Polls and Opinions: When the people are a metre...

During the week February 27th - March 3rd I ran a poll on this blog. Now, I don't necessarily trust or believe in polls because a long time ago a friend of mine whose area of research is qualitative data analysis in the social sciences told me polls can be manipulated. She also said, but people never believe you unless you can quote statistics and figures. So generally I avoid quantitative data and go for qualitative data: interviews, comments, observations etc. But, having spent some time in the trenches of academia, I know how much people prefer to see numbers over opinions....and in fact, numbers here have come to represent public opinion as is evidenced in the numerous polls used in print, audio and visual media on a daily basis. The tricky thing is that how you shape questions can easily shape how people respond and even whether people respond. I'm sure anyone who regularly views the TV6 News complains regularly about the wording of its questions; wonders at the sense of them at times; and probably even wonders how exactly are they meant to answer them.

Anyhow, the week of the No Confidence Motion I decided to use a poll on my blog to get an idea of what the numbers were like in terms of who was pro or anti the motion or the government from the readers of my blog.

I've since been taken to task for using a poll on my site and labelled cognitively dissonant for being against polls, yet using one on my blog.

What can I say....is my lot in life to be labelled. But I am going to fulfil my promise and give a breakdown of the polls results and discuss a few other things that I feel are pertinent.

I used the website Poll Daddy, which allowed me to choose from one of their templates. The decision to use the poll  as I stated above was to have a sense of numbers, to go with the comments and observations that I had taken note of on FB. In the months since starting the blog I had noticed a significant trend...fewer and fewer people actually clicked "like" or commented on my FB wall....but the traffic from FB to my wall was higher and higher. So I felt looking for figures on FB would be less successful because people would have to "openly" give an answer; whereas a poll allows for some level of anonymity.

On the day of the poll more than 1100 people actually visited the blog, and according to Poll Daddy's statistical report on the blog 851 people voted.

Now, the questions I used were as follows:

Do You Have Confidence in the Kamla Persad-Bissessar-led People's Partnership Government?

The response options were: No,  this government is not leading us or living up to its responsibilities.

Yes, I think they doing a great job. Three thumbs up!

We have a government or running on autopilot?

If you read those response options well you can see that my wording of it was geared to influencing opinion and how people decided to choose to respond. Only the most die-hard supporter of this government would click on yes....and I wanted to find out how many die-hards were out there because on my wall I was only hearing that people didn't have confidence in the government. But, I am aware that the blog gets a wide cross section of readers, both locally and internationally; so, I wanted their feedback.

Was the poll objective? No, but the ironic thing is, the response that I was hoping most people would select - option C as it were - was in the lead for all of 15 minutes. Thereafter, the poll seem to fall out of my control. I posted to several group walls on a daily basis for 6 days. Every day giving the the statistics for whichever response was in the lead. Friends posted to their wall and friends of friends posted to their wall as well. Many of these group walls had political activists or supporters from the three major parties here. By the Wednesday of that week, people in support of the Government were ahead by a significant number. I posted this figure first thing that morning. On a pro-PNM group wall, a gentleman who claims to be versed in statistics and polling said that the UNC blogging machine seemed to be working overtime and that give him an hour and he could reverse the results of the poll. Immediately I switched to inbox messages and asked him if he was serious, and he said yes, he had worked with polls and statistics for a long time and could influence any eletronic poll. He went even further to say ( and I am paraphrasing here)  that if the PNM was willing to pay him he could ensure that every politically-oriented media poll came out in their favour. He then told me give him an hour and go back to the poll.

As he predicted the poll had swung back to reflect that voters were not in favour of the government, and remained that way for the rest of the week. I wracked my brain to figure it out. I had adjusted the settings on the poll to ensure that no isp number could vote more than once. To date, I don't know how he did it...I just know it was done.

The poll's results, that was available to anyone viewing the blog went as follows: No Confidence in the Govt: 53% (451 votes)

Confidence in the Govt: 43.24% (368 votes)

We have a govt or running on autopilot: 3.76% (32 votes)

I had hoped that most people would have gone with the sarcastic "we have a govt???" response to show their lack of faith in the system entirely. But some thing struck me during the week. The voting on the blogpoll seemed to take on a tribal tone. It was almost as if political lives depended on the opinion of the blog. I saw it in how the link to the poll was posted along FB and even in some of the comments sent to me on the blog, on FB walls and in my inbox.

And it made me wonder about public opinion, and how important it is for people to feel that others agreed with them, and maybe how important it is for the government to feel that the population agrees with it as well.

A common assumption these days, ever since a Trinidad Express story last year buss the mark that the UNC hires people to blog for them is that many of the media polls (print, radio and visual) are influenced by paid bloggers. Indeed many of the TV6 People Metre questions come back with responses sometimes that baffle me. Take last Thursday's question that asked, in the aftermath of Minister Rambachan's comments about "good cultural values" whether the PM was indeed reflecting good cultural values, the result came back that 70% felt that the PM was exhibiting good cultural values, while 30% disagreed.

A friend of mine sent me a bbm message saying "Marsha's minions working overtime tonight, boy". And I wondered to myself, were the bloggers working over time? Or is it that people honestly don't see anything wrong in the PM's behaviour. It's hard to tell now.And it further leads to my apathy where polls are concerned.

Prior to Faine Richards' story about bloggers being paid to sway public opinion, I'd have said it was just poor/crafty sentence construction on the part of the media or pollsters that was influencing public opinion. Now that running polls and taking polls have become full time jobs for people, I'm even more confused and question their validity even more; even though I work in an environment that requires you have figures to back up what you're saying.

Then this weekend another poll of sorts came in and it had me wondering even more. The UNC, as part of a coalition government has been in power for 22 months. On the basis of all the polls I read and hear about around me, as well as on the basis of the comments you read on many blogs and group walls on FB they are still immensely popular and liked by their support base....but the support base I interact with are be a couple thousand, who have internet access. This weekend was the party's internal elections. This is a party that has a voter base of at least several hundred thousand, and a registered voters' list of 81,000. Yet, on elections day Saturday roughly 14,000 people actually made the effort to turn out and vote....if that isn't a public opinion poll. I don't know what is. And somehow, I doubt it was manipulated.

De Vice Cyah Done!