3.1 P13: Culturally Valid Diagnostic Measures
Bhui: Cultural Adaptation of Mental Health Measures: Improving Quality of Clinical
Practice & Research
Conceptual Issues
- Cultural relevance & validity
- Translating tests/measures
- Could use interview-based measures instead – but, expensive & time-consuming
- Mini International Neuropsychiatric Interview (MINI) = yes/no responses, but
still requires translation of questions
- Flaherty classified questionnaires according to their track record in cross-cultural
settings:
a) instruments w/a proven cross-cultural equivalence
b) instruments w/established validity & reliability in the original culture but not tested
in other cultures
c) instruments w/high face validity but untested in country of origin
Sartorius & Kuyken:
- Ethnocentric approach = total conceptual & psychometric equivalence, so
instruments used freely across cultures, don’t have to take cultural context into
consideration, instruments should be universal
- Pragmatic approach = advocated if there is a 60-70% overlap; limited adaptation, on
the assumption that psychometric properties are not significantly altered, and that
conceptual limitations exist but are not marked. ‘Derived etic’
- Where there is little conceptual overlap, this derived etic approach cannot guarantee
psychometric & conceptual equivalence – ‘emic plus one’ method recommended,
which advocates for use of original, literally translated questionnaire w/indigenous
instrument
- If there’s no conceptual overlap, then translation & validation are impossible.
Translation Issues
- Translators must have sufficient experience and know about cultural understanding of
mental distress & disorder
- Forward vs. backward translations: Edwards recommended choosing translators
who have learnt the language of the original version as a second language
Validation of Measures
- Face & content validity can be strengthened by focus groups, pilot studies &
consultation w/community agencies
- Ensuring criterion validity requires a gold standard – but may have no cultural
adaptation, e.g. some cultures don’t have gold standard to assess mental disorders
- Etic assumptions about universality of psychopathological disorders means we don’t
learn about cultural influences that influence measurement of disorder. So we don’t
learn about culturally constructed disorders that fall outside classifications of
psychological illnesses
, He: Bias & Equivalence in Cross-Cultural Research
Introduction
Bias = nuisance factors that jeopardise validity of instruments applied in different cultures.
Bias doesn’t refer to random errors but to systematic measurement anomalies that are
expected to be replicable if a study were to be repeated
Equivalence (lack of bias) = level of comparability of scores across cultures (e.g. km &
miles can’t be compared, so have to convert scales so they’re equivalent)
Equivalence is characteristic of cross-cultural comparisons & not to intrinsic property of
instruments; both km & miles are adequate units to measure distances and any lack of
equivalence issues arise only in comparison of both
Instrument Choice in Cross-Cultural Studies
Adoption
- Close translation of an instrument in a target language
- Simple to implement, cheap, high face validity – but can only be used when items in
source & target language have adequate coverage of construct measured
Adaptation
- Combination of close translation of some stimuli & change of other stimuli when
close translation would be inadequate for linguistic/cultural reasons
- Adaptation has become generic term to refer to translation process
- In the past, rendering an instrument in another language only required knowing the
other language, but now it also requires knowledge of target culture
Assembly
- Compilation of new instrument
- Only option if adoption & adaptation don’t work
- Assembly maximises cultural suitability of instrument, but will prevent any numerical
comparisons of scores across cultures
Selection Criteria
- If aim is to maximise opportunities for statistical comparisons in a study, adoption is
best choice
- If aim is to maximise ecological validity of instrument, adaptation/assembly is better
- Statistical tools like item response theory & structural equation modelling can deal
with instruments that are not completely identical across cultures. However, if number
of culture-specific items is large, comparability of construct or of remaining items
may be problematic and opportunities for cross-cultural comparisons are limited.
Taxonomy of Bias
Bias = when score differences on indicators of particular construct don’t correspond to
differences in underlying trait/ability (Van de Vijfer). This incomplete correspondence means
that a response in one culture could represent target construct, but responses in other
countries are due to other constructs/additional constructs
Bhui: Cultural Adaptation of Mental Health Measures: Improving Quality of Clinical
Practice & Research
Conceptual Issues
- Cultural relevance & validity
- Translating tests/measures
- Could use interview-based measures instead – but, expensive & time-consuming
- Mini International Neuropsychiatric Interview (MINI) = yes/no responses, but
still requires translation of questions
- Flaherty classified questionnaires according to their track record in cross-cultural
settings:
a) instruments w/a proven cross-cultural equivalence
b) instruments w/established validity & reliability in the original culture but not tested
in other cultures
c) instruments w/high face validity but untested in country of origin
Sartorius & Kuyken:
- Ethnocentric approach = total conceptual & psychometric equivalence, so
instruments used freely across cultures, don’t have to take cultural context into
consideration, instruments should be universal
- Pragmatic approach = advocated if there is a 60-70% overlap; limited adaptation, on
the assumption that psychometric properties are not significantly altered, and that
conceptual limitations exist but are not marked. ‘Derived etic’
- Where there is little conceptual overlap, this derived etic approach cannot guarantee
psychometric & conceptual equivalence – ‘emic plus one’ method recommended,
which advocates for use of original, literally translated questionnaire w/indigenous
instrument
- If there’s no conceptual overlap, then translation & validation are impossible.
Translation Issues
- Translators must have sufficient experience and know about cultural understanding of
mental distress & disorder
- Forward vs. backward translations: Edwards recommended choosing translators
who have learnt the language of the original version as a second language
Validation of Measures
- Face & content validity can be strengthened by focus groups, pilot studies &
consultation w/community agencies
- Ensuring criterion validity requires a gold standard – but may have no cultural
adaptation, e.g. some cultures don’t have gold standard to assess mental disorders
- Etic assumptions about universality of psychopathological disorders means we don’t
learn about cultural influences that influence measurement of disorder. So we don’t
learn about culturally constructed disorders that fall outside classifications of
psychological illnesses
, He: Bias & Equivalence in Cross-Cultural Research
Introduction
Bias = nuisance factors that jeopardise validity of instruments applied in different cultures.
Bias doesn’t refer to random errors but to systematic measurement anomalies that are
expected to be replicable if a study were to be repeated
Equivalence (lack of bias) = level of comparability of scores across cultures (e.g. km &
miles can’t be compared, so have to convert scales so they’re equivalent)
Equivalence is characteristic of cross-cultural comparisons & not to intrinsic property of
instruments; both km & miles are adequate units to measure distances and any lack of
equivalence issues arise only in comparison of both
Instrument Choice in Cross-Cultural Studies
Adoption
- Close translation of an instrument in a target language
- Simple to implement, cheap, high face validity – but can only be used when items in
source & target language have adequate coverage of construct measured
Adaptation
- Combination of close translation of some stimuli & change of other stimuli when
close translation would be inadequate for linguistic/cultural reasons
- Adaptation has become generic term to refer to translation process
- In the past, rendering an instrument in another language only required knowing the
other language, but now it also requires knowledge of target culture
Assembly
- Compilation of new instrument
- Only option if adoption & adaptation don’t work
- Assembly maximises cultural suitability of instrument, but will prevent any numerical
comparisons of scores across cultures
Selection Criteria
- If aim is to maximise opportunities for statistical comparisons in a study, adoption is
best choice
- If aim is to maximise ecological validity of instrument, adaptation/assembly is better
- Statistical tools like item response theory & structural equation modelling can deal
with instruments that are not completely identical across cultures. However, if number
of culture-specific items is large, comparability of construct or of remaining items
may be problematic and opportunities for cross-cultural comparisons are limited.
Taxonomy of Bias
Bias = when score differences on indicators of particular construct don’t correspond to
differences in underlying trait/ability (Van de Vijfer). This incomplete correspondence means
that a response in one culture could represent target construct, but responses in other
countries are due to other constructs/additional constructs
