Flashcards in Quant: Lecture 7 Deck (10):
What is the purpose of using a correlational method?
To assess the relationship between two numerical variables that cannot be manipulated but can be measured using different scales.
What's the purpose of a correlation coefficient?
What is another term for it?
To assess the direction and strength of the relationship, the score will always be between -1 and +1.
What is the difference between a positive correlation (0 to +1) and a negative one (0 to -1)?
With a negative correlation, as one variable increases the other decreases.
List the different strengths of a correlation
What is another term for the strength of the relationship?
0= no correlation
0.1-0.2= weak correlation
0.3-0.4= moderate correlation
0.5-0.9= strong correlation
1= perfect correlation
Name one things that a strong correlation suggests and one thing that it doesn't
It suggests that the two variables are associated with each other but it doesn't suggest that there is a causal relationship, as other variables maybe involved (also know as the 'third variable' or mediating variable).
What would a perfect positive correlation look like on a scatter graph?
No correlation? (4 answers)
It would be a strong diagonal line starting in the bottom left and ending in the top right.
A strong diagonal starting in the top left ending in the bottom right.
The line of best fit would show the same as a perfect positive but there would be more variability.
Data would be randomly distributed, in a bell shape, in the middle in a straight line and in the middle going downwards.
How do you calculate pearson's r?
Add up all the scores for variable 1 and then square the answer to get (Ex)2. Then square each individual value for variable one and total it to get Ex2. Now do the same for variable 2. Next, times all the values in variable one by the adjacent ones for variable 2 and total it to get EXY. Finally, put these values in the formula.
How do you calculate degrees of freedom for pearson's r?
How do you word your answer for pearson's r?
For a *tailed test, with *df, the calculated value of r (***) was *er than the critical value (***) at a significance level of pt a significant positive relationship.