Correlations Flashcards
What does a correlation measure?
The relationship between 2 co-variables.
What is the difference between a correlation and an experiment?
An experiment investigates the difference between 2 or more conditions to establish a cause and effect relationship. Whereas, a correlation investigates the relationship between two co-variables.
What is a positive correlation?
When both co-variables behave in the same way i.e. if one variable increases so does the other.
Give an example of a positive correlation.
An increase in sleep, increased performance on a memory test.
What is a negative correlation?
When each co-variable behaves in different ways i.e. if one variable increases the other will decrease.
Give an example of a negative correlation.
As exercise levels increase, anxiety levels will decrease.
What is no correlation?
When there is no significant relationship between two co-variables.
Give an example of an event with no correlation.
A student’s performance on their maths exam has no relationship with their performance on their English exam.
When writing a hypothesis in a correlation context which word MUST you include in your statement?
Relationship
What is used to display correlation data?
Scattergraphs
What is the scattergraph showing a perfect positive correlation?
All points go diagonally upwards.
What is the scattergraph showing a perfect negative correlation?
All points go diagonally downwards.
How does a scattergraph showing no correlation look like?
All points are scattered across the graph (there’s no set pattern).
True or False- Correlation doesn’t always mean causation?
True- there could be an additional variable that has influenced our findings, so we cannot be certain that the IV has caused the DV.
Discuss 3 strengths of using a correlation as a research method.
-Correlations are a great preliminary research technique which allow psychologists to link pieces of research together so that they can be further investigated in the future.
-Correlations allow us to investigate variables that were sensitive or unethical to research in the first place because the direct manipulation of variables is not required.
-Correlation data can be displayed conveniently in a graphical format which is also easy to interpret.