One of the key insights we use within The Happiness Index platform is ‘Key Driver Analysis’. The purpose is to help identify where there may be a relationship between questions in your survey. For example, there may be a relationship between the questions ‘How clear is the link between your role and the success of the organisation?’ and ‘Overall, how happy are you at work and please tell us why?’.
To create the Key Driver Analysis we perform a correlation calculation.
Please note our Key Driver Analysis is providing a correlation analysis and not a causation analysis (more on that later in the article).
So, what is correlation?
Correlation is a statistical calculation to see if there is a possible relationship between two or more variables.
Positive Correlation
Definition
A positive correlation occurs when there is a relationship between two variables in which both variables move in the same direction. Therefore, when one variable increases so does the other.
Example
As the temperature goes up, ice cream sales go up.
What would this look like visually?
If there is a high level of positive correlation a graph would look like this …
Negative correlation
Definition
A negative correlation occurs when there is a relationship between two variables in which an increase in one variable is associated with a decrease in the other.
Example
As the temperature goes up, coat sales go down.
What would this look like visually?
If there is a high level of negative correlation a graph would look like this …
No correlation
Definition
No correlation occurs when two or more variables do not appear to be statistically related. Therefore the value of one variable doesn’t increase or decrease in association with the increase or decrease of the other variable. The values of the variables appear to be random.
Example
There is no relationship between the intelligence of an individual and the amount of tea they drink.
What would this look like visually?
If there was no correlation a graph would look like this …
How do we visualise our Key Driver Analysis?
We display the results in a table (see the image below) with the results of the correlation calculation ranging from +1 (positive correlation) to -1 (negative correlation):
To help you interpret the results we use the following as a guide
- 0.50 or above = Positive correlation
- -0.49 to + 0.49 = No correlation
- -0.50 or below = Negative correlation
Using the results from the image above we’d say the top 3 questions all have a positive correlation with the question ‘Overall, how happy are you at work and please tell us why?’.
Therefore, if your goal was to improve the happiness of your workforce you’d focus your follow up activity on:
- Ensuring every employee is clear on the link between their role and the success of the organisation.
- Keeping employees informed of what is happening across the organisation.
- Making sure everyone is clear on the requirements of their job.
What is the difference between correlation and causation?
Causation means one thing causes another, in other words, action A causes outcome B. Whereas correlation is simply showing there is a relationship between action A and action B. Let’s look at an example:
- Sunny weather means people generally eat more ice cream and get sunburnt.
- There’s a correlation between eating ice cream and getting sunburned, showing there is a relationship.
- But eating ice cream doesn’t cause you to get sunburn and vice versa.
- Instead, both events are caused by something else, the sunny weather.
We use correlation, and not causation, as while we can statistically see a relationship or lack of relationship between the variables, without real world organisational context we cannot definitely say that one variable causes a change in the other or not. We suggest this correlation analysis is used to support your contextual understanding of your organisation and people.