Sampling Test Prep AK
1. D. The researchers did a good job by sampling from sites around the city. But, by sampling only people who shop at farmer's markets, they are surveying only people who care about fresh produce. People who shop at grocery stores will be less likely to want organic fruit. The rule to remember is: the site that you use to sample should not be related to the think you are surveying about. People with strong opinions about fruit shop at farmer's markets, so it's not a place to find people with random opinions about fruit.
2. B. It's better to have larger samples rather than smaller samples. But, it's also better to sample people who are more likely to have random opinions about the topic. Students at football games are sports fans and students in Honors Society are probably more likely to care about AP classes. Surveying in those sites will bias the sample. You are not trying to talk about parents, so their opinions do not matter. Since all students go to nutrition and because the team is trying to survey students who hang out at all different parts of campus, getting 40 more students that way will be the best way to increase the sample size.
3. C. Participants in the group that got the drug lost more weight than those who did, so it is reasonable to think that the drug works. As the drug was only tested on obese men, the results can only be generalized to them.
4. A. The signs seem to work well at some sites and not at all at others. More research is needed to figure out why some sites had effects and others didn't. Because half of the sites showed no (or negative) results, the researchers cannot generalize to all sites. The signs definitely do not cause accidents, but it's worth more research into what was going on at that one site (the increase may have been caused by something else going on at the same time!).
5. C. Almost all of the students improved their reading over the school year. It looks like preschool works, but the small difference in percents suggests that if phonics programs on iPads do anything, they do not do much. It's possible either that a) all iPad work or b) all preschools help low-income students improve their skills. Because there was no "no iPad" at all group, we can't tell the difference. But, the study does not provide evidence for A or B and because only low-income preschools were included, the researchers cannot conclude D.