Transition towards Open Access

Following the template in OpenAlex’s oa-percentage tutorial, this vignette uses openalexR to answer:

How many of recent journal articles from the University of Pennsylvania are open access? And how many aren’t?

library(openalexR)
library(dplyr)
library(tidyr)
library(ggplot2)

We first need to find the openalex.id for University of Pennsylvania. We can do this by fetching for the institutions entity and put “University of Pennsylvania” in display_name or display_name.search:

oa_fetch(
  entity = "inst", # same as "institutions"
  display_name.search = "\"University of Pennsylvania\""
) %>%
  select(display_name, ror) %>% 
  knitr::kable()

display_name	ror
University of Pennsylvania	https://ror.org/00b30xv10
California University of Pennsylvania	https://ror.org/01spssf70
Hospital of the University of Pennsylvania	https://ror.org/02917wp91
University of Pennsylvania Health System	https://ror.org/04h81rw26
Indiana University of Pennsylvania	https://ror.org/0511cmw96
University of Pennsylvania Press	https://ror.org/03xwa9562
Cheyney University of Pennsylvania	https://ror.org/02nckwn80

We will use the first ror, 00b30xv10, as one of the filters for our query.

Alternatively, we could go to the autocomplete endpoint at https://explore.openalex.org/ to search for “University of Pennsylvania” and find the ror there!

All other filters are straightforward and explained in detailed in the original jupyter notebook tutorial. The only difference here is that, instead of grouping by is_oa, we’re interested in the “trend” over the years, so we’re going to group by publication_year, and perform the query twice, one for is_oa = "true" and one for is_oa = "false" .

open_access <- oa_fetch(
  entity = "works",
  institutions.ror = "00b30xv10",
  type = "article",
  from_publication_date = "2012-08-24",
  is_paratext = "false",
  is_oa = "true",
  group_by = "publication_year"
)

closed_access <- oa_fetch(
  entity = "works",
  institutions.ror = "00b30xv10",
  type = "article",
  from_publication_date = "2012-08-24",
  is_paratext = "false",
  is_oa = "false",
  group_by = "publication_year"
)

uf_df <- closed_access %>%
  select(- key_display_name) %>%
  full_join(open_access, by = "key", suffix = c("_ca", "_oa")) 

uf_df
#>     key count_ca key_display_name count_oa
#> 1  2018     4473             2018     5756
#> 2  2015     4345             2015     4858
#> 3  2014     4263             2014     4715
#> 4  2013     4219             2013     4543
#> 5  2019     4191             2019     6577
#> 6  2020     4168             2020     8170
#> 7  2022     4166             2022     7694
#> 8  2025     4111             2025     8228
#> 9  2021     4087             2021     8249
#> 10 2016     4018             2016     5079
#> 11 2017     3948             2017     5303
#> 12 2024     3897             2024     8342
#> 13 2023     3477             2023     8571
#> 14 2026     2039             2026     3308
#> 15 2012     1316             2012     1122

Finally, we compare the number of open vs. closed access articles over the years:

uf_df %>%
  filter(key <= 2021) %>% # we do not yet have complete data for 2022 and after
  pivot_longer(cols = starts_with("count")) %>%
  mutate(
    year = as.integer(key),
    is_oa = recode(
      name,
      "count_ca" = "Closed Access",
      "count_oa" = "Open Access"
    ),
    label = if_else(key < 2021, NA_character_, is_oa)
  ) %>% 
  select(year, value, is_oa, label) %>%
  ggplot(aes(x = year, y = value, group = is_oa, color = is_oa)) +
  geom_line(size = 1) +
  labs(
    title = "University of Pennsylvania's progress towards Open Access",
    x = NULL, y = "Number of journal articles") +
  scale_color_brewer(palette = "Dark2", direction = -1) +
  scale_x_continuous(breaks = seq(2010, 2024, 2)) +
  geom_text(aes(label = label), nudge_x = 0.1, hjust = 0) +
  coord_cartesian(xlim = c(NA, 2022.5)) +
  guides(color = "none")