Emerald | Journal of Modelling in Management

Effects of Overconfidence and Dependence on Aggregated Probability Judgments

Shi-Woei Lin, Ssu-Wei Huang — 2012-03-16 00:00:00.0

Abstract

Purpose - We investigate how expert overconfidence and dependence affect the calibration of aggregated probability judgments obtained by various linear opinion-pooling models.Design/methodology/approach - We used a large database containing real-world expert judgments, and adopted the leave-one-out cross-validation technique to test the calibration of aggregated judgments obtained by Cooke’s classical model, the equal-weight linear pooling method, and the best-expert approach. Additionally, we rigorously tested the significance of the effects using linear models. Findings - We found significant differences between methods. Both linear- pooling aggregation approaches significantly outperformed the best- expert technique, indicating the need for inputs from multiple experts. The significant overconfidence effect suggests that linear pooling approaches do not effectively counteract the effect of expert overconfidence. Furthermore, the second-order interaction between aggregation method and expert dependence shows that Cooke’s classical model is more sensitive to expert dependence than equal weights., with hHigh dependence generally leadings to much poorer aggregated results;. By contrast, tThe eEqual- weight approach is more robust under different dependence levels.Research limitations/implications - Our results suggest that methods involving broadening of subjective confidence intervals or distributions may occasionally be useful for mitigating the overconfidence problem. An equal-weight approach might be more favorable when the level of dependence between experts is high. Although we found that the number of experts and the number of seed questions also significantly affect the calibration of the aggregated distribution, further research to find the minimum number of questions or experts required to ensure satisfactory aggregated performance would be desirable. Furthermore, other metrics or probability scoring rules should be used to check the robustness and generalizability of our conclusion.Originality/value - Provides empirical evidence of critical factors affecting the calibration of the aggregated intervals or distribution judgments obtained by linear opinion-pooling methods.

Improving inventory performance with clustering based demand forecasts

PRADIP KUMAR BALA — 2012-03-16 00:00:00.0

Abstract

Purpose - Purpose – The paper aims at developing forecasting model for retailers based on the customer segmentation to improve performance of inventory. Design/methodology/approach - Design/methodology/approach – The research makes an attempt to capture the knowledge of segmenting the customers based on various attributes as an input to the demand forecasting in a retail store. The paper suggests a data mining model which has been used for forecasting of demand. The proposed model has been applied for forecasting demands of eight SKUs for grocery items in a supermarket. Based on the proposed forecasting model, the inventory performance has been studied with simulation. Findings - Findings – The proposed forecasting model with the inventory replenishment system results in the reduction of inventory level and increase in customer service level. Hence, the proposed model in the paper results in improved performance of inventory. Practical implications - Practical implications – Retailers can make use of the proposed model for demand forecasting of various items to improve the inventory performance and profitability of operations. Originality/value - Originality/value – With the advent of data mining systems which have given rise to the use of business intelligence in various domains, the current paper addresses one of the most pressing issues in retail management as demand forecasting with minimum error is the key to success in inventory and supply chain management. The proposed forecasting model with the inventory replenishment system results in the reduction of inventory level and increase in customer service level. The proposed model outperforms other widely used existing models.

The Balanced Scorecard in Health Care: a Multilevel Latent Variable approach

Pietro Giorgio Lovaglio, Giorgio Vittadini — 2012-03-16 00:00:00.0

Abstract

Purpose - The aim of this article is to propose a practical conceptualization of the Balanced Scorecard to describe the mechanism producing creation of monetary value for Hospitals in the territorial context of Lombardy region (Italy). Design/methodology/approach - We propose a model building strategy that assigns Key indicators to Key performance areas, and identifies causal relationships between Key performance areas. Second, we utilize a suitable statistical approach to estimate causal relationships among involved latent variables taking into account the hierarchical structure of data. Utilizing a suitable data decomposition, the causal model is applied separately to the within data (Hospitals) and to the between data (Local Health Agencies).Findings - In the measurement model we found a new latent construct (Quality of the medical care) that resumes the amount of formal training and the performance of surgical staff in Hospitals. The estimated causal models reflect usual directional assumptions, supposed in a typical BSC causal scheme, with some differences. For Local Health Agencies, fruits (financial measures) are strongly related to Clinical Processes (leaves) for which the Quality of the medical care constitutes its unique trunk. Instead for Hospitals, fruits (financial measures) are directly linked to Clinical Processes and Patient Satisfaction. Research limitations/implications - The main limitations of this study are the lack of new independent data to validate the obtained causal structures and the limited number of indicators that reflects the deficiency of available information in regional administrative archives.Originality/value - The present study may be useful to guide further efforts which attempt to conceptualize BSC in Health sector. As more information can be made available, other performance indicators can prove to be linked with this structure using the same methodology.

Designing balanced scorecard for multi echelon repair inventory systems

Amik Garg, S G Deshmukh — 2012-01-09 00:00:00.0

Abstract

Purpose - A numbers of approaches have been suggested in literature for the maintenance performance measurement including use of balanced scorecard (BSC). However, its application towards assessing maintenance contribution to business objectives has been limited. A framework integrating maintenance activities in a supply chain environment has been lacking in the literature. This paper identifies these gaps and attempts to develop a maintenance performance measurement framework using BSC for multi echelon repair inventory systems (MERIS) comprising of modular electronic equipment. The paper also suggests a large number of performance measures/indicators (PMs/PIs) pertaining to both maintenance and supply chain components of such systems. Few BSC implementation issues pertaining to multi echelon repair inventory systems have also been covered.Design/methodology/approach - Paper analyses a base model of a large maintenance organization and after carrying out SWOT analysis, identifies strategic themes/objectives for performance measurement. Performance measures for various objectives are identified for MERIS and implementation strategy including cascading of BSC suggested. The developed performance measurement model has also been validated using an action research (AR) based approach.Findings - Performance measurement system (PMS) for MERIS using BSC approach has been developed and an implementation strategy suggested. Action research (AR) methodology is used for developing BSC for an organization using the selected performance measures. The importance of cause-and-effect relationships between various performance objectives and measures has also been established.Research limitations/implications - Quantification of each of the developed PMs/PIs and establishing an effective information system for MERIS has been the limitations of the present work and can be taken for future research. Practical implications - The paper shall be useful for practitioners in the field of MERIS of electronic repairable wherein relevant PMs/PIs can be selected for implementation as per the requirement of the application.Originality/value - The paper is original in value addition as no work towards performance measurement of MERIS using BSC concept as applicable for non-profit organizations like military, government, NGOs etc has been reported in literature so far. The paper is relevant in this context and its where its contribution in the area of performance measurement lies.

An Application of Analytic Hierarchy Process for the Selection of a Methodology to Improve the Product Development Process

Anand Gurumurthy, Rambabu Kodali — 2012-03-16 00:00:00.0

Abstract

Purpose - According to the literature, on an average, it takes around 3-4 years to develop a new product and about 50% of the costs incurred in Product Development (PD) are tend to be spent on wastes that occur during the PD process. Hence, organizations are implementing various alternative methodologies such as Concurrent Engineering (CE), Lean Product Development (LPD)/Lean New Product Development (LNPD), Agile Product Development (APD)/Agile New Product Development (ANPD), etc. to improve their existing PD process. However, it is not clear: how does an organisation or a PD manager choose between these alternative methodologies for improving their PD process?Design/methodology/approach - Since the above-mentioned problem requires multiple factors/criteria/elements (in short, it will be called as ‘attributes’ for the sake of simplicity) to be considered simultaneously; the use of a Multi-Attribute Decision Making (MADM) model is warranted. The most commonly used MADM model, namely the Analytic Hierarchy Process (AHP) is utilized to model the above situation using a hypothetical case situation. Findings - In this paper, the different attributes that are to be considered while making a decision of selecting a suitable PD methodology was identified. Furthermore, the results of AHP indicated that LPD is a better alternative for the case situation under consideration. Research limitations/implications - The results obtained from this study cannot be generalized for other organisations either in the same sector or in the other sector, as the case study is hypothetical. Also, the decisions by the managers may vary depending upon the circumstances that prevail within each organisation. Furthermore, it should be understood that these algorithms such as AHP can provide adequate support to the decisions being made. However, it requires the experience and judgements of the decision-makers in order to arrive at a particular decision. Practical implications - It is believed that this paper will enable the practitioners to appreciate how AHP can be used in such strategic decision-making process, especially in a product design and development department of an organisation.Originality/value - According to the authors’ knowledge, no paper exists in the literature of AHP or PD or LPD/LNPD that discusses the application of AHP for the selection of a product development methodology..