A Multivariate Analysis Computer Program

Abstract

Summary This paper describes an electronic computer program for the analysis of multivariate data. The program is interactive in the sense that its control structure is designed to facilitate a step‐by‐step process of analysis extending over several computer runs. Especially comprehensive input routines allow the user flexibility in data presentation. A group structure can be imposed and within‐ and between‐group analyses are possible. The program includes all the usual multivariate techniques, but special attention has been paid to providing a wide repertoire of operations useful in multiple‐regression analysis.

Coastal darkening substantially limits the contribution of kelp to coastal carbon cycles

“…To investigate the relationships between adult E. radiata density, total biomass m −2 and environmental variables (benthic irradiance, sea temperature, seawater nitrate + nitrite and ammonium concentrations, fetch, MWH, and SWH), distance-based linear modelling (DistLM) with BEST selection criteria was used. DistLM provides results for marginal and conditional tests (Anderson, 2004). The univariate analyses were carried out on Euclidean distance matrices.…”

DNA stable-isotope probing highlights the effects of temperature on functionally active methanotrophs in natural wetlands

“…The combined effects of soil parameters on the methanotroph community structure were tested using distance-based linear model multivariate analysis (DISTLM) (McArdle and Anderson, 2001). This method could minimize the autocorrelation between soil parameters, and then each soil parameter contribution was selected by the marginal tests, the cumulative effect of soil parameters was evaluated by the sequential tests (Anderson, 2003). Phylogenetic trees of pmoA and 16S rRNA genes were constructed and analyzed using FastTree (v1.0.0) and MEGA (v7.0.26) and visualized using the Interactive Tree of Life (iTOL, v5.5.1) (Letunic and Bork, 2011).…”

Morphological variation of the kelp Ecklonia radiata in northeastern New Zealand and across its Australasian range

“…To test multivariate relationships between E. radiata morphology (stipe length, lamina length, total length, percent stipe length) and the environmental variables (depth, wave exposure, turbidity and SST), DistLM with BEST selection criteria were used (Anderson, 2004) using the reduced Akaike's information criterion (AICc).…”