The category aggregation function determines how the values of different category items are combined into one value. This affects the summary value when a variable is collapsed, aggregating the variable in formulas and grouping by other categories (see Category modifiers). By default, Causal sums up all category items when aggregating a variable. To change this, right-click on the variable and select a different Category Aggregation. For example, if you've split Bonus % input by Department, it might be more appropriate to see the average of the Departments, not the total.
The options are:
No Aggregation: can be used when there is no meaningful way to aggregate a variable - it will just show a dash instead of any number.
Average: takes the average value of items in the category
Min and Max: takes the minimum or the maximum value of the category
Median: takes the median value of all the items within this category. Note, if the variable has many categories, the median value will be calculated over the "lowest level" cells. E.g. if you have Sales variable broken down by Profit Centre A / Profit Centre B and Product 1 / Product 2, the total median value for "Sales" will be calculated over 4 data cells that have both Profit Centre and Product specified
Formula: uses the variable's formula at the 'aggregate' level, to calculate an aggregation value.
A good example of this is to think of gross margin across two different centres:
Profit Centre A - $100K in Net Sales, $90K in Costs - Gross margin is 10%
Profit Centre B - $1M in Net sales, $500k in Costs - Gross margin is 50%
By default, Causal will display the Gross Margin variable (Net Sales-Revenue)/Net Sales as a “Sum” - so we would see 60% as our top-level Gross Margin.
A common way around this is to use the “Average” category aggregation, which would display a 30% Gross Margin. This value is also incorrect, as Profit Centre B has much higher net-sales, and its impact to margin should be greater than Profit Centre A.
Logically - the way to do this calculation appropriately would be to add the sales and costs of both profit centres, and then apply the calculation. This is what the “Formula” category aggregation helps us achieve, giving us the correct value of 46.4%