Using a commercially available visual graphical analysis tool, the National Science Digital Library (NSDL) has developed techniques to expedite evaluation of large batches of metadata. These techniques allow efficient and thorough review of large quantities of XML metadata, thus enabling the focus of limited resources on evaluation and manipulation tasks that are most important in our context. In the NSDL, metadata is evaluated for aggregation, but these techniques are applicable to any situation where batches of metadata need to be evaluated. This paper discusses the motivations for this approach and the techniques themselves.