We already know that a correlation shows the mutual relationship between two or more factors. Do not forget that correlation is not equal to causation.
However, what is a correlation matrix? A correlation matrix is a table that shows the correlation coefficients between variables. In the table, you will find the values of correlation and see which pairs have the highest correlation.
Tip: The value in the diagonal of the table is always 100% because the correlation between a variable and itself is 100%.
What Does a Correlation Matrix Look Like?
In our application, you will find the correlation matrix in the left sidebar under the Relations menu. Every column is a variable in the table, and the correlations are shown between them. It is like a relation map, but here you will see the strength of the relations as numbers. The higher the number, the stronger the correlation between the variables.
Tip: The darker the field, the stronger the relationship between the variables.
If you click on the icon at the top of a column, then that column will be ordered by strength. Doing so will help you quickly find the strongest relationship in each column. Just like our other features, the correlation matrix is more than a simple table.
You can click on any value in the table, and the correlation will be visualized as Suggested Charts on the right side of the screen.
Search and Filter
We know that you will sometimes have more than 100 columns in your spreadsheet, and you don’t want to see all of them in your correlation matrix. The good news is, that is not necessary.
On the left sidebar, you can pick only the columns that you would like to see in the correlation matrix. You can also search between your columns by typing the column’s name into the search field.
If you choose more than three variables, you will see only the certain variables on the screen, otherwise, the highlighted columns are shown in the first rows of the matrix.
Filtering is also available here. You can set one or more filters to analyze a preferred argument. In the matrix, only the filtered data will be shown. If you want to see the unfiltered data, just hit the Pause icon right next to the filter.
Any changes that you make here - selecting columns and filtering - will be used if you switch between menus (Charts, Relation Map, and so on).
A relation table is like a correlation matrix. It shows the correlation coefficients between your columns. However, there are some differences between the two visualizations.
With a relation table, the relations are listed, and a short explanation is next to each number. This feature helps you to visualize correlation values instantly and order them from minimum to maximum if you click on the Relation label.
The colors have also meanings here. Blue means a positive correlation, and orange indicates the presence of a negative correlation. Just like in correlation matrix, the darker the color, the stronger the relation.
If you are a man of numbers and you trust more in them, then this will be the most useful feature for you. The table shows you the strength of correlation in numbers, and you can visualize it on charts or in the relation map at any time.
Feel free to save and download your reports, or add them to your canvas and share them with your colleagues.
With AnswerMiner, you can do that instantly, and no coding is required.