If you are a regular visitor to The Brunei Times website (www.bt.com.bn) for news, you'll notice that they have on the right hand side panel a weekly online poll. Most news agencies have them to engage their readers and also to get act as an informal (but bearing in mind unreliable) thermometer to various issues and agendas.
The tricky thing with surveys and polls is that you need to come up with questions that will elicit an objective unbiased response and according to many research methodologists, the structuring of the question itself is EXTREMELY important to avoid response biases.
Take for example this week's question for The Brunei Times online poll:
Something must be done to prevent the quite often power outages in this rich nation.
If you were to ask any educated/sane/coherent adult to answer this question, then most people would agree that YES it makes sense that something SHOULD be done. After all, who in their right mind would even suggest that getting the problem of power outage fixed is non desirable. So if the survey questions logically points to one answer, and you expect everyone to answer a particular way, then why bother to ask the question in the first place.
To drill this point of response bias home, consider this question DebatingBrunei came up with:
It is in Brunei's best interest and future economic wellbeing to diversify from oil and gas.
Would ANY of you actually click or choose Disagree? Chances are that no one would choose No, because by disagreeing, you're effectively condemning Brunei to a post-oil Brunei with the economy of a third world country. Therefore the question becomes pointless and is simply a wasted exercise in asking people to state the obvious.
Let us look at another question from The Brunei Times online poll:
Do you agree with the Government that fuel subsidies should be reduced step by step?
This one contains what they call a double barreled question, i.e. a question that contains 2 issues that can be answered differently. In this particular example, one part of the issue is whether you agree/disagree that fuel subsidies should be reduced. The other issue is whether it should be reduced step by step. These are 2 entirely different issues. I can just as easily say I agree with reduction but NOT step by step and do it in one go. So the way the question is structured makes my response inaccurate as it assumes I have given the same response to 2 different issues altogether.
Survey and Polling questions should generally be designed with care so that responses are not biased, or are loaded to get a particular response so as to manipulate and redirect a particular response to an agenda or issue. And in the case of The Brunei Times, a media organisation, with good reach into the community with regards to social and economic issues, flawed survey and poll questions can have far reaching consequences.
Not sure who works on The Brunei Times online poll questions, but they need to get someone to design them better.
For more info on Response biases, click here.
The tricky thing with surveys and polls is that you need to come up with questions that will elicit an objective unbiased response and according to many research methodologists, the structuring of the question itself is EXTREMELY important to avoid response biases.
Take for example this week's question for The Brunei Times online poll:
Something must be done to prevent the quite often power outages in this rich nation.
If you were to ask any educated/sane/coherent adult to answer this question, then most people would agree that YES it makes sense that something SHOULD be done. After all, who in their right mind would even suggest that getting the problem of power outage fixed is non desirable. So if the survey questions logically points to one answer, and you expect everyone to answer a particular way, then why bother to ask the question in the first place.
To drill this point of response bias home, consider this question DebatingBrunei came up with:
It is in Brunei's best interest and future economic wellbeing to diversify from oil and gas.
Would ANY of you actually click or choose Disagree? Chances are that no one would choose No, because by disagreeing, you're effectively condemning Brunei to a post-oil Brunei with the economy of a third world country. Therefore the question becomes pointless and is simply a wasted exercise in asking people to state the obvious.
Let us look at another question from The Brunei Times online poll:
Do you agree with the Government that fuel subsidies should be reduced step by step?
This one contains what they call a double barreled question, i.e. a question that contains 2 issues that can be answered differently. In this particular example, one part of the issue is whether you agree/disagree that fuel subsidies should be reduced. The other issue is whether it should be reduced step by step. These are 2 entirely different issues. I can just as easily say I agree with reduction but NOT step by step and do it in one go. So the way the question is structured makes my response inaccurate as it assumes I have given the same response to 2 different issues altogether.
Survey and Polling questions should generally be designed with care so that responses are not biased, or are loaded to get a particular response so as to manipulate and redirect a particular response to an agenda or issue. And in the case of The Brunei Times, a media organisation, with good reach into the community with regards to social and economic issues, flawed survey and poll questions can have far reaching consequences.
Not sure who works on The Brunei Times online poll questions, but they need to get someone to design them better.
For more info on Response biases, click here.
No comments:
Post a Comment