The Benetech study was an effort to arrive at an accurate figure by comparing the reports provided by both Rahman’s and Ziadeh’s groups and four others as well as the Syrian government. The firm’s statisticians compared each of the databases with one another in an effort to weed out duplicates and insufficiently documented deaths. Researchers included only casualties that had been identified with a first and last name and a date and place of death.

The process yielded a list of 59,648 unduplicated death t reports from March 2011 through November. Of those, 76.1 percent were male and 7.5 percent were female. The sex of 16.4 percent could not be determined from the records, the report said.

But there were many questions that the report could not answer. For one, the analysis could not determine how many of those killed were civilians and how many were combatants. It also said that more than 70 percent of the records did not provide an age for the victim, meaning that the study could reach no conclusions about the death toll among children and the elderly.

The lack of information about whether the dead were bystanders or combatants also leaves open the debate over Syrian government tactics. Anti-Assad groups have consistently accused the government of targeting civilians in its bombardment of urban areas, a charge the Syrian government answers by claiming that the areas were occupied by armed rebels.

The way the various groups account for civilian casualties varies widely, underscoring the difficulty.

Ziadeh says his group’s numbers “indicate that 90 to 95 percent of those killed are civilians.” But Rahman’s Syrian Observatory sees a less lopsided ratio, with its numbers for November and December – 3,860 and 3,690, respectively – showing that only 42 percent of those were civilians.

Those variations exist even though both groups say they rely on the same basic methodology to gather their information: interviews with family members, photographic and video evidence, and evidence collected by activists on the ground to back up their statistics.

Both groups agree that violence peaked in August, when each counted for than 5,000 dead.

Rahman said, however, that he intends to present evidence to the United Nations that some of the death reports its study used included faked names and people who died from causes unrelated to the war.

He cited a recent attack on a gasoline station as example of the misrepresentation of some of the attacks that take place inside Syria.

“People said more than 30 people died,” Rahman said. “But no one had more than 12 names, or video of more than 12 bodies.”