Religion survey

[porchrat]

Two people equally educated - one religious the other not. The likelihood is based on the conclusion a person come to based on the information he/she receives. Not sure how research can come up with this notion of education level having a direct influence on a persons religious/non-religious state.

Let's once again take ToE as example. Two people oaf equal intelligence receive the information - one accepts the other shouts BS.

Someone doesn't know what "correlation" means.

[buyeye]

No you're not..

Why am I not?

[empirex]

Why am I not?

Because you're religious, and us religious folk tend not to be well-educated
I was being sarcastic

[buyeye]

Because you're religious, and us religious folk tend not to be well-educated
I was being sarcastic

Erm now that you mentioned it that darn emoticon was too far down for me to realise that.

Owell I'll stop skimming posts from now on.

[empirex]

Erm now that you mentioned it that darn emoticon was too far down for me to realise that.

Owell I'll stop skimming posts from now on.

Better?

[Rkootknir]

The sample group is about 1000 people per country which is far too small to be indicative of anything in my opinion. Especially when you're talking about as diverse a country as ours. I wouldn't read all that much into these results - the fact that they show double digit changes across the board indicates that either the world underwent some radical process (which didn't happen) or there is a problem with their dataset. I'm going with the latter...

The math says that 1 000 people sampled from a population of 50 000 000 will give results with a 3% margin of error on a confidence level of 95%. So, it can be quite accurate.

However, the 1 000 people in the sample have to be selected in such a way that they are representative of the entire population - these studies almost always assume a normal distribution. That would be quite difficult in SA because of the diversity you mention - but it can be allowed for.

[DJ...]

The math says that 1 000 people sampled from a population of 50 000 000 will give results with a 3% margin of error on a confidence level of 95%. So, it can be quite accurate.

However, the 1 000 people in the sample have to be selected in such a way that they are representative of the entire population - these studies almost always assume a normal distribution. That would be quite difficult in SA because of the diversity you mention - but it can be allowed for.

Which is why I mentioned the diversity aspect. It is nigh on impossible to draw any conclusions from this data-absent poll...

[buyeye]

Which is why I mentioned the diversity aspect. It is nigh on impossible to draw any conclusions from this data-absent poll...

Agreeing with you again , there are so many strata of our society .

[Nicko]

The math says that 1 000 people sampled from a population of 50 000 000 will give results with a 3% margin of error on a confidence level of 95%. So, it can be quite accurate.

However, the 1 000 people in the sample have to be selected in such a way that they are representative of the entire population - these studies almost always assume a normal distribution. That would be quite difficult in SA because of the diversity you mention - but it can be allowed for.

Yah finally someone who actually knows about statistics. People always moan when the sample size isn't near the actual population size, but it is near impossible to achieve that.

Which is why I mentioned the diversity aspect. It is nigh on impossible to draw any conclusions from this data-absent poll...

That is true, but you can't just automatically assume that they didn't try to get a accurate representation of South Africa's diverse populace. Anyway, it may not be such a drastic decrease as represented but I do feel that there are less religious people in the newer generations compared to the older, but that is just from personal experience, and most of the people I socialize with are from university which tends to be less religious.