Chapter 7 - Criminal Justice and Criminology Research Methods, 2nd Edition

Chapter 7 Experimental and Quasi-Experimental Crime and Justice Research 1 THE EXPERIMENT AS THE “GOLD-STANDARD” Experimental Group - the group that receives the treatment in experimental research. Control Group - the group that does not get the treatment in experimental research. Experimental research embodies the principles of a PSS approach more directly that any other research technique. It attempts to test the causal relationships between variables as stated in the hypotheses. 2 Causal Inference Cause must precede effect. There must be an association between cause and effect. The researcher must eliminate plausible alternatives. 3 Random Assignment – dividing subjects into groups at the beginning of experimental research using a random process, so the experimenter can treat the groups as equivalent. It is a way to increase one’s confidence that: The groups do not differ in a systematic way, Unknown, extraneous variables are controlled for, Statistical techniques and reasoning that relies on randomization are used. Importance of Random Assignment Parts of the Experiment Treatment or Independent Variable Dependent Variable Pretest Pretest - measurement of the dependent variable of an experiment prior to the treatment. Posttest Posttest - measurement of the dependent variable of an experiment after the treatment. Experimental Group Control Group Random Assignment 5 Steps in Conducting an Experiment Begin with a straightforward hypothesis that is appropriate for experimental research. Decide on an experimental design that will test the hypothesis within practical limitations. Decide how to introduce the treatment or create a situation that induces the independent variable. Develop a valid and reliable measure of the dependent variable. Set up an experimental setting and conduct a pilot test of the treatment and dependent variable measures. 6 Steps in Conducting an Experiment Locate appropriate subjects or cases. Randomly assign subjects. Gather data for the pretest measure. Introduce the treatment to the experimental group. Gather data for posttest measure. Debrief the subjects. Examine data collected and make comparisons between different groups. 7 Classical Experimental Design Classical Experimental Design An experimental design that has random assignment, a control group, an experimental group, and a pretest and posttest for each group. 8 Other Experimental Designs Two-Group Posttest Only Design - an experimental design that has all the parts of the classical design except a pretest. Latin Square Design - an experimental design used to examine whether the order or sequence in which subjects receive multiple versions of the treatment has an effect. Solomon Four-Group Design - an experimental design in which subjects are randomly assigned to two control groups and two experimental groups. Factorial Designs - a type of experimental design that considers the impact of several independent variables simultaneously. 9 Preexperimental Designs – experimental designs that lack random assignment or use shortcuts and are much weaker than the classical experimental design. One-Short Case Study Design – an experimental design with only an experimental group and a posttest, no prettest. One-Group Pretest-Posttest Design – a preexperimental design that has one group, a pretest, treatment, and a posttest. Preexperimental or Weak Experimental Designs Quasi-Experimental Designs Quasi-Experimental Designs - experimental designs that are stronger that preexperimental designs. Interrupted Time Series - an experimental design in which the dependent variable is measured periodically across many time points, and the treatment occurs in the midst of such measures, often only once. Equivalent Time Series Design - an experimental design in which there are several repeated pretests, posttests, and treatments for one group, often over a period of time. 11 Threats to Internal Validity Internal Validity - the ability of experimenters to strengthen the logical rigor of a causal explanation by eliminating potential alternative explanations. Threats include: Selection Bias History Effects Maturation Effects Testing Effect Instrumentation Experimental Morality Statistical Regression Diffusion of Treatment Compensatory Behavior Experimenter Expectancy 12 External Validity External Validity - the ability to generalize findings beyond a specific study. Population Validity Reactivity Reactivity Hawthorne Effect Demand Characteristics Placebo Effect 13 Practical Considerations Planning and Pilot Tests Instructions to Subjects Upholding Treatment Integrity Postexperiment Interview 14 A Word on Meta-Analysis Meta-Analysis - a study where all previous studies in a particular area are combined in an attempt to determine whether the causal factor studied has had an overall impact on the dependent variable. The objectives of meta-analysis are: Uncover robust empirical relationships. Identify weaknesses in the research. Provide summary statistics about overall effectiveness. To overcome the inherent external validity problems found in most experimental studies. 15