Several university students said Thursday that they fabricated survey results that factored into a judge's decision to move Scott Peterson's capital murder trial out of Modesto.

Judge Al Girolami learned of the phony poll allegation just before 5 p.m. and, through a court intermediary, declined to comment.

"We falsified the info," said a 20-year-old criminal justice student at California State University, Stanislaus. "The stuff we submitted wasn't true."

He referred to the 10-county Peterson bias survey compiled by 65 students and overseen by professor Stephen Schoenthaler.

Informed Thursday evening of the students' claims, Schoenthaler said, "I'm stunned, and I find it hard to believe. It seems impossible that I could have missed something like that."

University Vice Provost Diana Demetrulias said her office will begin an investigation today.

Chief Deputy District Attorney John Goold suggested that the revelation could cause Girolami to reconsider his decision.

"Oh, my God," Goold said when informed of the students' claims. "It certainly sounds like this would affect the underpinnings of the judge's decision."

The student and five others -- all seniors -- said Thursday that they made up every answer on all the surveys they submitted because they found it difficult to gather legitimate data.

They did it, they said, because they were short on time and money. They were required to participate in the survey for 20% of their grade and were given no money for dozens of lengthy long-distance phone calls, they said.

Another senior said she struggled to complete half of her required surveys, then gave up and faked the rest. Another said she refused to cheat but didn't have the resources to do the survey, so she didn't -- knowing that her grade could be lowered from an A to a C.

Three of the eight said they used answers from friends and relatives on some surveys, also in violation of survey ethics.

Goold said his office would discuss a course of action.

Regarding the judge, Goold said, "If he is aware of impropriety, he can notify the parties to be in court tomorrow" to address "falsity before the court."

Peterson's lead attorney, Mark Geragos of Los Angeles, said, "Hypothetically speaking, one should never put any credence in anonymous sources."

All students requested anonymity.

On the witness stand Thursday, Schoenthaler insisted that his methodology was sound when prosecutor Dave Harris questioned the survey's integrity.

"Is it possible that college students went home and simply made these numbers up?" Harris said later in court. "I think there is a significant likelihood of that."

Harris' arguments appeared to have little effect on Girolami, who said his decision to move the trial was heavily based on a "massive amount of publicity."

Schoenthaler said that he required the students who were conducting the survey to include the phone numbers they supposedly called when submitting data but that he had not verified any by calling them himself.

Formulas developed to detect fraud didn't alert him to anything unusual, he said.

Before The Modesto Bee published the survey results Sunday, Schoenthaler said he used 65 students to poll 1,175 prospective jurors randomly by telephone in late November and early December. He said 114 to 122 people responded in each of California's eight largest counties, split evenly between Northern and Southern California, plus Stanislaus and San Joaquin counties.

The survey suggested that more jurors without bias could be found in the Bay Area and Southern California than in Stanislaus County. Geragos submitted the survey as an official case exhibit, and the judge cited it when explaining his decision to move the trial.

Schoenthaler testified that no one paid him to do that survey or a smaller one in May. He has said he hoped to provide a public service and perhaps save taxpayer money.

In court Thursday, Harris questioned Schoenthaler's motives, suggesting that the professor sought public distinction. He said the survey was poorly designed and gave students the opportunity to falsify their data.

"You have to ask yourself what else is wrong with [the survey] when you ask college students for credit to go back to their house or their dorms to make long-distance calls," Harris said.

The students said Schoenthaler told them they could expect people farther away to know less about the case. They said they fabricated the surveys accordingly.

"You just make it up," said a 21-year-old student.

Schoenthaler gave the students survey materials two days before Thanksgiving, he and they said.

"It's just an asinine thing to make a student do a week before finals," a 22-year-old student said. "There is no way [Schoenthaler] can say this is legitimate because he wasn't there when we supposedly made it up."

A 21-year-old student said: "It's bogus."

Some students said they would have come forward earlier but they had no idea their fabrications would be used to help sway a judge making such an important decision.

A 35-year-old student said, "This is a death-penalty case. This guy's life is on the line. I'm absolutely outraged."

University spokesman Don Hansen said discipline for dishonest work can range from writing a paper or performing community service to suspension or expulsion.

Vice Provost Demetrulias said her investigation could take a week to complete.

"We take very seriously any scientific misconduct or suggestion of that," she said.

A class syllabus given to the students at the beginning of the fall semester states that 20% of their grade would be based on a class project.

The description:

"Each student will be assigned to survey public opinion attitudes and knowledge on the telephone from 20 people in various parts of California to test hypotheses that will be done in class. The survey typically takes five or six hours to complete and an hour of practice."

Stephen Lubet of Northwestern University School of Law in Chicago said, "The point is to teach students, not obtain their labor."