Correlations

The difference between correlations and experiments

• Experimental designs require manipulation of the independent variable and a measurement of the resulting change in the dependant variable

• In a correlational study, no variables are manipulated, two co-variables are measured and compared to look for a relationship

Co-variables

• The two factors/variables that are measured/collected by the researcher and then compared to each other

• E.g: Age, IQ, reaction time, bank account balance, number of pets, height etc.

Scattergrams

• A graph used to plot the measurements o two co-variables

• Scattergrams visually display the relationship between co-variables

Correlation types

• Positive: As one co-variable increases, the other co-variable increases

• Negative: As one co-variable increases, the other decreases

• Zero correlation: There is no relationship between the values of the two co-variables

Analysis of the relationships between co-variables

• The strength and direction of a correlation can be described visually with a scattergram, or numerically with a correlation coefficient

• A correlation coefficient represent both the strength and direction of the relationship between the co-variables as a number between -1 and 1

• Correlation coefficients are calculated using statistical tests such as Spearman’s rho or Pearson’s inter rater and test reliabilit

• A correlation coefficient equal to or greater than 0.8 is usually judged to show a strong correlation

Correlation evaluations

• Correlation does not show causation

• While a strong correlation may suggest a relationship exists between two variables, it does not show which co-variable led to the change in the other co-variable and there is the possibility that an unknown third variable caused the change in bot co-variables

• Correlational studies can highlight potential causal relationships which can then further be tested with experimental methods to discover cause and effect relationships

• Often the co-variable data already exists and is easily accessible which means that there is usually few ethical problems in data collection

• Correlation coefficient is useful tool in describing both the direction and strength f relationships between factors