Correlations

Features of a correlational study

• A correlation measures the relationship between two variables

• The relationship can be positive or negative

• The relationship can be strong or weak

• A correlation can be represented on a scattergram and analysed using a Spearman’s Rho or Pearson’s R to provide quanti

Positive correlation

As one variable increases, so does the other e.g. the hotter the weather, the more ice creams are sold

Negative correlation

As one variable increases, the other decreases e.g. the colder the weather, the more woolly jumpers are sold

Scattergrams

• Some variables are quantitative e.g. temperature, number of ice creams sold

• Some variables have to be given a quantitative value e.g. The amount of aggression scored on a rating scale from 1-10

• On a scattergram one variable is placed on the x axis, and the other is placed on the y axis

• By plotting variables against each other we can show their correlational relationship

Correlational coefficient

• Quantifies the strength of relationships in correlations

• A stats test needs to be conducted to calculate a correlational coefficient, either Spearman’s Rho or Pearson’s R

• If we have a relationship, we would expect a high correlation coefficient e.g. 0.8

• A + sign in front of a coefficient shows a positive correlation, a - showing it is negative

• In psychology, a correlation coefficient is around +0.8 or -0.8 is a strong correlation

• If it is 0, there is no correlation

Rules of writing a correlational hypothesis

• You should never use the words difference or effect

• You should alway use the words correlation or relationship

• You should not use the terms IV or DV as there are 2 co-variables

• The 2 co-variables should be fully operationalised

• You need quantitative data from both co-variables to be able to do a correlation

Directional hypothesis

• There will be a positive correlation between the number of hours of revision completed by a student in one week and % score in an EoTT

• There will be a negative correlation between the number of days a child spends in an orphanage and their IQ level measured by an IQ test

Non-directional hypothesis

There will be a correlation between the number of cups of coffee drunk before a psychology CAP and CAP performance measured by %

Null hypothesis

• The same rules apply, but predicts there will be no relationship

• There will be no relationship between the number of hours of revision completed y a student in one week and % score in an EoTT. Any relationship will be due to chance.

Strengths of correlations (+)

• Establishes a relationship between 2 variables when manipulation of variables is not possible e.g. a relationship between diet and heart disease. Looks as wider health issues that would not be ethical to experiment on

• Correlations allow researchers to conduct statistical analysis on situations where experimental manipulation would not be ethical or practical e.g. a psychologist could correlate the level of social deprivation with the physical health of infants however it would not be possible to conduct an experiment

• Fairly easy to analyse using scattergrams (and Spearman’s rho) compared to case studies where data is a lot more complex or difficult to interpret

Weaknesses of correlations

• A correlation does not establish cause and effect. We cannot be certain that one variable causes a change in the other

  • e.g. A psychologist could find a relationship between the number of hours of violent TV children watch and the amount of aggressive behaviour they display, however they couldn’t say that watching violent TV caused the children to be more aggressive