D&M Chapter 9

Theme: studies of metacognition and student education

General Models of Student Self-Regulated Learning

Non-metacognitive influences: motivation, goal setting, goal orientation
Metacognitive influences: self-reflection, control of learning

Winne & Hadwin's Model

• emphasizes monitoring-and-control processes; self-regulatory behavior
• Four stages
  • task definition
  • goal setting and planning
  • enactment
  • adaptation
• Inside each stage: COPES
  • Conditions (in which studying occurs)
  • Operations (using strategies)
  • Products (making something)
  • Evaluate (what you produced)
  • Standards (compare to your product)

Studies related to metacognition
Cognitive research -- tends to be laboratory based and focus on minute analysis of underlying mechanisms

Educational research -- tends to be conducted in classrooms using representative materials; investigates relationships between metacognition and student achievement.

Self-Efficacy

Definition: people's beliefs about their ability to successfully complete a given task.

• People who believe they can succeed set higher goals
• select strategies based on effectiveness -- not difficulty or time

Study: Chemers, Hu, Garcia.
Student questionnaire on self-efficacy. Higher self-belief correlates with higher performance (GPA).

Overall findings suggest improving self-efficacy boosts performance by improving self-regulation in learning. People who believe they can achieve will set more specific and more demanding goals, and use more effective strategies.

Quiz
• I know how to schedule my time
• I know how to take notes
• I know how to study to perform well on tests
• I am good at research and writing papers
• I am a very good student
• I usually do well in school and at academic tasks
• I find my university academic work interesting and absorbing
• I am very capable of succeeding at the university

Metacognitive Monitoring

"Monitoring and control are central to effective self-regulated learning and performance."
Overconfidence --> understudying, use of ineffective strategies, underachievement
Calibration of monitoring is related to effective self-regulated learning.
Observation of 'unskilled but unaware' -- tendency for students to be fairly confident, but in some that's unwarranted. Good performers got more accurate over the course of the semester, but poor performers did not recalibrate their judgments. Providing incentives for performance and judgment does seem to improve calibration.

Training accurate judgments --

• rated how well they understood
• described concept they had particular difficulty with
• answered practice questions
• judged their confidence

Results were compared against exams. People with calibration training performed the same on first exam as those who did not receive training. As the class progressed, judgment improved. Improvements in calibration were predictive of higher scores on the final test. More research needed into training methods, understand duration of impact, and transfer

Research - changing answers during a test.
Scores are mostly a result of preparation, intelligence, and motivation. Changing your mind is often beneficial -- trust your confidence on this. If you were guessing, go with the hunch/gut/initial reaction...but if you have a reason, go for it and make the change.

Metacognitive Control

Purpose: to allow people to control their cognitive processing. Ex: after an exam, using results to improve performance on next exam. Increase efforts, decrease efforts, etc.

Study -- Rocky's Boots. Problem-solving program with boolean logic. Question: will training students to monitor their progress and success during practice phases help them to better regulate their problem solving. Structure: some students get no training, one group gets problem-solving training, one group gets problem-solving training and cognitive monitoring prompts. Monitoring group outperformed the other two on difficult problems, and used less time. Thinking about thinking may have encouraged them to use more strategies -- monitoring improves control, which improves performance.

Domain Metacognition

Reading
Literacy is arguably the foundation of education. Major omissions in instructions experiment -- first graders didn't notice, third graders did. Comprehension monitoring -- internally enacting the instructions lets you evaluate whether your understanding allows you to obtain the goal of winning the game.

Comprehension monitoring includes detecting inconsistencies and omissions, evaluating understanding, and comparing to previous knowledge. Control activities might be re-study or using strategies.

Reading Model:

• to create an internal representation of the text, including the words/sentences and the meaning.
• to meet an evaluation standard (understand all, memorize crucial concepts, etc.).
• to use strategies to achieve these goals
  • prepare to read -- set goals, skim, etc.
  • construct meaning from the text -- identify main ideas, make inferences, interpret, monitor
  • reflect on reading -- ask questions, summarize

These phases are recursive and have no fixed order. Each may include monitoring and control activities.

Call this constructively responsive reading.

Ann Brown's work on Metacognition/Education
Gifted/slow was originally thought to be a matter of memory capacity. Brown argued instead that the issue is metacognition (knowing when, where, and how to remember).

One experiment had students read texts and then try to identify which parts were important versus what could be deleted without compromising the main ideas. They asked participants to judge this by first removing the least important ideas, then the next least, and so on. Youngest participants did not distinguish, older students did.

Another experiment had children learn to monitor their comprehension by summarizing, developing questions, predicting what will happen next, and take turns leading a discussion with these activities. Students improved in response to this, and transferred the skills readily.

Strategic skills for reading develop over time (re-reading, summarizing, slowing down, relating portions of texts)

Self-Explanation? Improving your performance by talking to yourself in the "right way". Good students tended to be those who did some "thinking out loud" about a problem, exploring a worked-out example and making inferences, examining trouble spots, etc. Explain your understanding as you proceed for a cognitive boost. Asking students "why do you think that?" is another prompt -- and, even better, ask them "how did I figure this out?" (explain a correct answer). Similar results shown in math, analogy, and science. This approach is thought to promote a more reflective strategy.

Writing

Evaluation and revision is similar to the comprehension & comprehension monitoring in reading. Initial writing models were informed by think-aloud protocols.

Hayes & Flower Model

• Planning
• Translating
• Reviewing

Bereiter & Scardamalia Finding Two differences between novice and expert:

• Novices list everything they know - knowledge telling
• Experts turn knowledge into a complete story - knowledge transforming

Math

Math students spend very little time on the problem, and focus more on the solution, even if it's a dead end -- mathematicians work mostly on the problem itself. Mathematicians also monitored their progress -- asking themselves if they are moving in a useful direction, and changing direction if not. this section was very strange -- geometry students characterized as knowing more about geometry than the mathematicians??? who are these people they did the study on??

De Corte et al recommend that students develop:

• domain knowledge - facts & rules of math
• general strategies for problem solving
• knowledge about one's own cognitive functioning
• self-regulatory skills

Fuchs et al intervention

• Large-scale self-regulation learning program
• Problem solving, monitoring, goal-setting, math self-efficacy, self-monitoring
• Compared to transfer training and no-training control group
• Self-regulation group outperformed others

Achievement is possible via high IQ and/or strong metacognition