Stat 2 Lab #11: Sample surveys Learning objectives: 0 Identify various types of bias in sample surveys 0 Critique the design of sample surveys and analyze the impact of a flawed survey design 0 Describe probability and non-probability sampling methods 1. A study conducted at Stanford University in 2005 wanted to evaluate what percentage of students had used the drug Ecstasy. Two assistants were stationed on the main campus plaza and instructed to interview all students who passed through at specified times. As it turned out, 19% of 369 students interviewed said they had used Ecstasy at least once. Does the investigator's procedure give a probability sample of Stanford students? Answer yes or no, and explain. If this is not a probability sample, what term best describes it? 2. Suppose that a simple random sample of 100 residential phone numbers is taken in a large city (you may assume that each house has exactly one phone line) and an interviewer receives 68 responses after calling back at a variety of different times. In order to avoid nonresponse, the interviewer chooses additional random houses (same method} until the desired 100 responses are obtained. The total number of people living at those 100 households is 287, so the interviewer estimates the average household size in the city as 2.87. Is this estimate likely to be biased to be too low, too high, or unbiased? Explain