Revision (2)

Question 1

What are the two ways in which linear regression and correlation are conceptually alike?

Answer 1

• Both of these concepts aim to identify linear relationships between two variables.
• Both of these concepts identify a line of best fit by using the method of least squares.

Question 2

To what does the method of least squares refer?

Answer 2

To the minimisation of the squared distances of datapoints to a line of best fit (used in tests of both linear regression and correlation).

Question 3

How are linear regression and correlation conceptually unalike?

Answer 3

Correlation does not treat the the variables it is investigating as separate entities, whereas linear regression does (it asks about the effect of one variable on another).

Question 4

Between what does linear regression distinguish (unlike correlation)?

Answer 4

An independent variable (the variable that may have an influence on another) and a dependent variable (the variable being influenced by another).

Question 5

While both linear regression and correlation analyse the covariance between two variables, how do they differ?

Answer 5

In how they standardise this covariance.

Question 6

How is covariance standardised for correlation?

Answer 6

By multiplying the standard deviation of the independent variable involved by the standard deviation of the dependent variable involved.

Question 7

How is covariance standardised for linear regression?

Answer 7

By squaring the standard deviation of the independent variable involved.

Question 8

What is the equation for the regression line?

Answer 8

Y = mX + c

Question 9

What are the ingredients necessary for the regression line equation?

Answer 9

• The gradient of the regression line (m)
• The intercept (c)

Question 10

To what does the intercept of the regression line refer?

Answer 10

To the value of Y when X = 0.

Question 11

What does it mean if the gradient of a regression line is significant?

Answer 11

That as the score on the independent variable involved changes, the score on the dependent variable involved also changes in a systematic way (the line representing their relationship is not flat).

Question 12

What does R2 measure?

Answer 12

The proportion of the variance observed in the dependent variable that can be explained by the independent variable.

Question 13

What is the formula for R2?

Answer 13

R2 = SSM/ SST

Question 14

Why is the R2 value multiplied by 100?

Answer 14

To generate a percentgae.