Generalization Is Everything, or Is It?

Abstract

1. IntroductionCase studies are a critical and even attractive research methodology 1 (Bartunek, Rynes & Ireland, 2006; Eisenhardt & Graebner, 2007; Sato, 2009). 2 However, there are sometimes questions about whether the case study method is rigorous enough as a research methodology.In addressing this question, Eisenhardt (1989) is one of the most frequently cited studies on the methodological case study fundamentals in theory construction. Yet most studies that cite Eisenhardt's method emphasize only generalizability. As a result, the potential for theory construction using the case study method has not been sufficiently investigated.2. Case Study Method in Eisenhardt (1989)According to Eisenhardt (1989), the case study is a research strategy which focuses on understanding the dynamics present within single settings. Eisenhardt (1989) focuses on the case study as a research methodology for theory construction in particular. According to Eisenhardt, theory construction consists of nine steps.a) Getting startedWhen conducting a case study, the first thing to do is to formulate the research question. This enables one to specify what kind of organization the survey should study or what data the survey should collect.b) Selecting casesThe next problem is case selection. In a case study that has the goal of theory construction, sample selection usually depends on theoretical sampling.In statistical sampling, the sample is selected randomly from a population. Theoretical sampling on the other hand reiterates previous cases and uses what amounts to an extreme example that reproduces an existing case, builds upon the theory, or corresponds to a theoretical category.c) Crafting instruments and protocolsResearch on theory construction is able to combine multiple data sources. In particular, it combines quantitative and qualitative data. The case study can therefore use both types of data.d) Entering the fieldsAn especially notable characteristic of the case study in theory construction is the overlap of data collection and data analysis.Overlap of data collection and data analysis leads to flexibility in data collection. It enables the researcher to change the data collection methodology in the course of the research when the researchers comes up with a new idea that may generate a new theory.e) Analyzing within-case dataFirst, independent analysis of various cases enables the researcher to avoid becoming overwhelmed by too much data.In addition, by analyzing individual cases before comparing the various cases, the researcher is able to detect the particular patterns of the various cases, thereby deepening his or her understanding of the cases in point and facilitating cross-case comparisons.f) Searching for cross-case patternsBias can occur in cross-case comparisons, which could lead to erroneous conclusions. There are three ways to prevent bias.The first is to separate the cases into categories and look at the similarities within each group and the differences between the groups. The second is to divide the cases into pairs for comparison, then look for differences between the cases that are similar and for similarities between the cases that appear to be different. The third is to divide the data according to the data source and try to gain insight into the particulars of the heterogeneous data.g) Shaping hypothesisThe first step in hypothesis development is clarification of constructs. This step consists of two processes: elaborating on the constructs' definitions and establishing proofs for measuring these constructs. These processes constitute a hypothesis-testing type of research, in that they are similar to formulating a construct from several metrics. However, a case study for theory construction differs in that it is the analysis itself that generates the constructs and their measurement, and that there is no way to group several metrics into one construct. …