CHAPTER 2 LITERATURE REVIEW
3.4 Problem Analysis and Questionnaire Survey
3.4.1 Selection of professional needs
Because too many professional needs items were collected, jewelry boutique owners and executive officers were invited to select the most critical ones and to identify the professional knowledge and skills that employees should receive training for first. A total of 32
questionnaires that divided the 15 professional needs into three characteristics: practicability, necessity, and learnability, were distributed to the experts. A 5-point Likert scale was
employed to measure the three characteristics. The scores were summed and combined into a scale indicating the sum obtained from experts’ opinions on professional needs (Table 3.5).
32
Table 3.5 Tallying expert opinions on professional needs
After obtaining the questionnaire scores, the expert weighting method was employed to analyze the 32 questionnaires to construct the relationship matrix and determine how the professional needs should be prioritized according to the jewelry boutique owners. Table 3.5 shows the relationship matrix T for each of the professional needs (derived from the data in Table 3.6), which was obtained using the expert weighting method.
course Gems Diamond color service fair Eloquence sale communicatioEnglish jewelry Psychology Fashion Precious MetaMetal Computer DMU Knowledge Grading collocation training trade training techniques skills listening design of customers boutique classification Craft Graphics
1 12 12 14 15 14 15 15 15 9 12 15 15 14 10 10
33
Table 3.6 Importance relationship matrix T obtained using the Expert weighting method
Table 3.6Importance relationship transposed matrix T obtained using the expert weighting method, as shown in Table 3.7
Table 3.7 Transposed matrix
32 32.03943279 34.82344 29.82454 33.89853 30.797478 30.83195 31.55765 37.95946 41.07551 33.22321 35.57626 34.45384 42.16252 47.0775 33.11088 32 35.55633 30.55677 34.54778 31.409982 31.66699 32.23652 38.72299 41.35257 33.96459 36.2883 35.0449 42.80945 47.57781 30.7786 30.34032357 32 28.03236 31.41084 29.040534 28.89879 29.58334 35.6655 38.01442 31.04941 33.35313 32.29953 39.61432 43.90187 35.29441 34.95415973 37.52071 32 36.54042 33.310956 33.2128 34.23135 41.27857 44.82874 36.0929 38.58689 37.31236 45.83553 51.00421 31.98154 31.3719586 33.56324 29.10116 32 30.039727 29.94633 30.65607 36.88052 39.53323 32.50367 34.94126 33.45956 40.79062 45.4737 34.13463 33.69475802 36.41442 31.21068 35.32203 32 32.25973 33.08543 39.93416 42.85501 34.80794 37.34429 36.13427 43.80812 49.1376 34.20929 33.9459485 36.3241 31.14848 35.20446 32.290443 32 32.95159 39.8174 43.04798 34.83237 37.14192 35.96523 44.05303 49.14875 33.38846 32.92132312 35.40642 30.55647 34.35474 31.544805 31.38356 32 38.87746 41.71616 33.86996 36.17443 35.08342 43.10101 48.02579 28.43137 28.08153236 30.20934 26.14738 29.23319 27.024908 26.86185 27.52414 32 35.43661 28.97604 30.99681 29.7457 36.13546 39.78627 27.44548 26.63327506 28.63752 25.1931 27.80181 25.68171 25.85507 26.32411 31.42928 32 27.40524 29.05012 28.42753 34.01111 37.30824 31.98366 31.66172716 33.90671 29.39622 33.23366 30.295638 30.26057 30.90571 37.38016 39.75974 32 34.64697 33.78279 41.1482 46.08158 30.36855 30.00990953 32.25633 27.76173 31.61441 28.688828 28.56166 29.25181 35.42796 37.28773 30.70847 32 31.72899 38.84333 43.1708 30.97314 30.45200078 32.85199 28.28919 31.83939 29.281568 29.0976 29.80769 35.6854 38.26108 31.47801 33.33525 32 39.32951 43.51491 26.15512 25.73485403 27.89308 24.09128 26.82689 24.540693 24.66898 25.32532 29.84613 31.68323 26.41914 28.14749 27.17515 32 35.63793 24.49336 23.973668 25.86658 22.46593 24.93274 23.06531 23.0208 23.65599 27.7559 29.23155 24.74809 26.326 25.31459 30.01353 32
32 33.11088356 30.7786 35.29441 31.98154 34.134632 34.20929 33.38846 28.43137 27.44548 31.98366 30.36855 30.97314 26.15512 24.49336 32.03943 32 30.34032 34.95416 31.37196 33.694758 33.94595 32.92132 28.08153 26.63328 31.66173 30.00991 30.452 25.73485 23.97367 34.82344 35.55632701 32 37.52071 33.56324 36.414419 36.3241 35.40642 30.20934 28.63752 33.90671 32.25633 32.85199 27.89308 25.86658 29.82454 30.55677101 28.03236 32 29.10116 31.210678 31.14848 30.55647 26.14738 25.1931 29.39622 27.76173 28.28919 24.09128 22.46593 33.89853 34.54778277 31.41084 36.54042 32 35.322031 35.20446 34.35474 29.23319 27.80181 33.23366 31.61441 31.83939 26.82689 24.93274 30.79748 31.40998168 29.04053 33.31096 30.03973 32 32.29044 31.54481 27.02491 25.68171 30.29564 28.68883 29.28157 24.54069 23.06531 30.83195 31.66699412 28.89879 33.2128 29.94633 32.259729 32 31.38356 26.86185 25.85507 30.26057 28.56166 29.0976 24.66898 23.0208 31.55765 32.23651904 29.58334 34.23135 30.65607 33.085426 32.95159 32 27.52414 26.32411 30.90571 29.25181 29.80769 25.32532 23.65599 37.95946 38.7229909 35.6655 41.27857 36.88052 39.934163 39.8174 38.87746 32 31.42928 37.38016 35.42796 35.6854 29.84613 27.7559 41.07551 41.35257243 38.01442 44.82874 39.53323 42.855014 43.04798 41.71616 35.43661 32 39.75974 37.28773 38.26108 31.68323 29.23155 33.22321 33.96458819 31.04941 36.0929 32.50367 34.807937 34.83237 33.86996 28.97604 27.40524 32 30.70847 31.47801 26.41914 24.74809 35.57626 36.28830336 33.35313 38.58689 34.94126 37.344286 37.14192 36.17443 30.99681 29.05012 34.64697 32 33.33525 28.14749 26.326 34.45384 35.04489954 32.29953 37.31236 33.45956 36.134271 35.96523 35.08342 29.7457 28.42753 33.78279 31.72899 32 27.17515 25.31459 42.16252 42.80945166 39.61432 45.83553 40.79062 43.808117 44.05303 43.10101 36.13546 34.01111 41.1482 38.84333 39.32951 32 30.01353 47.0775 47.57781385 43.90187 51.00421 45.4737 49.137604 49.14875 48.02579 39.78627 37.30824 46.08158 43.1708 43.51491 35.63793 32
34
From Tables 3.6 and 3.7, Table3.8 shows the order of importance for the professional needs (derived from the expert weighting method).
Table 3.8 Matrix showing professional needs, ranked by order of importance
In Table 3.8, for the matrix of the order of importance of the professional needs, the total values were acquired by calculating the sum each row. For the sequence of the second column from the right, the weights of service training, sales skills, and etiquette training were the highest, indicating the service industry’s emphasis on frontline services. Metal crafting and computer graphics were ranked 14th and 15th. These two had the lowest weights, indicating that jewelry boutique owners regarded customer service as critical under current economic conditions and with the high intensity of sales competition in business environments. Because computer graphics and metal crafting are difficult skills to acquire, learners may give up before mastering these skills, which is a waste of training resources and time. Therefore, the experts did not have a positive attitude toward learning these two skills.
3.4.2 Academic Course Design
Fullan (1993) indicated that transforming education is critical and indicated that experts and teachers must take action to carry out education transformation. However, scholars
1 0.9676 1.1314 0.8450 1.0599 0.9022 0.9013 0.9452 1.3351 1.4966 1.0388 1.1715 1.1124 1.6120 1.9221 17.4411 0.0736 0.8449 6 Gems Knowledge 1.0334 1 1.1719 0.8742 1.1012 0.9322 0.9329 0.9792 1.3789 1.5527 1.0727 1.2092 1.1508 1.6635 1.9846 18.0375 0.0761 0.8738 5 Diamond Grading 0.8838 0.8533 1 0.7471 0.9359 0.7975 0.7956 0.8355 1.1806 1.3274 0.9157 1.0340 0.9832 1.4202 1.6972 15.4072 0.0650 0.7464 10 Clor matching 1.1834 1.1439 1.3385 1 1.2556 1.0673 1.0663 1.1203 1.5787 1.7794 1.2278 1.3899 1.3190 1.9026 2.2703 20.6429 0.0871 1.0000 1 Service training 0.9434 0.9081 1.0685 0.7964 1 0.8505 0.8506 0.8923 1.2616 1.4220 0.9780 1.1052 1.0509 1.5205 1.8239 16.4720 0.0695 0.7979 8 Fair trade act 1.1084 1.0727 1.2539 0.9369 1.1758 1 0.9990 1.0488 1.4777 1.6687 1.1489 1.3017 1.2340 1.7851 2.1304 19.3422 0.0816 0.9370 3 Etiquette training 1.1095 1.0720 1.2569 0.9378 1.1756 1.0010 1 1.0500 1.4823 1.6650 1.1511 1.3004 1.2360 1.7858 2.1350 19.3583 0.0816 0.9378 2 sales skills 1.0580 1.0212 1.1968 0.8926 1.1207 0.9534 0.9524 1 1.4125 1.5847 1.0959 1.2367 1.1770 1.7019 2.0302 18.4341 0.0777 0.8930 4 Communication skills 0.7490 0.7252 0.8470 0.6334 0.7926 0.6767 0.6746 0.7080 1 1.1275 0.7752 0.8749 0.8336 1.2107 1.4334 13.0619 0.0551 0.6328 12 English listening 0.6682 0.6441 0.7533 0.5620 0.7033 0.5993 0.6006 0.6310 0.8869 1 0.6893 0.7791 0.7430 1.0735 1.2763 11.6097 0.0490 0.5624 13 Jewelry design 0.9627 0.9322 1.0920 0.8145 1.0225 0.8704 0.8687 0.9125 1.2900 1.4508 1 1.1283 1.0732 1.5575 1.8620 16.8373 0.0710 0.8156 7 Consumer psychology 0.8536 0.8270 0.9671 0.7195 0.9048 0.7682 0.7690 0.8086 1.1430 1.2836 0.8863 1 0.9518 1.3800 1.6399 14.9023 0.0629 0.7219 11 Conspectus of Fashion boutique 0.8990 0.8689 1.0171 0.7582 0.9516 0.8104 0.8090 0.8496 1.1997 1.3459 0.9318 1.0506 1 1.4473 1.7190 15.6580 0.0660 0.7585 9 precious metal classification 0.6203 0.6011 0.7041 0.5256 0.6577 0.5602 0.5600 0.5876 0.8260 0.9316 0.6420 0.7246 0.6910 1 1.1874 10.8192 0.0456 0.5241 14 Metal Craft 0.5203 0.5039 0.5892 0.4405 0.5483 0.4694 0.4684 0.4926 0.6976 0.7835 0.5370 0.6098 0.5817 0.8422 1 9.0844 0.0383 0.4401 15 Computer Graphics
237.108 1.0000
35
generally believe that their field of specialization is the most important. To prevent course planning from becoming overly subjective, resulting in such consequences of too many courses being introduced and employees being unable learn the knowledge and skills they require within a reasonable time, only the most relevant 14 courses for the jewelry boutique industry were selected (Table 3.9).
Table 3.9 Academic Course Design
Given adequate time for training, operators would willingly accept the course planning suggested by the academics. However, if the time for training employees is limited, then the owners are suggested to follow the CCR model (described in the following section) to select the acceptable courses with most importance.
Courses Courses
1 Gemology 8 Marketing strategies
2 The study of diamond 9 English conversation
3 Chromatics 10 Digital jewelry design
4 Quality management 11 Comsumer behavior 5 Operations management 12 Brand management 6 Interpersonal relationships 13 Materials science
7 Management 14 Basic jewelry inlay.
36
CHAPTER 4 EXPERT WEIGHTING METHODON PROFESSIONAL NEEDSAND CCR MODELON
ACADEMIC COURSE DESIGN
The order of importance of professional needs was evaluated according to expert
opinions. Depending on the professional needs, different courses provide different advantages.
Therefore, any evaluation should not be based solely on expert opinions; instead, the
CCR-input orientation model should be used to evaluate each course regarding its suitability for the professional needs.
4.1 CCR input-oriented model
To employ the CTC to minimize the gap between the industry and academia, this study selected the eight professional needs with the highest weights as the output variables from the ranked professional needs (Table 3.8). From the course design for academic teaching (Table 3.9), the 14 courses were DMUs and the CCR-input orientation model was employed. In addition, by incorporating the regulation for each weight of the variable emphasized in this study, the stochastic data envelopment analysis (SDEA) proposed by Hadad et al.(2008) was employed to obtain the minimum threshold of each weight as an objective efficiency value, and did not overemphasize the importance of other courses.
4.2 Guidelines for variable weight in the CCR efficiency evaluation model
Through the two-stage analysis and evaluation of professional needs and course design, the planned course design was optimized to correspond to industry needs; in addition, within
37
the time and energy constraints for learning, the most appropriate courses for training employees in the jewelry boutique industry were designed.
In this study, the CCR model was employed to evaluate the efficiency of each course design; specifically, each course design was viewed as a DMU, and the efficiency value determined the order in which courses be prioritized in the future. However, the greatest feature of adopting the CCR model for efficiency evaluation is that the greatest advantage of each DMU can be identified; however, a great disadvantage of applying this model is that an unapparent discrimination of the efficiency evaluation may frequently occur. The main reason for this unapparent discrimination of the efficiency evaluation is that there is no restriction on the weight assigned to each variable in each efficiency evaluation model; in other words, the efficiency evaluation model does not account for the limitations of the weight of each variable.
Hadad.et.al (2008) proposed an SDEA model to test the hypothesis that n DMUs existed and that each DMU has m input variables (j = 1,…,m) and s output variables (r = 1,…, s). The following criteria was proposed according to the model on the weight of each variable:
The two constraints of and Indicate the minimum threshold for the weight of each variable.
This study employed the lower limit of weight in the SDEA model to restrict the mutual weights to ensure an objective efficiency evaluation of each course. Consequently, the
CCR-input orientation model for efficiency evaluation was modified by incorporating (1) into (3):
38
Where the target of (3) was the efficiency value (θ) of each of the evaluated course.
4.2.1 Importance of academic course planning on professional needs
According to the professional needs of the jewelry boutique industry, six experts were invited to evaluate the importance of each course suggested by the academics. A 5-point Likert scale was employed to measure the questions. Table 4.1 shows the total scores of the experts.
Table 4.1 Importance of academic course planning on professional needs
Table 4.1 lists the DMUs in the first column and the following eight output variables across the top row: professional knowledge on gems, diamond grading, consumer psychology, service training, fair trade act, etiquette training, sales skills, and communications skill.
4.2.2 Training course ranking as determined by the academic and industry experts
Gems Diamond Consumer psychology Service Fair Etiquette Sales Communication Knowledge Grading of customers training trade act training skills skills
Gemology 25 23 19 15 19 11 17 19
The study of diamond 25 25 18 15 20 12 18 18
Chromatics 23 21 20 18 18 15 20 17
Quality management 22 22 19 17 20 13 17 18
Operations management 21 19 17 18 17 14 13 15
Interpersonal relationship 18 18 24 23 20 23 23 25
Management 19 19 20 19 19 17 20 20
Marketing strategies 24 23 24 23 22 22 24 23
English conversation 17 14 14 18 15 18 21 22
Digital jewelry design 21 21 21 21 20 20 22 21
Consumer behavior 15 15 15 14 13 12 18 16
Brand management 16 16 20 20 17 14 23 23
Materials science 19 18 19 15 16 9 16 15
basic jewelry inlay 24 22 21 15 10 10 10 10
39
After the data in Table 4.1 were input into the CCR-input orientation model in (3), Table 4.2 shows the efficiency value of importance, denoted as θ, and the order of importance for the training courses recommended by the academics and industry:
Table 4.2 CCR input-oriented efficiency values and ranking
According to Table 4.2, marketing strategies, interpersonal relationship, digital jewelry design, the study of diamond, and chromatics were the five highest priority courses
considered by training units and the academia and under conditions of time constraints. The other low-priority courses can be arranged as elective courses according to the requests of students and needs of jewelry boutique owners.
Conventionally, sales personnel are the frontline of business marketing in the jewelry boutique industry, and therefore they must possess a high level of professional knowledge on gems, and receptionists must be capable of drawing designs. Consequently, over the past several decades, curricula on gemstones worldwide have been focused mostly on teaching knowledge about gemstones. However, the major focus of customer relations is elevating
Course θ Sequence
Gemology 0.829929 8
The study of diamond 0.850512 4
Chromatics 0.839033 5
40
service quality and catering to customer needs in order to understand how their needs can be satisfied. This business model provides a path that promotes firm development and success (Stank et al.,1999).
4.3 The Equivalence Relationship of Two Common Weights Models in Date Envelopment Analysis
(Bao et al., 2015) Point out Data envelopment analysis (DEA) originated in economic theories. Previously, the majority of scholars estimated production possibility curves by using preset production functions to evaluate the productivity of various organizations. Farrell (1957) adopted a mathematical programming approach and substituted the commonly used
“preset production functions” with “non-preset production functions” to determine efficient frontier curves. Thereafter, Farrell (1957) then used these curves to evaluate the technology and price efficiency of various decision making units (DMUs). However, the model proposed by Farrell (1957) could only evaluate the efficiency of a single output and was incapable of concurrently evaluating the efficiency of multiple outputs, limiting its scope of application.
Subsequently, Charnes, Cooper, and Rhodes (1978) referenced Farrell’s efficiency
measurement method and developed an efficiency evaluation approach that could measure multiple inputs and multiple outputs under a fixed return. This approach was referred to as data envelopment analysis (DEA). The model that was used to measure the efficiency of fixed returns to scale was termed the CCR model.
The efficient and inefficient units among DMUs can be distinguished by measuring their efficiency through DEA. However, any external interference during the evaluation process may affect the evaluation results. The problem of sensitivity has often caused the
overvaluation of “efficient” DMUs in efficiency evaluation cases. This is the primary cause of
41
“inefficient” evaluations. Therefore, numerous studies have attempted to evaluate the efficiency scores of “efficient” DMUs through extensive investigations.
“Inefficient” evaluations mentioned in the preceding paragraph are often attributed to unrestricted variable weight, a problem commonly perceived by academic researchers.
Thannassoulis (1988) as well as Wong and Beasley (1990) asserted that unrestricted variable weight often causes DMUs to manifest efficiency scores that exceed their actual performance.
Without appropriate restriction, DEA efficiency evaluations often produce DMUs with an efficiency score of “1.” These results are the primary cause of “no significant” evaluation effectiveness.
Common DEA models for evaluating the efficiency of various DMUs maintain that the
“weight” estimations generated by the models most benefit the DMU efficiency evaluations.
Therefore, these models often produce several “efficient” units, reducing the accuracy of the evaluations. To enhance identification accuracy, many scholars began evaluating the
efficiency scores of DMUs by using “common weights.” A feature of this model is that the same weight is used for the efficiency evaluation of all DMUs. In other words, under a mutual evaluation standard for all DMUs, the occurrence of similar efficiency scores is greatly
reduced.
The concept of common weights was first introduced by Cook, Roll, and Kazakov in 1989. Cook et al. (1991) aimed to minimize the distance between the upper and lower limits of various weights to evaluate the common weights of different variables. Moreover, Roll et al. (1991) adopted common weights to evaluate the efficiency scores of DMUs and their nearest targets. Ganley and Gubbin (1992) developed common weights by using the sum of maximum efficiency ratios to distinguish the classes of various DMUs. Furthermore,
Sinuany-Stern (1994) employed linear discriminant analysis as the basis for common weights
42
evaluation to distinguish between efficient and inefficient DMUs. Additionally, Sinuany-Stern (1998) employed common weights and developed a novel method,
Discriminant Data Envelopment Analysis of Ratios (DR/DEA), to determine the optimal ranking of DMUs. Liu and Peng (2004a) proposed the Common Weights Analysis (CWA) Model by assuming that DMUs are all equally crucial and used this model to evaluate the common weights of different variables. Thereafter, Liu and Peng (2004b) proposed the Super CWA Model by assuming that the ranks of all DMUs are the same and used this model to evaluate the common weights of different variables. Kao and Hung (2005) developed the Compromised Weights Model to evaluate the financial performance of DMUs under a
common standard. Bao et al. (2011) introduced the concept of common weights compromise, which accounts for the compromise of each DMU variable. The aforementioned common weights methods have different degrees of managerial implication. However, most of the aforementioned common weights models (excluding the model proposed by Bao et al. (2011) are based on overall standards and fail to evaluate the efficiency of individual DMUs. In response to this problem, many scholars began researching efficiency evaluation models focusing on common weights. Makui et al. (2008) contended that the weight scores for variables evaluated by the CCR model differ significantly from the weight scores for the variables of different DMUs, demonstrating an irrational phenomenon, which prompted Makui et al. to propose a common weights approach. Thereafter, Makui et al. (2008) and Yang et al. (2010) further measured common weights by using multi objective programs.
Amirteimoori et al. (2009) developed a “three-stage approach” to evaluate common weights in order to more effectively rank the efficiency scores of various DMUs. Kao (2010) asserted that the evaluation outcomes of the CCR model differ from one another. Therefore, Kao proposed the concept of “absolute distance” to evaluate common weights. However, this
43
approach required linear programming to identify a model solution, thus increasing the difficulty of solving problems.
"Common weights” efficiency evaluations not only reinforce identification, they are also more effective for evaluating the overall efficiency of DMUs. Thus, Bao (2010) proposed the common weights model, and Bao et al. (2012) adopted DEA compromised weights to
examine common weights. Among the various common weights models, the model proposed by Bao (2008) is the most simple and practical. This model simplifies the solution
determination process without compromising rationality. Notably, the present study found that the model proposed by Bao (2008) manifests an equivalence association with the common weights model proposed by Karsak and Ahiska (2005). In other words, the present study verified through empirical analysis that, even though the external forms of the two models are distinct, the characteristics of and results obtained from the two models are similar.
This study primarily aimed to verify that the common weights model proposed by Bao (2008) manifests an equivalence association with that proposed by Alinezhad et al. (2009).
4.3.1The Common Weights Models of Bao (2008).
Bao’s model is presented as follows:
Where x represents the only input variable, v represents the weight setting of x, represents the weights of the various output variables ( ), and r =1,..s.
4.3.2The Common Weights Models of Alinezhad et al. (2009).
Alinezhad et al.(2009) proposed the minimax efficiency model to measure common weights. The model is expressed as follows:
44
Where x is the only permitted input variable, and represents the weights of This study employed (1) to analyze the common weights efficiency evaluation model of (2), verifying that (2) manifested an equivalence relationship with (1).
Prior to theorem verification, this study developed Lemma 1 to serve as an explanation for the verification of (1).
Lemma 1:
When (1) is applied to process the input and output variables of various DMUs, and when the variables are multiplied by k (k ≠ 0), the latter slack variables obtained using (1) are k-times the former slack variables, and the various weight scores remain unchanged.
Proof:
When the input and output variables of the various DMUs are multiplied by k (k≠0),let ∗ and ∗ . Subsequently, (1) becomes (3)
The constraint of (3) can be expressed as ∗ ∑ 0.This suggests that
′ ∗ and that the scores of and v remain unchanged. Thus, Lemma 1 is verified.
45
Theorem 1:
Thecommon weights model of Alinezhad et al. (2009) ((2)) manifests an equivalence relationship with (1).
Proof:
In (1), the input variables ( ) can be categorized into the following two conditions:
1; j = 1,…,n;
0 1; j = 1,…,nfor specific DMUs. The two conditions are discussed separately.
1; j = 1,…,n;
For(1),let ∗ ⁄ , ∗ / . Subsequently, (1) can be rewritten as follows:
Let ∑ ∗, ∗ ∗. Inevitably, ∑ ∗⁄ establishes 1 . Thus, in (2) is established.
If the first constraint of (4) is ∑ ∗ 0, then (4) can be rewritten as follows:
∑ 0or
∑ / 1, which is the second constraint of (2).
0 1; j 1,…,n.
Therefore, under similar constraint conditions, (2) determines min ∑ ∗, thus confirming that the common weights models of (1) and (2) equal.
When 1, the constraint ∑ ∗ ∗/ is not necessarily established. According to Lemma 1, when multiplying by constant k to obtain 1 and multiplying by constant k, the of (2) remain unchanged, suggesting that the sum of the output weights of the various DMUs and the efficiency ratio of the input variable multiplied by weight remain
46
unchanged. At this point, the various constraint conditions can be satisfied, verifying that the common weights model of (2) is similar to that of (1), which is proposed in the present study.
Many studies have focused on investigating DEA common weights models. These models differ from one another and are developed according to different arguments. The common weights model proposed by Bao (2008) is the most simple and practical of the common weights models. Although this model is simple, this study verified that this model ((1)) manifests an equivalence relationship with the common weights model proposed by Alinezhad et al. (2009; (2)), presenting an additional theoretical basis to support this model.
However, any external interference during the evaluation may affect the original evaluation results. Although, Alinezhad et al. (2009) the minimax efficiency model ((2)) manifested an equivalence relationship with the common weights model ((1)) proposed in the present study, the minimax efficiency model can only process a single input variable, which is a limitation of this model. Nevertheless, expanding (1) by referencing the study by Bao (2015), which established a model that could process multiple input variables, enables the mentioned limitation to be eliminated. This finding serves as a contribution of the present study (Bao et al., 2015).
47
CHAPTER 5 CONCLUSION AND RECOMMEDATION
CTC-based course training and planning has seldom been applied in the jewelry boutique industry. To date, most jewelry boutique owners select courses offered by training centers to train their employees according to their subjective personal experiences and needs. In addition, the training units and instructors at schools commonly offer courses that are determined based on the opinions of course designers that their field of specialization is the most important one.
Occasionally, some courses are offered in accordance to the preferences of jewelry boutique owners. Given these conditions, there is frequently a sizeable gap between the needs of professional and design of courses.
In this study, an expert weighting method and DEA were employed. The two-stage evaluation model revealed that jewelry boutique owners emphasized the importance of whether employees can demonstrate a high level of knowledge when providing customer service at the frontline, which can be acquired through courses such as service training and sales skills. Moreover, the three courses with the least importance are jewelry design, metal crafting, and computer graphics. According to the employment of CCR-input orientation model on the professional needs required in the industry and the course planning by the academia, the highest priority courses are marketing strategies and interpersonal relationships, which can be incorporated into service training and sales skills. Furthermore this article has developed “The Equivalence Relationship of Two Common Weights Models in Date Envelopment Analysis”. Make assessment more effective. And these results accord with industry expectations.
However, digital jewelry design was unexpectedly ranked third, perhaps because the academic experts regarded digital design as an imperative skill for relevant business