Does Size Matter? The Effect of Resource Base on Faculty Service Quality Perceptions in Academic Libraries

Abstract

We explore whether the relative size of an academic library’s resource base, as indicated by the Carnegie classification of the library’s parent institution, impacts faculty perceptions of library service quality. Using results from the 2006 administration of the LibQUAL+® survey, the study tests for statistically significant differences between research universities and masters-level universities in terms of faculty minimum, perceived, desired and adequacy gap scores for each of the three LibQUAL+® service dimensions (Information Control, Library as Place, and Affect of Service). Findings suggest that university type does impact expectations and perceptions of service quality, but does not impact ratings of service adequacy, the extent to which faculty perceive that a library meets their expectations. Introduction The current study tests for differences in faculty perceptions of library service quality between two types of academic libraries: those at large research universities and those at master’s-level colleges and universities. Our intention is to determine whether the relative size of an academic library’s funding level, as indicated by the Carnegie classification of its parent institution, matters to faculty perceptions of library service quality, as reflected in LibQUAL+® data. Our primary motivation for conducting this study was to seek evidence that larger funding allocations translate into higher faculty perceptions of library service quality, as such evidence might be useful fodder for library administrators to use in future funding requests to the central administrations of their parent institutions. The LibQUAL+® literature includes little in terms of comparisons between different types of academic libraries. In a fundamental grounding document for the LibQUAL+® instrument1, Cook compared user groups within Association of Research Libraries (ARL) and non-ARL institutions to see whether types of users differed in their perceptions of each service quality dimension identified by LibQUAL+®. She found small but significant differences between user groups for both types of institutions, and the patterns seemed slightly different for each institution type. Among institutions with ARL libraries, the largest difference among user groups was for the Library as Place dimension, whereas for non-ARL institutions, the largest differences among user groups was in the Affect of Service dimension. As Cook’s primary research focus was not on finding differences between institution types, she did not explicitly test to see whether these seemingly divergent patterns between ARL and non-ARL libraries were statistically significant. A second factor motivating us to question whether there might be variation across different types of academic libraries surfaced some years ago when one of this paper’s authors, Fred Heath, and his colleagues at Texas AM yet the perceived pattern raised some interesting questions. Should a significant difference between these types of libraries exist, would it be due to the fact that many of the faculty served by these non-ARL libraries earned their degrees at large research institutions with ARL libraries, which are usually better resourced than their non-ARL counterparts? Were faculty expectations for library service quality created during the doctoral research process at institutions with better-resourced ARL libraries being transported to these non-ARL institutions with smaller resource bases? Were these faculty members transferring their expectations and frustrations with library services to their graduates students and undergraduates at these smaller nonARL libraries? Were these smaller libraries being placed at a relative disadvantage in terms of perceived service adequacy by expectations created during their faculty’s training? Building on the work of Cook and the team at Texas AM representative examples include Columbia University, Clemson University, and the University of Texas at Austin. A masterslevel library is defined, for the purposes of the study, to be an academic library located at a Carnegie Masters L or M institution. These are institutions with larger and medium mastersgranting programs; representative examples include Humboldt State University, Gonzaga University, and the University of Texas at San Antonio. For research libraries our sample included fifty-six libraries with 8,215 faculty members surveyed. For masters-level academic libraries, our sample included sixty-six libraries with 5,664 faculty members surveyed. Measures LibQUAL+® is a set of services constructed in response to the Association of Research Libraries (ARL) New Measures Initiative. It is an assessment tool for collecting and analyzing customer perceptions of service quality in three areas: Affect of Service (questions in this category relate to the attitudes and abilities of employees when assisting others), Library as Place (questions in this category relate to the library facilities and use of space), and Information Control (questions in this category focus on collection breadth and scope, the ability of respondents to find information on their own, and the Libraries success in providing information).3 The survey consists of twenty-two service statements and a comment box. Respondents are asked to rate each service indicator on three levels (the minimum level of quality that is acceptable, the desired level of quality, and the current perceived level of service quality) using a Likert scale of 1-9. As noted above, adequacy gaps are calculated as the difference between perceived and minimum scores. Analysis We accessed summary data for each school; that is, the average faculty minimum, perceived, desired, and adequacy scores for each dimension. For each type of rating, we conducted a separate analysis of variance (ANOVA), containing the betweensubjects factor of Institution Type (Master’s vs. Research) and the within-subjects factor of Dimension of Service (Affect of Service, Library as a Place, and Information Control). Rating types (minimum, perceived, desired, and adequacy) were treated as separate dependent variables; because adequacy is calculated from minimum and perceived scores, it was not appropriate to include all four outcomes in a multivariate ANOVA. In each analysis, we focused on the main effect of Institution, and the interaction between Dimension and Institution (using Hotelling’s F). If the main