CHAPTER
CHAPTER
How Do We Know It's True? Methods of Research
"I'm not a statistic!" David insists. "You can't reduce people's lives to a table, chart, or graph!" Lisa insists. These assertions, which I've heard from students in my courses on drug use, are only half-right. David-and you-may not be a statistic, but both of you are included in a substantial number of statistics. For instance, during every year ending in a zero, the U.S. government conducts a nose count of the entire population of the United States. In the twenty ten census, David was counted, along with three hundred ten million or so other people residing in the United States; he added one person to the total population. Does that mean David was a population "statistic"? Of course it does! Does that mean all he is, is a statistic? Of course not; he is other things as well. And to the extent that, as a number, Lisa's characteristics or answers in a survey can be put in a table, chart, or graph, she too is part of one or more "statistics." Lisa happens to be an interesting, complex human being, but she also happens to contribute to the statistics demographers use to numerically depict the population.
"We're Losing the Drug War," screams one headline. "Increases in Students' Drug Use," announces another. "Roofies-The Date Rape Drug," intones a TV report. "Meth-the New Killer Drug," claims a magazine cover story. "Tidal Wave of Drugs Roll into the Suburbs," alleges the nightly news. "Drug Arrests at Record High," asserts our daily newspaper. How do we know when an assertion we read or hear is true? Specifically, how do researchers gather information about drug use? And how do they draw conclusions from the evidence they've gathered?
Researchers make use of a wide range of information to determine what the drug picture looks like. Indeed, there is something of an "embarrassment of riches" here, since there are so many data sources. But not all of them are equally valid, and all are flawed in one way or another. However, when we put several data sources together, we get a clearer sense of what that picture looks like. Researchers use the term triangulation to refer to using two or more sources of information to focus on a single phenomenon. If these data sources agree with one another, researchers call this multiple confirmation. And when two or more independent pieces of evidence say the same thing, our confidence that what the evidence says is true increases.
In Chapter two, we looked at some of the aspects of drugs and drug use in which pharmacologists and other medical and natural scientists are interested. Here, we focus on how social scientists study drug use. Sociologists and criminologists are interested, among other things, in rates of drug use, or how widespread it is. There are many different sources of information about how much drug use there is in the population. The systematic study of drug use entails making use of self-report surveys, as well as drug tests, arrest data, and hospital and medical examiner reports. In addition, the sale of legal psychoactive drugs (alcohol, tobacco, and prescription drugs) is recorded and is therefore publicly available for study. Illegal drug use poses special problems for the social researcher since it is, by its very nature, clandestine-hidden from public view. Hence, we must rely on a variety of indirect sources of information, including surveys. But before we make use of these sources, we need to know a bit about some basic principles of social research.
SOCIAL RESEARCH ON DRUG USE: AN INTRODUCTION
SOCIAL RESEARCH ON DRUG USE: AN INTRODUCTION
To repeat the title of this chapter: How do we know it's true? How can we feel confident that the conclusions we read in a study on drug use are reasonably valid, reliable, and accurate? Here are a few things we need to know about three matters in social research: lying, sampling, and statistics.