multicon (version 1.6)

Profile.r.rep: Profile Correlation Replicability

Description

Computes the replicability of both overall and distinctive Profile correlations.

Usage

Profile.r.rep(x.set, y.set, nomiss = 1, CI = 0.95, CItype = "xci")

Arguments

x.set

A data.frame or matrix of the same dimensions as y.set with rows corresponding to the rows in y.set

y.set

A data.frame or matrix of the same dimensions at x.set with rows correponsding to the rows in x.set

nomiss

A numeric between .00 and 1.00 indicating the proportion of pairs of scores that must be valid for a result to be computed.

CI

A numeric between .00 and 1.00 indicating the desired confidence level.

CItype

A character element of either "xci" or "aci" specifying the the type of confidence interval to compute based on Koning & Franses (2003).

Value

A matrix containing the replicability point estimate and its confidence intervals.

Overall

Replicability of Overall correlations between x.set and y.set

Distinctive

Replicability of Distinctive correlations between x.set and y.set

Details

Sherman and Wood (in press) describe a method for computing the replicability of a vector of correlation coefficients (see vector.alpha). They also discuss how this may be applied to profile correlations. This function applies the strategy outlined by Sherman and Wood (in press) and used by the vector.alpha function to profile correlations. The results include the replicability point estimate for both the overall profile correlations between x.set and y.set as well as the distinctive profile correlations. Confidence intervals are computed based on Koning and Frances' (2003) methods, choosing either asymptotic ("aci") or exact ("xci").

References

Sherman, R. A. & Wood, D. (in press). Estimating the expected replicability of a pattern of correlations and other measures of association. Multivariate Behavioral Research.

See Also

Profile.r vector.alpha

Examples

data(acq1)
data(caq)
	#Lets look at Profile correlations between self-report California Adult Q-Sort 
	#ratings of personality and Aquaintance #ratings
names(acq1)
names(caq)
	# The basic Profile agreements
Profile.r(caq, acq1)  
	# Both overall and distinctive agreements
Profile.r(caq, acq1, distinct = TRUE)$Agreement 
	# How replicable (reliable) are those agreement patterns?
Profile.r.rep(caq, acq1)