Transcript Slide 1
Bob Travica MIS 2000 Instructor: Bob Travica Class 11 Decision Making Processes, System Support & Decision Support System Updated May 2015 Outline • Decision making and problem solving process rational (scientific) model • MIS & DSS support to rational decision making • Decision Support System • Decision making under constraints (realistic) • Summary • Exercise of11 13 22of Decision making and problem solving process: Rational (Scientific) Model (H. Simon, 1950s) 1. Identify (define) problem ? 5. Evaluate/Adjust solution 4. Implement solution 1 % 2. Define optional solutions P 0 3 3. Evaluate optional solutions exhaustively and select best • Steps 1-3 are decision making; with steps 4-5 the whole process is problem solving. Read the case of sales drop in the chapter. • Desirable model in business and generally but its assumptions limit real applications. 3 of 11 Rational Decisional Making Process and Information Systems 1. IDENTIFY PROBLEM Manager (Business Analyst assisting) Define problem (may include negotiations) 2. DEFINE OPTIONS 3. SELECT OPTION Create optional solutions Evaluate options and define decision Store options and evaluation criteria Process & evaluate options/rank options Create reports & queries MIS TPS Create organization performance indicators & environment scan DSS The role of MIS is to inform the user so that a potential problem in operations can be identified. Also in Creating optional solutions. TPS may help. Model-driven DSS can support the entire DM process. Best option can be proposed by DSS, but the user should still make the ultimate choice. 4 of 11 Decision Support System (DSS) Executive managers’ decision making is supported by DSS. DSS Data modeling & Data mining Outputs (data modeling): • KPI* & drill-down • What-If analysis (fitting input values to desired output value) • Scenarios (testing effects of inputs of output value) • Statistical tests • Process simulation • New relationships & patterns (data mining) DSS Supports higher mgt. levels. Problems to solve are less structured than in routine situations where TPS & MIS can do the work. DSS uses outputs from MIS and data from sources outside the company. Time horizon of DSS: Future – from close (day, week) to far (one+ years). More 5 of 11 Management tasks: Planning and forecasting Visually representing key performance indicators (e.g., financial ratios); dashboards. DSS may have a “drill-down capability” to find what is behind aggregated figures 6 of 11 DSS Types • Model–driven: Processes data using different transformation methods (optimization, if-then analysis, statistical tests); cited in chapter’s case MIS Data transformation models - What-If - Process simulation - Statistical tests environment • Data-driven: Looking for patterns in large data sets. • Data mining and Big Data TPS MIS environment Data Warehouse - Events in sequence (buying this & then that) - Things that go together (contents of shopper’s cart) - Grouping of people (customer types) Data Mining software 7 of 11 Other Models of Decision Making Satisfycing (Good Enough) • Rational decision making not always possible • Organizational & human limitations (time pressure, resources, difficulties in understanding a situation) • Alternative: Making a choice that is good enough 1 Define problem under constraints 2 Define some optional solutions 3 Focus on key evaluation criterion 4 Select first solution that fits the criterion More 8 of 11 Other Models: Zig-Zag Problem Solving (“Muddling Through”) • How decision making typically runs in public organizations: - different interests that counter each other - nobody to “cut the knot” in struggles over budget - “endless” decision processes, like walking through mud. • Issues: What is the problem? Different things for different people! Negotiations and maneuvering in defining problem. • Optional decisions defined tentatively. • Decision maker implements just a part of a chosen option; If blocked, switches to optional solution… and over again… 9 of 11 Zig-Zag Problem Solving – Process Diagram Administrator Define problem Internal Stakeholder Define problem External Stakeholder Define problem Make a compromise definition Implement an option partly Evaluate situation Y React React Blocked? N Implement another part Evaluate situation Switch to another option React React Only without blocking to any part of a decision, it is completely realized. But there is usually some blocking and the loop may seem infinite. 10 of 11 Summary In Simon’s rational model, decision making (DM) takes 3 steps: (1) identify problem, (2) create optional solutions, and (3) evaluate optional solutions exhaustively and select the best one. The full problem solving (PS) process includes 2 more steps: (4) implement solution, and (5) evaluate effect & adjust. DSS supports higher mgt. levels in solving less structured problems related to upcoming activities in the near to far future. It can be modelor data-driven. TPS & MIS support start of decision process. Role of MIS in decision making processes is to inform the user at the problem identification step of rational DM. DSS can support whole DM process. People still make the final choice. Satisficing and Zig-Zag DM acknowledge constraints in problem solving. 11 of 11