in our defense

by | | 4 comments
It’s a topic that's attracted the attention of everyone, from the vendor peddling candy, to the office worker buying that candy. There have been lawsuits, media coverage, even t-shirts, about it. And no, I am not talking about the jejemon phenomenon.

Surveys.

This election season, surveys--specifically, election polls--have been given an incredible amount of attention. Most of this attention seems, to me, negative. It just always looks like survey firms are out to do their worst, to mess with the voter's mind.

To most people, surveys are those things that seem a regular part of the news these days.

To me, surveys are the way I make my living.

I am a Communication Research graduate from the University of the Philippines.
Simply put, my course delves into studying the ways people communicate with each other, and how these ways of communication manifest into all the other aspects of life. One of the ways we studied communication behavior is through surveys. From school, I learned how to conceptualize, conduct and analyze surveys.

Having recently graduated, I now work for a market research firm. Market research is a field that helps its clients find answers to certain questions they have regarding their products and services. One of the ways we help our clients achieve these answers, is through a survey. From work, I have seen how projects involving surveys are done, from project set up to reporting.

The recent attention given to surveys has been a little upsetting for me, and my friends who are in the know regarding surveys.

To borrow from a friend of mine; "I love what we do, and I'm sad that people don't understand it."

Because of this, I feel the need to atleast try and explain how surveys are done.

Let me start by saying I am no expert: I speak from the point of view of someone who has had the benefit of learning from experienced professors and skilled practitioners. I am someone who has experienced making and conducting surveys, someone saddened by the negativity they seem to have attracted.

So now, I will do my best to explain several aspects of surveys.

There is no way I can cover everything about surveys.

Perhaps the best I can do is give a little of the important stuff.

Surveys are tools of measurement, a set of questions about a certain topic that are put to a chosen person, that are based on self-reporting. That is, they ascertain a person's state of mind, based on what that person says. There are some limits to what surveys can and cannot measure. For example, it is easy to ask in a survey if a person likes or dislikes a certain television show. But it would be difficult to ask the reason why, as that would entail a longer explanation.

The first survey--the first recorded one, anyway--was done by Karl Marx. He asked a series of questions about the treatment of employees in a factory. The story goes that he sent out around three thousand questionnaires, though, sadly, none of them were returned.

Now. I just want to say that I do, on some level, understand where people are coming from. My father and mother, who have seen me struggle through countless projects involving surveys, still retain a more than healthy skepticism about them. And I get it: the idea that 1,200 people can represent 90 million is a little hard to swallow at first. It was difficult for me to get used to it, as well.

But I guess it is all in the way that you see it.

The idea of 1200 for 90 million comes from a basic survey principle: the principle of sampling.

Sampling is the reason, the basis, why we can say that the opinions of 1200 people reflect the rest of that 90 million.

Let's try a visual example.








Say you have a bowl of beautiful, obviously yummy champorado, sitting in front of you. (What? A girl gets hungry with all the explanations.)

You want to eat the champorado, but you aren't sure if you can. It might still be hot, it may already be cold. So what you do, is you grab a spoon and get a sample of that champorado, to find out if you can eat it already, or not.

In the same way, sampling is the process by which we select a part to test the waters of the whole.

Let's go further. Ofcourse, you can't just select any part of your champorado, right? Any experienced eater knows that the parts near the edge are always cooler than that of the middle. So, if you really wanted to find out your champorado's state, you would have to follow a few rules, to properly represent the whole.

This is the logic that is behind probability sampling, the sampling scheme behind most nationwide election surveys. Probability sampling provides every single one of the 90 million people an equal and unbiased chance of being selected to become a part of your survey. There are many types of sampling schemes under probability sampling, all ways that seek to make the selection of survey participants more accurate.

The common question I hear has always been, 'But how come I've never been selected to become a part of a survey?'

That is because of probability sampling: since everyone has a random, equal and unbiased chance of becoming selected, there is a chance that you will, and there is a chance that you won't be selected. Believe me, it is not because the survey firms are avoiding your house and your strong opinions.

Going back to our champorado, the question now is, where do we dip our spoon?

We can say that, to properly ascertain the state of our champorado, we'd divide our champorado bowl into four equal parts: upper and lower left, as well as upper and lower right. The plan is to get spoonfuls from each of these areas, to help us in our heat quest.

Comparing the Philippines to our champorado bowl, we divide into the three major islands. In most nationwide surveys, participants are selected from the islands of Luzon, Visayas and Mindanao, depending on area population. This is because, the more people in one area, the more participants are needed for that area. There are also safeguards in place for when people refuse to participate, or do not meet certain criteria.

Common reasons why a person can't participate in a survey include their refusal, their failure to complete the survey, and certain technical reasons, such as if they had participated in a survey before, or if they are related to someone who works for an organization that deals with surveys. In these cases, probability sampling provides additional rules so that survey interviewers can find an adequate replacement for the respondent.


Finally, having made our spoonfuls, we will now combine these impressions into the survey report. That is a long and infinitely interesting process, so we won't go into it.
Now, we go into reporting our spoonfuls.

Part of the reason why people seem to find surveys so astounding is because of the way they are reported and shown.

There are five important things to know about a survey: the number of respondents involved, the areas involved, the exact question asked, the confidence level and the margin of error. I do wish that they would report who commissioned the study, but I suppose that is just me quibbling.

The first three are easy: you need to know the number of people involved to learn confidence level, the areas to ensure representativeness, and the exact question, to know what these chosen people were agreeing or disagreeing with.

The concepts of confidence level and margin of error are related, and both are based on your sample size.

Let's stick with the champorado again. (Am I making anyone else hungry? I hope so.) When we selected our spoonfuls, there was always a chance that we did something wrong somewhere. Maybe we took a bit of the upper right when we were going for lower left, etc. Survey measurement has its pitfalls, after all.

The margin of error was conceived with this belief in mind. Except, with margin of error, we can ascertain exactly how our mistakes may or may not affect our survey's findings.

Ascertaining our margin of error can be done through our confidence level. The three: confidence level, sample size and margin of error all go together. There is a specific set of sample sizes and confidence levels related to a specific set of margins of error already in place. Like pi=3.1416, they are a constant in the survey research world (You can also compute manually for margin of error, but trust me. You do not want to do that.)

So, let's say you hear the report that a survey has a margin of error of +- 3 %. This means that a survey rating can be as right plus three, or as wrong minus three.

For example: candidate A is given a rating of 20 %. This rating can be as right as 23 % (your plus 3) or as wrong as 17. (your minus 3) I do not mean literally wrong; by this, I mean that his rating could be as high as 23, and as low as 17.

Now, a final question: Can surveys be cheated?

Yes, ofcourse, surveys can be cheated. I am of the opinion that anything can be cheated.
Personally, I find it demeaning to hear people say things like that. Having done survey fieldwork, I can tell you that it is not easy. The weather, the surly people, the commute, not to mention all the work that went into setting up that study. And then there are the people who commissioned your organization for what the survey.

To say that we would cheat our own results is not only insultingly unjust, it is irrational. Would anyone really go through the lengths of setting up their own downfall that way?

I wish to end with a personal opinion: while I value the art of surveys and the knowledge that they give us, it is important that we realize that surveys are just that.

They provide us with a spoonful of the current state of a person's thoughts and opinions at the time. But you and I are both human, and we know how often we change our minds.


Choosing your next president and other national and local leaders should not depend on what the rest of your country says. Even more, it should not depend on what the current showbiz personalities endorse.

From this, I implore you to rely on a more accurate measurement tool than what you find on the nightly news---your own thoughts and opinions. Do your best to become informed and aware, by reading candidate profiles, watching interviews and debates.

We all make the choice that we can best live with, regardless of popular opinion.

Only then can we say that we have given our right to suffrage proper justice.


***All knowledge a combination of all my Communication Research courses and on the job learning.
****Opinions are most welcome. :)
*****photo from: http://farm2.static.flickr.com/1159/700725161_58cea64c93.jpg

re-posts!

by | | 0 comments
I feel energized these days, so I am going to revive my blog.

When I tag a post as ye olde, it means it's an old memory, and that I want to preserve it.

Still dredging up memories for now.

a perfect non-starter

by | | 0 comments
I guess the Film Institute will always be special to our non-relationship. Taking my cue from Chuck, 'If our relationship were remotely real, FI would be our special place.'

But it's not, so FI will always be my special place.

It seems to be the only place you ever realize I'm alive.

Remember our last conversation, when you said hi to me during Gawad Plaridel? That was also in there. I had gone out to the restroom and when I came out, you were suddenly there saying hello. I couldn't do anything, so I nodded.

I nodded, but I wanted to sink to the floor and hit myself.

Afterwards, when even a glimpse of you breezing into the library was rare, I kept telling myself how irrational this was: no one, no person should have this much of an effect on someone else. Especially in the case of two almost strangers.

Which is what we are. I can't even be optimistic enough to use the word 'acquaintances' for us.

Because what we are, my dear, is nothing.

What we've shared, a dozen or so conversations, some shared walks, a library table...they only gain meaning from my hopeless infatuation.
Yes, I think I can now safely admit that I have an incredibly large crush on you.

Is it even necessary? I'm quite certain my blocmates' intrusion on our non-moments together would suffice. I do want to apologize for them; we seem to gain more immaturity as we delve deeper into research, so all their subtle (not) looks, gentle (not) jibes and soft voices (even more NOT) that alert me to your presence is something I can just chalk up to overwork.
All I am is thankful that we don't all share the same workspace...I doubt I could live it down if their overworked selves suddenly got bright ideas in an elevator.

But yesterday.

Yesterday, we weren't nothing.

And for once, there was no one there but us. No blocmates, no orgmates...us.

I wouldn't have talked to you. I don't know if you noticed, but I still have no idea how to behave around you.
I am able to when I immediately notice you are there. It gives me time to adjust, when I know you're there.
So I was ready, when we were lining up and being yelled at by the ushers outside. I knew you were in the vicinity, we are graduating together.
I had it all in my head, look at your course line, smile in your general direction, look away. I did it, too. Except you caught my eye and I had to look away.
And I thought, fine, last moment of humiliation. I am never seeing you again after this, so why not one last stroke to your ego? (Not that you have a large one...I think that's one of the reasons why I like you.)

And I just wanted to make sure that my mom and dad were comfy, so after that awkward little scene (to me, never you...it's always me, because in my AMAEverse, you know me.somewhat.), I had to go inside the theater.

But when you suddenly spring up in front of me, all barong and smiles, and touch my arm, and say, Congrats.

Just one word.

It's not even a real word. It's not even a world I like!

But you're there, you're smiling, we're still slightly touching.

So I summon up all my strength, and say it back.

Then you leave, and I leave.

And outside, the world rights itself.

You and I are nothing again.

But that moment, in the left-most corner of the theater, is one that will stay with me.

I thank you for these past five years, for being a constant inspiration, for being a part of the reason why I always think to dress my best, to actually care about my hair.

I thank you for having enough poise and good humor to ignore me and my blocmates' horrendous teasing, for having enough niceness in you to actually still speak to me even though I seem to do nothing but act like an idiot in your presence.

And most of all, I thank you for that moment of congratulations.

It was the perfect ending to something we never started.