Week 8 Lecture 8 - correlation and partial correlation Flashcards

1
Q

What does bivariate mean?

A

2 variables

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

What do relationships vary in terms of?

A
  • form
  • direction
  • magnitude/strength
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What are the assumptions for a linear correlation?

A
  • both variables should be continuous (level of measurement)
  • related pairs: each ppts should have a pair of values
  • absence of outliers
  • linearity –> points in the scatterplot should be best explained with a straight line
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

If there are outliers on your scatterplot, what may it be appropriate to do?

A

remove the outlier
as it may not be representative of the larger population that you are sampling from

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

True or false
correlations are sensitive to range restrictions

A

true
e.g., floor and ceiling effects

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

If the assumptions for a linear correlation are seriously violated, what non-parametric equivalent should you use instead?

A

Spearman’s rho (or Kendall’s Tau if fewer than 20 cases)

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

What letter is denoted to Pearson’s correlation coefficient?

A

r

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

What does covariance provide?

A
  • a measure of the variance shared between x an y variables
  • r = ratio of covariance to separate variances
  • we can obtain a measure of separate variances by multiplying the SD for x and y
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

true or false
r is a ratio

A

true

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

If covariance is large relative to the separate variances, what will r be?

A

further from 0

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

If covariance is small relative to the separate variances, what will r be?

A

close to 0

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

How does r reflect how well a straight line fits the datapoints?

A

if data points cluster close to the line, r will be further from 0
if data points cluster far from the line, r will be close to 0

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

Does SPSS, give dof output?

A

no, you have to work it out
dof = N-2

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

Do we get sampling error when considering correlations?

A

yes
we would get different r values for different samples within the same population

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

What are some features of the r distribution?

A
  • mean = 0
  • sample deviates from 0 expressed in standard error units
  • can determine probability of obtaining r value of a given magnitude when null is assumed to be true
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Does SPSS provide the 95% CIs on its output?

A
  • no
  • can obtain using a online calculator
17
Q

How many d.ps do you report r to?

A

3

18
Q

How many d.ps do you report CIs to?

A

2

19
Q

What does r^2 express?

A

the proportion of the separate variances that is shared
the remaining variance is split equally between x and y

20
Q

When asked about relative strength, what measure should you use?

A

r^2

21
Q

What is r^2 a measure of?

A

effect size

22
Q

What is a partial correlation?

A

relationship between 2 variables, whilst removing influence of a third given variable

23
Q

In partial correlation, if the 3rd variable decreases the r value and it is no longer significant, how would we interpret this?

A

3rd variable may explain relationship

24
Q

In partial correlation, if the 3rd variable decreases the r value but is still significant, how would we interpret this?

A

3rd variable may partially explain relationship

25
Q

In partial correlation, if the 3rd variable doesn’t decrease the r value, how would we interpret this?

A

3rd variable doesn’t influence the relationship