Final answer:
The meaning of a 2.3% margin of error is that the actual percentage of American adults who don't use the internet is within 2.3% of the reported 15%, higher or lower. Factors not covered by the margin of error include various biases and sampling issues. A decrease in confidence level from 99% to 90% results in a narrower confidence interval.
Step-by-step explanation:
The correct interpretation of the margin of error in the Pew Research report stating that 15% of American adults do not use the internet or e-mail with a margin of error of 2.3 percentage points is option D: They are pretty sure that the poll result differs from the actual percentage of American adults who don't use the internet or e-mail by 2.3% or less. This means that if multiple samples were taken and the same study was conducted numerous times, 95% of the confidence intervals created from these samples would contain the true proportion of American adults who do not use the internet or e-mail.
Factors that are not covered by the margin of error in a survey's outcome might include response bias, non-response bias, question wording effects, and sampling bias such as those introduced by using only landline telephones in a time when many people use cell phones exclusively.
If a confidence level were to decrease from 99 percent to 90 percent, the confidence interval would become narrower. This is because a lower confidence level means you are taking less of a 'confidence risk,' and as a result, the interval that estimates the true proportion is not as wide.