Correlation

Question 1

What is correlation

Answer 1

A way of determining how related two variables are to one another
- Statistical technique used to measure and describe relationship between two variable
There is positive, negative and none (none could be seen as a parabola)

Question 2

How is correlation often shown

Answer 2

Using Z scores
- standardise the scores of x and y and multiple standardised x score by standardised y score and add it up and divide it by the sample size

Question 3

What is the sampling distribution of r

Answer 3

Describes the distribution of sample correlation coefficients (like Pearson’s r) that would be obtained if you repeatedly sampled from a population

is distributed as t ; df = n - 2 it is not normal

Question 4

Way to write a conclusion for a null of correlation

Answer 4

at (alpha lvl) significance there is sufficient/insufficient evidence to conclude that the correlation between x and y differs/doesnt differ from 0 higher or lower (linear relationship - positive or negative)

Question 5

As the power goes up

Answer 5

So does the number of participants

Question 6

As the effect size increases

Answer 6

less participants are needed

Large effect don’t need that many participants
Sample size - power increases or decreases with it

Question 7

Effect size guides for correlations

Answer 7

small = .3
medium = .5
large = .7