In research, we often want to find out if there is a relationship between two or more variables, eg. is there a relationship between hard work and exam results? A **positive correlation** means that high values of one variable are associated with high values of the other variable, eg. the more hours of work student put into revision, the higher their exam score. A **negative correlation** means that high values of one variable are associated with low values of another variable, eg. the more hours spent in the pub, the lower the exam results. If there is no relationship, the variables are said to be **uncorrelated**.

A common mistake is to assume that when there is a strong connection between two variables, one must be causing the other to happen. **Correlation does not prove causation**. This mistake is often made by the media, politicians, etc. who make comments about the links between unemployment and crime or certain tastes in music and drug abuse. The correlations may exist, but they do not prove any causal relationship.

- Does unemployment cause crime?
- Does crime cause unemployment?
- Or does some third factor (poverty? poor education? discrimination?) cause both?

Consider: there is a proven correlation between church attendance in a neighbourhood and crime rates in that area. How can this be explained?

Among the core studies, Maguire uses the correlational method when looking at the brains of London taxi drivers. She found that, the longer they had been taxi drivers, the greaterthe alteration in the parts of their brains responsible for navigation and memory.

Strengths of correlational research | Limitations of correlational research |

Researcher can measure relationships between naturally occurring variables without manipulating or controlling them Correlations can predict the value of one variable when we only have information about the other one | Does not produce conclusions about cause and effect – some other factor may be responsible Some relationships between two variables do not show up as correlations |

Now download this slideshow about the different types of self-reports used in Psychology.

- This lecture handout might help you take notes.
- This handout explains the different types of correlational relationships

You are going to be collecting data from two independent measures and plotting a scattergram to see if there is a relationship between these measures. Some suggestions include:

- Shoe size and IQ
- Hours of sleep and alertness
- Size of breakfast and happiness
- Attitude to subject and success in subject
- Hours spent watching TV and memory recall

**Hypothesis**

When you are looking for a correlation between two variables, you are looking for a relationship. Words like “difference” and “effect” should not appear in a correlational hypothesis. You should predict the direction of the relationship you are looking for:

- A positive correlation means that as one variable increases, so does the other
- A negative correlation means that as one variable increases, the other decreases (and vice versa)

A correlational hypothis will look like this:

**There will be a correlation **[say whether positive or negative]** between variable X and variable Y**

**Measuring Correlations**

You will need to draw a SCATTERGRAM to show your data. This is where you have one variable plotted on each axis and points on the scattergram represent each participant’s score on both variables.

- If all the points are around a line moving from the bottom left to the top right of the graph, this suggests a
**positive correlation**between the variables. - If the pattern is from top left to bottom right, this suggests a
**negative correlation**

**Are the following correlations positive or negative?**

- The hotter the weather, the higher the ice cream sales
- The smaller the miniature pony, the more valuable it is
- The more overcrowded the tower block, the more vandalism reported
- The lower the calorie consumption, the less the weight
- The lower the food prices, the more business done by the supermarket

**Assertiveness and watching TV**

**Aim:**To see whether watching TV is related to assertiveness (because many programs encourage such behaviour)

**Method:** You need to obtain two pieces of data from everyone in your class.

- The number of hours they watched TV in the last 24 hours (presuming the more TV watched, the more “assertive” programmes were watched).
- An “assertiveness” score each person should rate themselves (or be rated by someone else) on a scale of 1 to 10 where 10 is very assertive.

**Results: **Put the scores for each person on a scatter graph. Calculate a correlation coefficient to assess the strength of the correlation. Use the Excel method described below

**Excel: **Open a new document (select * file new blank workbook*). Select

- This Correlations Proforma will help you organise your notes.
- This Evaluation Proforma will help you focus on the strengths and weaknesses of your research.
- This Excel Document has a sample scattergram already on it. Try changing the scores to see the scattergram change shape...