Data

Overview

Data can be anything. Data are all around us. Yet, it is important when doing research to consider what data are sought and how the data will be obtained for the research purpose and subsequent responses to hypotheses or research questions. Hypotheses are predictive statements about the relationship between variables, and research questions are statements about the relationship between variables without specific predictions (e.g., those which are exploratory and descriptive in nature). For instance, a hypothesis might state that women utilize more constructive conflict approaches, whereas a research question might question how sex affects conflict approaches. It is important to then draft the research design from the hypotheses and research questions. The variables dictate what researchers mean when they identify their concepts of interest. The design needs to obtain data that will appropriately reflect those variable conceptualizations articulated in the hypotheses and research questions. Once these are evident, then researchers can begin determining what data and analyses are relevant to their study.

Data can be derived either before or after the declaration of a theoretical framework. Commonly, data are structured in advance, meaning that researchers have an idea of the responses and answers that will be provided. For instance, in quantitative data, measurements are utilized to give data numerical structures. Thereby, priori is when data have preestablished values and often a theoretical framework established prior to data collection. Posteriori is when data are unstructured and there are no preestablished categories or codes. A theoretical framework might be applied following the data collection. Data cannot be utilized to both develop and confirm a theory because new data are necessary for each condition. The research purpose determines what data must be gathered and whether the data have methodological implications that can be utilized thereafter.

Sources of Data

Many individuals believe that data can only be numeric, but data can appear in many forms. Numeric research (quantitative) relies on the measurement or assignment of numbers to mark the characteristics of data. By contrast, it is difficult to account for nonnumerical data (qualitative, rhetorical, and critical) because anything a researcher can observe could count as data.

Most communication scholarship data comes from (1) self-reports, (2) other-reports, (3) observations, or (4) texts. Self-reports include all data [Page 336]collected by asking people to measure a target characteristic or behavior and comment or assess themselves. Examples include IQ tests, relationship breakup stories, and the Myers–Brigg personality type indicator. Other-reports ask people to describe or report how other people demonstrate particular characteristics or behaviors. Examples include subordinate evaluations, the partner jealousy scale, and student assessments of teacher immediacy. Observations represent data collected from watching human behaviors. For instance, interviews with Iraq veterans who are amputee, cultural participation in Yom Kippur, and 50th wedding anniversary videos may also be classified as observation-based data. Finally, texts include all written, spoken, or performed messages. Examples include stand-up comedy routines, comic strips, community graffiti, and popular music lyrics.

...