Good service and prompt attention do little to guarantee a big tip from restaurant-goers.

A new study conducted by researchers at Cornell University and the University of Houston suggests that the amount left as a tip by diners is influenced more by bill size and the fear of disappointing the server than by good service.

The study, "Tipping: An Incentive/Reward for Service?" was authored by Michael Lynn, associate professor of consumer behavior and marketing at Cornell's School of Hotel Administration and Jeffrey Graves of the Conrad N. Hilton College of Hotel and Restaurant Management.

"There is a rather weak relationship between the size of the tip and the level or quality of service one receives from their waiter or waitress," said Lynn, who has examined the social phenomenon of tipping for nearly a decade.

Lynn said the amount of the bill is still the dominant factor influencing how much tip one leaves and how much money a server earns.

The study's findings, Lynn believes, suggest that tipping may not be the incentive for excellent service that restaurant managers believe it is.

"Managers assume that servers are motivated to perform well to get good tips, so management doesn't make any effort to modify service performance," he said. "They don't give additional incentives for providing good service because they assume the tip is a sufficient incentive. Well, this study shows that's not the case."

In the study, a waitress at a Red Lobster in Columbia, Mo., was asked to log customers' responses to her questions relating to meal quality and service, such as "Is everything all right?" and to record how much money they left as a tip. A total of 178 tables were observed.

Customers who voiced praise for their dining experience tipped an average of 12 percent of bill size, compared with patrons who did not express praise, who tipped an average of 10 percent.

"Service quality had little influence on determining how much tip diners left for their server," Lynn said. "Those that told the waitress the service was fine offered little reward compared with those diners whose responses were noncommittal."

In a survey conducted at restaurants in Houston -- Bennigan's and The Olive Garden -- patrons were asked to rate the level of service they received on a scale of 1 to 5, with 5 being best.

Patrons who rated their dining experience 4.6 or higher tipped 16.6 percent of their bill size, while those who rated their dining experience 4.4 or lower tipped an average of 14.2 percent of their bill size.

"Here again, we found that while people who felt their service was, in this case, exceptional, rewarded servers with only minor increases over the normal practice of tipping 15 percent," Lynn said. "There was relatively little difference between a tip someone would leave for adequate service or for excellent service."

The findings show that bill size is the single largest predictor of tip size, suggesting that tipping is primarily a "norm-driven" behavior, Lynn said.

The study results, Lynn said, have serious implications for restaurant management. "If providing good service doesn't guarantee significantly larger tips, then there may be little incentive for wait staff to provide good service.

"Waiters and waitresses may start rushing their tables to get them to turn over faster and therefore be able to serve more dinners, which would lead to more income from tips, but also create a less desirable dining experience," Lynn said.

Restaurant managers may be better able to monitor the quality of service delivered by the wait staff by asking customers to file comment cards or by hiring "mystery diners" -- diners who are paid by management to assess the performance of waiters and waitresses, Lynn suggested.

The study will be published in a forthcoming issue of Hospitality Research Journal.