Multicriteria decision making on selection of decision analysis software.ABSTRACT The problem approached in this work dealt with the direct selection of Decision Analysis Software, looking for Looking for In the context of general equities, this describing a buy interest in which a dealer is asked to offer stock, often involving a capital commitment. Antithesis of in touch with. selecting the more appropriate tool for the organization purposes, providing them the best investment return. The complexity existent ex·is·tent adj. 1. Having life or being; existing. See Synonyms at real1. 2. Occurring or present at the moment; current. n. One that exists. Adj. 1. in this problem is due to the difficulty found by the decision makers to evaluate the several inherent aspects of this problem. This work presents a multi-criteria decision model for Decision Analysis Software selection. Keywords: Decision Analysis Software Selection, Multiple Criteria Decision Analysis, SMART. 1. INTRODUCTION The dynamism of the atmosphere in which the organizations are inserted, and the variations provoked pro·voke tr.v. pro·voked, pro·vok·ing, pro·vokes 1. To incite to anger or resentment. 2. To stir to action or feeling. 3. To give rise to; evoke: provoke laughter. in the direction of the businesses, such things make the processes of planning Information Systems (IS) and the way how the decisions about these investments are made make this more and more important, being object of several publications, once the investments on IS have been more and more significant. These systems selection can happen in two ways, as a stage of the IS planning process, related with the planning methodology, or through the direct selection. In the direct selection the IS context is not considered, in such a way that a selection method is applied directly on the alternatives set. The problem approached in this work is the direct selection of a Decision Analysis Software, aiming to select the tool that is more appropriate for the purposes of an organization and that provides a larger return to this organization. To identify the best investment an approach of Multiple Criteria Decision Analysis was used, because through this, it will be possible to measure intangible gains associated to this type of investment. Most of these gains are intangible and they cannot be quantified in a monetary way, being that the reason for which the techniques of financial analysis or investment analysis won't won't Contraction of will not. won't will not won't will be considered. The present work is resulted of the real need of a Decision Analysis Software purchase for an organization. The current dilemma is the definition of the criteria and weights that will be used for the choice of the tool to be acquired. Therefore, it is wanted to define criteria and weights that can best represent the strategic objectives of this organization, looking for a selection model aligned with the strategies of the organization. 2. THE PROBLEM The analyzed an·a·lyze tr.v. an·a·lyzed, an·a·lyz·ing, an·a·lyz·es 1. To examine methodically by separating into parts and studying their interrelations. 2. Chemistry To make a chemical analysis of. 3. organization develops academic activities and partnerships with companies through courses (trainings) and research and development projects. The performance areas of this organization are those related to Decision Systems and Information Management. This organization develops projects in areas such as Energy, Petroleum & Gas, Services, Civil Construction, Manufactures, Computer science, The situation comes in a way that the decision maker has a resource destined des·tine tr.v. des·tined, des·tin·ing, des·tines 1. To determine beforehand; preordain: a foolish scheme destined to fail; a film destined to become a classic. 2. exclusively for the purchase of such a tool, and they need, therefore, to select among the ones that are available in the market that that will be acquired. The decision of this problem turns out complex when the largeness of the activities developed by the organization is analyzed, the diversity of sections in which it acts and the strategic objectives that the decision maker should satisfy in your decisions. All these pointed factors contribute to a situation where a multiple criteria decision analysis approach would be very suitable. 3. ESTRUCTURING OF THE PROBLEM Through your experience in the organization, the decision maker looked for a family of criteria to represent the reality of the approached problem. The coherence coherence, constant phase difference in two or more Waves over time. Two waves are said to be in phase if their crests and troughs meet at the same place at the same time, and the waves are out of phase if the crests of one meet the troughs of another. of this family of criteria is obtained when the basic beginnings of exhaustivity, cohesion cohesion: see adhesion and cohesion. Cohesion (physics) The tendency of atoms or molecules to coalesce into extended condensed states. This tendency is practically universal. and no-redundancy are respected also for the independence of the criteria (VINCKE, 1992). After verified ver·i·fy tr.v. ver·i·fied, ver·i·fy·ing, ver·i·fies 1. To prove the truth of by presentation of evidence or testimony; substantiate. 2. the four coherence beginnings, the considered criteria were the following ones: * Content of the tool: this criterion represents an indicator so that the functionalities of each tool are punctuated (method, treatment of the uncertainty, analysis of probabilistic (probability) probabilistic - Relating to, or governed by, probability. The behaviour of a probabilistic system cannot be predicted exactly but the probability of certain behaviours is known. Such systems may be simulated using pseudorandom numbers. dependencies, among others). * User Capabilities: this criterion represents an indicator of how the tool would be capable to simplify your own use, reducing unnecessary works in the entrance and in the manipulation of the data and information used. * Restrictions and limitations of the tool: this criterion is to represent possible disadvantages observed among two tools, for instance, maximum number of variables. * Graphic availability: this criterion synthesizes the potentialities of each tool to store and to consult important information obtained during the decision process, to support elicitation e·lic·it tr.v. e·lic·it·ed, e·lic·it·ing, e·lic·its 1. a. To bring or draw out (something latent); educe. b. To arrive at (a truth, for example) by logic. 2. process, to illustrate results and to verify (1) To prove the correctness of data. (2) In data entry operations, to compare the keystrokes of a second operator with the data entered by the first operator to ensure that the data were typed in accurately. See validate. the robustness of the model of decision graphically developed. * The price of the tools: being considered the purchase of the selected tool, it becomes important to consider the cost of obtaining this tool, to incorporate this in the decision model, making possible to analyze the price of each tool associated to the functionalities and potentialities of each one. * Potential value to be applied in the activities developed by the organization. This last criterion was related by the decision maker as a form to measure the chances of using the tool in the organization and the gain associated by using the tool. A Multiple Criteria Decision Analysis Method is used when the need of prescribing objectives to be reached on the decision exists. In that way, an application of multiple criteria decision analysis approach has as objective to establish preference relations among the alternatives that it composes the set of actions by the light of several criteria. In agreement with Vincke (1992), the main factors to be analyzed in the choice of a multiple criteria decision analysis method are: the analyzed problem, context considered at this problem, the structure of preferences of the decision making and the problematic. Analyzing the preference structure of the decision maker to this problem, it was verified that he behaves in a compensatory way, because the earnings obtained in some criteria can compensate the losses in other criteria, it is admitted for instance to acquire a more expensive tool, as long as it possesses larger potential value to the activities of the organization, larger graphic availability or larger content. Considering the context of the problem and the existent difficulty for the elicitation process used in MAUT MAUT McGill Association of University Teachers MAUT Multiple Attribute Utility Theory (Multi--Attribute Utility Theory), Keeney & Raiffa (1976), the need for using a simplified approach for this stage was verified. The SMART method (Simple Multiattribute Rating Technique) is a compensatory approach based on MAUT that presents some simplifications, which simplifies the elicitation procedure. These simplifications can result in modeling mistakes (errors), however, these are not so significant as the possible mistakes (errors) that can happen in a complex process of elicitation (EDWARDS & BARRON Barron may refer to
4. SMART METHOD SMART (Simple Multiattribute Rating Technique) is a compensatory method of multiple criteria decision analysis, developed by Edwards in 1971. This method was designed to provide a simple way to implement the beginnings of Multi--Attribute Utility Theory. Along the years, failures in this method had been identified, and they were corrected (EDWARDS and BARRON, 1994), creating the methods SMARTS and SMARTER, that present two different forms of correcting these flaws, however, SMARTS and SMARTER follow the same procedures differing in the relative procedures for the weights determination. In this work the method SMARTS will be used, however, whenever we refer to the used method we will use the name of the first method SMART that was the name in which this method became popular. The simplicity of the questions done to the decision maker and the easiness of the analysis that it is done on the answers are the great advantage of SMART. This simplicity influences directly on the understanding of the decision maker about the process used in the solution of the problem. However, this simplicity presents a cost that consists in the simplifications done for the problem, through the hypotheses of linearity of the one-dimensional one-di·men·sion·al adj. 1. Having or existing in one dimension only. 2. Lacking depth; superficial. one-dimensional Adjective 1. having one dimension 2. utility functions and of independence addictivity. Edwards & Barron (1994) questioned the possible mistakes that happened during the elicitation process, in this discussion they consider the existent trade off among the consequences of the modeling mistakes and the mistakes generated in the elicitation process. In this sense they consider the modeling mistakes as the cost of the simplifications done through the hypotheses of linearity of the one-dimensional utility functions and of addictivity. However, the consequences of the elicitation mistakes are just identified when the decision maker doesn't does·n't Contraction of does not. consider the measures obtained as a reflex of his subjective preferences. SMART doesn't request preference or indifference Indifference Antoinette, Marie (1755–1793) queen of France to whom is attributed this statement on the solution to bread famine: “Let them eat cake.” [Fr. Hist. judgment among alternatives, as it is requested in most of MAUT derived methods. Edwards affirms that hypothetical Hypothetical is an adjective, meaning of or pertaining to a hypothesis. See:
1. Contraction of do not. 2. Nonstandard Contraction of does not. n. A statement of what should not be done: a list of the dos and don'ts. represent the real preferences, and it bores decision makers non instructed, leading them to reject the elicitation process or to accept any answer with the intention of finishing the questionnaire faster. The basic idea of the measurement of the multiattribute utility is that each consequence of an action can have a value in a number of different dimensions. MAUT tries to measure these values in one dimension per time and, it proceeds for the aggregation of these values on the dimensions, through the process of the weight obtaining. The simplest and more used aggregation rule it is to take an weighed lineal That which comes in a line, particularly a direct line, as from parent to child or grandparent to grandchild. LINEAL. That which comes in a line. Lineal consanguinity is that which subsists between persons, one of whom is descended in a direct line from the other. average, in other words Adv. 1. in other words - otherwise stated; "in other words, we are broke" put differently , value =[MATHEMATICAL EXPRESSION A group of characters or symbols representing a quantity or an operation. See arithmetic expression. NOT REPRODUCIBLE re·pro·duce v. re·pro·duced, re·pro·duc·ing, re·pro·duc·es v.tr. 1. To produce a counterpart, image, or copy of. 2. Biology To generate (offspring) by sexual or asexual means. IN ASCII ASCII or American Standard Code for Information Interchange, a set of codes used to represent letters, numbers, a few symbols, and control characters. Originally designed for teletype operations, it has found wide application in computers. ], where [w.sub.k] it's it's 1. Contraction of it is. 2. Contraction of it has. See Usage Note at its. it's it is or it has it's be ~have the kth dimension weight and [s.sub.jk] it's the measure of the jth alternative at the kth dimension. The weights of this lineal addictive ad·dic·tive adj. 1. Causing or tending to cause addiction. 2. Characterized by or susceptible to addiction. addictive ( function are obtained through a weights elicitation process known as swing change process. In this swing change procedure, the decision maker is invited to consider a hypothetical situation, where an alternative that possesses the smallest punctuation punctuation [Lat.,=point], the use of special signs in writing to clarify how words are used; the term also refers to the signs themselves. In every language, besides the sounds of the words that are strung together there are other features, such as tone, accent, and in all the involved criteria at the decision process (alternative NADIR nadir (nā`dər) [Arab.,=opposite], in astronomy, the point on the celestial sphere directly opposite the zenith, i.e., directly beneath the observer. according to according to prep. 1. As stated or indicated by; on the authority of: according to historians. 2. In keeping with: according to instructions. 3. Vincke (1992)). For this fictitious Based upon a fabrication or pretense. A fictitious name is an assumed name that differs from an individual's actual name. A fictitious action is a lawsuit brought not for the adjudication of an actual controversy between the parties but merely for the purpose of alternative, the decision maker is allowed to choose one of the attributes to elevate el·e·vate tr.v. ele·vat·ed, ele·vat·ing, ele·vates 1. To move (something) to a higher place or position from a lower one; lift. 2. To increase the amplitude, intensity, or volume of. 3. the performance of this alternative for the maximum value, like this, the decision maker attributes "100 points "for the criterion that he decided to elevate first to the largest possible punctuation. After attributing "100 points" for that criterion, the decision maker eliminates the chosen criterion of the following process, repeating this procedure for it is obtained an order for the weights (GOODWIN, 2004). In the weights obtaining process, Edwards & Barron (1994) explore the notion of importance, and the natural and correct idea that in an addictive model, the weights reflect the importance of a dimension related with the others. Obviously that the importance of a criterion will depend on the evaluations extension of the alternatives for that criterion. 5. SELECTION WITH SMART The alternatives set used in this work were composed by the main available tools in the market, most of them registered by Maxwell (2004). The evaluation matrix was composed by the utility of the tools for each criterion the value of the one-dimensional utility of each alternative was elicited e·lic·it tr.v. e·lic·it·ed, e·lic·it·ing, e·lic·its 1. a. To bring or draw out (something latent); educe. b. To arrive at (a truth, for example) by logic. 2. through questionnaires. During this process it was observed that the Restrictions and Limitations criterion should be retired of the model because the existent restrictions in the alternatives were not considered significant, in other words, these constraints CONSTRAINTS - A language for solving constraints using value inference. ["CONSTRAINTS: A Language for Expressing Almost-Hierarchical Descriptions", G.J. Sussman et al, Artif Intell 14(1):1-39 (Aug 1980)]. would not make unfeasible the use of the tool, consequently the decision maker is indifferent INDIFFERENT. To have no bias nor partiality. 7 Conn. 229. A juror, an arbitrator, and a witness, ought to be indifferent, and when they are not so, they may be challenged. See 9 Conn. 42. for all alternatives in this criterion. The utility of the tools in the criteria Content of the Tool User Capabilities and Graphic Availability were obtained from questionnaires based on the subjects that accompanied the criteria elaboration stage. The utility value for each criterion is presented in the table I. The following stage of the decision support method application consists of interviewing the decision maker so that the weights of the criteria can be obtained. The SWING process used in SMART to obtain the criteria weights is divided in two parts. With the objective of ordering the criteria, it's produced an imaginary Imaginary can refer to:
The imaginary alternative presents the worst evaluations done by the decision maker in all the criteria, starting from then questionings of such type are made: If you could just choose a criterion to make a change of the worst evaluation value for the best evaluation value, which criterion would you choose aiming to improve the global value of the imaginary alternative? In that way it is made the ordering of the criteria and in the second moment the final questionings are made for the weights obtaining. These questionings evaluate in a scale from 0 to 100 points how much it is worth for the decision maker the variation of the worst possible value to the best possible value in that criterion. For the criterion that occupies the first position in the ordering, it is supposed that this variation is worth 100 points. After these questionings are made it can be determined the values of the weights as scaling constants by the normalization In relational database management, a process that breaks down data into record groups for efficient processing. There are six stages. By the third stage (third normal form), data are identified only by the key field in their record. of the obtained values in the scale from 0 to 100. The obtained ordering for the criteria was: Content of the tool ([k.sub.1]) > Potential Value ([k.sub.2]) > User Capabilities ([k.sub.3]) > Graphic Availability ([k.sub.4]) > Price ([k.sub.5]). The answers obtained for the variation value in each criterion were: * Content of the tool: 100. * Potential Value: 85. * User Capabilities: 70. * Graphic Availability: 60. * Price: 40. These values were normalized so those were obtained the following scaling constants: [k.sub.1] = 0.282; [k.sub.2] = 0.239; [k.sub.3] = 0.197; [k.sub.4] = 0.169; [k.sub.5] = 0.113. Using the utility values of the table I, the global utility of each alternative was calculated through the average meditated by the scaling constants. The table II presents the overall utility of the alternatives. For the ranking presented in the table II, we observed that alternative Equity 3.2 presents the best acting, being the suitable alternative. The sensibility sensibility /sen·si·bil·i·ty/ (sen?si-bil´i-te) susceptibility of feeling; ability to feel or perceive. deep sensibility analysis done in the parameters provokes small alterations in the ranking; however it doesn't alter the position of the first three alternatives. In that sense the sensibility analysis reinforces the recommendation of the alternative Equity 3.2. 6. CONCLUSION This work presented a multi-criteria decision model for Decision Analysis Software Selection. The decision problem was structured, settling down a group of criteria aligned with the strategies of the organization. REFERENCES: Edwards, W.; Barton BARTON, old English law. The demesne land of a manor; a farm distinct from the mansion. , F.H., "Smarts and Smarter: Improved Simple Methods for Multi Attribute Utility Measurement". Organizational Behavior and Human Decision Processes, Vol. 60:, 1994, 306-325 Goodwin, P. & Wright, G. Decision Analysis for Management Judgment John Wiley John Wiley may refer to:
Keeney, R. L.; Raiffa, H., Decision with Multiple Objectives: Preferences and Value Trade-offs, John Wiley & Sons, New York New York, state, United States New York, Middle Atlantic state of the United States. It is bordered by Vermont, Massachusetts, Connecticut, and the Atlantic Ocean (E), New Jersey and Pennsylvania (S), Lakes Erie and Ontario and the Canadian province of , 1976. Maxwell, D.T., "Decision Analysis: Aiding Insight VII", OR/MS Todau, Vol 31 (5), 2004, 44-55. Vincke, P., Multicriteria decision-aid. John Wiley, New York, 1992 Adiel Adiel (ā`dīl), in the Bible, father of David's treasurer. T. de Almeida Filho, Federal University of Pernambuco Pernambuco (pərnəmb  `k ), state (1991 pop. 7,127,855), 37,946 sq mi (98,280 sq km), NE Brazil, on the Atlantic Ocean. ,
Recife Recife (rəsē`fĭ) [Port.,=reef], city (1991 pop. 1,298,229), capital of Pernambuco state, NE Brazil, a port on the Atlantic Ocean. It is also called Pernambuco by foreigners. , Pernambuco, BRAZILCristiano A. V. Cavalcante, Federal University of Pernambuco, Recife, Pernambuco, BRAZIL Ana Paula C. S. da Costa The surname da Costa derives from the Portuguese word for coast. It may refer to:
Mr. Adiel T. de Almeida Filho earned his MSc at The Federal University of Pernambuco in Brazil. He is a research at the Institute for Research in Information and Decision Systems (ata@ufpe.br) Dr. Cristiano A. V. Cavalcante earned his PhD at The Federal University of Pernambuco, Brazil in 2005. He is a Lecturer lecturer A person who is primarily–if not entirely—involved in the teaching activities of an academic center, who is not expected to perform research or Pt management; in general, lectureships are non-tenured positions at The Federal University of Pernambuco and deals with information and decision systems (cristiano@ufpe.br). Dr. Ana Paula Cabral Seixas da Costa earned her PhD at The Federal University of Pernambuco, Brazil in 2003. She is a Lecturer at The Federal University of Pernambuco and deals with information and decision systems (apcabral@ufpe.br).
TABLE I--UNIDIMENSIONAL UTILITY
UTILITY
CONTENT OF USER GRAPHIC
ALTERNATIVES THE TOOL CAPABILITIES AVAILABILITY
@Risk 0.714 0.315 0.758
Analytica 3.0 1.000 0.500 0.435
Crystal Ball
Premium Edition 0.429 0.250 0.329
Crystal Ball
Professional Edition 0.429 0.250 0.329
Crystal Ball
Standard Edition 0.286 0.250 0.329
Dea Solver Pro 1.000 1.000 1.000
Decisionpro 4.0 1.000 0.794 0.574
Decisiontools Suite 0.714 0.315 0.758
DPL 6.0 Professional 1.000 0.500 0.574
Equity 3.2 1.000 0.500 0.435
Frontier Analyst 0.143 0.199 0.143
Hipriority 0.571 0.315 0.574
Hiview 3 0.714 0.500 0.435
JBI Javabean Decision Tree 0.286 0.125 0.189
Netica 1.000 0.315 0.329
Onbalance 0.429 0.250 0.435
Precision Tree 0.714 0.315 0.082
Risk Sim 0.286 0.100 0.108
Tree Age Pro Suite 1.000 0.315 0.435
Tree Plan 0.571 0.315 0.249
PRICE POTENTIAL
ALTERNATIVES VALUE
@Risk 0.760 0,7700
Analytica 3.0 0.600 0,8100
Crystal Ball
Premium Edition 0.430 0,9200
Crystal Ball
Professional Edition 0.430 0,9600
Crystal Ball
Standard Edition 0.680 1,0000
Dea Solver Pro 0.530 0,1900
Decisionpro 4.0 0.850 0,6900
Decisiontools Suite 0.570 0,8500
DPL 6.0 Professional 0.570 0,5800
Equity 3.2 0.050 0,4600
Frontier Analyst 0.730 0,3800
Hipriority 0.700 0.5400
Hiview 3 0.520 0.4200
JBI Javabean Decision Tree 0.650 0.0000
Netica 0.800 0.6500
Onbalance 0.850 0.6500
Precision Tree 0.800 0.5800
Risk Sim 0.980 0.5800
Tree Age Pro Suite 0.550 0.6900
Tree Plan 0.980 0.0000
TABLE II--GLOBAL UTILITY
ALTERNATIVE OVERALL UTILITY
Equity 3.2 0.797
Tree Age Pro Suite 0.754
DEA Solver Pro 0.715
DecisionTools Suite 0.680
DecisionPro 4.0 0.661
@RISK 0.658
OnBalance 0.646
JBI Javabean Decision Tree 0.645
Hiview 3 0.570
Tree Plan 0.533
HiPriority 0.528
Crystal Ball Standard Edition 0.506
DPL 6.0 Professional 0.504
Analytica 3.0 0.502
Crystal Ball Premium Edition 0.496
Analytica 3.0 0.495
Frontier Analyst 0.376
Risk Sim 0.367
Precision Tree 0.278
Crystal Ball Professional Edition 0.210
|
|
||||||||||||||||||
Printer friendly
Cite/link
Email
Feedback
Reader Opinion