Linear Regression 1 Flashcards

1
Q

What do correlations tell us?

A

How strongly two continuous variables covary.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

To what is correlation not equal?

A

Causation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

To what is linear regression conceptually similar?

A

Correlation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is the difference between correlation and linear regression?

A

Unlike linear regression, correlation does not treat the two variables that it is evaluating differently.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Between which two variables does linear regression distinguish?

A

Independent and dependent variables.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What do we fit to data to describe a relationship between the two variables being evaluated in line with the linear regression concept?

A

A straight line.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is investigated after having fit a straight line to data in order to describe a relationship between two variables (in line with the linear regression concept)?

A

How much variance is explained by the straight line fit to the data, relative to the unexplained variance.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Which optimisation approach is widely used to find a line that best represents the relationship between two variables?

A

The method of least squares.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

How are all straight lines represented?

A

By Y = mX + c, where X is the score on the independent variable, Y is the predicted score on the dependent variable, m is the gradient of the line, and c is the intercept.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Which aspect of the equation representing straight lines is indicative of a positive, negative, or no relationship between two variables?

A

m

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What would m = 0 indicate?

A

The absence of a relationship between two variables.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

How is m calculated?

A

From the covariance of X and Y.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

How does the m calculation differ from the correlation coefficient calculation?

A

m is standardised according to the squared deviations in X only, rather than the X and Y standard deviations multiplied.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is the equation for m?

A

(COVXY/ s2x) = (Σ(X-X̄)(Y-Ȳ))/ (Σ(X-X̄)2)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

With what are regression equations also written on the end?

A

An error term (+ e).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What does an error term signify?

A

What we can’t model with a straight line.

17
Q

How is the predicted Y symbolised?

A

As Ŷ.

18
Q

Through which intercept will a regression line go?

A

Through the intercept (x = 0)(y = 0).

19
Q

Through which values will a regression line go, as well as through the intercept (x = 0)(y = 0)?

A

The means of x and y (x = X̄)(y = Ȳ).

20
Q

When do we have our straight line model?

A

Once we have calculated our parameters.

21
Q

What do we need to assess the goodness of fit of a linear regression model?

A

SST, SSR, and SSM.

22
Q

To what does SST refer?

A

To the total variability in data (the difference between each Y value and the mean Y value).

23
Q

To what does SSR refer?

A

To the differences between the observed Y values and those predicted by the model (the residual/ unexplained variance).