Correlations and coefficients

Define correlation

a mathematical technique in which a researcher investigates a relationship between two variables

What are the two variables known as?

co-variables

What is a correlation coefficient?

a number between -1 and +1 that represents the direction and strength of a relationship between co-variables

What does a correlation coefficient of +1 mean?

a perfect positive correlation between two covariables

What does a correlation coefficient of -1 mean?

a perfect negative correlation between two covariables

What does a correlation coefficient of 0 mean?

there is no correlation

What is a good correlation coefficient?

± 0.8

Give me an example of how you would write a directional hypothesis to relate the cold weather, to people’s increasing fuel bills

there is a negative correlation between the weather (measured in degrees celsius) and people’s fuel bills (measured in british pounds)

Give me an example of how you would write a non-directional hypothesis to relate the cold weather, to people’s increasing fuel bills

there is a correlation between the weather (measured in degrees celsius) and people’s fuel bills (measured in british pounds)

Give me an example of how you would write a null hypothesis to relate the cold weather, to people’s increasing fuel bills

there is no correlation between the weather (measured in degrees celsius) and people’s fuel bills (measured in british pounds)

What is the difference between a correlation and an experiment?

• correlations have two co-variables whilst experiments have an IV and DV

• correlations are used to establish a link between variables whereas experiments are used to see if one variable impacts another (cause and effect)

• correlations use scatter graphs whilst experiments can use bar charts, histograms, pie charts etc

• correlation hypothesis: there will be a/no/positive/negative correlation ; experimental hypothesis: there will be/not be an increase/ decrease/ difference

What are the strengths of correlations?

• do not require any manipulation of variables and are ‘safer’ than doing experiments on humans

• quick and economical to carry out as there is no need for a controlled environment or manipulation of variables

• can be used to direct future possible research as they provide a precise measure of how two variables are related

• secondary data can be used making them less time-consuming

What are the weaknesses of correlations?

• show linear relationships but don’t reflect curvilinear ones

• it says there may be a relationship between 2 sets of data but it doesn’t show what the cause is - not always 1 variable causing another

• ignore the fact that there could be a 3rd variable playing a part

• for them to be informative, there needs to be a large amount of data available for each variable so that a pattern can be seen