RM

• Between-subjects experimental design


  Determines whether differences exist between two+ treatment conditions
• Advantages of between-subjects design

  • Each score is independent from others and not influenced by contrast effects, fatigue, or experience
• Disadvantages of between-subjects design

  • Requires a large number of participants + individual differences
• Sources of confounds in between-subjects designs

  • Individual differences
  • Environmental differences
• Small vs. large variance within a sample

  Small = clustered up while large looks more evenly spread out
• Ways to minimize variability within treatments

  1. Standardizing procedures

  2. Limiting individual differences

  3. Random assignment

  4. Increasing sample size
• Differential attrition


  Participant dropout rate (Large differences between groups creates problems)
• Participant communication threats to between-subjects designs

  • Diffusion
  • Compensatory equalization/rivalry
  • Resentful demoralization
• Statistical analysis for two-group between-subjects design

  1. Independent samples T test

  2. Null hypothesis = nothing changes

  3. Alternative hypothesis = difference between groups

  4. Significant p-value = reject null and accept alternative
• Within-subjects experimental design


  Determines whether differences exist between two or more treatment conditions but all participants are exposed to all treatments or conditions
• Advantages of within-subjects design

  • Each person is their own control, few participants, reduces variance
• Disadvantages of within-subjects design

  • Scores are not independent from others + Participant attrition
• Potential sources of confounds in within-subjects designs

  • Environmental
  • Time Related variables
• Time-related threats in within-subjects designs

  • History
  • Maturation
  • Instrumentation
  • Regression towards the mean
  • Order effect
• Dealing with time-related threats and order effects


  Increasing the time between treatments
• Counterbalancing


  Changing the order in which treatment conditions are applied from one participant to another - Uses all possible order of treatments with an equal number of subjects participating in each sequence
• Factors differentiating between-subjects and within-subjects designs

  • Individual differences
  • Time related factors/order effects
  • Fewer Participants
• Statistical analysis for pretest-posttest design

  Paired sample t test
• Statistical analysis for repeated measures design

  Repeated measures ANOVA
• Factorial research design


  Research study that allows one to look at the impact of two or more variables acting together
• Factors


  An Independent variable in an experiment
• Advantages of factorial design

  • They create more realistic situations and allows researchers to see how each individual factor influences behavior
• Number of conditions in a 2 x 2 factorial design
• Main effect


  The mean differences among the level of one factor
• Interaction


  The production of a result different from that produced by either variable alone
• Differences between between-subjects, within-subjects, and mixed factorial designs

  • The number of treatments
  • The research designs used
• Statistical analysis for factorial design
• Correlation


  Attempts to establish a relationship between variables
• When to use correlational research

  To describe relationships
• Interpreting correlation coefficient


  Sign and magnitude
• Difference between linear and monotonic relationships
• Statistics for linear vs monotonic relationships
• Coefficient of determination (r2)

  The squared value of a correlation which measures variability
• Interpretation of r2 values

  • Small = .01, Medium = .09, Large = .25
• Null and alternative hypotheses for correlation

  Null = No relationship, Alternative = Relationship
• Statistically significant correlation


  There is a relationship between the two variables
• Predictor and criterion variables in regression

  Predictor = Variable used to predict, Criterion = Variable being predicted
• Strengths and weaknesses of correlational research

  • Strengths: Describes relationships between variables, nonintrusive, high external validity

  Weaknesses: Cannot assess causality, third variable problem, directionality problem