Nature of written langauge problems in children with Specific language impairment (SLI)

1. The Nature of Written Language
Deficits in Children With SLI
Claire Mackie and Julie E. Dockrell (2004)
Presenters: Xin Zhou & Hillary Sang

2. Outline
• Introduction
– def. of SLI
– review of previous studies and their implications
• Purpose
• Predictions
• Method
• Results
• Discussion & Conclusion

3. Introduction
• Def of SLI: children with language problems that cannot
be explained in terms of other cognitive, neurological or
perceptual deficits (Bishop, 1992)
• What does this mean?
– Protracted rate of language development
– Difficulties with particular sub-components of the
system
– >>>including processing of written text
• Leading to associated difficulties in
– reading decoding
– reading comprehension

4. Previous studies
• Several studies report difficulties in word decoding and
understanding of written text (Bishop & Adams, 1990;
Botting, Crutchley & Conti-Ramsden, 1998 ;)
• But, few studies have examined children’s written skills
with only 3 providing direct evidence of the written
language performance of children with LLD

5. Studies on written language skills of
Language impaired children
Study Groups Matching Task Results LLD deficit
Gillam &
Johnston
(1992)
LLD and
TD
CA, LA
(based on a
sentence-
imitation
task) and
reading age
Spoken and
written story
from a picture
Written*
narrative >
than oral
narrative for
all groups
Fewer
morphemes
and preps
More
grammatical
errors in
complex
sentences
relative to
CA & LA
peers
Scott &
Windsor
(2000)
LLD CA and LA
(based on
standardized
test- Test of
Language
development)
Oral and
written
summaries of
two
educational
videos
Written*
narrative >
than oral
narrative for
all groups
Higher % of
grammatical
errors in
written text
relative to
LA peers
Windsor et
al. (2000)
Same
participants
Same
matching
??? ??? More verb
composite
errors in
comparison
with peers

6. Implications of these studies?
• Performance of LLD children comparable with LA but not
with CA peers
• Specific limitation with written syntax; more errors than LA
peers
 So, is syntax the children’s core linguistic deficit?
But the studies:
• involved participants with LLD whose difficulties may not
be specific to language
• matching tasks i.e. sentence repetition (Gillam & Johnston,
1992) and overall LA measure(Scott & Windsor, 2000) do
not capture abilities to generate ideas in oral language

7. How does the present study address
these issues?
• Involved participants with SLI
• Matching on production of oral narratives – i.e. expressive
language
• Analysis of types and frequency of errors made
Purpose of the study
• Investigate written compositions produced by school-age
children with SLI in comparison with LA & CA matched
peers

8. Research questions and Predictions
• What are the patterns of performance across the 3 groups in
terms of fluency, content and accuracy of written text?
• Are there interrelationships among oral language, reading and
writing performance?
• What is the nature of the errors produced by children with SLI?
Predictions
• SLI children will be delayed in written language skills relative to
CA matched peers
• Content and length of compositions btwn SLI and LA matched
peers should be equivalent but less compared to CA peers
• SLI group would have specific difficulties with grammar and
spelling in relation to both LA & CA

9. The Nature of written language deficits
in children with SLI
Method and Results
By Clare Mackie & Julie Dockrell Part 2 presented by Xin ZHOU
EMCL Program

10. Method
Participants:
11 SLI children, 8m3f, (mean age 11 years, range 9;8–12;3)
from a special language school
11 TD children matched for CA and gender (mean age 11;2,
range 10;0–12;3) from a local primary school
11 TD children matched for LA and gender on the Bus Story
(mean age 7:3, range 6:0–9:8) from a local primary
school
SLI confirmed by CELF–R(Semel,Wiig, & Secord, 1987),
total score >2 SD below the mean

11. Instrument
Bus Story Test of Continuous Speech (Renfrew, 1985)
Clinical Evaluation of Language Fundamentals—Revised (CELF–R; Semel, Wiig, & Secord, 1987)
The Wechsler Objective Reading Dimensions (Rust et al., 1993)
The Picture Story Language Test (Myklebust, 1965)
Bus
Story
Test
Story told->look through pictures->retell->scored (40,
full mark , oral information score; sentence length
score)
CELF–R Language functions (e.g. phonology, syntax, semantics,
memory, and word finding), providing expressive and
receptive language score
WORD single-word reading accuracy (providing standard scores
and reading age equivalents)
RPM nonverbal ability (selecting the piece that can complete a
pattern)
PSLT standardized writing assessment (visual prompt leading the
generating of original text)

12. Procedure
PSLT:
• picture shown
• story writing (30 min max, time recorded) 
• meaning ascertained, misspell diciphered,
assessment counterbalanced (50% writing
1st, 50% L assessment 1st)

13. Writing Analysis
PSLT
(Factors of language
proficiency are ignored;
only the content or quality
of the ideation is
evaluated)
Productivity
(Number of words written)
Syntax
(total count of errors in
grammatical construction
and morphological
features)
1. additions
2. omissions
3. substitutions
Abstract–Concrete
(Content )
LEVEL 1: unrelated letters
or words
LEVEL2
LEVEL3
LEVEL4
LEVEL5
The picture is a point of
reference instead of being
central to the story itself

14. Fluency: (time for completing the PSLT)/ (total number of words produced)
Classification of spelling errors:
phonologically inaccurate error: not having a possible phoneme-to-
grapheme correspondence in English(e.g. clars / cloud, errors involving
substitution of wrong sounds, insertion of phonemes, or deletion of
phonemes)
orthographically inaccurate error : a misspelling containing a illegal
sequence of letters which would include positional restriction(e.g. wusz /
once)
Coding reliability
Reliablilty checks performed on all writing samples and scales by experienced
researchers. Interrater disagreement solved by taking Dr. Mackie’s score.
Mean reliability
Syntax: 89%
Abstract–Concrete: 92%.
Judgment of spell correctness: 100% (Across all spelling error categories :91%)

15. RESULTS Patterns of Differences in WritingTask
Performance Across the SLI,CA, and LA Groups
Statistically significant differences are found among the
3 groups
1. On the total number of words written, F(2, 30) = 7.35,
p = .003
2. On the Fluency of writing, based on the number of words
produced per minute, F(2, 30) = 7.42, p = .002
3. In the proportion of syntax errors produced, F(2, 30)
=9.98, p = .001
4. In the proportion of spelling errors produced, χ2(2, N =
33) = 9.88, p = .007
Surprisingly, no statistically significant difference
among groups on a measure of content, F(2, 30) = 3.54,
p = .042

16. Table 2. Performance on the
writing task: Length, fluency,
content,
syntax, and spelling errors.
Note. SLI = specific language
impairment; CA = chronological
age; LA
= language age.

17. RESULTS: Patterns of Relationships Between
Measures of Oral and Written Language
• To examine the relationship between the
written measures and the children’s oral
language ability, partial correlation analyses
were carried out separately for the 3 groups of
participants, controlling for age. (Bonferroni
correction with a probability level of .006).
• All correlations are presented in
Table 3 for the SLI group
Table 4 for the CA group
Table 5 for the LA group.

18. SLI GROUP : No statistically significant correlations between oral
language and any measure of writing, between reading and oral
language, or between the different measures of written language.
However, there was a statistically significant NEGATIVE
relationship between word reading and the proportion of spelling
errors produced, r(11) = –.82, p =.002

19. CA group: no statistically significant correlation between
the oral language measures and any measure of writing at
the .006 corrected level.
However, there was a trend for content to be associated
with total words written, r(11)= .76, p = .01

20. LA group: no significant correlations among word reading, oral
language, and any measure of writing.
However, a positive relationship was revealed between the total
number of words and fluency per minute, r(11) = .80, p= .005.
There are Trend 1 indicating positive relationships between content and
total words, r(11)= .76, p = .011
Trend 2 between scores on the two language measures: the Bus Story
Information score and the Bus Story Sentence Length score, r(11) = .77, p
= .009.

21. RESULTS: Spelling Errors
• SLI (M = .58, SD = .43) produced proportionately more
phonologically errors than CA (M = .09, SD = .18) & LA
(M = .37, SD = .34)
• one-way ANOVA revealed a significant difference
between the groups in the proportion of phonologically
inaccurate errors, F(2, 30) = 5.88, p = .007
• SLI (M = .34, SD = .38) produced proportionately more
orthographic errors than CA (M = .01, SD = .03) & LA (M
= .22, SD = .31)
• no statistically significant difference was revealed
among the groups in proportion of orthographically
inaccurate spelling errors, F(2, 30) = 3.75, p = .035, ( at
the corrected .01 level)

22. Results of Post hoc tests
Significant difference between the SLI and CA groups (p = .005)
in in the proportion of phonologically inaccurate errors
A trend toward a statistically significant difference between the SLI
group and the CA group (p = .03)

23. RESULTS: specific syntax measures

24. About Table 6 and a descriptive analysis
• SLI group: mean proportions of errors are more than CA
and LA groups in all types
• The Kruskal–Wallis test revealed that 3 groups differed
statistically significantly in the proportion of whole-word
addition errors produced(p = .005), in the proportion of
whole-word omission errors produced ( p = .006), but not
on the proportion of whole-word substitution errors ( p =
.27)
• Descriptive analysis for the whole-word omission:
SLI: 21 errors = 8 (different forms of be)+5 prepositions +5
pronouns+2 conjunctions +1 definite article
CA: 9 errors =5 prepositions+1 auxiliary v +1 vi+1 pronoun
+1 indefinite article
LA: 3 errors= 3 pronouns

26. About Table 7 and a descriptive analysis
• T test revealed no statistically significant difference
between the SLI and CA group t(20) = 1.87, p =.85, on
the proportion of word-ending addition errors.
• There was a difference among the 3 groups in the
proportion of omission-of-word-ending errors, F(2, 30) =
4.34, p = .022, (not statistically significant at the .01 level)
• Descriptive analysis for the omission-of-word-ending
errors:
SLI: 17 errors = 8 (-ing)+1 –ed +2 (3rd person –s) + 6( pl –s)
CA: 3 errors= 3 ( pl –s)
LA: 1 errors= 1 ( pl –s)

27. Discussion
1. Performance of written language across the 3 groups
• All groups spent similar amount of time producing there text – no
effect of time limitation
SLI:
Length
• produced texts that were shorter than matched CA & LA peers
>>>productivity is a problem for children with language difficulties
(Scott & Windsor, 2000) and learning difficulties (Graham, 1990)
• Produced fewer words per minute than their CA peers but not LA
>>attributed to reduced processing resources found in the oral
language of the children with SLI
• Evidence of difficulty in generating text

28. Content
• content of texts was equivalent across the 3 groups
>>>SLI children capable of producing abstract and
imaginative stories but are slower than peers
Syntax
• SLI had evident problems in written syntax
• >>>both LA & CA peers were more accurate
Spelling
• no significant difference between SLI and peers on the
proportion of spelling errors
>>>SLI children wrote words they could spell rather than
attempt words that they could not spell – this reflects a
strategy used by the children with SLI rather than a similar
level of spelling performance per se.

29. 2. Relationships among oral language, reading and
writing
• No relationship in any of the 3 groups btwn oral lge
measures and the written lge measures
>>>attributed to small sample size
• Reading accuracy was negatively associated with
spelling errors for the SLI group.
>>>link btwn reading and writing

30. 3. Nature of written errors - SLI group
Omissions
• whole words e.g. verbs such as “be“
• Word-ending (Inflections) e.g. -ed, -ing & -s
>>>errors in inflection were in equal proportion btwn
nouns and verbs (-ing ending and –s ending absent in
children’s writing)
• Prepositions, pronouns and conjunctions also ommitted
• NB: difficulties in grammar and morphology separate SLI
children‘s performance from both CA-matched and LA-
matched peers

31. Conclusion
• SLI children have specific deficits in written language in comparison
with CA- and LA-matched peers
• The nature of error patterns suggest that children with SLI have
specific problems with productivity, syntax and spelling but not
content.
• Use of verbs by the SLI group was compromised
– whole word omission of “be“
– Inflections e.g. -ed, -ing & -s
• Prepositions and pronouns were ommitted
• Reviewing skills of SLI children are reduced - a result of limited
processing resources
– However, SLI children have the ability to produce imaginative
stories but do not have the cognitive resources to translate their
ideas into written language.

32. Thank you
&