distribute_tasks: Divide a set of articles among two or more reviewers

Description

A common task in systematic review is to divide a dataset of articles located by a search (typically involving >1 databases) and distributing them amongst a team of reviewers for screening. This function takes a dataset divides it among the specified number of reviewers, returning the resulting data.frames either in a list to the workspace, or (by default) as a set of .csv files in the specified directory. The resulting files can be passed to any of the screening functions provided by revtools, i.e. screen_titles, screen_abstracts, or screen_topics.

Usage

distribute_tasks(data, reviewers, write_csv = TRUE,
    file_name = "reviewer.csv", return_data = FALSE, ...)

Arguments

data

a vector of strings

reviewers

Either a data.frame as returned by allocate_effort, an integer giving the number of reviewers, or a vector of strings giving reviewer names.

write_csv

Logical - should the function write a set of csv files (1 per reviewer)? Defaults to TRUE

file_name

a file path & name showing where .csv files should be saved. Ignored if write_csv is FALSE. Defaults to 'reviewer_[name].csv'.

return_data

Logical - should a list be (invisibly) returned, in which each entry is the data sent to a single reviewer? Defaults to FALSE.

...

Further arguments passed to allocate_effort

Value

Invisibly returns a list of data.frames, each with same columns as data but containing only a subset of rows.

Details

The dataset is allocated each author in the proportion of articles specified by allocate_effort, with the identity of articles passed to reviewer being chosen by rnorm. As a result, this function is very sensitive to the inputs provided to allocate_effort, so it is often worth running that function first and checking the results to be certain that effort is being distributed in a way that you are happy with.

Examples

Run this code

# NOT RUN {
# import some data
file_location <- system.file(
  "extdata",
  "avian_ecology_bibliography.ris",
  package = "revtools")
x <- read_bibliography(file_location)
result <- distribute_tasks(x, 4, write_csv = FALSE) # split evenly among 4 reviewers
# }