ABSTRACT
Up to this point, we have discussed how to quantify and test for association between variables, determing whether a relationship exists between two variables after accounting for their shared relationship with another variable or set of variables, and how to determine whether “chance” is the most parsimonious explanation for differences between groups or a relationship that we observe between two variables. But there are questions that need to be answered before we can say we really have gained some understanding of communication processes and theory from our research: “Why?” “When?” and “How?” A scientist can make a career of demonstrating that two variables are related, but the more memorable studies, the more impressive studies, and ultimately the more influential studies in the field go further by discovering or explaining why such relationships exist, under what circumstances, or for whom the relationship exists strongly as opposed to weakly or not at all. We truly understand some phenomenon if we are able to determine when the phenomenon will occur, why or how it occurs, and for whom it occurs or will occur. As you become increasingly knowledgeable about the discipline of communication and increasingly expert in your specialty area, you will discover that the most sensible answer to almost every question you will confront as a scientist is “it depends.” Such an answer is not a cop-out. What we study is often sufficiently complicated that it would be incorrect to say without condition or exception that X causes Y or that one group differs from another group on some outcome variable in all circumstances. Usually effects vary as a function of something else. For example, perhaps for some people exposure to televised violence causes aggressive behavior, but for others such exposure has no effect. Or perhaps the effect of such exposure differs depending on the consequences or form of the violence. Media violence that is perceived to be rewarded may lead people to engage in that behavior, whereas violence that is perceived to be punished may discourage such behavior. Or perhaps a message about the negative consequences of unsafe sexual practices could increase safe sexual practices among people of a certain background or age but decrease it or have no effect among people of a different background or age.
