Home

Compositional Analysis

How to Order Compos Analysis

Ecological Consultancy

About Us

Related Links

Contact Us

Contact:

Compos Analysis v6.3 software is designed to allow easy implementation of the logratio analysis method (Aitchison 1986) of statistical analysis of compositional data, also referred to as compositional analysis. Compositional Analysis is a technique that uses MANOVA to analyse two sets of data in which variables are represented as proportions. It is used to determine the statistical significance of differences and the rank order of differences between the variables. Applications include analysis of habitat preferences in wildlife studies where proportions of habitat use and availability are known.

*** The plus version (6.3+) includes the option to weight cases during analysis

This is valuable when the quantity of data collected varies from animal to animal

 

*** The pro version (6.3 Pro) handles very large datasets and includes the option to weight cases during analysis This will handle up to 20,000 cases with up to 30 variables. Note that the 64 bit version of Excel 2010 has a much increased RAM use limit over earlier versions of Excel and will allow data sets with 20,000 cases to be analysed, whereas versions of Excel with a 2 GB RAM limit may stop and report “out of memory” if more than about 5,000 or 10,000 cases are entered in the analysis.

 

· The software is an Add-In for use with Microsoft Excel. It works on a PC under Windows using Excel 2010, 2007, 2002  & 97 and on an Apple Mac under OS X using Excel 10 or any compatible versions of Excel.

· Compos Analysis (Smith 2004) implements fully the method for Compositional Analysis as described in the paper by Aebischer et al. (1993). [Testing for effects, which was not covered in that paper, is not included in the program, but can be achieved by entering log- ratio differences that are output from the program as intermediate data, into a standard MANOVA.]

· The significance of Wilk’s lambda and of t-values is determined by randomization tests as recommended by Aebischer et al. (1993) to overcome problems arising when the distribution of log-ratio differences is not multivariate normal.

· Missing data is dealt with by the preferred method recommended by Aebischer et al. (1993).

· The program fully automates the analysis with minimal input required from the user. Simply enter the two sets of data as percentages and start the program.

· The plus version includes the option to weight cases during analysis. This is valuable when the quantity of data collected varies from animal to animal.

 

For further details of how the program works upload the Compos Analysis 6.3 User Guide (Issue2). This is a PDF document that requires Adobe Acrobat Reader to view it.

For prices and ordering details for Compos Analysis 6.3 click here.

For previous versions of the Compos Analysis User's Guide click here.

References:

Aebischer, N. J., Robertson, P. A. & Kenward, R. E. 1993. Compositional analysis of habitat use from animal radio-tracking data. Ecology, 74: 1313-1325.

Aitchison, J. 1986. The statistical analysis of compositional data. Chapman and Hall, London, England, 416 pp.

Smith, P. G. 2004. Automated log-ratio analysis of compositional data: software suited to analysis of habitat preference from radio tracking data. Bat Research News, 45(1): 16.

 

Compositional Analysis Software