indexConflict3 {OpenRepGrid}

Conflict or inconsistency measure for grids (Bell, 2004) based on distances.

Description

Measure of conflict or inconsistency as proposed by Bell (2004). The identification of conflict is based on distances rather than correlations as in other measures of conflict indexConflict1() and indexConflict2(). It assesses if the distances between all components of a triad, made up of one element and two constructs, satisfies the "triangle inequality" (cf. Bell, 2004). If not, a triad is regarded as conflictive. An advantage of the measure is that it can be interpreted not only as a global measure for a grid but also on an element, construct, and element by construct level making it valuable for detailed feedback. Also, differences in conflict can be submitted to statistical testing procedures.

Usage

indexConflict3(
  x,
  p = 2,
  e.out = NA,
  e.threshold = NA,
  c.out = NA,
  c.threshold = NA,
  trim = 20
)

Arguments

x

repgrid object.

p

The power of the Minkowski distance. p=2 (default) will result in euclidean distances, p=1 in city block distances.

e.out

Numeric. A vector giving the indexes of the elements for which detailed stats (number of conflicts per element, discrepancies for triangles etc.) are prompted (default NA, i.e. no detailed stats for any element).

e.threshold

Numeric. Detailed stats are prompted for those elements with a an attributable percentage to the overall conflicts higher than the supplied threshold (default NA).

c.out

Numeric. A vector giving the indexes of the constructs for which detailed stats (discrepancies for triangles etc.) are prompted (default NA, i. e. no detailed stats).

c.threshold

Numeric. Detailed stats are prompted for those constructs with a an attributable percentage to the overall conflicts higher than the supplied threshold (default NA).

trim

The number of characters a construct (element) is trimmed to (default is 10). If NA no trimming is done. Trimming simply saves space when displaying the output.

Details

Status: working; output for euclidean and manhattan distance checked against Gridstat output.
TODO: standardization and z-test for discrepancies; Index of Conflict Variation.

Value

A list (invisibly) containing:

• potential: number of potential conflicts

• actual: count of actual conflicts

• overall: percentage of conflictive relations

• e.count: number of involvements of each element in conflictive relations

• e.perc: percentage of involvement of each element in total of conflictive relations

• c.count: number of involvements of each construct in conflictive relation

• c.perc: percentage of involvement of each construct in total of conflictive relations

• e.stats: detailed statistics for prompted elements

• c.stats: detailed statistics for prompted constructs

• e.threshold: threshold percentage. Used by print method

• c.threshold: threshold percentage. Used by print method

• enames: trimmed element names. Used by print method

• cnames: trimmed construct names. Used by print method

output

For further control over the output see print.indexConflict3().

References

Bell, R. C. (2004). A new approach to measuring inconsistency or conflict in grids. Personal Construct Theory & Practice, (1), 53-59.

See Also

See indexConflict1() and indexConflict2() for conflict measures based on triads of correlations.

Examples

# calculate conflicts
indexConflict3(bell2010)

# show additional stats for elements 1 to 3
indexConflict3(bell2010, e.out = 1:3)

# show additional stats for constructs 1 and 5
indexConflict3(bell2010, c.out = c(1, 5))

# finetune output
## change number of digits
x <- indexConflict3(bell2010)
print(x, digits = 4)

## omit discrepancy matrices for constructs
x <- indexConflict3(bell2010, c.out = 5:6)
print(x, discrepancies = FALSE)