Web Survey Bibliography
Title A Comparison of the validity of a paper based and an online Conjoint Analysis
Access date 20.09.2006
Abstract The data for a Conjoint Analysis is traditionally gathered by personal interviews. Hence, reaching a sample, which is big enough for a Conjoint Analysis, is highly expensive and time consuming. By changing the data collection process from a paper based personal interview to the Internet, it is easier to contact larger number of people within a shorter amount of time. In addition, collecting the data on the Internet is a more cost effective method compared to a paper based personal interview. However, it is questionable if the data, which is gathered by using the World Wide Web is of the same quality as data, which is gathered the traditional way. Therefore, a scientific study was conducted in order to compare the validity of two data sets, where one data set was the result of a paper based survey and the other data was the result of an online survey. To be able to compare the two methods an online survey and a paper based survey were conducted for four weeks each. However, the two samples did not consist of the same test persons. But the demographic, gender and the profession of the persons who took part in the surveys was similar. In addition, the test questions were identical. While taking the survey the test persons were asked to rank and rate nine specific MP3 players according to their preference. These specific products were derived from an orthogonal main effect plan. Since ranking and rating nine different stimuli is a rather complicated task, it had to be assured that the test persons could visually move the stimuli. This was accomplished by using special HTML code in order to move the nine products on a test person’s computer screen. Special product cards for the paper based survey assured that the test persons, who took part in the paper based survey also had the possibility to visually rank the nine products. At the end of the questionnaire several questions about the degree of difficulty had to be answered on a six point rating scale by the test persons. The comparison of the data is done in two ways. First the personal judgments of the test persons regarding the degree of difficulty are compared. Second certain validity criteria are used to compare the quality of the data on an individual as well as on aggregate level e.g. face validity, internal validity (e.g. Persons R and Kendalls tau) and prognostic validity (e.g. Hit Rate).
Access/Direct link Conference homepage (abstract)
Year of publication2006
Bibliographic typeConferences, workshops, tutorials, presentations
Full text availabilityNon-existant