Comparing online and offline modes of survey administration: VAA versus traditional survey administration

Abstract

This paper addresses the issue of potential bias in VAA generated data. Specifically, we report the findings from a study whereby a VAA questionnaire was administered through two different modes. One group consisted of participants who filled an online version of the VAA while the other consisted of participants elicited through a traditional survey. We find that there is differential respondent behaviour among the two groups suggesting that VAA-generated data is not fully comparable to traditional survey data - though this does not mean VAA data is better or worse but rather that both modes of administration are subject to different forms of methodological bias.