Coverage and consistency.

Share:

Description

Coverage and consistency.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
coverage(x, ...)

## Default S3 method:
coverage(x, y, alternative = c("less", "greater"),...)

## S3 method for class 'QCA'
coverage(x, data, type = c("raw", "unique"), which = 1, ...)

consistency(x,...)

## Default S3 method:
consistency(x, y, alternative = c("less", "greater"), ...)

## S3 method for class 'QCA'
consistency(x, data, which=1, ...)

Arguments

x

a numeric vector of fuzzy set score for default method of coverage and consistency. A QCA object for QCA method of coverage.

y

a numeric vector of fuzzy set score.

alternative

specify the relationship between x and y.

data

The original data which are used to generate the QCA object.

which

length-1 positive integer to select a solution from the QCA object.

type

The type of coverage.

...

not used currently.

Details

Given two fuzzy set score variables x and y, coverage returns the coverage of x by y, given their relationship; and consistency returns the consistency of x by y.

When alternative is "less", it returns coverage(x<=y). This is coverage for sufficent test. When "greater", it returns coverage(x>=y), it is coverage for necessary condition.

This value can be used to evaluate the importance or relevance of x as a necessary condition for Y. When coverage(x>=y) is small, the constraining effect of x on y is negligible. In other words, small value of coverage(x>=y) indicates en empirically trivial necessary condition (Ragin, 2008: 61).

Value

For default method of coverage and consistency, the value is a length-1 numeric vector.

Author(s)

Ronggui HUANG

References

Ragin, C. 2006. Set Relations in Social Research: Evaluating Their Consistency and Coverage Political Analysis,14(3):291-310.

Ragin, C. 2008, Redesigning social inquiry: fuzzy set and beyond. Chicago: The University of Chicago Press.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
## evaluate the congerage of necessary conditions for "no democratic consolidation"
## see help(CoD) for more information.
## coverage for necessary test for not-CoD (Schneider, 2009:77)

coverage(1-CoD$econdev,1-CoD$cod,'greater')
# not socioeconomically developed
coverage(1-CoD$demex,1-CoD$cod,'greater')
# no previous democratic experience
coverage(pmin(1-CoD$demex,1-CoD$econdev),1-CoD$cod,'greater')
# not socioeconomic developed society without prior democratic experience

## coverage and consistency for QCA object
rownames(CoD) <- CoD$label
tb6.4 <- reduce(CoD,'cod',c('econdev','eduhi','ethlihom','close','demex','nocom'),
                exp="positive",remainder='include',pre="fs")
coverage(tb6.4, CoD)
coverage(tb6.4, CoD,"unique")
consistency(tb6.4, CoD)
cbind(consistency(tb6.4, CoD),coverage(tb6.4,CoD),coverage(tb6.4,CoD,"unique")) 
## put together

## coverage method for QCA object
tb2<-reduce(policyChange~soc+pol+ind+are+foc,Yamasaki2009,cases="country")
coverage(tb2,Yamasaki2009)