Pierre Legendre
June 2000
Département de Sciences Biologiques
Université de Montréal
This program computes a t-test for a Pearson correlation coefficient corrected for spatial autocorrelation, following Dutilleul (1993). In this version of the program, the main input data file (text) must contain 4 columns: Coordinate X, Coordinate Y, Variable 1, Variable 2. The program assumes the coordinates to be in a Cartesian plane. One may provide files with one's own limits of distance classes, pre-computed Moran's I autocorrelation coefficients, as well as an upper-triangular matrix of distance classes, if one so wishes. These elements are not necessary, however. In the absence of such matrices, the program will do all calculations from the input data matrix of the previous paragraph, using equidistant classes. If the user provides the value of the Spearman correlation coefficient, the associated probability will be computed and printed after that of the Pearson correlation coefficient. The probabilities computed by the program are for two-tailed tests. It is obtained by linear interpolation between the values computed for the integer values of d.f. above and below the corrected real-valued degrees of freedom, which are printed as "df" on output.

Program availability

References:
Dutilleul, P. 1993. Modifying the t test for assessing the correlation between two spatial processes. Biometrics 49: 305-314.