The CSF-Alternatives Hasil Analisis dan Diskusi

Bab 4: Hasil dan Pembahasan 48

4.7.2. The CSF-Alternatives

One hundred and seven 107 responses were collected from project professionals through the questionnaire survey. They have been analysed for the relative importance of every factor of CSF-alternatives related to each element of the evaluation-criteria. Accumulated scores of forty variables CSFs were weighted into each related factor of eight CSF-alternatives in respect of four evaluation-criteria. The weights of each factor of eight CSF-alternatives were used to develop pair-wise comparisons of matrix algebra under four evaluation-criteria. They produced four different matrix algebras. These four different matrices are: the matrix algebra of eight CSF-alternatives under Project Type, under Project Phase, Project Monitoring and Project Deliveries. Similar procedures of matrix algebra analysis for the relative importance of evaluation-criteria discussed in Section 4.4 are applicable to CSF-alternatives for squaring, iterating, and comparing the last two iterated matrices, until the relative ranking of normalised eigenvectors provides none or minimum changes to determine the best results of the eigenvector solution. The best eigenvector solution of eight CSF-alternatives under the first evaluation-criteria Project Type was identified as the relative ranking of important CSFs and shown in Table 4-7. The provisional ranking of eight CSF-alternatives was attributed as: MARCON is placed on the second ranking with 14.81 eigenvector, DEVFOC fifth with 12.01, INVTEC seventh with 11.13, LOCRES sixth with 11.20, INTBEN eighth with 10.07, PROCOM third with 13.16, ROPRAC fourth with 12.08, and METOOL was first with 15.53 of this particular eigenvector solution. The other remaining three of the best eigenvector solutions and their provisional ranking positions of eight CSF-alternatives under the Project Phase, Project Monitoring and Project Deliveries are summarised in the left hand side matrix of Table 4-8. Consistency Ratio CR of matrices of CSF-alternatives under the four evaluation- criteria are mostly found very close to zero refer to the bottom row of the left hand side matrix of Table 4-8. The matrices of CSF-alternatives calculated under the Project Type represent a consistency ratio of CR= 0.019 10 ; Project Phase CR= 0.021 10; Project Monitoring CR= 0.014 10 ; and Project Deliveries CR= 0.022 10 . These Bab 4: Hasil dan Pembahasan 49 consistency ratios indicate that, the questionnaire survey was designed to observe several factors of CSF-alternatives using large numbers of variables to expect and collect greater responses, and consequently, more responses with detailed and actual weights are received. However, the ranking importance of eigenvector solutions for the priority CSFs in Table 4-8 would seem to be complicated and slightly inconsistent between each one. The ranking importance of CSF-alternatives under Project Type and Project Phase is different to the ranking of Project Monitoring and Project Deliveries. For example: LOCRES is positioned at 6 th under Project Type and project Phase, while at 7 th under Project Monitoring and Project Deliveries. ROPRAC is placed at 4 th ranking under Project Type and Project Phase, while at 5 th under Project Monitoring and 4 th under Project Deliveries. Fortunately, MARCON always appears in 2 nd ranking, and so on. This complex multi- criteria decision making was resolved through an AHP solution tree.

4.7.3. The AHP Solution Tree