Computer-Mediated Communication

Surveys

The goal of survey research is to collect information from respondents to describe their characteristics and/or to investigate the relationships among those characteristics. For instance, a survey can compare frequent Twitter users with “lurkers” (i.e., those who have a Twitter account, but do not tweet). Such a survey might measure interpersonal involvement on Twitter, number of Twitter followers, and other relevant factors to understand how those who tweet differ from lurkers. Survey research typically collects data from a large number of participants along with their demographic information such as age, gender, education, ethnicity, and income level. Surveys can ask about naturally occurring phenomena or they can be experimental (as will be described in the following section). A survey is designed as a series of questions that assess participants’ evaluations of information. The format of survey questionnaires can be open-ended (e.g., “How do you respond when someone comments on your posts on Facebook?”) or closed-ended with fixed response options (“How often do you comment on someone’s Facebook page?” 1 = Very often, 7 = Not at all).

When using surveys, the selection of sample (sampling) is very important. How much a sample resembles the characteristics of a population of interest influences the quality of data and generalizability of results. Performing online survey research is often less expensive than traditional pen-and-paper surveys, unless researchers need to have a fully random sample. Although it is preferred, random sampling usually costs more than using convenience sampling (often with college students). When researchers need to reach randomly selected participants, either by phone calls or mail, random sampling costs much more time, energy, and effort for both researchers and respondents than surveys collected online. A comparative analysis verifies that the practice of collecting survey data online provides results that differ very little (and perhaps not meaningfully) from those collected via paper-and-pencil. The following paragraphs detail an example of survey research.

Survey data have been the basis of a great deal of CMC research. Surveys allow researchers to examine either cross-sectional (i.e., at one point in time) or longitudinal (i.e., over an extended period of time) data. Nicole Ellison and colleagues have used surveys to study how college students’ Facebook usage is related to their social capital (i.e., the resources that become available through social connections to others). Their surveys measure variables including psychological well-being, communication behaviors, as well as Facebook usage. These variables have been analyzed in relation to three types of social capital: bridging, bonding, and maintenance capital.

...