Question

During election season you will constantly see polls estimating the percentage of people who will vote for candidate A and candidate B. These polls are misleading because they always show the point estimate, p-hat, when they are really confidence intervals. If you look closely, the graphics will show in small print away from the results a statement saying something like "this polls has a margin of error of +/- 3 points".

Consider the following: You see a news article stating that 10,000 people were surveyed across the country and proclaiming that Candidate A will receive 52% of the votes and will therefore win the election. However, the margin of error is +/- 3 percentage points.

• What does the 52% represent?
• Should you believe the proclamation made by the news organization?
• How could they be correct or how could they be wrong?

Answer 1

The 52% here represents the proportion point estimate with margin of error +/- 3 percentage on the basis of 10000 people surveyed.

We should not belive this type of statement because we did not know about quality of survey which raises many questions like is sample really a good representative of population? which type of sampling method is used to collect data? is the data collected by rationale person? is there any tweaks in data processing?and many more.

First of all they should publish all kind of questions we have asked above and apart from that to conclude the result with point estimate they should give people with confindence interval for eg in or case ( 52 - 3 , 52 + 3) is equal to

(49 , 55) .

Candidate A will receive between 49 and 55 percent of votes on the basis of 10000 people surveyed.

please like;)