World Doesn't Trust Politicians

Data: Global feelings about business and government are shown to be pretty poor in a recently conducted global survey.

Worldwide, 63 percent of the 50,000 people questioned believe politicians are dishonest while 43 percent think the same term applies to business leaders, according to the survey, titled "The Voice of the People."



Some 52 percent feel politicians behave unethically, and 39 percent believe the same of business chiefs. But while 39 percent think politicians are not capable or competent, only 22 percent viewed their business counterparts in the same way.



Least trusted by their peoples, the survey indicated, are the political leaders of Latin America, West Asia and Africa with dishonesty ratings of 87 percent for the first, 84 percent for the second and 82 percent for the third.



Although in Western Europe as a whole 46 percent of the survey sample described their politicians as dishonest, in Germany 76 percent held that view, while 70 percent of Germans thought business leaders were dishonest too.

Hmmm, I wonder what kind of confidence interval this study has. A quick and dirty calculation shows that they sampled 0.0000083% of the population size. In other words, each person surveyed represents 120,482(*) peoples' opinions.



This sounds more like research looking for trends rather then looking for a precise figure of public sentiment.



Update: Migurski comments:

One thing to remember is that the relative confidence level of a truly random sample increases quickly with population size at first, then much more slowly. Or, as my intro-to-stats professor explained in college: if you know your sample is random, you may never need more than 1000 data points to represent any size population.



So the question is: how did they choose survey participants?

Good point. I wonder how one goes about finding a truley radom sampling of the entire population of the planet? I'm glad I didn't work on that study. :-)

Comments

[Anonymous]

One thing to remember is that the relative confidence level of a truly random sample increases quickly with population size at first, then much more slowly. Or, as my intro-to-stats professor explained in college: if you know your sample is random, you may never need more than 1000 data points to represent any size population.

So the question is: how did they choose survey participants?

-migurski

[Anonymous]

Interesting...
Wonder if there's a correlation with political affiliation... what do people vote who think politicians are dishonest??