The effect of attitudes and experience on respondent answers to demographic questions in on-line surveys

\n\t\t\t\t\tSurvey response rates and response quality are declining in most markets, threatening the viability of many established research techniques. One attempt to combat this has been to make survey completion more convenient for respondents, with on-line surveys being an example of this. This paper looks at respondent attitudes to demographic questions in on-line surveys – questions that are crucial for many analysis practices and yet often have the highest non-completion rates. The exploratory survey conducted here (n=198) examined attitudes towards surveys and empirically examined the reasons behind non-completion and intentions to give misleading information. The results suggest that general attitudes towards providing demographic data in on-line surveys are related to the likely response behaviour, and the more uncomfortable someone claims to be with providing this data, the more likely they will be to either omit a response or exit the survey altogether. Past experience with on-line surveys increases the chances of respondents answering completely and accurately and in most cases, significantly reduces the likelihood of exiting the survey completely if asked these questions. Finally, there is some evidence that providing broad categorical response options is preferred to more specific categories or open text boxes, and this may represent the best technique for improving response rates and quality. \n\t\t\t\t