top of page

Ideas & Design Filtration - Brief guide to research methods

When you are working with multiple ideas or generating multiple ideas to resolve a problem, there are always constraints under which you intend to work or finalize the most relevant one. It's always a trouble to pick up the best from the best.

I myself was involved in a few projects, where I needed to pick up from multiple sources relevant ideas, process improvement, and techniques to enhance the efficiency of the process or boil down to the most relevant one.

Initially, it was a bit difficult and overwhelming to work with so many parameters, dependent or independent of each other. To make my life simpler and have a logical conclusion, I did research and came up with a few methods from various domains to help in the ideas selection and filtration.


Below is a brief on methods that I find pretty handy and helpful when working with multiple ideas and parameters.


Matrix Diagrams and Relationship*


It's a powerful visualization tool to show relationships between 2, 3, or 4 groups of information. They allow the clear visual display of complex data in multiple dimensions and this makes them very suitable for presenting information to decision-makers.

The below stated all the matrix serve the same purpose i.e. to compare items on the basis of specific criteria. The only difference is in the number of groups being compared and the type of relationship you are looking at.


Roof Matrix

Used to compare the effects of elements of a single set of data on the other elements of the same set, i.e. in the example diagram below;



A1 has +ve effect on A2 & A5 but A3 has -ve effect on A4 & A5.


L - Matrix

The is used to compare two sets of result (A and B) to each other or to compare 1 set of results to itself (A to A). The data can be anything from non-conformities to performance-related.

In the example A1 has a weak effect on B1 but A3 has a strong effect on B2.


T - Matrix

This is used to compare two sets of results to a third set, i.e. in the example, B and C can be compared to A but B and C cannot be compared.


This example shows the use of symbols instead of numbers and shows that A1 has a strong relationship to B2.


Y - Matrix

This is used to compare three sets of data to each other (but not simultaneously) i.e. in the example A can be compared to B or C.





C - Matrix

This is for full simultaneous comparison of 3 sets of data i.e. it is possible to see the variation in A simultaneously with variation in B and C. This is a 3D matrix and can be complex to use and see.




X - Matrix

This used to compare 1 sets of data i.e. in the example A an be compared to D & C and B can also be compared to D & C but A cannot be compated to B and D cannot be compare to C.


Other Methods**


C - Box


Datum Method


Harris Profile



Itemised response and PMI


vALUe












Weighted Objectives Method



*Tools for quality management and improvement

**Delft design guide


Hope this can be useful for the reader.


For further reading, and developing deeper insights into evaluating ideas and processes, I would suggest going through the concept of HOQ - House of Quality.


bottom of page