Using Samples to Compare Conflict Communication Processes
Once conflict situations are determined to be comparable across cultures in terms of the severity of the conflict, the relationship between the parties, the level of emotion involved, or other such considerations, researchers can begin to determine whether the processes involved in managing the conflict are comparable.
However, relatively equal variable means do not suggest that the processes that generated these means are the same, or that the means reflect the same point in the process, or that the process has equilibrated. Answering these questions involves determining the functional form of relationships between variables. The same functional form means employing the same equation (statistical or mathematical), with the same variables, with (for statistical models) the error term entering into the equation in the same way (e.g., additively vs. multiplicatively), and, finally, with the same estimate of parameter values.Differences in functional form may be approached as a theoretical question, a measurement question, or an analysis question. (This issue is dealt with in a detailed way in Fink et al., 2006.) As a theoretical question, functional form concerns whether the variables used in the analyses are sufficiently general and sufficiently interdependent. As a measurement question, functional form concerns how variables are measured. Zetterberg (1965) showed how conceptual and operational definitions of variables may overlap to different degrees; the metric for a variable— the numbers that correspond to the level of a continuous variable—is arbitrary in communication research (see Torgerson’s, 1958, notion of measurement by fiat). This arbitrariness in how variables are measured may result in different researchers reporting different functional forms relating the same theoretical variables. For example, conflict styles are measured with different instruments (Wilson, Cai, Campbell, Donohue, & Drake, 1995). To deal with this issue, either all researchers need to use the same measures (unlikely in the social sciences), or researchers need to provide the rules that relate (translate) their measures to standard ones. In this way, measures may be calibrated against the standard and, thus, to each other. If such measurement procedures are adopted, debates over differences in functional form become debates over differences in measurement rules.
As an analysis question, functional form addresses whether the relationships between the variables are analyzed so that we can determine that the processes are the same across groups. This discussion goes beyond the scope of this chapter. For an in-depth discussion of this issue, see Fink (2009).