Challenges of visualization and clustering are explored with respect to sequence data from a simulation-based assessment task. Visualization issues include representing progress towards a goal and accounting for variable-length sequences. Clustering issues focus on external criteria with respect to official scoring rubrics of the same sequence data. The analysis has a confirmatory flavor; the goal is to understand to what extent clustering solutions align with score categories. It is found that choices related to data preprocessing, distance metric and external cluster validity measures all impact agreement between cluster assignments and scores. This work raises key issues about clustering of educational data, especially in the presence of multidimensionality. Different clustering protocols may lead to different solutions, no one of which is uniquely best.