Skip to main content

Explanation Building

Encyclopedia
Edited by: , & Published: 2010
+- LessMore information
Download PDF

In case study research, an explanation—a good, successful, satisfactory, adequate, or acceptable explanation—is intended to act as an answer to a specific research question. What counts as an explanation depends on the interest of the researcher. Contrastive explanation is particularly valuable in case study. It shows why one thing rather than another (which might have been expected) occurred, or why one explanation of a given event is more acceptable than an alternative. One explanation of why a sports team does well over a season might be that it possesses better players than other teams; an alternative might be that the coach is so effective that he or she can get mediocre players to perform consistently at their best. Analysis of successful teams might lead to the conclusion that one explanation was comparatively better than the other.

Conceptual Overview and Discussion

The following examines the major considerations in building an explanation by addressing these questions: What is the relationship between explanation and understanding? In what sense is explanation objective? How do the questions asked influence the explanation? What is the nature of causality? How does a contrastive explanation provide benefit to a case study?

Explanation, Understanding, and Objectivity

To explain something means to contribute to fostering an understanding of it. Someone who can provide an explanation of something demonstrates that he or she understands it, but creating an explanation will also contribute to others' understanding.

Ludwig Wittgenstein's discussion of word games suggests that a major part of understanding something is that a person can accommodate it in his or her worldview, that complex of knowledge, beliefs, and values that is an individual's internal representation of what the world is. We avoid the apparent subjectivity of our individual worldviews because we have ideas in common with other people. Ideas shared with others possess objectivity. Thus we can share understanding through explanations that are related to commonly held ideas about the world. When such commonly held ideas are used explanatorily, they escape from subjectivity.

The Research Question

Posing a specific research question is fundamental to explanation building. An explanation does not exist in and of itself; it explains something. The formulation of the question identifies the interest of the researchers, the aspects of the case they are concerned with, and the direction of the research. It also suggests what would provide a satisfactory explanation.

A research question often concerns an event that occurred and that is not understood. Lack of understanding suggests that some other outcome was expected, so the question often implies an unstated “rather than something else.” For example, Why did the Soviet Union place strategic offensive missiles in Cuba (rather than directly negotiate with the United States for the removal of American missiles in Turkey)?

Causality

The question “why A?” or “how did situation A become situation B?” may invite a causal response. A scientific understanding of causality is “mechanical”: effect irrevocably follows cause. When we explain, for example, how lightning kills people, the outcome is a causal chain that is an inevitable result of the laws of physics. Such a deductive–nomological explanation is usually not feasible in human affairs because human actions are initiated by ideas and result from free will. As a result, the main mode of reasoning in explanatory case study is inferential and inductive rather than deductive, and causation is construed more broadly. We identify actions or ideas that have a strong causal influence on subsequent events as causes.

It makes sense to us to say that Hitler caused the Holocaust. In explaining how he did so, we would include the key Wannsee Conference in which the practical details of how to kill 11 million people were ironed out in what became an agenda for action. But this would not have occurred if Hitler had not become chancellor of Germany in 1933 and subsequently influenced others to accept his views and so create a society in which Jews were dehumanized and the unthinkable became unexceptional.

Contrastive Explanation

The contrastive explanation takes two forms. It may show that one explanation of a given situation is preferable to some other explanation. Alternatively, it may show why one state of affairs obtains rather than some other state of affairs.

As noted above, research questions often contain an implicit “rather than” clause. Because an anticipated or expected occurrence did not materialize, it is hypothetical or counterfactual. The explanation of what did occur is contrasted with the explanation of what might reasonably have been expected to occur.

Contrastive explanations are important because they can enable us to identify how the factors that influence the actual outcome differ from the factors that would result in some other outcome. John Stuart Mill's method of difference may be exploited when we identify that two different circumstances are the results of similar chains of causal influences and we infer that the difference in outcomes is largely due to the differences in the causal chains. Thus we recognize which factors had a particular influence on the outcome.

Not all explanations can be contrasted; they may be incommensurable. In such cases, if one explanation is to be preferred, its selection may depend on nonexplanatory factors such as the interest of the researcher. If alternative explanations are not commensurable, the acceptance of one may not require the rejection of the other.

Application

Ted K. Bradshaw's study of a military base closure provides a straightforward example of a contrastive explanation. Received wisdom was that the closure of a base would have major effects on the local economy. Lost jobs, falling house prices, and less money spent in local businesses are expected. This did not happen when Castle Air Force Base closed. The cleanup of the base itself created new jobs; jobs that had been held by family members of military personnel became available; retail sales grew (albeit more slowly than otherwise expected).

Received wisdom provided an explanation of what was expected, and suggested what data might be collected to support it. Because the data collected did not support received wisdom, the latter clearly provided an inadequate explanation. Bradshaw had to create an alternative contrastive explanation to show why the expected disbenefits did not occur.

Graham Allison and Philip Zelikow introduce their explanation of the Cuban missile crisis by identifying three models that were considered as the basis for an explanation of the crisis.

The rational actor model is a game-theoretical analysis that holds that the crisis was a game of chicken between Kennedy and Khrushchev in which one player's move is entirely determined by the other player's previous move(s). The organizational behavior model is that there are a number of semiautonomous groups within (and without) government, each with its own distinct interest in the situation, and that these groups produce decisions that are channeled upward and in turn influence presidential decisions. According to the government politics model, any government decision is a result of political bargaining between factions with different degrees of power, and in which interests are traded and favors given and called in. Each one of these models represents a particular academic discipline (game theory, organizational behavior, and political science) and shows how the interest relativity of the researcher(s) will favor a particular explanatory model.

In this well-known case, the authors demonstrate that any one explanation is not conclusively superior or preferable to others, and that they all inform aspects of the crisis. A more complete explanation than is provided by any one model is given by an aggregation of the three models, despite the interests of researchers (game theorists, organizational behaviorists, or political scientists) to provide complete explanations from their own fields. It is also possible that another explanation entirely, such as one focusing completely on the psychology of Kennedy and Khrushchev, may be more comprehensive.

Critical Summary

Explanation building is concerned with finding a robust explanation of why a particular state of affairs exists, often contrary to expectations. The expectations may be based on a theory/model or on observed previous similarities. This contrastive explanation requires the researcher to describe a hypothetical explanation and to show how actual causal influences differed from those projected by the model. A study that attempts to contrast explanations will not necessarily find that one explanation is preferable to another, but the contrasting explanation provides a methodology for case study research that will likely lead to the determination of an acceptable explanation.

Alan Belk
Further Readings
Allison, G.Zelikow, P.(2004).Essence of decision: Explaining the Cuban missile crisis. In R. K.Yin (Ed.), The case study anthology (
2nd ed.
, pp. 13–24).Thousand Oaks, CA: Sage.
Bradshaw, T. K.(2004).1999. Communities not fazed: Why military base closures may not be catastrophic. In R. K.Yin (Ed.), The case study anthology (pp. 233–249).Thousand Oaks, CA: Sage.
Cusimano, M. K. Essence of decision: Explaining the Cuban missile crisis (2nd ed.) by Graham Allison and Philip Zelikow [Review]. Political Psychology, (2000). 21 (3), 635–638.
Lipton, P.(2004).Inference to the best explanation (
2nd ed.
). London: Routledge.
Mill, J. S.(1970).A system of logic (
8th ed.
). London: Longman. (Original work published 1843)
Wittgenstein, L.(1968).Philosophical investigations (G. E. M.Anscombe, Trans.,
3rd ed.
). Oxford, UK: Blackwell.

Reader's Guide

  • All
  • A
  • B
  • C
  • D
  • E
  • F
  • G
  • H
  • I
  • J
  • K
  • L
  • M
  • N
  • O
  • P
  • Q
  • R
  • S
  • T
  • U
  • V
  • W
  • X
  • Y
  • Z

      Copy and paste the following HTML into your website