Online Data, Collection and Interpretation of

Survey Research

For surveys, the web has proven incredibly advantageous as a means of gathering responses quickly and efficiently and may contribute to higher response rates. A number of popular (and relatively inexpensive) software platforms have emerged, including Qualtrics and SurveyMonkey, which make creating a survey for online distribution both fast and user-friendly. Researchers can take advantage of these platforms to create custom survey instruments with a variety of question types (e.g., multiple-choice, scaled, multiple-response, open-ended) that can be distributed to a sample with little cost. Since only a web link is necessary to access an online survey, the survey can be distributed through e-mail or through online social networks. Typically, software packages log the Internet Protocol (IP) addresses of respondents as a means of preventing multiple responses from the same individual. Like many surveys, online survey research can suffer from self-selection bias and other concerns that may limit generalizability. Attempts should be made to mitigate these limitations, as is true with other, more traditional kinds of research. There is a great deal of literature, however, that suggests that the ease of access to online surveys increases response rates over what mail or telephone surveys would typically yield. In addition, with online survey research, the time between the creation of the survey instrument and distribution of the survey can be greatly compressed.

Online Collection Methods

Researchers in the communication fields can benefit from many data types that are readily available on the web, such as metadata (the material traces of web use that individuals leave). Metadata and archival traces of individuals’ Internet use have been the subject of controversial use by corporate interests seeking to harness information, including search engine histories, social network input, and visited websites for marketing purposes, a process that is referred to as “data mining.” While the use of archival information on the web is often tied to these kinds of purposes, academic researchers have used metadata and other kinds of data sets drawn from the Internet to answer many kinds of research questions that address use of the Internet and the communication patterns contained therein, as well as structural concerns about how websites and online communities are built. Archival data sets can be costly for researchers to obtain, though there are many collection tools available for purchase that can gather and analyze data, most commonly from social networking services like Facebook and Twitter. These software platforms can mine through vast collections of social media profiles and shared content based on parameters specified by the researcher. Moreover, there are literally hundreds of data collection tools in the marketplace. Data collection of this type does raise some ethical issues, particularly with regard to long-term effects of maintaining the data and the impacts on user privacy. This is particularly true since social network terms of service often grant wide latitude for site owners and their agents to reveal metadata, and this can go far beyond what a user expects when he or she signs up for the service. For researchers, basic tenets of anonymity (keeping identifiable information separate from the data) and informed consent should be observed as much as possible, and careful consideration must be given to measuring the value of the research against any potential harm to users.

...