Correlations simply describe the relationship between two variables in statistical terms, but it is not a research method in itself. Data collected from various research methods such as observations, questionnaires, and experiments can be analysed to see if there is a relationship between two variables. An example of a correlation is the relationship between hours spent revising for an exam and the grade attained.
Unlike experiments which have an independent variable and a dependent variable, correlations are described in terms of covariables. This is because both variables in a correlation vary (change) and are measured, and neither one is set or controlled by the researcher.
Positive and negative correlations
Correlations can be positive or negative. A positive correlation describes a relationship in which both variables increase together. A negative correlation describes a relationship in which one variable increases as the other decreases. For example, the relationship between hours spent revising for an exam and the grade attained is a positive correlation. This is because as hours spent revising increase, then the result attained also increases.
The strength of a correlation is described as a correlation coefficient. Coefficients range from -1.0 to +1.0, with a coefficient of less than zero describing a negative correlation and a coefficient above zero describing a positive correlation. A good rule of thumb is to consider of 0.0 to 0.3 as weak, 0.3 to 0.7 as moderate, and above 0.7 as strong.
Correlations are plotted on graphs called scattergrams (also called scatter plots or scatter graphs). A scattergram has one covariable on the x-axis and the other on the y-axis, and a straight line that best fits the points is plotted (a line of best fit). The gradient of the line is the same as the correlation coefficient.
A scattergram showing a positive correlation between hours spent revising for an exam and exam grade attained.
Strengths and weaknesses of correlational research
- Quick and easy. Correlations are a quick and easy way to see whether or not there is a relationship between two variables that is worth exploring further. They can use preexisting data (e.g. children’s SATs scores at the end of year 6 and their GCSE scores at the end of year 11) and if a correlation is found it may be worth investigating why there is a correlation.
- Describes the strength of a relationship. A correlation coefficient is a simple and objective way to describe the strength of a relationship between two variables. Expressing it as a precise number makes it clear and easy to understand.
- Correlations do not equal causation. This means that it is impossible to claim that one covariable actually causes the other covariable, as it could be that a third unknown variable (a mediating variable) is causing both variables to change together. In the example above of the positive correlation between hours spent revising and exam grade attained, it is impossible to be sure that hours spent revising actually causes a higher grade to be attained. It could be that children who revise more have more stable home lives, and the more stable home life could mean they do more homework and pay more attention at school, which in turn could lead to higher grades in exams. It is very difficult to establish cause and effect.
- Correlations can be misused. As finding a correlation between two variables tells us very little other than that a relationship exists, it is very difficult to make accurate conclusions about the causes of the relationship. Media, governments, and even sometimes scientists often make wild claims based on correlations that sound convincing to the public and support an argument, but in reality the relationship could mean something completely different.
A Level exam tips
Answering exam questions (PSYA1 AQA A specification)
Questions about correlations tend to provide a graph that has to be interpreted. It is essential to read the title and descriptions of the axis so that you know what variables the graph is describing. You must then decide whether the correlation is positive or negative. The question may also ask for one strength or one weakness of correlations, and to answer you should describe one of each and relate it to the topic the graph is describing.
Give one strength and one weakness of using correlations to describe the relationship between hours spent in day care and aggressive behaviour (2+2 marks)
One strength is that the relationship between hours spent in day care and aggression can be simply and easily described. You must then describe the relationship in the graph you have been given.
One weakness is that cause and effect cannot be established because correlations do not equal causation and there may be another mediating variable. For example, children who spend more hours in day care may have less caring mothers, and so the children may react aggressively in order to get attention. Maternal sensitivity is the mediating variable.