Interior Analysis
- Examines how predictor variables relate to the dependent variable to evaluate model quality.
- Identifies strong positive or negative associations to inform which predictors to include.
- Uses those relationships to build regression models that better reflect the data.
Definition
Section titled “Definition”Interior analysis is a method used in regression analysis to assess the quality of a model. It involves looking at the relationships between different predictor variables and the dependent variable, and determining how well the model captures these relationships.
Explanation
Section titled “Explanation”Interior analysis begins by examining the relationships between each predictor variable and the dependent variable. By identifying the direction and strength of these relationships (for example, strong positive or strong negative associations), the analyst determines which predictors should be included in the regression model. The model is then constructed to account for those identified relationships so that it more accurately predicts the dependent variable. Overall, interior analysis helps improve model quality and deepen understanding of the data.
Examples
Section titled “Examples”Example 1 — Age, Income, Education → Happiness
Section titled “Example 1 — Age, Income, Education → Happiness”Dataset variables:
- Age: the age of an individual in years
- Income: the income of an individual in dollars per year
- Education: the level of education an individual has attained (e.g. high school, college, graduate degree)
- Happiness: a measure of an individual’s overall happiness on a scale from 1 to 10
In this example, interior analysis examines relationships between the predictor variables (Age, Income, Education) and the dependent variable (Happiness). One might find a strong positive relationship between age and happiness and a strong positive relationship between income and happiness. Those identified relationships guide inclusion of predictors (for example, Age and Income) in the regression model used to predict Happiness.
Example 2 — Weight, Height, Exercise → BMI
Section titled “Example 2 — Weight, Height, Exercise → BMI”Dataset variables:
- Weight: the weight of an individual in pounds
- Height: the height of an individual in inches
- BMI: the body mass index of an individual
- Exercise: the amount of exercise an individual engages in per week
In this example, interior analysis examines relationships between the predictor variables (Weight, Height, Exercise) and the dependent variable (BMI). One might find a strong positive relationship between weight and BMI and a strong negative relationship between exercise and BMI. Those identified relationships guide inclusion of predictors (for example, Weight and Exercise) in the regression model used to predict BMI.