What Is Decision Analysis? Core Concepts, Elements, Types, and Steps

1 Concept of Decision Analysis

In modern management, decision is understood in two ways: a narrow view sees decision as simply making a choice among alternative actions (the so‑called "sign‑off"), while a broader view treats decision as a process—decision analysis—whereby a decision maker, based on available information and experience, generates feasible alternatives, applies scientific methods to compare, analyze and evaluate them against decision criteria, selects the most satisfactory option, and continuously adjusts the plan based on feedback until the goal is achieved.

2 Basic Elements of Decision

Decision maker: the individual or group responsible for the decision, influenced by social, political, economic, cultural, and psychological factors.

Decision goal: a clear objective the decision maker wishes to achieve, which may be single or multiple.

Decision alternatives: the set of possible courses of action. They can be explicit (a finite set) or implicit (described only by constraints, potentially infinite), with linear programming and other operations‑research techniques used to identify reasonable or optimal alternatives.

Nature state: the objective conditions of the decision environment that the decision maker cannot control but can anticipate; these may be certain or uncertain (discrete or continuous).

Decision outcomes: the results of each alternative under different nature states.

Decision criteria: the value standards used to evaluate whether an alternative meets the decision goal, usually reflecting the decision maker’s preferences or value orientation.

3 Classification of Decisions

Decisions can be classified according to various criteria.

3.1 Deterministic vs. Non‑Deterministic Decisions

A deterministic decision has only one possible nature state, with complete information allowing the outcome to be known in advance. A non‑deterministic decision involves two or more possible nature states whose occurrence is uncertain. Non‑deterministic decisions are further divided into competitive, risk, and fully uncertain types.

When one possible nature state is a competitor beyond the decision maker’s control, the decision is competitive (e.g., choosing a marketing strategy while considering rivals). If the probabilities of the possible nature states can be predicted, the decision is a risk decision. If the probabilities cannot be determined, it is a fully uncertain decision, which relies heavily on the decision maker’s personal preferences and value orientation, such as developing new technology or estimating market demand for a new product.

3.2 Single‑Stage vs. Sequential Decisions

Based on the continuity of the decision process, decisions are divided into single‑stage (static) and sequential (dynamic) decisions. A single‑stage decision addresses a specific point in time or a particular period, requiring only one action plan—for example, planning the annual production volume of a product.

A sequential decision consists of a series of interrelated decisions ordered over time, where each prior decision influences the subsequent one. The focus is on the overall rationality of the sequence rather than the benefit of any single decision—for instance, a company aiming to achieve a 90% market share within five years may break the total goal into yearly targets, devise corresponding actions, and adjust the plan each year based on the previous year’s performance.

4 Basic Steps of Decision Analysis

A rational and scientific decision‑making process generally includes five steps: (1) problem discovery and analysis; (2) determination of decision goals; (3) formulation of feasible alternative plans; (4) comparative analysis of alternatives to select the optimal one; and (5) implementation, feedback, and adjustment of the chosen plan.

Reference

Guo Wenqiang, Sun Shixun, Guo Lifu. Decision Theory and Methods (3rd Edition). 2020. Beijing: Higher Education Press.