Article Text

PDF

0132 Do participants who complete a telephone survey in a language other than English differ to those who complete the survey in English?
  1. Terry Boyle1,
  2. Renee Carey1,
  3. Susan Peters1,
  4. Deborah Glass2,
  5. Lin Fritschi1,
  6. Alison Reid1
  1. 1The University of Western Australia, Perth, Western Australia, Australia
  2. 2Monash University, Melbourne, Victoria, Australia

Abstract

Objectives Limited research indicates that using English-language only surveys in prevalence studies conducted in the general population or in specific ethnic populations may result in unrepresentative samples and biassed results. In this study we investigated whether participants from ethnic minorities who chose to complete a study interview in a language other than English (LOTE) differed from those who completed the interview in English.

Method This study was conducted within the Migrant Australian Workplace Exposure Study, a population-based telephone survey that assessed the prevalence of exposure to occupational carcinogens among 749 workers of Chinese, Vietnamese and Arabic ancestry. The study was conducted in Australia in 2013. Modified Poisson regression determined the demographic factors associated with completing the interview in a LOTE.

Results Participants who completed the interview in a LOTE differed from those who completed the interview in English on several demographic factors, including sex, city of residence and country of birth. They were more likely to have a post-school qualification and to speak a LOTE at home, and were also more likely to be exposed at carcinogens at work compared with those who completed the interview in English (40% compared with 29%, PDifference <0.01).

Conclusions The participants who choose to complete the study interview in their native language had several demographic differences to those participants who completed it in English, and were more likely to be exposed to carcinogens at work. Prevalence studies that offer only English-language study instruments are unlikely to produce representative samples of minority groups, and may therefore produce biassed results.

Statistics from Altmetric.com

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.