Government auditors told a House elections panel Friday that it may need to order a new batch of tests to determine whether electronic voting machines failed last year in Florida’s 13th district House race.

Nabajyoti Barkakati, a technology expert with the Government Accountability Office who appeared Friday before the House Administration’s special elections task force, said the GAO’s preliminary analysis suggests that equipment testing conducted by the manufacturer and state and local election officials was cursory and only included a small sample.

“Our analysis has found at least 112 different ways a voter could make his or her selection and cast the ballot in the Florida 13 race,” Barkakati said. “We found that the … tests [by state and local election officials] verified three ways to select a candidate … [other] tests verified 10 ways to select a candidate — meaning that of the 112 ways,13 have been tested.”

“We have not yet assessed whether this is significant,” he said.

The controversy, which the GAO has been ordered to settle, involves the November 2006 House race between Democrat Christine Jennings and now-Rep. Vern Buchanan (R). Jennings, who the state certified lost by 369 votes, alleges that the electronic voting machines contributed to 18,000 “undervotes” — when ballots are cast but a choice for a particular candidate goes unrecorded either because of ballot design, machine error or other factors.

Barkakati also told the three-Member panel that GAO experts are still determining whether the tests conducted by the manufacturer and election officials were rigorous enough. For example, it appears that voting equipment was inadequately tested in Sarasota County, where undervoting allegedly occurred.

“Our preliminary analysis has found that these sample sizes are too small to support generalization of the results to the overall populations,” Barkakati said. “The generalization of the results from the use of 10 machines for parallel testing cannot be supported because the sample drawn was not random and the sample size was too small.

“Our discussions with Florida officials indicate that such limitations resulted from court-imposed restrictions on machine access and resource considerations,” he said.

Rep. Vernon Ehlers (Mich.), the top Republican on the House Administration panel and a voting-machine expert, also recommended that the agency should look into whether something other than equipment malfunction could have contributed to missing votes. He said human error — or even poll-goers who purposely didn’t vote — are other factors that should be considered.

“That could be a crucial part,” he said.

Barkakati agreed the issue may have little to do with equipment malfunction. But re-creating the chaos of a polling place on Election Day, he admitted, would be impossible.

“It is true that we have not addressed the human errors,” Barkakati said. “It’s very much possible that the whole reason was ‘human factors.’”

“We couldn’t really figure out how to quantify that,” he said.

GAO officials also said lawyers are close to finalizing an agreement with voting equipment manufacturer ES&S to hand over blueprints and internal test results to its iVotronic line, the model at the center of the controversy. Since last year’s election, the company has successfully argued in court that the equipment’s inner workings are propriety information — a stance company officials appear to have abandoned since subpoena-power-wielding lawmakers stepped in earlier this summer.

Some election experts and House Democrats claim the company’s disclosure of the machines’ “source code” is crucial to determining whether 15 percent of voters didn’t cast a vote in last year’s House race between Buchanan and Jennings.

Earlier this summer, a Florida appeals court halted Jennings’ attempt to resolve the still-disputed election though the state judicial system. In May, the special three-Member elections panel agreed to have Jennings’ allegations explored by the GAO.

The agency has not given a firm timeline regarding how long its investigation may take, although a final inventory and to-do list could come before Members return from the August recess. The panel also agreed on Friday to discuss any developments that may arise during the recess.

Rep. Kevin McCarthy (R-Calif.), the task force’s lone Republican and a sometime critic of the process since House Democrats took over the investigation, said at the hearing he was encouraged with the panel’s progress.

“I am pleased with the progress that the GAO has made in its efforts to analyze the equipment and processes used in the election in Sarasota County,” McCarthy said. “I look forward to receiving additional updates from the GAO as it continues its work, including an analysis of the impact that ballot design may have had on the undervote in Florida’s 13th district.”