QCA: retention – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

retention

Compute the retention probability of a csQCA solution

Description

This function computes the retention probability for a csQCA solution, under various perturbation scenarios. It only works with bivalent crisp-set data, containing the binary values 0 or 1.

Usage

retention(data, outcome = "", conditions = "", incl.cut = 1, n.cut = 1,
         type = "corruption", dependent = TRUE, p.pert = 0.5, n.pert = 1)

Arguments

`data`	A dataset of bivalent crisp-set factors.
`outcome`	The name of the outcome.
`conditions`	A string containing the condition variables' names, separated by commas.
`incl.cut`	The minimum sufficiency inclusion score for an output function value of "1".
`n.cut`	The minimum number of cases for a causal combination with a set membership score above 0.5, for an output function value of "0" or "1".
`type`	Simulate corruptions of values in the conditions ("corruption"), or cases deleted entirely ("deletion").
`dependent`	Logical, if `TRUE` indicating DPA - Dependent Perturbations Assumption and if `FALSE` indicating IPA - Independent Perturbations Assumption.
`p.pert`	Probability of perturbation under independent (IPA) assumption.
`n.pert`	Number of perturbations under dependent (DPA) assumption.

Details

The argument data requires a suitable data set, in the form of a data frame. with the following structure: values of 0 and 1 for bivalent crisp-set variables.

The argument outcome specifies the outcome to be explained, in upper-case notation (e.g. X).

The argument conditions specifies the names of the condition variables. If omitted, all variables in data are used except outcome.

The argument type controls which type of perturbations should be simulated to calculate the retention probability. When type = "corruption", it simulates changes of values in the conditions (values of 0 become 1, and values of 1 become 0). When type = "deletion", it calculates the probability of retaining the same solution if a number of cases are deleted from the original data.

The argument dependent is a logical which choses between two categories of assumptions. If dependent = TRUE (the default) it indicates DPA - Dependent Perturbations Assumption, when perturbations depend on each other and are tied to a fixed number of cases, ex-ante (see Thiem, Spohel and Dusa, 2016). If dependent = FALSE, it indicates IPA - Independent Perturbations Assumption, when perturbations are assumed to occur independently of each other.

The argument n.cut is one of the factors that decide which configurations are coded as logical remainders or not, in conjunction with argument incl.cut. Those configurations that contain fewer than n.cut cases with membership scores above 0.5 are coded as logical remainders (OUT = "?"). If the number of such cases is at least n.cut, configurations with an inclusion score of at least incl.cut are coded positive (OUT = "1"), while configurations with an inclusion score below incl.cut are coded negative (OUT = "0").

The argument p.pert specifies the probability of perturbation under the IPA - independent perturbations assumption (when dependent = FALSE).

The argument n.pert specifies the number of perturbations under the DPA - dependent perturbations assumption (when dependent = TRUE). At least one perturbation is needed to possibly change a csQCA solution, otherwise the solution will remain the same (retention equal to 100%) if zero perturbations occur under this argument.

Author(s)

Adrian Dusa

References

Thiem, A.; Spoehel, R.; Dusa, A. (2015) “Replication Package for: Enhancing Sensitivity Diagnostics for Qualitative Comparative Analysis: A Combinatorial Approach”, Harvard Dataverse, V1. doi: 10.7910/DVN/QE27H9

Thiem, A.; Spoehel, R.; Dusa, A. (2016) “Enhancing Sensitivity Diagnostics for Qualitative Comparative Analysis: A Combinatorial Approach.” Political Analysis vol.24, no.1, pp.104-120.

Examples

# the replication data, see Thiem, Spohel and Dusa (2015)
dat <- data.frame(matrix(c(
    rep(1,25), rep(0,20), rep(c(0,0,1,0,0),3),
    0,0,0,1,0,0,1,0,0,0,0, rep(1,7),0,1),
    nrow = 16, byrow = TRUE, dimnames = list(
    c("AT","DK","FI","NO","SE","AU","CA","FR",
      "US","DE","NL","CH","JP","NZ","IE","BE"),
    c("P", "U", "C", "S", "W"))
))


# calculate the retention probability, for 2.5% probability of data corruption
# under the IPA - independent perturbation assuption
retention(dat, outcome = "W", incl.cut = 1, type = "corruption",
       dependent = FALSE, p.pert = 0.025)

# the probability that a csQCA solution will change
1 - retention(dat, outcome = "W", incl.cut = 1, type = "corruption",
       dependent = FALSE, p.pert = 0.025)

QCA

Qualitative Comparative Analysis

v3.11

GPL (>= 3)

Authors

Adrian Dusa [aut, cre, cph] (<https://orcid.org/0000-0002-3525-9253>), Ciprian Paduraru [ctb] (<https://orcid.org/0000-0002-4518-374X>), jQuery Foundation [cph] (jQuery library and jQuery UI library), jQuery contributors [ctb, cph] (jQuery library; authors listed in inst/gui/www/lib/jquery-AUTHORS.txt), lp_solve [cph] (http://lpsolve.sourceforge.net), Vasil Dinkov [ctb, cph] (jquery.smartmenus.js library), Dmitry Baranovskiy [ctb, cph] (raphael.js library), Emmanuel Quentin [ctb, cph] (raphael.inline_text_editing.js library), Jimmy Breck-McKye [ctb, cph] (raphael-paragraph.js library), Alrik Thiem [aut] (from version 1.0-0 up to version 1.1-3)

Initial release

2021-01-21