Correlations

Question 1

What is the difference between an experiment and a correlation

Answer 1

-Experimental design involves the manipulation of an independent variable and a measurement of the resultant change in the dependent variable
-However with correlational studies, there is no manipulation of variables and two co variables are measured and compared to try to find a relationship

Question 2

What are co variables + examples of co variables

Answer 2

-Co variables are two variables measured by the researcher that are then compared with each other
For example: Age, IQ, Bank account balance, height, aggressiveness, reaction time etc

Question 3

Scatter grams

Answer 3

-Scatter grams are graphs that are used to plot the measurements of co-variables
-Scatter grams visually represent the relationship between co-variables

Question 4

Positive correlation on scattergrams

Answer 4

-As one co-variable increases, the other co-variable increases

Question 5

Negative correlation on scattergrams

Answer 5

-As one co-variable increases the other co-variable decreases

Question 6

Zero correlation

Answer 6

-There is no relationship between co-variables

Question 7

Correlation coefficient

Answer 7

-Correlation coefficient numerically represents both the strength and direction of the relationship between co-variables as a number between -1 (perfect negative) and +1 (perfect positive)
-No correlation is a 0
-Strong correlation is 0.8 (+ or -)
-Moderate correlation is 0.5 (+ or -)
-Weak correlation is 0.2 (+ or -)

Question 8

How are correlation coefficients calculated and how are they used

Answer 8

-Correlation coefficient can be calculated using correlational tests such as Spearman’s rho or Pearsons
-Correlation coefficient can be used to test inter-rater reliability by testing the relationship between independently produced observations
-A correlation of 0.8 is generally accepted to be a strong correlation

Question 9

Negative evaluation of correlations (does not show causation)

Answer 9

-Correlation does not show causation
-Although a strong correlation suggests a relationship between co-variables, it does not show which co variable causes the change in the other, and there is a possibility that a third unknown variable cause the change in both co variables

Question 10

Positive evaluations of correlational studies (pre existing co variables,highlight causal relationships, shows strength of correlation)

Answer 10

-Correlational studies are useful in highlighting potential causal relationships between co-variables - researchers can then test this using experimental methods to discover a cause-and-effect relationship
-Correlational studies often use pre existing co variables which means that data collection is easy and has few ethical concerns
-Correlational studies are useful in showing the strength of the relationship between co variables as well as the direction