INDEX
1. Introduction
2. Early Foundations of Quantitative Operations Management
3. The Separation Between Theory and Practice
4. Types of Quantitative Model-Based Research
5. The Four-Phase Model of Research
6. Conducting Axiomatic Quantitative Research
7. Conducting Empirical Quantitative Research
8. Issues of Rigor
9. Issues of Relevance
10.Balancing Rigor and Relevance
11.Implications for Students
12.Implications for Managers
13.Future Directions for Operations Management Research
14.Conclusion
15.References
Quantitative approaches to the operations management
discipline
1. INTRODUCTION
Operations Management is the study of processes in firms. The processes involved
include the following: production, planning, inventory, flow, scheduling, purchasing, and
capacity planning. Managers utilize these different processes to deliver products and
services. Every decision has its constraints, pressures, and tradeoffs. To make sense of
these decisions, structure is necessary. Quantitative modeling supports structure.
Quantitative models incorporate variables and relationships that can be measured. The
purpose is to describe and forecast the behavior of a process. A process may consist of
machines, workers, materials, orders, customers, etc., all interacting among themselves.
A model is useful to the researcher in highlighting patterns and to managers in testing an
idea before action is taken.
This paper reviews the history, intent, and approaches of quantitative research in
Operations Management. Ideas are presented in a structure introduced by Bertrand and
Fransoo. The paper describes how early research grew and where it struggled, how it
evolved, and explains different types of research and how each supports learning. It also
addresses what makes research rigorous and relevant. The writing should avoid a heavy
academic tone. The goal is clarity.
Operations Management is a challenge to students due to the mixing of math and real
systems. Real systems have variation and human choices; math models want clean
assumptions. What this paper is going to do is try to bridge that gap by explaining
concepts in simple terms.
2. EARLY FOUNDATIONS OF QUANTITATIVE OPERATIONS MANAGEMENT
Scientific Management forms the early history. Frederick Taylor studied work in factories.
He measured each task in detail. He sought out the best method for every activity. He
advocated for standardized procedures. His concentration was on time, work division,
and worker output. He aimed for higher productivity.
Taylor's work quickly permeated into many industries. Managers widely implemented
time studies, instructions, and wage systems based upon output. Productivity gains
resulted in many firms. However, there are some significant limitations to Scientific
Management. It did not attempt to develop a general scientific theory. It looked only at
task-level improvements, not process-level behavior. It did not incorporate any
consideration of variation in systems. It did not explain how delays propagate or queues
form.
Then came the idealized problems introduced by universities. These were simple
enough to teach. Students learn concepts like reorder points, sequencing rules, and
routing decisions. Models like these helped students comprehend the logic behind the
decisions. For instance, the economic order quantity model represented the interaction
between ordering cost and holding cost. Scheduling rules showed the effects caused by
processing order on flow time.
These models did not take into account realistic factory conditions: breakdowns,
multitasking, worker learning, supplier variation, training needs, and constraints. They
could provide some insight but no real predictive power.
Simultaneously, the field of Operations Research was also developing. The researchers
used mathematics to study a number of problems related to optimization, queuing, and
statistics. Their work built solid tools. These tools formed the foundation of many
concepts related to Operations Management.
Even with strong mathematics, much of the research failed to intersect with practice. The
assumptions were often too simplistic. Real constraints were excluded from the models.
Managers ignored much of this work.
British researchers during wartime worked directly with units. They also solved problems
inside the factories, supply lines, and operations teams. They included human behavior
and context in their work. Their work produced practical results, not general theories.
The results stayed specific to each case.
The following three important developments gave shape to modern Operations
Management:
First came Forrester with his work on Industrial Dynamics. He demonstrated how
decisions cause feedback and delays. He researched what happens to future
performance when certain actions are taken today. Using this knowledge, he developed
a methodology that modeled dynamic behaviour over time. It helped explain the
inventory swings, backlogs, and instability.
The second was queuing theory. Queuing theory described waiting lines and delays. It
showed how arrival rates and service rates influence system behavior. It showed how
variation affects waiting time. It became important to manufacturing, services, health
care, and logistics. These developments forged a deeper basis for the field.
3. THE SEPARATION BETWEEN THEORY AND PRACTICE
It is between the 1960s and the 1980s that Operations Research failed to have much influence
on industry. The field became enamored with math. Many papers produced correct formulas.
But the results solved abstract problems. Managers needed solutions for real constraints. They
needed models that worked when data changed suddenly. They needed systems that handled
variation.
Management perceived academic models as being unrealistic. They felt that the models were
not relevant to their real life problems. The general feeling was that academics lived in a world
of their own. Ackoff, in particular argued very strongly the discipline was solving the wrong
problems.
During this time, the industry moved ahead with absolutely no help from the universities. The
first significant example was Material Requirements Planning. MRP permitted firms to plan their
production on a tight schedule. It integrated demand, supplier lead times, and inventory. This
was a major leap forward in synchronization. MRP was the rage not only in manufacturing but
also elsewhere in the economy. Many of the researchers dismissed it because it did not fit in
with their paradigm.
The other major development was the Toyota Production System. Toyota focused on the flow of
goods, reliability of operations, and gradual improvement. They targeted problem-solving,
machine reliability, and worker involvement. They used pull systems, quality at source, and
small batch sizes. These concepts later caught on as lean techniques. These practices went
global and raised the output of many factories
The third development was workload management. Several factories realised that the control of
work-release would raise their production rates. They applied things like CONWIP to even out
the flow in the system. These systems kept the area free of excess work in progress—no
bottlenecks and no waiting. Conventional queuing theory did not account for this phenomenon,
thus leading to the need for a study that would explain these systems' working principles.
The situation was typical of that pattern. The theory could not keep pace with the practice. The
actual processes could be explored only with the help of more modern approaches by the
researchers. They had to combine modeling with real data.
4. TYPES OF QUANTITATIVE MODEL-BASED RESEARCH
Quantitative Operations Management research may be classified under four major types. These
types are a function of two dimensions. One of the dimensions is axiomatic versus empirical.
The other is descriptive versus normative.
The axiomatic approach is based upon logical steps and presuppositions. The researcher is
going to create a model of the situation. The model will be of either a mathematical or a
simulative nature. The model is not derived from data. It is developed from the underlying
structure.
Axiomatic Descriptive literature describes how a system works. As an example, queueing theory
predicts that waiting time increases with utilization. With the help of the mathematical model, the
researcher retrieves system properties.
Axiomatic Normative research gives advice on the choice. For instance, a certain policy of
ordering may lead to minimum cost. A rule of scheduling may lead to less flow time. The
objective of the model is to reveal the optimal choice.
Empirical research is based on observation. The researcher conducts a study of a live system.
The researcher gathers data, does measurement of variables, and brings out patterns.
Empirical Descriptive research provides insight into actual behavior. For example, the
researcher might note that delays in decision-making in the plant cause fluctuations in inventory
levels. This pattern is explained by the researcher drawing on data. Empirical Normative
research judges the decisions in actual environments. An example is a factory trying out a new
scheduling system. The researcher looks into the outcome. If the system does not cause any
additional waiting time, the efficiency gain is documented.
All the types of research contribute value in their own ways. Axiomatic research gives the
benefit of being clearer. It yields knowledge that is valid in general. Empirical research makes it
more lively. It validates the assumptions and reveals the voids.
Operations Management research that is quantitative in nature can be categorized into four
major types. These types are determined by two dimensions. One of the dimensions is
axiomatic vs. empirical. The other is descriptive vs. normative.
The axiomatic approach is based on logical steps and presuppositions. The researcher is going
to create a model of the situation. The model will be of a mathematical or a simulative nature.
The model is not derived from data. It is developed from the underlying structure.
Axiomatic Descriptive literature explains the operation of a system. For instance, the queueing
theory predicts an increase in waiting time with the increase in utilization. Making use of the
mathematical model, the researcher obtains system properties.
Axiomatic Normative research gives advice on the choice. For instance, a certain policy of
ordering may lead to minimum cost. A rule of scheduling may lead to less flow time. The
objective of the model is to reveal the optimal choice.
Empirical research is based on observation. The researcher conducts a study of a live system.
The researcher gathers data, does measurement of variables, and brings out patterns.
Empirical Descriptive research provides insight into actual behavior. For example, the
researcher might note that delays in decision-making in the plant cause fluctuations in inventory
levels. This pattern is explained by the researcher drawing on data. Empirical Normative
research judges the decisions in actual environments. An example is a factory trying out a new
scheduling system. The researcher looks into the outcome. If the system does not cause any
additional waiting time, the efficiency gain is documented.
All the types of research contribute value in their own ways. Axiomatic research gives the
benefit of being clearer. It yields knowledge that is valid in general. Empirical research makes it
more lively. It validates the assumptions and reveals the voids.
5. THE FOUR PHASE MODEL OF RESEARCH
Mitroff describes research as consisting of four phases: conceptualization, modeling, solving
and implementation.
The first phase is conceptualization, in which one defines the problem. The researcher outlines
the boundaries. The researcher identifies the variables. This step sets direction. Weak
conceptualization creates confusion later on. Strong conceptualization is clear and simple.
The second phase, modeling, involves structure. The researcher forms the vaguely conceptual
into a scientific model. This can include variables, relations, and assumptions. For example,
set-up time, processing time would be represented in a scheduling model.
The solution phase employs the tools to compute results. These tools range from mathematics
through algorithms, simulation, and the numerical methods. The output can be formulae,
numbers, or measures of performance.
Implementation is the final phase, which concerns practice. This stage determines whether the
results help managers. Many academic models never reach this stage because their
assumptions fail in real settings.
The confusion between solution and implementation easily arises. Several researchers often
think that a solved model represents a practical solution. That is not the case. A solution matters
only when it fits real constraints.
Empirical researchers spend more time on conceptualization and measurement. Axiomatic
researchers spend more time on modeling and solving. Strong reasoning is needed by both.
6. CONDUCTING AXIOMATIC QUANTITATIVE RESEARCH
Strong axiomatic research moves along a disciplined path.
The first step has been to describe the real problem: in standard Operations Management
terms, this ensures clarity. For example, any lot sizing problem must define demand, lead time,
holding cost, ordering cost, and capacity limits.
The second step is to study past work. This helps identify gaps. It prevents repetition. It helps
position the study.
The third is to construct the scientific model. The model must be exact. It should spell out the
assumptions clearly. It should define all the variables. The assumptions should be realistic
enough to provide insight.
The fourth step is to solve the model. Using math or simulation, the researcher solves the
prepared model. He checks for logical consistency. The solution should be complete; any
shortcuts are explained.
The fifth step involves interpreting results. The researcher connects the results with the actual
problem. For instance, the model could indicate that safety stock increases with the variability of
demand. Such a revelation would help management understand the tradeoff.
Many models are too complicated for math. In such cases, simulation is helpful. Simulation
needs a good justification. It should be evident that math cannot solve the model.
A proper simulation study needs to involve several steps in detail. Input factors shall be defined.
Ranges shall be specified. Replications shall be run. Output shall be measured with clarity.
It is where the researcher will analyze the results by using statistical tools that can help in
comparing scenarios to spot patterns and test ideas. Simulation does not generate proof. It
does generate insight. Insight needs to be explained with clarity.
7. CONDUCTING EMPIRICAL QUANTITATIVE RESEARCH
Empirical research studies real systems. It tests assumptions. It observes patterns. It challenges
theory when needed.
The first step is to identify assumptions in the theoretical model. For example, many inventory
models assume fixed lead time. The researcher must test this assumption. Real lead times often
vary.
The second step is to identify the type of system. A job shop behaves differently from a flow
shop. A retail store behaves differently from a warehouse. Each system needs different
definitions.
The third step is to create operational definitions. Operations Management lacks standard
measures. This creates confusion. For example, cycle time might include waiting or exclude it.
The definition affects results.
The fourth step is to create hypotheses. These hypotheses must be clear. They must involve
measurable variables. For example, higher variability might increase waiting time. The
hypothesis must state this clearly.
The fifth step is to design measurement systems. These systems must record accurate data.
They must track time, flow, utilization, and output. They must reduce noise.
The sixth step is to collect data. This step is difficult. Real systems change. Human choices
affect outcomes. Data might be incomplete or noisy.
The seventh step is to analyze data. Statistical methods help test hypotheses. They reveal
patterns. They help validate or challenge a theory.
The final step is interpretation. The researcher explains what the results mean. The researcher
identifies limits.
Empirical research requires access to firms. It requires trust. It requires an understanding of
processes.
8. ISSUES OF RIGOR
Rigor means clarity, accuracy, and consistency. Axiomatic research is often rigorous. Math is
exact; models are formal; assumptions explicit.
But rigor is not enough. A model can be rigorous yet useless when its assumptions fail. Most
early models employed unrealistic assumptions. They possessed rigor but no relevance.
Empirical research lacks rigor. Real systems have noise. Measurements are not replicable.
Construct validity is a problem. For instance, two plants could measure setup time differently.
Researchers need to establish unequivocal definitions. They need to devise sound systems of
measurement. They need to carefully process data.
Rigor helps to build strong conclusions; it supports trust in results.
9. ISSUES OF RELEVANCE
Relevance means usefulness. A relevant model matches real-world constraints. It helps
managers make decisions.
Many early academic models had no relevance; they avoided real issues, did not represent
variation or human behavior, were easy to solve, and had little value.
A relevant model need not have perfect realism. It must capture the most influential factors. For
instance, a lot sizing and scheduling model should incorporate setup times if the setup times are
dominant.
Industry breakthroughs like MRP and lean systems worked because they solved a real issue.
The theory followed practice in those cases.
Relevance does depend on context. It depends on how managers use the model.
10.BALANCING RIGOR AND RELEVANCE
The future of Operations Management depends on balanced research. Axiomatic models offer
structure, while empirical data offer realism.
A robust study uses both. It uses models to generate ideas. It uses data to test assumptions. It
refines models based on findings.
For instance, a firm may test a theoretical scheduling rule. Data could indicate that worker
flexibility influences performance. The model may now be revised to include that factor. This
cycle builds deeper knowledge.
Balanced research avoids extremes. Pure theory, without data, has little value. Pure
observation, without structure, has little use generally.
11.IMPLICATIONS FOR STUDENTS
Students often feel that models in Operations Management appear simple. They find gaps
between textbooks and real systems. It is quite natural.
Models aid students in thinking clearly. They show tradeoffs. They teach structure. They do not
replace judgment.
The students should learn using the models with care. They should know about assumptions.
They must adapt decisions with respect to the situation.
Students also need to learn to read data. Real systems contain variation. Learning how to study
variation is key.
Students attending these classes in research methods are thus better prepared for the
implementation of management work. The students will develop a structured way of thinking,
learn how to break down problems and to measure processes.
12. IMPLICATIONS FOR MANAGERS
Models guide decisions, managers use them to understand demand, inventory, flow, and
capacity; then they adjust them for the real world constraints.
Managers need lucid insights. They want models that can help them act. Tools must be simple
and practical. Complex formulas are avoided unless supportive of decisions.
Managers learn from empirical studies. They learn through experiments within their firms. They
learn from past performance. They learn from data.
Strong research assists managers in the formulation of improved processes. It helps them
reduce delays and thus helps them to plan better and respond to uncertainty.
13.FUTURE DIRECTIONS FOR OPERATIONS MANAGEMENT RESEARCH
This has meant that, with increased amounts of digital data and modeling within the industry,
there has been a subsequent increase in 'big data' captured by business regarding process
flows. Such data supports empirical study and is used in conjunction with a range of
computational and simulation-based techniques for establishing and confirming the validity of
comprehensive models using these new tools. Machine learning methods for pattern recognition
data, and data for empirical studies, will be important aspects of future research into digital
supply chains, automation and use of robotics and new service delivery methodologies and how
organisations manage and coordinate across their networked supply chains, and how
organisations decide to act based on the volume of digital flow data available. Future research
will increasingly focus on both quantitative models/forms of empirical research, as well as
develop a continued emphasis on empirical research to validate and refine theories associated
with these activities.
14.CONCLUSION
Many different stages have contributed to the development of Operations Management
research. It has grown from simple time study methods through developing ideal types of
models, to the addition of Operations Research methods. Also, it has benefited from
understanding industrial applications, and application of new concepts from both System
Dynamics and Queuing Theory, while being heavily influenced by the emergence of Lean
Production and Workload Control processes.
A fundamental understanding derived from this field is that a trade-off must exist between
structure, that is, model building, and realism, that is, how models are actually utilized.
Axiomatic Research brings in the level of detail in constructing well-defined and coherent
models while Empirical Research validates and strengthens these assumptions. Good
Operations Management research, therefore, should involve both Axiomatic and Empirical
research approaches.
Quantitative modeling of Operations Management allows managers to understand how well a
process is working, facilitates informed decision-making, and identifies tradeoffs between
options. It also allows them to understand how much variation is occurring within their
processes.
The rigor in which this field engages with both accepted research methods and what problems
are being solved will determine the future of it. It will also require precise measurements and
logical reasoning. Foremost, Operations Management requires its practitioners to be committed
to clearly reasoned articulation.
This paper has given a structured and simplified explanation of the Quantitative Methods in
Operations Management to assist students or Managers who require assistance with
understanding Quantitative Methodology and an overview of what these methods will continue
to foster in the future by creating increased processes and improved Operations performance.
REFERENCE
Bertrand J W M, Fransoo J C. Operations management research methodologies using
quantitative modeling. 2002.
Forrester J W. Industrial Dynamics. 1961.
Taylor F W. The Principles of Scientific Management. 1911.
Ackoff, R. The future of operational research is past. 1979.
Hopp W, Spearman M. Factory Physics. 1996.
Goldratt E, Cox J. The Goal. 1984.
Ohno T. Toyota Production System. 1988.
Womack, JJ.P.P, Jones D D.T. Lean Thinking. 1996.
Spearman M L, Woodruff D. CONWIP systems in manufacturing. 1990.
Little J. A proof for the law of average waiting, often called Little’s Law. 1961.
Shingo S. A Study of the Toyota Production System. 1981.
Schmenner, R. Service Operations Management. 1995.
Chase R. The customer contact approach to service system design. 1978.
Ward P, Zhou H. Inventory and performance in manufacturing firms. 2006.
Skinner W. Manufacturing strategy concepts. 1969.