Where does mental model collide?

In the theory of reasoning that Philip Johnson-Laird set out, a mental model is a small internal representation of a situation that a person builds in order to draw a conclusion. On this account people do not reason by applying formal rules of logic to the shape of sentences; they imagine the possibilities that a set of premises allows, read off what holds across those possibilities, and look for cases that would overturn a tentative conclusion. Johnson-Laird's 1983 Mental Models laid out the framework, and his 2010 review Mental models and human reasoning gathered the evidence, including the systematic errors that appear when a task forces people to hold many possibilities in mind at once. The idea traces back to Kenneth Craik's 1943 The Nature of Explanation, which proposed that the mind carries a small-scale model of reality to predict events. Here the referent is a transient structure built to settle a particular inference, not a user's picture of a machine or a manager's standing assumptions.[2][3][1]

Unit

A representation of the possibilities that a set of premises allows

Recovery means

Not applicable

Measured by

Inferred from which conclusions people draw, the errors they make, and how difficulty rises with the number of models a problem requires

In human factors and interaction design, a mental model is the picture a user forms of how a device or system works, which they use to predict what it will do and to plan their actions. Donald Norman's 1983 chapter Some observations on mental models, in Gentner and Stevens's Mental Models, separated the user's model from the designer's model and from the system's actual workings, and noted that trouble arises when these drift apart. His The Design of Everyday Things named the gulf of execution, the gap between what a person wants to do and the actions a device allows, and the gulf of evaluation, the gap between the device's state and what the person can perceive of it. The design aim is to shape a system, through its visible controls and feedback, so that the model users infer matches how the thing actually behaves. The referent is a person's working understanding of a particular artifact, judged by whether it lets them use the thing well, not a structure built to settle a logical inference.[4][5]

Unit

A user's working understanding of a particular device or system

Recovery means

Not applicable

Measured by

Probed by asking users to predict, explain, or troubleshoot a device, and by observing where their expectations and its behaviour diverge

In systems thinking and organisational learning, a mental model is the set of deeply held assumptions, generalisations, and images that shape how a person understands their world and decides how to act. Peter Senge's 1990 The Fifth Discipline made working with mental models one of the five disciplines of a learning organisation, alongside systems thinking, personal mastery, shared vision, and team learning. The claim is that managers often act on outdated or untested pictures of their market and their organisation without noticing, so the discipline is to surface these assumptions, hold them up for examination, and test them against evidence. The emphasis falls on tacit beliefs that guide behaviour and frequently go unstated, rather than on a representation a person deliberately builds for a single problem. The referent is a standing worldview held by a person or shared across a group, and the practical aim is to make it visible enough to question.[6]

Unit

A person's or group's standing assumptions about how their world works

Recovery means

Not applicable

Measured by

Drawn out through dialogue, reflection, and exercises such as the ladder of inference rather than through a formal scale