Variables and Their Classifications Lecture

Definition & Core Idea of a Variable

• A variable is any characteristic, attribute, or factor that
  • May vary or change from one individual, situation, or time point to another.
  • Can affect or alter the results of a study if not properly identified and controlled.
  • Assumes different values, which can be numerical (quantitative) ​or categorical (qualitative).
• Significance in research
  • Variables embody the unknowns researchers wish to measure, compare, predict, or explain.
  • Correct identification prevents confounding and strengthens internal validity.

Illustrative Examples of Common Variables

• Sexual identity
• Socio-economic status
• Educational attainment
• Height, weight, age, temperature, number of children, group size, Likert-scale attitude ratings, contest rankings.

Master Classification Framework

1. Numerical (Quantitative) Variables
2. Categorical (Qualitative) Variables
3. Experimental Variables
4. Non-experimental Variables
5. Variables by Number Studied (Scope)

1. Numerical (Quantitative) Variables

• Describe measurable quantities; answer “how many?” or “how much?”
• Provide data suitable for arithmetic operations & most statistical tests.

a. Continuous Variables

• Can take any fractional or decimal value within an interval (including negatives).
• Examples: time, age, temperature, height, weight.
• Practical note: Use when precision is required (e.g., $37.48\,^\circ C$ body temperature).

b. Discrete Variables

• Restricted to countable whole numbers; no fractional or negative values between points.
• Examples: number of children in a family, frequency of visits, group size.

Two Measurement Levels for Numerical Data

• INTERVAL SCALE
  • Equal differences between values carry meaning.
  • Zero is arbitrary (does NOT mean “none”).
  • Example: Celsius or Fahrenheit temperature (0 °C ≠ no temperature).
• RATIO SCALE
  • Has all properties of an interval scale plus a true, meaningful zero that indicates complete absence.
  • Permits statements like “twice as much.”
  • Examples: age, height, weight.
  • Statistical implications: allows use of coefficients of variation, geometric means.

2. Categorical (Qualitative) Variables

• Capture attributes, qualities, or group membership; answer “what type?” or “which category?”

a. Nominal Variables

• Purely labels with no intrinsic ranking.
• Examples: gender, eye color, marital status, blood type.

b. Ordinal Variables

• Have ordered categories; distances between rankings are unknown or unequal.
• Examples: education level (elementary, high school, college), Likert scale (strongly disagree → strongly agree), contest winners (1st, 2nd, 3rd).

c. Dichotomous vs. Polychotomous (extension)

• Dichotomous: only two categories (e.g., pass/fail, yes/no).
• Polychotomous: three or more categories (e.g., eye colors, academic ranks).

3. Experimental Variables (Used in True/Controlled Experiments)

• Independent Variable (IV)
  • The factor purposely manipulated by the researcher.
  • Hypothesized cause of change.
  • Example: dosage of a new drug (low, medium, high).
• Dependent Variable (DV)
  • The outcome measured; expected to depend on the IV.
  • Example: blood pressure reduction after drug intake.
• Extraneous Variables
  • Additional factors that may influence the DV but are not the primary focus.
  • Must be controlled or randomized to avoid confounding (e.g., participants’ caffeine intake, room temperature).

4. Non-experimental Variables (Observational / Predictive Studies)

• Predictor (Explanatory) Variable
  • Observed variable believed to predict or explain variance in another variable.
  • Example: study hours as a predictor of exam score.
• Criterion (Response / Outcome) Variable
  • The variable being predicted or explained.
  • Example: exam score itself.
• Key distinction: Predictor ≠ manipulated; relationships are correlational, not causal.

5. Variables by Number Studied (Scope of Analysis)

• Univariate Study
  • Investigates one variable.
  • Example: average age of students.
• Bivariate Study
  • Examines two variables to determine a relationship/correlation.
  • Example: hours of sleep vs. GPA.
• Polyvariate (Multivariate) Study
  • Involves three or more variables simultaneously.
  • Example: predicting job performance using motivation, IQ, and years of experience.

Ethical & Practical Implications of Variable Selection

• Misclassification (e.g., treating ordinal data as interval) can lead to invalid statistical inferences.
• Control of extraneous variables upholds research integrity and participant safety.
• Socio-economic status and sexual identity variables require sensitive handling to maintain privacy and avoid bias.

Quick Skill-Check Matrix (Self-Test)

• Identify each data example and classify:
  • Body temperature: Interval (or Ratio under Kelvin).
  • Annual income: Ratio.
  • Eye color: Nominal.
  • Satisfaction rating (1–5): Ordinal.

Connecting to Prior Statistical Principles

• Variable type determines permissible descriptive statistics (mean vs. median vs. mode) and inferential tests (e.g., $\chi^2$ for nominal, ANOVA for interval/ratio).
• Understanding scales anchors the concept of levels of measurement (Nominal < Ordinal < Interval < Ratio) introduced in earlier lectures.

Real-World Relevance

• Marketing: Discrete counts of product purchases & categorical brand preference.
• Medicine: Continuous dosage vs. categorical adverse-effects rating.
• Education: Ordinal class rank alongside ratio study-hours.

Summary Cheat-Sheet

• Variable = changeable characteristic; must know “type” to choose correct analysis.
• Numerical → Continuous (fractions) vs. Discrete (counts) → Interval vs. Ratio scales.
• Categorical → Nominal (labels) vs. Ordinal (ranked) → may be Dichotomous or Polychotomous.
• Experimental → IV, DV, control extraneous.
• Non-experimental → Predictor, Criterion.
• Scope → Univariate, Bivariate, Polyvariate.
• Ethical precision protects validity, replicability, and participant welfare.