Users' behaviour patterns in academic libraries' OPACs : a multivariate statistical analysis

Abstract

This paper aims to present real time user searches in a Greek academic library OPAC (University of Macedonia Library) in relation to user profile.

Design/methodology/approach: Using as a test bed a Greek academic library and its OPAC's transaction logs along with a system implanted questionnaire, data were gathered, processed and analyzed using multivariate statistical analysis techniques.

Findings: In making a synthesis of the analyzed data, a series of questions related to everyday library work were answered, giving libraries a tool to apply the gained knowledge in order to make decisions regarding their OPAC, their user education programs and their reference services.

Research limitations/implications: The present paper focuses on the analysis of those variables that were considered to be the most representative for constructing a user profile.

Originality/value: This paper builds upon the techniques of data collection and presents a new tool for analyzing them statistically. Data derived from libraries were processed and analyzed statistically using the classical descriptive statistics. The suggested multivariate statistical method is designed to become a tool for analyzing qualitative data and to be used in a variety of library applications. It is also particularly helpful in analyzing cross-tabular data in the form of numerical frequencies and allows all associations amongst pairs of variables to be analyzed as well as each association between a variable and itself.