Personal interest for a survey topic has proved to ubstantially influence the decision to participate in mail (Dommeyer 1985, Heberlein/Baumgartner 1978, Martin 1994) as well as web surveys (Galesic 2006, Marcus et al. 2007, Porter/Whitcomb 2005). According to the ‘Leverage Salience Theory’ of Groves, Singer and Corning (2000) the decision to take part is not only based on the different attributes of the given survey (e.g. topic, sponsor, length etc.) but will also be influenced by how salient those attributes are made to the invitee. Invitations for web surveys in online access panels therefore usually leave this information out to reduce the risk of an expert bias (Bethlehem/Stoop 2007). This study aims to bring forward whether online access panel members use the information about the survey topic in e-mail invitations in their decision on participating in web surveys.
In a preliminary study data about personal interest, attitudes and other psychographic variables of 1.670 panel members were collected. Panelists were then randomly assigned to participate in one of two surveys about different topics (insurances and holidays). The e-mail invitations to these surveys were manipulated. Panelists either received an e-mail with the topic stated in the subject line as well as the invitation text, an e-mail message where only the text referred to the topic or an e-mail without any information about the survey topic. The influence of announcing the topic in the e-mail was measured on response rate and different indicators of data quality (response latency, item omission, answer length to open-ended questions, non-differentiation in grids).