Concept Explication


                            CONCEPT EXPLICATION

Concept explication involves the in-depth defintion of a single
idea that is the focus for your research.   Concepts are building
blocks for creating theories and models, which are used to
describe or predict relationships between two or more concepts.

This is a compilation of several handouts developed by Professor
Steven H. Chaffee at Stanford's Institute for Communication
Research, formerly a faculty member member at the University of Wisconsin.   
A fuller treatment is contained in his book Explication (Sage 
Communication Concepts 1),  Newbury Park, CA: Sage, 1992.


                         Explication At a Glance 

1.  Is it a variable?
    a. What's the unit of analysis?
    b. How does it vary?  Over time?  Across people?  etc.
    c. Is an independent, intervening or dependent variable?

2.  How is it defined in the literature?
    a.  What are the different conceptual names and meanings?
    b.  What operational definitions?
    c.  How are a and b above related?
    d.  Which of above are most promising for you?

3.  After choosing an operationalization, describe its empirical
    nature specifically
    a.  What values will it have?
    b.  What's the range of variation, and it that most useful?
    c.  What are the major correlates?
    d.  What are the antecedents?  Consequences?

4.  Alternative approaches to definitions 
    a.  Nominal definition
    b.  Meaning analysis 
    c.  Empirical definition (definition by rule)

5.  Characteristics of the concept
    a.  Is it a property or relational term?
    b.  How, specifically, is it measured?
    c.  Is it real?  (Avoid reification)
    d.  Is invariance of usage maintained?
    


CONCEPT EXPLICATION Steven H. Chaffee The key to useful research is careful definition of the major concepts in the study. In communication study we have many terms, and many measures. Explication is the procedure that provides linkages between the terms we use in discourse and the measures we use empirically. Without explication, our words are nothing more than words and our data add nothing to them. The following steps should be covered in an explication, although not necessarily in the order listed. This can be considered an outline or checklist for the student explicating a concept for research. 1. Preliminary identification of the concept. Most research starts by focusing on one term that interests the student. The first question to ask is whether this is a variable, and if not, whether variables can be derived from it. Once a variable concept has been identified, further key questions include these: a. What is the unit of analysis (e.g. persons, communities, events)? b. What is the nature of variation in relation to time? For instance, does it vary across persons at a given point in time, or across times for a given person? c. How would this variation be utilized in research? For example, would you study variation in this concept alone, or its association with variation in other concepts? Would you study its antecedent causes, it subsequent effects, or its intervening role in relationships between variables? Chaffee Comment: At this preliminary stage, about all you can decide is what you want to focus on. Your thinking about that "focal variable" should change quite a bit as you study it. Try to select a concept that is a) amenable to empirical observation, and b) likely to fit into relationships that are important for communication theory. Generally avoid adopting "canned" operational definitions from other people's research. You can make your best contribution by a fresh start that might lead to innovative studies. 2. Literature search. Once you have formalized your ideas of the concept at a preliminary level, you can begin organizing the research literature you are finding on it. Here are some useful questions to ask in structuring your analysis of the literature: a. What are the different conceptual meanings that have been assigned to this concept? What have been the research purposes of each (see 1c above)? b. What are the different operational definitions that have been used? c. What are the different names under which the concept has been studied operationally? d. What have been the operational contingencies (features of time, sampling, setting, topic, data collection, measurement) in these studies? How are these variations from study to study associated with different definitions that have been used (see 2a, 2b and 2c above)? e. In view of the prospective research purposes (1c above), which of the usages of the concept is most promising? Chaffee Comment: Once you have decided roughly what your focal variable is to be, scour the research journal, books, etc. in search of studies that have dealt with it. Your purpose is to find the various definitions that have been used. (Note that at this stage is not particularly relevant to catalog the various relationships that have been found between your concept and other variables.) Keep a list of all the ways that the concept has been defined for research purposes. You can ignore purely abstract definitions (i.e. instances where the concept is given some meaning that can't be tied to empirical observations in the real world) or cases where the term is used without definition. You might also find cases where your concept has been given some other name; keep those on your list, too. It is the empirical usage of the concept that is important, not the label that is put on it. 3. Empirical description. Review in detail the properties of the operational definitions representing the reduced number of concepts selected from the literature (2e above). a. What values does this variable take on in different populations (means, etc.)? b. What is the range of variation across units of analysis? c. What is its range of variation across time? Does it tend to increase, decrease, or merely fluctuate? d. What are the major correlates within studies? Across studies (2d above)? e. How is it distributed in various populations (e.g. normality, skewness)? f. What appear to be its immediate antecedents? Effects (if any)? Hallahan Comments: Chaffee's emphasis here is largely on quantitative research. However, many of these concerns relate to qualitative research as well. For example, if we are studying the effects of gender portrayals in TV commercials, it is important to understand, with some precision, the various ways in which consequences have been measured, the variation in those consequences, the populations involved, and all possible explanations. The specifics are important. 4. Define. Sort out the various definitions you have found. There are three levels of definition in empirical science, of progressively greater utility for research: a. Nominal definitions. Set aside those (many) cases where you have found the term used but have not been able to answer questions listed (in 1 and 2 above) because it has simply been a name arbitrarily stipulated for a without any linking statement. Instructor Comment: Some examples of nominal definitions include: "Public opinion" is what public opinion goals measures. "Consensus" consists of a majority vote. b. Real definition: meaning analysis. Two forms of meaning analysis are found in the literature. They consist of attempts to answer the following questions: i. What does this term mean conceptually? What are its essential elements, empirical research aside? (Sometimes called Theoretical Analysis) ii. What lower-order concepts does this general concept subsume? For example, "mass media use" is often defined as reading newspapers, watching TV, etc. This constitutes definition by listing as many as possible of the events that are included and specifying which ones are excluded. Chaffee Comment: Such lists expand and contract historically, but are much more helpful than undefined terms, nominal definitions, or purely abstract analyses of meaning. c. Real definition: empirical analysis. The "most scientific" type of definition consists of a set of rules that can be applied to determine whether a given event is an instance of the concept. Rules for both inclusion and exclusion are needed. The advantage of rules is that they do not change historically, being equally applicable across a wide range of circumstances. Useful questions sometimes take these forms: i. What are the necessary and sufficient conditions for inferring that an event of this class has occurred? ii. What observations are required, of what units of analysis, at what times, under what conditions? iii. What formal operations should be applied to the observations to create operational measures that correspond to the concept? Chaffee Comment: [Empirical analysis] is the most useful type of definition for scientific purposes, since changes in the lower- order concepts do not change the nature of the higher-order concept. But empirical analysis is usually the end product, rather than the starting point, of research. In a way, empirical definitions are "hypotheses" subject to modification as we learn more about the concept. In communication, this type of definition is rare. Some examples of meaning analysis: -- "Communication" requires that a symbol be transmitted by one person and received by a second person; and a signal (represented by the symbol) must be shared, at least in part, by the transmitter and the receiver" -- "Information seeking" consists of a person undertaking some action to increase his [or her] input of a specific type of communication content; that he [she] be, to some extent, uncertain what content he [she] will receive; and that his [her] action is to some extent motivated by uncertainty. Example of empirical analysis: -- Various writers have described the adoption process as a series of stages including awareness, evaluation, trial and adoption. An explicator looks at this list and decides that, whatever the heuristic value of the total model, it includes too much. All that needs to be necessary are awareness and adoption. The evaluation and trial stages may occur in some cases, but not in all, so they are surplus elements of the concept and can be dropped. -- A study shows that the strength of an expressed opinion can be increased by "reinforcing" it through social approval. The author's conclusion is that reinforcement is a necessary condition for opinion formation. A later study demonstrates that there are conditions under which opinions change without reinforcement. So the definition is waterered-down, in that reinforcement becomes a sufficient condition, rather than a necessary one. Finally, it is found that in some instances opinions shift in a direct opposite to the pattern of reinforcement. So, the element of reinforcement is eliminated from the definition of opinion formation, because it is neither necessary nor sufficient. 5. Review definition With a real definition (4b or 4c above), you can apply various criteria to it. Some important questions: a. Property or relation? Is this a concept that is observable for a single unit of analysis, in isolation from other units (property term)? Or is it only observable in the context of other units, such as a comparison between communities or an interaction between people (relational term)? (Caution: Almost all concepts in communication research are relational; unfortunately most data collection and much theory and analysis proceed as if they were properties.) Chaffee example: Income is a property term, but socioeconomic status is a relational term. So if you are interested in SES, but have data only on income, you should be treating that data as relational, e.g. as "relative income." b. Specificity. How explicitly is the definition spelled out, at each level from higher-order concepts to lower-order constituent concepts, to operational procedures? The more specific, the better. When terms are used vaguely, there is more danger that operational definitions are inappropriate to the concept intended, and consequently that the research will be inconsistent with the theory behind it. Hallahan example: In examining the effects of "television viewing," it makes a difference whether the viewer saw "Beavis and Butthead" or "60 Minutes." c. Non-reification. Does the term refer to real events? Do the operational measures represent real events? Our capacities to gather data, and to talk about communication, greatly exceed the incidence of communication events in the real world of human experience. Such terms are called opaque or reifications. ... Chaffee comment: Insofar as possible, avoid giving names to attributes that you might image exist, but that cannot be observed. You may suspect that there is a key factor that has not been observed, but that could be given empirical meaning by careful research. In that case you are proposing a hypothetical construct, the hypothesis being that it exists; the first task of your research should then be a "validity check" on its existence. When you provide operational evidence of a hypothetical construct, it attains the more secure status of an intervening variable. If a hypothetical construct remains unobserved, it is considered a reification, and other researchers are unlikely to be persuaded by your references to it. -- Some common reifications in communications research are the terms "catharsis," "dissonance," "group cohesiveness," "coorientation" and "attitude." So far none of these things has ever been observed, yet they hold important positions in certain theoretical formulations. The danger is that they may not exist, except in the minds of the theoreticians. -- Some hypothetical constructs that have gradually been converted into intervening variables by careful research include "empathy," "understanding," "learning," and "conformity." However, these concepts are tied to very specific operational definitions, and when they are used to cover other kinds of situations they are simply reified terms. d. Invariance of usage. The general goal of explication is so that one, and only one, concept will be attached to an empirical meaning. Several questions need to be asked: i. Do does this researcher use a single definition for the concept and the same name for that definition consistently (intra-observer invariance) ii. Do different researchers use the same combination of concepts and meanings? (inter-observer invariance). Are these also consistent over time? This happy state of consistent usage cannot be legislated, but it can be facilitated by careful explication. Chaffee example: The term "generation" is a term used appropriately for analyzing families and other kinship systems. However it is misapplied to differences between age-groups in society as a whole in the notion of a "generation gap." 6. Modify definition. Your review of the definition at stage 5 (above) may well lead you to redefine the concept. Recycle this definition through all of the steps above (1...5). Then: With a adequately explicated research concept, you are prepared to move on to a research design. The key steps include: Explication of additional concepts in the study Operationalization (creating methods for observing or measuring the concepts) Choosing a sample to study Testing the concept, and observational and analysis methods Refining the concepts and operationalizations Conducting the research Analyzing the results Validation, refinement of the concept Writing and reporting the results