AAPS-3 - Pacific University

advertisement
Psychometric Characteristics of Single Word Articulation Tests
Peter Flipsen Jr., Ph.D.1 & Diane Ogiela, Ph.D.2
1Pacific
ASHA Convention, 2013
INTRODUCTION
• This study involved the compilation and
evaluation of the psychometric properties of 11
currently available single word articulation tests.
• The goal was to assist busy clinicians with test
selection.
METHOD
• The two authors independently evaluated the test
manuals and forms and then reached consensus.
The following tests were evaluated:
AAPS-3 - Arizona Articulation Proficiency Scale
(Third Edition; 2000).
ALPHA - Assessment Link between Phonology and
Articulation (2000?).
BBTOP - Bankson-Bernthal Test of Phonology
(1990).
CAAP- Clinical Assessment of Articulation and
Phonology (2002).
DEAP-A - Diagnostic Evaluation of Articulation and
Phonology (American Edition; 2006).
GFTA-2 - Goldman-Fristoe Test of Articulation
(Second Edition; 2000).
HAPP-3 - Hodson Assessment of Phonological
Patterns (Third Edition; 2004).
LAT - Linguisystems Articulation Test (2010).
PAT-3 - Photo Articulation Test (Third Edition;
1997).
SHAPE - Smit-Hand Articulation and Phonology
Evaluation (1996).
SPAT-D - Structured Photographic Articulation Test
Featuring Dudsbury (2001).
2Idaho
University, Forest Grove, OR
Chicago, IL
State University, Meridian, ID
RESULTS: Basic Analysis Does each test meet the basic criteria for standardized tests as set forth by McCauley & Swisher (1984) along
with some additional criteria (highlighted in red)?
AAPS-3 ALPHA BBTOP CAAP
Clearly defined standardization sample?
√
√
Adequate subgroup size (n > 100)? 1
√
√2
Did norming sample include those with SSD?
?
√
DEAP-A GFTA-2
√
√
HAPP-3
LAT
√
√
√
√
3
√
4
√
√
√
√
5
√
√
√
√
√
Item analysis used to select final stimuli?
Provides Means and SDs for subgroups?
√
Standard error of measurement data provided?
√
Evidence of concurrent validity presented?
√
√
PAT-3
√
√
√
√
√
√
√
√
√
√
√
SHAPE SPAT-D
√
√
√
?
√
√
√
√
√
√
√
√
√
√
√
Evidence of predictive validity presented?
√
Test-retest reliability reported?
Inter-examiner reliability reported?
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
Detailed administration instructions provided?
√
√
√
√
√
√
√
√
√
Examiner qualifications specified?
√
√
√
√
√
√
√
√
√
√
√6
√
√
√
√
11
10
9
10
10
6
√
Diagnostic accuracy data reported?
TOTAL CRITERIA MET (of 13 possible)
8
8
7
8
√
9
√ = meets criterion. ? = Not specifically mentioned. 1 For actual comparison groups in normative tables. 2 For 11/12 subgroups. 3 Only for 6/10 subgroups.
4 Only for 15/22 subgroups.
5 Only for about half of the sample.
6 Includes predictive power and sensitivity/specificity findings in addition to group differences.
RESULTS: Analyses Available Does each test allow for systematic analysis of both traditional (initial, medial, final position) accuracy
and phonological patterns (processes)? Did it include formal analysis of vowels? Does it examine multisyllabic words?
AAPS-3
Traditional – consonants
√
Traditional – vowels
√
ALPHA BBTOP
√
√
Patterns – atypical 1
√
√
√
Multisyllabic words 2
As defined by the test authors.
4 Only on two subtests.
√
DEAP-A GFTA-2
√
HAPP-3
√
√
LAT
√
√
Patterns – typical
1
√
CAAP
2
√
√
√
4
Four or more words containing at least three syllables.
PAT-3
√
SHAPE SPAT-D
√
√
3
√
√
√
√
√
3
√
√
√
Not included in the score calculation.
√
DISCUSSION
• Several tests excluded children with SSDs from the
normative sample yielding a somewhat “purified”
comparison group.
• This presumes that “normal” and “disorder” are
unique categories rather than part of a continuum.
• None of the tests provided any evidence of predictive
validity (as defined by McCauley & Swisher, 1984).
• Is this criterion an unreasonable burden (see
AAPS-3 manual)? Long-term studies of outcomes
prior to publishing are not commercially practical.
They may also require withholding treatment.
• Diagnostic accuracy is related to predictive validity and
may be a more appropriate criterion. Most of the tests
(7/11) presented diagnostic accuracy findings but used
group differences; only the DEAP-A also included
specific findings for sensitivity and specificity.
• McCauley & Swisher (1984) suggest that correlation
values for test-retest and inter-examiner reliability should
be 0.90 or higher. The current study only determined
whether should correlations were reported.
• None of the tests reviewed met all 13 criteria, but all met
at least six. Four tests met 10 or more criteria.
• None of the five articulation tests reviewed by
McCauley & Swisher (1984) met even 4/10 criteria
suggesting that our tests have improved since then.
REFERENCES
McCauley, R.J., & Swisher, L. (1984). Psychometric review of language and articulation tests
for preschool children. Journal of Speech and Hearing Disorders, 49, 34-42.
Download