PSY173 Week 9 Notes

Acknowledgment of Country

• Murdoch University is situated on the lands of the Whadjuk and Binjareb Noongar people.
• Acknowledgment of Noongar elders' enduring culture and contributions.
• Emphasis on Murdoch University's role in a long tradition of learning.

Goals of the Session

• Understand survey design.
• Understand the correlational approach.
• Reading assigned: Chapter 7 of "Discovering the Scientist Within".
• Quiz 3 will relate to content from Chapter 7 and this session.

PART 1: Understanding Survey Design

Surveys

• Definition: A quantitative research strategy for systematically collecting information from individuals, generalizing it to a larger population.
• Key Benefits:
  • Collects large amounts of data quickly.
  • Correlation is often used in survey research.

Psychometric Tests

• Definition: Psychometric implies a measure of the mind.
• Uses: Quantification of psychological characteristics.
• Types of information measured:
  • Personality traits.
  • Mental disorders.
  • Attitudes.
  • Intelligence.

Questionnaire Design Considerations

• Important questions to consider:
  • How many items?
  • Which items?
  • What response format?
  • Item ordering?
  • Phrasing of items?
  • Minimizing response biases?
  • Instructions provided?

Item Design Considerations

• Avoid:
  • Leading questions (e.g., “Wouldn’t you say that…”).
  • Jargon and colloquialisms.
  • Double-barreled questions (e.g., “Do you enjoy heavy metal and ballet?”).
  • Double negatives (e.g., “Is it important to never stop never stopping?”).

Process for Designing a Questionnaire

• Important iterative steps:
  • Operationalize your construct.
  • Conduct interviews, focus groups, and observations.
  • Identify key themes.
  • Create items based on informed understanding.
  • Pilot the questionnaire.
  • Test reliability and validity, then refine items as necessary.

Response Bias

• Common biases include:
  • Yeah-saying: A tendency to give similar responses across questions.
  • To minimize:
  • Vary item phrasing.
  • Alternate positive and negative phrased items.
  • Error of central tendency: Participants may avoid using extreme responses.
  • Address by using a wider response scale (5-9 points preferred).

Types of Response Formats

1. Yes/No Responses: Clear binary format for questions.
2. Likert Scales: Measure level of agreement with a statement.
3. Semantic Differential Scales: Measure attitudes using opposites.
4. Ranking Statements: Rank statements by importance.
5. Open-ended Specific Questions: Allow for detailed responses.

Scoring a Questionnaire

• Basic scoring procedures include:
  • Assign numerical values to responses (e.g., Yes=1, No=0).
  • Reverse scoring for certain items.
  • Combine scores for overall constructs.
  • Presentation of scores can include summed totals or averages for clarity.

Normative Values

• Comparison of individual scores to group scores based on:
  • Sample demographics (e.g., age, gender).
  • Establish mean scores and range (high/low).

PART 2: Understanding Correlational Approach

Overview of Correlation

• A research approach investigating relationships between variables.
• It's crucial to remember correlation does not imply causation (e.g., ice cream sales vs shark attacks).

Selection of Statistical Tests

• Determining the appropriate test based on the nature of variables (nominal, ordinal, interval/ratio).
• Examples of tests:
  • Correlation Tests: Pearson's for interval/ratio, Spearman's for ordinal.
  • Chi Square for nominal data.

Correlation Examples

• Examples of correlational studies include relationships between:
  • Happiness and income.
  • Urbanization and mental health concerns.
  • Trust levels and voting behavior.

Scatterplots and Direction of Association

• Scatterplots visually represent relationships:
  • Positive correlation: Both variables increase together.
  • Negative correlation: One variable increases while the other decreases.

Pearson's Correlation Coefficient

• Measures the strength and direction of a linear relationship between two interval/ratio variables, denoted as r.
• Interpretation:
  • Range from -1 (perfect negative) to +1 (perfect positive).
  • 0 indicates no correlation.
  • The closer to -1 or +1, the stronger the relationship.

Assumptions of Correlation Coefficients

• Must be on a continuous scale.
• Ensure no outliers are present.
• There should be a linear relationship between variables.

Reporting the Results

• Example:
  • "The mean attendance score was 6.40 (SD = 2.61), with lab report marks averaging at 4.27 (SD = 2.99). There is a significant negative correlation (r(28) = -0.93, p < .001), indicating that higher attendance correlates with lower marks."