Statistical Significance For The Correlation Coefficient

Question 1

Where do we find significance in correlation when we have a population?

Answer 1

If the coefficient within the extreme 2.5% of each end of the distribution we can reject the null hypothesis.

Question 2

How do you calculate significance when we only have a sample?

Answer 2

1. Take sample variance and estimate the population variance.
2. From this we can estimate a distribution of correlations in samples for a null population. This is constrained by sample size, the smaller the sample, the stronger the correlation needed to achieve significance. (Usually done through SPSS)
3. Once this is done, we will be given a significances tables, with sample sizes. It shows the minimum coefficient needed for each sample to make it significant. These tables are two-tailed.

Question 3

How do we do this in SPSS?

Answer 3

Analyze -> correlate -> bivariate

Move each variables into the box

Question 4

How do you write significance with correlation?

Answer 4

R (DofF) = correlation, p

Question 5

What is a null hypothesis in the correlation coefficient?

Answer 5

0