Note: this page is currently under construction. Please do not cite.
Self-efficacy is a person's perception (whether accurate or not) of their own capability to attain certain goals or complete a certain task (Schunk, Pintrich, & Meece, 2008; p. 8). Although there is some debate about how specific self-efficacy beliefs tend to be, they are often taken to relate to relatively specific (and therefore possibly changing) abilities. Not so much the general belief that I am competent in algebra, but that I am competent in a specific (type of) algebraic task (e.g., solving linear equations or even linear equations of a certain type). [VA: This latter part needs to be checked. There is some literature about how specific these beliefs tend to be, I think.] Also, it might be better to stick to established definitions, though the problem is that there isn't a single authoritative one.]
Although self-efficacy is an inherently subjective notion, in many academic learning situations people's "subjective" self-efficacy beliefs may be heavily influenced by "objective" measurements of their competence, such as for example their results on tests and quizzes.
Self-efficacy (and similar constructs) are an important construct in multiple theories of motivation (Bandura's social-cognitive theory of motivation, Eccles and Wigfield's expectancy-value theories, attribution theory (Weiner). Dweck? Perhaps also in theories of self-regulated learning? It is related to but different from constructs such as feeling of knowing (FOK), self-concept, self-competence, and self-esteem. (Need to point out how these concepts are different. And whether each deserves its own place on the PSLC wiki.) (Is learned helplessness a related construct? Has it been interpreted as a lack of self-efficacy?)
Question: have researchers distinguished between beliefs about one's ability to _perform_ in a certain domain and one's ability to _learn_ in a certain domain? One probably expects these to be highly correlated and perhaps often not clearly distinguished in people's minds ... has that distinction/correlation been studied? Would it be interesting to study? Would it be a PSLCish thing to study?
Measurement: self-efficacy is typically assessed by means of questionnaires (MSLQ?) and is often related to a specific task (e.g., a subject is presented with a specific task instances and asked "how well do you think you will do on this task?". (Need to check this and give real not made-up examples.) Teacher assessment is sometimes used.
Sources of self-efficacy beliefs Some studies find (I believe) that particular interventions enhance self-efficacy and learning (Schunk & Ertmer, 2000). Are there findings that show that people's attributions of success and failure in learning affect their self-efficacy?
Effects of self-efficacy beliefs Many studies find that self-efficacy is highly correlated with learning (or performance???). Some studies find that self-efficacy is a better predictor of learning (future academic achievement - probably fair to view that as learning???) than prior knowledge.
The often-observed correlations between self-efficacy and learning (learning??? or performance???) could simply reflect the fact that people are (somewhat) accurate in their assessment of their own competence, although there is probably more to it than that. (E.g. I personally find it rather likely that self-efficacy would lead to greater persistence. Conversely, learned helplessness - if we may interpret it for now as a manifestation of low self-efficacy - would lead one to not try or give up easily.) But a causal link between self-efficacy and learning is hard to establish, because it is different to manipulate self-efficacy without also manipulating other factors of the learning environment (but see Schunk & Ertmer (2000); also, perhaps attribution re-training is an effective way of fairly directly "manipulating" self-efficacy).
Accuracy of self-efficacy beliefs. (Need to cite some research about how accurate people's self-efficacy beliefs tend to be.) It is generally believed among motivation researchers that underestimating one's capabilities tends to be worse - from a viewpoint of future academic achievement (??? check!) - than overestimating one's capabilities.
Recently, researchers have started to look at ways of automatically measuring self-efficacy within computer-based learning environments. For example, Lester, McQuiggan (sp?), Boyer at NC State [REFS] have started to create machine-learned detectors that (unobtrusively and automatically, in real time) detect behaviors that reflect high/low self-efficacy within a serious game. These types of detectors open up novel opportunities for learning sciences research. It is still an open question to what degree such detectors can be used to enhance the effectiveness of the learning environment (for example, by helping the learner develop greater self-efficacy, or by adapting in other ways to learners' self-efficacy). Further validation efforts may be needed to ensure that these types of automated detectors honor the original notion of self-efficacy.
Schunk D., & Ertmer P. (2000) Self-regulation and academic learning: Self-efficacy enhancing interventions., In M. Boekaerts, P. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 631-649). San Diego: Academic Press.
Schunk, D. H., Pintrich P. R., & Meece, J. L. (2008). Motivation in education: Theory, research, and applications (3rd Ed.). Upper Saddle River, NJ: Pearson Prentice Hall.