Table 1 Steps for developing, translating and validating questionnaires for use with migrant women
From: Cross-border movement and women's health: how to capture the data
1) Identification of the variables (based on the literature and input from migration/health experts) | Advisory committees consisting of healthcare professionals, representatives from non governmental organizations, and government officials all interested in migration and the reproductive health of migrant women are consulted |
2) Identifying/drafting questionnaires | Questionnaires are identified by an extensive review of the literature |
3) Assessing readability | Readability is assessed by counting the number of polysyllable words in each question1 |
4) Translation | To maximize the application of each translation to a broad population using the same language, translators with different backgrounds for each language are identified (e.g., Colombian and Mexican backgrounds for Spanish). The purpose and context of the studies in which the tool(s) will be administered are explained to translators Translators are instructed to use simple language and to avoid idioms and regional terms/expressions.2 |
5) Blind back translation | "Blind back-translation" is translation back into the source language by an independent translator unfamiliar with the original version of the questionnaire; Back-translated versions are compared to the original language version and discrepancies in wording noted and each item discussed and debated until agreement is reached on the optimal wording for the translated versions. When clarity is lacking in the original questionnaire, adjustments are made across all versions (original and translated). |
6) Discussion groups with representatives from different ethno cultural communities | Migrant women representing a mix of ethno-cultural communities (Asia, South America, Africa, Europe) are asked to qualitatively assess the content validity and acceptability of the questions (i.e., feasibility to complete and cultural appropriateness); Groups generally consist of 5-10 participants |
7) Administration of the translated questionnaires with monolingual individuals | Individuals who speak one of the 'translation languages' but not the original source language of the questionnaire, are asked to assess grammar, and ease of understanding of the translated version; They also assess the practical aspects of administration; 5-10 participants per language |
8) Reliability testing (test retest and internal consistency as appropriate) of each language version as well as between the English and translated versions. | Reliability testing is completed via administration of the original and translated versions to persons fluent in both languages to ensure all language versions are understood in the same way and are equivalent |