Working hard to hardly work using Quarto parametrised reports
This report describes the state of scientific articles that have been retracted, focusing on researchers from India
Author
R4DEV
Published
March 19, 2024
Tracking retractions as a window into the scientific process
Retractions worldwide ⚠️
Retraction Watch plays an important role in the scientific community by tracking and reporting on retractions of research papers. They help maintain the integrity of the scientific record by highlighting instances where research may be flawed, fraudulent, or unethical, preventing the proliferation of incorrect information. Additionally, by analyzing and reporting on the reasons for retractions, Retraction Watch shows some common pitfalls in research practices, encouraging researchers, publishers, and institutions to adopt higher standards.
The number of retractions identified by Retraction Watch (@Oransky and Marcus (2024)) reached 52219. In this report we will focus on describing retractions of articles published by authors from India.
The data recently became public thanks to CrossRef (Rosa-Clark (n.d.)) and you can directly download it from this link.
Retraction Watch has recently been featured in many articles (“Why fake research is rampant in china” (n.d.), Van Noorden (2023)). Their work has been Check out the map of retractions below (Figure 1).
Figure 1
What’s going on in India?
Retractions for authors from India reached 15634, and represented 0.2993929 % of all retractions registered. How does India compare to other countries? What are the main reasons articles written by authors from India are retracted? Who are the most retracted authors in India?
Putting India in an international perspective. India ranks 3 in retractions (Figure 2).
Figure 2
Why were the articles retracted? The most frequent reason authors from India have their articles retracted is +Investigation by Journal/Publisher (Figure 3)
Figure 3
Who are the most retracted authors in India. Let’s look at specific researchers (Figure 4).
Figure 4
Which institutions host the most retracted authors in India? (Figure 5).
Figure 5
References
I. Oransky and A. Marcus. (18 March 2024), “Retraction watch. Retraction watch,”https://retractionwatch.com/ (accessed March 19, 2024).
R. Van Noorden. (2023), “More than 10,000 research papers were retracted in 2023 — a new record,”Nature, Vol. 624/7992, pp. 479–481, https://doi.org/10.1038/d41586-023-03974-8.
---title: "RetractionWatch"subtitle: "Working hard to hardly work using Quarto parametrised reports"description: "This report describes the state of scientific articles that have been retracted, focusing on researchers from `r params$country`"author: "R4DEV"date: last-modifiedparams: country: "France"bibliography: RetractionWatch.bibcsl: oecd.csllink-citations: true---> *Tracking retractions as a window into the scientific process*```{r raw-data}#| include: falselibrary(tidyverse)# Data ####retractionwatch_raw <- readr::read_csv("retractions.csv") |> janitor::clean_names() # Separate authors by country and reasons for retraction ###retractionwatch_df <- retractionwatch_raw |> tidyr::separate_rows(country ,sep=";") |> tidyr::separate_rows(reason, sep=";") |> dplyr::mutate(iso3c = countrycode::countrycode(sourcevar = country, origin = "country.name", destination = "iso3c")) |> dplyr::mutate(country = str_remove_all(country,"'"))# Summaries ####total_retractions <- n_distinct(retractionwatch_raw)retractions_country <- retractionwatch_df |> dplyr::count(country) |> dplyr::filter(country==params$country) |> dplyr::select(n) |> as.integer()country_prop <- retractions_country/total_retractions |> round(digits=2)country_rank <- retractionwatch_df |> dplyr::count(country) |> dplyr::mutate(rank = dense_rank(desc(n))) |> dplyr::filter(country==params$country) |> dplyr::select(rank) |> as.integer()country_top_reason <- retractionwatch_df |> dplyr::filter(country==params$country) |> dplyr::filter(reason!="", !is.na(reason)) |> dplyr::count(reason) |> dplyr::slice_max(order_by = n, n = 1) |> dplyr::select(reason) |> as.character()```# Retractions worldwide ⚠️**Retraction Watch** plays an important role in the scientific community by tracking and reporting on retractions of research papers. They help maintain the integrity of the scientific record by highlighting instances where research may be flawed, fraudulent, or unethical, preventing the proliferation of incorrect information. Additionally, by analyzing and reporting on the reasons for retractions, **Retraction Watch** shows some common pitfalls in research practices, encouraging researchers, publishers, and institutions to adopt higher standards.The number of retractions identified by [**Retraction Watch**](https://retractionwatch.com/) (\@@oransky_retraction_2024) reached `r total_retractions`. In this report we will focus on describing retractions of articles published by authors from `r params$country`.The data recently became public thanks to CrossRef (@rosa-clark_news_nodate) and you can directly download it from [this link](https://www.crossref.org/blog/news-crossref-and-retraction-watch/).**Retraction Watch** has recently been featured in many articles (@the_economist, @van_noorden_more_2023). Their work has been Check out the map of retractions below (@fig-map).```{r map}#| label: fig-maplibrary(ggiraph)library(sf)library(tidyverse)library(countrycode) library(ggthemes)library(gapminder)library(rnaturalearth)library(ggiraph)library(ggthemes)library(viridis)# Map ####world_sf <- rnaturalearth::ne_countries(returnclass = "sf", scale="medium") |> dplyr::select(iso3c=adm0_a3, geometry)# Data for parametrised country ####retractionwatch_param_country <- retractionwatch_df |> dplyr::count(country) |> dplyr::filter(country==params$country) |> dplyr::mutate(iso3c = countrycode::countrycode(sourcevar = country, origin = "country.name", destination = "iso3c")) |> dplyr::left_join(world_sf, by=c("iso3c")) |> dplyr::mutate(centroid = st_centroid(geometry)) |> dplyr::mutate(coord = map(centroid, ~st_coordinates(.) |> as_tibble())) |> tidyr::unnest_wider(col = coord, names_sep = "_")# ggplot ####retractionwatch_map_gg <- retractionwatch_df |> dplyr::count(country) |> dplyr::mutate(iso3c = countrycode::countrycode(sourcevar = country, origin = "country.name", destination = "iso3c")) |> dplyr::left_join(world_sf, by=c("iso3c")) |> ggplot()+ ggiraph::geom_sf_interactive( aes(geometry=geometry, fill=n, data_id=country, tooltip=paste0(country, "<br>Authors involved in retractions: ", n |> round(digits=0) |> format(big.mark=" "))))+ ggiraph::geom_label_repel_interactive( data = retractionwatch_param_country, show.legend = FALSE, aes(x = coord_X, y = coord_Y, label = country ))+ ggiraph::scale_fill_gradient_interactive( transform = "log10", label = scales::label_number(), name = NULL, low = "royalblue1", high = "red3", na.value = "grey50")+ ggthemes::theme_map()# Girafe ####ggiraph::girafe(ggobj = retractionwatch_map_gg)```# What's going on in `r params$country`?Retractions for authors from `r params$country` reached `r retractions_country`, and represented `r country_prop` % of all retractions registered. How does `r params$country` compare to other countries? What are the main reasons articles written by authors from `r params$country` are retracted? Who are the most retracted authors in `r params$country`?::: {.panel-tabset .nav-pills}## 🗺️ ComparisonsPutting `r params$country` in an international perspective. `r params$country` ranks `r country_rank` in retractions (@fig-top).```{r top}#| label: fig-toplibrary(ggiraph)library(tidyverse)library(countrycode) library(ggiraph)# First build the pipe for ggplot ####retraction_country_gg <- retractionwatch_df |> dplyr::count(country) |> dplyr::filter(country==params$country) #retraction_top_countries_gg <- retractionwatch_df |> dplyr::count(country) |> dplyr::mutate(iso3c = countrycode::countrycode( sourcevar = country, origin = "country.name", destination = "iso3c")) |> dplyr::slice_max(order_by = n, n = 10) |> ggplot(aes(x=n,y=country |> reorder(n),fill=country))+ geom_col_interactive( aes(fill=n, data_id=country, tooltip=paste0(country, "<br>Authors involved in retractions: ", n |> round(digits=0) |> format(big.mark=" "))))+ geom_col_interactive( data = retraction_country_gg, aes(fill=n, data_id=country, tooltip=paste0(country, "<br>Authors involved in retractions: ", n |> round(digits=0) |> format(big.mark=" "))))+ scale_fill_distiller_interactive( palette = "Set1", na.value = "grey50")+ labs(y=NULL)+ theme_classic()+ theme(legend.position = "none")# Then ggiraph #### ggiraph::girafe( ggobj = retraction_top_countries_gg)```## 🤔 ReasonsWhy were the articles retracted? The most frequent reason authors from `r params$country` have their articles retracted is *`r country_top_reason`* (@fig-reasons)```{r reasons}#| label: fig-reasonslibrary(ggiraph)library(tidyverse)library(countrycode) library(ggiraph)# First build the pipe for ggplot ####retraction_reasons_gg <- retractionwatch_df |> dplyr::filter(country==params$country) |> dplyr::count(reason) |> dplyr::filter(reason!="", !is.na(reason)) |> dplyr::slice_max(order_by = n, n = 15) |> ggplot(aes(x=n,y=reason |> reorder(n),fill=reason))+ geom_col_interactive( aes(fill=n, data_id=reason, tooltip=paste0(reason, "<br>Reasons for retractions: ", n |> round(digits=0) |> format(big.mark=" "))))+ scale_fill_distiller_interactive( palette = "Set1", na.value = "grey50")+ labs(y=NULL)+ theme_classic()+ theme(legend.position = "none")# Then ggiraph #### ggiraph::girafe( ggobj = retraction_reasons_gg)```## 👨🔬 AuthorsWho are the most retracted authors in `r params$country`. Let's look at specific researchers (@fig-authors).```{r authors}#| label: fig-authorslibrary(ggiraph)library(tidyverse)library(countrycode) library(ggiraph)# First build the pipe for ggplot ####retraction_country_authors_gg <- retractionwatch_df |> dplyr::filter(country==params$country) |> tidyr::separate_rows(author ,sep=";") |> dplyr::count(author) |> dplyr::slice_max(order_by = n, n = 10) |> ggplot(aes(x=n,y=author |> reorder(n),fill=author))+ geom_col_interactive( aes(fill=n, data_id=author, tooltip=paste0(author, "<br>Retractions: ", n |> round(digits=0) |> format(big.mark=" "))))+ scale_fill_distiller_interactive( palette = "Set1", na.value = "grey50")+ labs(y=NULL)+ theme_classic()+ theme(legend.position = "none")# Then ggiraph #### ggiraph::girafe( ggobj = retraction_country_authors_gg)```## 🧑🔬 InstitutionsWhich institutions host the most retracted authors in `r params$country`? (@fig-institutions).```{r institutions}#| label: fig-institutionslibrary(ggiraph)library(tidyverse)library(countrycode) library(ggiraph)# First build the pipe for ggplot ####retraction_inst_gg <- retractionwatch_df |> dplyr::filter(country==params$country) |> tidyr::separate_rows(institution ,sep=";") |> dplyr::count(institution) |> dplyr::slice_max(order_by = n, n = 10) |> dplyr::filter(!is.na(institution), institution!="") |> ggplot(aes(x=n,y=institution |> reorder(n),fill=institution))+ geom_col_interactive( aes(fill=n, data_id=institution, tooltip=paste0(institution, "<br>Retractions: ", n |> round(digits=0) |> format(big.mark=" "))))+ scale_y_discrete(label = scales::label_wrap(50))+ scale_fill_distiller_interactive( palette = "Set1", na.value = "grey50")+ labs(y=NULL)+ theme_classic()+ theme(legend.position = "none")# Then ggiraph #### ggiraph::girafe( ggobj = retraction_inst_gg)```:::# References::: {#refs}:::