Week 7: Experimental Designs

• Confounding Variable


  Variables that are intertwined with another variable so that you cannot determine which of the variables is operating in a given situation.
• What is the relevance of a confounding variable?

  We cannot determine whether the treatment condition or the confounding variable is responsible for the effect; this opens the study up to alternative explanations.
• Internal Validity

  A study that effectively meets all three criteria for cause and effect is said to have high internal validity.

  1. Establishes temporal precedence.
  2. Establishes covariation.
  3. Eliminates most alternative explanations.
• What are the three steps to a basic experiment?

  1. Obtain two equivalent groups.
  2. Introduce the independent variable.
  3. Measure the effect of the independent variable on the dependent variable.
• What are three basic ways of obtaining two equivalent groups?

  1. Independent Groups Design (Within-Subjects Design)
  2. Repeated Measures Design (Between-Subjects Design)
  3. Matched Pairs Design
• Independent Groups Design (Between-Subjects Design)

  Randomly assign only to one level of the independent variable.
• Repeated Measures Design (Within-Subjects Design)

  Assign to all levels of the independent variable.
• Matched Pairs Design


  First, matching two people on a characteristic and then randomly assign one of each pair to only one level of the independent variable.
• What are the two forms that the independent group design takes on?

  1. Posttest-Only Design: Participants are randomly assigned to independent variable groups and are tested on the dependent variable once.
  2. Pretest-Posttest Design: Participants are randomly assigned to independent variable groups and are tested on the dependent variable twice – once before and after exposure to the independent variable.
• What are the advantages of the pretest-posttest design?

  • Sample size is small; allows you to assess the equivalency of groups – you know the starting point.
  • When you need specific participants for the experiment – testing the participants before the start of the study.
  • When the likelihood of dropout rate is high.
• What are the disadvantages of the pretest-posttest design?

  • Can be time-consuming and awkward to administer.
  • Can sensitise the participants to what is being studied.
  • Can reduce the extent to which results generalise to people who did not receive a pretest.
• Solutions to the disadvantagesof pretest-posttest design.

  Using deception and embedded the pretest with other irrelevant measures.
• Solomon Four-Group Design


  Used to assess impact of the pretest by treating the pretest as a second independent variable. Half the participants receive the pre-test and the post-test; half receive only the pretest.