HIERARCHICAL MULTIPLE REGRESSION

Give three examples of continuous dependent variables.

Examples: Height (cm), Salary (USD), Reaction time (milliseconds), Weight (kg), or Test performance (0–100).

How many independent variables are needed for multiple regression?

Two or more independent variables.

What levels of measurement can independent variables have?

Independent variables can be measured at either the continuous or nominal level.

Give examples of nominal independent variables.

Gender (male/female), Ethnicity (Caucasian, African American, Hispanic), Profession (doctor, nurse, dentist).

What is a dichotomous variable?

An independent variable with only two categories (e.g., male/female).

What is a polytomous (or multinomial) variable?

An independent variable with three or more categories (e.g., profession: surgeon, dentist, nurse).

How are ordinal variables handled in multiple regression?

They can be entered if treated as continuous or nominal variables (e.g., Likert scale items).

What does "independence of errors" mean?

The residuals (errors) for each observation should be independent of each other.

What is the assumption of linearity?

There should be a linear relationship between each predictor (and their combination) and the dependent variable.

What does homoscedasticity mean?

The residuals (errors) should have equal variances across all levels of the independent variables.

What is multicollinearity?

When two or more predictors are highly correlated, making it hard to determine their unique effects.

What assumption is violated if there are extreme data points that distort results?

The assumption of no significant outliers, high leverage points, or highly influential points.

What assumption refers to the shape of residuals’ distribution?

The errors (residuals) should be approximately normally distributed.

How many total assumptions are required for multiple regression?

Eight assumptions.

What is another term for “residuals”?

Errors — the differences between observed and predicted values.