Class Notes: Experimental Design and Manipulation

Chapter 9: Conducting Experiments

• Focus on the basic premise of conducting experiments, emphasizing the importance of appropriate sample size.

• The relationship between sample size and data representation discussed:
  
  

  • Larger sample sizes lead to better representation of the population.
    
    

  • Importance of statistically significant results highlighted.
    
    

  • Smaller groups may not yield realistic conclusions about broader populations.

• Types of errors in hypothesis testing:
  
  

  • Type I Error: False positive (finding a difference when there isn't one).
    
    

  • Type II Error: False negative (not finding a difference when there is one).

• Power analysis mentioned as a method to determine appropriate sample size.

• Demonstration of straightforward manipulation:
  
  

  • Explanation of independent and dependent variables illustrated through an example relating alcohol consumption and casual sex:
    
    

  • Ethical concerns with manipulation (not giving alcohol to participants).
    
    

  • Suggested observational study setup: monitoring alcohol consumption at a club.
    
    

• Types of Manipulations:
  
  

  • Straightforward manipulations involve direct questioning about behaviors (e.g., frequency of one-night stands).
    
    

  • Example of using visual stimuli in questions (e.g., ice cream preference).

• Importance of gender bias in self-reporting sexual behavior highlighted:
  
  

  • Women often underreport sexual experiences while men may overreport them.

• Discussion of confederates: individuals involved in the study who may influence outcomes.

Sample Size and Measurement

• Discussion on measuring dependent variables:
  
  

  • Different methods:
    
    

  • Self-report measures: Surveys assessing behaviors or states.
    
    

  • Behavioral measures: Counting occurrences of behavior (e.g., sleep duration on medication).
    
    

  • Physiological measures: Medical indicators, e.g., heart rate, blood pressure.

• Importance of avoiding ceiling effects (too high of expectations for success) and floor effects (too easy tasks):
  
  

  • Ceiling Effect: Too difficult for participants to meet expectations (e.g., measuring achievement against unrealistic stats).
    
    

  • Floor Effect: Too easy of a task results in everyone succeeding and no variances recorded.

• Control for participant behavior and researcher effects important:
  
  

  • Mention of demand characteristics affecting responses.

Placebo and Research Design

• Concept of placebo groups introduced: control for treatment effects.

• Placebo-controlled studies used in drug trials to compare groups receiving treatment versus those not.

• Single-blind and double-blind designs: discussion on eliminating bias:
  
  

  • The researcher or participants may not know about group assignments to mitigate expectation effects.

Chapter 10: Complex Studies

• Complexity of designing experiments with multiple variables discussed.

• Considerations for factorial designs: combination of drugs, levels, and their potential impacts on the outcome.

• Explanation of main effects (impact of individual independent variables) and interaction effects (how independent variables combine to affect outcomes).
  
  

  • Real-world analogy made using performance in sports (e.g., good batting averages in baseball hall of fame).

• Independent versus repeated measures designs discussed:
  
  

  • Independent Group Design: Different participants for different conditions.
    
    

  • Repeated Measures Design: Same participants across conditions.
    
    

  • Mixed Designs combining elements of both.

Chapter 11: Single Case Design

• Focus on individual case studies: investigating the impact of variables on single subjects.

• Single-case designs: - monitoring singular issues with baseline evaluation:
  
  

  • Example of ABA design introduced (baseline-intervention-return to baseline).
    
    

  • Usefulness in behavioral studies and testing interventions like relationship dynamics.

• Consistency and manipulation accuracy across subjects, behaviors, and situations emphasized.

Final Notes

• Overview of quasi-experimental designs when full control isn’t possible:
  
  

  • Discussed in the context of navigating real-world complexities like dietary restrictions before testing.

• Testing designs explained: one-group post-test, pre-test/post-test, non-equivalent control groups, and longitudinal versus cross-sectional studies.

General Class Information

• Reminder that the upcoming final will follow a similar format to previous tests but with potential variations in questions.