Chapter. 3: Defining and Measuring Variables

Defining and Measuring Variables

Overview of Variables

  • Importance of operational definitions in research: Operationally defining constructs is critical for the proper measurement and manipulation of variables.

Steps for Conducting Research

  1. Find a Research Idea

    • Select a topic to explore and review existing literature to discover unanswered questions.

  2. Form a Hypothesis

    • Develop a hypothesis, which is a tentative answer to the research question.

  3. Determine How You Will Define and Measure Your Variables

    • Identify specific procedures for defining and measuring all research variables.

    • Plan to evaluate the validity and reliability of measurement procedures.

  4. Identify Participants or Subjects

    • Decide the number of participants and their required characteristics.

    • Plan for ethical treatment of participants.

  5. Select a Research Strategy

    • Consider both internal and external validity.

    • Choose between various research strategies: experimental vs. descriptive, correlational, nonexperimental, or quasi-experimental.

  6. Select a Research Design

    • Decide among designs such as between-subjects, within-subjects, factorial, or single-case designs.

  7. Conduct the Study

    • Collect data according to the established procedures.

  8. Evaluate the Data

    • Use appropriate statistical methods (both descriptive and inferential) to summarize and interpret results.

  9. Report the Results

    • Adhere to guidelines for formatting and style while ensuring accurate reporting. Protect anonymity and confidentiality of participants.

  10. Refine or Reformulate Your Research Idea

    • Utilize findings to modify or expand upon the original research idea, or to generate new hypotheses.

Operational Definitions of Constructs

  • The need for operational definitions arises when measuring/manipulating a variable:

    • Example Constructs to Definitions:

    • Gratitude toward partner:

      • Asking individuals if they agree with the statement, "I appreciate my partner."

      • Observing couples and counting how often they thank one another.

    • Wealth:

      • Surveying participants to report their income.

      • Coding participants’ vehicles based on value: higher vehicle value indicates increased wealth.

    • Intelligence:

      • Administering an IQ test.

      • Recording brain activity during complex problem-solving tasks.

    • Recommendation: Consult previous research for ideas on operational definitions.

Evaluating Measurement Quality

  • Criteria for Measurement Quality:

    1. Validity of Measurement: Determines if a measurement accurately captures what it claims to measure.

    2. Reliability of Measurement: Assesses whether the scores remain stable/consistent across different circumstances.

  • Key Point: A measure can be reliable without being valid; however, it cannot be valid unless it's also reliable.

Types of Validity

  1. Face Validity:

    • Looks like it measures what it is supposed to measure at first glance.

  2. Concurrent Validity:

    • New measure scores correlate directly with scores from an established measure of the same variable.

  3. Predictive Validity:

    • Measure accurately predicts an outcome it is intended to predict (e.g., a college entrance exam predicting college GPA).

  4. Construct Validity:

    • Measurement reflects the behavior of the variable itself; e.g., weight should correlate with food intake.

Types of Reliability

  • Measurement inconsistency may arise from various factors such as changes in the observer, environment, or participants.

  • Key Metrics of Reliability:

    1. Test-Retest Reliability:

    • A measure should yield consistent results over repeated trials.

    1. Inter-Rater Reliability:

    • Measurements should be consistent regardless of who conducts them.

Scales of Measurement

Scale

Characteristics

Examples

Nominal

Qualitative distinctions without quantitative distinctions

Nationality, Ethnicity, Ice cream flavor

Ordinal

Rank-ordered sequence with identified differences and directionality

Clothing sizes (S, M, L, XL), Olympic medals

Interval

Ordered with equal intervals but no absolute zero point

Temperature (Fahrenheit & Celsius), Golf scores

Ratio

Ordered with equal intervals and an absolute zero point

Number of correct answers, Length, Weight

Modalities of Measurement

  1. Self-report measures

  2. Physiological measures

  3. Behavioral measures

    • Example Questionnaire Scale:

      • Very often, Often, Sometimes, Rarely

      • Indicators: ☐ ☐ ☐ ☐

Measurement Issues

  • Ceiling Effect:

    • All scores are clustered together at the high end, limiting output variability.

  • Floor Effect:

    • All scores cluster at the low end, similarly restricting variability.

  • Artifacts:

    • External factors may distort measurements.

    • Experimenter Bias: Expectations of the experimenter affecting results. Mitigate through blind designs.

  • Demand Characteristics:

    • Cues that may reveal the research purpose to participants.

  • Reactivity:

    • Participants may change their behavior when they know they are being observed; thus, reassure them of anonymity and confidentiality and encourage honest responses.