A preferred approach for the cognitive testing of translated materials: testing the source version as a basis for comparison

In recent years, a growing amount of research has demonstrated that parallel development of different language versions of a survey instrument is a preferred method to enhance the collection of equivalent data across language groups. This method involves the simultaneous development and pretesting of different language versions of a questionnaire in order to allow for two-way feedback and to maximize the likelihood that each version is culturally and linguistically appropriate for its target population. Unfortunately, it is common practice among many large survey organizations to pretest and finalize source language versions of survey questions and other materials prior to having them translated. Pretesting of translations does not always take place and when it does, the source language documents have often been finalized and the results of the translation pretesting cannot have any impact on further revisions to the source documents. This paper describes a methodological approach for the cognitive testing of a translation that can be used when pure parallel development of different language versions of a survey instrument is not feasible. In short, we examine the advantages of cognitively testing both source language materials and translations concurrently, even when source materials have already been finalized. Methodological, cost, and logistical implications are discussed.