Headline evaluation experiment results

This technical report describes an experiment intending to show that different summarizationtechniques have an effect on human performance of an extrinsic task. The task is document
selection in the context of information retrieval. We conclude that the task was too difficult and
ambiguous for subjects to perform at the level required in order to make statistically significant
claims about the relationship between summarization techniques and human performance. An
alternate interpretation of the experimental results is described and plans for future experiments are
discussed.