Scientific journals operate over a broad spectrum of publishing strategies, from strictly for-profit, to non-profit, and in-between business models (e.g. for-profit but academic friendly journals). Scientific publishing is increasingly dominated by for-profit journals, many of which attract prestige and submissions through high impact factors. In contrast, non-profit journals – those that reinvest revenue into the academic community – struggle to maintain visibility despite offering more equitable publishing models.
The R package fairpub aims to provide a user-friendly
toolbox to investigate the fairness of a research (article,
bibliographic list, citation list, etc.). The fairness is measured
according to two dimensions:
A journal with a non-profit business model is fairer than an academic friendly journal with a for-profit business model. But the later is still fairer than a non-academic friendly journal with a for-profit business model.
This information comes from the DAFNEE initiative, a Database of Academia Friendly jourNals in Ecology and Evolution.
Note: The package fairpub provides a
subset of the original DAFNEE database. We are currently working to
increase this list of journals.
The package fairpub also implements the method proposed
by Beck et al. (in revision): the strategic citation. By
deliberately choosing to cite relevant articles from non-profit journals
when multiple references would be equally valid, researchers can
contribute to increasing their visibility and future impact factor. This
method is implemented in the fp_compute_ratio() function
and can answer the question How fair am I when I cite previous
works? by computing the fairness ratio on the references cited
in a manuscript.
The fp_compute_ratio() function can also be used to
answer the question How fair is my publication list? by
computing the fairness ratio on the bibliography of an author.
Let’s explore the features of fairpub but first let’s
load and attach the fairpub package:
Some functions of fairpub query the OpenAlex bibliographic database by
using the openalexR
package. As recommended by the OpenAlex API documentation,
it’s a good practice to send your email address to each query made by
openalexR.
Use the function options() to store your email address
as a global variable available for the current session:
options(openalexR.mailto = 'your_email@mail.com')Note: No registration is required on the OpenAlex website.
The function fp_journal_fairness() of the package
fairpub queries the DAFNEE database to retrieve the
fairness of a journal.
fp_journal_fairness(journal = "Science")
#> journal fairness
#> 1 Science Non-profit and academic friendlyThis function returns a two-column data.frame with the
name of the journal and the fairness status of the journal. This
fairness status can have the following three values:
This function searches for an exact match (case insensitive) but if the journal is not found, a fuzzy search returns the three best candidates (low distance between characters). For instance:
fp_journal_fairness(journal = "Science of Nature")
#> No exact match found!
#> The fuzzy search returns these three best candidates:
#> 'The Science of Nature'
#> 'Science Advances'
#> 'People and Nature'Then, user can run this same function with the correct spelling:
fp_journal_fairness(journal = "The Science of Nature")
#> journal fairness
#> 1 The Science of Nature For-profit and non-academic friendlyAs mentioned before, the list of DAFNEE journals included in
fairpub is not exhaustive, and it’s possible that many
journals are missing.
The function fp_article_fairness() of the package
fairpub is dedicated to retrieve the fairness of a
scientific article. This function first queries the OpenAlex bibliographic database to find
the journal name from the DOI (digital object identifier) of the
article. Then, the journal name is used to query the DAFNEE database to
get the fairness status.
Note: Even if this function works at the article level, this fairness status is associated to the journal. In other words, all articles published in a journal have the same fairness status.
fp_article_fairness(doi = "10.1126/science.162.3859.1243")
#> journal fairness
#> 1 Science Non-profit and academic friendlyThe output is identical to the output of
fp_journal_fairness(), i.e. a two-column
data.frame.
Note: if the DOI is not found in OpenAlex, the output will be:
fp_article_fairness(doi = "10.xxxx/xxxx")
#> journal fairness
#> 1 NA Record not found in OpenAlexNote: if the DOI is found in OpenAlex but the journal is absent from DAFNEE, the output will be:
fp_article_fairness(doi = "10.21105/joss.05753")
#> journal fairness
#> 1 The Journal of Open Source Software Record not found in DAFNEE databaseOne core function of fairpub is
fp_compute_ratio(). It computes the fairness ratio of a
list of references cited in a manuscript. This ratio can be used by the
author to replace some for-profit citations by fairer citations (e.g.
citations of articles published in non-profit journals). Of course, if
the findings of two studies are similar.
This function uses one argument: a vector of DOI. The
fairpub package provides the
fp_doi_from_bibtex() function that reads a BibTeX file and
extracts all available DOI.
Let’s use the function fp_doi_from_bibtex() with the
BibTeX file provided as an example in fairpub.
# Path to the BibTeX provided by 'fairpub' ----
filename <- system.file(
file.path("extdata", "references.bib"),
package = "fairpub"
)
# Extract DOI from this BibTeX file ----
doi_list <- fp_doi_from_bibtex(filename)
doi_list
#> [1] "10.1098/rsos.160384" NA
#> [3] "10.1126/science.1212540" "10.9745/ghsp-d-21-00145"
#> [5] "10.1126/science.adk9900" "10.1016/j.ecolecon.2021.107082"
#> [7] "10.1177/014107680609900316" "10.1093/reseval/rvad012"
#> [9] "10.1111/ele.14395" "10.3998/ptpbio.3363"
#> [11] "10.4000/proceedings.elpub.2018.30" "10.1093/scipol/scs093"
#> [13] NA "10.1002/leap.1102"
#> [15] "10.1087/0953151053584975" NA
#> [17] "10.1177/0263395719858571" "10.1017/s1062798709000532"
#> [19] "10.1371/journal.pone.0127502" "10.1162/qss_c_00305"
#> [21] "10.1007/978-3-030-02511-3_1" NA
#> [23] "10.1371/journal.pbio.1002264" "10.3389/frma.2016.00007"
#> [25] "10.48550/arxiv.2407.16551" NA
#> [27] "10.1080/08109028.2014.891710" "10.1007/s11192-022-04586-1"
#> [29] "10.1098/rspb.2019.2047" "10.5281/zenodo.4558704"
#> [31] "10.1162/qss_a_00272" "10.3917/inno.063.0095"
#> [33] "10.1371/journal.pone.0243664" "10.1257/jep.15.4.183"
#> [35] "10.1073/pnas.0305628101" "10.1016/j.amjmed.2019.07.028"
#> [37] "10.32614/rj-2023-089" "10.5534/wjmh.230001"Note: Make sure to check your references, especially
the DOI field. In this example, some references don’t have a DOI (book,
book chapter, etc.). The fp_doi_from_bibtex() function
don’t remove these references to provide a full report (see below the
output of the fp_compute_ratio() function).
Now we have a vector of DOI, we can use the
fp_compute_ratio() function to compute the fairness ratio
of this list.
fp_compute_ratio(doi = doi_list)
#> $summary
#> metric value
#> 1 Total references 38
#> 2 References with DOI 33
#> 3 Deduplicated references 33
#> 4 References found in OpenAlex 30
#> 5 References found in DAFNEE 10
#> 6 Non-profit and academic friendly references 9
#> 7 For-profit and academic friendly references 1
#> 8 For-profit and non-academic friendly references 0
#>
#> $ratios
#> Non-profit and academic friendly For-profit and academic friendly
#> 0.9 0.1
#> For-profit and non-academic friendly
#> 0.0This function returns a two-element list:
summary: a two-column data.frame with the
number of references remaining after each step of the workflowratio: a three-element vector with the
value of the three fairness ratiosIn this example, this list of references has a fairness ratio
(Non-profit and academic friendly) of 90%. But this value
must be interpreted with caution. Indeed this ratio has been computed on
26% (10 over 38) of the references, because the journal of 20 articles
is not indexed in the DAFNEE database.
As mentioned before, we are currently working to improve the list of journals indexed in the DAFNEE database.
It’s possible to use the fp_compute_ratio() function to
compute the fairness ratio of all publications of an author. In that
case, the list of DOI should contain all articles published by this
author.