Blossom (version 1.4)

sp: Multiresponse sequence procedure

Description

Multiresponse sequence procedure (MRSP) is a special case of MRPP where first-order sequential pattern of data is tested against the null hypothesis of no sequential pattern. Univariate analyses are analogous to the Durbin-Watson test for first-order serial pattern and bivariate analyses are analogous to Schoener's $t^2/r^2$ statistic (Solow 1989). Permutation versions of these two tests can be done. Options allow you to select the sequencing variable and to turn off multivariate commensuration.

Usage

sp(data, expon = 1, commens = TRUE, number.perms, exact = FALSE,
save.test,sequence,variables)

Arguments

data

an object of class matrix (or an object coercible by as.matrix) with columns representing response variables.

expon

allows selection of alternative exponents in distance calculations.

commens

a logical value indicating whether to perform multivariate commensuration.

number.perms

number of permutations used if a Monte Carlo resampling procedure is to be used.

exact

logical indicating whether to perform an exact test or use a Monte Carlo resampling procedure.

save.test

logical indicating whether to return Monte Carlo resampled test statistic values.

sequence

a numeric vector specifying how the data should be ordered for the analysis.

variables

a character string of the names to be used in the analysis. These should match names found in the data.frame.

Value

sp returns an object of class MRSPObj.The functions summary as well as print can be used to obtain a summary of the test.Generic accessor functions pvalue and ResampVals can be used to obtain the p-value and Monte Carlo resampled test statistic values respectively.

Details

In this analysis of ungrouped data, the agreement measure (1 - average Euclidean distance between ordered observations/average Euclidean distance among all possible pairs of observations) is a statistic describing first-order serial dependency. Significance of the null hypothesis of no first-order serial dependency is provided by the Pearson Type III approximation on the first 3 exact moments of the permutation distribution by default, optionally by exact enumeration for small samples by specifying the exact option, or by a Monte Carlo resampling procedure by specifying the number.perms option.

References

Solow, A.R. 1989. A randomization test for independence of animal locations. Ecology 70, 1546--1549.

See Also

MRSPObj

Examples

sp(blue162[,4:5], expon = 1,number.perms = 1000,save.test = TRUE)