callback {callback}

Data formatting

Description

Creates the data set used in the callback package.

Usage

callback(data, cluster, candid, callback, comp = "ref")

Arguments

data

A data frame.

cluster

A variable name, identifying the test (e.g., a job offer number).

candid

A list of factor names defining the candidates (e.g., gender, origin).

callback

A Boolean variable, equal to TRUE for non negative callbacks.

comp

An option, equal to "all" or "ref" (the default). "ref" give the comparisons with the reference candidate, and "all" the pairwise comparisons.

Value

A 'callback' object containing the formatted data set (fds), the averaged formatted data set (afds, for multiple tests), the list of the paired formatted data sets (pfds), the offer-level formatted data set (ofds), the names of the candidate variables (candid) and the name of the callback variable (callback).

fds contains the following variables:

• cluster: the cluster variable.

• candid: the concatenation of the candidate variables.

• callback: the callback variable.

afds contains the following variables:

• cluster: the cluster variable.

• candid: the concatenation of the candidate variables.

• callback: the aggregated callback dummy.

pfds data frames containing the following variables:

• callback1: TRUE if candidate 1 had a callback.

• callback2: TRUE if candidate 2 had a callback.

• c00: TRUE if neither candidate was called back.

• c10: TRUE if candidate 1 was the only one called back.

• c01: TRUE if candidate 2 was the only one called back.

• c11: TRUE if both candidates were called back.

• callback: TRUE if either candidate was called back.

• calldif: callback difference.

ofds contains variables corresponding to the candidates' codes and the following variables:

• cluster: the cluster variable.

• ncall: the number of callback for the current offer.

Author(s)

Emmanuel Duguet

Examples

data(inter1)
m <- callback(data=subset(origin1,reput=="U"),cluster="offer",candid="origin",
     callback="callback",comp = "ref")
str(m)