PART I
SUMMARIZING DATA 1
DATA ORGANIZATION
1.1 INTRODUCTION
High school math teachers must cringe when they hear the age-old question āWhen am I ever going to need to know this?ā Social scientists learn the answer to this question during their first attempts at social research. Early stages of research, including developing a research hypothesis, performing a literature review, creating data-gathering instruments, and actually gathering data certainly challenge novice researchers like you. However, the greatest anxiety seems to surround the anticipation of data analysis.
Those who have become familiar with data analysis, though, would tell you to relax. The challenges posed by data analysis pale in comparison to those already encountered by one who has designed and implemented a means of gathering data. Statistical analysis follows a relatively structured plan that, once recognized, provides a basis for evaluating data in any form. In fact, at the point of statistical analysis, the topic of oneās study becomes somewhat irrelevant. The same protocols and techniques apply to all data, regardless of the issues to which the data pertain or the method used to collect them.
1.2 CONSIDERATION OF VARIABLES
You can refer to anything that changes as a variable. In the research context, a variable is an entity about which you gather data. These entities can change over time, for different people, in different situations, and for many other reasons. In your analysis, you attempt to determine whether these changes follow any particular pattern.
1.2.1 Units of Analysis
Before beginning the analysis process, you must acknowledge the origin points of your data, called the units of analysis. Each data point describes a particular unit of analysis. For social research, the units of analysis are most often human beings. Data indicating the responses to survey questions, behaviors observed during field studies, and performances on pretests and posttests of experiments all pertain to individuals. Social researchers refer to these individuals as subjects and to the compilation of their subjects as a sample. Proper ways to select your sample are discussed in Chapter 4.
Example 1.1: Human Units of Analysis
A researcher who wishes to determine whether a relationship exists between the placement of oneās tattoo on oneās body and the cost of the tattoo, for example, would gather information about individuals who have tattoos. By speaking with these individuals or by observing them while they receive and pay for the tattoos, the researcher would obtain the information that he or she needs. Each data point originates with one individual person and, after data collection the researcher can associate each person with a tattoo placement and cost. Thus, people serve as the unit of analysis.
Like many other aspects of the social sciences, however, the identification of analysis units does not always remain so straightforward. Rather than evaluating individuals, some social research compares and contrasts social institutions or settings. Data points in these situations do not correspond to people. The origin of the data and, thus, the units of analysis, reflect the nonhuman entity that the data describe.
Example 1.2: Nonhuman Units of Analysis
Slightly changing the focus of the study described in Example 1.1 to one that compares the prices of tattoo parlors in urban and in rural areas provides an example of nonhuman units of analysis. A researcher conducting this study would obtain prices from various randomly selected tattoo parlors and would characterize each as located in an urban or a rural area. In this case, the data pertain to locations of and prices at tattoo parlors, making these establishments the units of analysis.
1.2.2 Variables
Data analysis begins with the recognition of variables. In a general sense, the term variable describes anything that changes. This definition provides a foundation for understanding the concept of variables for social research. In this context variables are issues that the researcher measures. Each piece of data (datum) collected by a researcher provides information about a particular unit of analysis. The term variable applies because the information gathered generally addresses behaviors, attitudes, and characteristics that change from subject to subject.
Example 1.3: Variables
For example, a researcher pursuing the study proposed in Example 1.1 would, at the very least, need to note the part of the body on which each individual receives a tattoo as well as the cost for receiving the tattoo. The information recorded about placement of tattoos on the body and cost of tattoos changes with each individual who provides information. These two aspects, then, are variables.
Some studies use more than two variables. The complexity of your study and your intentions determine the number of variables that you need to consider. Some scenarios involving more than two variables receive attention in Section 1.4 and in Chapters 8 and 10 of this book. However, developing an understanding of these situations rests on your recognition and description of the two main variables.
Roles of Variables.
Even if you didnāt realize it, you were likely aware of your studyās independent and dependent variable(s) even before collecting data. However, you must formally address the distinction between the independent and dependent variables at the data analysis stage. When first introduced to the concept of research, you may have learned to regard the independent variable as the causal factor and the dependent variable as the effect of that causal factor. Although these associations may hold true for research in the natural sciences, social scientists should avoid causal terms when describing the roles of the independent and dependent variables. The section of this chapter entitled āVariable Relationshipsā further explains the importance of doing so.
You should think of the independent variables as a predictor of behaviors, attitudes, and characteristics. The independent variable describes a given condition, either already existing or created by the researcher before the start of data gathering. The dependent variable, then, refers to the behaviors, attitudes, and characteristics predicted by the independent variable.
With this understanding and previous identification of the two main variables for a study, you can simply insert variable names into the sentence, āData about ______ predict data about ______.ā Placing the variable names into the incorrect positions leads to an illogical statement. Once the sentence accurately portrays the researcherās goal in the study, you know that the variable inserted into the first blank is the independent variable and the variable inserted into the second blank is the dependent variable.
Example 1.4: Independent and Dependent Variables
This technique works well with the tattoo example. To determine whether tattoos cost more in urban or rural areas, the researcher wishes to investigate whether ādata about location predict data about costā and thus, will designate location as the inde...