Are America's hotels measuring their performance accurately? No, according to a recent Cornell University study, which shows that industry performance averages, commonly used by hotels throughout the United States, are not reliable as the only gauge of how hotels are doing.

The study was conducted by Cathy A. Enz, Linda Canina and Kate Walsh, faculty members at Cornell's School of Hotel Administration, under the aegis of the Hotel School's Center for Hospitality Research in alliance with Smith Travel Research, which supplied data from its database of name-brand hotels in the United States.

The researchers scrutinized the average daily rate (ADR), revenue per available room (RevPAR) and occupancy averages at virtually all name-brand hotels in the United States from 1988 through 2000. Their findings: Overall industry averages are skewed by the extreme values of some hotels. The averages for ADR and RevPAR are distorted upward by the extremely high rates charged by certain hotels. At the same time, occupancy is skewed downward by the exceptionally low occupancy rates recorded by some hotels. The findings may be especially important to luxury and upscale hotels, where the greatest distortions occur.

"On a practical level, what those distortions in the averages mean is that the typical hotel actually has a lower average daily rate and revenue per available room, and a higher occupancy rate, than found in the industry average," says Enz.

The researchers made their determinations by examining two other measures -- the median, or middle value, and the mode, or most common value. They found that the average occupancy for all name-brand hotels in the U.S. was lower than that for the typical hotel (the mode) and for the middle value of all hotels (the median). Their conclusion: hotel industry averages commonly understate occupancy.

But while the researchers found that this analysis held true for individual lodging segments, they also note that the greatest distortions of the averages occurred in the luxury and upscale segments. On the other hand, the statistics for economy and budget hotels were more consistent. That is, for high-price hotels the average, the median and the mode varied greatly. For low-price hotels, those three figures were closer. The researchers warn that, as a result, operators and analysts of luxury and upscale hotels must be more careful in their use of overall industry statistics than those who are dealing with economy and budget properties.

Because many of the extreme values are found in the top 25 markets, the researchers also analyzed the statistics from those markets and once again found that the overall RevPAR statistics from each market were distorted by the high-price hotels in each market. Of equal interest is their finding that each of the top 25 markets had its own particular RevPAR distribution. Even the markets that commonly were considered similar in RevPAR distributions were in reality distinct from each other, their study revealed.

The Center for Hospitality Research seeks to shape the global knowledge base for hospitality managers through in-depth research on and for the industry. Current partners and sponsors include Four Seasons Hotels; Smith Travel Research; Willowbend Golf Management; Windsor Hotels; Richmond Events; and Cornell Hotel Society Foundation.

For a copy of the study or more information, contact hosp_research@cornell.edu or (607) 255-9780 or visit this web site: http://www.hotelschool.cornell.edu/chr/ .