Language Outcomes, Growth, and
Predictors of Success: A Multi-State
(NECAP) Perspective
WREIC
June 16, 2017
Presenter
Allison Sedey, Ph.D., CCC-SLP, CCC-A
[email protected]
University of Colorado-Boulder
Colorado School for the Deaf and the Blind
Additional Author
Christine Yoshinaga-Itano, Ph.D.
University of Colorado-Boulder
[email protected]
Acknowledgements
• Early interventionists in NECAP states
• Participating families
• Assessment Coordinators: Clare Neville,
Anna Clark, and Mallene Wiggin
• A long list of SLHS undergraduate and
graduate student employees who assisted in
scoring test protocols and data entry
Today’s Topics
• Brief description of NECAP
• Sample characteristics
• Factors associated with more
successful language outcomes
• Language outcome data over time
• Strengths and limitations in specific
language skills
What is NECAP?
• All families are participating in NECAP
• National Early Childhood Assessment
Project
• CDC-supported project examining
language outcomes at a national level
• Birth to 3
• Children who are deaf or hard of hearing
Motivation for NECAP
2007 JCIH Position Statement (Goal 6 of 2013
Supplement)
“All children who are D/HH should have their
progress monitored every 6 months from
birth to 36 months of age, through a protocol
that includes the use of standardized, normreferenced developmental evaluations”
Motivation for NECAP
Using a common set of instruments across
states allows us to:
• Examine outcomes nationally (rather than
from a single program or state)
• Improve the validity of outcome results by
increasing sample size
• Include children who are participating in
different amounts and types of intervention
Demographic Characteristics of
Participants with Bilateral Loss
• English, Spanish and/or ASL used in
the home
• With and without additional disabilities
• All degrees of hearing loss
• Demographics that follow based on 448
children in predictive study
• Individual studies vary slightly (by a few
percentage points)
Participant Characteristics:
EHDI Guidelines
Age at…
Mean (mos)
Identification
4.1
Intervention
7
*58% of children met the EHDI 1-3-6 guidelines
Participant Characteristics:
Degree of Hearing Loss
Degree of hearing loss
%
Mild
35%
Moderate
22%
Moderately severe
15%
Severe
8%
Profound
20%
Participant Characteristics:
Type of Amplification
Type of Amplification
%
None
11%
Hearing aid
68%
Cochlear implant
16%
Bone conduction aid
5%
Participant Characteristics:
Mother’s Level of Education
Highest degree
completed
% of primary
caregivers
Less than HS
13%
High school diploma
38%
Vocational or Associates
20%
Bachelor’s degree
22%
Graduate degree
7%
Participant Characteristics:
Communication Mode
Communication mode used
with child by family
Primarily spoken language
% of primary
caregivers
74%
Spoken only
30%
Very occasional sign used
44%
Spoken + sign language
Sign only
22%
4%
Assessment Instruments
Assessments
• Minnesota Child Development Inventory
• 12 months to 6 years
• Expressive Language and Language
Comprehension subscales
• MacArthur-Bates Communicative
Developmental Inventory
• 8 to 30 months
• Expressive vocabulary subscale (in spoken
and/or sign language)
Assessment Characteristics
All 3 of the assessments are:
• Standardized
• Norm-referenced (normed on children
who are hearing)
• Parent report
Minnesota Child Development
Inventory
• Parents respond “yes” or “no” to a variety of
statements about their child
• Examples:
• “Has a vocabulary of 20 or more words”
• “Uses the word ‘not’ in sentences
• “Expresses likes and dislikes in words”
MacArthur Bates Communicative
Development Inventory
• List of 680 words divided into semantic
categories (toys, food, action words,
descriptive words, etc.)
• Parents indicate words their child produces
•
•
In spoken language
In sign language
Determining Language
Quotient
Language Age/Chronological Age x 100
If LQ = 100, Language Age = CA
If LQ < 100, Language Age < CA
If LQ > 100, Language Age > CA
Average Range
Minnesota: LQs of 80+
MacArthur: LQs of 75+
Three Separate Studies
• Titles of the 3 studies
• “Predictive Analysis”
• “Growth Analysis”
• “Item Analysis”
• What is different
• Number of participants
• Demographic characteristics (but very similar
across studies)
Predictive Analysis: MacArthur
Expressive Vocabulary
MacArthur Predictive Analysis:
Participant Criteria
• Bilateral, pre-lingual hearing loss
• (all degrees -- mild to profound)
• No diagnosis of auditory neuropathy
• English or Spanish is written language
of the home
• Children with and without additional
disabilities included
MacArthur Predictive Analysis:
Number of Participants
• 448 children
• Child’s most recent assessment used
• Chronological age
• Range = 8 to 39 months
• Mean = 25 months
States Represented in
Predictive Analysis
•
•
•
•
•
•
Arizona
California
Florida
Idaho
Indiana
Maine
•
•
•
•
•
•
North Dakota
Oregon
Texas
Utah
Wisconsin
Wyoming
Predictors of Better Vocabulary
Outcomes: Multiple Regression
• Significant predictors of MacArthur
expressive vocabulary quotient
•
•
•
•
•
•
Presence of additional disabilities
Meeting EHDI 1-3-6 guidelines
Higher mother’s level of education
Lesser degrees of hearing loss
Parent who is deaf or hard of hearing
Quotient decreases with age so gap between CA
and Language Age is widening over time
Group Comparisons: Children with
no additional disabilities
100
Vocabulary Quotient
90
Deaf/HOH
B.A.+
meets
Mild-Mod
80
Doesn’t
meet’
< B.A
Mod/SevProf
Hearing
70
60
50
EHDI 1-3-6
Mother's Ed
Degree of Loss
Parent Hearing Status
Relationship of vocabulary
quotient and chronological age
Group Comparisons: Vocabulary
Quotient and Chronological Age
100
Vocabulary Quotient
< 24 months
90
80
24+ months
70
60
50
Chronological Age
Exploring the widening
language gap
Exploring the widening gap between
chronological age and language age
• When does the gap begin?
• How much does it widen over time?
• Which general areas of language
demonstrate the biggest gap?
• Which specific language skills contribute
least and most to the gap?
Growth Analysis:
Minnesota Language and
MacArthur Expressive Vocabulary
Growth Analysis: Participant
Criteria
• Bilateral, pre-lingual hearing loss
• (all degrees -- mild to profound)
• No diagnosis of auditory neuropathy
• English is written language of the home
• No other disabilities that would affect
speech or language development
Minnesota CDI Growth Analysis:
Number of Participants/Assessments
• Outcomes are a combination of crosssectional and longitudinal data
• 457 children
• Assessed on 1 to 5 occasions
• Total assessments = 739
MacArthur CDI Growth Analysis:
Number of Participants/Assessments
• Outcomes are a combination of crosssectional and longitudinal data
• 634 children
• Assessed on 1 to 5 occasions
• Total assessments = 1,066
States Represented in Language
Growth Analysis
•
•
•
•
•
•
•
Arizona
California
Florida
Idaho
Indiana
Maine
Minnesota
•
•
•
•
•
•
•
New Mexico
North Dakota
Oregon
Texas
Utah
Wisconsin
Wyoming
Growth Analysis:
Participant Characteristics
• Chronological age
• Range = 12 to 35 months
• Minnesota mean CA = 24 months
• MacArthur mean CA = 23 months
Minnesota CDI: Expressive
Language Subscale
Language Age
35
Hearing
Deaf/Hoh
30
25
20
15
10
12
14
16
18
20 22 24 26 28
Chronological Age
30
32
34
Minnesota CDI: Language
Comprehension/Conceptual Lang
Language Age
35
Hearing
Deaf/Hoh
30
25
20
15
10
12
14
16
18
20 22 24 26 28
Chronological Age
30
32
34
MacArthur-Bates CDI: Expressive
Vocabulary
Language Age
35
Hearing
Deaf/Hoh
30
25
20
15
10
12
14
16
18
20 22 24 26 28
Chronological Age
30
32
34
MacArthur-Bates CDI: Expressive
Vocabulary Size – 50th Percentile
Number of Words
600
Hearing
Deaf/Hoh
500
400
300
200
100
0
12
14
16
18
20 22 24 26 28
Chronological Age
30
32
34
Item Analysis: Minnesota
Expressive Language and
Language Comprehension
Item Analysis: Participant Criteria
• Bilateral, pre-lingual hearing loss
• (all degrees -- mild to profound)
• English is written language of the home
• No other disabilities that would affect
speech or language development
Minnesota CDI Item Analysis: Number
of Participants/Assessments
• Outcomes are a combination of crosssectional and longitudinal data
• 535 children
• Assessed on 1 to 6 occasions
• Total assessments = 935
Minnesota CDI Item Analysis:
Participant Characteristics
• Chronological age
• Range = 12 to 35 months
• Mean = 24 months
Primary Contributing States
(each > 5% of the database)
•
•
•
•
•
•
Colorado – 39%
Idaho – 20%
Arizona
Indiana
Texas
Wisconsin
Secondary Contributing States
(< 5% of the database)
•
•
•
•
•
California
Florida
Maine
Minnesota
New Mexico
•
•
•
•
North Dakota
Oregon
Utah
Wyoming
Minnesota CDI Item Analysis:
Results: Language Quotients
Expressive Language
Median = 95
LQ of 80 and above = 79%
Comprehension and Conceptual Language
Median = 87
LQ of 80 and above = 65%
Determining Item Age Level
• Group children into 3-month age brackets
(e.g., “12 month group” = 12 to 14 months)
• Find the first age group at which 75% of the
children achieve a given item
• These same procedures followed in the
test’s normative sample
Linguistic skills at age level
Non-verbal communication
• “Points”
• Asks for drink or food using words or sounds
Early verbal communication (first 20 to 40
words)
• Uses at least 10 words
• Greets people with “hi” or similar expression
Linguistic skills at age level
Basic comprehension
• Follows simple instructions
• Understands the meaning of “up” and “down”
Rote language
• Says/signs “please” and “thank you”
• Greets people with “hi” or similar expression
Linguistic skills: Mildly delayed
Comprehension of early conceptual vocabulary
• Understands what “off” & “on” mean
• Understands what “open” & “close” mean
Concrete, single word utterances
• Names a few familiar objects in picture books
• Uses 50 or more different words
Linguistic skills:
Moderately delayed
Early number concepts
• Understands “one” and gives you just one when
you ask for “one”
Use of first pronoun
• Says/signs “mine” or “my”
Linguistic comprehension skills:
Severely delayed
Prepositions
• Understands the meaning of at least 3 location
words (“in,” “on,” “under,” etc.)
Complex comprehension
• Follows 2-part instructions
Linguistic production skills:
Severely delayed
Pronoun use
•
Use of negative forms (not, don’t, won’t, can’t,
etc.)
Conceptual vocabulary
• “Points to/names the bigger of 2 objects”
• “Expresses feelings in words”
• Says/signs when something is heavy
Linguistic production skills:
Severely delayed
Grammar
• Uses plural words (e.g., “girls,” “cats”)
• Asks questions with ‘what’ or ‘where’
• Uses the words “a,” “an,” and “the”
Summary
• Better language scores are associated with:
•
•
•
•
•
Absence of additional disabilities
Meeting the EHDI 1-3-6 guidelines
Higher maternal level of education
Less significant degrees of hearing loss
Parent who is deaf/hoh
Summary
• One of the strongest predictors of language
quotients is chronological age
 As CA increases, Language Quotient decreases
Summary
• As language demands increase over time,
gap between CA and Language Age widens
• Divergence from age expectations starts at:
• 25 mos for general, surface structure expressive
language
• 20 mos for cognitive-linguistic skills
• 16 mos for expressive vocabulary
Summary
• By 35 months of age the gap between
language age and chronological age is:
• 7 mos for general, surface structure
expressive language
• 8 mos for cognitive-linguistic skills
• 12 mos for expressive vocabulary
Summary
• The majority of 12- to 18-month level
language items are demonstrated at or close
to on schedule
• Children who are d/hoh demonstrated
delays on almost all language items at or
above the 21-month level
Summary
• Beginning at 19 mos, hearing children
produce 35 to 40 new words per month
• Beginning at 19 months, children who are
deaf or hard of hearing average 15 new
expressive words per month
• It is important for parents and interventionists
to keep typical development in mind when
assessing progress and setting goals
Summary
• Average language scores for a group of
children from birth to 3 typically
underestimate the skill level of children
below 2 years of age and overestimate the
skills of children above 2 years of age
• Research and accountability data for this
age range should divide the children into two
age groups