Chapter 6 - class notes PDF

Preview

Summary

class notes...

Description

1. Causal explanation a. A cause is an explanation for some characteristic, attitude, or behavior of groups, individuals, or other entities (such as families, gangs, police departments) or for events 2. Research process a. Research question b. Conceptualization c. Operationalization i. Measurement validity d. Sampling i. Generalizability e. Research design i. Casual validity 3. Criteria for causation a. Empirical association (required) b. Time order (required) i. Temporal order c. Nonspuriousness (required) d. Mechanism e. Context 4. Empirical association a. Variation in one variable is related to variation in another i. Correlation ii. Cannot be a constant b. Relationship must be observable -- cannot be only assumed or believed c. The independent variable and the dependent variable must vary together 5. Time order a. The change in the independent variable must occur before the change in the dependent variable b. It is often too difficult to establish cause-effect relationships in social research, because it can be difficult to determine which came first i. Juvenile delinquency and the number of delinquent peers 6. Nonspuriousness a. A relationship between two variables that is due to variation in a third variable b. When this third variable (an extraneous variable) causes the variation, it creates a spurious relationship between the independent and dependent variable c. Random assignment can reduce threats of spuriousness d. Remember the old adage: “correlation does not prove causation” 7. Mechanism

a. Process that creates the connection variation in an independent variable and the variation in the dependent variable it is hypothesized to cause 8. Context a. The set of circumstances surrounding an event or situation b. No causal relationship can be separated from some larger context in which it occurs c. A cause is one of a set of interrelated factors required for the effect 9. Research designs and causality a. Experimental designs i. True experiments b. Nonexperimental designs i. Cross-sectional ii. Longitudinal 1. Repeated cross-sectional designs 2. Fixed-sample panel designs 3. Event-based designs 10. True experiments a. Three requirements i. An experimental group (aka. Treatment group) that is receiving the experimental condition (i.e., treatment or intervention) and the control group that is not receiving the treatment or receiving another form of it ii. Random group assignment iii. Assessment of change in the dependent variable for both groups after the experimental condition has been received (posttest) b. A prerequisite for meeting each of the three criteria is maintaining control over the conditions subjects are exposed to after group assignment 11. Causality and true experimental designs a. Association i. Having a treatment and control group ii. Comparing the dependent variable b. Time order i. Can be established by pretest and posttests c. Non Spurious relationships i. Random assignment eliminates many extraneous influences that can create spurious relationships d. Mechanism i. Experimental designs cannot directly address this factor e. Context in which change takes place i. Difficult to control context in field experiments 12. Nonexperimental designs:

a.

Cross-sectional i. Observations are made at one time point, a snapshot ii. Three special circumstances that can increase confidence in drawing conclusions about time order: 1. The independent variable is fixed at some point prior to the variation in the dependent variable a. Demographics 2. We believe that respondents can give us reliable reports of what happened to them or what they thought at some earlier point in time 3. We know that the value of the dependent variable was similar for all cases prior to the treatment b. Longitudinal i. Repeated cross-sectional designs (trend) a. Data are collected at two or more points in time from different samples of the same population 2. Appropriate design when the goal is to determines whether a population has changed over time 3. Example: national polls that are conducted annually to track sentiment on social issues ii. Fixed-sample panel designs 1. Data are collected from the same individuals-the panel-at two or more points in time i. Example Sampson and Laub (1990) studied a sample of white measles in Boston when the subjects were between 10 and 17 years old and then followed up when the subjects were in their adult years b. Two issues with this design: i. Cost ii. Subject attrition 1. Subjects drop out of study or provide invalid information (subject fatigue) iii. Event - or - Cohort - Based designs 1. Data are collected at two or more points in time from individuals in a cohort 2. Types of cohorts: a. Birth b. Seniority c. School

3. Can be a type of repeated cross-sectional design or a type of panel design a. Or a fixed sample 13. Causality in nonexperimental designs a. Association can be established but other criteria of causality are more difficult: i. Time order 1. Can be difficult to establish that change in the independent variable preceded change in the dependent variable ii. Nonspurious relationships 1. Can use statistical controls, meaning that one variable is held constant so the relationship between two or more other variables can be examined apart from influence of “control” variable, to reduce threat of spuriousness 2. Does not eliminate it iii. Mechanism 1. The use of intervening variables can help determine how variation in the independent variable affects variation in the dependent variable iv. Context 1. Context can be developed, especially when surveys are administered in different settings and with different individuals 2. Replication...