UPI Study › Courses › Foundations of Leadership › Free lessons › Lesson 7.8 Free with a free account
Quantitative Analysis
Foundations of Leadership · Chapter 7: Decision-Making ModuleSection 7.8 · lesson 45 of 9011 min · free with a free account
Free preview
Create a free account to keep watching
It takes a minute and costs nothing. Your free account unlocks all of Chapter 1 and 3 more lessons of Foundations of Leadership.
Lesson complete
Earn the 3 college credits for Foundations of Leadership
Enrolling adds a quiz after each lesson, the graded work and the official transcript. Recommended for credit by NCCRS.
Next lesson: Ethics in Decision Making →← The Garbage Can ModelNext lesson: Ethics in Decision Making →
In this lesson
What the methods can settle, and the three things they will always get wrong.
- Separate descriptive, predictive and prescriptive analytics
- Match method to question across five families
- Build and stress-test a weighted scoring model
- Name the limits that judgement must cover
Free study notes
Get this lesson's notes by email
The key points and 4 practice questions with answers, in your inbox. One email, free.
One email. Unsubscribe in one click.
Transcript
Everything said in the video, section by section (1,466 words). Read along, or search it.
Decision-Making Module
Quantitative Analysis
Modern organisations have access to more data than ever before. So the interesting question is no longer whether to use data in decisions. It is how. Which methods, on which problems, with what cautions. Quantitative analysis is the family of techniques built to answer that. It began in the operations-research work of the Second World War. Scientists were asked to optimise convoy routings and bomber crew sizes. It has since expanded into linear programming, forecasting, simulation, and optimisation. And into the broader field now called business analytics. We will cover the major methods and their appropriate uses. And, just as importantly, the limits that make them complements to judgement rather than replacements for it.
Three Phases
How the Toolkit Accumulated
Quantitative analysis in business has gone through three distinct phases. The first began with the operations-research successes of the nineteen forties. Its focus was optimisation. Methods for finding the best allocation of resources under stated constraints. The second ran from the nineteen sixties through the nineteen nineties. Its focus was statistical inference. Regression analysis, hypothesis testing, and forecasting models. Those let firms make defensible predictions about uncertain quantities. The third phase runs from the early two thousands to now. The era of business analytics and machine learning. Large-scale computational methods that find patterns without the analyst specifying them first. Each phase added new tools without eliminating the older ones. So a mature decision-maker today draws on methods from all three.
Three Layers
What Kind of Question Are You Asking?
Modern analytics is usually classified by what kind of question it answers. Descriptive analytics asks what happened. It summarises past performance through dashboards, reports, and key performance indicators. Predictive analytics asks what is likely to happen. It uses regression, time-series, and machine-learning models to forecast future quantities. Prescriptive analytics asks what we should do. It combines predictions with optimisation methods to recommend specific courses of action. The important structural point is that the three families build on each other. An organisation that lacks reliable descriptive analytics is rarely well-served by prescriptive recommendations. If you cannot say what happened last quarter, a recommendation engine is not your next problem.
Optimisation
Linear Programming Still Does the Work
Linear programming was developed by George Dantzig in nineteen forty-seven. It remains the workhorse of optimisation in business. Every linear programming problem has three elements. Decision variables, which are the quantities the decision-maker actually controls. An objective function, which is the thing to be maximised or minimised. And a set of constraints, which the solution has to satisfy. The applications are the classic operations problems. Production planning. Which products to make, in which quantities. Transportation. Which routes to use. And workforce scheduling. Which shifts to assign. Where the relationships are not linear, related methods extend the approach. Integer programming, nonlinear programming, dynamic programming. And where the outcomes themselves are uncertain, stochastic programming adds probabilistic elements.
Forecasting
Optimisation Needs Inputs That Do Not Exist Yet
Optimisation methods require inputs, and this dependency is easy to overlook. Those inputs are themselves uncertain, and have to be forecast first. Time-series forecasting uses past values of a variable to predict its future values. The methods range from simple moving averages, through exponential smoothing, to ARIMA models. Regression analysis goes further than that. It models the relationship between a dependent variable and one or more explanatory variables. It is the standard tool for questions like this one. How much will sales change if we raise prices by five per cent? Or, which factors most strongly predict customer churn? The strength of regression is its interpretability. The coefficients carry meaning you can explain to a non-technical decision-maker. Its weakness is the flip side of that. It requires strong assumptions about the form of the relationship, and the independence of the variables.
The Families
Matching Method to Question
It is worth holding the major families side by side. Linear programming answers what allocation maximises profit subject to capacity. It is provably optimal, but only under linearity, and real problems are often non-linear. Regression answers which factors drive an outcome, and by how much. Its coefficients are interpretable, but it is sensitive to specification choices. Time-series forecasting answers what demand will be next quarter. It rests on well-developed statistical theory, and on one large assumption. That the future resembles the past. Simulation answers how a complex system behaves under varied conditions. It handles non-linearities and uncertainty, at the cost of computation and input sensitivity. And weighted scoring answers which alternative best fits multiple criteria. It is transparent and easy to communicate. Its weights are entirely subjective. Read across that last column and you have the real skill in this area. Not running the method, but knowing which limit you have just accepted.
Simulation
A Portrait, Not an Answer
When a system is too complex for closed-form mathematical analysis, simulation is the alternative. A simulation model is a computational representation of the system. Its behaviour can then be observed under different assumptions and inputs. Monte Carlo simulation is the version most often used in business. It runs a model thousands of times. Each run takes random draws from the probability distributions of the uncertain inputs. What comes out is a distribution of outputs, not a single figure. That shows not just the most likely outcome, but the full range of possibilities. Supply-chain managers use it to test inventory policies. Financial analysts use it to value complex options. And epidemiologists use it to model disease spread under different policy interventions. Simulation does not produce a single right answer. It produces a portrait of how a system behaves. Which somebody still has to interpret.
Multi-Criteria
When There Is No Single Objective
Not all decisions reduce to a single objective. Many real choices require trading off cost, time, risk, environmental impact, and ethical fit. Weighted scoring models are the simplest widely used method for that. Each criterion is assigned a weight reflecting its relative importance. Each alternative is scored on each criterion. The total weighted score is computed for each alternative. And the highest score is the recommended choice. The technique's transparency is its real strength. Both the weights and the scores can be discussed, contested, and revised by stakeholders. More sophisticated methods exist. The Analytic Hierarchy Process, developed by Thomas Saaty. It uses pairwise comparisons to derive the weights more rigorously. But it rests on exactly the same underlying logic.
Case
The Conversation the Model Forced
A case where the model's real contribution was not the answer it produced. A Bengaluru-based e-commerce firm needed a new third-party logistics provider in twenty twenty-four. Five providers were shortlisted, on five criteria. Cost at thirty per cent. Delivery time at twenty-five. Reliability at twenty. Reverse-logistics capability at fifteen. And technology integration at ten. A procurement committee scored each provider on each criterion, on a one to ten scale. Two providers ranked within a single point of each other. So the committee ran a sensitivity analysis on the cost weight. The rankings flipped if cost was weighted at thirty-five per cent rather than thirty. So the committee stopped scoring, and held a discussion about how much cost should actually weigh. That conversation would not have happened without the weighted scoring model. They chose the slightly more expensive provider, on the strength of better technology integration. The value of quantitative analysis often lies less in the final number. And more in the structured conversation it forces people to have.
Three Limits
Where the Numbers Stop Helping
Quantitative methods are not a substitute for judgement. Three limits deserve emphasis. The first is the garbage-in, garbage-out rule. A model is only as good as the data and assumptions on which it rests. Sophisticated mathematics applied to poor data produces sophisticated nonsense. And sophisticated nonsense looks exactly like the real thing. The second limit is the illusion of objectivity, and it is the more dangerous one. Numbers feel more authoritative than judgement does. But every number in a model rests on judgements underneath it. Judgements about scope, about definition, and about measurement. All of which should be open for examination, and usually are not. The third limit is that quantitative methods favour what is measurable. That systematically excludes considerations that resist measurement. Culture. Ethics. Long-term reputation. Morale. None of them unimportant. Only unmeasured. So a mature decision-maker treats the analysis as one input among several.
Consolidate
Why the Regulator Requires a Human in the Loop
Indian banking gives a clear picture of quantitative analysis used correctly. Public-sector banks run a battery of methods over their loan portfolios. Credit-scoring models, which are essentially logistic regressions, predict the probability of default. Loan-pricing models combine that probability with the cost of funds and a target return. Which is how the interest rate gets set. And stress-testing simulations explore portfolio performance under macroeconomic shocks. None of these methods makes the credit decision by itself. Each one feeds into a structured judgement. A judgement that also incorporates relationship history, collateral assessment, and qualitative factors. The Reserve Bank of India's supervisory frameworks require that combination explicitly. And the reason is worth sitting with. Fully algorithmic credit decisions tend to embed historical biases. In ways that purely quantitative validation cannot detect. A model cannot audit the assumption it was trained on.
About Foundations of Leadership at UPI Study
Foundations of Leadership (Business 205) is a self-paced online course of 90 video lessons in 12 chapters, from the roots of leadership theory and the trait, situational and transformational approaches to power, decision-making, communication, teams, conflict, diversity and global leadership. It carries a NCCRS credit recommendation for 3 semester credits.
Every lesson transcript is free on pages like this one, and the videos are free with a free UPI Study account. Enrolling adds a practice quiz after each lesson, the graded work and final exam that earn the credit, lifetime access, and an official transcript on completion. UPI Study has helped over 130,000 students with college credits. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.
Enroll now → Course details Download the syllabus All Business courses Every course with PRO →
Compare: Management and leadership courses · Foundations of Leadership vs Leadership and OB
Why take it here
Why students take Foundations of Leadership with UPI Study
Where the credit goes
Universities that work with NCCRS
Foundations of Leadership carries a NCCRS credit recommendation. These universities work with NCCRS; some are UPI Study's signed partners. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.
Thomas Edison State UniversitySigned partner
University of the PeopleSigned partner
Charter Oak State CollegeSigned partner
Strayer UniversitySigned partner
Columbia Southern UniversitySigned partner
Indiana TechSigned partner
Penn StateWorks with NCCRS
Southern New Hampshire UniversityWorks with NCCRS
SUNYWorks with NCCRS
+many more cooperating universitiesFind my college →Find my college → Not listed? Ask Athena in the corner of this page, or book a 15-minute call.
FAQ
Questions students ask us
About Quantitative Analysis
What is the difference between descriptive, predictive, and prescriptive analytics?
What is linear programming used for in business?
What is Monte Carlo simulation in business decision making?
What are the limitations of quantitative analysis in decision making?
About the Foundations of Leadership course
How many college credits is Foundations of Leadership?
How much does Foundations of Leadership cost and what is included?
Which universities accept UPI Study credits?
How long does Foundations of Leadership take?
How are the exams done?
How do my credits reach my college?
Ready for the full course? Earn 3 credits in Foundations of Leadership.
90 video lessons, a quiz after each one, and the graded work that earns the credit. Recommended for credit by NCCRS.
3 credits · NCCRSEnroll →