The seven steps provided by Sousa and Rojjanasrirat (2011) for the cultural adaptation of health care assessment instruments primarily focuses on translation. This does include some cultural considerations as well, and they suggest it best to use translators who are not only fluent in both languages, but also from the culture in question, in order to capture meanings accurately rather than a direct word-for-word translation.
The seven steps are:
- Initial translation from the original language of the instrument into the language sought for adaptation by two translators. One of translators should also possess fluency in the scientific terminology of the discipline being used, while the other should not be, but still familiar with common expressions related to the discipline.
- Initial synthesis – comparing both the original and translated instruments by a separate translator for accuracy of meanings. Any discrepancies would be handled by a meeting between all translators to decide on the best meaning to use.
- The newly updated translated instrument from the previous step is then back-translated to the original language by two other translators who have never seen the original instrument. These translators should possess similar qualities to the initial two translators.
- Secondary synthesis – The two back-translated instruments are compared, and any discrepancies are similarly resolved as the initial synthesis by consensus, and all versions of both the translated versions and back-translated versions are compared.
- Pilot test the first draft of the instrument with individuals who speak the target language only and are not bi-lingual. The participants rate the questions, instructions, and items on the assessment itself for being clear or unclear. The use of an expert panel who are within the discipline or have knowledge of it is also recommended for this step to evaluate the clarity.
- Preliminary stage testing with participants who are bilingual. This often is skipped over in testing instruments, but it can be a valuable step if it is incorporated.
- Complete full testing with the final draft of the translated instrument using a sample from the target population. This test can help to iron out any final inconsistencies or issues, and can be used to determine the general validity and reliability of the instrument.
Raghavan (2018) shares recommendations when conducting assessments for survivors of torture, but, these suggestions are also highly useful to consider for any cross-cultural adaptation of assessments and instruments. These recommendations are based in the premise that cultural contextual factors are critically important to understanding, assessing, and treating how individuals from differing backgrounds express, conceptualize, and experience mental illness. Some researchers believe that inconsistencies in assessments and instruments cross-culturally are not due to actual differences in the rates and underlying symptomology of a mental illness, but rather that the measurement tool does not accurately portray definitions or options which reflect the cultural viewpoint of those being assessed.
These strategies include:
- Cultural idioms of distress. How mental illness manifests within a culture, how it is commonly understood and described, and experienced. Many collectivistic cultures use somatic symptoms to express mental distress, for example.
- Impact of Interpreters. When clients or participants must use an interpreter to communicate with a provider or researcher, the true meaning of what they try to convey may be misconstrued or lost. Furthermore, the skills and bias of the interpreter may alter both the meaning of the client/participant and the clinician/researcher, and the interpreter may suffer from secondary trauma. It would be best to use a researcher/clinician who is already fluent in the client’s language, but, if this is not possible, using an interpreter is still a better option than not using one.
- Cross-cultural equivalence of measures. Five criteria are proposed to determine if measures are equivalent in differing cultures, such as being contextually relevant, differences in the cultural understanding of constructs, or if the method of data collection itself creates a response bias, or is inaccessible to some.
- Adaptation and Translation of Measures. Here Raghavan (2018) seems to agree with the steps of translation proposed by Sousa and Rojjanasrirat (2011).
- Use of culture-specific normative data. This consists of the baseline sample to which the assessment would be comparing. For example, the normative data would be the general rate of PTSD among community members in Guatemala, but the assessment would be measuring rates of PTSD of Guatemalan refugees. Rather than comparing to PTSD rates among worldwide populations, this ensures that there is accuracy within a cultural framework.
In reviewing a study by Oe et al. (2020), the researchers did make use of the above steps by Sousa and Rojjanasrirat (2011) and recommendations by Raghavan (2018). They used a Japanese-developed trauma screen, the TEC-J, and compared this with the Global Psychotrauma Screen (GPS) which was developed elsewhere and then modified/translated for use in Japan using the guidelines by Sousa and Rojjanasrirat (2011). The use of the TEC-J is important because it was developed internal to the culture in question, and therefore included culturally relevant considerations in its design (Oe et al., 2020). However, this may be outdated, as the TEC-J was developed in the 1990s and therefore avoided asking about highly taboo topics in Japan such as childhood sexual abuse. It is unclear whether this would still be such a taboo today to the point that it would be avoided on an instrument all together. In their analysis, Oe et al. (2020) consider cultural factors which may have impacted the scores particularly on the GPS, including response bias as mentioned by Raghavan (2018). One limitation, which is mentioned by the authors (Oe et al., 2020), is that the sample was skewed to those with severe trauma and who were seeking help; in other words, no normative data for comparison (Raghavan, 2018).
References
Raghavan, S. S. (2019). Cultural Considerations in the Assessment of Survivors of Torture. Journal of Immigrant and Minority Health, 21(3), 586-595. http://dx.doi.org/10.1007/s10903-018-0787-5
Sousa, V.D. & Rojjanasrirat, W. (2011). Translation, adaptation and validation of instruments or scales for use in cross-cultural health care research: A clear and user-friendly guideline. Journal of Evaluation in Clinical Practice, 17, 268–274.
Oe, M., Kobayashi, Y., Ishida, T., Chiba, H., Matsuoka, M., Kakuma, T., Frewen, P. & Olff, M. (2020). Screening for psychotrauma related symptoms: Japanese translation and pilot testing of the Global Psychotrauma Screen. European Journal of Psychotraumatology, 11(1).