Is there a way to create a summary statistic of the overall percent agreement between the two datasets?