Recent Submissions

  • INDICATORS OF IMPROVING LIFE SCIENCES RESEARCH PERFORMANCE AT US ACADEMIC INSTITUTIONS

    Vernon, Marlo; Department of Interdisciplinary Studies (8/3/2018)
    The academic research institution has long been recognized as a source of innovation and scientific advancement. The goal of this study is to determine how external and internal influences on university research can best contribute to and benefit society through science, economics, and public health. A systematic review of university ranking systems first outlines the current metrics used to evaluate the productivity of research and their validity for assessing research quality and translation of ideas. A total of 24 ranking systems were identified and 13 eligible ranking systems were evaluated. Ranking systems rely on singular indicators, reputation surveys, and tend to be non-replicable. Rankings influence academic choice yet research performance measures are the most weighted indicators. A new multi-dimensional framework of indicators for evaluation of academic research is then proposed across three factors: contributions to science, public health, and economics is then proposed. Data on faculty size, research expenditure, publications, citations, intellectual property outcomes, clinical trials registration and results, and contributions to clinical practice guidelines were included. National benchmarks are reported for the top ten percentile averages of each indicator. One of the proposed public health indicators utilizes clinical trials reporting. At 167 universities, 16,787 clinical trials were evaluated for planning, execution, and overall quality between 2001 and 2016. Over time, significant quality improvement was observed, however execution quality was much lower than planning quality. For completed intervention trials after 2007, only 21% (95%CI: 20%, 21%) had uploaded results – this is required under certain conditions within one year of completion. NIH funded trials had significantly better quality scores than all others (p<0.001). Finally, latent profile analysis (LPA) identified three university profiles determined by the proposed indicators, which significantly predicted research expenditure and income generated from licensure. The profiles were also linearly associated with the 2015 Carnegie Classification, providing convergent validity. Considering the large ratio of non-reproducible research, and the increasing societal pressure to demonstrate value, broader and more practical indicators for evaluation, as proposed, may better support improvement and attract public trust in research.