The EMA tool successfully identified that higher patient census was associated with lower perceived teaching quality, but the results were also influenced by the type of teaching service. We found that, regardless of census or teaching service, attendings can improve their teaching by diversifying the number of methods used in daily rounds.