Evidence-Based Librarianship in E-Resources Acquisition Decision: Construct Validity and Reliability of Instrument

Abstract

This paper reports on the assessment of construct validity and reliability of an instrument used to measure librarians' adoption and implementation of Evidence-based Librarianship (EBL) in the acquisition decision of electronic resources. The instrument is developed based on well-established theories/models of Technology Organization and Environment (TOE), Innovation Diffusion Theory (IDT), and a Concern-based Adoption Model (CBAM) with a newly added construct, user needs, and preferences. The instrument was evaluated using the Partial Least Squares (SMART- PLS) software applications using structural equation modelling (SEM) to determine its validity and reliability through) analysis of the measurement model (outer model) and ii) analysis of the structural model (inner model). A total of 278 participants were identified from the 1040 research population. This research applied probability sampling using proportionate stratified techniques to gather responses from librarians in library and information center management. The results of both analyses indicated that the construct validity and reliability of the instrument were acceptable and moderate, respectively. The internal consistency reported a Cronbach's alpha value of 0.718 to 0.956, composite reliability of 0.70 and 0.90, average variance extracted value above 0.50. The model prediction accuracy and relevancy revealed R2 (adoption: 0.300, implementation 0.399) and Q2 (adoption 0.213, implementation 0.227), which were acceptable and moderate, respectively. The model has also been reported to be free from collinearity issues. Thus, the instrument is ready for use in e-resource and evidence-based library acquisition research.