Purpose:
This proposal is submitted under the “from a differing perspective” strand, as it represents a library-related portion of a broader PhD dissertation. The overall purpose of the dissertation is to understand how universities can best support the research productivity of their faculty. In approaching this topic, it is crucial to examine how investment in library resources and services factors into fostering a high-performance research environment. Focusing on the discipline of biomedical engineering (BME), this study analyzes how various library measures and other institutional factors correlate with the quantity and quality of scholarly output generated by BME programs at research-intensive US doctoral institutions. Using quantitative methods, these programs will be ranked in terms of their productive efficiency for the purpose of identifying a handful of programs at the upper and lower extremes of the rankings. These contrasting subsets of BME programs then form the basis of a comparative analysis – using both quantitative and qualitative methods – designed to illustrate which institutional factors most distinguish high productivity programs from their less effective peers, with a particular focus on the library characteristics that surround programs at each end of the spectrum.
Design, methodology or approach:
Borrowing from the field of production theory in economics and, more specifically, the concept of total factor productivity, this methodology is designed to calculate how much high-quality scholarship a BME program ought to produce given it resources (e.g., level of grant funding, departmental faculty count, teaching loads, etc.). The true purpose is to identify which programs are actually producing more scholarship than should be expected, with the assumption that those who exceed expectations by the greatest margin have achieved the highest productive efficiency in the field.
To begin, the number of peer reviewed articles produced by each BME program from 2013 to 2016 are aggregated and weighted by journal impact factor to serve as a dependent variable measuring the quantity and quality of each program’s scholarly output. Using OLS regression analysis, this number is then regressed against an array of programmatic and institutional inputs presumed to influence scholarly output, including a variety of library expenditure and staffing levels measures. The resulting regression equation describes the relationship between the inputs of the research process and the scholarly output for the field of biomedical engineering. Using this equation, it is possible to calculate the expected amount of scholarship that each BME program should have been able to produce given it inputs. As state above, the difference between this expected value and actual output forms the basis for establishing each program’s productive efficiency – with those who out produce expectations being considered high productivity and those producing below expectations being considered low productivity – enabling the comparative analysis.
Findings:
The quantitative analysis for this study is currently in progress, but it is scheduled to be completed well in advance of the conference.
The intent is to divide the presentation into three equal segments. The first segment will profile libraries associated with low productivity research environments. This will be contrasted with second segment that examines library characteristics associated with high productivity environments. The primary focus of these segments will be to discuss correlations between various library variables and scholarly productivity, as well as to define any cross-institutional trends that characterize libraries associated with either environment. The final segment will consist of the presentation of a qualitative framework designed to understand how researchers in both types of environments experience and interact with their libraries. Relying on group interviews with the researchers themselves, the assumption is that more can be learned by talking to researchers in these opposite and contrasting environments that could be learned by simply interviewing researchers in general.
The qualitative portion of the dissertation will not be completed by the time of the conference, but qualitative framework has been designed and will be discussed in detail.
Note: no identifiable information associated with any institution or program will be disclosed.
Conclusions:
As mentioned above, the analysis is scheduled to be completed later this spring and conclusions will be developed accordingly.
Originality and value:
At its most basic level, this study attempts demonstrate library value empirically by correlating library investment with scholarly productivity – although it remains to be seen whether the correlations will prove to be positive and statistically significant. Nonetheless, it is important to make this attempt simply because supporting scholarly research is so central a function of university libraries. The approach of studying both high- and low-productivity environments as a means of further illustrating the role the libraries can play in the advancing research is possibly unique. This, too, is intended to make manifest the value that a well-resourced library offers to its research community. The most fundamental reason for conducting this empirical analysis, however, is simply to set the stage for the qualitative investigation that will follow. Identifying high- and low-productivity BME programs will enable the collection of valuable testimonies, stories, and anecdotes from researchers who occupy drastically different environments about their interactions with their libraries. Although the qualitative study will unfortunately not be completed in time for the conference, it is hoped that sharing its qualitative framework will spur discussion about what and how we might learn by engaging our research community.
