Data Analysis: X-Y Matrix
Christopher D. Cole
Vice President & Lead Product Manager | Designer | MIT-Trained in AI Product Design at Wells Fargo
The X-Y matrix is used to analyze and score all process inputs as related to events that occur during a product or service lifecycle. This usually takes place during the analysis of root causes. At first, when the Ishihara (Fishbone) diagram is produced, all possible causes (X) are listed after meeting with subject matter experts to decompose them from high to low level inputs. The possible causes (X) are not 100% vetted as the #1 problem, but could be one of many possibilities related to the negative effects (Y).
In the image above by www.six-sigma-material.com, the elements needed to create a X-Y matrix is shown. As the Ishihara diagram captures inputs (X), that relates across to the process map, and then to the voice of the customer. The X-Y matrix can be filled out with all this information to be used for further analysis. The employees most knowledgeable to the process provide further guidance on how much weigh each input has against the effect (Y). Here is an example below of the X-Y matrix:
Based on values deemed most critical to customers, its determined they care most about the text highlighted in red. After meeting with subject matter experts, the inputs most likely connected to the areas in red are highlighted in green. Each input is given a weighted score, based on their strength to the Y’s (red text). As shown, for example, “Good quality ink” has a strong connection to "Clearly readable print" and "Good quality photo", but not "Harmless to health", while "Paper quality" has a weak relationship to "Clearly readable print" "Good quality photo" and both are scored a “1”. The score of “3” is also weighted as a weak relationship as well. Each input score (X) is multiplied by the weighted (Y) score, for example, [Good quality ink (15 x 9 + 10 x 9 + 10 + 1) = 235]
Key Takeaways: Data must be separated into what is critical and/or not critical to the effects. In the end, the voice of the customers displays what inputs are deemed necessary for successful products and services. It's an important tool to have when faced with abundant amounts of process information, that when combined with additional analytics provide a great objective way of identifying the most important root cause.