ASU Learning Sparks

Abductive Reasoning in Problem Solving

What is abductive reasoning and how is it helpful? While deductive and inductive reasoning are useful for certain questions, abductive reasoning, commonly used in the design process, allows for making the best guess with available information. Design problems often require gathering insights rather than deriving absolute truths, and a small number of user testers can provide ...

What is abductive reasoning and how is it helpful? While deductive and inductive reasoning are useful for certain questions, abductive reasoning, commonly used in the design process, allows for making the best guess with available information. Design problems often require gathering insights rather than deriving absolute truths, and a small number of user testers can provide sufficient information to proceed in a meaningful and cost-effective direction.

How does a design approach help individuals and organizations thrive in the face of uncertainty and scarce information?

It helps to understand what kinds of reasoning are well suited to tackle what kinds of questions. From science and math you might be familiar with deductive and inductive reasoning, but design turns to abductive reasoning to solve some of its trickiest questions. 

So what are deductive, inductive and abductive reasoning? Let’s look an example question to illustrate:

For instance let’s say you wanted to answer:

“What proportion of a carbon dioxide molecule is carbon by mass?”

A deductive approach would be to start with the understanding that carbon dioxide is one carbon and two oxygens. I know the mass of each of those kinds of atoms so I can divide the mass of carbon by the mass of one carbon + two oxygens to get the percentage of CO2 that is carbon by weight.

An inductive approach might have you take several samples of carbon dioxide, experimentally separate and directly measure their constituent elements and then average the outcome of your experiments to come to a conclusion.

Deductive reasoning is top-down: take the general principles and apply them to the particular case. Inductive is bottom-up: take all of the particular cases and use them to form a generalization. 

These are both great approaches to a question like this one and characterize reasoning processes in various research activities of the physical sciences. 

But what if you had a question that escaped this kind of reasoning? Imagine you are a doctor with a patient with a complex pattern of symptoms for which all of the best fitting diagnoses do not have definitive testing. Or you are a small business owner who has to decide whether to hire more employees in economically uncertain times and no way to know the future.

For situations like these you might be better served by what is called abductive reasoning, which is a mainstay of design process.

In abductive reasoning you make your best guess with available information. 

For example in design you might face the challenge of testing the usability of the website or product you designed. There is no existing data or generalized theories on the performance of the particular thing you just invented, because it did not exist until you invented it so deductive reasoning is limited in its usefulness. 

And if your purpose was to derive truth statement insights about the nature of human condition or create an absolutely optimized design solution through an inductive, statistical analysis with a sample size that allowed your claims to be statistically significant you would need to spend a fortune on recruiting participants. But your purpose is not to make truth claims, it is to make a satisfactory product that can incorporate as much insight as possible. 

For many design problems the marginal returns to insight on each additional usability testing participant is diminishing, so often a surprisingly small number of user testers can give you enough information to proceed in a meaningful, beneficial, and economical direction.