Cluster 197: trial confound best abnormality results interfere participate laboratory duration current 
Cluster 2336: confound results study situation subject condition significantly participation risk opinion