PSYCH212_Ch10_Hayward_ONLINE_2slides_2ab93c8708dde67581e0169ca991341b

Chapter Overview

• Introduction to Simple Experiments

• Key Elements:

  • Examples: Two Simple Experiments

  • Experimental Variables

  • Causal Claims Support

  • Study Designs: Independent-groups vs. Within-groups

  • Validity Analysis of Causal Claims

Examples of Simple Experiments

• Example 1: Taking Notes

• Example 2: Motivating Babies

  • Investigates how different conditions affect attention and retention

Experimental Variables

• Independent Variables (IV):

  • Manipulated by the experimenter

  • Includes conditions (levels of the IV)

• Dependent Variables (DV):

  • Measured outcome of the experiment

• Control Variables:

  • Variables held constant to prevent confounding

Why Experiments Support Causal Claims

• Covariance:

  • Experiments show that changes in the IV correlate with changes in the DV

• Temporal Precedence:

  • Establishes that the IV precedes the DV in time

• Internal Validity:

  • Well-designed experiments rule out alternative explanations for the results

Establishing Covariance

• Comparison groups are essential:

  • Comparison Group: Used to evaluate effects of the IV

  • Types:

    • Control Group (no treatment)

    • Treatment Group(s) (varied treatment conditions)

    • Placebo Group (receives a placebo)

Establishing Temporal Precedence

• Clear evidence that the cause precedes the effect

Internal Validity in Designs

• Design Confounds:

  • Problems of systematic variability affecting results

• Selection Effects:

  • Bias in participant selection for different conditions

  • Solutions:

    • Random Assignment

    • Matched Groups Approach

Independent-Groups Designs

• Comparison of Independent-Groups vs. Within-Groups Designs

• Posttest-Only Design:

  • Measures outcomes after the treatment

• Pretest/Posttest Design:

  • Measures outcomes before and after the treatment

• Which design is more effective depends on the research situation

Within-Groups Designs

• Repeated-Measures Design:

  • Same participants engage in all levels of the IV

• Concurrent-Measures Design:

  • Participants are exposed to two different conditions simultaneously

• Advantages:

  • Participants act as their own controls

  • Fewer participants are needed compared to other designs

Validity in Within-Groups Designs

• Order Effects:

  • Impact of exposure to one condition on responses to others

• Counterbalancing to avoid order effects:

  • Full Counterbalancing: All possible orders

  • Partial Counterbalancing: Limited set of orders (e.g., Latin square)

• Disadvantages:

  • Potential for carryover effects

  • Demand characteristics affecting participant behavior

Pretest/Posttest in Repeated-Measures Design

• Definitions and distinctions between independent-groups and within-groups

Interrogating Causal Claims with Validities

• Construct Validity:

  • Quality of measurement and manipulation

  • Use of manipulation checks and pilot studies

• External Validity:

  • Generalizability of causal claims to wider populations and situations

• Statistical Validity:

  • Effect size, precision of estimates, and replication studies

• Internal Validity:

  • Alternatives explanations and controls for confounding variables