LogSeq: LogSeq

In DySeq: Functions for Dyadic Sequence Analyses

Description

Implementation of Bakeman & Gottman (1997) for sequence analysis. Kenny, Kashy & Cook (2006) provide further examples.

Usage

LogSeq(x, delta = 0.5, subgroups = NA, single.case = FALSE)

Arguments

x	a state.trans object or a list of 4*2 state-transition tables
delta	constant added to every cell, required if zero frequencies occur!
subgroups	an optional vector containing groupmembership if estimates should be compared between groups
single.case	should p-values be computed for single case analysis

Details

• Runs logit models over a multiple number of state-transition tables, see: StateTrans

• Aggregates coefficients of all logit models and tests them against zero.

• If subgroups are defined, coefficients are tested to be different between groups.

• Print-function displays mean logit-coefficients and p-values.

References

• Bakeman, R., & Gottman, J. M. (1997) <DOI: 10.1017/cbo9780511527685 >

• Kenny, D. A., Kashy, D. A., & Cook, W. L. (2006) <DOI: 10.1177/1098214007300894>

Examples

## Not run: 
data(CouplesCope)
my.states<-StateExpand(CouplesCope, 2:49, 50:97)
my.trans<-StateTrans(my.states, FALSE)
my.logseq<-LogSeq(my.trans, single.case=TRUE)
my.logseq

plot(my.logseq) # interaction can be plotted

single.LogSeq(my.logseq, 41) # for single case analysis

## End(Not run)

DySeq documentation built on May 1, 2019, 10:14 p.m.