Correlations

What is a correlation?

A technique that measures whether or not there is a relationship between two variables (known as co-variables).

What type of technique is a correlation?

A statistical technique, it’s used to measure or quantify the strength of a relationship between two variables

How can correlation be used in research?

It can be used as a method in its own right OR to analyse data gathered from another research method.

How is correlation represented?

Graphically and numerically (using a correlation coefficient).

How is data displayed and interpreted in a graphical correlation study?

The data is plotted as points on a scattergram, and a line of best fit is drawn through the points to show the trend in the data.

How can correlations be represented mathematically?

As a correlation coefficient.

What does the correlation coefficient show?

The correlation coefficient (represented by r) is a number between +1 (perfect positive correlation) and -1 (perfect negative correlation). It shows how closely two variables are related.

What does the strength of a correlation depend on?

The closer the correlation coefficient is to +1 or –1, the stronger the relationship between the two variables. If it’s close to 0.0, there is little or no correlation.

What does the sign of the correlation coefficient (+ or –) tell us?

The sign shows the direction of the relationship:

• A positive correlation (+) means that as one variable increases, the other also increases.

• A negative correlation (–) means that as one variable increases, the other decreases.

What does it mean if a correlation coefficient is –0.52 compared to +0.52?

Both show a correlation of the same strength, but –0.52 indicates a negative correlation, while +0.52 indicates a positive correlation.

What is positive correlation?

This means that as one variable increases, the value of the other variable will also increase e.g. the more time spent studying, the higher the exam results.

• When the correlation coefficient is calculated, it is between 0 - +1, the nearer the value is to +1, the stronger the relationship

• Anything measured above 0 on the correlation coefficient becomes a positive correlation

What is negative correlation?

This means that as one variable increases, the other variable being measured decreases e.g. the more time spent socialising with friends, the less revision gets done for an exam.

• When the correlation coefficient is calculated, it’s between 0 - -1, the nearer the value is to -1, the stronger the relationship

• Anything measure below 0 on the correlation coefficient becomes a negative correlation

What is zero correlation?

This means that the two variables are not related at all e.g. the earnings of a deep-sea fisherman and the size of a portion of chips. 

• Zero correlation shows no relationship between the two variables

What does a significant correlation tell us?

It simply shows that there is a relationship between two variables.

Why can we not infer causation from a correlation?

Because correlation does not show why or how variables are related — it does not mean that one variable caused the other to change.

What are the 3 possible explanations for why or how there is a relationship between two variables?

1. Causality - one variable caused the other to change e.g. there’s a positive correlation between time spent revising and exam score (it’s likely that revising more causes better exam results)

2. Chance - variables that happened to be related e.g. there’s a correlation between the number of people wearing sunglasses and ice cream sales (this might happen by chance, without any real link between the two)

3. Third factor involved - another variable is causing the relationship e.g. intervening variable e.g. there is a positive correlation between wearing sunglasses and ice cream sales, but the real reason is a third variable — hot weather, which increases both

What is the main difference between correlations and experiments?

• In an experiment, the researcher manipulates the independent variable to see its effect on the dependent variable, allowing cause and effect to be established

• In a correlation, variables are not manipulated, so cause and effect cannot be determined — a relationship may exist, but one variable doesn’t necessarily cause the other, there may be all sorts of ‘other variables’ causing this effect, which are called intervening variables

What are the strength of correlations?

• They allow you to measure things that cannot be manipulated experimentally because it would be unethical, since they make use of existing data, meaning psychologists can carry out research without raising any ethical issues

• They can suggest trends that may lead to further experiments, as they are quick, cheap, and help show whether further investigation is worthwhile.

What are the weaknesses of correlations?

• They can’t show cause and effect — a third (intervening) variable may explain the relationship, which can lead to misinterpretation/serious public misunderstandings 

• They depend on the data collection method (e.g. questionnaires), which can reduce validity due to people not answering honestly on a questionnaire (because they want to look good - called social desirability bias) or them misremembering things