linear.hierarchy.test: Testing Linearity in dominanace hierarchies

In DLEIVA/DyaDA: Dyadic Data Analysis

Description

Function to estimate statistical significance for Landau's h index.

Usage

   linear.hierarchy.test(X,rep=9999)

Arguments

X	Original sociomatrix.
rep	Number of randomizations. A number between 1 and 1000000 should be specified.

Details

linear.hierarchy.test estimates statistical significance for Landau's h index by means of a two-step randomization procedure proposed by de Vries (1995). If number of individuals is greater or equal to 10 it also provides with statistical significance for Kendall's linearity index by means of a Chi-squared test.

Value

nind	Number of individuals.
dyads	Number of dyads.
total	Total number of interactions.
desctable	Description of the dyadic relationships in the input sociomatrix.
randomizations	Specified number of randomizations.
landauh	Landau's h index.
improved.landauh	Improved Landau's index h'.
expected.h	Mathematical expectancy of Landau's h under the assumption of no linear hierarchy.
variance.h	Variance of Landau's h under the assumption of no linear hierarchy.
right.pvalue	Right-tailed p-value.
left.pvalue	Left-tailed p-value.
circ	Number of circular triads.
expcirc	Expected number of circular triads.
maxcirc	Maximum number of circular triads.
rho	Kendall's rho index.
rho.test	Statistical test for Kendall's rho.

Author(s)

David Leiva <[e-mail]>, Antonio Solanas <[e-mail]>, Han de Vries <[e-mail]>, & David A. Kenny <[e-mail]>.

References

de Vries, H. (1995). An improved test of linearity in dominance hierarchies containing unknown or tied relationships. Animal Behaviour, 50, 1375-1389.

Kendall, M. G., & Babington Smith, B. (1940). On the method of paired comparisons. Biometrika, 31, 324-345.

See Also

getexpech, getimplandau, getlandau, getvarh, getcirc, getexpeccirc, getkendall, getmaxcirc.

Examples

  set.seed(123)
  X <- matrix(floor(runif(20*20,min=0,max=20)),nrow=20,byrow=TRUE)
  diag(X)<-0
  linear.hierarchy.test(X)

DLEIVA/DyaDA documentation built on May 6, 2019, 1:17 p.m.