More than 85% INDIANS back BJP over Aamir Khan : Times of India

[Nilgiri]

From where the samples were taken, what % do they represent?
How was the analysis done? Where they politically biased?
Who took part? Age group?
How did they filter out those with repeated accounts? or those with clicking habits?

How would a yes/no be politically biased? The bias would come by the TOI online readers being a particular subset of Indian population. There are no metrics regarding those since that would require a completely separate study, and then you have to correlate and create coefficients for correction etc. You lose the whole utility of doing an online survey.

This is the tradeoff between doing things quickly/cheaply as to more thoroughly. But the exact margin of bias is also unknown. In the end we have a "number"...kinda take it or leave it.

This brings up a conversation a bunch of us had about poverty rate measurement in India. A more detailed comprehensive survey method compared to the one before has created much interest from WB given the impact it had on the absolute poverty rate in India at: What state in India does what best? | Page 3

But that is not a binomial yes/no so it has many degrees of freedom (and variable measurement) for such a difference to happen in the first place.

As for how the "analysis" was done? There was no "analysis"....just a clear tally counter for yes and no's running on some server.

As for the who? We don't know, online polls are anonymous.

Repeated polling, I personally am not sure how they weed them out. Probably implementation of some IP based cookie. If its something more secure....it would be an IP logger + filter with some sort of firewall to detect poll spamming etc.

Both can be circumvented to some degree if you know how (cookie dump and proxy/VPN respectively)....if you really want to know I can test it for a TOI poll....or maybe another member already knows.

As for how many people are knowledgeable enough + have enough time/motivation to engage in this "cheating" ...who knows for sure. I would assume not many....and TOI may factor this in already in its results too.

 

[Dubious]

How would a yes/no be politically biased?

wait let me re read what I wrote...Coz i meant those analyzing and making sense of the "survey"

Are surveys in india direct and only give certain options? When we made surveys it always had to be AT LEAST 1 page long and extra questions to tweak out the people who would answer crap...

 

[Nilgiri]

Are surveys in india direct and only give certain options? When we made surveys it always had to be AT LEAST 1 page long and extra questions to tweak out the people who would answer crap...

Its an online yes/no based survey from what I understand and have seen of TOI surveys in general.

The long detailed surveying that you describe is generally not done online (unless some kind of reward/motivation is given)...since attention/time is at a premium online.

Yes/no is pretty much the only way to go.

 

[Butchcassidy]

I think he and his wife would feel safer in the caliphate. With all the love floating there, maybe he might get some

 

[Dubious]

Its an online yes/no based survey from what I understand and have seen of TOI surveys in general.

then i bail out and take all i said back!

The long detailed surveying that you describe is generally not done online (unless some kind of reward/motivation is given)...since attention/time is at a premium online.

for us it is always a long one to tweak out irregularities and biasedness

 

[Dragon4]

Its because he hurt sentiments of so called non-bhakt people too. Bad choice of words used by Amir.

 

[Nilgiri]

then i bail out and take all i said back!

Its a yes/no/cant say. They got one running right now on india-pak proposed cricket series.

for us it is always a long one to tweak out irregularities and biasedness

Well companies, govts etc...want better quality, "secure" data....and are willing to implement the time and resources regarding that...since much money/decision-making is based on it at some point.

TOI just wants a number to put other there to attact readership/debate (just like this thread). That is how they generate revenue for minimal cost.

I just checked, its cookie based repeater control. Not the most secure system, people can bypass it (I simply did so by just using another browser and I assume a cookie dump would have the same effect).....there may be an underlying IP logger/filter in post poll processing so again one would have to use a proxy to cheat to make absolutely sure that TOI can't detect it. But then again why would TOI really care beyond a point like I said, they are gonna make a money off a result anyway it goes.

 

[xyxmt]

I love it, I wish it reaches 90% soon.
I always believed India is for Hindus and everyone else should either leave India or change their religion

 

[PakEye]

Nope, they are already hurt because of the violence and pogrom which is openly allowed in India's west neighboring country, and what is happening to their Shia bretheren, hindus and Sikhs, Ahmedias has filled them with despise.

babey too kitni bari hogaai,look inside east neighbouring country the Dallit are equal to puppy according to BJP and shiv sena they are burned the children in fire and waht are NAMO doing for their protection.

 

[deckingraj]

I say it is questionable

Sure...that's what sampling is all about...It can never give you 100% accurate result, however if sample is well diverse then it can give you very good insight...

I asked this in a previous post however let me repeat....Post Poll Sample for General Election in India constitute somewhat 60-80K Sample size...If you see in that light 22K is a very decent size sample....So questioning sample size might not be good...however what you should question is how diverse the sample is... As said the sample is pretty much an urban sample...and thus put lot of holes in claims(which is very much urban in nature) that India has become intolerant...

Now complement this survey about logical reasoning that i gave in my last post which you chose to ignore for some reason....I am not sure i can share anything more to convince you...unless and until you have some contradiction in this context please consider this as my last reply.

 

[Dubious]

however if sample is well diverse then it can give you very good insight...

fair enough but how will one know it is diverse?

 

[Nilgiri]

fair enough but how will one know it is diverse?

Ensuring complete randomness. Easier said than done.

 

[deckingraj]

fair enough but how will one know it is diverse?

Exactly...there is no way!!...the only thing you can conclude is that it was an open forum and TOI having decent spread across country....it is highly likely that sample is decently diverse...Having said that nothing can be said for sure...but think this way.....atleast there is a sample which is highly likely to be diverse...on the other hand there is a brigade which say's India is becoming intolerant without any study to prove that....see the difference??

 

[GR!FF!N]

A security threat doesnt need to be defined...If you dont feel secure based on things happening in your surrounding - things happening to people in your village, of your caste, of your religion or what not.......ANY parent would feel threatened!

Wrong.What you're suggesting,we can term it "Insecurity without any reason".Like I'm feeling insecure in India as some fathead raised ISIS flag in Kashmir and want to leave it right now.This kind of insecurity knows no bound.And I can assure you,in that case,You can't find any safe refuge in this world,be it Desert of Sahara or Manhattan.Crime will be committed everywhere any you'll freaked out without any reason.

 

[livingdead]

one should read times of india comment section.. to know their readers..
in this specific case though i think its just a hyperbole.. nobody thinks that a family which belongs to top 1% privileged class really faces danger.