Research Methods: Variables and Measurement (Topic 2)

Variables and Measurement

• Definition of Variables:
  • Symbol able to assume various numerical values.
  • Property of an organism/event that can be measured.
  • Changeable aspect in different testing conditions.
  • Attribute of a phenomenon.

Types of Variables

• Independent and Dependent Variables:

  • Independent variable: Condition manipulated by the researcher with at least two levels. It's the antecedent variable.
  • Dependent variable: Behavior measurement reflecting effects of the independent variable.
  • Variable classification may change depending on the study context.

• Continuous and Discrete Variables:

  • Continuous variable: Falls along a continuum (e.g., distance, time).
  • Discrete variable: Falls into categories with no intermediate values (e.g., male/female).
  • Naturally discrete variables are inherent (e.g., biological sex), while artificially discrete variables are created (e.g., age categories).
  • Statistical procedures differ based on variable types (e.g., Pearson's correlation for continuous variables).

• Quantitative and Qualitative Variables:

  • Quantitative variable: Varies in amount (e.g., reaction time).
  • Qualitative variable: Varies in kind (e.g., college major).
  • Quantitative variables are generally continuous while qualitative are discrete.

Measurement

• Definition of Measurement:

  • Assignment of numerical values to objects/events based on rules.
  • Involves operational definitions to quantify constructs for study.

• Levels of Measurement:

  • Labels: Numbers used for tracking, not for analysis (e.g., participant IDs).
  • Nominal Scale: Grouping without quantitative relationships (e.g., coding genders).
  • Ordinal Scale: Rank ordering without absolute measures (e.g., rankings in polls).
  • Interval Scale: Measures the extent of an attribute without true zero (e.g., satisfaction ratings).
  • Ratio Scale: Includes true zero and properties of previous scales; rare in behavioral sciences (e.g., temperature in Kelvin).

Evaluation of Measurement Methods

• Reliability: Consistency of scores over time and conditions.
• Validity: Extent to which a method accurately measures what it intends to measure.

Types of Test Validity

• Criterion-related Validity: Predictive effectiveness of a test in forecasting future behavior; includes:

  • Concurrent validity.
  • Predictive validity.
  • Postdictive validity (based on timing of data collection).

• Content-related Validity: Assesses adequacy of content sampled by the test (e.g., classroom tests) based on expert judgment.

• Construct-related Validity: Measures alignment with theoretical constructs; includes:

  • Convergent validity: Measures of the same construct are correlated.
  • Divergent validity: Measures of different constructs are not correlated.

• Face Validity: Subjective assessment of whether a test appears to measure its intended variable.

Multi-Trait/Multi-Method Matrix

• A framework for assessing construct-related validity, ensuring measures of the same constructs converge while measures of different constructs diverge.

This quiz covers the definitions and concepts related to variables and measurement in research methods. You will explore the different types of variables and their significance in the context of research. Test your understanding of how variables can change under different conditions.