When most people focus on the digital divide they think of those without the technology requisite to becoming part of the sampling frame. However, there is another dimension to the digital divide that is not commonly discussed. How many people have internet and a phone, but will not participate in an internet survey while they will participate in the phone survey? This non-response error could be just as devastating as coverage error to a well run research design (Groves, 1989). In the years that Western Wats and Opinion Outpost has conducted phone recruit to web panels and phone recruit to web surveys, we have encountered lower response rates than to regular RDD phone surveys. We hypothesize that
there is a significant aspect to the digital divide that is social rather than technological in nature. To test our hypothesis we conducted a RDD phone survey of the individuals in the United States with both internet access and a phone. These respondents were called and asked to complete a web survey. Those that agreed to participate were sent an email invitation to take a 10 minute survey about attitudes towards online surveys, shopping, and social media sites. Respondents who did not agree take the web version were given a shortened 5 minute. Those that agree to take the web version and did not take after several reminders were contacted again on the phone and given the 5 minute version of the survey. The sample contained 9,879 valid residential numbers with a contact rate of 38% and a cooperation rate of 15% which resulted in 581 interviews. Data differences in attitudes and behaviors are highlighted in this presentation.