Introductory Psychology: Research Methods

Research Methods in Psychology

Importance of Scientific Methods

  • Scientific Method: A structured approach used to study psychology, aimed at reducing bias and systematically collecting evidence.

    • Helps formulate a clear and observable basis for supporting or refuting theories.

Learning Retention Strategies

  • Spacing Effect: Learning spaced over time results in better retention compared to cramming.

    • Example: Students who spaced out their studying before exams perform better than those who crammed.

Characteristics of a Good Theory

  • Falsifiability: A key characteristic of a good theory where predictions can be tested.

    • Example: Darwin’s Theory of Evolution is falsifiable, whereas Freud’s psychoanalytic theory is less so.

    • Confidence in theories increases if findings consistently support the hypothesis.

Steps in Scientific Investigation

  1. Formulate a testable hypothesis

  2. Select the research method and design the study

  3. Collect the data

  4. Analyze the data and draw conclusions

  5. Report the findings

Research Design Outline

  • 1. Steps in a Scientific Investigation

    • Correlational designs

    • Experimental designs

  • 2. Other Considerations

Variables and Operationalization

  • Variable: Any property that can vary across individuals or over time.

    • Examples: Amount of Sleep, Eye Color.

  • Operational Definition: A concrete description of a property in measurable terms.

    • Example: Wealth could be operationalized by $ in bank account, number of properties owned, or net worth.

Formulating a Hypothesis

  • A good hypothesis includes:

    1. Your variables

    2. The predicted relationship between those variables

    • Example Hypothesis: ( V1: \text{Number of Friends} ) vs ( V2: \text{Loneliness} ).

    • Prediction: Having more friends results in feeling less lonely.

Research Method Selection

  • Choose how to measure your variables.

    • Types of designs: Correlational or Experimental.

Correlational Designs

  • Purpose: Measure how closely two variables vary together or predict changes in one variable from another.

    • Definitions:

    • Predictor Variable: The one that is used to predict.

    • Criterion/Outcome Variable: The one that is measured.

  • Correlations:

    • Positive (+): Both variables increase or decrease together.

    • Negative (-): One variable increases as another decreases.

    • Zero Correlation: No predictable relationship.

Correlation Coefficients

  • Pearson r: A statistical measure that determines the strength and direction of a correlation, ranging from -1 to 1.

    • Values close to 1 or -1 indicate strong correlations.

Designing a Correlational Study

  • Example Hypothesis: "Internet use in class is negatively related to class performance." (Ravizza, Uitvlugt, & Fenn, 2017)

    • Operationalize variables: Internet use can be measured by pages visited and time spent browsing; class performance can be measured by exam grades.

  • Data Collection: Include metrics like data usage (GB), grades, and course grades.

Analyzing Data and Conclusions

  • After data collection, analyze for trends and determine correlation strengths.

    • Example correlation coefficient: ( r = -0.25 ) for non-academic internet use relative to grades.

Issues with Directionality and Causality

  • Example: "Higher caffeine consumption relates to poorer sleep quality."

    • Potential variables that could affect the relationship include work schedule/hours.

  • Another Example: "Ice cream sales rise with an increase in murders."

    • Both events may correlate with warm weather, showing that correlation does not imply causation.

Questioning Causation

  • Can we assert that internet use causes poorer exam performance?

    • Important considerations: Potential third variables may mislead causation conclusions.

  • Correlational Design Limitations:

    • Cannot infer causation due to issues like directionality and third variable risks.

  • Correlational Design Benefits:

    • Often more ethical and useful for making predictions that can lead to future experiments.

Testing Hypotheses

  • Two methods: Correlational designs and Experimental designs.

Designing an Experimental Study: Bandura's Social Learning Theory

  • Theory: Individuals learn behaviors such as aggression through observation, imitation, and modeling.

  • Hypothesis: "Children who watch adults display aggressive behavior will act more aggressively."

Steps in Experimental Design

  1. Formulate a testable hypothesis

  2. Select the research method and design the study:

    • Experimental methodology where the independent variable (IV) is manipulated (e.g., aggression model) while observing the dependent variable (DV) of aggressive behavior.

Independent and Dependent Variables

  • Independent Variable (IV): The behavior that is varied or manipulated (e.g., exposure to an aggressive adult model).

  • Dependent Variable (DV): The behavior measured as a response (e.g., aggressive behavior in children).

Controlling Variables

  • Keep all other factors constant across experimental conditions such as setting and experimenter to eliminate alternative explanations.

Random Assignment and Participant Selection

  • Random Assignment: Involving random distribution of participants across conditions to ensure equality.

  • Population vs. Sample:

    • Population: The entire group of interest;

    • Sample: A subset selected to represent the population.

  • Importance of generalization in study results.

Sampling Techniques

  • Random Sampling: Each member of the population has an equal chance of selection.

  • Convenience Sampling: Participants are chosen based on ease of availability, which can affect study generalizability.

Data Analysis and Conclusions

  • Analyze collected data to draw meaningful conclusions regarding the IV and DV.

Reporting Findings

  • Important to publish findings in reputable journals to contribute to the scientific community and validate research.

Ethical and Practical Considerations in Research

  • Certain questions may be impractical or unethical for experimental designs; controlled settings may also hinder real-world application of findings.

Advantages and Disadvantages of Experimental Designs

  • Benefits: Can infer causation with precise control over variables.

  • Drawbacks: Some experimental methods may not be feasible or ethical, and findings may not generalize to real-world scenarios.

Comparison of Research Designs

Correlation

Experiment

Measures relationship strength (-1 to 1)

Manipulates IV to see its impact on DV

Observational

Controls outside variables or assigns participants to conditions

Cannot infer causation

Can infer causation (IV change affects DV)

Additional Considerations

  • Placebo Effect: Changes in behavior are influenced by the belief in treatment.

  • Experimenter Bias: Expectations of the experimenter can affect outcomes.

  • Demand Characteristics: Participants modify their behavior based on their interpretation of the study’s purpose.

Examples of Biases

  • "Smart rats vs. Dull Rats": Experimenters perceived smart rats as performing better based on biases (Rosenthal & Fode, 1963).

  • Priming Studies: Example of how bias in interpreting stereotypes can alter behavior (Bargh, Chen, Burrows, 1996).

Limitations and Solutions

  • Researchers often mitigate flaws by replicating studies and refining methodologies, ensuring findings are reliable and valid.

  • Double-Blind Experiments: Both the researcher and participant are blinded to conditions to minimize biases.

Replication in Research

  • The process of repeating studies to confirm results and enhance reliability in psychological research.

Conclusion: Understanding Research Design

  • Ability to critically distinguish between types of research design and identify potential biases is essential for psychological science.

1) Differentiate between correlational and experimental designs.
2) Assess sources of bias in research design and implement corrective measures.