This chapter focuses on quantitative methods for the study of conflict communication for several reasons.
First, qualitative methods, such as participant observation and unstructured interviews, are often used to determine the following: the universe of meaning (i.e., the full range of connotations in addition to the denotations) of a phenomenon, the types of individuals relevant to a domain of investigation, the operative categories employed by cultural participants (i.e., the emic constructs; see Pike, 1967), and the causal attributions provided by individual participants or observers (for discussion of these methods, see Denzin & Lincoln, 2000).
For this chapter, we view the goal of conflict research as the creation of general and abstract social science theory, which entails prediction as well as description. To this end, qualitative investigations, however valuable, serve as precursors to studies that explicitly take into account measurement validity, internal validity, and external validity, or, in short, quantitative investigations. Furthermore, although there exist naive theories that can be well explicated by qualitative methods (see Heider, 1958), quantitative investigations are needed to check the naive causal attributions made by informants (see, e.g., Nisbett & Wilson, 1977). Finally, the literature on research methods germane to conflict communication research is so extensive that even restricting our focus to quantitative methods leaves us with little opportunity to cover them all (see Jiang & Buzzanell, this volume, for qualitative methods and conflict).Second, we chose to highlight some methods that are especially relevant to cross-cultural and other group comparisons. Like many areas of investigation, the study of conflict communication began with obvious and important applied questions: How do we avoid war? Can we minimize strife within families? Can labor and management cooperate in the modern corporation? These questions lend themselves to group comparisons: warring versus nonwarring countries, conflictual versus nonconflictual families, and companies with cooperative labormanagement relations versus those with records of labor-management clashes.
In this chapter, we examine how quantitative methods are used to address such comparisons.Some of the material included here is reprinted, with some revision, from the chapter presented in the first edition of this Handbook (Fink, Cai, & Wang, 2006). However, because the previous and current chapters are somewhat different in their emphases, we recommend that both chapters be consulted.
Areas of Quantitative Methods
Quantitative research methods can be considered to consist of five areas:
1. Issues of measurement and observation, which include the assumptions that different measurement systems provide, emic versus etic considerations in measurement (see, e.g., Pike, 1967), measurement reliability, and measurement validity
2. Research design, which includes the distinctions between experimental and nonex- perimental designs (see Cook & Campbell, 1979) and between cross-sectional and longitudinal designs (including panel and time-series designs), as well as issues of internal and external validity
3. Sampling, for purposes of exploration, description, and conceptual explication, or for purposes of inference and the enhancement of external validity; this distinction logically leads to the differentiation between probability and nonprobability sampling
4. Data analysis, which includes distributional assumptions and data transformations, descriptive versus inferential statistics, parametric versus nonparametric tests, univariate versus multivariate data structures, and exploratory versus confirmatory investigation
5. Ethics, which consists of two related ideas:
Internal ethics: Issues designed to protect participants from (a) threats to respecting them as persons, (b) harm in the course of their participation, or (c) injustice in the relation of the benefits versus the burdens that participants bear (see the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1978, also known as the Belmont Report)
External ethics: Who sponsors the research, and toward what end? Will individuals, or groups to which participants belong, be harmed not by the investigators but by the use to which the investigators’ research is put?
Our goal here is to indicate important methodological areas for the scholar studying or engaging in research on conflict communication and to explicate a few of the critical intersections of quantitative methods and the study of conflict communication.
Ethics, although crucially important to research on conflict communication, will not be discussed in depth here, but it is suggested that the reader see the following valuable resources on this subject: Kelman (1982, 2001), Sjoberg (1967), and Lewis (1975). For data analysis, we will restrict our discussion to explaining statistical dependence of data, moderation, and mediation. This chapter ends with an excursus on culture, which is important to current trends in cross-cultural research on conflict communication.