Task representation: meaning and application

Someone considering solar panels sees mainly 'such a hassle' and puts the brochure aside, whether or not that picture is accurate. This article explains what a task representation is: the inner picture of what a task involves, costs and delivers. You read how to research it and how communication can adjust an incomplete or overly heavy picture.

A task representation is the mental picture a person has of a task or action: what it involves, what it costs and what it delivers.

It is a specific form of mental model: a simplified inner representation of how something works. When someone has to start doing something, such as adopting a product, applying for a service or taking on an additional role, that picture can carry weight. A task representation cannot be observed directly. You study the description someone gives of it, which is neither behaviour nor brain activity.

Mental models: from Craik to models of systems

The idea that people think with inner models is usually traced to the British psychologist Kenneth Craik. In The Nature of Explanation he argued that an organism carrying a 'small-scale model' of external reality and of its own possible actions can try out alternatives, choose the best one and react to situations before they arise (Craik, 1943). A task representation does exactly that: it runs through an action in advance.

In Mental Models, Philip Johnson-Laird turned the idea into a theory of comprehension and reasoning (Johnson-Laird, 1983). He later summarised that a mental model represents a possibility and that its structure corresponds to what it represents. He also noted a limit: in fast, intuitive reasoning people often build one model and do not spontaneously seek alternatives (Johnson-Laird, 2010).

Also in 1983, Gentner and Stevens published the volume Mental Models, including chapters on mental models of electricity and of interactive devices and a chapter by Donald Norman (Norman, 1983). An interdisciplinary review describes mental models as personal, internal representations of external reality used to anticipate events, reason and decide. They are incomplete and change through learning (Jones et al., 2011).

Task representation in cognitive control

In laboratory research on task switching the meaning is narrower. There, a task set is the organisation of cognitive processes and mental representations that enables a person to perform a task as instructed, including relevant stimuli and responses and which response belongs to which stimulus. Switching tasks costs time and produces more errors than repeating one (Kiesel et al., 2010). It says nothing about how people evaluate a service.

Simulation and planning

A third line studies how people mentally rehearse future tasks. Buehler, Griffin and Ross described the planning fallacy: expecting your own project to go as planned, although most similar projects ran late. In their first study, 37 students predicted on average 33.9 days to finish their thesis; it took 55.5. In a think-aloud study, 74 per cent of thoughts concerned future scenarios and 7 per cent relevant past experience (Buehler, Griffin & Ross, 1994). Such scenarios tend to be idealised, schematic and simplified (Wiese, Buehler & Griffin, 2016).

Taylor and colleagues distinguished process simulation (imagining the steps towards a goal) from outcome simulation (imagining the result) (Taylor et al., 1998). Students who mentally rehearsed a school project step by step more often finished it when they had said they would than those imagining the outcome. Taylor notes that process simulation does not always work better (Taylor, 2011). With advertisements, Escalas and Luce found that focusing on the process of product use increased the influence of strong arguments on behavioural intentions under low to moderate involvement; under high involvement this reversed (Escalas & Luce, 2004).

Neurofactor uses a working definition with three components. It is an aid, not a validated measure.

• Content: the steps someone expects, their order, who is involved and what knowledge or resources are needed.
• Costs: what the task seems to demand in time, effort, money, attention and risk.
• Benefit: what the task seems to deliver, for whom and when.

The division fits research on technology acceptance. Davis found that perceived usefulness and perceived ease of use were related to the use of a system; usefulness was more strongly related to use, and ease of use may have mattered mainly through usefulness (Davis, 1989). These were correlations among 152 users, not proof that task pictures cause behaviour.

Task pictures are often incomplete: steps are missing. They can also be skewed: too light, as in the planning fallacy, or too heavy. That people can overestimate an unfamiliar task is a working hypothesis here; the cited planning studies concern underestimation.

Anyone installing an app, applying for insurance or taking on an additional role also assesses the path to the result. Three situations are common.

1. The picture is incomplete. Someone sees only the first step and misses what follows. That can mean a quick yes and dropping out later.
2. The picture is heavier than reality. Someone counts on steps the provider already handles. The task then becomes a barrier, even when the result is attractive.
3. Costs and benefit are out of balance. Someone knows the steps but does not see what the effort yields, or only much later.

Whether a group is ready to act is a related question; see Readiness segmentation.

1. Define task and decision context. Name the action ("requesting a quote", not "going green") and the group.
2. Start with a free description. "What do you think it all involves?" Note steps, people involved, expected time and benefit in the participant's words.
3. Collect associations. They show the charge of the picture, such as "hassle" or "security"; see How do you measure associations?.
4. Have people think aloud at a concrete step, such as an application page. Concurrent thinking aloud reflects what someone is attending to at that moment; asking for information people would not normally attend to can change the thought process (Ericsson & Simon, 1980).
5. Have people create or sort a step list. This reveals gaps and is a form of direct elicitation of a mental model (Jones et al., 2011). Breaking a task down can itself influence the estimate; the findings on this are mixed (Wiese et al., 2016).
6. Compare the picture with the actual route. The gaps are your starting points.
7. Test behaviour separately. Only behavioural data show whether people apply or finish more often.

These are editorial recommendations to test in each situation.

• Complete an incomplete picture. Show the whole route and be honest about possible disappointments.
• Lighten an overly heavy picture, if accurate. Show which steps the provider takes over and what people do themselves. An inaccurate lighter picture damages trust.
• Show the process, not only the outcome. Research on mental simulation suggests process attention helps in some situations, not all.
• Make the benefit concrete. If costs come now and benefit later, say when people will notice something.

This example is fictional. An installer sees that many homeowners request solar panel information, but few request a quote.

• Decision question: what picture do homeowners have of "having solar panels installed", and is there a barrier in it?
• Available information: website data on where people drop out; nothing yet on their task picture.
• Appropriate approach: interviews with homeowners who requested information but no quote: a free description, associations, a self-made step list and finally thinking aloud on the quote page. The expected route is compared with the actual one. There are no results; this is a research plan.
• Possible interpretation: if participants mention steps the company already handles, the picture is heavier than necessary. If steps causing later delays are missing, it is incomplete. Each needs a different message.
• Next step: compare a page showing the full route with the current one, using outcome measures set in advance.

• Equating the picture with the description. A mental model is not directly observable. What someone says may differ from what guides their actions: an espoused theory versus a theory in use (Jones et al., 2011).
• Equating the picture with behaviour or brain activity. Someone with a heavy picture may still do the task. A brain measurement shows no task picture.
• Assuming a single picture. Compare groups, such as experienced users and newcomers, before generalising.
• Asking leading questions. How you ask shapes what people mention. Ask open questions before presenting steps.
• Translating lab findings directly. Task-switching and planning studies often used students and specific tasks. They provide a framework, not a prediction for your target group.

A task representation is the picture someone has in advance of what a task involves, costs and delivers. It is a concrete form of mental model, rooted in research on mental models, cognitive control and planning. When people have to start doing something, it helps you examine where the picture is incomplete or too heavy and tailor communication accordingly. Treat the picture as a hypothesis about the choice, not a measure of behaviour.