12. RM Correlations Flashcards
Define a correlation.
-A correlation is a way of measuring the relationship between two co-variables.
-The term ‘correlation’ is used to refer to a study that uses a correlation analysis.
Why are they called co-variables?
As both variables are important, so they are called co-variables (not the IV and DV).
What research design is correlational analysis?
-As there is no IV being manipulated, correlational designs are therefore not repeated measures, independent groups or matched pairs. They involve the same participants providing data for the two, measures, the co-variables.
-Correlational designs are often used when it is inappropriate or ethically unacceptable to use an experimental design. This is because the IV is not being directly manipulated.
What are the 3 types of correlation?
- Positive /
- Negative \
- No correlation ~
How can the strength of the relationship be represented on a graph?
Gradient of the line, how steep is it.
What do the correlation coefficients tell us?
The sign indicates the direction and the number indicates the strength.
+1➡️Perfect positive /
-1➡️Perfect negative \
What are some strengths of correlational analysis?
-Correlational analysis provides a means of looking at relationships between continuous variables and determining whether the relationship is significant. The alternative is to look at differences as in experimental research.
-Correlation is a useful way to conduct a preliminary analysis on data. If a correlation is not strong then we can rule out a casual relationship. We can’t demonstrate a casual relationship using a correlation but if there is no correlation between co variables then there can’t be a casual relationship.
-If the correlation is strong then further investigation is justified because there may be a casual link.
-It is an ethical way of investigating the relationship between 2 variables as they are not directly manipulated.
What are some weaknesses of correlational analysis?
-Cannot show a cause and effect relationship because there is no independent variable that has been deliberately altered. People often misinterpret
correlations and assume that a cause and effect have been found whereas this is not possible.
-If co-variables are correlated one may be causing the changes in the other but we do not know the direction of the possible effect. For example, research studies have shown a positive correlation between amount of violent videos watched and aggressiveness. It might be that watching violent videos is increasing aggressiveness, or it could be that more aggressive people choose to watch violent videos.
-There may be intervening variables that can explain why the co-variables being studied are linked. For example, research studies have shown a positive
correlation between amount of TV watched and aggressiveness. However, it is wrong to conclude that
watching TV is directly related to aggressiveness because it could be that a low boredom threshold was the cause
of both of them, an intervening factor.
-The method used to measure either co-variable may lack reliability or validity. For example, one co-variable may be measured using a questionnaire (such as when measuring aggressiveness). The reliability and validity of the questionnaire would affect the reliability and validity of the research using a correlation.
When would a Spearman’s Rho test of correlation be used?
-If the researcher wants to find a correlation/relationship.
-Their data is ordinal (can be put into order from smallest to largest) OR interval (data measured using units of equal measure).
In a Spearman’s Rho statistical test, how can we determine if the results are significant?
Observed > Critical to be significant.
If results are significant this suggests there is a strong positive relationship between the co-variables.
(If there’s an R in the name, the observed needs to be GREATER than critical).
