4
Application of the International Test Commission (ITC) Guidelines in reducing bias in test adaptation: An illustration from the Addenbrooke’s Cognitive Examination III for the Bengali speaking population in India
Ranita Nandi, Duttanagar Mental Health Centre, Kolkata, India
Rahul Venkatesh, Duttanagar Mental Health Centre, Kolkata, India
Sulakshana Rao, Duttanagar Mental Health Centre, Kolkata, India
Pallavi Bhargava, Kokilaben Dhirubhai Ambani Hospital, Munbai, India
Arpita Bose, University of Reading, Reading, United Kingdom
Amitabha Ghosh, Apollo Multispeciality Hospitals, Kolkata, India
Jonathan Evans, University of Glasgow, Glasgow, United Kingdom
Objective:
Adapting neuropsychological tests for use in different languages and cultures require a systematic approach to ensure the tool is free from construct, item and method bias. However, rigorous and uniform methods are often not implemented during test adaptation across different languages and cultures. We undertook a culture-specific approach to adapt the Addenbrooke’s Examination (ACE) III subtests for Bengali speakers in India by implementing the International Test Commission (ITC) guidelines and test adaptation frameworks to add to diagnostic accuracy of the Bengali version of the ACE II (BACE III) for early detection of cognitive impairment and dementia.
The BACE III was adapted through six phases. The adaptation process included translation of instructions, stimulus adaptation utilizing culture, language or familiarity driven frameworks, item selection, and updating scoring and administrative instructions based on the ITC guidelines for translating and adapting tests. A judgmental design was used for forward and backward translation. Stimulus adaptation initially involved categorizing subtests based on the underlying construct and test principles. This was followed by adaptation of stimuli based on expert consensus only through Delphi technique or development of corpus of items through Delphi technique prior to item selection. Pre-piloting was conducted to evaluate the understanding and acceptability of the items among the Bengali speakers. For subtests requiring item selection, item difficulty and discrimination statistics were used during the piloting phase, based on responses from 30 patients, diagnosed with Vascular Mild Cognitive Impairment, vascular dementia and probable Alzheimer's disease, and 60 healthy controls matched for age and education.
Out of the 18 guidelines outlined in the ITC, 12 guidelines from ‘pre-condition’, ‘test development’, ‘administration’ and ‘documentation’ were implemented to translate and adapt the ACE III for the Bengali speaking population. The adaptation process led to thirty-six changes, including adaptation of stimuli in the orientation, memory and language subtests, and changes in instruction (phonemic fluency) and scoring (verbal fluency; clock drawing test), BACE III subtests showed good psychometric properties, with item difficulty ranging from .47 to .98, item discrimination ranging from .60 to .87 and internal consistency ranging from .66 to .92.
Drawiing on the ITC Guidelines was found to be helpful in ensuring that the ACE III was adapted in a way that is appropriate to the linguistic/cultural context in which it is intended to be used. We recommend use of the neuropsychological application of the ITC guidelines to ensure effective adaptation of neuropsychological tests to new cultural/linguistic contexts both across and within nations.