Mr Wong wants to get fit and decides to start running five kilometers a day, but after his first two kilometers, he is exhausted. Clearly the goal was too ambitious. What does he do now?

The field of goal setting is of interest to consumer researchers who want to know not only how goals are set but how consumers set about striving towards them. To date, the research has focused almost entirely on a scenario where the consumer has a single goal and either finds a way to achieve it or gives up.

But what if the goal is not a static thing? What if it changes over time based on performance? What if, for instance, Mr Wong revised his goal downwards to four kilometers a day (or, if he ran better than expected, revised it upward to six kilometers)?

These questions are posed by Chen Wang and Anirban Mukhopadhyay who devise a model that can account for just such a scenario.

"Other research has pointed out that as people accumulate experience in a domain, they frequently adjust the expected pacing of their progress. Our research suggests they may also dynamically adjust their goal targets," they say.

They draw from an existing concept - the "Test-Operate-Test-Exit" model - which has been used to describe how people go about achieving their goals. In Mr Wong's case, this model would allow for him to adjust the pace towards his goal, but not the goal itself.

The authors adapt this to take into account goal adjustment and call their model "Test-Operate-Test-Adjust-Loop (i.e., TOTAL)". It is based on four principles that govern how individuals update their goals over time.

One principle is monotonicity, in which people will revise their goals in proportion to the discrepancy between goal and performance so the bigger the discrepancy, the bigger the goal revision. So in Mr Wong's case, the worse his performance the lower the revised target, or the better the performance the higher the target.

Second is diminishing sensitivity, which is about the limits of goal revision. Large discrepancies will have relatively smaller impacts on goal revision than ones that are smaller in magnitude. Let's say Mr Wong was also trying to save money. If he aimed to save $110, he would be less sensitive to having saved $120 than if the goal was $10 and he had saved $20.

The third and fourth principles are related to whether the motivation behind the goal is intrinsic, meaning it is pursued for inherent satisfaction, or extrinsic, meaning it is pursued for some external reward.

The authors argue intrinsically-motivated goals should lead to aspiration maximization - the principle that positive discrepancies will lead to significant upward revisions whereas negative ones will lead to relatively smaller downward revisions. "When people are intrinsically motivated, they are averse to lowering their standards even if they have failed in the past."

Extrinsically motivated goals, on the other hand, do not have the same "sticky" quality. People are much more comfortable revising them downwards if they fail, or to stay with the same goal if they succeed.

"These goals are for the sake of external awards as opposed to inherent satisfaction. As a result if the incentives are linked to goal attainment, people may not desire to improve on their performance when progress is going well. This essentially implies downgrading the target to one that is expected to be comfortably achieved. As a result negative discrepancies should see significant downward adjustments in response."

The authors carried out four experiments that provided initial confirmation for the predictions of their model, which incorporates elements from consumer behavior, psychology, engineering and mathematical modeling.

"We believe this is a complete model of goal directed behavior that encapsulates pursuit, achievement, failure and abandonment. It advances our understanding of people's goal-directed behaviors by being the first to present a comprehensive framework for self-regulation over multiple periods," they say.