Abstract Demography of Language is an inter-disciplinary research area that draws on the methodology and data of demography as well as sociology and linguistics. Its basic methodology is survey based and it relies on data collected by means of censuses of population and large-scale surveys. Despite the widespread use of language-related census and survey data in research and policy studies, their methodological underpinnings have not been systematically evaluated. In this paper, it is argued that census agencies, when they chose to include sociolinguistic topics in their questionnaires, use their own ad hoc measures, and provide little description of how they were developed. Problems arising from the variable wording, format or response categories of language related questions in censuses call into question the reliability and validity of data collected with census methods. The objective of this paper is to investigate the way in which respondents are influenced by the way language proficiency questions in censuses are shaped and worded in a large set of censuses relating to one language (Irish) in the Republic of Ireland and the Northern Ireland region of the UK. The analysis draws on earlier research on this topic, but expands the analytical scope of previous work by applying the theoretical insights of recent methodological advances in social research and applied linguistic research.
|
|
|
|
|