By STEVE LOHR

Since the 16th century, the ideal of education has been the tutorial system pioneered at Oxford and Cambridge, nurturing young minds one to one, inquiring, prodding and encouraging. The tutorial method, research shows, is a proven winner.

But it is also highly elitist, hardly a system for educating the masses. So the drive for public education, in America and elsewhere, required a very different model — of one to many, with the teacher standing in front of a classroom, working from a textbook and lecturing. Education moved from a bespoke craft to a more industrial approach.

Today, though, 21st-century technology carries the potential to nudge mainstream education back toward the 16th-century vision of one-to-one tutoring.

The Internet, high-speed networks, powerful and lighter computers, and clever software for video, collaboration and simulations on the Web all help. Equally important is a maturing understanding of how to use wisely the new digital tools in education. The goal, proponents say, is to open the door to more engaged, interactive and personalized learning.

“The promise of technology is to take us back to the past, toward one-to-one learning,” said Monica Martinez, president of the New Tech Network, a nonprofit group that trains teachers and designs high schools that use computing extensively. “But this is returning back to that concept in a very different way.”

Skeptics question such enthusiasm. They say there are better uses for scarce budget dollars than high-technology gear, and point out that previous waves of enthusiasm for computerizing classrooms have come and gone, with little to show for it.

Yet there encouraging signs, and fresh evidence,that things are different this time, proponents say.

The Department of Education recently announced that it was developing a new National Educational Technology Plan to provide a “vision of how information and communications technologies can help transform American education.” The plan, the agency said, will include “concrete goals,” with a draft expected early next year.

There is also new evidence that education technology delivered outside of schools — online courses, for example — is steadily improving. A study conducted by SRI International for the Department of Education examined the research on online versus traditional classroom teaching from 1996 to 2008, winnowing the data down to measurable, fair comparisons. Its conclusion: “On average, students in online learning conditions performed better than those receiving faceto-face instruction.”

The analysis for the Education Department found that students doing some or all of their course work online ranked in the 59th percentile in tested performance, compared with the average classroom student scoring in the 50th percentile. That is a small but statistically meaningful difference, said Barbara Means, the study’s lead author.

Online education used to be mostly correspondence courses put on the Web. But no more, as interactive simulations for trial-and-error experiments become routine. An example, Ms. Means said, might be Web-based software to teach elementary school students the concept of density by testing if virtual objects float or sink in water. Size and weight alone, she noted, determines whether an object floats — and students can test their predictions against online simulations.

“Students are not repeating something they learned by rote, but making if-then judgments,” said Ms. Means, an educational psychologist at SRI International. “The more of that you can do, the more real learning goes on.”

Online courses alone, education experts say, will be used mainly to fill niches and will be most popular in university graduate-level and continuing education courses. No one really expects classrooms to go away or K-12 and college students to learn in isolation.

But online resources, experts say, will increasingly be used to supplement and transform classroom education, moving from stand-and-lecture formats to project-based learning.

“It’s a world apart from the old factory model of the high school with its rows of desks, textbooks and memorization,” said Ms. Martinez, whose organization has helped design 40 high schools in nine states and hopes to double that number by next year. Students in the New Tech schools typically outperform comparable schools in standardized tests.

For all its promise to improve education, technology is still no match for one human tutoring another — which, of course, cannot be used to educate large numbers of students and is expensive.

Still, one-to-one tutoring is the learning method proven time and again to sharply improve a student’s measured performance. A good human tutor can deliver a “home run,” educationally and statistically, explained Christopher J. Dede, a professor of education at Harvard University.

“With technology,” said Mr. Dede, an expert in learning technologies, “we don’t aspire to home runs, but good solid singles.”