Pew Research Report on Religion in India

[Aupmanyav]

This is in reply to Jainarayan's topic, "Ok, so when are western “scholars” going to stop?" in Hinduism forum. I replied that they are never going to stop.

Read the following report to get some idea about religion in India. They have taken a sample of 30,000 people. This is not considered sufficient for a population as large as in India (1.4 billion). Statisticians said the minimum should have been 70,000. So, there can be doubts about the findings of the report.

Key findings about religion in India

 

[SalixIncendium]

Which statisticians? What parts of the report do you find to be inaccurate?

 

[Vinayaka]

This is in reply to Jainarayan's topic, "Ok, so when are western “scholars” going to stop?" in Hinduism forum. I replied that they are never going to stop.

Read the following report to get some idea about religion in India. They have taken a sample of 30,000 people. This is not considered sufficient for a population as large as in India (1.4 billion). Statisticians said the minimum should have been 70,000. So, there can be doubts about the findings of the report.

Key findings about religion in India

Statistically, 30 000 is indeed sufficient. (Far more than sufficient)

How to choose a sample size (for the statistically challenged) - tools4dev

 

[Jayhawker Soule]

Statistically, 30 000 is indeed sufficient. (Far more than sufficient)

The assumption, I assume, is that the sample is representative.

 

[Muffled]

Statistically, 30 000 is indeed sufficient. (Far more than sufficient)

How to choose a sample size (for the statistically challenged) - tools4dev

I would suppose a sample from Mumbai would give a good cross section but one from Goa might be skewed more Christian. According to the 2011 census, in a population of 1,458,545 people, 66.1% were Hindu, 25.1% were Christian, 8.3% were Muslim and 0.1% were Sikh. - Wikipedia.

 

[Vinayaka]

I would suppose a sample from Mumbai would give a good cross section but one from Goa might be skewed more Christian. According to the 2011 census, in a population of 1,458,545 people, 66.1% were Hindu, 25.1% were Christian, 8.3% were Muslim and 0.1% were Sikh. - Wikipedia.

I'm sure PEW had enough brains to take sufficient sampling from each state, proportional to the % of Indian population. If not, then they're not worth their weight in BS. Indians, because they don't get around India much, have an even stronger local bias than Americans or Canadians do. A person from Mumbai figures all of India is like Mumbai. That's why the influence of Hindi was surprising to me. In Mumbai the main language is Marathi, with Hindi being first in 'second language'.

 

[Muffled]

I'm sure PEW had enough brains to take sufficient sampling from each state, proportional to the % of Indian population. If not, then they're not worth their weight in BS. Indians, because they don't get around India much, have an even stronger local bias than Americans or Canadians do. A person from Mumbai figures all of India is like Mumbai. That's why the influence of Hindi was surprising to me. In Mumbai the main language is Marathi, with Hindi being first in 'second language'.

I suppose you mean this: The religious groups represented in Mumbai as of 2011 include Hindus (65.99%), Muslims (20.65%), Buddhists (4.85%), Jains (4.10%), Christians (3.27%) and Sikhs (0.49%) - Wikipedia.

I was surprised that Sikhs were so low. Punjab state is as expected.

About 57.7% of the population of Punjab state is Sikh, 38.5% is Hindu, and the rest are Muslims, Christians, and Jains. - Wikipedia

 

[Aupmanyav]

Which statisticians? What parts of the report do you find to be inaccurate?

It was the former head of the Indian Department of Statistics. I have not read the report in a way to find faults. You mean, I should give my views, comment on the report? I will surely do that in time. I will also post the comments made by that Statistician gentleman.

 

[Vinayaka]

I suppose you mean this: The religious groups represented in Mumbai as of 2011 include Hindus (65.99%), Muslims (20.65%), Buddhists (4.85%), Jains (4.10%), Christians (3.27%) and Sikhs (0.49%) - Wikipedia.

I was surprised that Sikhs were so low. Punjab state is as expected.

About 57.7% of the population of Punjab state is Sikh, 38.5% is Hindu, and the rest are Muslims, Christians, and Jains. - Wikipedia

There are very few Sikhs outside of the north. Punjab, and some in neighbouring states, and Delhi, due to it being the federal government place. What a lot of people don't realise is the high number of Punjabi Hindus. I met one here in Edmonton just the other day. I think the false assumption is that if you're Punjabi, you're Sikh.

 

[GardenLady]

As someone who spend a 36-year career on social science research, I question why a sample of 30K is not enough. And I'd like to know what statisticians suggested 70K.

 

[epronovost]

I don't see either what is objectionable in this survey. The sample size might be on the shorter end of the stick, but it's not small enough to make the results unreliable. It seems to have been well done and follow good protocols to avoid selection bias.

 

[Vinayaka]

As someone who spend a 36-year career on social science research, I question why a sample of 30K is not enough. And I'd like to know what statisticians suggested 70K.

The only reason I can see it being as large as it is is the diversity within India. It's comparable to Europe that way. What you get in Tamil Nadu is very different from Punjab or Bengal. But statistically, 30 000 is huge. In the footnotes, there is also reference to using the country's own census, which of course would have a 'sample size' of 100 % of the population. But essntially, a person with no background in statistics and numbers 'intuitively' feels it should always be higher. Election polls here, for example, in a population of 34 million, is usually around 1000. It does take a wee bit of mathematical understanding though.

 

[stvdv]

This is in reply to Jainarayan's topic, "Ok, so when are western “scholars” going to stop?" in Hinduism forum. I replied that they are never going to stop.

Read the following report to get some idea about religion in India. They have taken a sample of 30,000 people. This is not considered sufficient for a population as large as in India (1.4 billion). Statisticians said the minimum should have been 70,000. So, there can be doubts about the findings of the report.

Key findings about religion in India

Difficult to get a proper picture for the whole of India with only 29.999 sample size when there are 1400 million people (ca. 2E-3 %), even 70.000 would be only 5E-3 %. Depends how they selected the sample. IF they took random people in various cities that could be very inaccurate still.

BUT it did give a nice result, showing quite some tolerance towards other religion then one's own religion. I like that. That was also my impression of India Hinduism, but nice to see other religions have a similar few. So, maybe the 30.000 was sufficient.

 

[Moonjuice]

I would suppose a sample from Mumbai would give a good cross section but one from Goa might be skewed more Christian. According to the 2011 census, in a population of 1,458,545 people, 66.1% were Hindu, 25.1% were Christian, 8.3% were Muslim and 0.1% were Sikh. - Wikipedia.

If you read the report it says they took a sampling across all of India, in 17 different languages.

 

[Aupmanyav]

I think the false assumption is that if you're Punjabi, you're Sikh.

Also it is wrong to assume that if they are Punjabi Hindus, they would not have reverence for Sikh gurus. For many, Sri Guru Nanak Sahib is the only guru and 'Sri Guru Granth Sahib' is the only scripture.

The assumption, I assume, is that the sample is representative.

Yeah, that the sample is not restricted only to cities. That is not where India lives.

Please note that I have no comments on PEW survey yet. I have yet not read the report critically. I was only happy to note that people other than Hindus are satisfied with the religious freedom that they have in India and all were generally adjusting well with each other. Here is what Pronob Sen, the former Chief Statistician of India said about the report:
Exclusive Interview: Pew Survey Bias In Non-US Countries Cannot Be Ruled Out, Says Former Chief Statistician Pronab Sen
His biodata on Wikipedia: Pranab K. Sen - Wikipedia
Pronob Sen is an American citizen and a Professor of Statistics since 1965 at University of California, Chapel Hill.