Flashcards
Multiple linear regression
Cards in this deck (17)
- 1What can a “multiple linear regression” be used for?- Examining how strong the relationship is between one response variable and two or more predictor variables
- Estimating the value of the response variable at certain values of the predictor variable(s) (e.g., the amount of soil erosion at a certain level of rainfall and temperature) within the range of our dataset!. - 2When do we use a “multiple linear regression”?When these assumptions are met:
- Two or more predictor variables and one response variable
- Homogeneity of variance (homoscedasticity)
- Independence of observations:
- Normal distribution:
- The relationship between the predictor(s) and response variable is linear (Linearity)
- No Multicollinearity
- Continues data / numeric data - 3What is ” Homoscedasticity”?Homoscedasticity: The variance of the response variable does not change with the value of the predictor variable.
- 4What is “Independence of observations”?Independence of observations: Each observation is independent of the others - the value of observation x is not affecting/being affected by the value of observation y.
- 5What is ”Normal distribution”?Normality: “The response variable or the residuals of the model are normally distributed”
- 6What is “linear relationship”?Linear relationship: “That there is a linear relationship between the predictor and response variable i.e. follow a straight line”
- Check visually using a scatterplot with linear relationship - 7How can we check for the assumptions of a simple linear model in R?Plot(lm)???
- 8How can we create a multiple linear model in R?lm(response variable ~ predictor variable1 + predictor variable2, data = Dataset)
- 9Show me an example of the output from plot(lm)? Explain what you are looking for?USE WEBPAGE FROM MICHAEL??
- 10What is a “multiple linear model”?
- 11How do we interpretate the results from a simple linear model in R?Summary(lm_model)
- 12Show me an example of the output from summary(lm_model) and explain the interpretation:- P-value < 0.05
- Coefficient —> (intercept) = estimate for response variable
- USE Equitable equations or so from Youtube
- 13What is “residual”?Residual: “The difference between the actual value and the value predicted by the model for any given point”
- 14What is “Multicollinearity”?Multi collinearity: Several predictor variables in a model are correlated
- Check the correlation coefficient: Variance Inlfation Factor (VIF), correlation matrix - 15How do I find “Variance Inflation Factor (VIF)” in R?X
VIF values above 10 indicating problematic multicollinearity.
Solutions: Centering the data, or removing the variables causing multicollinearity - 16How do we find the “Correlation matrix” in R?X
- Correlation coefficient should ideally be below 0.8 - 17What does it mean to center data?X
Information
- Description
- European species Statistical and Geospatial Modelling (Regression Analysis) Flashcards on Multiple linear regression, created by Marie Skalshøi on 13/03/2025.
- Created by
- Marie Skalshøi
- Views
- 0
- History
-
- Created by Marie Skalshøi · Mar 13, 2025
- Copied by Marie Skalshøi · Mar 13, 2025
- Rating