Understanding Kathleen Eisenhardt’s Approach to Multiple Case Studies

Qualitative research methods are diverse and often require expertise to navigate and justify. Among these, the multiple case study is a very common approach. When researchers in management started looking at which qualitative studies were being published in top journals, two main methods emerged as popular templates: the Eisenhardt method and the Gioia method. This article focuses on the approach pioneered by Kathleen Eisenhardt.

Unlike quantitative research, which often has clear “right and wrong” choices for methods, qualitative research is more flexible. Because of this flexibility, researchers using qualitative methods need to be very clear and transparent about exactly how they conducted their analysis.

What is Eisenhardt’s Approach About?

The Eisenhardt method is sometimes called “post-positivist,” or “realist”. The core idea is to focus on facts and events as people describe them. This is different from the Gioia method, which is more focused on understanding how people interpret events. In Eisenhardt’s approach, the goal is to be more objective and to see people’s interpretations of events as potential sources of bias that need to be managed. You are trying to understand the objective reality, not just subjective viewpoints.

While it’s a qualitative method, Eisenhardt’s approach often includes a quantitative focus. It seeks associations between different variables or concepts.

Key Features and Steps

1. Multiple Cases are Needed: This method requires studying multiple cases, typically between four and 10 different cases. Often, the number ends up being between eight and 10 when following this approach.
2. Starting with Concepts: Unlike some qualitative methods that start with a blank slate, the Eisenhardt approach often begins with some concepts already in mind that the researcher wants to study. For example, you might decide beforehand that you want to study “power” and “politics”.
3. Selecting Cases: Cases are selected using theoretical sampling. This means you choose cases specifically because they are likely to provide interesting insights into the concepts you are studying and, importantly, because they show variation in those concepts. If you are studying power and politics, you might choose some organizations where politics is applied a lot and others where it’s applied less. You might seek out companies where power is centralized and others where it is not. If your initial cases lack variation, you would then specifically seek out cases that provide that missing variation.
4. Collecting Data: Data is collected using multiple methods. While interviews are common, you might also use surveys (including giving informants a paper survey before an interview) or data from databases, like financial data.
5. Overlapping Data Collection and Analysis: In qualitative research, including this method, data collection and analysis happen at the same time, not one after the other. As you start analyzing early data and begin to understand processes, you can then focus later data collection on the most important parts of what you’re discovering.
6. Analyzing the Data: The analysis process is described as being very transparent. It starts by organizing and sometimes quantifying your data, looking for patterns related to the concepts you chose earlier.
   • You might code the data to find evidence of your concepts and different levels of those concepts.
   • You can develop profiles or short descriptions for each case or the key units within it.
   • Timelines are often created to map key events and understand processes.
   • The analysis always begins with within-case analysis. This means you analyze and understand each case thoroughly on its own first.
   • Then comes cross-case analysis. This involves comparing the cases, often in pairs (like a high-performing company versus a low-performing one, or two similar companies). The goal is to find patterns that explain similarities and differences between cases.
   • This process of seeking patterns, then looking for evidence of what might cause them (causality), and comparing across cases is iterated many times. You look for patterns across cases, but if a pattern only shows up in a couple of cases, it might not be the best one to build your theory on.
7. Dealing with Conflicts: People observing the same event might describe it differently based on their own interpretations. In Eisenhardt’s approach, these conflicts are highlighted. When you find conflicting accounts, you investigate why they differ and use that investigation to try and figure out what the actual reality was.
8. Patterns, Not Laws: The associations or patterns you find are not seen as absolute laws. Just because most cases showing politics also had centralized power doesn’t mean centralized power always leads to politics. The correlations or associations might not be perfect or even always strong.
9. Theory Emerges: Through this iterative process of analyzing cases, comparing them, and seeking patterns and causal explanations, a theory emerges. The goal is to develop propositions, which are statements about potential cause-and-effect relationships between two concepts. Importantly, you need to explain how, why, and when these causal processes work, not just claim that a relationship exists.
10. Comparing with Prior Research: Once the theory starts taking shape, you compare it with existing research to see where it aligns or conflicts.
11. Knowing When to Stop: You continue collecting data and adding cases until you reach theoretical saturation. This means you stop when adding a new case no longer provides significant new information that helps develop or refine your theory. If a new case doesn’t add much value, you’ve likely reached saturation.

The Goal: Generalizable Theory

Ultimately, the aim of this approach is to come up with a generalizable theory and propositions based on the analysis of the multiple cases. It’s generally used to find associations and causal relationships between concepts that were chosen before the study began, rather than developing entirely new concepts (which might be better suited to other methods like grounded theory).

The final output often includes tables showing associations between concepts, a written narrative explaining the relationships, and quotes from the data to support the findings and show the evidence for the processes described.

In summary, Eisenhardt’s approach is a rigorous qualitative method that uses multiple cases to build theory. It emphasizes looking for objective facts and events, systematically analyzing individual cases before comparing them, seeking patterns and causal explanations between predefined concepts, and iterating this process until the theory is well-developed and supported by the evidence, while also managing the potential bias from people’s interpretations.