Correlation

• Correlation is a method of sales forecasting
• Correlation looks at the strength of a relationship between two variables
• For marketing, it might be useful to know that there is a predictable relationship between sales and factors such as advertising, weather, consumer income etc.
• Correlation is usually measured by using a scatter diagram, on which data points are plotted
• For example, a data point might measure the number of customer enquiries that are generated per week against the amount spent on advertising
• Independent variable is the factor that causes the other variable to change, plotted on the x-axis
• Dependent variable is the variable being influenced by the independent variable on the y-axis
• Positive correlation is a positive relationship exists where as the independent variable increases in value, so does the dependent variable
• Negative correlation is a negative relationship exists where as the independent variable increases in value, the dependent variable falls in value
• No correlation is where there is no discernible relationship between the independent and dependent variable
• Strong correlation means that there is little room between the data points and the line
• Weak correlation means that the data points are spread quite wide and far away from the line of best fit
• If the data suggests strong correlation, then the relationship might be used to make marketing predictions
• The big danger with correlation is of believing there is a causal link between two variables, when, in fact, they are not related
• Confidence Interval is a measure of the likely accuracy of the results of a sample -> with a 95% confidence internal, there is a 0.95 probability that the true average will be where the sample believes it will lie (19/20 correct)