Abstract Reasoning

Glossary series

This is part of the Test Partnership glossary series for everyday learning.

PAGE CONTENTS

What is abstract reasoning?
Other videos

section one

What is abstract reasoning?

Abstract reasoning represents a person’s ability to solve problems, identify patterns, and work with logical systems. It refers to the capacity to think conceptually, grasp complex ideas, and identify patterns and relationships without relying on concrete, specific examples. Abstract reasoning is a major component of general cognitive ability, and is closely aligned to the G-factor in cognitive ability testing.

0:53 Lead consultant, Ben Schwencke, explains what is abstract reasoning.

Also known as inductive, abstract thinking, logical, or non-verbal reasoning, the format and layout of abstract reasoning tests will differ between publishers, they all aspire to be language free, relying on images, patterns, and diagrams instead of text. They also aim to minimise any numerical reasoning component in their assessments, retaining a diagrammatic format and layout.

In the academic literature, abstract reasoning is closely aligned to “Fluid Intelligence”, showing less association with “Crystallised Intelligence”. Whereas crystallised intelligence relates to learned knowledge and experience, abstract reasoning tests are more pure fluid intelligence tasks. Fluid intelligence pertains to a person’s problem-solving ability based solely on logic, rather than any prior knowledge.

Because abstract reasoning involves no verbal or numerical component, abstract reasoning is a particularly pure measure of fluid intelligence.

Often, stand-alone measures of fluid intelligence will comprise solely abstract reasoning tasks, ensuring they are both language free and culture fair.

Unfortunately, there seems to be little consensus in the commercial space regarding nomenclature for these types of assessments. Assessments which measure abstract reasoning are referred to as inductive, logical, diagrammatic, or non-verbal reasoning tests, but they all fundamentally measure the same psychological construct. Therefore, employers can treat these assessments as roughly equivalent from a construct validity perspective, smoothing the transition from one publisher to another.