Superfast SIMs for expats & tourists
Explore Japan without limits with Mobal. Stay connected with reliable SIMs and eSIMs, with solutions for both expats and tourists. Get a real Japanese phone number from the best Japan SIM service for English-Speakers - it’s easy to use, there are no contracts, and delivery is free! Join Mobal today.
9 Comments
Login to comment
Skeptical
Important and reliable?
Let’s take important first.
To be important at all, it must disclose information over how it was done. And who paid for it. Not disclosed? Then not important. And then we move on to reliable.
Reliability? If they disclosed how it was done, and who paid for it, next we will give small value to 'instant polls.' Anything that asks anyone to “vote on our website,” or “use your smart phone to call,” are meaningless. Full stop.
Next it must be random. How were potential participants selected? If a specific person is selected for a particularly anticipated response, based on past polling answers, then we must move on from it as well. Not random? Then useless.
Next, was it representative of the desired population? For example, if 60% of the voters in a prefect are under the age of 40, but the poll only had 15% of those from that group, then it is not representative. And it will not be accurate.
Next, how was it designed? What were the questions. Anyone in sociology or political science will tell you how (very easy) it is to put the fat thumb on a scale when you are measuring public opinion. Extreme example, perhaps, but when you ask a voter, “do you prefer a thoughtful, family oriented candidate like _ , or would you whether vote for a confused, nearly broke, felon-in training candidate like __ ." This matters most, so if it is fixed, it is worthless, and we move on.
Size matters. If you found a truly representative demographic, say for an upcoming election in a voting location, but the sampling size was pathetically small, then it is likely useless, and we move on.
Factors such as margins of error, statistical and discrepancies between multiple surveys, and variability and bias, also matter.
And the people who conduct polls, and the people who pay for them, know all of the above. So if they won’t disclose the details, do yourself a favor, and ignore them. And move on.
virusrex
No poll is perfect but as long as they come with methods and the people responsible of doing it it can be judged how reliable they are. Contradictory polls that are both valid can happen depending on how they are conducted so the important thing is not to assume any single poll applies universally but take the small amount of time to evaluate how they were conducted and take the reported results according to that.
In the same way, If no details about the poll are available then there is nothing to gain from it.
Sven Asai
They are important for making general plans and scheduling in society or get an idea of how the different opinions are distributed among citizens. Most of them are also reliable in a statistical sense. The percentages published only represent the average or center value of a certain interval in which the real values are between a lower and upper limit. But now the real problem. The questions are often misleading and often exclude the choices that cannot be chosen as an answer because they are intentionally omitted.
Speed
About as reliable as this one.
Aly Rustom
depends on the poll and who's conducting it.
Tom San
Here at JT?
Do I really need to elaborate?
Eastman
these are normally manipulated ones in most cases so I just pass.
ZALD
Given the fact that some opinions are sometimes manipulated by one authority, it is helpless to be said that they are not so reliable. At the same time, thanks to wider use of commenting system such as SNS, we are individually able to express our opinions or convey what is occurring in a real time, being secured by an anonymity. Therefore, if we have a moral of how to use the Internet, opinion polls might be no longer important than they used to be.