Expectancy-value theory holds that motivation for an activity rests on two questions a learner answers, usually without noticing: Can I do this? and Do I want to? The first is expectancy of success, the learner's belief about how well they will do; the second is subjective task value, how much the task is worth to them. Effort, persistence, and choice follow from the combination, and crucially from a combination, not a sum: a task a learner is sure they could ace but cares nothing about gets neglected just as surely as one they value but expect to fail. Jacquelynne Eccles and Allan Wigfield built and refined this model over four decades into the leading account of why students choose the courses and tasks they do.
Expectancy answers the competence question and behaves much like a close relative of self-efficacy: a forecast of success on the task ahead. The richer machinery sits in value, which Eccles and Wigfield split into four parts.
Attainment value is how central the task is to the learner's identity, doing well at it because it matters to who they are. Intrinsic value is the straightforward enjoyment of the activity. Utility value is usefulness toward some other goal, the way a learner studies English not for its own sake but because it opens a job or a degree. Against these sits cost, the negative term: what taking on the task drains away. Cost itself has three faces, namely effort cost (how much work it demands and whether that is worth it), emotional or psychological cost (the anxiety and fear of failure it provokes), and opportunity cost (the other valued things the learner gives up by spending time here).
In 2020 Eccles and Wigfield renamed the framework situated expectancy-value theory (SEVT) to correct a misreading the old name invited. Expectancies and values are not stable traits a learner carries around; they form in the moment, shaped by the specific situation, the immediate social context, and a long developmental and cultural history of prior experiences. The same student can hold high utility value for English in a job interview and low value for the same English in a grammar drill, because the situation supplying the value has changed. The relabelling also folded cost firmly inside the model, treating subjective task value as a net figure that weighs the benefits of an option against its costs.
The model maps neatly onto why learners pick up, stick with, or drop a language. Utility value carries much of the weight in instrumental cases, where English is a means to migration, employment, or study, which connects EVT to the instrumental and ideal-self currents in L2 Motivational Self System thinking about who a learner imagines becoming. The cost term explains drop-off that pure motivation accounts miss: a learner may value English highly yet quit a course when the emotional cost of speaking anxiety or the opportunity cost against work and family grows too steep. Because SEVT treats value as situated, it predicts that the same learner's commitment will swing across tasks and settings, which matches the volatility teachers actually see. The expectancy side links outward to Attribution Theory, since how a learner explained past success or failure feeds directly into what they now expect.
For most of the theory's life the cost component was underexplored, the "forgotten" term that the value half of the model named but rarely measured; Flake and colleagues (2015) made the case that cost deserved its own validated instrument, and serious empirical work on it is recent. That late start left measurement unsettled, with an open debate over whether cost belongs inside subjective task value or stands as a separate construct alongside it, a disagreement that changes how studies model the data and compare findings. Expectancy and value are also strongly domain-specific, so a learner's profile in mathematics tells you little about their profile in a language, which complicates any general claim and demands that measures be tied to a particular subject. Finally, cultural generalisability is a live question: because expectancy and value beliefs are shaped by culture and situation, the value hierarchies the model documents in one setting cannot be assumed to transfer, a caution the situated reframing acknowledges more than it resolves.
Both levers must be worked, since either alone stalls. Raising expectancy means structuring tasks so success is genuinely reachable and visible, building the well-founded belief "I can do this." Raising value is more than making lessons fun: utility-value interventions that ask learners to articulate, in their own words, why the material matters to their goals have moved engagement, and connecting language work to a learner's real purposes does the same. The cost term is the neglected lever worth pulling deliberately, because lowering the felt cost (easing speaking anxiety, trimming busywork, protecting against the sense that the time could be better spent) can lift motivation as much as adding value does.