The Eisenhardt Method is an approach to building new theories by studying multiple real-world cases. Unlike methods that rely solely on abstract theorizing, it emphasizes gathering detailed, direct insights from carefully chosen examples.
Let’s break down each step clearly and thoroughly:
Step 1: Core Aim: Theory Building
What it means:
- Your primary goal is to create a theory, not just tell a good story or describe facts.
- Theory here means clearly defined concepts (constructs) linked together by clearly explained relationships, backed by arguments (mechanisms) explaining why these relationships hold.
How you achieve it:
- Start with broad curiosity, but quickly identify key concepts.
- Constantly articulate emerging concepts clearly. Draw diagrams showing links between these concepts early and often.
- Aim to explain why and how things happen—not just that they do.
Step 2: Research Questions Without Obvious Answers
What it means:
- Pick research questions that haven’t been answered yet, or where existing answers are conflicting or unclear.
- Look for topics with unclear processes (“black box” problems), contradictions in literature, or novel phenomena not previously studied.
Why this matters:
- These questions provide opportunities to develop truly novel theories. They’re likely to yield insights other methods can’t easily provide.
How you identify such questions:
- Questions should prompt curiosity and be challenging to resolve using existing literature.
- Example: “How do startup teams make strategic pivots under extreme market uncertainty?” rather than “Do pivots lead to success?” (already studied extensively).
Step 3: Careful Case Selection (Theoretical Sampling)
What it means:
- You don’t select cases randomly; you carefully choose cases that offer the greatest insight into your research question.
- You pick cases strategically to see differences or similarities that help your theory emerge clearly.
How you select cases:
- Choose settings where the phenomenon clearly occurs.
- Aim to enhance focus, reduce alternative explanations, and improve generalizability.
Strategic types of case selections:
- Common antecedents:
Pick cases similar at the beginning, see why they diverge later.
(Example: Similar startups, different outcomes) - Matched pairs:
Pick closely similar cases, then analyze subtle differences that caused varied outcomes. - Racing:
Pick cases starting simultaneously under similar conditions, observe differences over time, useful for dynamic studies.
(Example: Several startups launched at the same incubator at the same time.) - Polar types:
Cases at extremes (very successful vs unsuccessful), helps sharply highlight contrasts. - Common process, varied settings:
Same phenomenon but in very different contexts to increase generalizability. - Serendipitous opportunities:
Remain flexible; unexpected findings might offer valuable theoretical insights.
Practical considerations:
- No perfect formula; judgment is critical.
- Keep flexibility to add or remove cases if insights emerge differently than expected.
Step 4: Explicitly Developing and Defining Constructs and Measures
What it means:
- Clearly define the concepts (constructs) you’re studying.
- Develop measures—specific indicators—that make constructs observable, grounded, and testable.
How you do this practically:
- Regularly review your data (interviews, observations) to group related findings.
- Move iteratively between raw data and more abstract ideas (constant comparison).
- Regularly refine and clearly define constructs to ensure they accurately reflect your data.
- Use clear labels (e.g., “decision-making anxiety” or “innovation urgency”) and give precise definitions.
Why it’s essential:
- Clearly defined constructs make your theory reliable, understandable, and replicable.
Step 5: Emphasizing Explicit Theoretical Arguments (Mechanisms)
What it means:
- Explicitly state the logic (mechanism) that explains why relationships between constructs occur.
How you achieve this:
- Continuously ask “why?” for observed relationships.
- Use logic from:
- Data itself: “Participants explained they pivoted due to financial pressure.”
- Logical reasoning: “Organizations experiencing financial pressure might take bigger risks.”
- Existing theories from other disciplines: psychology, economics, biology, etc., which offer deeper, surprising insights.
Importance:
- Clearly explained mechanisms make your theory strong, convincing, and internally consistent.
Step 6: Identifying Boundary Conditions and Alternative Explanations
Boundary Conditions:
- Define clearly where your theory applies and where it might not.
(Example: Theory of rapid pivots might apply only to high-uncertainty industries.)
Alternative Explanations:
- Actively consider and test explanations different from your own theory.
- Clearly argue why your theory explains phenomena better than others.
Why this matters:
- Enhances the credibility and accuracy of your theory.
- Clarifies precisely when and where your theory works.
Step 7: Analysis using Constant Comparison, Replication Logic, and Cross-Case Analysis
What it means practically:
- Constant Comparison:
Continuously compare your emerging theory against fresh data within each case and between cases. Update theory accordingly. - Replication Logic:
Each case is treated like an independent experiment. See if your emergent theory fits across cases independently, rather than aggregating data prematurely. - Cross-Case Analysis:
Examine patterns and variations systematically across cases. Helps confirm or adjust your theoretical conclusions.
Techniques:
- Visual displays (matrices, timelines) to clarify patterns.
- Tables to systematically compare cases.
- Writing individual detailed mini-cases first, then comparing carefully to build general theory.
Why this matters:
- Increases theoretical robustness, accuracy, and creative insights.
The Ideal Outcome of Eisenhardt Method:
- Parsimonious (simple yet powerful): Easy to grasp.
- Accurate: Closely fits reality (grounded strongly in evidence).
- Generalizable: Useful beyond studied cases.
- Logically coherent: Internally consistent and convincing.
- Surprising: Often reveals unexpected insights.
Common Misunderstandings:
- It does not require a specific number of cases (commonly 4–10, but flexibility is key).
- It is not limited to qualitative data only; quantitative or mixed methods are valid too.
- It is not restricted to performance-oriented topics; can explore processes, relationships, or other phenomena.
- It can create both process and variance theories.
If Eisenhardt Were to Update the Method Today, She’d:
- Encourage even more creative research designs and choices.
- Emphasize clearly how phenomena evolve over time (temporality).
- Allow for a broader variety of scientific philosophies (beyond strict positivism).
However, the core of the method—close, rigorous theory building—remains the same.
Bottom Line:
The Eisenhardt Method is practical, systematic, and adaptable, designed explicitly to produce clear, insightful theories closely grounded in rich, empirical evidence.
Leave a Reply