Web Survey Bibliography
Title Participation in Web Surveys
Author Lozar Manfreda, K.
Access date 14.07.2004
Abstract Web surveys are computer-assisted self-administered interviews with computer questionnaires based on HTML, (usually) presented in a standard Web browser. Responses are (usually) immediately transferred to the server through the electronic network, (usually) the Internet. At the end of the millenium these types of questionnaires can be found on thousands of Web pages and Web survey panels already include millions of Internet users. However, nonresponse and noncoverage are the most critical features of these surveys what makes statistical inference in the vast majority of today’s Web surveys impossible. The aim of the paper (disseration study) is to identify factors influencing the participation of respondents in the Web surveys, from the likelihood to be included in the sample, being contacted, through their willingness and ability to cooperate, completely answering the questionnaire and finally willingness to participate in future surveys. These are stages in the Web survey participation process. The factors incluencing the survey participation in general are social environment, respondent's characteristics and survey design. These factors are somewhat different and differently interact in the case of the Web surveys. We are going to give an overview of the literature on the survey participation in general and compare it to experience from the Web surveys. In addition, we are going to study the above mentioned factors with two empirical research projects: 1. The factors, such as respondents' characteristics (social-demographics, techniqual equipment, psychological predisposition) and questionnaire design issues will be studied on data from an experimental study on participation in a Web survey in Slovenia. Respondents' participation will be analysed using survival anaysis, structural equation models and logit regression. 2. Other survey design issues and characteristics of the social and technical environment will be studied with a meta-analysis of Web surveys reported in the literature. We are interested in how do the design characteristics of the Web survey (type of the target population, strategies for contacting potential respondents, survey topic, questionnaire design, survey organization, etc.) and specifics of the social and the technical enivornment in different countries and cultures influence response rates. With our research we want to contribute to the evaluation of the use of the World Wide Web for conducting survey research. By identifying the respondents' characteristics and characteristics of the social and the technological environment we can estimate and predict who are people who usually participate in Web surveys. In this way the appropriateness of statistical inference from such surveys can be evaluated and adequate weighting strategies potentially developed. In addition, with identifying the survey design issues influencing participation, we want to contribute to the development of adequate Web survey design strategies aiming at achiving the sufficient unit response rates and data quality (low item nonresponse, low partial response, etc.) in Web surveys.
Access/Direct link Homepage - conference (abstract)
Year of publication1999
Bibliographic typeConferences, workshops, tutorials, presentations