Internet surveys should be viewed as a complement to other data collection methods, and not necessarily as a replacement.
The advent of the Internet has clearly changed the face of survey research and posed many new challenges, as well as provided new opportunities for reaching respondents in alternative ways. The use of Internet surveys has grown substantially and preference for Internet surveys continues to increase in popularity.
In CMOR’s 2003 Respondent Cooperation & Industry Image Study, preference for Internet surveys continues to increase. In 2003, Internet survey preference (18%) increased 44% from 1999 (10%). Although mail surveys remain the preferred methodology, this is most likely due to its convenience. Convenience was also a key factor in the rise in popularity of Internet surveys, combined with the general increase in Internet popularity and accessibility.
As in the previous two waves of this longitudinal study, telephone respondents’ attitudes and behaviors were compared to an Internet sample. Since Internet respondents were part of a panel and received a small incentive, their responses are not directly comparable in all categories to the telephone respondents who were contacted via RDD. However, it does give us some insight on how to reach different types of respondents.
The demographics of the Internet respondent show that they were generally younger and better educated. There was only a slight difference in income. More Asians were reached through the Internet, while more African Americans were contacted via telephone.
Internet respondents tend to enjoy technology and use it, as seen in the high number that own their own computers and call screening devices. They prefer using a computer over using a telephone. Clearly, they take extra precautions against unwanted telephone calls as seen by the higher use of all three types of special services: Caller ID, Call Blocking, and Distinctive Ringing. Interestingly though, the Internet respondents of this survey were less likely to have their telephone number unlisted.
In general, Internet respondents had more positive and less negative attitudes toward the survey research industry compared to telephone respondents. This attitude could be related to fact that they often chose to participate in an Internet panel, and can complete a survey at their leisure or even over the course of several days. They are not bothered at “dinnertime” or during a favorite television show or during any other important event in the home, as compared to a telephone call for telemarketing or legitimate survey research. Fewer Internet respondents felt that survey research was an invasion of privacy probably because they feel they are in control of their survey participation and their own personal time schedule. Regarding telephone surveys, the Internet respondents usually don’t answer the phone and allow their calling devices to screen the calls. However, reported past year refusal rates were similar for both phone and Internet respondents, but over time, telephone respondents still refuse more on average.
So what have we learned about Internet respondents? These respondents embrace surveys as a high-involvement activity. After all, most are members of online survey panels, reflecting their agreement to participate in multiple studies. The good news is that they appear to be unusually interested in taking surveys. They are covetous of their time and privacy, so the challenge will be to reach them by telephone. One consideration is to offer multiple modes of completing a survey if you can actually reach someone by phone. But take heart. Once you reach these respondents, there’s a better chance they’ll participate. However, Internet surveys should be viewed as a complement to other data collection methods, and not necessarily as a replacement.
The research industry has an opportunity to begin using its industry identifier for all communications with respondents. This symbol and slogan which was recently developed by CMOR with its partners MRA and ARF, will be tested for use on questionnaires, as well as, call screening devices which may help elusive respondents to quickly identify a legitimate survey. But as always, surveys must be well-designed, interesting and short, and cooperation should not be the issue.