BIOL 3P96 Lecture Notes - Lecture 18: Cicindela, Observational Error, Squared Deviations From The Mean
Document Summary
Biol 3p96 lecture 18: multiple explanatory variables. A mathematical representation of the relationship between a response variable y and one or more explanatory variables. Anova and linear regression are linear models. Response variable can be represented by a linear model plus random error. Scatter of y measurements around the model (random error) results from chance and various effects not included in the model. The model of linear regression is a straight line. Least squares yielded the best fit of the model to the data. Hypothesis testing compares the null model to the regression model. Regression model is a superior fit to the data. Data points lie closer to the line. If the f-ratio is significantly large, then the reduction in the magnitudes of residuals represents a significant improvement in fit. General linear models (glm) is an all-purpose method that extends the linear regression approach in two ways: Allows for more explanatory variables to be included in the model.