JOHNSTON — In a computer lab at Nicholas A. Ferri Middle School, 21 sixth graders on laptops scrolled through a series of questions — not the kind asked in English, math or science.

“Do you think you will drink alcohol within the next six months?”

“How often in the last 30 days has your family … encouraged each other to stay away from cigarettes?”

“How much do you agree or disagree with the following statement: I will have more fun if I don’t drink alcohol?”

Statewide, 3,400 sixth graders — roughly a third of public school sixth graders — are participating in a $3.5-million research experiment, Project BEST. Designed by researchers at the University of Rhode Island, the computerized program aims to guide pre-teens away from smoking and drinking before those risky behaviors take root, as well as encourage them to eat right and exercise.

The program, financed by a five-year grant from the National Institute of Drug Abuse, is being run by URI’s Cancer Prevention Research Center, which conducted similar projects for older students and adults in the 1990s that showed improvement, according to the center.

The computer survey asks the students a series of questions and provides support in curbing risky behavior. The students’ answers trigger follow-up questions specific to that student. Researchers hope Project BEST’s personal approach will help it succeed where other programs targeted toward preventing risky behaviors among middle school students have failed.

“We haven’t had a lot of success so far with classroom approaches in health class or with DARE officers talking to kids,” said Wayne Velicer, a URI psychology professor who is principal investigator for the project. “We hope a more private and individualized approach will prove more effective.”

National studies show the middle school years are critical, a time when many young people fall into bad habits that can stay with them into adulthood.

A third of smokers started by age 14, according to the American Lung Association. In a 2005 Rhode Island survey, Rhode Island Kids Count reported that 21 percent of middle school students said they had used alcohol in the previous month, as did 44 percent of high school students.

The URI team tailored the project’s interactive computer survey to the tastes of 11-, 12- and 13-year-olds, adding music and visual images that appeal to adolescents. They also designed the survey to adapt information according to each student’s answers.

“This is not a one-size-fits-all approach,” said Karin Oatley, research project manager. “Each child has their own intervention, based on how they answer the questions.”

In all, there are about 10,000 possible “interventions” generated by the computer survey, Oatley said.

By asking sensitive questions electronically and assuring students their answers will be kept private, the research team hopes students will answer honestly. The program is also designed to help keep them honest, said Velicer, a developer of computer-based interventions to promote healthy behaviors and prevent disease. For example, if a boy, who is identified by a number only seen by researchers, answers he doesn’t smoke and 10 questions later responds that he has had 14 cigarettes in the past week, the program recognizes the inconsistency and adjusts accordingly.

Students in middle schools in Bristol/Warren, Burrillville, Central Falls, Cranston, Foster/Glocester, Jamestown, Johnston, Middletown, Newport, North Kingstown, Portsmouth, South Kingstown, Tiverton, West Warwick and Westerly are participating in the project for the next four years.

At Ferri alone, more than 200 sixth graders are taking part, with their parents’ permission. Less than 2 percent of parents of students at participating schools declined to sign off on the project, say the researchers, who will track the students through ninth grade.

After an initial assessment that surveyed students about their behavior in all four areas, half the students are asked questions related to alcohol and tobacco use, while the other half will be asked questions about diet and exercise.

Teachers have told the researchers that students seem to enjoy the surveys and like coming to the sessions, Oatley said.

The researchers will visit the participating schools twice this year, while the students are in sixth grade. The researchers will visit three times next year when the students are seventh graders to gather information about their behaviors and provide more “interventions,” and visit them twice in eighth grade. At the start of the students’ freshman year of high school, the researchers will visit for the last time to gauge how successful the program has been.

“Each year, we will be asking the student new information, as people change over time,” Velicer said. “We also give them information about how they responded the previous year and how they compare to other students. If this succeeds, I hope this would become available to lots of schools throughout the country.”

Researchers plan to publish several reports as they analyze student responses over the next few years, as well as at the conclusion of the project.

“Let’s say the ordinary smoking rate would be around 18 to 20 percent by the ninth grade,” Velicer said. “We are hoping to cut that in half. Same with drinking.”

jjordan@projo.com