Back-Translation Guidelines

Introduction

Translation from one language to another is a fairly complicated exercise. There are obstacles associated with 1) the different structures of the languages – these are linguistic obstacles; 2) the different psychosocio-ethnological differences – these are cultural obstacles. At the core of the exercise is the fact that there is an underlying philosophy for each language.

The process of translation and adaptation of research instruments seeks to create different language versions of the English instrument that are conceptually equivalent in each of the target countries/cultures. The instrument should be equally natural and acceptable and should practically perform in the same way. The focus is on cross-cultural and conceptual, rather that on linguistic/literal equivalence. A well-established method to achieve this goal is to use forward-translations and back-translations.

Several steps are involved in the process, as follows:

1. Forward translation
2. Use of expert panel
3. Back translation
4. Harmonization
5. Pre-testing
6. Preparation of the final version

The following are guidelines for the translation of the GENACIS questionnaire for each step of the
process.

General Guidelines for Translation

1. Forward translation

One translator, preferably a health professional familiar with terminology of the area covered by the
instrument and with interview skills, should be given this task. The translator should be knowledgeable of the English-speaking culture but his/her mother tongue should be the primary language of the target culture.

The following principles should be considered in the process of translation:

• Translators should always aim at the conceptual equivalent of a word or phrase, not a word-for-word translation, i.e., not a literal translation. They should consider the definition of the original term and attempt to translate it in the most relevant way.
• Translators should strive to be simple, clear and concise in formulating a question. Fewer words are better. Long sentences with many clauses should be avoided.
• The target language should aim for the most common audience. Translators should avoid addressing professional audiences such as those in medicine or any other professional group. They should consider the typical respondent for the instrument being translated and what the respondent will understand when she or he hears the question.
• Translators should avoid the use of any jargon. For example, they should not use: technical terms that cannot be understood clearly; and colloquialisms, idioms or vernacular terms that cannot be understood by common people in everyday life.
• Translators should consider issues of gender and age applicability and avoid any terms that might be considered offensive to the target population.
• Translators should avoid terms that have connotations, i.e., terms with the same meaning but with pejorative or positive dimensions or aspects not present in the source language.

2. Use of expert panel

The goal of this step is to identify and resolve the inadequate expressions/concepts of the translation, as well as any discrepancies between the forward translation and the existing or comparable previous versions of the questions, if any. The expert panel may question some words or expressions and suggest alternatives. Experts should be given any materials that can help them to be consistent with previous translations. Principal investigators and/or project collaborators will be responsible for providing such materials. The number of experts in the panel may vary. In general, the panel should include the original translator, experts in health, as well as experts with experience in instrument development and translation.

The result of this process will produce a complete version of the questionnaire.

3. Back-translation

Using the same approach as that outlined in the first step, the instrument will then be translated back to English by an independent translator, whose mother tongue is English and who has no knowledge of the questionnaire. That is, this translation should be done without the help of the original version. The attempt here is to reproduce an original version of the items using only the translation that was produced.

Again, emphasis in the back-translation should be on conceptual and cultural equivalence and not
linguistic equivalence. Discrepancies should be discussed and further work (forward translations,
discussion by the bilingual expert panel, etc.) should be carried out as many times as needed until a
satisfactory version is reached.

The validity of the new English version (produced by the back-translation) will be determined by the
degree to which this new version reproduces truly the original version. It thus becomes fairly easy to identify the weaknesses of the translation.

4. Harmonization

When adapting one instrument in several countries (as in GENACIS), it is important to check if all the country versions have made similar decisions with regard to culturally sensitive items. Since some cultures share the language of origin (e.g., Latin), it is helpful to review if the versions of the instrument in those countries have followed the same directions (i.e., if the same alternative has been chosen when several exist in each culture/language). This important step will be the responsibility of a central body, for example, the WHO in the case of studies funded by that organization.

5. Pre-testing

The objective of this test is to determine if the items that compose the experimental version of the survey questionnaire are clear, written without ambiguity and in a language that is well understood by the target population, in our case, the general population.

The following guidelines should be considered in administering the pre-test.

• Pre-test respondents should include individuals who are representative of those to be administered the questionnaire.

• Pre-test respondents should be administered the instrument and be systematically debriefed. This debriefing should ask respondents what they thought each question was asking, whether they could repeat the question in their own words, what came to their mind when they heard a particular phrase or term. It should also ask them to explain how they chose their answer. These questions should be repeated for each item.

• The answers to these questions should be compared to the respondent's actual responses to the experimental questionnaire for consistency.

• Respondents should also be asked about any word they did not understand, as well as any word or expression that they found unacceptable or offensive.

• Finally, when alternative words or expressions exist for one item or expression, the pre-test respondent should be asked to choose which of the alternatives conforms better to their usual language.

• This information is best accomplished by in-depth personal interviews, although the organization of a focus group may be an alternative.

• It is very important that these interviews be conducted by an experienced interviewer.

The following methods are suggested for the pre-testing of the questionnaire:

• In the test-retest technique, two interviewers question the same participants (minimum N = 20) at different times. If the items of the survey are clear, the results should be similar.

• Alternatively, respondents can be asked to circle items that are not clear so that the circled items can be re-examined several times afterwards.

• A variant of this procedure is to ask several participants to read each item and to judge its clarity on a scale from 1 to 7. Those with the lower score (4 and less, for instance) should be reexamined.

At the conclusion of the pre-test, a written report of the exercise, together with selected information regarding the participating individuals, should be provided.

6. Preparation of the final version

The final version of the instrument in the target language should be the result of all the steps described above. It is important that a serial number (e.g., 1.0) be given to each version.

Documentation

All the cultural adaptation procedures should be traceable through the appropriate documents. These include, at least:

• initial forward version
• a summary of recommendations by the expert panel
• the back-translation
• modifications agreed to during the international harmonization
• a summary of problems found during the pre-testing of the instrument and the modifications proposed
• the final version

It is also necessary to describe the samples used in this process (e.g., the composition of the expert panel and the pre-test respondent samples). For the latter, the number of individuals as well as their basic characteristics should be described, as appropriate.

Conclusion

The translation of items is more than a simple activity. It is a dynamic process in which the new version is submitted to evaluations to verify if the contents are a valid representation of items in the original version. Though this seems very demanding, a bad translation is more costly at the end of the process since the comparative nature of the GENACIS study will be compromised. If we cannot guarantee that the items are similar, then the comparative results could be erroneous.