WHODAS 2.0 Statistics, intenral consistency statistical analysis

Download a professional PowerPoint presentation on the World Health Organization Disability Assessment Schedule (WHODAS 2.0) focusing on internal consistency and reliability. This slide deck explains how WHODAS 2.0 measures disability across domains, highlights psychometric properties, and demonstrates Cronbach’s alpha findings for clinical and research use. Ideal for educators, clinicians, researchers, and students in psychology, counseling, rehabilitation, and public health.

1. STATISTICS IN
PSYCHOMETRICS
WHODAS 2.0
Brief overview of internal reliability,
subscale reliability and construct
validity.

2. STAT
INSIGHTS
01
Item, Subscale, Construct Reliability –Ratio
of variance to the total variance.
02 Cross cultural applicability
03
Field studies –reliability and validity –
Gen pop., Pop. Physical prob., pop. with
mental or emotional prob., Pop. With prob.
w/ SUD or AUD.
04 Concurrent validity
05 Responsiveness Models -Rasch Scale –
Item difficulty vs. item discrimination

3. GLOBAL
100 countries -
96 items to 36 items to 12 items.
Studied on N=≥18
Work sample –studies used “gainfully
employed” vs. unemployed
Focus groups and concurrent application
Identified factor structures
that would allow measure to
be shortened (high
reliability items)

4. APPLICABLE
MEASURES
1.Medical outcomes study 12 item shofrtform
health survey (SF-12)
2.And 36 item version (SF) 36
3.London Handicap Scale (LHS)
4.WHOQOL(23)
5.WHOQOLBrief Scale (WHOQOL-BREF)
6.Optional use of ICFChecklist
Used concurrently with applications:

5. VERSIONS
1.Most detailed
2.Generates scores for six domains
3.Also allows calculates overall functioning score
4.20 minute application time
5.36 items
6.3 different forms – interviewer, self, or proxy administered
7.3 “30 day severity questions [record number of days these difficulties
were present.”
36 item

6. VERSIONS
1.Brief
2.Generates scores for six domains
3.Overall functioning Sum score, compared to subscale and overall
4.5 minute application time
5.12 items
6.3 different forms – interviewer, self, or proxy administered
7.3 “30 day severity questions [record number of days these difficulties
were present.”
12 item

7. VERSIONS
1.Uses same rules – is “item response theory” dependent
2.Uses 12 items to screen for problematic domains
3.Based on positive responses to the initial 12 items, respondents may
give un up to 24 additional questions if positive in one domain
4.Can only be administered by interview or computer adaptive testing
(CAT).
5.Average time 20 minutes.
12+24 item

8. RESPONSIVENESS
1.Measured against disorder severity (by clinician
judgment) or standardized measures (Clinical
global impression, Hamilton depression Rating
scale) with another disability measure (LHS or
SF-36)
2.Tested in dichotomized version
3.Against Rasch model – both samples and both
versions (work and excluding work)
4.Polytomous items

9. ICF
CHECKLIST
1.https://cdn.who.int/media/docs
/default-
source/classification/icf/icfche
cklist.pdf?sfvrsn=b7ff99e9_4&d
ownload=true

10. TEST RETEST
RELIABILITY Cognition = .94
Mobility = .96
Self Care = .95
Getting along = .93
Life activities = .94
Household = .94
Participation = .95
Test-retest reliability had
intraclass coefficient (ICC) of
- .69-.89 at item level;
- .93-.96 at domain level
-.98 at overall level
-ICC -0-1 How strong
measurements in the same
group resemble each other
Internal consistencies used
Cronbach’s alphas (measure
of how well a set of variable or
items measures a single,
unidimensional latent construct
[disability]).

11. TEST RETEST
RELIABILITY
Test-retest reliability had
intraclass coefficient (ICC)
Wave 2 study – Use
analysis of variance (ANOVA)
to assess consistency and
agreement among different
ratings, such as those made by
different institutions
Wave 1 factor analysis – Two
level hierarchical structure with
one general disability factor
feeding into the six domains.
(Exception was “leisure:
belonged to “participation”)
Cognition = .94
Mobility = .96
Self Care = .95
Getting along = .93
Life activities = .94
Household = .94
Participation = .95

12. UNIDIMENSIONALITY
•Data fits the model it indicates that the scale is truly
unidimensional – it measures a single underlying trait

13. CRONBACH’S
ALPHA VALUES
ITEM-DOMAIN-
CONSTRUCT
Item-Domain-Construct

14. RELIABILITY
SUBGROUP

15. CROSS CULTURAL
SENSITIVITY TO
CHANGE
•Responsiveness studies- evaluating how well a
measurement tool accurately detects changes in the
underlying concept it is meant to measure over time
•AKA Longitudinal Validity
•Cross cultural applicability found high – summary change
scores were unaffected by socio demographic factors –
applicable across cultures
•Responsiveness studies across different scales
•At least as sensitive to change as other measure of social
functioning
•Effect sizes range – outpatient care depressed elderly vs.
outpatient care of newly referred schizophrenia cases -
•Multilevel analysis summary change scores were unaffected
by sociodemographic factors
•Is applicable across cultures
•Polytomous items fit extension Rasch scale (discrimination
vs,. Difficulty) – construct, measure single latent trait,
•Polytomous version was compatible with partial credit model
(identifying single latent trait can account for “partial credit”
on differing items.

16. RESOURCES •Columbia University Mailman School of Public Health.
(n.d.-a). Rasch modeling. In Population Health Methods.
Columbia University Mailman School of Public Health.
Retrieved August 26, 2025, from
https://www.publichealth.columbia.edu/research/popula
tion-health-methods/rasch-modeling