Under the mango tree: SWS fudging the surveys?

Thirty-six-year-old Janna (not her real name) and her friend were killing time on April 17 in a fast-food restaurant at an MRT station in Pasay City when they noticed six people seated at a table beside them busy filling out forms that bore the logo of the Social Weather Stations (SWS).

Janna, a professional taking up law studies, decided to take three photos with her iPad, and shared them with a chat group. “Guys, look!” she said. Then Janna and her friend proceeded to a reflexology center at the mall.

Her massage had just begun when one of her friends sent a message on their chat group, urging Janna to go back to the restaurant and take a video of the group with the SWS forms.

For her friends, something did not seem right with a group of people filling out what looked like stacks of preelection survey forms. The stacks were bundled according to districts and barangay in Pasay City.

Janna did as told and was lucky to find a table right beside the group. Only a glass panel separated them. She angled her iPad so that she could take a clear shot of the group while she pretended to read a book.

A copy of the video—all of 11 minutes—was obtained by the Inquirer.

Meeting with the Inquirer on Wednesday night, Janna recounted the incident, the questions that popped in her mind, and why she did not want to upload it on Facebook.

“I am an active supporter and volunteer of one presidential candidate. I do not want to be bullied on social media,” Janna told the Inquirer, explaining why she does not want her full identity revealed.

Like a sample ballot

The video does not show the group shading answers, or candidate preferences, but the women appeared busy filling out forms labeled “Call Sheet.” On the same form were the numbers “04-16.”

For preelection surveys, the SWS uses a form similar to a sample ballot, thus the shaded circles that could be seen on the video.

One woman, whose face could not be seen on the video, shuffled through one set of papers, then with a purple pen, wrote down a respondent’s name, address and the time that the interview supposedly started and ended. The woman wrote figures that looked like “3:15.”

The watch on the woman across her showed it was 4:40 p.m.

Janna said she wondered how the women worked with so much ease, as if knowing exactly what they were doing. They shuffled through the papers, knew which stack to get, knew what to write down, knew what answers to encircle.

Janna was curious how the woman filling out the forms seemed to have memorized what to encircle on the sheets of paper. For example, the respondent’s house was made of stone.

And why would the women work at a busy fast-food restaurant, which for Janna appeared to be a cavalier attitude if preelection surveys were considered very important given the impact of their results on the electorate and the elections as a whole?

“Shouldn’t these forms be sealed in an envelope and sent to the office for encoding?” Janna asked.

Under the tree

In gathering data for quantitative surveys, there is what is called “working under the mango tree,” a disparaging way of describing how field workers fill out questionnaires even without respondents just to complete the work.

The Inquirer showed the photos and video to SWS officials Jay Sandoval and Leo Laroza.

Laroza, SWS spokesperson, said the photos appeared similar to what is now being circulated on Facebook, with the allegation of cheating.

But he would not comment on the Facebook post because the SWS does not respond to “insinuations.”

But Laroza and Sandoval, SWS director for Sampling, Processing, and Data Archiving, said they would investigate the video.

They said the SWS has to determine if the forms seen on the video are authentic, adding the agency has caught several fake survey sheets in the past. Sandoval also said the SWS does not subcontract the hiring of field workers.

Sandoval and Laroza walked the Inquirer  through the survey process undertaken by SWS, the same procedure followed by their field workers whether these are preelection surveys or not.

Data transfer

Looking at the video, Sandoval and Laroza said that it appeared that the women were transferring initial notes they wrote on the other pages of the set of survey forms.

“Assuming that is our questionnaire, the time of the interview and (other) information were already recorded in the first part of the questionnaire where you started the interview,” Sandoval said, explaining the transferring of notes as seen in the video.

Asked if the “04-16” on one of the papers meant April 16 or April 2016, the SWS officials said they would have to determine first if the forms were authentic SWS forms.

The April 18 to 20 SWS preelection survey showed Davao City Mayor Rodrigo Duterte getting 33 percent of voter preference followed by Sen. Grace Poe with 24 percent, Mar Roxas with 19 percent, Vice President Jejomar Binay with 14 percent and Miriam Defensor-Santiago with 2 percent, and 3 percent remained undecided.

READ: First post-‘rape’ poll shows Duterte widening lead / Duterte still up after rape joke; Binay biggest loser

While the SWS could not show the Inquirer  the exact preelection survey forms due to a confidentiality clause with the respondents, they showed other declassified survey sheets.

That forms are being filled out after the interview and without the respondents present did not seem to alarm Sandoval and Laroza, as they gave logical explanations, if indeed, these were legitimate SWS questionnaires.

They explained that every survey consists of several forms to be filled out in a set of preelection survey questionnaires, not just the sheet for the voter preferences.

Interviewer’s report

Among the other forms that are included in the set of preelection surveys are the interviewer’s report, which notes whether the respondent was cooperative, etc.; the call sheet or the record of attempts or calls made to the respondent; and respondent’s demographics.

“In the questionnaires, there is a portion that would be filled out after the interview. For example, the name of the respondents, we have what we call a call sheet which you would (fill out) that while you are in the area coverage,” Sandoval said.

 Selection table

There is also the probability selection table where all the members of the household are listed down and those who are qualified to participate in the survey are identified. Field workers are usually assigned to five respondents per barangay.

“The selection table is in the inner part of the questionnaire,” Sandoval said, which could explain the shuffling of sheets by the women for every set of questionnaire in the video.

Sandoval said that SWS implements “very strict control procedures,” adding that the agency follows “zero tolerance for dishonesty.”

“We have a one-strike policy,” Laroza said, “especially if you are caught doing your work under the mango tree.” TVJ