DECISION SUPPORT & ARTIFICIAL INTELLIGENCE The New Science of Management Decision young field TEDSS Case still dynamic Herbert Simon (1960) Wall Street Journal -Exodus (30Sep05) Wall Street Journal -Terrorist (17Feb06) Caspian Sea Pipeline Project (2012) ProQuest assignment: analytic hierarchy process TEDSS, Decision Types, Phases & DSS day.1 1b TEDDS illustrates decision types supported by DSS (details: More Efficient Exodus, WSJ) 13 benefits of building DSS (similar: regression) S-1 the journey into decision making will lead to a major revolution in management and organization (observed in CIO cases) 2 TERRAIN MAP of decision making is needed because: S-2 different (IT) techniques are required for different phases of the journey 3 TYPES OF DECISIONS S-3a Programmed (routine sequence of responses so studied / programmed) S-3b opportunity use ITS to program decisions (TPS, MIS JIT) S-3a Nonprogrammed (unique & ill-structured) S-3c require flexible / interactive (IT) techniques (DSS, AI) 4 In reality, decisions exist in continuum requiring a complement of ITS: PG &Vanity Fair used DSS (Stonyfield : Access) to access & analyze information from TPS & IOS 5.a PHASES OF DECISIONS S-6.2 complex / unstructured problems are typically encountered in design & choice phases of nonprogrammed decisions DSS(AI) useful Build:Buzzsaw; Dock:Sparks; Home:TEDDS,PredictTerrorist,N-Site; Indian:Covisint,DBMS; Jackpot:WiNet,Celtics; Rx:CPOE-EMR S.4 Assignment: proquest search (analyic hierarchy process) day.2 2 WHAT IS A DSS? 6.1 flexible/interactive ITS support complex/ill-structured DM 6.2 typically applied to design/choice in nonprogrammed decisions 6.3 AHP is an example 8 momentous changes for (complex/ill-structured) DM ProQuest shows applied to many diverse decisions 9 AHP developed to assist DM evaluate (design / choice phases of) complex judgmental (involving qualitative criteria) problems 7 Analytic (decompose into components) design Hierarchy (organize into hierarchical structure) Process (simpler pair-wise comparisons) choice 13 Simple, Intuitive, Powerful 9 understanding (components of decision) 10 management of complexity (hierarchical structure, simpler choices) 11 sophistication / speed of information processing (infers priorities from series of simple choices) 14a EC applied to spouse choice demonstrates DSS & components assignment: use EC to weight presentation criteria 16 integrated into GDSS (Ranking Technical Managers) DSS extended by groupware & telecommunications GDSS 16.a What is GDSS? 17 18 22 Hiring technical manager is complex, ill- structured, team decision Design and choice phases of GDSS H relate to table handouts Benefits of GDSS day.3 A Decision Support System to Save Lives day.1 1. You cannot see, hear, or smell nuclear radiation, but it’s deadly all the same. In 1986 about 3.5 million people in Belarus, Russia, and Ukraine were exposed to it when there were two explosions in Unit 4 of the Chernobyl Nuclear plant. Some of the effects are only now emerging, like the high incidence of thyroid cancer in children. Experts agree that many health problems would have been avoided if an evacuation plan had been in place. 2. To avoid such a human tragedy in Virginia, power plant administrators commissioned the development of TEDSS to assist in creating an evacuation plan in case of a nuclear power plant disaster there. TEDSS is a special type of IT system called a decision support system. If a nuclear disaster were to occur, planners can use TEDSS to quickly (speed) determine the best evacuation routes (force) and how best to notify the public of those routes 3. For TEDSS to help determine the best evacuation strategies, it uses information on multiple variables that include The behavior of radioactive gasses, such as dispersion rates Highway system characteristics, such as number of lanes Population distributions, such as densities and the location of people with disabilities Current weather conditions, such as wind direction 4. If this information were static, administrators would have no need for a decision support system because the evacuation route would be developed once and the job would be done. However, the information fluctuates constantly depending on the time of day, time of year, and meteorological and economic conditions. Population densities also change, as does the highway system. The task, then, is to analyze all the information given the specific conditions of the moment, find the best solution, and find it quickly. 5. Some of the information that TEDSS needs resides within the power plant organization, such as the layout of the power plant, information on deadly gasses, and the escalation rate of the accident. Other information is supplied by state agencies and other external sources. 6. By processing this information in its simulation models, TEDSS provides output — some of it in graphic form — on the following factors: Evacuation routes and paths from any origin to assigned shelters Projected volumes of traffic on the highway system Highways that may become severely blocked by radiation The time that will have elapsed before the last vehicle clears the area 7. With these outputs produced by TEDSS, planners can evaluate traffic management strategies, such as one-way operation of highways, shoulder use, and flashing signals to reduce traffic congestion and to improve evacuation time. A Decision Support System to Save Lives 1. You cannot see, hear, or smell nuclear radiation, but it’s deadly all the same. In 1986 about 3.5 million people in Belarus, Russia, and Ukraine were exposed to it when there were two explosions in Unit 4 of the Chernobyl Nuclear plant. Some of the effects are only now emerging, like the high incidence of thyroid cancer in children. Experts agree that many health problems would have been avoided if an evacuation plan had been in place. 2. To avoid such a human tragedy in Virginia, power plant administrators commissioned the development of TEDSS to assist in creating an evacuation plan in case of a nuclear power plant disaster there. TEDSS is a special type of IT system called a decision support system. If a nuclear disaster were to occur, planners can use TEDSS to quickly determine the best evacuation routes and how best to notify the public of those routes 3. For TEDSS to help determine the best evacuation strategies, it uses information that includes The behavior of radioactive gasses, such as dispersion rates Highway system characteristics, such as number of lanes Population distributions, such as densities and the location of people with disabilities Current weather conditions, such as wind direction 4. If this information were static, administrators would have no need for a decision support system because the evacuation route would be developed once and the job would be done. However, the information fluctuates constantly depending on the time of day, time of year, and meteorological and economic conditions. Population densities also change, as does the highway system. The task, then, is to analyze all the information given the specific conditions of the moment, find the best solution, and find it quickly. 5. Some of the information that TEDSS needs resides within the power plant organization, such as the layout of the power plant, information on deadly gasses, and the escalation rate of the accident. Other information is supplied by state agencies and other external sources. 6. By processing this information in its simulation models, TEDSS provides output—some of it in graphic form—on the following factors: Evacuation routes and paths from any origin to assigned shelters Projected volumes of traffic on the highway system Highways that may become severely blocked by radiation The time that will have elapsed before the last vehicle clears the area 7. With these outputs produced by TEDSS, planners can evaluate traffic management strategies, such as one-way operation of highways, shoulder use, and flashing signals to reduce traffic congestion and to improve evacuation time. DSS-1.b Introductory Case A Decision Support System (DSS) to Save Lives illustrates types of decisions supported by DSS Evacuation planning for nuclear power plant disaster multi-factor TEDSS.3&5 complex + High Cost & Benefit dynamic TEDSS.4 non-routine unstructured TEDSS.2&7 best solution Force TEDSS Speed quickly at the moment TEDSS.6 simulation: regression color graphics output (N-GAGE, Sparks, Celtics-HO) Details: Modeling…more efficient exodus. WSJ.30Sep05 Figure 2: The presentation of the Surry Power Station area consists of the highway network and the protective action zones (PAZs). NBC10.com/traffic Provides similar graphic output (based multi-variate model) that assists decision maker in selecting best route. TEDDS: 2,7 DSS-1.b Introductory Case A Decision Support System (DSS) to Save Lives Evacuation planning for nuclear power plant disaster (terrorism) + multi-factor TEDSS.3&5 complex High Cost & Benefit dynamic N-Site (AHP) TEDSS.4 non-routine unstructured best solution DSS benefits Force TEDSS TEDSS.2&7 Speed quickly at the moment color graphics output (N-GAGE, Sparks, Celtics-HO) Tools to Predict Likely Teroritst Moves. WSJ.30Sep05 DSS-13 DSS What are the benefits of a DSS? involves interaction between decision maker and IT system that supplements & enhances human decision making by increasing: V understanding of the problem/decision by examining force intuition/judgment/experience of decision makers (spouse choice, accounting policy, terrorist activity, pipeline path) V management of complexity & lack of structure Excel V sophistication of processing capabilities Access N-GAGE(AHP) V speed of access to & processing of information u for an organization, the enhancement of human decision making contributes to: V increasing flexibility & productivity (Dock:Sparks, DHS:N-Site) V decreasing costs (Jackpot : Celtics, Rx:CPOE) u Introduction S-1 New Science of Management Decision Herbert Simon (1960) u …this journey (into the decision making process) is going to lead…(to) a major revolution in the art and science of management and organization (observed in CIO cases) S.7.4 u …the computer and the new decision-making techniques associated with it are bringing momentous changes like machinery brought to manual jobs S.xi.1 Proquest u In 1960, described a terrain map of decision making that provides a framework the discussion in new millennium S.7.5 DSS.2 Types of Decisions DSS-3 What types of decisions do you face? Structured Processing specific info. in specified way Recurring repeatedly or periodically Programmed TPS & MIS Non-recurring infrequently Unstructured No precise way to get a right answer AI (DSS) CIO assessments (N-Gage) Celtics Analytics Slam Dunk Non-programmed Plant Layout Line Balancing Spouse Choice (N-Gage) Hiring Technical Managers … or University Faculty Accounting Policy Caspian Sea Pipeline Path Types of Decisions S-2 Different Processes for Programmed vs Nonprogrammed Decisions u My reason for making the distinction is that different techniques are used for handling the programmed and the nonprogrammed aspects of decision making S.5.3 S-3.a Types of Decisions Two Polar Types of Decisions u Programmed V to the extent repetitive and routine so that a definite procedure has been developed S.5.4 V the name is from the computer field indicating a detailed sequence of responses exists to a complex task environment S.6.2 u Nonprogrammed V to the extent novel and unstructured so that no specific procedures have been developed to respond to the situation S.6.1 V general problem solving capacities (judgment, creativity, heuristics) are used to respond S.6.3 V frequently ineffective and high cost so focus of DSS and AI S.6.4 S-3.b Types of Decisions Different Processes for Programmed vs Nonprogrammed Decisions u Programmed decisions are frequently observed so they can be studied and better understood (sales, purchases, inventory, registration, grades) V traditionally, organizations use SOP (habit) and structure (departments & committees) S.9-10 V opportunity to use IT (TPS, MIS, JIT) so focus on nonprogrammed (Gresham’s Law) S-3.c Types of Decisions Different Processes for Programmed vs Nonprogrammed Decisions u Nonprogrammed decisions occur infrequently so they are less studied and less understood (spouse choice, team project assessments, accounting policy decisions, Caspian Sea pipeline path) Vtraditionally, organizations use general problems solving capabilities (judgment, heuristics, task forces, selection /training) Vopportunity to use IT (DSS &AI ) to study & improve Human thinking, problem solving & learning have been mysterious processes that have been labeled but not explained S.13.1 nonprogrammed decision making will soon undergo as fundamental a revolution as the one currently transforming programmed decisions in business organizations S.21.1 DSS-4 Continuum of decisions, Complement of ITS Recurring Nonrecurring Structured Nonstructured Which supplier to use, based Probably recurring only on price In between, if consider factors other than price Which car insurance to buy at renewal time In between (Continuum / Not discrete) Mostly unstructured Recurring Whether to expand business Nonrecurring into Eastern Europe What plants to include in the Nonrecurring landscaping around a new building How to use tax regulations Recurring to fill out an income tax form How many lanes to put into Nonrecurring a new bowling alley In between (Continuum / Not discrete) Unstructured if consider ethical & legal issues In between Types of Decisions S-4 A Continuum of Decisions, A Complement of ITS u They are not really distinct types, but a whole continuum. V We can find decisions of all shades of gray along the continuum, the terms programmed and nonprogrammed are concepts that define the range S.5.3 V The obvious reason why repetitive decisions tend to be programmed, and vice versa, is that if a particular problem recurs often enough, a routine procedure will usually be developed for solving it. S.6.0 u Not all IT systems and strategies are appropriate for every company or set of circumstances, but they are complementary BroadVision @CIO.com (092101) DSS-5.a Figure 4.3 Phases of the Decision Making Process Spouse Choice N-GAGE (AHP) GDSS [ DSS (AI )] Complex Ill-structured Subjective Examples Hire technical manager Accounting Choices Access Assessment Spouse Choice N-Site Pipeline DSS-5.b Decision Process What is the Decision Making Process? u Intelligence - find what needs fixing V recognizing a threat or opportunity u Design - find fixes V developing possible solutions u Choice - pick a fix V selecting the best solution u Implementation - apply the fix V carrying out the solution, monitoring results & adjusting Figure 4.3 S-5 Decision Process Three Principal Phases of Decision Making u Decision Making Comprises Three Phases S.2.1 The first phase of the decision-making process searching the environment for conditions requiring a decision - I shall call intelligence (borrowing the military meaning). The second phase inventing, developing and analyzing possible courses of action - I shall call design activity. The third phase - selecting a particular course of action - I shall call choice activity. S-6 Decision Process Phases of Decision Making u Generally, intelligence activity precedes design, and design precedes choice. S.3.2 u The cycle of phases is, however, far more complex than this sequence suggests. Figure 5.3 V Each phase in making a particular decision is itself a complex decision-making process. V The design phase may require new intelligence activities V Problems at any given level generate subproblems that, in turn, have their intelligence, design and choice phases. DAY 2 N-GAGE ENGAGEMENT for SPOUSE CHOICE day.2 1.b WHAT IS A DSS? 6.1 flexible/interactive ITS support complex/ill-structured DM 6.2 typically applied to design/choice in non-program. decisions (Spouse Choice) 6.3 AHP is an example 8 momentous changes for (complex/ill-structured) DM ProQuest shows applied to many diverse decisions 9 AHP developed to assist DM evaluate design / choice phases of complex judgmental (qualitative criteria) problems 7 Analytic (decompose into components) design Hierarchy (organize into hierarchical structure) Process (simpler pair-wise comparisons) choice 13 Simple, Intuitive, Powerful 9 understanding (components of decision) 10 management of complexity (hierarchical structure, simpler choices) 11 sophistication / speed of information processing (infers priorities from series of simple choices) 14a EC applied to spouse choice demonstrates DSS & components assignment: use EC to weight presentation criteria 16 integrated into GDSS (Ranking Technical Managers) DSS extended by groupware & telecommunications Psychological Foundation: Approach – Avoidance principle Analytic: decompose into components (more complete analysis) Hierarchical Structure (smart people organize) ABCDEF ABCDEF ABCDEF ABCDEF Process of pairwise comparisons: (easier 2 @ time, on all criteria) N-GAGE PROCEDURES ► ► ► ► ► ► Go to kennedyonline.us Select N-Gage Run Run File New ► Enter file name & description ► Next ► Enter # of levels under the goal ( goal is level 0) ► Next ► Enter goal name & description ► Done ► Select Design tab ► Select Criteria Select level 1 ► Name & describe criterion ► OK ► Name & describe other criteria as previously until finished level 1 Select Criteria ► Select next level (2) ► Name & describe elements (alternatives) ► OK ► Name & describe other alternatives as previously until finished level 2 ► Select goal ► Select arrow button to draw relationships ► Draw arrows from the goal to each criterion in level 1 ► When finished drawing arrows to define relationships at level 1, select a criterion ► Draw arrows from the criterion in level 1 to each alternative in level 2 to be rated on that criterion ► Repeat the process for each criterion N-GAGE PROCEDURES ► Select the Compare tab ► Double click the Goal (a pair-wise comparison – PWC – table appears) Select the first row in the table ► Compare (approach : avoidance) the criteria using the number line ► Select the next row in the table ► Repeat the comparison process ► OK when the comparisons are finished in the last row of the table ► Double click a criterion in level 1 (a pairwise comparison table appears) ► Select the first row in the table ► Repeat the comparison process in the table just as previously ► Repeat the process for each criterion ► Select the Goal when finished the comparison process for each criterion ► Select the Solution tab ► The hierarch of criteria & alternatives appears with the weights derived from the PWC process on the arrows and an IR (inconsistency ratio) in each box ► If the IR > 0.10 a “Revise IR!” statement appears in the box ► To revise the inconsistent comparisons, select the compare tab ► Then, double click on the box with the inconsistency message ► Redo the comparison process as before ► When all the inconsistencies have been resolved, select the Goal and the select the Solution tab ► Print the hierarchy (if the hierarchy does not print on a single page, use print screen PrtSc – to copy and past into PowerPoint or Word) DSS-5.a Figure 4.3 Phases of the Decision Making Process Spouse N-GAGE Choice (AHP) GDSS [ DSS (AI )] Subjective Complex Ill-structured Examples Hire technical manager Accounting Choices Access Assessment Spouse Choice N-Site Pipeline DAY 3 DECISION SUPPORT SYSTEMS 1.b WHAT IS A DSS? 6.1 flexible/interactive ITS support complex/ill-structured DM 6.2 typically applied to design/choice in non-programmed decisions (Spouse Choice) 6.3 AHP is an example 8 momentous changes for (complex/ill-structured) DM ProQuest shows applied to many diverse decisions 9 AHP developed to assist DM evaluate design / choice phases of complex judgmental (qualitative criteria) problems 7 Analytic (decompose into components) design Hierarchy (organize into hierarchical structure) Process (simpler pair-wise comparisons) choice 13 Simple, Intuitive, Powerful 9 understanding (components of decision) 10 management of complexity (hierarchical structure, simpler choices) 11 sophistication / speed of information processing (infers priorities from series of simple choices) 14a EC applied to spouse choice demonstrates DSS & components assignment: use EC to weight presentation criteria 16 integrated into GDSS (Ranking Technical Managers) DSS extended by groupware & telecommunications DSS-2 What is a DDS? Artificial Intelligence Excel Access Expert Choice ( AHP) N-Site networks Recurring Programmed Structured { Non-recurring Non-structured } Nonprogrammed (qualitative) DSS-6 DSS What is a Decision Support System (DSS)? u u u DSS is a flexible and interactive IT system designed to support decision making when problem is complex / unstructured V frequently includes AI models like N-GAGE (AHP) V must be flexible / interactive in response to / because of problems that are complex / unstructured usually involves qualitative criteria typically in design / choice phases of decisions ACCESS Examples: PROJECTS V V V Excel – statistical & what-if analysis (flexible budgets) MS Access –data mining tools like queries and reports N-GAGE – employs AHP to evaluate complex hierarchical problems involving qualitative criteria DSS-7 DSS-6 DSS What is a Decision Support System (DSS)? u u u DSS is a flexible and interactive IT system designed to support decision making when problem is complex / unstructured V frequently includes AI models like N-GAGE (AHP) V must be flexible / interactive in response to / because of problems that are complex / unstructured usually involves qualitative criteria typically in design / choice phases of decisions ACCESS Examples: PROJECTS V V V Excel – statistical & what-if analysis (flexible budgets) MS Access –data mining tools like queries and reports N-GAGE – employs AHP to evaluate complex hierarchical problems involving qualitative criteria DSS-7 A-12 Select Access Projects Choice: order CIO.1 presentations (Rx Tentative dates T TH CIO Team Access Project 800 A: Build) 930 Data Definition (cradling : relevance) B: Application Generators (capturing : reliably) BRING C: Application Generators (capturing : reliably) FLASH D: Application Generators (capturing : reliably) DRIVE E: Data Manipulation (creating : analysis) F: Data Manipulation (creating : analysis) THURSDAY 200 DSS-7 DSS: AHP Analytic Hierarchy Process u is an AI that models how experts approach complex hierarchical decisions involving qualitative criteria H H H Analytic - decompose complex problems into components (criteria / alternatives) Hierarchy - organize into meaningful structure Process of pair-wise comparisons involving trade-offs assigns weights to the criteria and preferences to the alternatives more natural to compare two things than numerous elements (Saaty 2000, 1990, 1977) Analytic: decompose into components Hierarchical Structure B D H I J Rx B D H I J Rx B D H I J Rx Process of pairwise comparisons B D H I J Rx ASSIGNMENT: N-GAGE (ASSESS CIO.1) DSS-8 DSS: AHP Simple, Intuitive…Powerful AHP has been applied to many diverse decisions: (ProQuest) V Strategic Planning for a Caspian Sea Pipeline Project 2012 V AHP-Delphi GDSS for Locating Whey Processing Facility 2008 u V N-Site: An Anti-terrorist Distributed Consensus Building and Negotiation Support System across the WWW 2006 ► Evaluating Characteristics of Financial Information 2004, 1995 V TQI Benchmarking Tools for Evaluating TQM programs 2003 V Translating Financial Phrases into Numerical Probabilities 1997 V Consensus Ranking of Technical Manager Candidates 1996 Qualitative Characteristics of Financial Information Decision Usefulness design Analytical: decompose into components RELEVANCE RELIABILITY Hierarchical Structure Predictive Value Feedback Value Timeliness Verifiability Neutrality Process of pairwise comparisons choice FASB: SFAC2 understanding Comparability (including Consistency) Representational Faithfulness manage complexity Similar to: Choice spouse & CIO/Access assessments DSS-10 AHP QUESTIONNAIRE Concept A o o Extreme o Very strong o o Strong o o o Moderate o Equal o o o Moderate o Strong o o Very strong o o Extreme Concept B Reliability o o o o o o o o o o o o o o o o o Relevance Reliability o o o o o o o o o o o o o o o o o Cost Reliability o o o o o o o o o o o o o o o o o Materiality Reliability o o o o o o o o o o o o o o o o o Comparability Relevance o o o o o o o o o o o o o o o o o Cost Relevance o o o o o o o o o o o o o o o o o Materiality Relevance o o o o o o o o o o o o o o o o o Comparability Cost o o o o o o o o o o o o o o o o o Materiality Cost o o o o o o o o o o o o o o o o o Comparability Materiality o o o o o o o o o o o o o o o o o Comparability DSS-12. DSS: EC Analytical Hierarchy Proces: Decision Support for complex judgmental problems u While the applications of AHP appear different, each “incorporated judgments on intangible criteria and other elements alongside tangible ones which have known measurements.” Saaty 1987 p.157 u N-GAGE & Expert Choice are an expert systems (artificial intelligence) based on AHP designed specifically to make explicit the judgments of experts in evaluating complex unstructured problems involving qualitative criteria DSS-13 DSS What are the benefits of a DSS? u involves interaction between decision maker and IT system that enhances & supplements human decision making by increasing: V understanding of the problem/decision by examining intuition/judgment/experience of decision makers V management of complexity and lack of structure V sophistication of processing capabilities V speed of access to and processing of information u for an organization, the enhancement of human decision making contributes to: V increasing flexibility and productivity V decreasing costs DSS-14.a What are the Components of a DSS? Expert Choice Prepares Auditors Users Graphics & prompts What makes accounting data decision useful? What accounting alternative should be used to report these events? AHP Figure 4.5 Pairwise comparisons DSS-14.b What are the Components of a DSS? Expert Choice graphics & prompts AHP pairwise comparisons Figure 5.6 DSS-14.c DSS What are the components of a DSS? DSSs vary in application and complexity, but all share specific components: u user interface V permits user to enter information, commands and models V should be simple, flexible, consistent u model management V stores and manages the DSS models u data management V stores and maintains information Relate to Expert Choice DSS-15 DSS What is the process of developing DSS? u Intelligence V examine the problem to determine if a DSS is needed DSS if the problem is complex / unstructured TPS, IOS or MIS if the problem is structured / routine DSS.2-3 u Design identify what is available to buy as an alternative to building V DSS generators: Excel, Expert Choice (Research) V u Choice V u compare build to buy considering cost, fit with the problem/decision, ease of use Implementation V test, evaluate and revise the DSS (Research) Consensus Ranking Technical Manager: similar spouse choice & CIO.1 ranking DSS-16a Facilitators: Researchers Team Members: Nursing Directors Nurse Managers Staff Nurse IT Tools DSS supports team DM when problem is complex / unstructured (qualitative criteria) as in CIO.1 RANKING } GDSS supports TEAM STRATEGY Netscape/ Outlook N-GAGE LAN / internet DSS-16.b GDSS What is Group Decision Support System (GDSS)? DSS that is designed to support decision making by a team especially when decision is complex /unstructured DSS-2 typically involves qualitative criteria DSS.17 H.135.6 GASB - setting accounting policies N-Site – multi-national response to terrorism Assign rankings to CIO.1 team presentations Consensus Ranking of Technical Manager Candidates Omega v24 n5 pp523-538 DSS-16.c GDSS What are the components of a GDSS? People Team Members H united by a common goal & task interdependence H experience, perspective, judgment Facilitators H nontechnical role to conduct the meeting H technical role related to IT tools IT Tools Omega Groupware permit team to provide input simultaneously/anonymously and view by others DSS capabilities to classify, analyze, and rank ideas Telecommunications hardware/software to network team members (LAN, WAN, Internet) DSS-17 Consensus Ranking Consensus Ranking of Technical Manager Candidates u Hiring technical managers is: Complex H broad range of skills is required H candidates have different skill sets Unstructured H qualitative criteria have no inherent scale / metric for trade-offs between criteria or comparisons between candidates H DMs have implicit criteria that may not be appropriate (e.g. religion, sex, race) and should be challenged Team / Group Decision H perceived importance of required skills (hiring criteria) will vary within and between organizational levels DSS-18 Consensus Ranking What are the phases of team/group decision making? I Develop explicit functional criteria that reflect organizational perspectives Brainstorming (Table 1, Table 2) Issue categorization & analysis (Table 3, Figure 1) 19.a-b II Assign priority to criteria & preferences to candidates Voting (Expert Choice, Figure 2) 20.a-b III Achieve a consensus while minimizing the dysfunctional effects of groups (groupthink, conflict, status) Ranking of the Candidates (Tables 6 & 8) 21.a-b DSS-19.a Consensus Ranking: Phase I (I) Developing the Hiring Criteria Hierarchy Each Decision Maker (DM) develops a list of criteria. Facilitators aggregate individual lists of criteria and develop a comprehensive list for each DM group. Table 1 Facilitators do a comprehensive literature search to prepare operational definitions of the criteria. Table 2 Omega: 525-8 DSS-19.b Consensus Ranking: Phase I (I) Developing the Hiring Criteria Hierarchy Each DM identifies related criteria sets and rank orders the criteria within each set. Table 3 Facilitators collect the rankings and develop a synthesized hierarchy of criteria for each DM group. Figure 1 Facilitators and each DM group meet to finalize each group’s criteria hierarchy. Omega : 525-8 DSS-20.a Consensus Ranking: Phase II (II) Assigning Importance to Criteria & Preferences to Candidates Facilitators familiarize DMs with principles of AHP and suitable AHP software such as Expert Choice (EC). DMs use EC for pairwise comparisons between criteria in the hierarchy in Figure 1. EC alerts DMs to logical inconsistencies and encourages DMs to repeat the pairwise comparisons process in the previous step. HR does initial screening and identifies several eligible candidates. Omega: 528-30 DSS-20.b Consensus Ranking: Phase II (II) Assigning Importance to Criteria & Preferences to Candidates DM groups interview candidates using their criteria hierarchy. DMs use EC for pairwise comparisons to evaluate candidates on each criterion at lowest level of hierarchy. EC alerts DMs to logical inconsistencies and encourages DMs to repeat the comparisons process in the previous step. pairwise Table 4 Omega: 528-30 DSS-21.a Consensus Ranking: Phase III (III) Identifying the Consensus Ranking of the Candidates Facilitators provide anonymous feedback to all the DMs about each group’s average weights and preferences. Figure 2 If the DMs are satisfied that there is adequate consistency within the group, the process proceeds to Step 2 below Otherwise, DMs to go back to Step 6 of Phase 2. Table 5 Omega: 530-34 DSS-21.b Consensus Ranking: Phase III (III) Identifying the Consensus Ranking of the Candidates Facilitators input the final data from Step 1 to MAH and obtain a consensus ranking of the candidates. Tables 6 Facilitators meet with DMs. Unless unforeseen dissatisfactions, the process of employment offers begins in the order prescribed by the GDSS (AHP & MAH). Tables 8 Omega: 530-34 DSS-22.a GDSS What are the benefits of a GDSS? u Achieves the benefits of a team approach to a decision by: V compressing time /space so individuals can interact V brainstorming to share ideas, facilitate consensus and achieve acceptance V perspectives from different organizational levels revealing required expertise in candidates (nursing directors, nurse managers, staff nurses) V explicit functional criteria are encouraged rather than implicit / inappropriate criteria such as sex, race, Omega religion. H.138.2 DSS-22.b GDSS What are the benefits of a GDSS? u minimizes the problems of group processes including: V groupthink in which different ideas are suppressed V conflict by keeping focus on problem not personalities V status differences that restrict participation (nursing directors, nurse managers, staff nurses) V sequential interaction in face-to-face meeting, Omega H.138.2