Content area
Abstract
Self-administered, online survey and questionnaire instruments are ubiquitous in research. They are highly visual and involve spatial information processing in addition to the cognition involved in response formulation. Despite its wide use, the linguistic focus of survey research is still largely oriented to Eurocolonial and English-speaking contexts. There is evidence that the writing direction of a person’s language produces an effect on the processing and execution of spatial tasks. This is referred to as spatial agency bias. For readers of right-to-left (RTL) languages, such as Arabic, Persian, and Urdu, this spatial agency bias means that we can expect to see an effect in directionally-dependent visuo-spatial or visuo-motor tasks. Online questionnaire and survey instruments are developed for and by left-to-right (LTR) language readers. Given an increasingly diverse, diasporatic global population, it is important to consider how research methods developed within one linguistic context affect data quality when used in more diverse populations. This dissertation takes an experimental approach to explore the relationships between unconscious bias and other effects produced by an interaction between writing system direction and response scale direction in self-administered, online questionnaire instruments. Three experiments attempt to identify any interaction between two independent variables: 1) writing system direction for questionnaire response language) and 2) response scale category order (direction) on three dependent variables: 1) score on scale items, 2) duration of time spent in instrument, and 3) trust in the questionnaire interface. Participants are Arabic or English speakers responding to an instrument presented in one of these two languages. The analysis found no interaction effect between the independent variables on either trust or time on response, but did find a significant interaction effect on mean scale score for horizontal response scale items. This study has implications for multilingual, international, and cross-cultural survey and questionnaire design. This work contributes to efforts to incorporate more diverse populations in research through better understanding how language context affects data collection.
