Journal and Database Review, 2018–2019

The Caltech Library is currently undertaking a collections review process in the context of the necessity to reduce our expenditures for journals, databases and books by the amount of our reduced buying power due to cost increases, which is $175,000 for calendar year 2018. Through this website, we present data on journal usage and seek community feedback on individual titles. The Library requests that comments be submitted by October 31 but will continue to accept feedback indefinitely.

On the recommendation of the Library Committee, we are planning for both 2018 and 2019 reductions through this process.

As with the journal cuts taken in 2016, an additional $50K per year will be cut so that the library has some flexibility to restore or add journals based on new information ($25K) and increase DocuServe's budget to meet anticipated increased demand ($25K). Thus, the target reduction for 2018-19 is $450K ($225K each year). Reductions for 2019 do not need to be finalized until fall 2018 and those journals or databases will not be canceled until October 2018 at the earliest. The Web of Science database will be assessed during 2018, as its usage has declined significantly in the past five years.

The journal subscription reductions should be seen in light of a rapidly evolving context for accessing the scholarly literature. Increasingly, articles are acquired outside of library subscriptions, through open access copies available on Google Scholar, PubMed Central or arXiv, or through growing research sharing sites such as ResearchGate.

The Caltech Library continues to leverage its place as a node in a network of research libraries committed to resource sharing. DocuServe plays an increasing role in meeting Caltech's needs for research materials. The number of DocuServe requests has increased more than 150% over the past two years without any increase in turnaround time (in fact, a slight decrease). In addition, this past year we introduced a near-real-time, unmediated delivery option, RUSH, including 10 free Rush requests per person per year. ​See DocuServe and Interlibrary Loan for more detail. We have analyzed the impact of the 2016 journal cuts and found that they had minimal impact on Caltech's increase in borrowing.

The linked journal, ebook package and database title lists are searchable and sortable by division as well as each of the data fields. The titles are as comprehensive as possible, reflecting what the library acquires and how much it costs. The titles that appear on these lists do not reflect what we intend to cancel.

The data includes several relevant data points for each title:

  • Price: Projected 2018 price.
  • 2016 Downloads: Number of PDF downloads of 2016 content in 2016 (last full year). Note that this is not all downloads of articles in the journal in 2016. Downloads of current year content corresponds to what we pay for, and thus is the only meaningful download figure to use in a cost-per-download calculation (unless the library does not retain access to prior years' content after cancellation; in that case, we use a download figure that represents all usage in 2016).
  • CPD and OA-Adj CPD: Cost per download and Open Access Adjusted Cost Per Download.
  • Caltech papers: Number of peer-reviewed documents indexed in the Web of Science between 2007 and 2016 for which at least one author was affiliated with Caltech. Source: Compiled by Science-Metrix using data from the Web of Science (Thomson Reuters). Methodological note: Anomalies between the reported number of papers and results of a direct WoS search are the result of several factors: these numbers include only peer reviewed journals and DOC TYPE article and review. In addition, identification of Caltech (and not JPL) authors uses zip code, which is occasionally missing in the WoS data.
  • Caltech references: Number of references made by researchers of your institution to articles from a given journal between 2007 and 2016. See above for source and methodology notes.
  • Additional data points (3-year OA average level and database prices) are available on the "more information" link, which requires authentication because the data on OA levels by title is proprietary (provided by 1science).

At the end of the year, we will amend these lists with final disposition status for each title for 2018. The website will remain live, and continuously updated with new data, through at least 2018, in anticipation of 2019 reductions. Please feel free to contact Kristin Antelman, University Librarian, or your liaison librarian with any questions.

Notes on title lists: 

  • Many titles are purchased as part of a package and therefore cannot be canceled individually. The "more info" page will describe package contents and subscription constraints.
  • Accessing some data, as well as commenting, requires authentication.
  • The level of open access for titles has been incorporated in the decision-making process because studies show that approximately 50% of articles accessed by researchers are open copies (Archambault, Simon Inger Consulting) . Therefore, the value of a subscription to a journal with high open access content is declining relative to a journal with low open access.
  • Download data is provided by the publishers and includes downloads of publisher-hosted copies of open access articles; it does not include downloads of articles posted to sites such as ResearchGate or repositories such as PubMedCentral or arXiv. Thus, journals' official usage data is not a good indicator of the actual usage of those journal titles by Caltech researchers.