Nonexperimental Research Designs: Case Studies & Single Variable Research Chapter 6 James A. Van Slyke Importance of Case Studies The Curious Case of Phineas Gage Before accident Intelligent capable worker excellent manager responsible family man After Accident Maintained his general intelligence Unreliable and volatile Socially Inappropriate Lost his family and fortune R econ stru ction of th e S k u ll of P h in eas G age F rom H ana D am asio et al., “T he R eturn of P hineas G age: C lues about the brain from the skull of a fam ous patient” Science. 1994, 264, 1 1 0 2-1105. Galvanic Skin Response Iowa Gambling Task Nonscientific Case Study Method Case study: An intensive description and analysis of a single individual. Data: clinical observations, self-report, archival data (e.g., medical records) Case studies typically report the results of a treatment. Major problem: Lack scientific control simultaneous treatments, extraneous variables Case Study Method (continued) Advantages of the Case Study Method: rich source of ideas for developing hypotheses, opportunity for clinical innovation, method for studying rare events, possible challenge to theoretical assumptions, tentative support for a psychological theory, and complement to the nomothetic study of behavior Case Study Method (continued) Disadvantages of the Case Study Method: difficulty drawing cause-and-effect conclusions (limited internal validity), possible biases when interpreting outcomes due to observer bias and biases in data collection (e.g., due to poor memory), and problem of generalizing findings from a single individual (limited external validity) Case Study Method (continued) Case studies provide great anecdotal evidence and “testimonials.” Case studies that appear in the popular press are rarely scientific. People want to believe that the treatment in these testimonials will work for them, but often they do not. It’s better to pay attention to the results of single-subject experimental designs. Scientific Case Studies Single-subject experimental designs have their roots in B. F. Skinner’s approach called applied behavioral analysis. Single-subject designs improve on nonscientific case studies, because the researcher attempts to gain more scientific control. Characteristics of Single-Subject Experiments Critical feature of single-subject designs: An independent variable is examined: treatment no treatment control (baseline stage) Researchers compare treatment conditions for one individual whose behavior is continuously monitored (repeated measures). Single-Subject Experimental Designs (continued) The baseline condition is used to describe behavior before treatment is provided, and predict what behavior will be like in the future without treatment. Single-Subject Experimental Designs (continued) Compared to baseline, the behavior decreases after treatment is implemented. Although this pattern of data suggests the treatment was effective, some other factor that occurred at the same time as the treatment could have caused the frequency of behavior to decrease. The ABAB Design: Baseline and Treatment conditions are contrasted in the ABAB Design Illustration of a Treatment Effect: The frequency of the behavior decreases during treatment (B), reverses when treatment is withdrawn (second A), and reverses again during treatment (second B). This design is also called a reversal design. ABAB Design (continued) Illustration of no reversal: When the frequency of behavior does not reverse when treatment is withdrawn (second A), it is very difficult to determine whether the treatment was effective. ABAB Design (continued) Methodological Issues Associated with ABAB Designs If behavior does not reverse back to baseline levels after treatment is withdrawn, researchers cannot conclude that treatment caused the initial behavior change. A variable other than treatment may have caused the behavior to change. ABAB Design (continued) Methodological Issues Associated with ABAB Designs Treatment may have promoted change, and then other variables (e.g., positive attention) may persist to maintain behavior change. Some behaviors may not be logically expected to change once improved (e.g., when new skills are learned). ABAB Design (continued) Ethical considerations Is it ethical to remove a treatment that appears to be beneficial (i.e., implement the second “A” baseline stage)? Dilemma between goal of understanding and goal of creating change. Problems with All Single-Subject Designs Baseline Records If baselines demonstrate unstable, increasing, or decreasing trends in behavior, the effects of treatment are hard to interpret. Problems with All Single-Subject Designs Baseline Records When baseline behavior shows extreme variability, it’s difficult to detect a clear discontinuity in behavior when treatment is implemented. Solutions: Look for factors that may contribute to variability, wait for baseline behavior to stabilize, average baseline data points across observations. Problems with Single-Subject Designs (continued) Baseline Records, continued Whether increasing or decreasing baseline trends are a problem depends on the desired direction of behavior change Problems with Single-Subject Designs (continued) Baseline Records, continued Suppose the goal is to increase the frequency of a behavior. If the baseline shows an increasing frequency of behavior, determining whether behavior increases following treatment will be difficult. However, if the baseline shows a decreasing trend and treatment reverses this trend, we can be confident about the effect of the treatment. Problems with Single-Subject Designs (continued) External Validity Single-subject designs are frequently criticized for their limited external validity. Will treatment effects observed for one individual generalize to other individuals? Problems with Single-Subject Designs (continued) External Validity Reasons why external validity may not be limited: Treatments are usually powerful. Multiple-baselines designs can be used to demonstrate generality of effects. Group treatment can be used to demonstrate effectiveness of treatment.