Missing Data in Hierarchical Classification – a study with Personality developpment data

In this work we analyse the effect of missing data in hierarchical classification of variables according to the following factors: amount of missing data, imputation techniques, similarity coefficient, and aggregation criterion. We have used two methods of imputation, a regression method using an ordinary-least squares method and an EM algorithm. For the similarity matrices we have used the (unweighted) basic affinity coefficient ca ∑ , where x. j ∑ xij and x. j' ∑ xij', as defined for instance in i= 1 x. j' i= 1 i= 1 Bacelar-Nicolau(2000).and the Bravais-Pearson correlation coefficient In this work we are interested in the classification of variables. We use the following = n xij x. j xij' hierarchical aggregation criteria as defined in Anderberg(1973): Average linkage C