140 likes | 254 Vues
This overview explores the key differences between laboratory and natural experiments in psychological research, focusing on the concepts of internal and external validity. It discusses the control of variables, practical limitations, and the goals of both basic and applied research. Key threats to internal validity, such as history, maturation, and selection biases, are examined alongside the challenges of generalizability in external validity. The text also highlights the importance of various research designs, including single-subject and psychophysical approaches, emphasizing correlation versus causation.
E N D
PSYC 235: Research Methods ALT DESIGN
Differences between lab and natural experiments • Control • Internal validity • External validity • Basic and applied goals • Consequences • Practical limitations • GOALS – to ENHANCE internal and external validity
Control – threats to internal validity • History (events other than tx) • Maturation • Repeated testing • Instrumentation changes • Regression to the mean • Selection and interaction with selection • Subject mortality and attrition
External Validity - generalizability • Subjects • Situations • Task • Time
Subject variables: e.g. developmental research • Correlation does not equal causation!!!! • Cross sectional • Longitudinal • Time lag • Cross sequential • Major confounds of each
Combining Quasi and True IV • E.g. age + ritalin, diagnosis + task difficulty, gender + video content • Importance of the interaction
Natural Treatments – ex post facto • Smoking in the psychiatric ward • Improving the basic design • Non-equivalent conrtol group
Interrupted time series • Will lowering the BAC from .1 to .08 reduce drunk-driving fatalities? • Do people drink more in college than at other times )or other places)? • Do women who wear more makeup make more money than women who don’t? • Would eliminating PSYC 235 from the curriculum increase the number of psychology majors?
Small n research • Nomothetic versus Idiographic • Advantages of each • Disadvantages of each
Small n: Applied Behavior Analysis • Skinnerian versus statistical traditions • Designs • AB • ABA • ABAB • Alternating treatments (e.g. ABCA) • Example: Behavioral Contrast • Multiple baseline designs
Psychophysical designs • Signal detection theory • Discriminability (d’) • Bias (B)