Quasi-Experimental Research

Quasi-Experimental Research

  • Quasi-Experimental Design: Lacks random assignment; prevalent in applied settings with less control.

  • Obstacles to True Experiments: Access, permissions, and perceived unfairness of random assignments.

Types of Quasi-Experimental Designs

  • **Designs: **

    • Nonequivalent Designs

    • Archival Research

    • Time Series

Internal Validity Threats

  • History

  • Testing

  • Maturation

  • Instrumentation

  • Regression to the Mean

  • Experimental Mortality

  • Differential Selection of Participants

Differential Groups Design

  • Individual differences as independent variables (IV).

  • Confounds make causal inferences difficult.

Pre-Post Test Design

  • Structure: O1 X O2

  • Issues: History, Maturation, Testing, Instrumentation, Regression to the Mean, Demand Characteristics, Observer Bias.

Non-Equivalent Group Design

  • Structure: N X O1 / N O2.

  • Problems: Selection bias affecting comparisons.

Archival Research

  • Examines existing data sources; investigates effects over time, e.g., temperature's effect on aggression.

Time Series Design

  • Observations taken over time with some manipulation for analysis.

Program Evaluation

  • Four Types of Evaluations:

    1. Needs Analysis (Planning)

    2. Formative Evaluation (Monitoring)

    3. Summative Evaluation (Outcome)

    4. Cost-Effectiveness Analysis (Weighing Costs)

Ethical Considerations

  • Informed consent, confidentiality, perceived injustice, conflict with stakeholders.

Developmental Research Designs

  • Longitudinal Study: Repeated measures over lifespan.

  • Cross-Sectional Study: Compares different ages at one time.

  • Cross-Sectional Longitudinal Design: Combines both methods for mixed age comparisons and retesting.