Skip to content

agbarnett/league.tables

Repository files navigation

University league table of good research practice

Creating an international university league table based on a good research practice.

The key R files are:

  • data.management.R takes the raw data from Scopus and attempts to create consistent names for institutions.
  • WhoCitesEquator.Rmd is the Rmarkdown file for the complete analyses.
  • ui.R and server.R make the Shiny page with the interactive results.

The two key data sets (both available in csv and RData format) are:

1. Papers.for.Analysis.500 which are a random sample of 500 of the 47,876 papers used to create the league tables, with the variables:

  • doi
  • affiliation
  • affid, Scopus affiliation ID number
  • country
  • year, 2016 or 2017
  • Region
  • weight, fraction count of authors in the range (0,1]

2. Processed.papers.for.Shiny which is the summary data used to create the interactive league tables in Shiny, with the variables:

  • year, 2016 or 2017
  • country
  • Region
  • affiliation
  • wsum, sum of fractional papers per university which determines the rank
  • n, number of papers
  • rank, rank per year
  • cluster.member, estimated cluster based on wsum from 1 (lowest) to 5 (highest)
  • p10, bootstrap probability of being in the top 10
  • rlower, bootstrap lower 95% confidence interval
  • rupper, bootstrap upper 95% confidence interval

3. pubmed.frame which details the EQUATOR papers used to count citations:

  • type, type of EQUATOR paper: CONSORT, PRIMSA or STROBE
  • title, paper title
  • journal, journal title
  • date, date of publication
  • pmid, pubmed ID number
  • year, year published