Impact-ED : A new model of digital library impact evaluation

Digital library development is a relatively new area of research and previously focus has been mostly on developmental issues or usability. However more recently the focus has started to shift towards impact evaluation both in the digital library and traditional library domains. But what do we mean by impact and how can we measure it? Does information retrieved from a library help to confirm existing user knowledge/decisions, provide new knowledge for the user to make a decision/action, or contradict the user's existing knowledge to prevent a wrong decision or action? Or does the information have no impact on user knowledge, attitudes and behaviour? This thesis identifies a need for a digital library impact evaluation model that allows for evaluation of real digital libraries in real world settings and of real use by real users. It suggests that development of an impact score for comparing library impact over time or between libraries is a feature missing in previous evaluations. It then develops a model and framework to meet this need and shows how the model can be applied on an actual library by using the National Resource for Infection Control (NRIC) as a case study. The model uses a combination of pre and post visit questionnaires, study beginning and end questionnaires, web server log analysis and interviews. The subsequent data analysis leads to the development of the impact score to show the NRIC's overall impact score to be 0.58 (maximum possible score of 1). The model was then refined following discussion of how it met the requirements for digital library impact evaluation. The thesis ends with a discussion of how the project has added value to the digital library research domain and with suggestions for future research to develop the model and its application further