Year: 2002
Type: book-chapter
Abstract: The quality of data collection on the Internet depends in part on potentially biasing surface characteristics of Web pages, such as pop-up menus. Also, answering behavior to survey questions is highly sensitive to cognitive contexts evoked by other survey questions. A Web experiment with two experimental phases is used to identify six potentially biasing factors. Manipulations of scale type, reading directionality, cursor entry position, question order, and – in contrast to findings from offline research – type of numerical labeling do not show effects. However, it is shown that it makes a difference whether survey questions are presented together on one Web page or separately on consecutive Web pages. Results are discussed with regard to general as well as online specific implications for the design of survey instruments and multiple response scales.
Source: KOPS (University of Konstanz)
DOI: 10.5167/uzh-19732
Author: Reips, U D
Language: English
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W840683036
