William F. Morris, Ruth A. Hufbauer, Anurag A. Agrawal, James D. Bever, Victoria A. Borowicz, Gregory S. Gilbert, John L. Maron, Charles E. Mitchell, Ingrid M. Parker, Alison G. Power, Mark E. Torchin, and Diego P. Vázquez. 2007. Direct and interactive effects of enemies and mutualists on plant performance: a meta-analysis. Ecology 88:1021–1029.

Supplement 2

Matlab code to perform factorial meta-analyses using Hedges' d and the log response ratio.
Ecological Archives E088-063-S2.

Copyright

Authors
File list (downloads)
Description

Author(s)

William F. Morris
Department of Biology
Duke University
Box 90338
Durham, NC 27708-0338 USA
E-mail: wfmorris@duke.edu

File list

meta_fact.zip -- zip file containing the following eight MATLAB function files:

fact_hedges_d.m -- A Matlab function that returns the individual, overall, and interaction effect sizes for 2 "agents" in a 2 × 2 factorial experiment, where effect size is measured using Hedges' d; the sampling variances of each effect size are also returned.

fact_logRR.m -- A Matlab function that returns the individual, overall, and interaction effect sizes for 2 "agents" in a 2 × 2 factorial experiment, where effect size is measured using the log response ratio; the sampling variances and degrees of freedom of each effect size are also returned.

J.m -- A Matlab function that computes the small-sample size correction factor J.

Q.m -- A Matlab function that computes a weighted sum of squares.

mean_effect.m -- A Matlab function that returns a weighted mean effect size and its 95% confidence limits, where the weights include the among-study variance if it is significant at P < 0.05. Best used when effect sizes are measured using Hedges' d; for the log response ratio, use mean_effect_L.

mean_effect_L.m -- A Matlab function that returns the weighted mean log response ratio effect size, its SE, and its 95% confidence limits, where the weights include the among-study variance, the significance of which (from a chi-square test on the sum of squares) is returned as well.

test_Qb_mixed_2.m -- A Matlab function that tests for a significant between-class sum of squares in a mixed-model meta-analysis comparing two classes.

test_Qb_mixed_n.m -- A Matlab function that tests for a significant between-class sum of squares in a mixed-model meta-analysis comparing n classes.

Description

This supplement includes Matlab code to compute individual, overall, and interactive effects using Hedges’ d and the log response ratio, to calculate weighted mean effect sizes, and to perform mixed-model homogeneity tests.

Functions mean_effect, mean_effect_L, test_Qb_mixed_2, and test_Qb_mixed_n all use the function chi2cdf from the Matlab Statistics Toolbox. Additional documentation appears as comments at the beginning of each function file; once the files have been downloaded into a folder in the Matlab path, typing help function_name (e.g., help fact_logRR) at the Matlab command prompt will display the descriptive comments.