An abstract model for a learning strategy
I model a learning strategy as a coordinated set of events with two features. The first feature is that a strategy includes at least two tactics. A tactic is an operation a learner applies if particular conditions arise. It can be abstractly represented as what computer scientists call a production: lF[condition(s)]-THEN[operation(s)]. I point out now and elaborate shortly that every operation generates products which cause the state of the task to be updated.
The second feature of my model for a strategy is the presence of one or more checkpoints between conditions at the start of the task and those at the final state of the task. Checkpoints are spots along the timeline of work on a multistep task where a learner has an opportunity to monitor the state of the task and proceed as planned, per the strategy, or adapt. States of tasks change because operations introduce new products. Those products contribute to new conditions setting the stage for the next If-Then production in the strategy. Checkpoints afford options about what to do next. They can be represented as elaborating the earlier and simpler If-Then model to take on the form lF[condi-tion(s)]-THEN[operation(s)]-ELSE[different operation(s) if conditions are different].
How does a learner decide which learning strategy to use?
The If-Then-Else model of a learning strategy begs an important question: how does a learner decide what to do when a tactic that would usually be enacted at a checkpoint in a task is judged inappropriate or is unavailable (e.g., forgotten, precluded by a condition in the environment)? Newton (this volume) references Star, Caronongan et al. (2015, p. 26) who characterize a model like mine as having a “rationale behind the use and effectiveness of these steps.” What makes up that kind of rationale?
One element is knowledge about the domain in which a tactic operates. For example, as Newton describes, a learner may know several methods for working with fractions when manipulating arithmetic expressions. How is one chosen over another? I model learners’ decision making using five features that blend metacognitive knowledge, motivation, and decision making to form a judgment about utility.
Throughout the timeline of work on a task, learners choose among tactics and learning strategies in the context of current and successively updated conditions. I posit these checkpoints are viewed in light of a personal history. That history includes four fundamental categories of information.
The first category is what the learner predicts will result if a particular learning strategy is applied. Products generated when a learner carries out a learning strategy are not limited to the space of the task’s disciplinary domain, e.g., a reduced form of a complex fraction or an inference about the main idea of a text. Carrying out a learning strategy also generates perceptions that are personal. Effort, pace, and affect are examples of personally relevant kinds of information inherently associated with generating a product. As well, social status may be a factor if the learning strategy and its product can be observed by peers or a teacher. Bandura (1977) labeled this kind of information - what a learner predicts will be the result of a tactic or learning strategy - an outcome expectation.
A second category of information enfolded in decisions about using or adapting a learning strategy is the learner’s judgment of skill to enact it. The probability an expected product actually will materialize when a learning strategy is applied is inversely proportional to the learner’s skill to enact that strategy. This kind of information was labeled an efficacy expectation by Bandura (1977).
A third kind of information I suggest a learner examines is the incentive or “payoff” (see Graham & Weiner, 2012) associated with the predicted product generated by enacting a particular learning strategy. Incentives are reasons to behave, including applying a learning strategy. They can be social (Baars et al., this volume) as well as something “interior” to the learner, what the learner prefers (Karabenick & Col-lins-Eaglin, 1997) or, in the case of a disincentive, what the learner doesn’t like (Bartels, Magun-Jackson, & Kemp, 2009).
The fourth category of information in my list of elements comprising the learners personal history is an explanation about why a learning strategy succeeds, falters, or fails. According to attribution theory (Weiner, 2010), different attributions, e.g., to luck or to effort, automatically give rise to affect, e.g., worry when a learning strategy seems to have generated the right product by luck, or pride if the learner judges hard work (effort) made a key contribution to generating the right product. Memories about associations between attributions and affect bear on choices about learning strategies in the moment. Procrastination is a good example.
Altogether, these four categories of information - outcome expectations, efficacy expectations, incentives associated with particular products, and attributions explaining success and failure - are input to an unknown calculus the learner uses to determine the utility of a particular learning strategy in particular conditions. The strategy with the greatest marginal utility “wins.”
Isolating the first letter of each category of information, including the utility a learner assigns to each strategy, forms an easy to recall mnemonic: AEIOU. Like vowels that add warmth to cold consonant sounds in speech, attributions, efficacy expectations, incentives, outcome expectations, and utilities related to learning strategies imbue cold decision making with warmth when learners choose a strategy to use (Winne & Mar-zouk, 2019).
Decision-making learners carry out to choose learning strategies is very little addressed in chapters within this section. This is because almost all the primary research available to be reviewed for these chapters investigated effects of only a single learning strategy. To be sure, it is costly and sometimes practically difficult to secure large samples willing to participate in statistically powerful research that examines multiple strategies. Notwithstanding, life in classrooms is more complex than in the lab or in field studies where just one strategy was a focus for research. Learners fortunate enough to know multiple strategies must choose which one to use. Future research should investigate how learners’ decisions unfold and how learners can be supported to evolve more productive decision making about learning strategies. Such research is essential to advancing comprehensive theories and fruitful applications of self-regulated learning.