Toolkit 2 simple linear regression Flashcards
What is the model structure of linear regression?
Y=3x
What is a bi variate data set?
When there is 1 x and 1 y variable
What is a multi variate data set?
1 y variable but multiple x variables
What is Linear Regression?
It is the concept of ordinary least squares in the evaluation of b0 and b1
It introduces the idea of samples as estimates of population
Interpretations of the Regression Table and the Analysis of variance table
residuals and concept of model specification
Catch all plots
How is the linear data set collected and shown?
Simple scatter plot
What is the correlation coefficient ?
the way of quantifying how close the scatter of points was to a linear form. -1
Where b0 and b1 are the best unbiased estimates of the true unknown values Y hat is e best estimate of the true Y. What is the equation?
y hat = b0 + b1X
What is the objective you aim to find looking at the linear equation scatter plot?
It is to find a line through the scatter that minimises this error for all points
How many lines are there for any given set of points that gives the minimum clue for the error which occurs in a linear regression scatter graph?
1
What computer programmes offer statistical software?
SPSS and others
What are the coefficients ?
These give the coefficients of the equation and can be used to plot the line through the centre of the data
What does the R value =?
The correlation coefficient
If the f value is greater than 4 and the associated p value less than 0.05 then what is the interpretation?
The model is statistically sound and it explains a significant amount of the variation of Y
Do large r squared and significant t and F value ensure the data is fitted well?
No. The results from regression are valid and have meanings only in so far as the assumptions concerning the residuals in the model are satisfied.
Should the residual values show any pattern or not ?
No
