Identifying library user characteristics and looking beyond the demographics

<p>Title: Identifying library user characteristics and looking beyond the demographics</p> <p>Objective: Discuss how the commonly used demographic information collected through polls and user surveys, while descriptive, may have limitation in identifying library user characteristics. Describe statistical models that can improve our knowledge about library users.</p> <p>Background: Library surveys often ask demographic questions such as age, user type or library use. However these demographic questions are created by researchers' preconceived notions which were formed based on existing library statistics, user profiles, or simply by the researchers' observation of the survey population. What if the demographics is not a factor in survey responses? For example, if the average library facilities satisfaction scores from two demographic groups (Graduate Students and Faculty) are similar, does that indicate the Graduate Student and Faculty respondents are satisfied with library facilities the same way? What if there are significant differences among the Graduate Students and Faculty respondents are not on group level (Graduate Students vs. Faculty) but on an individual level? How can researchers test and discover the unobserved user characteristics?</p> <p>Method: In spring 2014 the University of Illinois at Urbana-Champaign conducted the LibQUAL+(r) Lite Survey, an online survey for assessing library service quality. Initially descriptive statistical analysis was conducted (e.g. cross-tabulation by demographics). However, it was unclear if there were significant relationships between User Groups and particular questions under investigation. General Linear Models were used to identify correlation in the data and some hidden user characteristics (latent variables) emerged.