Chapter 7 – Variables and the Structure of Research (pp. 137-159)

Overall teaching objective: To introduce undergraduate criminal justice research methods students to the functions of variables, attributes and hypotheses in criminal justice research.

·  In social science, we use variables to describe the different characteristics of individuals, groups, organizations and social phenomena.

·  The manner in which we describe things can sometimes help us understand a problem or phenomenon more precisely.

·  A variable is any characteristic of an individual, group, organization or social phenomenon that changes.

·  A hypothesis is a statement that predicts how a change in one or more variables will cause a change in another variable.

Making Research Real 7.1 – Aggravated Assaults in the Midwest (p. 137)

·  The chief of a small town tries to understand why aggravated assaults are increasing among Hispanic residents

·  After visiting with a community leader (local priest) he learns of a recent increase in Hispanics from Central and South America.

·  Conflict between these groups might be the cause of the increase in aggravated assaults

·  Because the chief was able to more narrowly define the issue (using a more precise variable for Hispanic) he was able to address the issue more intelligently.

Types of Variables (p. 139)

·  Generally speaking, there are three types of variables. Each functions a bit differently within a causal relationship.

·  An independent variable is the causal variable, or the variable that a researcher predicts will be the cause of a change in another variable.

·  A dependent variable is the effect, or the variable that a researcher predicts will change as a result of a change in another variable or set of variables.

·  An intervening variable is any variable that occurs between the independent and dependent variables, changing how, or even if, the independent variable affects a dependent variable.

Figure 7.1 – A Causal Relationship Between an Independent and Dependent Variable (p. 140)

Figure 7.2 - A Causal Relationship Between an Independent, Dependent and Intervening Variable (p. 140)

Making Research Real 7.2 – An Intervention to Reduce Traffic Fatalities Caused by Intoxicated Drivers (p. 140)

·  A community attempts to reduce traffic crash fatalities with ordinances regulating the number of hours alcohol can be served.

·  They know that an increase in serving hours results in an increase of traffic crash fatalities.

·  However, their ordinance does not have its intended effect.

·  They learn that the availability of a taxi service intervenes in the relationship between serving hours and fatalities and, to the extent it is used, reduces fatalities.

Variable Attributes (p. 142)

·  Attributes are the different characteristics or values that a variable can take on.

·  Exhaustiveness refers to the completeness of the list of a variable’s attributes.

·  Mutual exclusivity refers to the capacity for a list of attributes to provide one, and only one, option for each respondent.

Making Research Real 7.3 – Learning About Religious Preference, With a Little Help from Our Friends (p. 144)

·  A researcher attempts to develop a set of attributes for the variable ‘religious preference’

·  The researcher learns that this is much more difficult than anticipated.

·  Eventually, the researcher learns that knowledge of the variable (in this case the differences between religious preferences) is essential to developing an exhaustive and mutually exclusive set of attributes.

Elements of a Good Research Question (p. 145)

·  Once you have identified the variables and their attributes in your research, you are poised to construct a good research question.

·  Research questions may come from mere curiosities about social phenomena or from a casual observation of some behavior. But in all cases, a research question is an interrogative statement. This simply means that research questions are actual questions, as opposed to statements.

·  Good research questions should be measurable, unanswered, feasible and disinteresting.

o  Research questions should be measurable. In other words, you should be able to measure the variables in the research question.

o  Research questions should also be unanswered. Most questions in the social sciences have been asked and answered by another researcher. This does not mean that we cannot ask them again or in different ways. Societies change constantly and sometimes dramatically. Thus, old research needs to be replicated, lingering questions from past research need to be explored and new techniques need to be applied.

o  Third, research questions should be doable. Money and time are always finite. So researchers need to ask whether a particular research project is practical or feasible.

o  Finally, research questions should be disinteresting. When we say that a research question is ‘disinteresting,’ we mean that it should be indifferent to the outcome. Researchers should never try to ‘prove’ anything. Researchers, like criminal investigators, should be prepared for any answer, even if the answer turns out to be different from what they suspected, or hoped, it would be.

Making Research Real 7.4 – Measuring Jesus (p. 146)

·  A student attempts to test a hypothesis that inmates who experience a religious conversion during incarceration are less likely to recidivate.

·  The student confronts a problem when he attempts to measure the extent to which an inmate experiences a religious conversion.

·  It is also clear that the student researcher is not disinterested in this question.

Hypotheses in Criminal Justice Research (p. 147)

·  A hypothesis is a predictive statement that alleges a plausible connection between two or more variables.

·  By ‘predictive’ I mean that the hypothesis makes a specific prediction about how two or more variables are connected.

·  By ‘plausible connection’ I mean that the hypothesis must describe the nature of the connection between the variables.

·  An alternative hypothesis is what the researcher hopes to prove as true during the research. But before a researcher can prove an alternative hypothesis, he or she must first disprove a competing hypothesis. We refer to these competing hypotheses as null hypotheses.

·  A null hypothesis is a statement that alleges no plausible connection between two or more variables.

Making Research Real 7.5 – Can’t We Just Agree on the Question?(p. 149)

·  A police chief and university professor are evaluating the effectiveness of a robbery suppression program.

·  Conflict exists between them because of the way they present the research question.

·  They are asking the same question, but the professor developed null and alternative hypotheses while the chief just asked whether the program worked or not.

·  Both are correct in their approaches but the professor’s approach provides more detailed insight into the problem.

Making Research Real 7.6 – Sherlock the Researcher

·  A woman reports a burglary.

·  The police respond and a detective makes an arrest.

·  The detective later learns that the person accused of burglarizing the residence had permission to use the ‘stolen’ property.

·  This story illustrates the similarities between the structure of research (null and alternative hypotheses) and the criminal investigation process.

·  Separate from the distinction between the null and alternative hypotheses, a hypothesis can also be categorized into one of two types.

·  The first type is association. Hypotheses of association allege a constant and predictable correlation, or relationship, between independent and dependent variables.

·  The second type is difference. Hypotheses of difference allege that the independent variable(s) makes groups different in some respect.

Figure 7.3 - Illustrating the Data in a Hypothesis of Association (p. 153)

Ha: More weekly hours of exercise among police officers is associated with a longer life expectancy.

Figure 7.4 - Illustrating the Data in a Hypothesis of Difference (p. 154)

Ha: Police officers who actively participate in the Department’s wellness program have a lower resting heart rate than police officers who do not.

Getting to the Point (Chapter Summary) (p. 155)

·  A variable is any characteristic of an individual, group, organization or social phenomenon that changes.

·  A hypothesis is a statement that predicts how a change in one or more variables will cause a change in another variable.

·  An independent variable is the causal variable, or the variable that a researcher predicts will be the cause of a change in another variable.

·  A dependent variable is the effect, or the variable that a researcher predicts will change as a result of a change in another variable or set of variables.

·  An intervening variable is any variable that occurs between the independent and dependent variables, changing how, or even if, the independent variable affects a dependent variable.

·  Attributes are the different characteristics or values that a variable can take on.

·  Exhaustiveness refers to the completeness of the list of a variable’s attributes.

·  Mutual exclusivity refers to the capacity for a list of attributes to provide one, and only one, option for each respondent.

·  Good research questions should be measurable, unanswered, feasible and disinteresting.

·  An alternative hypothesis is a predictive statement that alleges a plausible connection between two or more variables. A null hypothesis is a statement that alleges no plausible connection between two or more variables.

·  A hypothesis of association alleges that a change in the independent variable(s) is associated with a change in the dependent variable. As one increases/decreases, the other increases/decreases.

·  A hypotheses of difference alleges that the independent variable(s) makes groups different with respect to the dependent variable.

·  If the independent variable is ordinal, interval or ratio, the hypothesis will be one of association. Hence, the data used to test a hypothesis of association can be illustrated in a linear graph. If the independent variable is nominal, the hypothesis will be one of difference. Hence, the data used to test a hypothesis of difference can be illustrated in a bar graph.