Week 8 Power Point (Research Methods)

• Cause-and-effect relationships


  The goal of the experimental research strategy is to establish the existence of a cause-and-effect relationship between two variables
• Internal validity


  If a study produces a single, unambiguous explanation for the relationship of the two variables. Any alternative explanation is a threat to internal validity.
• External validity


  The extent to which the results of a study can be generalized to and across other situations, people, stimuli, and times. Any non-random assignment of participants or conditions is a threat.
• Statistical significance


  Before you can interpret the differences between treatments as a cause-and-effect relationship, you must conduct a hypothesis test and demonstrate that the difference is statistically significant. A significant result means the difference is large enough and consistent enough for a hypothesis test to rule out chance as a plausible explanation, therefore the difference must have been caused by the treatments.
• Independent variable (IV)


  The variable that is manipulated by the researcher. The different levels of the independent variable are called conditions.
• Dependent variable (DV)


  The variable that is observed for changes to assess the effects of manipulating the IV.
• Jane Elliot's Experiment


  • Jane divided her class into two groups: blue-eyed and brown-eyed children. She gave the blue-eyed children extra privileges and emphasized how superior they were to the brown-eyed, who were now a "minority group." As a result, the brown-eyed children saw a drop in confidence, academic performance and an increase in bullying. However, when she later labelled the blue-eyed group as the inferior, these effects were reversed.
• Extraneous variables


  All other variables in the study other than the IVs and DVs, i.e. the noise, confounding variables.
• Third-variable problem


  A researcher has observed that children who eat more sugar tend to show a higher level of activity than children who eat less sugar. However, the researcher suspects that the apparent relationship may be explained by the fact that some children have a higher rate of metabolism, which causes them to eat more and to be more active compared to children with a lower rate of metabolism who eat less and are less active.
• Directionality problem


  A research study may establish a relationship between two variables, but the existence of a relationship does not always explain the direction of the relationship. The remaining problem is to determine which variable is the cause and which is the effect.
• Basic goals in demonstrating a cause-and-effect relationship


  • Demonstrate that the "cause" happens before the "effect" occurs
  • Value of the IV is followed by a change in the DV
  • Establish that one specific variable is responsible for changes in another variable
  • An experiment must rule out the possibility that changes are caused by an extraneous variable
• Manipulation


  Allows researchers to determine the direction of the relationship by manipulating one variable (IV) and observing the second variable (DV) to see if it is affected. Researcher determines specific values of the IV to examine. Helps researchers control the influence of outside variables (confounding variable).
• Three primary techniques for controlling extraneous variables

  • Holding variables constant
  • Matching values across treatment conditions
  • Control by randomization
• Goal of an experiment

  Show that differences in the dependent variable are caused by the independent variable. The purpose of control is to ensure that no variable could be responsible for causing the scores to differ.
• Holding constant and matching


  Require extra effort or extra measurement
• Randomization


  Controls a wide variety of variables simultaneously, but is not guaranteed to be successful in balancing variables
• Experimental group


  The treatment condition in an experiment. Observations of the DV are compared across different levels of the IV.
• Control group


  The no-treatment condition in an experiment.
• No-treatment control group


  A condition in an experiment in which the participants do not receive the treatment being evaluated. A placebo control group is an ineffective treatment that produces an effect, and the "placebo effect" occurs simply because the individual thinks the medication is effective.
• Simulation


  The creation of conditions within an experiment to simulate or closely duplicate the natural environment in which the behaviours being examined would normally occur.
• Field study


  Research conducted in a place that the participant or subject perceives as a natural environment (psychological realism).
• Between-Subjects Designs


  • Experimental research wherein a researcher manipulates an IV and then measures the DV for each participant. Goal is to determine whether differences exist between two or more treatment conditions. Requirement is only one score for each participant is obtained, with separate groups of participants used for different treatment conditions.
• Between-Subjects Designs


  • A researcher with a sample of 100 university students might assign half of them to write about a traumatic event and the other half write about a neutral event.
  • A researcher with a sample of 60 people with severe agoraphobia (fear of open spaces) might assign 20 of them to receive each of three different treatments for that disorder.
• Advantages of Between-Subjects Designs

  • Each individual score is independent from all other scores. The individual's score is not influenced by practice or experience gained in other treatments, fatigue or boredom from participating in a series of different treatments, or context effects that result from comparing one treatment to another.
• Disadvantages of Between-Subjects Designs

  • Require a relatively large number of participants. Each score is obtained from a unique individual who has personal characteristics different from all of the other participants, which may produce high variability in scores.
• Assignment bias

  Applies exclusively to research designs with between-subjects designs. Confounding variable: any extraneous variable differentiating the groups. Groups with different characteristics threaten the internal validity of study. When an experiment is confounded, it is impossible to draw any clear conclusions.
• Confounding variables in between-subjects designs

  • Confounding from individual differences (assignment bias): participants in one group may be older, smarter, taller, or have higher socio-economic status than the participants in another group.
  • Confounding from environmental variables: testing one group in a large room and another group in a smaller group.
• Three primary techniques to limit confounding by individual differences

  • Random assignment (randomization)
  • Matching groups (matched assignment)
  • Holding variables constant
• Threats to internal validity of between-subjects designs

  • Differential attrition (participant dropout rate) from one group to another
  • Communication between groups (diffusion, compensatory equalization, compensatory rivalry, resentful demoralization)
• Advantages and disadvantages of the two-group design

  • Primary advantage: simplicity, provides best opportunity to maximize the difference between the two treatment conditions. Primary disadvantage: provides little information, researcher obtains only two real data points for comparison.
• Within-Subjects Experiments


  • Uses a single group of participants and tests or observes each individual in all of the different treatments being compared. Often called a repeated-measures design. Study repeats measurements of the same individuals under different conditions.
• Threats to internal validity in within-subjects experiments


  • Confounding from environmental variables: characteristics of the environment may change from one treatment condition to another.
  • Confounding from time-related variables: between the first measurement and the final measurement, participants may be influenced by factors other than the treatments being investigated.
• Five time-related threats to internal validity

  • History (environmental events which change across treatment)
  • Maturation
  • Instrumentation (change in standards of observation)
  • Regression toward the mean
  • Order effects (practice, fatigue, and carry-over effects)
• Order effects


  Order effects can produce changes in the scores (not caused by the treatments) from one treatment condition to another. The order effect varies systematically with the treatments and contributes to the second treatment, but never the first. The group mean shows higher scores in second treatment.
• Dealing with time-related threats and order effects

  • Controlling time: the possibility of a time-related threat is directly related to the length of time required to complete the study. Shortening the time between treatments increases the likelihood that order effects will influence results.
  • Switch to a between-subjects design: between-subjects design is a better choice for research conditions that are prone to order effects, such as conditions leading to fatigue or boredom.
  • Counterbalancing: changing the order in which treatment conditions are applied from one participant to another, with the goal of using every possible order of treatments with an equal number of subjects participating in each sequence, to eliminate time-related confounding.
• Advantages of within-subjects design

  • Requires relatively few participants, eliminates problems based on individual differences, reduces variance, and increases the chances of detecting a treatment effect.
• Disadvantages of within-subjects designs

  • Each participant usually goes through a series of treatment conditions, often with each treatment administered at a different time, which may lead to time-related factors influencing the subjects' scores. Participant attrition: some of the subjects who start the study will drop-out before the study is completed.
• Disadvantages of multiple-treatment designs

  • Researcher may create too many treatments, the distinction between treatments may become too small to generate significant differences, may increase attrition if more time is required of participants, and counterbalancing is more difficult as number of treatments increases.