Study Notes on Experimental Design in Psychology

Design of Experiments

Overview of Chapters 5-8

  • Chapters focus on the design of experiments in psychology.

Chapter 5: Essential Features of Experimental Research

Key Components of Experiments
  • Independent Variables: Factors varied in the experiment.
  • Dependent Variables: Outcomes measured in the experiment.
  • Extraneous Variables: Other factors controlled to isolate the effect of the independent variables.
Objectives of Chapter 5
  • Understand essential features of experiments.
  • Assess how design impacts validity of studies as outlined below:
    • Impact of Robert Woodworth's Experimental Psychology (1938).
    • Definition and examples of manipulated independent variables.
    • Distinction between experimental and control groups.
    • Application of John Stuart Mill's rules of inductive logic.
    • Recognition and implications of confounding variables.
    • Identification of independent and dependent variables in various experiments.
    • Differentiation between manipulated and subject variables.
    • Factors affecting statistical conclusion validity.
    • Construct validity in experiment design.
    • Importance of internal and external validity.
    • Threats to internal validity and their implications.
    • Ethical guidelines in running subject pools.
Historical Context: Woodworth’s Contribution
  • Robert Sessions Woodworth's Experimental Psychology published in 1938 influenced the definition of experiments:
    • Defined experiments with rigorous standards contrasting them from correlational research.
    • Introduced terminology like "independent variable" and "dependent variable."
  • Origin of methods established by John Stuart Mill which aid in establishing causality in experiments.
John Stuart Mill's Inductive Logic
  • Method of Agreement: Identifies conditions where a factor (X) consistently leads to an outcome (Y).
  • Method of Difference: If X does not occur, the outcome (Y) is absent.
  • Joint Method: Combining both methods enhances confidence that a relationship exists.
Establishing Independent Variables
  • Defined as the manipulated factor in an experiment.
  • Must have at least two levels for comparison.
  • Example: Study on caffeine effects requiring two dosage levels.
Categories of Manipulated Independent Variables
  1. Situational Variables: Environmental features affecting behavior.
    • Example: Number of bystanders in a helping behavior study.
  2. Task Variables: Variations in the tasks assigned to participants.
  3. Instructional Variables: Different instructions given to participants.
Control Groups
  • Experimental Group: Receives treatment or manipulation.
  • Control Group: Does not receive treatment, serving as a baseline.
  • Importance of ensuring both groups are as identical as possible to isolate effects of treatment.
Example Study: Superstition and Luck
  • Experiments by Damisch, Stoberock, and Mussweiler (2010) demonstrated how beliefs in luck affect performance on tasks.

Controlling Extraneous Variables

  • Confounding: Occurs when extraneous variables co-vary with independent variables, affecting interpretation of results.
  • Need to hold extraneous factors constant to avoid confounds.
  • Real-life examples of confounding are elaborated using hypothetical studies.

Measuring Dependent Variables
  • Dependent variables are the behavior outcomes measured in the study.
  • Decisions regarding how dependent variables are defined affect both construct and statistical validity.
  • Operational definitions must be precise for reliability.

Subject Variables

  • Variables not manipulated by the researcher, existing prior to the study.
  • Common in studies on group differences, e.g., gender, age.
  • Conclusions about causality can only be assumed for manipulated variables.
  • Example: Study comparing anxiety levels in different natural groups.

Validity of Experimental Research

  • Types of validity:
    • Statistical Conclusion Validity: Accuracy in statistical analyses and conclusions drawn.
    • Construct Validity: Appropriateness and accuracy of operational definitions used.
    • External Validity: Generalization of results to other populations, environments, times.
    • Internal Validity: Methodological soundness and control from confounding variables.
Threats to Internal Validity
  • Addressed through several potential pitfalls:
    • History
    • Maturation
    • Testing
    • Instrumentation
    • Selection effects
Conclusion on Internal and External Validity
  • Internal validity relates to methodological rigor, while external validity concerns generalizability.
  • Importance of maintaining a balance between internal and external validity in research design.

Ethical Considerations in Research

  • Essential to meet ethical standards when running experiments, especially concerning subject pools.
  • Ensure participant rights and well-being are prioritized.