The steady growth and mediocre performance of Georgia’s virtual charter schools tell us two things: Parents want online education options for their children. And Georgia’s not yet giving them great ones.

As a recent The Atlanta Journal-Constitution investigation found and studies suggest, students in computer-based virtual schools don’t perform as well as peers in traditional classrooms. Georgia Cyber Academy — the largest online program enrolling 14,200 of the 20,000 full-time students in statewide virtual schools — earned a D on the state’s report card.

After evaluating Georgia Cyber Academy for the Governor’s Office of Student Achievement, Georgia State University researcher Tim Sass said, “They seem to be performing poorly relative to kids who look like their students. The school is educating a lot of kids in Georgia. A lot of resources have gone into a program that, at least on the surface, doesn’t seem to be performing very well.”

The online industry rightfully counters that research thus far hasn’t been able to fairly compare the outcomes/grades of virtual students with those who applied to virtual schools but were turned down. (With unlimited capacity, online schools don’t turn down many kids.) Those comparative results would tell us whether the poor performance owes to deficits in the students or the schools.

That doesn’t mean the state of Georgia should ignore the troubling evidence available today about online schools while working toward better evidence tomorrow.

For whatever reason students choose to learn alone in front of a computer — academic, social, personal or medical — the current approach is not serving most of them well. A study released earlier this month found Ohio’s 35,000 online students perform worse on state tests than similar students in brick-and-mortar schools, even accounting for prior achievement. The findings have sparked discussions over whether virtual learning ought to target kids who are independent learners, disciplined and self-motivated.

But Georgia Cyber Academy is not allowed to hand-pick its students, and June Ahn, the New York University learning technology researcher behind the Ohio study, said selective admissions could be seen as undermining the principle of a public education system open to all. “The second option is to think really hard about how to set up these schools to serve kids well. My conjecture is that we need to focus on increasing interactions with teachers and peers and have more active learning,” said Ahn in an interview.

Ohio is debating moving to outcome-based funding of its e-schools, a system whereby providers are paid by student competency rather than mere enrollment. That’s already the funding lever in New Hampshire, which developed an online school that emphasizes teacher-student relationships and has produced performance similar to or slightly higher than the state’s brick and mortar schools. Georgia ought to also consider the New Hampshire model.

Jay Greene, the head of the Department of Education Reform at the University of Arkansas, sees online schools as serving a niche group, saying, “I am skeptical of the claim that virtual education will transform education because I think most people benefit from in-person interactions and social connections to motivate us to learn more. I am perfectly willing to believe that some students benefit from virtual education even if the average student does not.”

But it’s too early, said Greene, to declare that online learning in Georgia is good or bad. His sensible advice: “Be careful about it. Commission a rigorous study by setting an enrollment cap. If the evidence is encouraging, lift the cap.”