Lecture on Types of Variables and Types of Data

Overview

• Variables describe the measurements made in an experiment or collected data.
• Variables can be explanatory (independent) or response (dependent).
• Explanatory variables are manipulated to observe effects on response variables.
• Response variables are the outcomes of interest in an experiment.

Explanatory vs. Response Variables

• Explanatory (Independent) Variables:
  • Example: Drug dosage in clinical trials.
  • Considered the cause in cause-effect relationships.原因
• Response (Dependent) Variables:
  • Example: Blood pressure in response to a drug.
  • Considered the effect in cause-effect relationships.結果

Exercise Examples

• Sleep Deprivation and Math Ability:
  • Explanatory: Sleep deprivation
  • Response: Math ability
• Glucose Levels and Drug Dosage:
  • Explanatory: Drug dosage
  • Response: Glucose levels
• Tumor Size and Radiation Therapy:
  • Explanatory: Radiation therapy
  • Response: Tumor size

Types of Variables

• Quantitative Variables: Numeric measurements (e.g., height, BMI).
• Qualitative Variables: Descriptive characteristics (e.g., hair color, course difficulty).
• Variables can vary over time for individuals or groups.

Data Types

• Categorical Data: Falls into distinct categories.
  • Nominal: No natural order (e.g., sex, ethnicity).
  • Ordinal: Natural order or ranking (e.g., education levels, maturation stages).
• Numeric Data: Whole or real numbers.
  • Discrete: Countable, whole numbers (e.g., number of mice).
  • Continuous: Any value, depends on measurement precision (e.g., height, age).

Conversions and Misconceptions

• Continuous data can be converted to categorical (e.g., age groups).
• Nominal data can be assigned arbitrary numbers.
• Ordinal data can be treated as continuous due to inherent order.

Examples

• Discrete Data: Car sales numbers by month.
• Continuous Data: Height or weight measurements.
• Conversion: Age as continuous, ordinal, or nominal data depending on categorization.

Key Points

• Pay attention to the data type (categorical vs. continuous) rather than the variable label.
• Data categories are not absolute and can be interchanged based on research needs.