Syrian activists have sharply criticized a new opinion poll that found most people in the city of Raqqa, which is controlled by the extremist group Islamic State, believe that IS has had a positive influence on the country.
The poll was commissioned by the BBC from market research firm ORB International and examined public opinion in Iraq and Syria.
It asked 1,365 people across Syria's 14 provinces about various issues, including whether they thought IS had a positive or a negative influence on the country.
While most Syrians said IS had a somewhat negative or completely negative influence, a majority -- some 70 percent -- of the 53 people polled in IS-controlled Raqqa said IS had a somewhat positive or a completely positive influence on events in Syria.
Johnny Heald of ORB International told the BBC that before polling in IS-controlled areas, his team visited the head of the town and asked his permission to randomly interview people.
Heald told the BBC that IS agreed to the poll because "as the data verifies, many of those living in Raqqa now are happier since IS took over."
"They welcome the security, they see IS trying to help the people with electricity, with food, with petrol. In many respects it is a story they are keen to tell," Heald added.
But Abu Ibrahim Raqqawi, an activist with Raqqa Is Being Slaughtered Silently (RBSS), a group that documents IS's abuses, reacted angrily to the BBC poll and Heald's comments.
"This is crap," Raqqawi told RFE/RL. "It's a lie. The people in Raqqa are not happy."
Raqqawi also challenged Heald's assertion that IS has provided Raqqa residents with basic necessities and security.
People in Raqqa "can't find food and electricity and water," he told RFE/RL.
Water in Raqqa Province is "undrinkable" and has been so since IS took over the city, Raqqawi said.
And RBSS reported on September 18 that drinking water in al-Tabaqa city west of Raqqa has been brown and infested with worms lately.
Local Syrians pay inflated rates for water and electricity, according to activists.
But maybe it depends who you ask.
After all, IS provides free electricity and health care to foreign militants living in Raqqa.
Too Scared To Complain?
Why would people in Raqqa tell pollsters they thought IS was a positive influence, if they did not believe this to be the case?
Raqqawi said residents are simply too afraid of IS to say what they really think.
If they refuse to say they are happy in Raqqa, IS would simply behead them, he added.
"People are scared of everything," Raqqawi told RFE/RL.
IS has cracked down hard in Raqqa against anyone who speaks out against its rule.
The militant group has murdered activists who have opposed it, including at least one RBSS member.
RBSS said this week that the militants are scouring internet cafes in Raqqa city to find people who are working against them.
As part of its reign of terror, IS has taken steps to cut off residents from the outside world, restricting Internet access in Raqqa city and banning "dangerous" foreign TV channels in areas under its control.
Polling In Raqqa
So just who did ORB ask in Raqqa? And given the climate of fear in IS-controlled areas, could respondents really say what they truly think about IS?
RFE/RL attempted to contact ORB's Heald for further details on the polling methodology used in Raqqa but had not received a response from him by late on September 18.
Henry Potts, a chartered statistician at University College London, set out a few of the questions that he would put to ORB in order to better assess the validity of the poll.
Potts said it was important to know more about how ORB tried to randomly interview people in Raqqa.
Even though pollsters try to contact a random sample of people, those who respond will probably not be random -- and that can introduce bias.
"We know this is a problem in the U.K. and it will only be worse if people are concerned for their safety," Potts said.
Potts also wanted to know if IS guided ORB's interviewers to specific areas in Raqqa.
If the interviews were done in public spaces, that would have implications for who the pollsters met and also what people were willing to say, said Potts.
People being dishonest on polls is problematic even in the West, because people sometimes feel socially pressured to respond with answers they think the pollster wants to hear.
"Clearly the situation in IS-controlled areas is far more dangerous," Potts said.
"People may lie outright, give the answer they think is wanted or is safe -- and people can also decline to take part, which biases the result."