When it comes to Web surveys, the type of attribute rating scale used does matter. Based upon data collected during the Burke Strategic Consulting Group's recent Workforce Perspectives® study, this paper shows how responses to drop-down box, radio button, and fill-in-the-blank survey interfaces can differ significantly from each other. Distributions of responses will differ depending upon the scale used. What might have an ever greater impact on utilization of survey findings is the fact that a radio button scale can drive respondents to more frequently use the same response option.
With the increasing use of the Internet as a medium for survey data collection, survey program managers might be encouraged to transition programs to the Internet that have previously been conducted using the telephone. They might also consider using a mixed-mode approach, combining data collected from both the Internet and telephone. Some caution, however, should be exercised. This paper describes a study designed to isolate differences in responses to Internet surveys and phone surveys relating specifically to differing survey methods. Results indicate that scale usage can vary considerably between the two methods, with phone respondents more likely to use endpoints of scales in which all response alternatives are verbally labeled. By contrast, responses are much more similar when a numerical scale is used and only the endpoints are verbally labeled. The study did indicate, however, that surveys can take less time for respondents using the internet and that internet respondents express more willingness to participate in future survey projects.