RM - Correlations Flashcards
What is a correlation?
A method used to analyse data, not a research method.
Is correlation a research method?
No
What is a correlation used for?
To analyse the association between two variables, in this case called co-variables.
A study that uses a correlational analysis should be called what?
‘A study using a correlational analysis’.
Or,
‘a correlation’ or a ‘correlational study’.
What does a correlational hypothesis do?
States the expected association between the con-variables.
What is a strong positive correlation on a scatter graph?
The dots form a line from bottom left to top right.
What is a strong negative correlation on a scatter graph?
The dots form a line from top left to bottom right.
How do you know whether the pattern of dots on a scatter graph represents a meaningful, systematic association?
Amateur - You can eyeball the graph and decide whether it looks like a (strong) +/- correlation.
Researcher - Use a statistical test to calculate the correlation coefficient and then deciding whether it is significant using a table of significance.
What are the strengths of (using) correlations?
Used to investigate trends in data and if a correlation is significant, then further investigation is justified (such as experiments). If a correlation isn’t significant, then you can probably rule out a causal relationship.
The procedures in a correlation can usually be easily repeated meaning that the findings can be confirmed.
What are the limitations of (using) correlations?
Variables are simply measured and no deliberate change is made so no conclusion can be made about one co-variable causing the other.
People assume causal conclusions which is a problem because such misinterpretation of correlations mean that people design programmes for improvement based on false premises. (e.g. improving attendance will improve exam results).
The supposed causal connection may actually be due to intervening variables, unknown variables which can explain why the co-variables being studied are linked.
As with experiments, a correlation may lack internal/external validity. The co-variables may not be able to be generalised.
Therefore, the strength of the correlational research lies in investigating the extent of relationships between variables, which can be particularly useful in the early stages of research. However, it is imperative to avoid any causal inferences.
What are co-variables?
The two measured variables in a correlational analysis. The variables must be continuous.
What is a continuous variable?
A variable that can take on any value within a certain range. Liking football (on a scale of 1 to 10) is continuous whereas the football team a person supports isn’t. The latter could be arranged in any order.
Define correlation:
Determining the extent of an association between two variables; co-variables may not be linked at all (zero correlation), they may both increase together (positive correlation), or as one co-variable increases, the other decreases (negative correlation).
What is a correlation coefficient?
A number between -1 and +1 that tells us how closely the co-variables in a correlational analysis are associated.
What is a curvilinear correlation?
A non-linear relationship between co-variables.
