Psych 2300 Week 11 Notes

Week 11 of Psych 2300: Social Science Research Methods

Final Exam Information

  • Date: Tuesday, December 16

  • Time: 12:00 PM to 2:00 PM (2-hour exam, despite typical 3-hour slots)

  • Location: Center for Engineering Innovation, Room 1100 (across from the business building)

  • Format: In-person exam for the entire class; previously lab group format for midterm.

Exam Details

  • Question Count: 70 multiple-choice questions

  • Cumulative Nature: Primarily focuses on material covered since the midterm, but includes foundational concepts from before the midterm.

    • Essential for understanding more complex topics from later in the course.

    • Example question type: Scenario-based questions requiring identification of research design and operational definitions.

  • Study Preparation: Use the same study guide as for the midterm; a new study guide for post-midterm materials will be provided.

  • ID Requirement: Bring student ID; IDs will be checked upon entry.

  • Sign-Out Sheet: Required, participants must sign out upon completion of the exam.

  • Latecomers Policy: No entries allowed after the first person leaves the exam. Generally, no entry once 30 minutes have passed since the start.

Important Reminders

  • Lab Attendance: Required for this week to review important details for Lab Assignment #3.

Key Concepts Covered in Prior Lectures

  • Reliability and Validity: Introduction and overview, including construct validity.

Internal Validity

  • Definition: The extent to which conclusions can be made about causal relationships in a study. It considers how the different measures work together to answer questions about cause and effect.

  • Causality Criteria: To assert causation, three criteria must be met:

    1. Co-variation: There’s an association or correlation between the variables.

    2. Temporal Precedence: The cause precedes the effect.

    3. Ruling Out Rival Hypotheses: Controlling for alternative explanations.

  • Increasing Internal Validity:

    • Design studies to better meet the causality criteria.

    • Well-constructed experiments generally boast higher internal validity than non-experimental designs.

  • Threats to Internal Validity: These can significantly impact the study's findings and can generally be categorized as time-related or group-related threats.

Time-Related Threats

  1. History Threat: A coincidental event that occurs during the study impacting outcomes.

    • Example: Studying teen stress during September 11th can skew results due to heightened stress.

  2. Maturation Threat: Normal developmental changes in subjects that may affect outcomes (e.g., children’s language development).

    • Example: Change in vocabulary scores due to maturation rather than interventions.

  3. Testing (Test Reactivity): Responses can shift due to the act of being measured.

    • Example: Sensitivity training may alter participants’ responses on follow-up surveys.

    • Hawthorne Effect: Change in behavior when individuals know they're being observed.

  4. Instrument Decay: Changes in measurement instruments or human error over time.

    • Example: A change in ADHD behavioral criteria between DSM-IV and DSM-V may affect data collection.

  5. Attrition: Dropout from the study can skew results. Attrition rate reflects the percentage of participants who drop out, impacting the reliability of data.