WEEK 5: Correlation and cross sectional studies Flashcards
What is a cross sectional design
Snapshot of information at a given point in time
What methods can be used instead of a cross sectional design?
Longitudinal study
Cross-sequential design ( combination of longitudinal and cross sectional design)
Does correlation mean causation?
No, just because two things are found to be related does not mean one causes the other
Example: I always have bad luck on Friday the 13th does not mean the number 13 causes bad luck
Correlational design
What is a correlation?
What is the best way to represent a correlation?
What is a correlation coefficient?
What is the correlation co efficient when there is no relationship?
What is the correlation co efficient for a perfect correlation?
a correlation is the extent to which two variables are related
The best way to represent a correlation is by using a scatter diagram
Correlation coefficient is a single number which represents the direction and strength of a relationship between two variables
Correlation coefficient of 0.00 shows that there is no relationship
A perfect correlation would be 1 or -1
When to use a correlational design
Determine?
estimate?
What kind of study does it support
What can it help avoid?
Determine whether a significant relationship exists between two variables
Estimate the strength of the relationship
Performing a naturalistic study
Avoid ethical problems which may arise if the variables were tested in an experiment
Is there manipulation in a correlational design?
What variables are used in a correlational design? examples? (what levels of data?
There is no experimental manipulation in a correlation design
Variables must be continuous scores e.g. ordinal, interval or ratio, in order to be used in a correlational design.
