Through his elaboration on the sociological imagination, C.
The Sociological Imagination C.
Students will learn that sociologists collect their data through a number of research methods. One of the most common is the social survey, in which a sample of people respond to a questionnaire that is administered on paper, in a personal interview, by telephone, or over the internet. Sociologists may also engage in a participant observation, in which they become part of a group they seek to understand. Some sociologists, like psychologists, conduct experiments, while others rely principally on historical or archival data to test their hypotheses. The choice of data collection methods depends upon the kind of data that are needed to test a hypothesis. Some hypotheses may be tested through multiple methods. Students will learn how sociologists tabulate their data using statistical methods, some of which are highly sophisticated. It is common to report measures of central tendency for each variable, for example, the mean or median values. It is also common to report a measure of the spread from the mean, such as the standard deviation or interquartile range. Measures of association indicate whether and/or how closely two variables are related to each other. A measure of association such as chi-square can show if the relationship of two variables might have happened by chance or if it is a significant relationship; it is also possible to calculate the strength of an association through the use of a correlation coefficient. When sociologists measure one variable taking into account the effect of a second variable, they are said to “control” for the second variable, and multivariate statistics are sophisticated means to control simultaneously for the effects of many independent variables.
I. The Sociological Perspective
In developing explanations, students will learn how sociologists are careful to distinguish the types of variables they are investigating. In general, a dependent variable is the variable being studied. An independent variable is a variable believed to vary with the dependent variable; the independent variable is often relatively fixed (such as one’s gender), or it occurred earlier than the dependent variable (such as childhood experience). In the example in the preceding paragraph, the dependent variable is juvenile delinquency rate, and the independent variables are family income and the occupational prestige of workers in the family. Notice, however, that high parental income may be associated with a low juvenile delinquency rate, but it does not necessarily cause a low juvenile delinquency rate. Instead, the relationship may be mediated in various ways. For example, wealthier parents may be able to provide more activities for their teens, or they may be able to hire better lawyers if their teens do get into trouble. Drawing on theoretical foundations, students will learn that to assess a causal relationship between variables, it is necessary 1) to establish the time order of the variables (with the independent variable coming before the dependent variable), 2) to establish that the variables are correlated, and 3) to rule out any competing hypotheses. Suppose, for example, that a researcher finds that ice cream consumption is inversely related to juvenile delinquency rates. This finding does not prove that ice cream prevents juvenile delinquency. Instead, this hypothesis may be misspecified because the wrong independent variable has been named. Perhaps parents with higher income can buy more ice cream, so that higher income co-varies positively with more ice cream and also co-varies negatively with the juvenile delinquency rate.