TBI, age 42:

Well, normally on vacation, like we take a long trip. I got a van. I bought a Dodge van. We ' re going down the road. I 'll say to my wife, " U h . . . aren ' t we supposted to turn here?" She kinds of looks at the map. "No, I 'm not sure. Let 's see." Couple times we had to pull in, ask a policeman at the rest area. But normally, we don ' t get lost. We kind of look it over before we leave. But, that 's about it, just on long trips maybe, but kind of, I doubt it."

Male adult control, age 40:

Well one time when I was driving to North Carolina. I was supposed to get off at this one exit. And I missed the exit. The real problem was that I d idn ' t realize that I missed it until about 20 miles from that point. So, I had to go all the way back. And I made sure to pay real good attention to the roads the rest of the way.

Female adult with TBI, age: 41

Well, I 've gotten lost even coming here. It was probably the second time I came here. I, uh, went down, uh, 27, no 96, I think. And I came up . . . I remember they said 14 mile. I thought ended. Well, anyways, I just went around and around in cir- cles. And u h . . . so I got lost there. And it does s e e m . . . I do drive myself when I come here but I still get confused. I don ' t get lost but I get scared. I see Southfield Road and then I j u s t . . . Southfield and Greenfield. And still, after coming since February, I 'm still not sure whether for a few minutes, a few seconds there if I 'm supposed to take Green-

continued

462 BIDDLE et al.

Table 3. Continued

field or Southfield, you know. And then I don't know Southfield, you know. And, u h . . . I did get lost the second time I came here

Female adult control, age: 44

Ok One day I went shopping to Southland. And we have two cares, a black car and a white car. I 'm really terrible when I park in the parking lot. I have a habit of not really watching where I park. And this was around the holidays. So I parked my car. I run in. I 'm doing all my shopping, running out. Time to leave the shopping center and I 'm getting in the parking lot and it's like, "Where did I park my car?" I can't remember where I parked my car. And I 'm thinking, "OK, I always drive the white car, so 1 have the white car." I think I know in what general a rea . . , and this probably isn't about being lost, but I lost my car (laughing). So here I am walking up and down the aisles looking for a white car. And I have all these people who think I 'm going to be jumping in my car and they will get my spot. Only, I can't find my car. I was really embarrassed. So, after 10 minutes of walking up and down, it dawns on me that I don't have the white care. I have the black car!

Adult male with TBI, age: 31

(Long pause). About the only realy good time I can remember getting lost was in West Virginia. Me and my brother had been walking into the woods one day and we found a pig in the woods. A pig. Pig got out of my u n c l e ' s . . , pig, u h . . . So we found a pig around. And then we found out that we did get lost. So we went walking through the woods for about two hours, until it got really dark. Then we found a house and went back home. But during the time that we were lost, we were really scared cause we didn't think we were going to make it back. And I got all upset and started crying like a little wimp. And I didn't think we were goin make it back. But we did. It was a little traumatic for m e . . . cause I didn't think we were going to make it.

Adult control male, age: 35

I was coming back from Buffalo and took the road, my sister was driving. She was supposed to be coming back and got to the exit sign coming 1-75 North and she went past the exit. And we ended up close to Indiana before she wook me up, cause I was sleeping. So we had to go up and turn around and come back. It took us a little while but we did it.

present study quant i fy the impress ion of repeti t ious del ivery style to which McDona lds ' (1993) raters referred.

Redundancy may be another indicat ion of insuff icient moni to r ing by indi- viduals with TBI. Alternat ively, the repeti t ions made by the part icipants with TBI could represent a pseudodeve lopment strategy. That is, individuals with TBI may have used repeti t ions as a way to elaborate their stories when they were unable to organize and produce addit ional information.

In this study, as in others (Hartley & Jensen, 1991, 1992; Marsh & Knight , 1991), the individuals with TBI produced narratives which were less f luent than their peers. Their dysf luency resulted from using signif icant ly more fillers and more false starts than the control group. The present study supports Hart-

NARRATIVE SKILLS 463

ley and Jensens' (1991, 1992) findings and documents persistent problems with fluency in the population with TBI beyond an initial phase of recovery.

These dysfluencies have several implications. First, although it appears that the individuals with TBI were using the fillers as a strategy to compensate for possible problems with word retrieval or memory, this strategy disrupted the flow of their narratives and made them difficult to follow. Second, the abundance of false starts used by children and adults with TBI indicated prob- lems with the planning and organization of narrative discourse (Peterson & McCabe, 1983). That is, these individuals were experiencing significant diffi- culty organizing and executing complex discourse on demand.

The findings of this study are contrary to the conventional wisdom which posits that children fare better than adults following brain injury (See Finger's historical review, 1991). The children in this study did as poorly or worse than the adults with TBI. Specifically, the children used more fillers and omitted more information from their stories than control participants or adults with TBI. Levin et al. (1982) and others (Fletcher, Miner, & Ewing-Cobbs, 1987) have indicated that there is a lack of convincing evidence that children have improved chances for recovery following TBI, when compared with adults. More recently several researchers (Dennis & Barnes, 1990; Mateer & Will- iams, 1991; Ylvisaker, 1993) have suggested that prior optimism concerning recovery may need to be replaced with the acknowledgment that deficits per- sist following TBI in children. Furthermore, these deficits may not be appar- ent until much later in the child's development (Grattan & Eslinger, 1991). The trend documented in this modest sample warrants further examination.

The narrative impairments of the population of adults and children with TBI in this study appeared to be the result of problems with planning, produc- ing, and monitoring discourse. The results of the present study support prior descriptions (Brooks, 1987; Levin et al., 1982) of deficits in cognitive plan- ning and executive control in individuals who have sustained a TBI. It is pos- sible that the disruptions evident in the narratives of the children and adults with TBI were related less to language impairments than to difficulties with the executive processes utilized in discourse production (Ylvisaker, 1986, 1993; Ylvisaker & Szekeres, 1989).

The fact that the communication deficits of this group of individuals with TBI appeared to be related to impairments in executive processes may help to explain why Jordan, Ozanne, and Murdoch (1988) were not able to define the communication deficits of the children with TBI within the parameters of a recognized developmental language impairment. In fact, the results of this study supported other researchers (Glosser & Deser, 1990; McDonald, 1993) who have noted that the language problems of individuals with TBI are differ- ent in form from other language disordered populations.

In conclusion, qualitative and quantitative differences in the discourse of individuals following TBI were demonstrated in this study. The measure-

464 BIDDLE et al.

ments supplied by dependency analysis provided the "communication assess- ment" for which Ylvisaker has called (1986, p. 52). That is, dependency anal- ysis accounted for the dynamic interaction of linguistic and cognitive factors in the narrative performance of individuals with TBI. The results of this study have supported and extended McDonald's (1993) assertion that CHI discourse is disrupted in b o t h . . . "informational content and its organization" (p. 44).

By assessing the spontaneous personal narrative production of this popula- tion, it was possible to identify problems which may not have been discovered through traditional assessment measures. The individuals with TBI in this study did not produce poorly formed sentences, nor did they appear to have lost the basic structure for a narrative. In other words, their deficits appeared less related to basic language skill deficits than to impairments in higher order processes required for the planning and organization of language. Depen- dency analysis quantified those impairments and has been demonstrated to be a reliable index of narrative ability following TBI.

Clinical Implications The results of this study yielded important information for clinicians and edu- cators working with individuals with TBI. It would appear that speech-lan- guage pathologists need to utilize a dynamic approach to language assessment and therapy which accounts for the bidirectional influence of language and cognition on communicative competence. Clients with TBI need assistance in planning and producing discourse. They also need strategies to help them monitor their language and alert them to discrepancies in their speech. The use of conversational partners who could scaffold discourse for the adults with TBI may relieve some of the frustration these individuals experience and may help them to more effectively accomplish their communicative goals.

For children with TBI, discourse deficits may impede their academic as well as their social development. Ylvisaker (1993) in his comprehensive re- view of children's language outcomes post injury has noted that many chil- dren with TBI do not qualify for academic support services, yet they struggle to meet the academic and social demands of school. The results of this study confirm Ylvisaker's (1993) belief that the assessment of children following brain injury must be ongoing. These assessments must include tasks, such as narrative analysis, which measure the children's performance in real world settings, where complex processing demands may elicit the information clini- cians need to devise appropriate intervention plans.

C O N C L U S I O N The results of this study support the inclusion of dependency analysis in future investigations of the functional language outcome of children and adults fol-

NARRATIVE SKILLS 465

lowing TBI. Bruner (p. 77, 1990) noted, " . . . one of the most ubiquitous and powerful discourse forms in human communication is narrative." Loss of one's narrative ability will likely have negative real world consequences whether in academic, vocational, or social settings. It is critical to include the evaluation of narrative ability in future explorations of the consequences of TBI for children and adults.

The present study has added to the existing literature concerning communi- cation outcomes post injury in children and adults by demonstrating that nar- rative deficits persist beyond the initial period of injury recovery. In addition, this study has established dependency analysis as a reliable tool for distin- guishing impaired from intact narrative ability. Finally, through dependency analysis, the narrative skills of a group of children and adults with TBI were both quantified and characterized.

This research was supported by a grant to the first author from the National Institute of Child Health and Human Development, Grant No. T32 HDO 7415.

REFERENCES Alexander, M.P., Benson, D.F., and Stuss, D.T. (1989). Frontal lobes and lan-

guage. Brain and Language 37, 656-691. Benton, A.L. (1987). Thoughts on the application of neuropsychological tests.

In H.S. Levin, J. Grafman, and H.M. Eisenberg (Eds). Neurobehavioral re- covery from head injury (pp 111-115). New York: Oxford University Press.

Brooks, D.N. (1987). Measuring neuropsychological and functional recovery. In H.S. Levin, J. Grafman, and H.M. Eisenberg (Eds). Neurobehavioral re- covery from head injury (pp 57-72). New York: Oxford University Press.

Bruner, J. (1990). Acts of Meaning. Cambridge: Harvard University Press. Chapman, S.B., Culhane, K.A., Levin, H.S., Harward, H., Mendelsohn, D.,

Ewing-Cobbs, L., Fletcher, J.M., and Bruce, D. (1992). Narrative discourse after closed head injury in children and adolescents. Brain and Language 43, 42--65.

Deese, J. (1984). Thought into speech: The psychology of a language. Engle- wood Cliffs, N.J.: Prentice-Hall, Inc.

Dennis, M. (1991). Frontal lobe function in childhood and adolescence: A heuristic for assessing attention and regulation, executive control, and the intentional states important for social discourse. Developmental Neuropsy- chology 7, 327-358.

Dennis, M., and Barnes, M.A. (1990). Knowing the meaning, getting the

466 BIDDLE et al.

point, bridging the gap, and carrying the message: Aspects of discourse fol- lowing closed head injury in childhood and adolescence. Brain and Lan- guage 39, 428-446.

Dennis, M., & Lovett, M. (1990). Discourse ability in children after brain damage. In Y. Joanette & H.H. Brownell (Eds). Discourse ability and brain damage. Theoretical and empirical perspectives (pp 199-223). New York: Springer-Verlag.

Erlich, J.S. (1988). Selective characteristics of narrative discourse in head in- jured and normal adults. Journal of Communication Disorders 21, 1-9.

Ewing-Cobbs, L., and Fletcher, J.M. (1987). Neuropsychological assessment of head injury in children. Journal of Learning Disabilities 20, 526-535.

Feagans, L. (1982). The development and importance of narratives for school adaptation. In L. Feagans and D. Farran (Eds). The language of children reared in poverty. New York: Academic Press.

Finger, S. (1991). Brain damage, development, and behavior: Early findings. Developmental Neuropsychology 7, 261-274.

Fletcher, J.M., Miner, M.E., and Ewing-Cobbs, L. (1987). Age and recovery from head injury in children: Developmental issues. In H.S. Levin, J. Graf- man, and H.M. Eisenberg (Eds). Neurobehavioral recovery from head in- jury (pp 279-291). New York: Oxford University Press.

Glosser, G., and Deser, T. (1990). Patterns of discourse production among neurological patients with fluent language disorders. Brain and Language 40, 67-88.

Grattan, L.M., and Eslinger, P.J. (1991). Frontal lobe damage in children and adults: A comparative review. Developmental Neuropsychology 7, 283-326.

Hartley, L.L., and Jensen, P.J. (1992). Three discourse profiles of closed head-in- jury speakers: Theoretical and clinical implications. Brain Injury 6, 271-282.

Hartley, L.L., and Jensen, P.J. (1991). Narrative and procedural discourse af- ter closed head injury. Brain Injury 5, 267-285.

Jordan, F.M., and Murdoch, B.E. (1990). Linguistic status following closed head injury in children: A follow-up study. Brain Injury 4, 147-154.

Jordan, F.M., Cannon, A., and Murdoch, B.E. (1992). Language abilities of mildly closed head injured children 10 years post injury. Brain Injury 6, 39--44.

Jordan, F.M., Murdoch, B.E., Buttsworth, D.L. (1991). Closed-head-injured children's performance on narrative tasks. Journal of Speech and Hearing Research 34, 572-582.

NARRATIVE SKILLS 467

Jordan, F.M., Ozanne, A.E., and Murdoch, B.E. (1988). Long-term speech and language disorders subsequent to closed head injury in children. Brain Injury 2, 179-185.

Katz, D.I. (1992). Neuropathology and neurobehavioral recovery from closed head injury. Journal of Head Trauma Rehabilitation 7, 1-15.

Labov, W. (1972). Language in the inner city. Philadelphia: University of Pennsylvania Press.

Levin, H.S., Benton, A.L., and Grossman, R.G. (1982). Neurobehavioral con- sequences of closed head injury. New York: Oxford University Press.

Liles, B.Z. (1993). Narrative discourse in children with language disorders and children with normal language: A critical review of the literature. Jour- nal of Speech and Hearing Research 36, 868-882.

Liles, B., Coelho, C.A., Duffy, R.J., & Zalagens, M.R. (1989). Effects of elic- itation procedures on the narratives of normal and closed head-injured adults. Journal of Speech and Hearing Disorders 54, 356-366.

Marsh, N.V., and Knight, R.G. (1991). Behavioral assessment of social com- petence following severe head injury. Journal of Clinical and Experimental Neuropsychology 13, 729-740.

Mateer, C.A., and Williams, D. (1991). Effects of frontal lobe injury in child- hood. Developmental Neuropsychology 7, 359-376.

McCabe, A., and Peterson, C. (1990). Keep them talking: Parental styles of interviewing and subsequent child narrative skill. Paper presented at the Fifth International Congress for the Study of Child Language, Budapest, Hungary.

McCabe, A., and Peterson, C. (Eds). (1991). Developing narrative structure. Hillsdale, N.J.: Lawrence Erlbaum Associates.

McDonald, S. (1992). Communication disorders following closed head injury: New approaches to assessment and rehabilitation. Brain Injury 6, 283-292.

McDonald, S. (1993). Pragmatic language skills after closed head injury: Ability to meet the informational needs of the listener. Brain and Language 44, 28-46.

Mentis, M., and Prutting, C.A. (1987). Cohesion in the discourse of normal and head-injured adults. Journal of Speech and Hearing Research 30, 88-98.

Mentis, M., and Prutting, C.A. (1991). Analysis of topic as illustrated in a head-injured and a normal adult. Journal of Speech and Hearing Research 34, 583-595.

468 BIDDLE et al.

Peterson, C., and McCabe, A. (1983). Developmental psycholinguistics: Three ways of looking at a child's narrative. New York: Plenum Press.

Peterson, C., and McCabe, A. (1991). Linking children's connective use and narrative macrostructure. In A. McCabe and C. Peterson (Eds). Developing narrative structure (pp 29-53). Hillsdale, N.J.: Lawrence Erlbaum Associ- ates.

Preece, A. (1987). The range of narrative forms conversationally produced by young children. Journal of Child Language 14, 353-373.

Teasdale, G., and Jennett, B. (1974). Assessment of coma and impaired con- sciousness: A practical scale. Lancet 2, 81-84.

Wyckoff, L.H. (1984). Narrative and procedural discourse following closed head injury. Unpublished doctoral dissertation. The University of Florida.

Ylvisaker, M. (1986). Language and communication disorders following pe- diatric head injury. Journal of Head Trauma Rehabilitation 1, 48-56.

Ylvisaker, M., and Szerkeres, S.F. (1989). Metacognitive and executive im- pairments in head-injured children and adults. Topics in Language Disor- ders 9, 34-49.

Ylvisaker, M. (1993). Communication outcome in children and adolescents with traumatic brain injury. Neuropsychological Rehabilitation 3, 367-387.

C O N T I N U I N G E D U C A T I O N Narrative Skills Following Traumatic Brain Injury in Children and Adults QUESTIONS

1. Dependency analysis focuses on: a. Informational content b. Syntax c. Short-term memory for retelling d. Cohesion e. Pragmatics

2. The personal narrative of children and adults with TBI are _ _ their non-injured peers: a. Less fluent b. More repetitive c. Lacking in critical information d. a & b only e. a,b, & c

than

NARRATIVE SKILLS 469

3. An overabundance of false starts in a spontaneous personal narrat ive m ay be indicat ive of problem s with: a. Mem ory b. Im pulsivity c. Planning and organizat ion d. Word retrieval e. Phonological processing

4. Com m unicat ion deficits in children with TBI a. Are less severe than for adults b. Are a lways quantifiable using tradit ional language measures c. Are apparent im m ediate ly following injury d. May not becom e apparent until m uch later in the child's developm ent e. Do not persist beyond the initial recovery period

5. Traum at ic brain injury frequent ly includes injuries to the re- gion of the brain which m ay result in _ _ im pairm ents: a. Parietal/spatial b. Tem poral/aphasia-like c. Frontal/execut ive processing d. Fronta l/m em ory e. Tem poral/ re t rieval

All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.