Coverage of DOAJ journals in WoS and Scopus dataset

Author

Affiliation

Philippe Mongeon

Dalhousie University

Published

March 18, 2024

Data sources

DOAJ data dump (downloaded on March 6th, 2024).
Scopus journal list (downloaded on March 6th, 2024) - last updated February 2024.
WoS journal list (downloaded on March 6th, 2024 - last updated Feb 19th, 2024).
OpenAlex (February 2024 data dump).
World Bank classification of countries by income group (downloaded March 18th): https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending-groups.

Create database

Code

# 
# # ----------------------------
# # countries
# # ----------------------------
# 
# dbExecute(db,
#           "CREATE TABLE countries (
#             country_code CHAR(2),
#             economy TEXT,
#             region TEXT,
#             income_group TEXT,
#             PRIMARY KEY (country_code)
#             );")
# 
# # ----------------------------
# # doaj_countries_publishers
# # ----------------------------
# 
# dbExecute(db,
#           "CREATE TABLE doaj_countries_publishers (
#             id SMALLINT,
#             country TEXT,
#             country_code CHAR(2),
#             PRIMARY KEY (id),
#             FOREIGN KEY (country_code) REFERENCES _countries (country_code)
#             );
#           ")
# 
# # ----------------------------
# # doaj
# # ----------------------------
# 
# dbExecute(db,
#           "CREATE TABLE doaj_journals (
#             id TEXT,
#             title TEXT,
#             oa_start_year SMALLINT,
#             country_publisher_id SMALLINT,
#             oa_type TEXT,
#             PRIMARY KEY (id),
#             FOREIGN KEY (country_publisher_id) 
#               REFERENCES _doaj_countries_publishers (id)
#           );")
# 
# # ----------------------------
# # doaj_issns
# # ----------------------------
# 
# dbExecute(db,
#           "CREATE TABLE doaj_issns (
#             doaj_id TEXT,
#             issn TEXT,
#             type TEXT,
#           PRIMARY KEY (doaj_id, issn),
#           FOREIGN KEY (doaj_id) REFERENCES _doaj (id));")
# 
# # ----------------------------
# # doaj_languages
# # ----------------------------
# 
# dbExecute(db,
#           "CREATE TABLE doaj_languages (
#             doaj_id TEXT,
#             language TEXT,
#           PRIMARY KEY (doaj_id, language),
#           FOREIGN KEY (doaj_id) REFERENCES _doaj (id));")
# 
# # ----------------------------
# # journals_disciplines
# # ----------------------------
# 
# dbExecute(db, 
#   "CREATE TABLE journals_disciplines (
#     doaj_id text,
#     discipline varchar(15),
#   PRIMARY KEY (doaj_id, discipline),
#   FOREIGN KEY (doaj_id) REFERENCES doaj (id));")

Data processing

Country income level

Code

countries <- read_csv("data/country_classification.csv")

DOAJ

Import DOAJ journal list

Code

# Load DOAJ journal list
doaj_data <- read_csv("data/journalcsv__doaj_20240306_1820_utf8.csv") %>% 
  select(id = `URL in DOAJ`,
         title = `Journal title`,
         oa_start_year = `When did the journal start to publish all content using an open license?`,
         alternative_title = `Alternative title`,
         issn = `Journal ISSN (print version)`,
         eissn = `Journal EISSN (online version)`,
         language = `Languages in which the journal accepts manuscripts`,
         country_publisher = `Country of publisher`,
         former_issn = `Continues`,
         new_issn = `Continued By`,
         apc = APC) %>% 
  mutate(id = str_remove(id, "https://doaj.org/toc/"))

Publishing countries

Code

doaj_countries_publishers <- doaj_data %>% 
  left_join(countries, by=c("country_publisher"="economy")) %>% 
  select(country_publisher, country_code) %>% 
  unique() %>% 
  rownames_to_column("id")

DOAJ journals

Code

doaj_journals <- doaj_data %>%    
  mutate(id = str_remove(id, "https://doaj.org/toc/")) %>%    
  mutate(oa_type = ifelse(apc == "Yes","Gold","Diamond")) %>% 
  left_join(doaj_countries_publishers %>% 
              rename(country_publisher_id = id),
            by="country_publisher") %>% 
  select(id, title, oa_start_year, country_publisher_id, oa_type)

Language

Code

doaj_languages <- doaj_data %>% 
  select(doaj_id = id, language) %>% 
  separate_rows(language, sep = ",") %>% 
  mutate(language = str_squish(language)) %>% 
  unique()

ISSNs

Code

   doaj_issns <- bind_rows(doaj_data %>%
                         select(doaj_id = id, issn) %>%
                         mutate(type = "issn"),
                       doaj_data %>%
                         select(doaj_id = id, issn = eissn) %>% 
                         mutate(type = "eissn"),
                       doaj_data %>%
                         select(doaj_id = id, issn = former_issn) %>% 
                         mutate(type = "former issn"),
                       doaj_data %>%
                         select(doaj_id = id, issn = new_issn) %>% 
                         mutate(type = "new issn")) %>% 
     unique() %>% 
     mutate(doaj_id = str_remove(doaj_id, "https://doaj.org/toc/"))

Journal titles

Code

doaj_titles <- doaj_data %>% 
  select(doaj_id = id, title) %>% 
  mutate(type = "main_title") %>% 
    bind_rows(doaj_data %>% 
                select(doaj_id = id, title = alternative_title) %>% 
                mutate(type = "alternative_title")) %>% 
  drop_na() %>% 
  unique()

Scopus

We do not exclude journals that were discontinued from Scopus in January 2024, and do not include the newly accepted journals in the process of being added to Scopus.

Import Scopus journal list

Code

scopus_data <- read_xlsx("data/ext_list_February_2024.xlsx", sheet = 1) %>% 
  select(id = 1,
         title = 2,
         issn = 3,
         eissn = 4,
         active = `Active or Inactive`,
         discontinued_quality_issues = 7,
         language = 8,
         oa_status = 10,
         type = 13,
         related_title1 = 15,
         related_title2 = 16,
         related_title3 = 17,
         related_title4 = 18,
         publisher = 19,
         life_sciences = 22,
         social_sciences = 23,
         physical_sciences = 24,
         health_sciences = 25
         ) %>% 
  mutate(life_sciences = as.logical(ifelse(is.na(life_sciences),F,T))) %>% 
  mutate(social_sciences = as.logical(ifelse(is.na(social_sciences),F,T))) %>% 
  mutate(physical_sciences = as.logical(ifelse(is.na(physical_sciences),F,T))) %>% 
  mutate(health_sciences = as.logical(ifelse(is.na(health_sciences),F,T))) %>% 
  mutate(active = as.logical(ifelse(active == "Active",T,F))) %>% 
  mutate(active = as.logical(ifelse(discontinued_quality_issues == "Discontinued by Scopus",T,F)))

Journals

Code

scopus_journals <- scopus_data %>% 
  select(id, title, active, discontinued_quality_issues, type, publisher,
         life_sciences, social_sciences, physical_sciences,
         health_sciences)

ISSNs

Code

scopus_issns <- bind_rows(scopus_data %>%
                            select(scopus_id = id, issn) %>%
                            mutate(type = "issn") %>%
                            drop_na() %>% 
                            mutate(issn = ifelse(str_length(issn)==7,
                                                 str_c("0",issn),
                                                 issn)) %>% 
                            mutate(issn = str_c(str_sub(issn,1,4),
                                                "-",
                                                str_sub(issn,5,8))),
                          scopus_data %>%
                            select(scopus_id = id, issn = eissn) %>% 
                            mutate(type = "eissn") %>% 
                            drop_na() %>% 
                            mutate(issn = ifelse(str_length(issn) ==7,
                                                 str_c("0",issn),
                                                 issn)) %>% 
                            mutate(issn = str_c(str_sub(issn,1,4),
                                                "-",
                                                str_sub(issn,5,8)))) %>%
  unique()

Journal titles

Code

scopus_titles <- bind_rows(scopus_data %>% 
                             select(scopus_id = id, title) %>% 
                             mutate(type = "title"),
                           scopus_data %>% 
                             select(scopus_id = id, title = related_title1) %>% 
                             mutate(type = "related_title"),
                           scopus_data %>% 
                             select(scopus_id = id, title = related_title2) %>% 
                             mutate(type = "related_title"),
                           scopus_data %>% 
                             select(scopus_id = id, title = related_title3) %>% 
                             mutate(type = "related_title"),
                           scopus_data %>% 
                             select(scopus_id = id, title = related_title4) %>% 
                             mutate(type = "related_title")) %>% 
  drop_na() %>% 
  unique() 

# Manually clean journal titles with square brackets
scopus_titles %>% 
  filter(str_detect(title, "\\[")) %>% 
  writexl::write_xlsx("data/scopus_journal_titles_to_clean.xlsx")

scopus_titles <- bind_rows(scopus_titles %>% 
  filter(!str_detect(title, "\\[")),
  read_xlsx("data/scopus_journal_titles_cleaned.xlsx")) %>% 
  unique()

Web of Science

Import WOS journal list

Code

wos_data <- bind_rows(read_csv("data/wos-core_AHCI 2024-February-19.csv") %>%
                        mutate(index = "ahci"),
                      esci<-read_csv("data/wos-core_ESCI 2024-February-19.csv") %>%
                        mutate(index = "esci"),
                      ssci<-read_csv("data/wos-core_SSCI 2024-February-19.csv") %>%
                        mutate(index = "ssci"),
                      scie<-read_csv("data/wos-core_SCIE 2024-February-19.csv") %>%
                        mutate(index = "scie")) %>% 
   select(title = `Journal title`,
         issn = ISSN,
         eissn = eISSN,
         publisher =`Publisher name`,
         language = Languages,
         wos_category = `Web of Science Categories`,
         index) %>% 
  group_by(title, issn, eissn, publisher, language, wos_category) %>% 
  mutate(index, str_c(index, collapse = ",")) %>% 
  unique() %>% 
  rownames_to_column("id")

Journals

Code

wos_journals <- wos_data %>% 
  select(-index)

Indexes

Code

wos_journals_indexes <- wos_data %>%
  select(wos_id = id, index) %>% 
  separate_rows(index, sep=",")

ISSNs

Code

wos_journals_issns <- bind_rows(wos_data %>% 
                                  select(wos_id = id, issn) %>%
                                  mutate(type = "issn") %>%
                                  drop_na(),
                                wos_data %>%
                                  select(wos_id = id, issn = eissn) %>% 
                                  mutate(type = "eissn") %>%
                                  drop_na()) %>% 
  unique()

Openalex

Import OpenAlex journal data

Code

openalex_data <- read_csv("data/openalex_Feb_2024.csv")

OpenAlex journals

Code

openalex_journals <- openalex_data %>% 
  select(id,
         title = display_name,
         is_oa,
         is_in_doaj,
         type,
         country_code,
         publisher)

ISSNs

Code

openalex_issns <- bind_rows(openalex_data %>% 
                              select(openalex_id = id, issn = issn_l) %>% 
                              mutate(type = "issn_l") %>% 
                              drop_na(),
                            openalex_data %>% 
                              select(openalex_id = id, issn) %>% 
                              separate_rows(issn, sep = ",") %>%
                              mutate(type = "issn") %>% 
                              drop_na() %>% 
                              mutate(issn = str_remove_all(issn, "\"")) %>% 
                              mutate(issn = str_remove_all(issn, "\\[")) %>% 
                              mutate(issn = str_remove_all(issn, "\\]"))) %>% 
  unique()

Journal matching

DOAJ and WoS

Code

doaj_wos_issn_match <- doaj_issns %>% 
  inner_join(wos_journals_issns, by="issn") %>% 
  select(doaj_id, wos_id) %>% 
  mutate(matched_by = "issn")

doaj_wos_title_match <- doaj_titles %>% 
  filter(!doaj_id %in% doaj_wos_issn_match$doaj_id) %>%
  mutate(title = str_to_lower(title)) %>% 
  inner_join(wos_journals %>% 
               select(wos_id = id, title) %>% 
               mutate(title = str_to_lower(title)), by="title") %>% 
  select(doaj_id, wos_id) %>% 
  mutate(matched_by = "title")

doaj_wos_matches <- bind_rows(doaj_wos_issn_match, 
                              doaj_wos_title_match) %>% 
  unique()

DOAJ and Scopus

Code

doaj_scopus_issn_match <- doaj_issns %>% 
  inner_join(scopus_issns, by="issn") %>% 
  select(doaj_id, scopus_id) %>% 
  mutate(matched_by = "issn")

doaj_scopus_title_match <- doaj_titles %>% 
  filter(!doaj_id %in% doaj_scopus_issn_match$doaj_id) %>%
  select(doaj_id, title) %>% 
  mutate(title = str_to_lower(title)) %>% 
  inner_join(scopus_journals %>% 
               select(scopus_id = id, title) %>% 
               mutate(title = str_to_lower(title)), by="title") %>% 
  select(doaj_id, scopus_id) %>% 
  mutate(matched_by = "title")

doaj_scopus_matches <- bind_rows(doaj_scopus_issn_match, 
                              doaj_scopus_title_match) %>% 
  unique()

DOAJ and OpenAlex

Code

doaj_openalex_issn_match <- doaj_issns %>% 
  inner_join(openalex_issns, by="issn") %>% 
  select(doaj_id, openalex_id) %>% 
  mutate(matched_by = "issn")

doaj_openalex_title_match <- doaj_titles %>% 
  filter(!doaj_id %in% doaj_openalex_issn_match$doaj_id) %>%
  mutate(title = str_to_lower(title)) %>% 
  inner_join(openalex_journals %>% 
               select(openalex_id = id, title) %>% 
               mutate(title = str_to_lower(title)), by="title") %>% 
  select(doaj_id, openalex_id) %>% 
  mutate(matched_by = "title")

doaj_openalex_matches <- bind_rows(doaj_openalex_issn_match, 
                              doaj_openalex_title_match) %>% 
  unique()

Checks and manual matches

Journals listed as in_doaj = T in Openalex but that were not matched in our process
Journals listed as OA journals in the Scopus list that were not matched in our process.

OpenAlex

Code

library(kableExtra)
openalex_in_doaj <- openalex_data %>%
  filter(is_in_doaj == T)

openalex_data %>% 
  left_join(doaj_openalex_matches , by=c("id"="openalex_id")) %>% 
  mutate(matched = ifelse(is.na(doaj_id),0,1)) %>% 
  group_by(is_in_doaj) %>%
  mutate(total = n()) %>% 
  mutate(n_matched = sum(matched)) %>% 
  select(is_in_doaj, n_matched, total) %>% 
  unique() %>% 
  kbl()

is_in_doaj	n_matched	total
FALSE	1334	273680
TRUE	19476	19691

Code

openalex_data %>%
  filter(is_in_doaj == T) %>% 
  anti_join(doaj_openalex_matches, by = c("id"="openalex_id")) %>% 
  writexl::write_xlsx("data/openalex_journals_to_match.xslx")

dbReadTable(db, "doaj") %>% 
  left_join(dbReadTable(db, "doaj_issns"), by=c("id" = "doaj_id")) %>%
  group_by(id) %>% 
  mutate(issn = str_c(issn, collapse = ", ")) %>% 
  unique() %>% 
  ungroup() %>% 
  writexl::write_xlsx("data/doaj_for_matching.xlsx")

Add manual matches to data

Code

openalex_manual_matches <- read_xlsx("data/openalex_journals_manually_matched.xlsx")

doaj_openalex_matches <- bind_rows(doaj_openalex_matches,
                                   openalex_manual_matches %>% 
                                     filter(doaj_id != 0) %>%  
                                     select(doaj_id, openalex_id = id) %>% 
                                     mutate(matched_by = "manual"))

doaj_openalex_matches <- doaj_openalex_matches %>% 
  mutate(openalex_id = str_remove(openalex_id, "https://openalex.org/S"))

Scopus

Code

scopus_data %>% 
  left_join(doaj_scopus_matches, by = c("id" = "scopus_id")) %>% 
  mutate(matched = ifelse(!is.na(doaj_id),1,0)) %>% 
  group_by(oa_status) %>% 
  mutate(total = n()) %>% 
  mutate(n_matched = sum(matched)) %>% 
  select(oa_status, n_matched, total) %>% 
  unique() %>% 
  kbl()

oa_status	n_matched	total
NA	972	38074
Unpaywall Open Acess	7230	7405

Code

scopus_data %>%
  filter(!is.na(oa_status)) %>% 
  anti_join(doaj_scopus_matches, by = c("id"="scopus_id")) %>% 
  writexl::write_xlsx("data/scopus_journals_to_match.xslx")

Add manual matches to data

Code

scopus_manual_matches <- read_xlsx("data/scopus_journals_manually_matched.xlsx")

doaj_scopus_matches <- bind_rows(doaj_scopus_matches,
                                   scopus_manual_matches %>% 
                                     filter(doaj_id != 0) %>%  
                                     select(doaj_id, scopus_id = id) %>% 
                                     mutate(matched_by = "manual"))

Discipline classification

Code

# doaj_issns <- dbReadTable(db, "doaj_issns")
# old_data <- dbReadTable(db, "data")
# 
# dbWriteTable(db, "journals_disciplines", 
#              doaj_issns %>%
#                filter(!doaj_id %in% dbReadTable(db,"journals_disciplines")$doaj_id) %>%
#                inner_join(old_data %>% filter(!is.na(issn)) %>% select(issn, nsf), by="issn") %>%
#                select(doaj_id, discipline = nsf) %>% 
#                separate_rows(discipline, sep = ";") %>% 
#                mutate(discipline = str_squish(discipline)) %>% 
#                unique() %>% 
#                drop_na(),
#              row.names = F,
#              append = T)
# 
# dbWriteTable(db, "journals_disciplines", 
#              doaj_issns %>%
#                filter(!doaj_id %in% dbReadTable(db,"journals_disciplines")$doaj_id) %>%
#                inner_join(old_data %>% filter(!is.na(eissn)) %>% select(issn = eissn, nsf), by="issn") %>%
#                select(doaj_id, discipline = nsf) %>% 
#                separate_rows(discipline, sep = ";") %>% 
#                mutate(discipline = str_squish(discipline)) %>% 
#                unique() %>% 
#                drop_na(),
#              row.names = F,
#              append = T)
# 
# dbWriteTable(db, "journals_disciplines", 
#              dbReadTable(db, "doaj_titles") %>%
#                filter(!doaj_id %in% dbReadTable(db,"journals_disciplines")$doaj_id) %>%
#                inner_join(old_data %>% filter(!is.na(journal_title)) %>% select(title = journal_title, nsf), by="title") %>%
#                select(doaj_id, discipline = nsf) %>% 
#                separate_rows(discipline, sep = ";") %>% 
#                mutate(discipline = str_squish(discipline)) %>% 
#                unique() %>% 
#                drop_na(),
#              row.names = F,
#              append = T)

Automatic classification based on previous classification

Code

# library(tidytext)
# library(textstem)
# subject_discipline <- dbReadTable(db, "doaj") %>% 
#   separate_rows(subject, sep = ",") %>% 
#   mutate(subject = str_squish(subject)) %>% 
#   mutate(subject = str_sub(subject,1,str_locate(subject,":")[,1]-1)) %>% 
#   select(id, subject) %>% 
#   inner_join(dbReadTable(db, "journals_disciplines"), by=c("id" = "doaj_id")) %>% 
#   group_by(subject, discipline) %>% 
#   summarize(n = n()) %>% 
#   ungroup() %>% 
#   group_by(subject) %>% 
#   mutate(total = sum(n)) %>% 
#   mutate(pct = n/total) %>% 
#   drop_na()
#   
# journal_discipline <- dbReadTable(db, "doaj") %>% 
#   filter(!id %in% dbReadTable(db, "journals_disciplines")$doaj_id) %>% 
#   separate_rows(subject, sep = ",") %>% 
#   mutate(subject = str_squish(subject)) %>% 
#   mutate(subject = str_sub(subject,1,str_locate(subject,":")[,1]-1)) %>% 
#   select(id, title, subject) %>%
#   inner_join(subject_discipline, by="subject") %>% 
#   group_by(id, discipline) %>% 
#   mutate(total_score = sum(pct)) %>% 
#   ungroup() %>% 
#   group_by(id) %>% 
#   mutate(best_match = ifelse(total_score == max(total_score),T,F)) %>%
#   filter(best_match == T) %>% 
#   select(doaj_id = id, discipline) %>% 
#   unique()

Create file with journals to manually classify

Code

# dbReadTable(db, "doaj") %>% 
#   filter(!id %in% dbReadTable(db, "journals_disciplines")$doaj_id) %>% 
#   writexl::write_xlsx("data/journals_to_classify.xlsx")

update data

Code

 # 
 # journals_disciplines <- dbReadTable(db, "journals_disciplines") %>% 
 #   bind_rows(read_xlsx("data/journals_to_classify.xlsx") %>% 
 #               select(doaj_id = id, discipline) %>% 
 #               unique())

Authorship data

Getting the authorship data from Google Big Query

Code

-- Query executed on March 8th 2024
/*
SELECT a.source_id,
       c.publication_year,
       d.institution_id,
       count(author_id) as n_publications
FROM `global-timer-415013.diamon_oa.openalex_sources`as a
JOIN insyspo.publicdb_openalex_2024_02_rm.works_primary_location as b
  on b.source_id = a.source_id
JOIN insyspo.publicdb_openalex_2024_02_rm.works as c
  on c.id = b.id
JOIN insyspo.publicdb_openalex_2024_02_rm.works_authorships as d
  on d.work_id = c.id
WHERE c.publication_year between 2019 and 2023 and c.type='article'
GROUP BY a.source_id, c.publication_year, d.institution_id"
*/

Import OpenAlex authorship data

Code

openalex_authorship <- read_csv("data/openalex_institutions_work_count_per_year.csv") %>% 
  mutate(source_id = str_remove(source_id, "https://openalex.org/S")) %>% 
  mutate(institution_id = str_remove(institution_id, "https://openalex.org/I"))

Import OpenAlex institution data

Code

openalex_institutions <- read_csv("data/openalex_institutions.csv") %>% 
  select(id, name = display_name, country_code, country) %>% 
  mutate(id = str_remove(id, "https://openalex.org/I"))

Analysis

Code

journals_disciplines <- read_csv("data/journals_disciplines.csv") %>% 
  bind_rows(read_xlsx("data/journals_to_classify.xlsx") %>% 
               select(doaj_id = id, discipline) %>% 
               unique())

doaj_coverage <- doaj_journals %>% 
  select(id) %>% 
  unique() %>% 
  mutate(coverage = case_when(
    id %in% doaj_scopus_matches$doaj_id ~ "WoS or Scopus",
    id %in% filter(doaj_wos_matches %>% 
                     inner_join(wos_journals_indexes, by="wos_id"),
                   index == "esci")$doaj_id ~ "WoS - ESCI",
    id %in% filter(doaj_wos_matches %>% 
                     inner_join(wos_journals_indexes, by="wos_id"),
                   index != "esci")$doaj_id ~ "WoS or Scopus",
    .default = "Not in WoS or Scopus"))

Coverage of DOAJ journals in OpenAlex, WoS and, Scopus

Overlap

Code

library("ggVennDiagram")
library("ggvenn")


plot_overlap_diamond <- list(
  OpenAlex = unique(doaj_journals %>% 
                      filter(oa_type == "Diamond") %>% 
                      inner_join(doaj_openalex_matches, 
                                 by=c("id" = "doaj_id")))$id,
  WoS = unique(doaj_journals %>% 
                      filter(oa_type == "Diamond") %>% 
                      inner_join(doaj_wos_matches, 
                                 by=c("id" = "doaj_id")))$id,
  Scopus = unique(doaj_journals %>% 
                      filter(oa_type == "Diamond") %>% 
                      inner_join(doaj_scopus_matches, 
                                 by=c("id" = "doaj_id")))$id
  ) %>% 
  ggvenn(stroke_size = 0.5) +
  ggtitle("Diamond OA") +
  theme(plot.title = element_text(size = 20, hjust = .5))


plot_overlap_gold <- list(
  OpenAlex = unique(doaj_journals %>% 
                      filter(oa_type == "Gold") %>% 
                      inner_join(doaj_openalex_matches, 
                                 by=c("id" = "doaj_id")))$id,
  WoS = unique(doaj_journals %>% 
                      filter(oa_type == "Gold") %>% 
                      inner_join(doaj_wos_matches, 
                                 by=c("id" = "doaj_id")))$id,
  Scopus = unique(doaj_journals %>% 
                      filter(oa_type == "Gold") %>% 
                      inner_join(doaj_scopus_matches, 
                                 by=c("id" = "doaj_id")))$id
  ) %>% 
  ggvenn(stroke_size = 0.5) +
  ggtitle("Gold OA") +
  theme(plot.title = element_text(size = 20, hjust = .5))

ggarrange(plot_overlap_diamond, plot_overlap_gold)

Code

ggsave("tables and figures/overlap.tiff", device = "tiff", width = 8, height = 6, units = "in", dpi=300)

Coverage

Code

plot_coverage <- doaj_journals %>%
  inner_join(doaj_coverage, by="id") %>% 
  inner_join(journals_disciplines, by=c("id"="doaj_id")) %>% 
  filter(discipline != 'n/a') %>% 
  mutate(group = str_c(oa_type, " - ", discipline)) %>% 
  unique() %>% 
  group_by(group) %>% 
  mutate(total = n()) %>% 
  group_by(group, coverage) %>% 
  mutate(n = n()) %>% 
  select(group, coverage, n, total) %>% 
  unique() %>% 
  mutate(p = n/total) %>% 
  ggplot() +
  aes(y=p, x=group, fill=coverage) +
  geom_bar(stat="identity") +
  ylab("Share of journals") +
  xlab("Domain and OA type") +
  theme_minimal() +
  theme(plot.title = element_text(size = 10, hjust = .5)) +
  theme(legend.title=element_blank()) +
  theme(legend.position = "top") + 
  scale_y_continuous(labels = scales::percent) +
  scale_fill_brewer(palette = "Dark2")

plot_coverage

Code

ggsave("tables and figures/coverage.tiff", device = "tiff", width = 8, height = 5, units = "in", dpi=300)

Publishers

Publisher region

Code

plot_publisher_region <- doaj_journals %>% 
  select(id, country_publisher_id, oa_type) %>% 
  unique() %>%
  inner_join(doaj_coverage, by="id") %>% 
  inner_join(doaj_countries_publishers, by=c("country_publisher_id"="id")) %>% 
  inner_join(countries, by="country_code") %>% 
  inner_join(journals_disciplines, by=c("id" = "doaj_id")) %>% 
  filter(discipline != "n/a") %>% 
  select(id, coverage, region, discipline, oa_type) %>%
  mutate(group = str_c(oa_type, " - ", discipline)) %>% 
  group_by(group) %>% 
  mutate(total = n()) %>%
  drop_na() %>% 
  group_by(group, region, coverage) %>% 
  reframe(n_journals = n(),
            p_journals = n_journals/total) %>% 
  unique() %>% 
  group_by(group, region) %>% 
  mutate(x=sum(p_journals)) %>% 
  ungroup() %>% 
  ggplot() +
  aes(x=p_journals, y=reorder(region, x), fill=coverage) +
  scale_x_continuous(labels = scales::percent, breaks=c(0,0.2,0.4,0.6,0.8,1)) +
  geom_bar(stat="identity") +
  ylab("Region") +
  xlab("Share of journals") +
  theme_minimal() +
  theme(panel.grid.minor = element_blank()) +
  theme(plot.title = element_text(size = 10, hjust = .5)) +
  theme(legend.title=element_blank()) +
  theme(legend.position = "top") +
  scale_fill_brewer(palette = "Dark2") +
  facet_wrap(facets = "group")


plot_publisher_region$data %>% 
  writexl::write_xlsx("tables and figures/publisher_region_data.xlsx")

plot_publisher_region

Code

ggsave("tables and figures/publisher_region.tiff", device = "tiff", width = 8, height = 5, units = "in", dpi=300)

Publisher income group

Code

plot_publisher_income_group <- doaj_journals %>% 
  select(id, country_publisher_id, oa_type) %>% 
  unique() %>%
  inner_join(doaj_coverage, by="id") %>% 
  inner_join(doaj_countries_publishers, by=c("country_publisher_id"="id")) %>% 
  inner_join(countries, by="country_code") %>% 
  inner_join(journals_disciplines, by=c("id" = "doaj_id")) %>% 
  filter(discipline != "n/a") %>% 
  select(id, coverage, income_group, discipline, oa_type) %>%
  mutate(group = str_c(oa_type, " - ", discipline)) %>% 
  group_by(group) %>% 
  mutate(total = n()) %>%
  drop_na() %>% 
  group_by(group, income_group, coverage) %>% 
  reframe(n_journals = n(),
            p_journals = n_journals/total) %>% 
  unique() %>% 
  group_by(group, income_group) %>% 
  mutate(x=sum(p_journals)) %>% 
  ungroup() %>%
  mutate(income_group = factor(income_group, 
                               levels=c("Low income", 
                                        "Lower middle income", 
                                        "Upper middle income",
                                        "High income"))) %>%

  ggplot() +
  aes(x=p_journals, y=income_group, fill=coverage) +
  scale_x_continuous(labels = scales::percent) +
  geom_bar(stat="identity") +
  ylab("Income group") +
  xlab("Share of journals") +
  theme_minimal() +
  theme(panel.grid.minor = element_blank()) +
  theme(plot.title = element_text(size = 10, hjust = .5)) +
  theme(legend.title=element_blank()) +
  theme(legend.position = "top") +
  scale_fill_brewer(palette = "Dark2") +
  facet_wrap(facets = "group")

plot_publisher_income_group

Code

plot_publisher_income_group$data %>% 
  writexl::write_xlsx("tables and figures/publisher_income_group_data.xlsx")

ggsave("tables and figures/publisher_income_group.tiff", device = "tiff", width = 8, height = 5, units = "in", dpi=300)

Publishing language

Here we divide journals in three groups:

English journals only publish in English
Multiple languages journals publish in English and at least one more language.
Non-English journals do not publish in English.

Code

n_lang <- doaj_languages %>% 
  group_by(doaj_id) %>%
  summarize(n_lang = n())

journal_language<-doaj_languages %>% 
  filter(language == "English") %>% 
  filter(doaj_id %in% filter(n_lang, n_lang == 1)$doaj_id) %>% 
  select(doaj_id) %>% 
  mutate(language = "English") %>% 
  unique()

# Classify journals in one of the three language groups
journal_language<-bind_rows(journal_language,
                            doaj_languages %>% 
                              filter(language == "English") %>% 
                              filter(doaj_id %in% filter(n_lang, n_lang > 1)$doaj_id) %>% 
                            select(doaj_id) %>% 
                            mutate(language = "Multiple Languages") %>% 
                            unique())

journal_language<-bind_rows(journal_language,
                          doaj_languages %>% 
                            filter(!doaj_id %in% journal_language$doaj_id) %>% 
                            select(doaj_id) %>% 
                            mutate(language = "Non-English") %>% 
                            unique())
                            

plot_publishing_language <- doaj_journals %>% 
  select(id, oa_type) %>% 
  unique() %>% 
  inner_join(doaj_coverage, by="id") %>% 
  inner_join(journal_language, by=c("id"="doaj_id")) %>% 
  inner_join(journals_disciplines, by=c("id"="doaj_id")) %>% 
  filter(discipline != "n/a") %>% 
  select(id, coverage, language, discipline, oa_type) %>%
  mutate(group = str_c(oa_type, " - ", discipline)) %>% 
  unique() %>% 
  group_by(group) %>% 
  mutate(total = n()) %>%
  ungroup() %>% 
  drop_na() %>% 
  group_by(group, language, coverage) %>% 
  reframe(n_journals = n(),
            p_journals = n_journals/total) %>% 
  ungroup() %>% 
  unique() %>% 
  group_by(language) %>% 
  mutate(total_language=sum(n_journals)) %>% 
  arrange(desc(total_language)) %>% 
  ungroup() %>% 
  ggplot() +
  aes(x=p_journals, y=language, fill=coverage) +
  scale_x_continuous(labels = scales::percent) +
  geom_bar(stat="identity") +
  ylab("Language") +
  xlab("Share of journals") +
  theme_minimal() +
  theme(panel.grid.minor = element_blank()) +
  theme(plot.title = element_text(size = 10, hjust = .5)) +
  theme(legend.title=element_blank()) +
  theme(legend.position = "top") +
  scale_fill_brewer(palette = "Dark2") +
  facet_wrap(facets = "group")

plot_publishing_language

Code

plot_publishing_language$data %>% 
  writexl::write_xlsx("tables and figures/publisher_language_data.xlsx")

ggsave("tables and figures/publishing_language.tiff", device = "tiff", width = 8, height = 5, units = "in", dpi=300)

Publisher countries (Gini)

Code

table_publisher_countries_gini <- 
  data <- doaj_journals %>% 
  select(id, country_publisher_id, oa_type) %>% 
  unique() %>% 
  inner_join(doaj_coverage, by="id") %>% 
  inner_join(doaj_countries_publishers, by=c("country_publisher_id"="id")) %>% 
  inner_join(journals_disciplines, by=c("id"="doaj_id")) %>% 
  filter(discipline != "n/a")


write_xlsx(table_publisher_countries_gini %>% 
             group_by(discipline, oa_type, coverage, country_code) %>% 
             reframe(n_journals = n()),
           "tables and figures/data_journals_by_country.xlsx")


# Creates a combination of every journal and country code
x<-expand(countries,
          select(table_publisher_countries_gini, 
                 id, coverage, discipline, oa_type),
          country_code) %>% 
  unique()

table_publisher_countries_gini <-x %>% 
  left_join(table_publisher_countries_gini %>%
              select(id, country_code) %>% 
              mutate(n = 1), 
            by=c("id", "country_code")) %>% 
  mutate(n = ifelse(is.na(n),0,1)) %>% 
  unique() %>% 
  group_by(coverage, oa_type, discipline, country_code) %>% 
  reframe(n = sum(n)) %>% 
  group_by(coverage, oa_type, discipline) %>% 
  summarize(gini = ineq(n,type="Gini")) %>% 
  pivot_wider(names_from = "coverage", values_from = "gini") %>% 
  drop_na()

table_publisher_countries_gini %>% 
  kbl()

oa_type	discipline	Not in WoS or Scopus	WoS - ESCI	WoS or Scopus
Diamond	BM	0.9081304	0.9382768	0.9181212
Diamond	NSE	0.9039658	0.9169850	0.8948191
Diamond	SSH	0.9086776	0.9378867	0.9055282
Gold	BM	0.9461987	0.9705903	0.9615593
Gold	NSE	0.9524853	0.9605911	0.9496604
Gold	SSH	0.9591767	0.9583750	0.9440742

Authorship

Author region

Code

plot_authorship_region <- doaj_journals %>% 
  inner_join(doaj_openalex_matches, by=c("id"="doaj_id")) %>% 
  left_join(openalex_authorship, by=c("openalex_id"="source_id")) %>% 
  left_join(openalex_institutions, by=c("institution_id" = "id")) %>% 
  left_join(countries, by="country_code") %>% 
  inner_join(doaj_coverage, by="id") %>% 
  inner_join(journals_disciplines, by=c("id" = "doaj_id")) %>% 
  filter(discipline != "n/a") %>% 
  mutate(group = str_c(oa_type, " - ", discipline)) %>%
  group_by(group) %>% 
  drop_na() %>% 
  mutate(total = sum(n_publications)) %>%
  group_by(group, region, coverage) %>% 
  reframe(n_authorships = sum(n_publications),
            p_authorships = n_authorships/total) %>% 
  unique() %>% 
  group_by(group, region) %>% 
  mutate(x=sum(p_authorships)) %>% 
  ungroup() %>% 
  ggplot() +
  aes(x=p_authorships, y=reorder(region, x), fill=coverage) +
  scale_x_continuous(labels = scales::percent, breaks=c(0,0.2,0.4,0.6,0.8,1)) +
  geom_bar(stat="identity") +
  ylab("Region") +
  xlab("Share of authorships") +
  theme_minimal() +
  theme(panel.grid.minor = element_blank()) +
  theme(plot.title = element_text(size = 10, hjust = .5)) +
  theme(legend.title=element_blank()) +
  theme(legend.position = "top") +
  scale_fill_brewer(palette = "Dark2") +
  facet_wrap(facets = "group")

plot_authorship_region

Code

plot_publisher_region$data %>% 
  writexl::write_xlsx("tables and figures/authorships_region_data.xlsx")


ggsave("tables and figures/authorship_region.tiff", device = "tiff", width = 8, height = 5, units = "in", dpi=300)

Author income group

Code

plot_authorship_income_group <- doaj_journals %>% 
  inner_join(doaj_openalex_matches, by=c("id"="doaj_id")) %>% 
  left_join(openalex_authorship, by=c("openalex_id"="source_id")) %>% 
  left_join(openalex_institutions, by=c("institution_id" = "id")) %>% 
  left_join(countries, by="country_code") %>% 
  inner_join(doaj_coverage, by="id") %>% 
  inner_join(journals_disciplines, by=c("id" = "doaj_id")) %>% 
  filter(discipline != "n/a") %>% 
  mutate(group = str_c(oa_type, " - ", discipline)) %>%
  group_by(group) %>% 
  drop_na() %>% 
  mutate(total = sum(n_publications)) %>%
  group_by(group, income_group, coverage) %>% 
  reframe(n_authorships = sum(n_publications),
            p_authorships = n_authorships/total) %>% 
  unique() %>% 
  group_by(group, income_group) %>% 
  mutate(x=sum(p_authorships)) %>% 
  ungroup() %>% 
  ggplot() +
  aes(x=p_authorships, y=reorder(income_group, x), fill=coverage) +
  scale_x_continuous(labels = scales::percent, breaks=c(0,0.2,0.4,0.6,0.8,1)) +
  geom_bar(stat="identity") +
  ylab("Income group") +
  xlab("Share of authorships") +
  theme_minimal() +
  theme(panel.grid.minor = element_blank()) +
  theme(plot.title = element_text(size = 10, hjust = .5)) +
  theme(legend.title=element_blank()) +
  theme(legend.position = "top") +
  scale_fill_brewer(palette = "Dark2") +
  facet_wrap(facets = "group")

plot_authorship_income_group$data %>% 
  writexl::write_xlsx("tables and figures/authorships_income_group_data.xlsx")

plot_authorship_income_group

Code

ggsave("tables and figures/authorship_income_group.tiff", device = "tiff", width = 8, height = 5, units = "in", dpi=300)

Author countries (Gini)

Code

countries_in_data <- openalex_authorship %>% 
  inner_join(openalex_institutions, by=c("institution_id"="id")) %>% 
  select(country_code) %>% 
  drop_na() %>% 
  unique()

authorship_gini_plot <- doaj_journals %>% 
  inner_join(doaj_openalex_matches, by=c("id"="doaj_id")) %>% 
  left_join(openalex_authorship, by=c("openalex_id"="source_id")) %>% 
  left_join(openalex_institutions, by=c("institution_id" = "id")) %>% 
  left_join(journals_disciplines, by=c("id"="doaj_id")) %>% 
  left_join(doaj_coverage, by="id") %>% 
  select(id, discipline, oa_type, coverage, country_code, n_publications) %>% 
  group_by(id, discipline, oa_type, coverage, country_code) %>% 
  reframe(n = sum(n_publications))

write_xlsx(authorship_gini_plot %>% 
             group_by(discipline, oa_type, coverage, country_code) %>% 
             reframe(total_authorship = sum(n)),
           "tables and figures/data_authorships_by_country.xlsx")

x<-expand(countries,
          doaj_journals %>% select(id),
          country_code) %>% 
  inner_join(countries_in_data, by="country_code")

authorship_gini_plot <-x %>% 
  left_join(authorship_gini_plot, by=c("id","country_code")) %>% 
  mutate(n = ifelse(is.na(n),0,n)) %>% 
  group_by(id) %>% 
  mutate(gini = ineq(n,type="Gini")) %>% 
  drop_na() %>% 
  filter(discipline != "n/a") %>% 
  mutate(discipline = as.factor(discipline),
         coverage = as.factor(coverage),
         oa_type = as.factor(oa_type)) %>% 
  unique()

authorship_gini_plot %>%
  select(id, coverage, oa_type, discipline, gini) %>% 
  filter(discipline != 'n/a') %>% 
  unique() %>% 
  mutate(coverage = as.character(coverage)) %>% 
  mutate(coverage = ifelse(coverage == "Not in WoS or Scopus",
                           "OpenAlex only",
                           coverage)) %>% 
  ggplot() +
  aes(x=coverage, y = gini, fill = oa_type) +
  geom_boxplot() +
  facet_wrap(facets = "discipline") +
  ylab("Gini coefficient") +
  xlab("coverage") +
  theme_minimal() +
  theme(legend.position = "top") +
  scale_fill_brewer(palette = "Dark2") +
  theme(axis.text.x = element_text(angle = 45)) + 
  theme(axis.text.x = element_text(margin = margin(t = 20))) +
  theme(legend.title = element_blank()) +
  theme(axis.title.x = element_text(margin = margin(t = -15)))

Code

ggsave("tables and figures/authorship_gini.tiff", device = "tiff", width = 8, height = 5, units = "in", dpi=300)

Kruskall-Wallis test

Code

authorship_gini_plot<-authorship_gini_plot %>%
  ungroup() %>% 
  mutate(group = str_c(oa_type, " - ", coverage)) %>% 
  mutate(group = as.factor(group))

authorship_gini_plot %>% 
  group_by(discipline) %>% 
  kruskal_test(gini ~ group) %>% 
  select(discipline, n, statistic, df, p) %>% 
  inner_join(authorship_gini_plot %>%
               group_by(discipline) %>% 
               kruskal_effsize(gini ~ group) %>% 
               select(discipline, effsize, magnitude),
             by="discipline") %>% 
  kbl()

discipline	n	statistic	df	effsize	magnitude
BM	140356	34698.79	5	0.2471948	large
NSE	127369	27503.77	5	0.2159086	large
SSH	99600	35145.65	5	0.3528391	large

Pairwise Wilcoxon test

Code

authorship_gini_plot %>% 
  filter(discipline == "BM") %>% 
  wilcox_test(gini ~ group, p.adjust.method = "bonferroni") %>% 
  select(-1) %>% 
  kbl(caption = "Pairwise Wilcoxon test for BM journals")

Pairwise Wilcoxon test for BM journals
group1	group2	n1	n2	statistic	p.adj.signif
Diamond - Not in WoS or Scopus	Diamond - WoS - ESCI	10551	1876	13696012	****
Diamond - Not in WoS or Scopus	Diamond - WoS or Scopus	10551	21620	165080846	****
Diamond - Not in WoS or Scopus	Gold - Not in WoS or Scopus	10551	8705	67648676	****
Diamond - Not in WoS or Scopus	Gold - WoS - ESCI	10551	3683	33416252	****
Diamond - Not in WoS or Scopus	Gold - WoS or Scopus	10551	93921	899097033	****
Diamond - WoS - ESCI	Diamond - WoS or Scopus	1876	21620	22568556	****
Diamond - WoS - ESCI	Gold - Not in WoS or Scopus	1876	8705	10267316	****
Diamond - WoS - ESCI	Gold - WoS - ESCI	1876	3683	5391716	****
Diamond - WoS - ESCI	Gold - WoS or Scopus	1876	93921	149528659	****
Diamond - WoS or Scopus	Gold - Not in WoS or Scopus	21620	8705	107084083	****
Diamond - WoS or Scopus	Gold - WoS - ESCI	21620	3683	56381925	****
Diamond - WoS or Scopus	Gold - WoS or Scopus	21620	93921	1588441118	****
Gold - Not in WoS or Scopus	Gold - WoS - ESCI	8705	3683	19549852	****
Gold - Not in WoS or Scopus	Gold - WoS or Scopus	8705	93921	562692469	****
Gold - WoS - ESCI	Gold - WoS or Scopus	3683	93921	206964867	****

Code

authorship_gini_plot %>% 
  filter(discipline == "NSE") %>% 
  wilcox_test(gini ~ group, p.adjust.method = "bonferroni") %>% 
  select(-1) %>% 
  kbl(caption = "Pairwise Wilcoxon test for NSE journals")

Pairwise Wilcoxon test for NSE journals
group1	group2	n1	n2	statistic	p.adj.signif
Diamond - Not in WoS or Scopus	Diamond - WoS - ESCI	11329	1623	10518550	****
Diamond - Not in WoS or Scopus	Diamond - WoS or Scopus	11329	25271	224415720	****
Diamond - Not in WoS or Scopus	Gold - Not in WoS or Scopus	11329	9673	73156915	****
Diamond - Not in WoS or Scopus	Gold - WoS - ESCI	11329	1565	14653520	****
Diamond - Not in WoS or Scopus	Gold - WoS or Scopus	11329	77908	779928144	****
Diamond - WoS - ESCI	Diamond - WoS or Scopus	1623	25271	32241307	****
Diamond - WoS - ESCI	Gold - Not in WoS or Scopus	1623	9673	10300465	****
Diamond - WoS - ESCI	Gold - WoS - ESCI	1623	1565	2128887	****
Diamond - WoS - ESCI	Gold - WoS or Scopus	1623	77908	114633754	****
Diamond - WoS or Scopus	Gold - Not in WoS or Scopus	25271	9673	104319919	****
Diamond - WoS or Scopus	Gold - WoS - ESCI	25271	1565	22707667	****
Diamond - WoS or Scopus	Gold - WoS or Scopus	25271	77908	1349522587	****
Gold - Not in WoS or Scopus	Gold - WoS - ESCI	9673	1565	9622416	****
Gold - Not in WoS or Scopus	Gold - WoS or Scopus	9673	77908	549418991	****
Gold - WoS - ESCI	Gold - WoS or Scopus	1565	77908	75215917	****

Code

authorship_gini_plot %>% 
  filter(discipline == "SSH") %>% 
  wilcox_test(gini ~ group, p.adjust.method = "bonferroni") %>%
  select(-1) %>% 
  kbl(caption = "Pairwise Wilcoxon test for SSH journals")

Pairwise Wilcoxon test for SSH journals
group1	group2	n1	n2	statistic	p	p.adj	p.adj.signif
Diamond - Not in WoS or Scopus	Diamond - WoS - ESCI	36878	6617	153026523	0.000	0	****
Diamond - Not in WoS or Scopus	Diamond - WoS or Scopus	36878	29555	831561731	0.000	0	****
Diamond - Not in WoS or Scopus	Gold - Not in WoS or Scopus	36878	8390	156609211	0.078	1	ns
Diamond - Not in WoS or Scopus	Gold - WoS - ESCI	36878	1478	44793631	0.000	0	****
Diamond - Not in WoS or Scopus	Gold - WoS or Scopus	36878	16682	570289852	0.000	0	****
Diamond - WoS - ESCI	Diamond - WoS or Scopus	6617	29555	131037690	0.000	0	****
Diamond - WoS - ESCI	Gold - Not in WoS or Scopus	6617	8390	22912151	0.000	0	****
Diamond - WoS - ESCI	Gold - WoS - ESCI	6617	1478	7386899	0.000	0	****
Diamond - WoS - ESCI	Gold - WoS or Scopus	6617	16682	99053391	0.000	0	****
Diamond - WoS or Scopus	Gold - Not in WoS or Scopus	29555	8390	78124342	0.000	0	****
Diamond - WoS or Scopus	Gold - WoS - ESCI	29555	1478	27412302	0.000	0	****
Diamond - WoS or Scopus	Gold - WoS or Scopus	29555	16682	398847420	0.000	0	****
Gold - Not in WoS or Scopus	Gold - WoS - ESCI	8390	1478	9274796	0.000	0	****
Gold - Not in WoS or Scopus	Gold - WoS or Scopus	8390	16682	121138229	0.000	0	****
Gold - WoS - ESCI	Gold - WoS or Scopus	1478	16682	17471720	0.000	0	****

Data sources

Create database

Data processing

Country income level

DOAJ

Import DOAJ journal list

Publishing countries

DOAJ journals

Language

ISSNs

Journal titles

Scopus

Import Scopus journal list

Journals

ISSNs

Journal titles

Web of Science

Import WOS journal list

Journals

Indexes

ISSNs

Openalex

Import OpenAlex journal data

OpenAlex journals

ISSNs

Journal matching

DOAJ and WoS

DOAJ and Scopus

DOAJ and OpenAlex

Checks and manual matches

OpenAlex

Add manual matches to data

Scopus

Add manual matches to data

Discipline classification

Automatic classification based on previous classification

Create file with journals to manually classify

update data

Authorship data

Getting the authorship data from Google Big Query

Import OpenAlex authorship data

Import OpenAlex institution data

Analysis

Coverage of DOAJ journals in OpenAlex, WoS and, Scopus

Overlap

Coverage

Publishers

Publisher region

Publisher income group

Publishing language

Publisher countries (Gini)

Authorship

Author region

Author income group

Author countries (Gini)

Kruskall-Wallis test

Pairwise Wilcoxon test

References