Psychology: Research Methods

Research Methods

For at least 10,000 years, humans have beenmanipulating their own brains by drinking alcohol. And for at least the last few decades, researchers have wondered whether alcohol had a positive effect on physical health. Study after study seemed to suggest that people who imbibed one alcoholic beverage per day—a 12-ounce beer, a 6-ounce glass of wine, or a 1.5-ounce shot of spirits—had healthier hearts than did people who abstained from drinking altogether. A drink a day, it seemed, kept the cardiologist away.

Yet these studies may be flawed. When Kaye Fillmore, a researcher at the University of California, San Francisco, and her team analyzed 54 published studies on how moderate drinking affects the heart, they found that most of the drink-a-day studies had not used random assignment. In studies with random assignment, researchers use coin tosses or the like to decide into which condition—the control group or various experimental groups—each study participant should go. By letting chance dictate who goes into which group, researchers are more likely to end up with truly comparable groups.

Instead of randomly assigning participants to drinking and nondrinking groups, though, 47 of the 54 studies compared people who were already having one drink daily to people who were already teetotaling. Why is this design a problem? Think about it: In the United States, where most of these studies took place, it’s fairly nor-mal to have a drink occasionally. Usually, people who never drink abstain for a rea-son, such as religious or moral beliefs or medical concerns.

In fact, Fillmore and her team found that many of the nondrinkers in these studies were abstaining from alcohol for medical reasons, including advanced age or a history of alcoholism. In other words, the nondrinking groups in most of the studies included more unhealthy people to begin with, compared to the drinking groups. As a result, these studies didn’t show that drinking alcohol led to better health. Instead, they showed that better health often leads to, or at least allows, a moderate level of alcohol consumption.

Fillmore and her colleagues also highlighted seven studies that avoided this methodological shortcoming by excluding participants with a history of certain med-ical problems. These studies found that the drinkers were no healthier than longtime nondrinkers—evidence against the drink-a-day hypothesis. But because this conclu-sion rests on very few studies, researchers can’t yet say whether a drink a day is good for your heart.

Why didn’t all 54 studies use random assignment to avoid these complications? The investigators were simply coping with the realities of human research. To conduct an experiment with random assignment, a researcher would have to convince some wine lovers to give up their daily Bordeaux and persuade some teetotalers to abandon their objections to alcohol. This job would be hard enough, but ethical concerns would still remain: What if alcohol really did damage the drinkers’ hearts or, alterna-tively, what if the nondrinkers suffered from not tossing back their daily dram?

In all of their research, psychologists have to address these issues of scientific rigor, cost, and ethics—and they do so with careful attention to research methods. In this, we’ll introduce the basics of collecting, analyzing, and interpreting the data of psychological science. We’ll start by describing how psychologists make observations and how they make sure their data are reliable and unbiased. We’ll then turn to how psy-chologists summarize what they’ve observed and use the data to draw conclusions. We’ll also look at the two major types of research—observational studies and experimental studies. And we’ll discuss how psychologists balance the demands of scientific research with ethical concerns, to protect the rights and dignity of research participants.

With our arsenal of sound research methods, you’ll be ready to test your own ideas about how, why, and what people sense, perceive, think, and act. And with some prac-tice, you too, can begin contributing to the field of psychological science.