Analytic models in strategy, organizations, and management research: A guide for consumers

• Author: Douglas P. Hannah,Ron Tidhar,Kathleen M. Eisenhardt
• Date: 01 February 2021
• Published date: 01 February 2021
• DOI: http://doi.org/10.1002/smj.3223

RESEARCH ARTICLE

Analytic models in strategy, organizations, and

management research: A guide for consumers

Douglas P. Hannah

1

| Ron Tidhar

2

| Kathleen M. Eisenhardt

2

1

Questrom School of Business, Boston University, Boston, Massachusetts

2

Department of Management Science and Engineering, Stanford University, Stanford, California

Correspondence

Douglas P. Hannah, Questrom School of

Business, Boston University, Boston, MA.

Email: dphannah@bu.edu

Abstract

Research summary: Analytic models are a powerful

approach for developing theory, yet are often poorly

understood in the strategy and organizations commu-

nity. Our goal is to enhance the influence of the method

by clarifying for consumers of modeling research how to

understand and appreciate analytic modeling and use

modeling results to enhance their own research. Our pri-

mary contribution is a guide for reading analytic models.

Using comparisons with other methods and exemplar

analytic models, we explore key features as well as coun-

terintuitive aspects and common misconceptions. We

also add by illuminating strengths and weaknesses of

analytic modeling relative to other theory-building

methods. Finally, we identify under-exploited opportuni-

ties for pairing analytic models with complementary

methods. Overall, our aim is enhancing the influence of

analytic modeling by better-informing consumers.

Managerial summary: In this paper, we explore the

use of analytic (mathematical) models for developing

strategy and organizations theory. Analytic modeling is

common in related fields like economics but is often

poorly understood among the broader of strategy and

organizations community. Whereas existing resources

on analytic modeling are geared towards modelers, our

aim is to enhance understanding and appreciation of

the method among potential consumers of modeling

Received: 28 August 2018 Revised: 23 July 2020 Accepted: 24 July 2020 Published on: 2 September 2020

DOI: 10.1002/smj.3223

Strat Mgmt J. 2021;42:329–360. wileyonlinelibrary.com/journal/smj © 2020 John Wiley & Sons Ltd 329

research. We offer three specific contributions in this

regard. The first is a guide for reading analytic models,

including key features, counterintuitive aspects, and

common misconceptions. Second, we clarify the

strengths and weaknesses of analytic modeling relative

to other theory-building methods. Finally, we discuss

promising opportunities for pairing methods.

KEYWORDS

formal theory, game theory, modeling, research methods, theory

development

1|INTRODUCTION

As social scientists, our ability to explain and predict depends on the quality of our theories

(Sutton & Staw, 1995; Weick, 1989). One of the most powerful tools in our pursuit of better the-

ory is analytic modeling. An analytic model is an abstract rendering of a more complex reality

into a set of mathematical equations based on concepts, relationships, and assumptions (Adner,

Polos, Ryall, & Sorenson, 2009). As a means for developing theory, analytic models are common

in finance, economics, political science, and to a lesser extent, sociology (Swedberg, 1990;

Debreu, 1991; O'Rand, 1992; Stokes, 2005). In contrast, however, their use in the strategy and

organizations literature remains more limited. While some scholars use the method to great

effect (e.g., Alacer, Dezso, & Zhao, 2015; Casadesus-Masanell & Yoffie, 2007; Jia, 2013; Kaul &

Luo, 2018; Panico, 2017; Sakhartov & Folta, 2014), the method is nonetheless poorly understood

by many in the broader community, especially relative to its potential and in contrast to related

fields.

A number of indicators reflect the state of analytic modeling in the strategy and organiza-

tions literature. One is low use. For example, we reviewed the 6,966 articles published in six

premier strategy and organizations journals from 2005 to 2019 and found that analytic model-

ing was the primary research method in just 4% (see Appendix). A second indicator is low cita-

tion rate. In particular, our analysis reveals that papers that rely on analytic models receive

fewer citations on average (20.6 per paper) than either econometric analyses (34.5) or qualita-

tive research (37.2)—and far fewer than verbal theory and reviews (68.4).

1

A third and perhaps

most critical indicator is the perception of analytic modeling in the broad scholarly community.

To understand these perceptions, we interviewed 35 strategy scholars across a range of disci-

plines and surveyed 34 strategy PhD students at seven schools across the United States and

Europe. Scholars without training in modeling research (the majority) reported that they rarely

read analytic modeling research, regard analytic models as harder to interpret than other

methods, and believe analytic models are rarely relevant to their own work.

1

This pattern holds even when controlling for author affiliation, journal, and year of publication (see Appendix).

Moreover, it offers a striking comparison to economics, where as Knudsen, Levinthal, and Puranam (2019) note, formal

methods have been a dominant research method since Samuelson (1947).

330 HANNAH ET AL.

These indicators reflect a significant, but we believe addressable, challenge: a gap between

the producers of analytic modeling research and potential consumers of that work. This gap is

exacerbated by diverse doctoral training (i.e., many scholars have little exposure) and the rela-

tively small number of modelers in strategy and related fields. Moreover, potential consumers

lack resources to improve their understanding of the method. On the one hand, economics texts

(Bolton & Dewatripont, 2004; Gibbons & Roberts, 2013; Tirole, 1988) and modeling texts

(Kemeny & Snell, 1962; Mershon & Shvetsova, 2019) discuss the mathematical apparatus of

analytic models. Yet, these are often inaccessible or at least little read by general audiences, and

assume familiarity with tools and concepts like game theory, equilibrium reasoning, and linear

programming. On the other hand, appeals for more modeling (e.g., Adner et al., 2009;

Ghemawat & Cassiman, 2007) indicate the strengths of the method, but do not (and are not

intended to) provide readers with tools to understand and use modeling research. Thus, a gap

exists for those potential consumers who want to better understand analytic models and use

published modeling results to inform and enhance their own research.

2

We aim to enhance the usefulness of analytic models for potential consumers of analytic

modeling research, and so address this gap. We do so by clarifying how to understand and

appreciate analytic models, as well as how to interpret their results. In developing our insights,

we draw on an extensive review of well-cited strategy and organization exemplars and inter-

views with 17 experienced modelers. We also rely on prior work on theory development

(e.g., Knudsen et al., 2019; Makadok, Burton, & Barney, 2018; Montgomery, Wernerfelt, &

Balakrishnan, 1989; Pfeffer, 1993) as well as our own work on theory development using other

methods (e.g., Davis, Eisenhardt, & Bingham, 2007; Eisenhardt, Graebner, & Sonenshein, 2016;

Tidhar & Eisenhardt, 2020). Overall, we rely on a broad set of sources to fill the gap for con-

sumers between technical modeling texts and appeals to the model.

We offer several contributions. Our primary contribution is a guide for reading, understand-

ing, and appreciating analytic models. Since our focus is the consumer perspective, we articu-

late their essential features from this lens. Like any good travel guide, we provide the highlights

for how (and why) to read and appreciate analytic models, but not every detail of how to model.

For example, we indicate common modeling approaches, and clarify their strengths, weak-

nesses, and types of insights that consumers might expect. Given our aim, we also emphasize

the counterintuitive aspects of analytic models and common misconceptions such as the role of

strong assumptions (like rationality) and omitted variables, which we argue are a strength of the

method rather than a weakness. We also highlight the relevance of what we term the “concep-

tual narrative,”and the specific research questions that analytic models are often used to

address. By emphasizing similarities, differences, and complementarities with familiar methods,

we use comparison to further illuminate analytic modeling. By using interviews with experi-

enced modelers, we capture some of the art of analytic modeling as well.

A second contribution is positioning analytic modeling within the broad repertoire of theory

development methods in strategy, organizations, and management research more broadly—that

is, what the method is and when it is useful. In particular, we clarify the value that analytic

models bring to theory building, including their unique strengths (precision, internal consis-

tency, and transparency) and weaknesses (external validity and role of the modeler). In doing

so, we sharpen what consumers can (and cannot) expect to learn from analytic models.

2

One recent effort to fill this gap is Csaszar (2020), which although geared to potential producers of modeling work

contains valuable insights for consumers of that work as well. Knudsen et al. (2019) is similarly a parallel effort.

HANNAH ET AL.331