Intro to Business Statistics: Correlation and Regression Dummy Variables Coding Categorical Predictors

What This Is

Dummy variables, also known as indicator variables, are used to code categorical predictors in regression analysis. A retail chain wants to know if average daily sales exceed $10,000 based on the day of the week (Monday to Sunday). By using dummy variables, the chain can analyze the effect of day of the week on sales while controlling for other factors.

Key Formulas & Symbols

• Dummy Variable (D): A binary variable that takes the value 1 if a condition is met and 0 otherwise. For example, D = 1 if the day is Monday, and D = 0 otherwise.
• Multiple Regression Equation: Y = β₀ + β₁X₁ + β₂D + ε, where Y is the dependent variable, X₁ is a continuous predictor, D is a dummy variable, β₀ is the intercept, β₁ and β₂ are coefficients, and ε is the error term.
• Coefficient of a Dummy Variable (β₂): The change in the dependent variable for a one-unit change in the dummy variable, holding all other predictors constant.
• F-Statistic: F = (R² - R²₀) / (1 - R²) / (n - k - 1), where R² is the multiple R-squared, R²₀ is the R-squared from a model without the dummy variable, n is the sample size, and k is the number of predictors.
• p-Value for F-Statistic: The probability of observing an F-statistic at least as extreme as the one observed, assuming the null hypothesis is true.
• Null Hypothesis (H₀): The hypothesis that the coefficient of the dummy variable is equal to zero, i.e., β₂ = 0.
• Alternative Hypothesis (H₁): The hypothesis that the coefficient of the dummy variable is not equal to zero, i.e., β₂ ≠ 0.

Step-by-Step Procedure

1. State Hypotheses: Specify the null and alternative hypotheses for the dummy variable.
2. Choose Test: Select the appropriate test statistic (e.g., F-statistic) and determine the degrees of freedom (e.g., n - k - 1).
3. Compute Test Statistic: Calculate the F-statistic using the formula above.
4. Find p-Value or Critical Value: Determine the p-value associated with the F-statistic or find the critical value from the F-distribution table.
5. Compare to α: Compare the p-value or critical value to the significance level (α = 0.05).
6. Conclude: Reject the null hypothesis if the p-value is less than α or the critical value is exceeded, and conclude that the dummy variable is statistically significant.

Common Mistakes

• Mistake: Misinterpreting the p-value as the probability that the null hypothesis is true.
• Correction: The p-value is the probability of observing the data (or more extreme) if the null hypothesis is true. It does not provide information about the probability of the null hypothesis being true.
• Mistake: Failing to center the dummy variable before including it in the regression equation.
• Correction: Centering the dummy variable ensures that the intercept is meaningful and that the coefficient of the dummy variable represents the change in the dependent variable for a one-unit change in the dummy variable.
• Mistake: Ignoring the assumption of linearity between the dummy variable and the dependent variable.
• Correction: The assumption of linearity should be checked before including the dummy variable in the regression equation.

Quick Practice Problems

1. A marketing firm wants to know if the average daily clicks on a website exceed 100 based on the day of the week (Monday to Sunday). By using dummy variables, the firm can analyze the effect of day of the week on clicks while controlling for other factors. Calculate the F-statistic for the dummy variable.

F-statistic = (R² - R²₀) / (1 - R²) / (n - k - 1) = (0.05 - 0.02) / (1 - 0.05) / (100 - 2 - 1) = 0.03 / 0.95 / 97 = 0.0032

1. A quality control team wants to know if the average defect rate exceeds 5% based on the production line (A or B). By using dummy variables, the team can analyze the effect of production line on defect rate while controlling for other factors. What is the p-value associated with the F-statistic?

p-value = 0.01

1. A sales manager wants to know if the average daily sales exceed $10,000 based on the region (North, South, East, or West). By using dummy variables, the manager can analyze the effect of region on sales while controlling for other factors. What is the coefficient of the dummy variable for the South region?

Coefficient = 0.05

Last-Minute Cram Sheet

• Dummy Variable (D): A binary variable that takes the value 1 if a condition is met and 0 otherwise.
• Multiple Regression Equation: Y = β₀ + β₁X₁ + β₂D + ε.
• Coefficient of a Dummy Variable (β₂): The change in the dependent variable for a one-unit change in the dummy variable.
• F-Statistic: F = (R² - R²₀) / (1 - R²) / (n - k - 1).
• p-Value for F-Statistic: The probability of observing an F-statistic at least as extreme as the one observed, assuming the null hypothesis is true.
• Null Hypothesis (H₀): The hypothesis that the coefficient of the dummy variable is equal to zero, i.e., β₂ = 0.
• Alternative Hypothesis (H₁): The hypothesis that the coefficient of the dummy variable is not equal to zero, i.e., β₂ ≠ 0.
• Centering Dummy Variable: Ensures that the intercept is meaningful and that the coefficient of the dummy variable represents the change in the dependent variable for a one-unit change in the dummy variable.
• ⚠️ p-Value is NOT the probability that H₀ is true – it’s the probability of observing the data (or more extreme) if H₀ is true.
• Assumption of Linearity: The assumption that the relationship between the dummy variable and the dependent variable is linear.
• F-Distribution Table: A table that provides critical values for the F-statistic.
• Degrees of Freedom: The number of observations minus the number of parameters estimated.