Mokhtar Abdullah

Evaluating functional relationship between image, customer satisfaction and customer loyalty using general maximum entropy

Much has been written in the past few years on customer satisfaction. Traditional approaches to customer satisfaction, however, have been too narrow in scope in the way satisfaction is measured. Most instruments to measure customer satisfaction are not sensitive to the fact that a customer’s voice has a wide spectrum and goes beyond the single note of satisfaction or happiness. Most approaches fail to link customer satisfaction to drivers of loyalty and market share and to processes that contribute to pro® tability. Traditional approaches to customer satisfaction also do not oVer a comprehensive way to measure the range of perceptions that in ̄ uence customer behaviour. Organizations that want to improve quality signi® cantly by aligning internal processes to the voice of the customer must look beyond customer satisfaction to market-oriented measures for gauging progress. This requires shifting attention away from customer satisfaction to the more comprehensive concept of customer loyalty. Loyal customers stick with their suppliers or service providers over the long run. They also express their loyalty by giving a greater share of their wallets to their high-value brands or service providers and by generating word-of-mouth referrals. All of these behaviours will directly aVect pro® tability. But it takes a lot more than customer satisfaction to make a customer loyal and to make internal processes truly responsive to customer needs, market share opportunities and pro® t objectives. Recent studies have demonstrated mixed results in analysing the relationship between satisfaction and loyalty. Several studies have indeed found satisfaction to be a (and often the) leading factor in determining loyalty (e.g. Anderson & Fornell, 1994; Oliver & Linda, 1981; Pritchard, 1991). Other studies, however, suggest satis® ed customers may not be suYcient to create loyal customers (e.g. Cronin & Taylor, 1992; Fornell, 1992; Oliva et al., 1992). These studies tend to support Reichheld’s (1993) argument that customer satisfaction is not a surrogate for customer loyalty, and thus increasing customer satisfaction does not necessarily lead to increased customer loyalty to a product, service or organization. To gain further insight into the development of customer loyalty toward a service or product, this study looks

Evaluating public sector efficiency with data envelopment analysis (DEA): A case study in Road Transport Department, Selangor, Malaysia

It is the responsibility of most organizations in the public sector to allocate, monitor and evaluate their annual expenditure and service delivery. However, over the years, there have been no proper system and procedures to address the question of overall eYciency within an organization unit or between organizations or sectors. A number of performance measures were used in the public sector, with the most prominent being the centrally evaluated set of cost-weighted activity indices. This paper presents a comprehensive measure of how well resources are combined and utilized in an organization in order to accomplish the most optimized set of results. A technique known as data envelope analysis (DEA) is used. The measure of an organization’s productivity in the public sector is actually similar to the measure of a system’s eYciency. It can be de® ned as a ratio of output to input. By using DEA one will be able to

The Malaysian Total Performance Excellence Model: A conceptual framework

The growing interest in total performance excellence measurement (TPEM) has led to the emergence of a number of frameworks such as the European Quality Awards (EQA) model, Malcolm Baldrige, Kanjis model and several others. However, these models are based on enablers that are restricted to quality management perspectives. There are other enablers that are deemed to affect performance, as found in other studies outside the total quality management framework (see Flynn et al., 1994, Journal of Operation Management, 11, pp. 339–366; Larrache (2000), http://www.insead.fr/Research/compfit). This paper presents a new framework for TPEM for organizations in Malaysia. The framework consists of a set of dimensions that focuses on, among other things, organizational values and culture, change management, best practices, innovation and productivity. Other dimensions, typical to those in the EQA, Malcolm Baldrige and Kanjis models, are also incorporated in the TPEM framework. These include leadership, policy and str...

On service improvement capacity index: A case study of the public service sector in Malaysia

The ultimate goal of the total quality management (TQM) eVort in service industries is customer satisfaction. Measuring service quality of an organization that focuses on customers and satisfying customers’ requirements is thus very vital. Service features that conform to the needs of the customers should be designed and concentrated upon. Through customer surveys, information was obtained that was able to indicate to a certain extent whether ® xed quality standards are actually adequate. The public service sector, being a service provider to the public, should not be immune from pressure that drives an organization to be successful with quality services that satisfy their customers. According to Randall and Senior (1994), public sector services are experiencing pressures as a result of ® nancial constraints, legislative changes, criticism of standards and political tension. Together with internal pressure, the desire for improved quality of service has increased. Thus, service quality initiatives in the public service, although funded through budget allocation resources, need to focus and contribute to social good. There are two types of customers to the public services according to Donnelly et al. (1995): the recipients or users of services which make either little or no ® nancial contribution towards their provision; or those who have to pay for a particular public service but do not experience its bene® t through direct use. To improve the service of the public sector, methods for evaluating customer services need to be adapted suitably. An approach for assessing the quality of service is through customer satisfaction surveys, which articulate customers’ perceptions of the service delivery and expectation of service quality. This is particularly important since customers’ feedback is very useful to determine which feature in the service needs to be improved. According to Curry and Herbert (1998), Ovretveit (1991) took into account the three quality categories in public service organizations, which consist of

Benchmarking institutions of higher education

Market growth in the new millenium is expected to expand and spread world-wide. An organization therefore has to be aggressive and competitive to establish its position domestically and internationally by being staVed with the best employees. As a feeder, Institution of Higher Educations (IHEs) need to produce employees that are able to respond to the demands and to give con® dence to society. The best employees will be produced through good students and therefore the competitiveness of IHEs in attracting good students is an important issue. IHEs should therefore change to meet the needs by improving their quality in every aspect by exposing students to the best training in working skills and ethics and being capable of critical thinking (Albino, 1993). These can be achived quickly by emulating the best practices of world-class IHEs that have already proven their best performance. Adoption of best practices is known as benchmarking, which was de® ned by Camp (1989), Zairi (1994), Fram and Camp (1995), Cook (1995) and Murphy (1995) as a systematic and continuous process to identify, determine, measure, compare, learn, adopt and implement the best practice obtained through internal and external evaluation of an organization so that performance of a higher standard can be achieved and improved. Benchmarking helps IHEs to understand their strengths and weaknesses so that quality improvement can be implemented eVectively. The use of benchmarking will help speed up the improvement of best practices and performance of IHEs (Voss, 1997). The direction of IHEs as manifested in their mission statements and supported by their objectives must be dealt with in a proper manner and with good care in striving for excellence. The presence of the critical success factors (CSFs) and best practices speeds up IHEs in achieving excellence. The existence of similarities between IHEs in implementing its missions and objectives through CSFs and best practices would enable them to produce the best performance. This paper explores the existence of similarities and attempts to compare similarities between IHEs through their missions, objectives, CSFs and best practices as measurement dimensions. The similarity measures will result in, as well as encourage, the practice of benchmarking in IHEs.

Detection of influential observations in functional errors-in-variables model

This paper presents several methods for detecting influential observations in a functional errors-in-variables (FEIV) model. Since the FEIV model is a regression-like model, similar approaches to those in regression model can be adopted. We focus on influence diagnostics based on leverage values, influence curve and case deletion approaches. Some numerical examples are given to illustrate the methods.

A NEW ROBUST BIVARIATE CONTROL CHART FOR LOCATION

This paper has developed a new robust Shewhart-type control chart for monitoring the location of a bivariate process and examine its behavior based on the Hodges–Lehamnn and Shamos–Bickel–Lehmann estimators. A numerical example is given to illustrate the use of the proposed method. Its performance is investigated using a simulation study.

Box-chart: Combining x and S charts

Among the techniques that form the core of statistical process control (SPC), control charts are perhaps the most important and widely used tools. First developed by Shewhart (1925), the use of control charts has become standard practice in industrial applications. Although over the years other control charts have been developed, the variables control charts, xÅ and R (or xÅ and S), remain the most popular. These charts are used to monitor simultaneously both the process mean and process variability. The xÅ chart monitors changes in location, while the R (or S) chart monitors changes in spread or dispersion. The xÅ and R chart can identify special causes of variation while using relatively small sample sizes. Though the xÅ and R charts are easy to construct and easy to interpret, at times they `over react’ to variations. The xÅ and S charts, on the other hand, work well for larger sample sizes and estimate the variation more eYciently (Gitlow, 1995). Recent developments deviate from early ones, most notably on the emphasis placed on target values on simultaneously monitoring the process mean and process variability (Yeh & Lin, 1999). Control charts based on these target values help determine whether the existing process is capable of meeting the desirable standards. Furthermore, they also help management set realistic goals for the existing process. One must be cautious when interpreting control charts based on target values. Sample observations can still fall outside the control limits even though no special causes are present in the process, since the desirable standards may not be consistent with the process conditions. Time and resources might be wasted looking for special causes that actually do not exist. The focus of this paper is to develop new variable control charts which will maintain the ability to monitor simultaneously, on a single chart, the process mean and the process variability. First, the Shewhart xÅ and S charts are constructed using target values for a univariate process (Mitra, 1993). Next, the same data are used to construct the proposed chartÐ the box-chart. Then the performance of the box-chart is compared with the Shewhart xÅ and S charts.

NEW ROBUST STATISTICAL PROCESS CONTROL CHART FOR LOCATION

The application of robust methods in construction of statistical quality control charts is demonstrated. Two estimators that are more robust than the sample mean and sample standard deviation are introduced. They are the Hodges-Lehmann and the Shamos-..

Robust estimation of a linear functional relationship

This paper presents the M—estimators of a simple linear functional relationship (SLFR). A compuyational scheme called the “Iteratively Reweighted Functional Relationship” (IRFR) method used to compute the robust estimates. As an iterative procedure the IRFR requires good starting values. A modified Ll—norm criterion is proposed for the starting values and its improvement over the classical maximum likelihood SLFR estimates is examined. The difficulty in dealing with a possible outlier in the SLFR is indeterminacy in the direction of the outlier. In this study we investigate the extent to which the effect of the indeterminacy problem may have on the robust estimates.