While social media can provide real time insight into knowledge, attitudes and behaviors, twin challenges for health communication researchers are gathering accurate, detailed demographic information about users and grappling with a large quantity of qualitative information. Surveys aptly handle these challenges, but may be too slow in an emerging public health crisis such as an infectious disease outbreak. Further, surveys often underrepresent young, urban participants and minorities, groups which can readily be captured via social media platforms. Using the 2014-15 influenza season as a test case, we employ rapid, computational Big Data techniques that use advanced natural language algorithms and geolocation to parse demographics and psychographics of social media. We compare our social media findings to published CDC survey data about vaccination, reporting where the data complement and/or overlap each other, where they diverge, and importantly, for which demographic groups. Our results suggest opportunities for researchers to track emerging epidemics more quickly and cheaply via social media, as well as suggesting other instances where social media may be insufficient and survey data is necessary to provide depth of detail in easily quantifiable and analyzable formats.