Multivariate data refers to a situation where you record
several response variables as the outcomes of your research.
This may occur in many different ways, e.g.
Questionnaire results typically give responses to a set of different questions for each person who completes the questionnaire.
Several characteristics (e.g. height, number of leaves, etc) may be measured for each of different plants in a growth trial.
Different methods of collecting fingerprints may be assessed by recording several qualities (e.g. sharpness, visibility) in each print.
Several variables (e.g. cadmium, manganese, BOD, etc) may be recorded to assess pollution levels.
Multivariate data can usually be set out in a table, with the columns giving the different variable values that you record and the rows giving the subject being measured, e.g. the person in the questionnaire, the plant being measured, the particular fingerprint being analysed, pollution sample, etc.
ACTION: You now need to
decide whether you have multivariate data, and
whether the particular analysis that you require will be a univariate analysis or a multivariate analysis.
When performing your data analysis:
Multivariate analysis of your data would involve analysing two or more of your response variables together as joint response variables.
Univariate analysis of your data would involve performing individual analyses on each of your response variables separately.
Note that testing for correlation between two or more variables (e.g. does weight and height increase together) does not imply multivariate analysis.