week3 - Lecture 6 Nancy Pfenning Stats 1000 Designing...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Lecture 6 Nancy Pfenning Stats 1000 Designing Observational Studies The main advantage of an experiment is that it gives researchers as much control as possible over the explanatory variable of interest. One disadvantage is a possible lack of ecological validity if the setting is too unrealistic. Example The article When your hair’s a real mess, your self-esteem is much less reports: “A Yale University study of the psychology of bad hair days found that people’s self-esteem goes awry when their hair is out of place. They feel less smart, less capable, more embarrassed and less sociable.” The study is described as follows: “For the study, researchers questioned 60 men and 60 women ages 17 to 30, most of them Yale students...The people were divided into three groups. One group was questioned about times in their lives when they had bad hair. The second group was told to think about bad product packaging, like leaky containers, to get them in a negative mindset. The third group was not asked to think about anything negative. All three groups then underwent basic psychological tests of self-esteem and self-judgment. The people who pondered their bad hair days showed lower self-esteem than those who thought about something else.” Is thinking about bad hair in this artificial setting truly comparable to experiencing bad hair in real life? Interesting to note is the fact that the study was paid for by Procter & Gamble, which makes shampoo. We must also be aware that there are many “treatments”—such as income or smoking—which are im- practical or unethical for researchers to impose. Thus, observational studies are a very commonly used alternative for gathering statistical data. Example In 1939, an experiment was carried out on 22 orphans in Iowa: one group of 11 was given positive speech therapy, and the other 11 were induced to stutter by constant badgering on the part of their speech therapist. Out of those 11, 8 became chronic stutterers. One of them, interviewed over sixty years later, said, “It’s affected me right now. I don’t like to read out loud because I’m afraid of making a mistake. I don’t like talking to people because of saying the wrong word.” The therapist herself came to deeply regret her role in this experiment, which eventually led to a theory that helped thousands of children overcome stuttering. Most people today would agree that the price of this knowledge was simply too high for the unwitting victims. Thus, many problems must be researched via an observational study, rather than an experiment. There are many approaches that may be taken in such a study, and many pitfalls to be avoided. Parents putting more time into kids is the headline of an article published in 2001. “The University of Michigan research showed that children spent four to six more hours a week with their parents in 1997 than they did in 1981...The gains recorded were significant: In 1997 children ages 3 to 12 spent about 31 hours a week with their mothers, a gain of six hours over 1981, and 23 hours a week with their fathers, a gain...
View Full Document

This note was uploaded on 02/15/2012 for the course STAT 1000 taught by Professor Taeyoungpark during the Fall '06 term at Pittsburgh.

Page1 / 7

week3 - Lecture 6 Nancy Pfenning Stats 1000 Designing...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online