3 ms·

There's something fishy about this infographic. "The Culture Score Index Series is based on further analysis of the NOP World Roper Reports Worldwide(TM) surve

by Link- 13y ago

There's something fishy about this infographic. "The Culture Score Index Series is based on further analysis of the NOP World Roper Reports Worldwide(TM) survey, which includes in-depth personal interviews with more than 30,000 people age 13 and older in 30 countries between December 2004 and February 2005." This is a relatively minuscule sample compared to India's population which is ~= 1.3Billion in 2004/2005. Same applies for many other countries. The results are not indicative at all.

thisisnotclear 13y ago
I am from India and I am pretty sure that Indian people are not the best readers in the world.
Wohui 13y ago
As disappointing as the sample size is (average 1k people per country), more important is the absence of info of where they all come from (pessimism says 1 city). (optimism says 2)
shalmanese 13y ago
It's easy to think that but it's not true. Regardless of population size, providing a sample is truly random, a sample size of 300 is enough to measure with an accuracy of + or - 3%. Increasing sample size much beyond this does not meaningfully increase accuracy. An analogy I've found useful thinking about this is, if you have a pot of soup, tasting a teaspoon is enough to figure out whether the soup is oversalted or not, regardless of the size of the soup pot, providing the soup is well mixed.
- Link- 13y ago
  Assuming the set is homogeneous! Which in this case, is not. A sample extracted from a more or less financially stable region of the country will have drastically different output than a sample extracted from a poorer area. Your assumption about 'the soup well mixed' is less probable to be true, given the lack of any evidence supporting the claim that this study is either meaningful or properly executed.
  - shalmanese 13y ago
    Right, if there's a flaw in the study, it's in the representativeness, not the sample size.

by Link- 13y ago

thisisnotclear 13y ago

I am from India and I am pretty sure that Indian people are not the best readers in the world.

Wohui 13y ago

As disappointing as the sample size is (average 1k people per country), more important is the absence of info of where they all come from (pessimism says 1 city). (optimism says 2)

shalmanese 13y ago

It's easy to think that but it's not true. Regardless of population size, providing a sample is truly random, a sample size of 300 is enough to measure with an accuracy of + or - 3%. Increasing sample size much beyond this does not meaningfully increase accuracy. An analogy I've found useful thinking about this is, if you have a pot of soup, tasting a teaspoon is enough to figure out whether the soup is oversalted or not, regardless of the size of the soup pot, providing the soup is well mixed.

Link- 13y ago
Assuming the set is homogeneous! Which in this case, is not. A sample extracted from a more or less financially stable region of the country will have drastically different output than a sample extracted from a poorer area. Your assumption about 'the soup well mixed' is less probable to be true, given the lack of any evidence supporting the claim that this study is either meaningful or properly executed.
- shalmanese 13y ago
  Right, if there's a flaw in the study, it's in the representativeness, not the sample size.