Thinking beyond quality: meeting the challenges of reliability analysis in library management

Abstract

The role of a reliability analysis for improving the performance of libraries and information services has become extremely crucial. The research question considered here is simple and can be summarized as follows: “How reliable is your library?”. Indeed, reliability considerations go beyond the issue of quality since they incorporate the dimension of time, i.e. they involve a number of different activities and operations that take place throughout the life-cycle of a library system or service. This paper seeks to address these issues. Design/methodology/approach – Inference modelling techniques based on both parametric and nonparametric methods are reviewed and presented together with illustrative examples. From a wide number of non-parametric methods that have been made available for reliability estimation, the ones examined here are the Kaplan-Meier and the Cumulative-hazard methods. Furthermore, parametric methods and applications, which are based on the Weibull distributional model, are examined. Findings – The nonparametric reliability modelling methods presented are simple to use and suitable for estimating the reliability of information systems and services directly from the available life data. The Weibull analysis is quite useful for a wide range of library management methods, financial applications, for modelling user behaviour within many library settings and systems i.e. digital information systems. Originality/value – This is among the first implementations of reliability analysis in libraries and information services. This paper provides essential insight to library management researchers and practitioners on how they might incorporate into the library management agenda reliability considerations. The reliability modelling techniques presented can be employed with all types of information services, and not just libraries.