Notes: Chapter 1 – Introduction to Quantitative Analysis and Decision-Making - decision analysis

Problem Solving and Decision Making

• Chapter focus: Introduction to quantitative analysis and decision making; modeling cost, revenue, and profit; practical examples; overview of quantitative methods in practice.

The Problem Solving Process

essential steps for decision making

• Seven steps of problem solving:
  1. Identify and define the problem.
  2. Determine the set of alternative solutions.
  3. Determine the criteria for evaluating alternatives.
  4. Evaluate the alternatives.
  5. Choose an alternative (make a decision).
  6. Implement the selected alternative.
  7. Evaluate the results.
• Note: The first five steps constitute the decision-making process itself.
• Related framing: Define the problem, identify alternatives, determine criteria, evaluate alternatives, and choose an alternative as core steps; this is the “Quantitative Analysis and Decision Making” framework.

Quantitative Analysis and Decision Making

• Define the problem; identify alternatives; determine criteria; evaluate alternatives; choose an alternative; structure and analyze the problem; decision-making process.

Models of Cost, Revenue, and Profit (Intro to Modeling)

• Models are representations of real objects or situations.
• Mathematical models represent real-world problems with formulas and expressions based on key assumptions, estimates, or statistical analyses.
• Advantages of models:
  • Generally require less time to experiment with than real systems.
  • Usually less expensive.
  • Involve less risk.
  • The closer the model resembles the real situation, the more accurate the conclusions and predictions.

The Quantitative Analysis Process (Modeling Cycle)

• The process consists of four main stages:
  • Model Development
  • Data Preparation
  • Model Solution
  • Report Generation
• These steps turn a problem into a solvable mathematical representation and communicate results to decision makers.

Model Development

• Models are representations of real objects or situations; mathematical models use a system of mathematical formulas based on key assumptions, estimates, or data.
• Key takeaway: The quality of decisions depends on the quality of the model and its assumptions.
• Advantages reiterated: time-saving, cost-saving, risk reduction; accuracy improves as model fidelity improves.

Mathematical Models

• Objective Function: describes the problem’s objective (e.g., maximize profit, minimize cost).
  • Example (production): If x denotes units produced and sold per week, with a profit of $10 per unit, then the objective is to maximize profit: $ext{Profit} = 10x.$
• Constraints: restrictions or limitations (e.g., production capacity).
  • Example: If each unit requires 5 hours of production and total available hours are 40 per week, a capacity constraint is: $5x \,\le\, 40.$
  • The term 5x represents total time to produce x units; the constraint enforces time availability.
• Uncontrollable Inputs: environmental factors not under the decision maker’s control.
• Decision Variables: controllable inputs or decision alternatives (e.g., the number of units to produce).

Transforming Model Inputs into Output

• A complete mathematical model maps inputs to outputs: uncontrollable inputs (environmental factors) and controllable inputs (decision variables) yield the output (projected results).

A Complete Mathematical Model (Simple Production Problem)

• Objective (example): maximize profit (from the optimistic viewpoint) or minimize cost (from the cost perspective).
• In the simple production example, a possible model is:
  • Decision variable: $x = ext{units produced and sold per week}$
  • Parameters: profit per unit, hours per unit, total available hours, etc.
  • Objective (maximize profit): $ext{Maximize } P(x) = 10x.$
  • Constraint (time): $5x \le 40, \, x \ge 0.$
• This yields the basic structure of the model (objective + constraints).

Deterministic vs Stochastic Models

• Deterministic Model: all uncontrollable inputs are known and fixed.
• Stochastic (Probabilistic) Model: uncontrollable inputs are uncertain and may vary.
  • Stochastic models are typically harder to analyze.
  • Example: If hours of production per unit could vary from 3 to 6 hours depending on material quality, the model becomes stochastic (not guaranteed of a fixed 5 hours per unit).
• Cost/benefit considerations must guide the choice between deterministic and stochastic models.

Data Preparation

• Data refers to the values of uncontrollable inputs to the model.
• Data preparation is non-trivial due to the time required and potential data collection errors.
• Example magnitude: a model with 50 decision variables and 25 constraints can have over 1300 data elements.
• Often, a fairly large database is needed; information systems specialists may be required.

Model Solution

• The analyst seeks the alternative (i.e., the set of decision variable values) that yields the best output for the model.
• Definitions:
  • The best output is the optimal solution.
  • An alternative that violates any model constraint is infeasible (rejected regardless of objective value).
  • An alternative that satisfies all constraints is feasible and a candidate for the best solution.
• Example structure (illustrative): a production problem with candidate solutions and their feasibility and profit values shows the iterative evaluation toward the optimal solution.
• Software options for solving models include: Microsoft Excel and LINGO.

Model Testing and Validation

• Model accuracy is often unknown until solutions are generated.
• Use small test problems with known or expected solutions to test and validate the model.
• If the model yields expected results, apply to full-scale problem.
• If inaccuracies are found, take corrective actions such as:
  • Collect more accurate input data.
  • Modify the model.

Report Generation

• A managerial report should be prepared based on model results.
• The report should be easily understood by the decision maker.
• Include:
  • The recommended decision.
  • Other pertinent information (e.g., sensitivity of the solution to assumptions and data).

Implementation and Follow-Up

• Successful implementation of model results is critical.
• Secure as much user involvement as possible throughout the modeling process.
• Continue to monitor the model’s contribution.
• It may be necessary to refine or expand the model.

Example 1: Break-Even Analysis

• Break-even analysis shows the point where total revenue equals total cost; below break-even, losses occur, above, profits accrue.

Example 2: Ponderosa Development Corp. (PDC)

• Context: A small real estate developer building one house style.
• Revenue per house: $R(x) = 115{,}000\,x.$
• Costs per house (variable costs): $V = 55{,}000 + 28{,}000 + 20{,}000\,= 105{,}000.$
• Fixed monthly costs: Office lease $= 2{,}000,$ utilities and equipment $= 3{,}000,$ salaries for seven permanent office employees total $= 35{,}000.$
• Total fixed costs: $F = 40{,}000.$
• Total cost function: $C(x) = F + V\,x = 40{,}000 + 105{,}000x.$
• Break-even condition: $R(x) = C(x)$ → $115{,}000x = 40{,}000 + 105{,}000x.$
  • Solve: $(115{,}000 - 105{,}000) x = 40{,}000 \Rightarrow 10{,}000x = 40{,}000 \Rightarrow x = 4.$
  • Break-even point: 4 houses per period (month in this example).

Using Excel for Break-Even Analysis (Ponderosa)

• Data organization in a spreadsheet:
  • Fixed Cost: $40{,}000$
  • Variable Cost per Unit: $105{,}000$
  • Selling Price per Unit: $115{,}000$
• Model section:
  • Sales Volume (x)
  • Total Revenue: $R = 115{,}000 \times x$
  • Total Cost: $C = 40{,}000 + 105{,}000 \times x$
  • Total Profit: $\text{Profit} = R - C$
• Question: What is the break-even point in monthly sales of houses?
• Spreadsheet solution using Goal Seek:
  1) Data → What-If Analysis → Goal Seek
  2) Set cell: Total Profit (e.g., the cell displaying Profit) to 0
  3) By changing cell: Sales Volume (x)
  4) Click OK to obtain the break-even sales volume.

Goal Seek: Steps (as described in the transcript)

• Step 1: Select Data on the menu
• Step 2: Choose What-If Analysis in Data Tools submenu
• Step 3: Choose the Goal Seek option
• Step 4: In the dialog box:
  • Set cell: the cell with Total Profit (e.g., B9)
  • To value: 0
  • By changing cell: the cell for Sales Volume (e.g., B6)
• Complete to obtain the solution (break-even x).

Management Science Techniques (List of Tools)

• Linear Programming
• Integer Linear Programming
• PERT/CPM
• Inventory Models
• Waiting Line Models
• Simulation
• Decision Analysis
• Goal Programming
• Analytic Hierarchy Process
• Forecasting
• Markov-Process Models
• Dynamic Programming

Practical/Philosophical/Ethical Considerations

• The model-building process requires transparency about assumptions and limitations.
• Decision makers should understand sensitivity and risk when relying on model results.
• Stakeholder involvement is crucial to ensure the model addresses real decisions and gains buy-in for implementation.
• Ethical considerations include data quality, fair assumptions, and avoiding overstated precision in uncertain contexts.

Quick Reference Formulas and Concepts

• Profit function (per-unit profit example): $P(x) = p \cdot x$ where p is the profit per unit.
• Time constraint example: $a x \leq T$ where a is hours per unit and T is total available hours.
• Break-even condition: $R(x) = C(x)$ where $R(x)$ is total revenue and $C(x)$ is total cost.
• Total Revenue: $R(x) = p x$ (p = price per unit)
• Total Cost: $C(x) = F + v x$ where F is fixed cost and v is variable cost per unit
• Optimization objective examples:
  • Maximize Profit: $\max P(x) = p x$ subject to constraints
  • Minimize Cost: $\min C(x) = F + v x$ subject to demand or capacity constraints