Amanda B. Click & Rachel Borchardt American University Follow the Money An exploratory study of open access publishing funds’ impact Amanda B. Click & Rachel Borchardt American University Good afternoon! My name is Rachel Borchardt, and I’m the science librarian at American University. “I’m Amanda Click and I’m the business librarian at American”. American University recently started to pilot test an open access fund through the library, which got us interested in learning more about trends in open access funds as well as the impact that these funds have.

Background - purpose (2) While researching trends in OA funds, we noticed that there have been discussions of trends in the amount of funding and policy restrictions placed on these funds. However, less research has been done on the funded articles themselves - that is, the published articles that were supported by an open access fund. https://www.lib.fsu.edu/page/open-access-publishing-fund

Background - methodology Contacted 63 schools SPARC list Sent spreadsheet template with categories 16 schools responded Response rate: 25% Mixture of spreadsheets received Some filled in our template Some sent existing spreadsheets (3) To study trends in the published articles, we first contacted SPARC, who collect yearly data from 63 schools related to open access funds. We emailed these schools asking for a list of articles funded by their OA funds, and sent a spreadsheet template. 16 schools responded, some using our template, but most sending existing spreadsheets to us. https://schoolofpermaculture.com/wp-content/uploads/2017/06/chaos-and-order1.jpg

Background - methodology cont. Data manipulation: Mapped self-reported categories to ours Example: School and Discipline -> Discipline Looked up category information where applicable Example: Carnegie classification Categorized subject data into larger fields Journal subject and author’s discipline Categories based on Library of Congress (DOAJ) (4) After collecting spreadsheets, we started to normalize categories to standardize terminology as much as possible. This was particularly important for author’s discipline, since universities variously listed schools, departments, and research centers. We also looked up information like Carnegie classification codes, and categorized the journals into subjects based on LC classifications listed in DOAJ. https://schoolofpermaculture.com/wp-content/uploads/2017/06/chaos-and-order1.jpg

Participants George Mason University Johns Hopkins University University of California, Irvine University of California, San Francisco University of California, Santa Barbara University of California, Santa Clara University of Colorado Boulder University of Iowa University of Massachusetts Amherst University of North Carolina at Greensboro University of Oklahoma University of Pennsylvania University of Pittsburgh University of Rhode Island Virginia Tech Wake Forest University (5.1) So who participated in our data collection study? This is a list of the 16 participating institutions.

Participants - public George Mason University Johns Hopkins University University of California, Irvine University of California, San Francisco University of California, Santa Barbara University of California, Santa Clara University of Colorado Boulder University of Iowa University of Massachusetts Amherst University of North Carolina at Greensboro University of Oklahoma University of Pennsylvania University of Pittsburgh University of Rhode Island Virginia Tech Wake Forest University (5.2) Of the 16 universities, 13 of them are public institutions.

Participants - private George Mason University Johns Hopkins University University of California, Irvine University of California, San Francisco University of California, Santa Barbara University of California, Santa Clara University of Colorado Boulder University of Iowa University of Massachusetts Amherst University of North Carolina at Greensboro University of Oklahoma University of Pennsylvania University of Pittsburgh University of Rhode Island Virginia Tech Wake Forest University (5.3) The remaining 3 are private.

Participants - R1 classification George Mason University Johns Hopkins University University of California, Irvine University of California, San Francisco University of California, Santa Barbara University of California, Santa Clara University of Colorado Boulder University of Iowa University of Massachusetts Amherst University of North Carolina at Greensboro University of Oklahoma University of Pennsylvania University of Pittsburgh University of Rhode Island Virginia Tech Wake Forest University (6.1) Sorting by Carnegie classification, 12 are considered R1 institutions.

Participants - R2 and other George Mason University Johns Hopkins University University of California, Irvine University of California, San Francisco University of California, Santa Barbara University of California, Santa Clara University of Colorado Boulder University of Iowa University of Massachusetts Amherst University of North Carolina at Greensboro University of Oklahoma University of Pennsylvania University of Pittsburgh University of Rhode Island Virginia Tech Wake Forest University (6.2) 3 universities have an R2 classification, with one “other” designation for a four-year medical school focus.

Participants - medical schools George Mason University Johns Hopkins University University of California, Irvine University of California, San Francisco University of California, Santa Barbara University of California, Santa Clara University of Colorado Boulder University of Iowa University of Massachusetts Amherst University of North Carolina at Greensboro University of Oklahoma University of Pennsylvania University of Pittsburgh University of Rhode Island Virginia Tech Wake Forest University (6.3) Speaking of medical school, 11 institutions have medical schools, a designation we will return to later.

How big are these spreadsheets? (7) To get a sense of the distribution of articles between these schools, this graph shows all 1196 articles included in our study divided by institution. It’s important to note, however, that not every universities sent us their entire collection of funded research, but you can see the number of years funded in the chart on the right. Total articles: 1176

How big are these spreadsheets?: med schools (8) One of the first big trends we noticed was that the universities with medical schools, highlighted in pink, tended to fund more articles than schools without medical schools. UCSB is the only exception, having funded more articles than 3 universities with medical schools, although over a longer period of time.

Articles funded per year? (9.1) If we sort by the average number of articles published per year of funding, we see a fairly broad distribution between 3 and over 46 articles funded per year.

Articles funded per year? - R1 R2, and other (9.2) However, when we apply Carnegie classifications to the number of articles funded per year, we see the less research-heavy institutions, the R2s, are funding fewer research articles per year than the other universities.

Reimbursement amounts? Data for 885 total articles Smallest amount: $38 Largest amount: $3750 Average: $1394 Total reimbursement amount: $1.2M+ Looking at reimbursement or expenditure amounts from these OA funds, the average reimbursement was $1394, a bit less than PLOS One’s current APC of $1495. The largest amount is likely due to limits on total requests for funds and policies against funding hybrid APCs, which can be pricier for prestigious journals. Image from https://revcycleintelligence.com/images/site/article_categories/Reimbursement.jpg

Who was funded? (11.1) The vast majority of the funded authors were classified as faculty - 71%. This included a variety of labels - everything from adjunct to emeritus. Students were almost all graduate level, only three undergrads. We included a category for researcher, anyone whose classification included researcher but NOT faculty/professor.

Who was funded? 71% (11.2) The vast majority of the funded authors were classified as faculty - 71%. This included a variety of labels - everything from adjunct to emeritus. Students were almost all graduate level, only three undergrads. We included a category for researcher, anyone whose classification included researcher but NOT faculty/professor.

Who was funded? Medicine/Health 372 Science 359 Social Science 132 Engineering 110 Agriculture 57 Education 17 Humanities 10 Law 2 TOTAL 1059 (12.1) We were able to identify the author’s discipline for 1,059 of the 1,196 funded articles (89%). Almost 70% came from medicine/health or the sciences. These categories are loosely based on the DOAJ categories for journal discipline although we changed the technology category to engineering.

Who was funded? 69% Medicine/Health 372 Science 359 Social Science 132 Engineering 110 Agriculture 57 Education 17 Humanities 10 Law 2 TOTAL 1059 69% (12.2) We were able to identify the author’s discipline for 1,059 of the 1,196 funded articles (89%). Almost 70% came from medicine/health or the sciences. These categories are loosely based on the DOAJ categories for journal discipline.

What about tenure? UC Irvine UC San Francisco UC Santa Barbara UNC Greensboro University of Pennsylvania University of Rhode Island Virginia Tech (13) We had tenure status information for 7 of the 16 institutions – 449 total articles. 250 tenured, 199 non-tenured. We were a bit surprised that the numbers were as even as they were, and expected that the tenured number would be higher.

Where were the articles published? (14) DOAJ uses Library of Congress classifications for journals and articles. 43% were categorized as medicine and 24% as science. The N/A 13.9% were not listed in the DOAJ. Note that these are the 144 journals that published more than two of the articles in our dataset.

Where were the articles published? Journal Title # of Articles Discipline 1. PLOS ONE 220 Science 2. Scientific Reports 40 3. Frontiers in Psychology 21 Social Science 4. Nature Communications 16 5. Optics Express N/A 6. Frontiers in Microbiology 15 7. BMC Public Health 14 Medicine 8. Ecosphere 13 Journal Title # of Articles Discipline 9. Ecology and Evolution 11 Science 10. Nucleic Acids Research 11. Acta Crystallographica Section E 10 12. Cell Reports 13. Journal of Medical Internet Research Medicine 14. Physiological Reports 15. PLoS Genetics (15) 144 journals published more than one article from this dataset, and here you can see the top 15. The discipline categories come from DOAJ and break down to 11 science, 2 medicine, 1 social science and 1 N/A – Optics Express which is a physics journal but was not in DOAJ.

Where were the articles published? (16.1) We have publisher data for 882 (or 74%) of the articles. This graph shows the 8 most popular with number of articles published. PLOS published 20% of the articles in this subset. Note that these publishers are predominantly STEM.

Where were the articles published? 20% (16.2) We have publisher data for 882 of the articles. This graph shows the 8 most popular with number of articles published. PLOS published 20% of the articles in this subset. Note that these publishers are predominantly STEM.

Summary of key findings Medical college and R1s both seem to fund more OA research Some institutions were able to fund more articles with a smaller budget Mostly faculty use these funds Authors tend to be in science and medicine/health (17) Image from http://salesbox.com/wp-content/uploads/2016/11/18-Key-Findings-Revealed.jpg

What’s next for our study? Metrics: h-index, impact factor, citation counts Co-authors, international Time in profession Limits on use of funds (18) Today we’re shared our preliminary findings. We’re interested in looking more closely at journal and article level metrics. We’re also interested in collaborations – looking at co-authorships and international partnerships. We’d like to know more about author’s length of time in the profession as well. Image from http://www.momoney.co.nz/wp-content/uploads/Looking-to-the-future-472x355.png

Why does this matter? Are there ties between open access and impactful research? How do we measure success in open access funds? What can libraries do to better ensure success of their funds? (19) Why does this work matter? We’re exploring the following questions: … We hope our findings can inform the development and revision of library OA funds

Thank you! And we are very interested in your thoughts about this project, so please tell what you’d like to know. Amanda Click aclick@american.edu Rachel Borchardt borchard@american.edu (20) Thank you for listening! We are very excited about this dataset, and are interested in your input. What sort of questions would you like to see answered? How could this data support the work that you do?