STAT 7: Analysing Relationships Flashcards
1
Q
What is the correlation coefficient?
A
- denoted as ‘r’
- ranges between -1 and 1
- use cor() function in R
- tells us the direction and strength of the association
2
Q
What is r2?
A
- this value tells us how much variation in one variable can be explained by the other
- e.g. if r is +0.53, r2 is 0.28
- therefore, 28% of the variation can be explained by the other variable
- 72% comes from another source
3
Q
What are the two types of experiment where you would compare continuoud variables?
A
- Looking for an association between two variables when neither are experimentally manipulated
- Experimentally manipulating one variable and looking to see whether another variable changes too
4
Q
What is linear regression used for?
A
- using the line of best fit, you can estimate one variable from another
- use lm() in R
- only suitable for two continuous variables that have a visually linear relationship, so the goodness of fit needs to be assessed
5
Q
Is a linear regression equation suitable for this data?
A
6
Q
Is a linear regression equation suitable for this data?
A
7
Q
Is a linear regression equation suitable for this data?
A
8
Q
Is a linear regression equation suitable for this data?
A
9
Q
Is a linear regression equation suitable for this data?
A
10
Q
What does ~ mean?
A
- tilde
- means ‘is explained by’
- e.g. a~b means ‘a is explained by b’
- a is response variable, b is explanatory
11
Q
What does plot (a~b) make on R?
A
12
Q
What function do we use to predict unknown values in linear regression in R?
A
13
Q
How can we predict unknown values in a linear regression model on R?
A
