Correlations Flashcards
What is a correlation?
Correlations investigate the strength of a relationship between 2 variables (co-variables)
What are the three types of correlation:
- positive
- negative
- no correlation
Positive correlation
As one variable increases, so does the other
Negative correlation
As one variable increases, the other decreases
What are the two ways in which correlation data is presented?
- Scatter graph
- Correlation coefficients
If you get asked to draw a scatter graph, what are marks allocated for?
(1) Clear and specific title outlining both variables - “A scatter graph displaying the relationship between V1 and V2”
DONT FORGET TO APPLY
(1) Labelling X and Y axis + use units if applicable
(1) Accurate plotting
Correlation coefficients
The strength and direction (positive/negative) can be expressed by a numerical value known as correlation coefficient.
- negative correlation coefficient indicates a negative correlation
- positive correlation coefficient indicates a positive correlation
Indicates strength of correlation:
1-0.7
Strong correlation
Indicates strength of correlation:
0.7-0.4
Moderate correlation
Indicates strength of correlation:
0.4-0.1
Weak correlation
Indicates strength of correlation:
0.1-0
No correlation
Interpreting correlations exam q:
1) identify direction of correlation and strength if applicable
2) state what the direction of correlation is between and be specific with covariables
3) explain what happens to variables in the type of correlation identified, e.g, “As _____ increases, so does ______”
2 Strengths of correlations:
+ good preliminary tool for research- good starting point. By helping identify rs between 2 variables, allows grounds to further investigate the variables through methods like lab experiments to investigate C+E between 2 variables
+ since correlation only involves gathering/ measuring data + no manipulation- useful for variables which may be unethical to manipulate in exp setting
One weakness of correlations:
- Variables not controlled + cannot be manipulated in any way. Only informs us that they are related, not if variable A causes variable B or vise versa. There may be a 3rd variable affecting both variables