In a previous blog, I suggested three charts to analyze a conjoint analysis, with my last suggestion being a radar chart. Knowing that it was a simple chart to build in excel, I thought it must not be very difficult in Tableau. Unsurprisingly, there is content out there showing different approaches to building a Radar Chart, including the Information Lab and Tableau, but here is how I implemented these into a Conjoint Analysis.

 

Step 1: Understand your purpose

As mentioned in my previous blog, Radar Charts are great for showing rating scores for different categories. For me, I wanted to know at a quick glance which measures had the biggest magnitude at a glance.

 

Step 2: Understand your measure

In my instance, I will be using utility score as a measure, and while the magnitude of this score is important, it is equally important knowing whether it is a positive or negative value. So my solution to this was to color based on the direction (see Step 7) and calculate an absolute value to get it looking like a conventional radar chart.

 

Step 3: Create an “Angle” Equation

Essentially a radar chart is a polygon revolving around a circle. So your first step is to distinguish exactly this, and the equation that you will use is:

RUNNING_SUM(

(2*PI()) / MIN({COUNTD([Attribute])})) + (PI()/2

)

Where [Attribute] is the labels you want on the points of the radar. You notice that RUNNING_SUM is a quick table calculation, so specify that you want to compute using your Attribute variable.

 

Step 4: Calculate the distance from Zero

You next need to create the calculation: AVG([Attribute]) in order to create the size of the bars.

 

Step 5: Create Spatial X and Y Points

In Step 3 you create a direction from the origin, and Step 4 you create a distance from the origin, so now by multiplying these both together, you will get the points in the right position. The calculation then is:

X = [Distance from center] * COS([Angle])

Y = [Distance from center] * SIN([Angle])

 

Step 6: Dual Axis Dots

Personally, it looks nice to include circles at the end of the radar to show that they are points, so I created a dual-axis. Remember to synchronize and change the marks to circle/shape. You can also do your circle formatting in this step to make it look visually appealing

 

Step 7: Create A Color Equation (Optional)

As alluded to in the intro, it was not good enough for me to know the magnitude, I wanted to know the direction visually too. So, I create this simple equation:

If [Measure] < 0 then “Negative” Else “Positive” END

When you drag this on color for your circle dual-axis, you will notice that now you can quite simply see both the direction and the magnitude

.