In this work we propose two solutions to the problem we consider when we have multivariate experimental data of individuals with structure of groups that have been repeated at different occasions or experimental situations. We obtain the first solution applying dual version of canonical STATIS proposed by Vallejo-Arboleda et al. (2007), and the second with doubly multivariate repeated measures model. We use the data of SWALE project (Stephen et al. 2004), with 7 physical-chemical variables, measured in 4 treatments, each one with 9 repetitions and during 4 weeks (selected between 10 original weeks in the project) to compare these solutions.