Hypothesis testing describes a range of methods which attempt to help researchers make a decision concerning the truth or falsity of a given hypothesis using data from an appropriate experimental design. For example, the hypotheses that two groups of samples have equal mean abundances of a given set of OTUs or that an environmental variable is unable to account for the variation in OTUs across a gradient may be tested and the likelihood any association is simply due to random chance examined.
Hypothesis testing is a form of confirmatory analysis. Rather than exploring data for patterns, it evaluates if the data confirms or refutes a given notion. Thus, the method you choose will depend on the question posed and the design of your study. Numerous methods exist, however, this guide includes some of the more widely-used, multivariate testing procedures in ecology.
Would you like to...
test for association between response and explanatory variables? (e.g. OTU abundances) and explanatory variables (e.g. environmental parameters)