Question

A local charity organization held an event to celebrate its work for the past five years. A local newspaper reporter who attended the event asked 75 of the attendees about their contributions. The reporter computed the mean amount of money raised by the 75 attendees to be $18,500. In an article published by the local newspaper, the reporter was quoted as stating, "The members of the charity organization were successful in raising money for their cause. Last year, they raised a mean amount of $18,500!"

Part A: What is a statistical advantage of using the median of the reported amounts, rather than the mean, as the estimate of the typical amount raised? (4 points)

Part B: The newspaper editor felt the individuals who attended the event may be different from the organization as a whole. A more detailed survey of the members was planned to determine a better estimate of the amount of money raised. The staff developed two methods based on the available funds to carry out the survey.

Method 1: Send out an e-mail to all 10,530 members of the organization and ask them to complete an online form. The staff estimates at least 2,000 members will respond.

Method 2: Select a simple random sample of members of the organization and contact the selected members directly by phone. Follow up to ensure all responses are obtained. Because method 2 requires more time than method 1, the staff estimates that only 1,000 members of the organization can be contacted using method 2.

Which of the two methods would you select for estimating the average yearly amount raised by all 10,530 members of the charity organization? Explain your reasoning by comparing the two methods and by describing the effect of each method on the estimate.

Answer 1

Answer:

Part A. The statistical advantage of using the median is that the median is less effected by skewed data and outliers.

Part B. I would select method two because method one has a non-response bias and the sample individuals are not selected at random.

Step-by-step explanation:

Higher values pull averages up where medians just select the middle value and aren't necessarily pulled by outliers.

Non response biases happen when the individuals chosen choose not to answer or can volunteer to answer.

Sending out emails to everyone at the org is not choosing them at random.