See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/229971208

Self‐Regulated Learning

Chapter · January 2006

DOI: 10.1002/0470018860.s00671

CITATIONS READS
41 33,595

3 authors:

Gregory Schraw Douglas Kauffman

University of Nevada, Las Vegas
72 PUBLICATIONS 1,784 CITATIONS
88 PUBLICATIONS 18,337 CITATIONS
SEE PROFILE
SEE PROFILE

Stephen Lehman
Independent Researcher
9 PUBLICATIONS 1,612 CITATIONS

SEE PROFILE

All content following this page was uploaded by Douglas Kauffman on 26 March 2019.

The user has requested enhancement of the downloaded file.

 Self-regulated Learning 1063

Self-regulated Learning Intermediate article

Gregory Schraw, University of Nevada, Las Vegas, Nevada, USA
Douglas F Kauffman, University of Oklahoma, Norman, Oklahoma, USA
Stephen Lehman, Utah State University, Logan, Utah, USA

CONTENTS
Definitions and background Self-regulation and strategies
Components of self-regulated learning Comprehension monitoring
Processes and models Summary

Self-regulated learning relates to our ability to individuals learn to become self-regulated by

understand and control our learning environments. advancing through four levels of development, in-
cluding observational, imitative, self-controlled,
and self-regulated levels (Schunk, 1996; Zimmer-
DEFINITIONS AND BACKGROUND man, 2000). Learning at the observational level
Self-regulated learning refers to our ability to focuses on modeling, whereas learning at the imi-
understand and control our learning environ- tative level focuses on social guidance and feed-
ments. To do so, we must set goals, select strategies back. Both of these levels emphasize a reliance on
that help us achieve those goals, implement those external social factors. In contrast, as students de-
strategies, and monitor our progress towards velop, they rely more and more on internal, self-
our goals (Schunk, 1996). Few students are fully regulatory skills. Thus, at the self-controlled level,
self-regulated; however, those with better self- students construct internal standards for accept-
regulatory skills typically learn more with less able performance and become self-reinforcing
effort and report higher levels of academic satisfac- via positive self-talk and feedback. At the self-
tion (Pintrich, 2000; Zimmerman, 2000). regulatory level, individuals possess strong self-
Self-regulated learning theory is a relatively efficacy beliefs, as well as a large repertoire of cog-
recent development in cognitive psychology, with nitive strategies, that enable them to self-regulate
its origins dating back to the social-cognitive learn- their learning.
ing theory of Albert Bandura (1997). At the heart of Contemporary self-regulated learning theory
Bandura's theory is the idea of reciprocal determin- differs in two important ways from Bandura's
ism, which suggests that learning is the result of social-cognitive learning theory. One is the em-
personal, environmental, and behavioral factors. phasis that self-regulated learning theory places
Personal factors include a learner's beliefs and atti- on the construction and management of cognitive
tudes that affect learning and behavior. Environ- strategies to control one's academic learning, other-
mental factors include things such as the quality of wise known as meta-cognitive control. A second
instruction, teacher feedback, access to informa- change is the inclusion of broad motivational con-
tion, and help from peers and parents. Behavioral structs, such as causal attributions, goal orienta-
factors include the effects of prior performance. tions, and intrinsic motivation, that extend far
Reciprocal determinism states that each of these beyond the boundaries of social-cognitive motiv-
three factors affects the other two factors. ational constructs such as self-efficacy.
Since the 1980s, researchers have applied Ban- The remainder of this article is divided into five
dura's (1997) social-cognitive theory to many set- sections. The first of these describes three main
tings, including school learning. These attempts components of self-regulated learning, including
led to the development of self-regulated learning cognitive, meta-cognitive, and motivational com-
theory, which states that learning is governed by ponents. Section two summarizes three different
a variety of interacting cognitive, meta-cognitive, process models of self-regulated learning based
and motivational components (Butler and Winne, on the work of Michael Pressley, Philip Winne,
1995; Zimmerman, 2000). Social-cognitive ap- and Barry Zimmerman. Section three summarizes
proaches to self-regulated learning postulate that recent research on strategy instruction and
1064 Self-regulated Learning

provides a seven-step guide to implementing strat- capacity, whereas long-term memory is thought
egy instruction. Section four reviews recent re- to be permanent and unlimited in capacity
search on comprehension monitoring. Section five (Neath, 1998).
summarizes our main themes. Organization refers to how information is sorted
and arranged in long-term memory. Most experts
agree that information in long-term memory is or-
COMPONENTS OF SELF-REGULATED
ganized into knowledge structures called schemata
LEARNING and scripts (Neath, 1998). A schema is an organized
Experts agree that self-regulated learning includes body of declarative knowledge (i.e. stateable fac-
three main components, including cognition, meta- tual or conceptual knowledge). One example is a
cognition, and motivation. Cognition includes skills politics schema in which government is divided
necessary to encode, memorize, and recall infor- into three branches (i.e. executive, judicial, and le-
mation. Meta-cognition includes skills that enable gislative) headed by the President, Supreme Court,
learners to understand and monitor cognitive pro- and Congress respectively. Schemata are crucial to
cesses. Motivation includes beliefs and attitudes self-regulated learning because they enable us to
that affect the use and development of cognitive organize, store, and recall large amounts of infor-
and meta-cognitive skills. Each of these three mation very quickly. Indeed, without schemata, the
main components is necessary for self-regulation. automated processing on which expertise depends
Those who possess cognitive skills, but are unmoti- would be impossible. Similarly, a script is an or-
vated to use them, for example, do not achieve ganized body of procedural knowledge that en-
at the same level of performance as individuals ables us to perform a task or skill automatically.
who possess skills and are motivated to use them Examples include scripts for getting dressed,
(Zimmerman, 2000). Similarly, those who are mo- ordering food at restaurants, riding a bicycle, and
tivated, but do not possess the necessary cognitive most other routine procedural activities. Collect-
and meta-cognitive skills, often fail to achieve high ively, schemata and scripts enable us to organize
levels of self-regulation. and access huge amounts of information in
The three main components of self-regulation memory quickly.
can be further subdivided into the subcomponents Elaboration refers to our ability to embellish new
shown in Figure 1. We describe each of these com- information by linking it to information in long-
ponents below, as well as several finer-grained term memory. Elaboration can occur at a shallow
subcomponents. or deep level. Shallow elaboration is often ref-
erred to as maintenance rehearsal. For example, stu-
dents may memorize the great lakes by repeating
Cognition
them over and over. In contrast, students could
The cognitive component includes encoding, organ- engage in several types of elaborative rehearsal such
ization, elaboration, and inferencing subcomponents. as creating an acronym (e.g. HOMES: Huron, On-
Encoding refers to our ability to process infor- tario, Michigan, Erie, and Superior), or constructing
mation currently in working memory in order a mental image based on a map. Students could
to store it in long-term memory. Working memory encode this information even more deeply by asso-
is thought to be temporary and limited in ciating different lakes with different colors. For
example, Lake Ontario could be remembered as
orange, while Lake Michigan is remembered
Self-regulated learning as magenta.
Inferencing refers to our ability to infer new
information from existing knowledge and informa-
Cognition Meta-cognition Motivation
tion. Inferencing is crucial to self-regulated learn-
Examples: Rehearsal Knowledge Self-efficacy ing because it enables us to go beyond what we
of cognition
know, constructing what we need to know to per-
Organization Regulation Attribution form at a higher level of proficiency. The role of
of cognition
inference generation has been studied extensively
Elaboration Goal
orientation by cognitive psychologists, particularly as it per-
tains to reading. Researchers know that self-regu-
Inferencing Intrinsic
motivation
lated readers combine the facts and main ideas in a
text into themes that are never stated explicitly, but
Figure 1. Components of self-regulated learning. are essential for comprehension.
 Self-regulated Learning 1065

Meta-cognition appraising the products and regulatory processes

of one's learning. Typical examples include re-
Meta-cognition includes two main subcomponents
evaluating one's goals, revising predictions, and
generally referred to as knowledge of cognition and
consolidating intellectual gains.
regulation of cognition (Schraw and Moshman, 1995). Experts believe that self-regulatory processes, in-
Knowledge of cognition refers to what we know
cluding planning, monitoring, and evaluation, may
about our cognition, and usually includes three
not be conscious or explicit in many learning situ-
subcomponents. The first, declarative knowledge,
ations. One reason is that many of these processes
includes knowledge about ourselves as learners
are highly automated, at least among adults. A
and what factors influence our performance. For
second reason is that some of these processes may
example, most adult learners know the limitations
develop without any conscious reflection and
of their memory system and can plan accordingly
therefore are difficult to report to others.
based on this knowledge. Procedural knowledge, in
contrast, refers to knowledge about strategies and
other procedures. For instance, most adults possess
Motivation
a basic repertoire of useful strategies such as note- The motivation component shown in Figure 1 in-
taking, slowing down for important information, cludes four important subcomponents, consisting
skimming unimportant information, using mne- of self-efficacy, attributions, goal orientations, and in-
monics, summarizing main ideas, and periodic trinsic motivation.
self-testing. Finally, conditional knowledge includes Self-efficacy: refers to the degree to which an
knowledge of why and when to use a particular individual is confident that he or she can perform
strategy. Individuals with a high degree of condi- a specific task or accomplish a specific goal
tional knowledge are better able to assess the (Bandura, 1997). Self-efficacy is extremely import-
demands of a specific learning situation and, in ant for self-regulated learning because it affects the
turn, select strategies that are most appropriate for extent to which learners engage and persist at chal-
that situation. lenging tasks. Previous research indicates that stu-
Research suggests that knowledge of cognition is dents with higher self-efficacy are more likely to
late developing and explicit (Alexander et al., 1995). engage in a difficult task and more likely to persist
Adults tend to have more knowledge about their at a task even in the face of initial failures compared
own cognition and are better able to describe that to low-efficacy students (Pajares, 1996). Higher
knowledge than children and adolescents. How- levels of self-efficacy are related positively to
ever, many adults cannot explain their expert school achievement and self-esteem. The trends
knowledge and performance, and often fail to observed with respect to student self-efficacy also
spontaneously transfer domain-specific knowledge generalize to teachers and even schools. Teachers
to a new setting. This suggests that meta-cognitive with higher levels of teaching self-efficacy, for
knowledge need not be explicit to be useful and, in example, set higher goals and standards, give
fact, may be implicit in some situations (Butler and more autonomy to students, and help students
Winne, 1995). reach higher levels of achievement than do teachers
Regulation of cognition typically includes at least with lower levels of self-efficacy (Goddard et al.,
three components, planning, monitoring, and 2000).
evaluation (Schraw and Moshman, 1995). Planning Self-efficacy is affected by a number of variables,
involves the selection of appropriate strategies and but especially vicarious learning and modeling.
the allocation of resources. Planning includes goal Vicarious learning occurs when individuals learn
setting, activating relevant background knowledge, by observing others perform a skill or discuss
and budgeting time. Previous research suggests a topic. Vicarious learning is advantageous to
that experts are more self-regulated compared to learners because they are not expected to perform
novices largely due to effective planning, particu- the task, and therefore experience less anxiety, and
larly global planning that occurs prior to beginning can also focus all of their resources on observing
a task. Monitoring includes self-testing skills neces- experts. Modeling occurs when learners learn inten-
sary to control learning. Research indicates that tionally from other individuals such as teachers
adults monitor at both the local (i.e. an individual and students. Modeling typically includes the
test item) and global levels (i.e. all items on a test). teacher breaking a complex task into manageable
Research also suggests that even skilled adult parts and asking students to demonstrate each
learners are poor monitors under certain conditions part separately in sequence. Bandura (1997) pro-
(Pressley and Ghatala, 1990). Evaluation refers to posed that modeling is effective because it raises
1066 Self-regulated Learning

expectations that a new skill can be acquired, in affective responses, whereas internal, uncontrol-
addition to providing a great deal of knowledge lable causes such as ability may create negative
about the skill. Peer models are usually the most emotions such as shame and guilt.
effective because they are most similar to the Fortunately, students may be helped to change
learner. Indeed, students are most likely to in- negative attributional responses through observa-
creases their own self-efficacy when observing a tion and training. A review of the attributional
model of similar ability level perform the skill retraining literature found that the majority of at-
(Schunk, 1996). tribution retraining programs are successful. Suc-
There are two main ways to increase students' cessful programs included the following three
self-efficacy. One is to use both expert (e.g. teacher) components: (1) individuals are taught to identify
and non-expert (e.g. student peers) models. Re- desirable behaviors such as effort and strategy use,
search demonstrates that models improve cogni- (2) attributions that support positive behaviors are
tive skills and self-efficacy. The second is to evaluated, (3) favorable attributional responses
provide as much informational feedback to stu- are rewarded. Overall, the attributional retrain-
dents as possible. Feedback should indicate not ing literature provides evidence that individuals
only whether the skill was performed acceptably, can learn to make more adaptive attributional re-
but provide as much information as possible about sponses that improve motivation and achievement-
how to improve subsequent performance. Given related behaviors such as effort, help seeking, and
detailed informational feedback, performance and persistence.
self-efficacy can increase even after students ex- Goal orientations: refer to beliefs about ability
perience initial difficulty performing a skill. and how those beliefs affect learning. Dweck and
Attributions: refer to causal explanations of Leggett (1988) proposed that learners adopt either
events that happen in our lives. For example, two performance goals or learning goals based on personal
students may do poorly on a test. One student may beliefs about the stability of intelligence. Students
attribute her poor performance to bad luck, while who believe that intelligence is fixed and un-
the other student attributes her poor performance changeable adopt performance goals, in which
to lack of effort. These attributions provide very they seek to prove their competence in academic
different explanations of the same event. Attribu- settings. Those who believe that intelligence is mal-
tion theory states that it is not an event per se that leable and changeable adopt learning goals, in
affects us, but our interpretation of that event which they seek to improve their competence. A
(Graham and Weiner, 1996; Weiner, 1986). number of studies suggest that students who
Weiner (1986) proposed that attributions vary adopt learning goals are more adaptive and satis-
along three dimensions. The first is locus of control, fied than students who adopt performance goals.
which defines the cause of an outcome as either Learning-oriented students typically achieve more
internal or external to the individual. Mood and because they seek challenge, persist, use strategies,
emotions are examples of internal causes, whereas attribute success to effort, and demonstrate positive
teachers are external causes. A second dimension is responses to periodic failure. In contrast, perform-
stability, which pertains to whether an attributional ance-oriented students often adopt maladaptive re-
cause is permanent or temporary. Ability is stable, sponse patterns characterized by avoidance of
whereas effort tends to be less stable. A third di- challenge, quitting after initial failure, use of in-
mension is controllability, which refers to whether appropriate strategies, helplessness, and attribut-
an event is under the student's control or is uncon- ing success to uncontrollable causes such as
trollable. Controllable causes of academic success ability and luck (Ames and Archer, 1988).
include effort and strategy use, whereas uncontrol- Learning- and performance-oriented students
lable causes include luck and task difficulty. differ with respect to academic self-efficacy and
Researchers have considered the separate effects self-regulated learning (Midgley et al., 1995). Schunk
of locus of control, stability, and controllability; (1996) reported that students with learning goals
however, of greater importance is how the three report higher levels of self-efficacy which, in turn,
dimensions contribute simultaneously. Internal, is related to higher levels of academic achievement.
controllable, stable causes such as effort promote Bouffard et al. (1995) found that college students
positive academic responses, whereas external, un- who reported strong learning goals also attained
controllable, unstable causes such as luck produce the highest level of academic self-regulation.
frustration or undermine academic confidence. Learning-oriented students also appear to
Weiner (1986) reported that internal, controllable have better relationships with their teachers. In
causes, such as strategy use, promote positive fact, in a study by Ames and Archer (1988),
 Self-regulated Learning 1067

learning-oriented students considered teachers to includes the three main components described
be more important than effort, ability, or strategy above (i.e. cognition, meta-cognition, and motiv-
use. Surprisingly, however, learning-oriented stu- ation). However, experts differ in terms of the rela-
dents did not attribute their failure to teachers, tive contribution of each of these three components.
whereas performance-oriented students did! In this section, we describe three overlapping yet
Bruning et al. (1999) have suggested a number of distinct models of self-regulated learning. The first
ways to foster adaptive goals. One is to promote a model is based on the work of Michael Pressley
flexible attitude about the role of ability. Students and colleagues and is known as the Good Informa-
should be encouraged to make the most of their tion Processor model (Pressley et al., 1989). This
existing ability rather than focus on how much model places special emphasis on the role of cogni-
ability they have compared to other students. tive strategies. The second model is based on the
Second, teachers and parents should concentrate work of Philip Winne and colleagues and is known
on rewarding effort. Third, teachers should stress as the Self-regulated Learning model. This model
that mistakes are a normal part of learning and are emphasizes the interactive relationships among
best dealt with by persistence, help seeking, and cognitive, meta-cognitive, and motivational com-
strategy use. ponents, although it differs from the good infor-
Intrinsic motivation: refers to behaviors that are mation processor model in that it more strongly
engaged in for their own sake (Deci and Ryan, emphasizes the role of meta-cognition, and espe-
2000). When an individual is intrinsically motiv- cially the role of monitoring and feedback. The
ated, tasks are performed for internal reasons third model is based on the work of Barry Zimmer-
such as joy and satisfaction, rather than for external man and is often referred to as the Phases of Self-
reasons such as rewards, obligation, or threat. Ex- regulation model. Though quite similar to Winne's
trinsic motivation refers to behaviors that are self-regulated learning model, the phases of
performed to achieve some externally prized con- self-regulation model differs in that it has fewer
sequence, not out of interest or personal desire for phases and emphasizes the role of personal
mastery. Studies reveal that performing a task be- volition.
cause of intrinsic motivation results in satisfaction
and a desire to perform the task again. In contrast, Pressley's Good Information Processor
performing a task due to extrinsic motivation may
lead to indifference or displeasure, and may de-
Model
crease the desire to perform the task again. The good information processor model was de-
veloped initially to explain effective strategy use.
The model includes five main characteristics: (1) a
Summary
broad repertoire of strategies, (2) meta-cognitive
Self-regulated learning refers to learners' abilities knowledge about why, when, and where to use
to understand and control their learning environ- strategies, (3) a broad knowledge base that is rele-
ments. Self-regulated learning involves a combin- vant to the task at hand, (4) the ability to eliminate
ation of cognitive strategy use, meta-cognitive unwanted distractions, (5) automaticity in the four
processing, and motivational beliefs. Cognitive components mentioned previously.
strategies take the form of encoding, organization, Regarding the first of these characteristics, Press-
elaboration, and inference-making. Meta-cognitive ley et al. (1989) distinguished between two different
processing refers to knowledge and control of cog- types of strategies. The first of these include
nitive skills, and usually involves planning, moni- domain-specific strategies that are appropriate
toring, and evaluating. Finally, the motivational only for a specific task such as solving a quadratic
component refers to students' beliefs in their cap- equation. The second type is a higher-order strat-
acity to learn. Motivation takes many forms includ- egy, which is used to control other lower-level
ing self-efficacy, attributions, goal orientation, and strategies. One example of a higher-order strategy
intrinsic motivation. is sequencing the use of several domain-specific
strategies while reading; tactics such as skim-
ming before reading, drawing conclusions, then
PROCESSES AND MODELS
reviewing. Using higher-order strategies to orches-
Self-regulated learning is a relatively new field of trate lower-level strategies is crucial to one's ability
study in cognitive psychology. Different theorists to self-regulate.
view self-regulated learning in different ways. The second characteristic corresponds closely to
Most experts agree that self-regulated learning what experts call conditional knowledge; that is,
1068 Self-regulated Learning

knowledge about when, why, and where to use Winne's Self-Regulated Learning Model
strategies in an optimal fashion. Conditional know-
Winne's Self-Regulated Learning model makes
ledge is important because knowing how to do
three broad assumptions about self-regulated
something is of little practical use unless one also
knows when to do it. For example, one can study learning. First, unlike the Good Information
Processor model, it emphasizes the sequence of
the correct information for a test but do so quite
self-regulated learning over the five individual
poorly. Knowing how to study is at least as import-
components described by Pressley and colleagues.
ant as knowing what to study.
Second, Winne conceptualizes self-regulated learn-
The third characteristic is a broad knowledge
ing as the ability to bridge the gap between setting
base. Researchers agree that learning is extremely
and achieving learning goals (Winne and Perry,
difficult and time-consuming without supporting
2000). Third, he highlights the importance of self-
knowledge already in long-term memory. Indeed,
a number of studies report that average-ability stu- generated feedback as a mechanism that supports
self-regulated learning. Winne and colleagues em-
dents with high levels of knowledge about a topic
phasize four stages in the self-regulated learning
generally outperform higher-ability students with
process, including (1) defining the task, (2) plan-
low levels of background knowledge (Bruning et al.,
ning and goal setting, (3) enacting tactics, (4)
1999). In addition, background knowledge is re-
adapting meta-cognition (Butler and Winne, 1995;
lated to the effective use of memory resources and
Winne and Perry, 2000).
the ability to construct integrated internal represen-
Phase one of the self-regulated learning model
tations of a task.
The fourth characteristic is the ability to elimin- consists of defining the task. This phase can be
broken down into two major subcomponents, task
ate unwanted distractions. Pressley et al. (1989)
conditions and cognitive conditions that support
refer to this as action control. Students with action
self-regulated learning. Task conditions refer to
control are able to motivate themselves in several
factors external to the learner such as time, instruc-
ways. One is to allocate effort to the task and persist
tional cues, and availability of resources that affect
when the task is difficult. A second is to attribute
the learner's ability to perform a task successfully.
their success to controllable causes such as effort
Cognitive conditions refer to factors internal to the
and strategy use. A third is to tune out unwanted
distractions. learner that affect performance. Examples include
personal beliefs such as self-efficacy and attribu-
The fifth characteristic is automaticity of the four
tions, domain knowledge, knowledge of the task,
previous characteristics. Automaticity refers to being
conditional knowledge, and personal motivational
able to perform a task or retrieve information from
factors such as goals and intrinsic motivation.
memory with little conscious effort. Automaticity is
According to Winne, both task and cognitive con-
important for two interrelated reasons. First, auto-
ditions conjointly influence the learner's ability to
mating lower-level cognitive skills conserves our
evaluate the task and formulate outcome expect-
resources for higher-level cognitive tasks that are
less likely to be fully automated. Second, because ations. After defining the task, the self-regulated
learner must plan and set goals.
fewer resources are consumed, students have more
Phase two consists of planning and goal setting.
resources available to engage in complex informa-
In this phase, learners evaluate task and cognitive
tion processing. Automaticity is one of the key
conditions information in order to establish their
components of self-regulation because our effort
main goals. Learners may have multiple goals, each
can be devoted to planning and monitoring the
with its own standard for performance. Learners
outcome of our performance, rather than perform-
set these standards based on their knowledge about
ing the task.
Collectively, the five skills characteristics of good the task domain, automaticity performing the task,
and an assessment of how well they can monitor
information processors enable students to self-
their performance.
regulate learning with a great deal of efficiency.
Phase three consists of selecting and coordinat-
Needless to say, each of the five components is
ing a wide variety of cognitive learning strategies
necessary and must work in synchrony with the
based on the goals and standards that have been set
others. For this reason, Pressley and colleagues
previously. These include information search strat-
(Pressley and Wharton-McDonald, 1997) suggest
egies such as retrieving information from long-
teaching the five components described above in
an integrated fashion in which all components are term memory, information management strategies
such as identifying important information and
addressed simultaneously.
 Self-regulated Learning 1069

summarizing, and help-seeking strategies such as observation. Self-control includes attention focusing
working in groups or asking peers or teachers for strategies that enable students to tune out un-
help. The purpose of enacting strategies is to pro- wanted distractions. This component also includes
duce cognitive products such as organized infor- the use of a wide variety of study strategies to
mation in memory or written reports. In turn, control learning. In addition, students may also
products can be evaluated against the goals and use what Zimmerman refers to as self-instruction,
standards set in phase two. in which individuals vocalize instructions to them-
Phase four consists of using meta-cognitive selves explaining how to perform a task or monitor
knowledge, particularly monitoring, to evaluate their comprehension. The self-observation com-
one's performance. Winne and Perry (2000) distin- ponent includes a variety of record keeping activ-
guished between meta-cognitive knowledge and ities in which learners keep track of their cognitive
meta-cognitive monitoring. Meta-cognitive know- progress and emotional reactions.
ledge includes conditional knowledge about The self-reflection phase includes both self-
cognitive strategies, knowledge about the task, judgments and self-reactions. Self-judgments include
knowledge about one's current knowledge base, monitoring one's cognitive performance, evaluat-
and one's own interests. Meta-cognitive monitor- ing affective reactions to performance, and making
ing includes judgments of one's available re- appropriate causal attributions. During this phase,
sources, assessing the relative difficulty of the learners monitor whether they have met their pre-
task, evaluating current performance, and generat- viously established learning goals. Progress is
ing feedback to correct comprehension errors. evaluated on a number of dimensions, including
Meta-cognitive knowledge and monitoring are whether basic goals have been mastered, how well
used to assess the fit between students' initial they have performed relative to others, and how
goals and final performance. The extent to which well they have performed compared to their previ-
a disparity exists necessitates a return to phase one ous performance. Self-reactions pertain mainly to
of the cycle to eliminate this disparity. This self- judgments of their affective engagement. Zimmer-
regulated learning process continues until per- man (2000) argues that task satisfaction depends on
formance matches one's learning goals. one's ability to meet previously established goals
and standards.
Zimmerman's Phases of Self-
Regulation Model Summary
Zimmerman and colleagues proposed a cyclical The three models of self-regulation described
model of self-regulated learning that consists of above agree on the need for cognitive, meta-cogni-
three distinct phases: (1) forethought, (2) perform- tive and motivational components. They differ in
ance control, (3) self-reflection (Zimmerman and the extent to which these main components inter-
Kitsantas, 1999; Zimmerman, 2000). Zimmerman's act. Pressley's Good Information Processor Model
model is similar to Winne's sequential self-regu- focuses more on the componential makeup of
lated learning model in many respects, yet differs self-regulated learning, while the other models
in three ways. First, motivational factors such as emphasize the sequential (Winne) or the cyclical
self-efficacy play a more influential role during (Zimmerman) nature of the self-regulated learning
the forethought phase. Second, self-affect plays a process. In Winne's model, the process of bridging
more influential role during the self-reflection the gap between setting and achieving goals, as
phase. Third monitoring plays a smaller role in well as the importance of self-generated feedback
Zimmerman's model than in the Pressley and is crucial. In contrast to Winne, Zimmerman em-
Winne models. phasizes the importance of motivational variables,
The forethought phase of Zimmerman's model such as self-efficacy as well as suggesting that self-
includes two main components consisting of task reactions play an integral part in the learner's con-
analysis and motivational beliefs. Task analysis vari- tinued regulatory processes.
ables are similar to those described by Winne and
Perry (2000), and include planning and goal setting.
SELF-REGULATION AND STRATEGIES
However, Zimmerman includes more motivational
variables, emphasizing the role of self-efficacy, goal
orientations, and intrinsic motivation.
Research on Strategy Instruction
The performance phase likewise includes two Research on strategy instruction has boomed since
components consisting of self-control and self- the early 1980s. Unfortunately, implementing a
1070 Self-regulated Learning

strategy training program is time-consuming and finally automated. The following seven-step se-
expensive; thus, most interventions have focused quence is typical of effective strategy instruction
on the effectiveness of only one or two strategies. programs:
Researchers have conducted meta-analyses to better 1. Discuss and explain the value of strategies. Strategies
understand the effectiveness of strategy instruc- increase efficiency, save time, and enhance deeper
tion. A meta-analysis is a procedure that aggregates processing.
similar studies to determine their overall effective- 2. Introduce a limited number of strategies. Most pro-
ness. Two analyses by Hattie et al. (1996) and grams recommend four or five general strategies such
Rosenshine et al. (1996) supported the following as summarizing and comprehension monitoring.
four claims: 3. Practice each strategy over an extended period of time
until it becomes automated.
1. Strategy instruction typically is moderately to highly 4. Model strategies extensively so students acquire not
successful. only the strategy, but conditional knowledge about
2. Strategy instruction appears to be most helpful for how, when, and where to use the strategy.
younger and under-achieving students. 5. Provide feedback to students about strategy use. This
3. Programs that combine several interrelated strategies information helps them evaluate the effectiveness of
are more effective than those that include only one strategies and monitor their comprehension more ef-
strategy. An interrelated repertoire of four or five fectively.
strategies seems optimal (Pressley and Wharton- 6. Promote transfer by encouraging students to use strat-
McDonald, 1997). egies in new settings or by adapting them to new
4. Strategy interventions are more effective when they tasks. Previous research suggests that strategies
teach conditional knowledge. learned in one setting do not transfer unless students
Strategy research has also addressed whether are instructed to use them in different settings.
strategy instruction is more effective in teacher- 7. Encourage reflection on strategy use. Students who
reflect on strategy use acquire more meta-cognitive
centered versus student-centered classrooms. Nei-
knowledge and are more apt to use strategies in a
ther type of setting appears to increase the flexible way to self-regulate. One way students can
effectiveness of interventions, although it should self-reflect is through journals. A second way is by
be noted that few studies have compared different comparing the advantages and disadvantages of dif-
instructional approaches directly. ferent strategies with peers.
Another important question addressed by strat-
egy instruction research is what strategies should Overall, research suggests that effective strategy
be taught to students. Experts generally agree that a use is critical for self-regulated learning. Effective
limited number of general strategies are most ef- strategy use is accomplished most efficiently
fective (e.g. 4±8). Based on their analysis, Hattie through the extended instruction, modeling and
et al. (1996) suggested the following set of strat- practice of a small repertoire of general strategies
egies: self-checking, creating a good study environ- such as planning, inferencing, and monitoring. In
ment, planning and goal setting, reviewing, addition to a repertoire of strategies, increasing
summarizing, and seeking teacher and peer assist- self-regulated learning is dependent on the
ance. Similarly, in a comprehensive review of the learner's ability to monitor comprehension.
strategy instruction literature, Dole et al. (1991) rec-
ommended a similar set of five core learning strat- COMPREHENSION MONITORING
egies that included determining what is important
to learn, summarizing, drawing inferences, gener- Comprehension monitoring refers to evaluating the
ating questions before and during studying, and ongoing state of one's understanding. Monitoring
monitoring one's comprehension. takes place during or after a learning activity and
provides information about the effectiveness of that
activity. Monitoring is important because it pro-
How to Teach Strategies vides self-generated feedback to the learner. With-
Strategy instruction should be an integral part out accurate monitoring, efficient control of one's
of every class (Pressley and Wharton-McDonald, performance is impossible.
1997). Prior to strategy instruction, teachers should Monitoring studies typically require individuals
help students understand the value of strategies to make subjective judgments of learning or test
and decide which strategies to teach their students. performance during or after an initial study
Most experts recommend an instructional sequence phase. Four types of judgments have been used in
that stretches from 10 to 20 weeks in which strat- the adult monitoring literature, including ease of
egies are introduced, modeled, practiced, and learning (i.e. judgments of encoding difficulty),
 Self-regulated Learning 1071

judgments of learning (i.e. the degree to which infor- rather than recognition tests. Calibration of per-
mation was learned during the study phase), feeling formance was also related to level of test perform-
of knowing (i.e. the degree to which one has access to ance. In addition, individuals monitored with less
previously learned information in memory), and bias when judging their performance on easy rather
performance judgments (i.e. assessments of perform- than more difficult items.
ance accuracy). Third, feedback, incentives, practice, and
Studies measuring these four types of judgments training positively affect monitoring proficiency.
indicate that adults monitor their learning and Schraw (1994) reported that pre-experimental esti-
performance with a moderate degree of success, mates of monitoring proficiency were related to
although results vary from study to study. Surpris- both local (i.e. the accuracy of item-specific per-
ingly, monitoring proficiency does not appear to be formance judgments made during testing) and
related strongly to relevant domain knowledge global (i.e. judgments of overall performance
or academic achievement (Pressley and Ghatala, made after testing) monitoring accuracy. The ac-
1990). These conclusions have been supported in curacy of local monitoring was correlated posi-
the children's monitoring literature as well, al- tively to the accuracy of global monitoring. In
though there is considerable debate regarding addition, the change in monitoring accuracy be-
whether children monitor as accurately as adults tween local and global monitoring improved sig-
(Alexander et al., 1995). nificantly among good monitors, but did not
These studies indicate that there are three spe- improve among poor monitors.
cific factors that affect monitoring proficiency. Monitoring training also improves performance.
First, situational constraints affect estimates of Delclos and Harrington (1991) examined fifth- and
monitoring proficiency. One constraint is the sixth-graders' ability to solve computer problems
point in the learning-test sequence in which moni- after assignment to one of three conditions. The
toring judgments are made. A number of studies first group received specific problem-solving
indicate that calibration of comprehension (i.e. the training, the second received problem-solving plus
correlation between pre-test judgments and actual self-monitoring training and practice, while the
test performance) is often quite poor, with most third received no training. The monitored prob-
studies reporting correlations in the 0.00 to 0.25 lem-solving group solved more of the difficult
range (Pressley and Ghatala, 1990). In contrast, problems than either of the remaining groups and
calibration of performance (i.e. the correlation be- took less time to do so. The group receiving prob-
tween post-test judgments and actual test perform- lem-solving and monitoring training also solved
ance) appears to be much better in both children complex problems faster than the control group.
and adults, often ranging from 0.30 to 0.50 (Pressley The monitoring research summarized above
and Ghatala, 1990). leads to a number of conclusions. Overall, adults
Second, specific testing conditions affect moni- monitor their performance with a moderate degree
toring proficiency. For example, calibration of com- of accuracy. Monitoring accuracy improves as tests
prehension can be improved under the following become easier and more factual. Second, monitor-
circumstances: (1) when adjunct questions similar ing proficiency appears to be independent of intel-
to post-test questions are provided during study, lectual ability (Alexander et al., 1995) and academic
(2) when periodic feedback is provided to test achievement (Pressley and Ghatala, 1990). Third,
takers, (3) when expert knowledge about the to- monitoring proficiency may be independent or
be-learned material is minimized, (4) when test even negatively related to domain knowledge, in-
takers generated missing text information. Sur- dependent of ease of comprehension judgments,
prisingly, calibration of comprehension does not but correlated with other types of meta-cognitive
appear to improve when learners are specifically knowledge. Fourth, one's ability to monitor one's
requested to monitor their comprehension or when performance may improve with practice (Delclos
they are given the opportunity to re-study the to- and Harrington, 1991).
be-learned materials, or when they are given prac-
tice questions prior to study.
SUMMARY
Like calibration of comprehension, calibration of
performance improved under a number of testing Self-regulated learning theory evolved from Ban-
conditions, especially when adjunct questions were dura's (1997) social-cognitive learning theory. In
provided during the study phase, when test takers 2002 self-regulated learning theory focuses on the
received external incentives to improve monitoring transition from social to self-directed learning pro-
accuracy, and when test takers received recall cesses. Several main themes emerge from this
1072 Self-regulated Learning

research. The first is that self-regulated learners Butler DL and Winne PH (1995) Feedback and self-
rely on an integrated repertoire of cognitive, meta- regulated learning: a theoretical synthesis. Review of
cognitive, and motivational skills. Second, self- Educational Research 65: 245±281.
regulated learners use these skills to plan, set Deci EL and Ryan RM (2000) Self-determination theory
and the facilitation of intrinsic motivation, social
goals, implement and monitor strategy use, and
development, and well being.
evaluate their learning goals. Third, self-regulated Delclos VR and Harrington C (1991) Effects of strategy
learners use a wide variety of strategies in flexible monitoring and proactive instruction on children's
ways, augmenting these strategies with a variety problem-solving performance. Journal of Educational
of adaptive motivational beliefs such as high self- Psychology 83: 35±42.
efficacy, attributions to internal controllable causes Dole JA, Duffy GG, Roehler LR and Pearson PD (1991)
of academic success, learning goals, and intrinsic Moving from the old to the new: research on reading
motivation. comprehension instruction. Review of Educational
A review of the strategy instruction literature Research 61: 239±264.
suggests that a repertoire of four or five general Dweck CS and Leggett ES (1988) A social-cognitive
strategies such as summarizing and comprehen- approach to motivation and personality. Psychological
Review 95: 256±273.
sion monitoring skills can be taught to students of
Goddard RD, Hoy WK and Woolfolk Hoy A (2000)
all ages. Strategy instruction is most effective when Collective teacher efficacy: its meaning, measure and
it extends over a 10- to 20-week period, includes impact on student achievement. American Educational
extensive modeling and feedback, teaches condi- Research Journal 37: 479±508.
tional knowledge necessary for effective strategy Graham S and Weiner B (1996) Theories and principles of
use, and helps students recognize the cognitive motivation. In: Berliner DC and Calfee RC (eds) The
and motivational benefits of strategy use. In add- Handbook of Educational Psychology, pp. 63±84. New
ition to strategies, self-regulated learners monitor York, NY: Macmillan.
their comprehension and debug learning problems Hattie J, Biggs J and Purdie N (1996) Effects of learning
when they occur. Research suggests that monitor- skills interventions on student learning: a meta-
ing is unrelated to ability and improves with analysis. Review of Educational Research 66: 99±136.
Midgley C, Anderman EM and Hicks L (1995)
practice.
Differences between elementary and middle school
Overall, there is strong agreement that self-regu- teachers and students: a goal theory approach. Journal of
lated learning is necessary for academic success Early Adolescence 15: 90±113.
and is attributable to meta-cognitive knowledge Neath I (1998) Human Memory: An Introduction to
and a repertoire of learning strategies, rather than Research, Data, and Theory. Pacific Grove, CA:
ability per se. These skills are learned through ob- Brooks-Cole Publishing.
servation and modeling, feedback from others, and Pajares F (1996) Self-efficacy beliefs in academic settings.
consistent practice. Models also provide important Review of Educational Research 66: 543±578.
motivational support. Pintrich P (2000) The role of goal orientation in self-
regulated learning. In: Boekaerts M, Pintrich P
and Zeidner M (eds) Handbook of Self-Regulation,
References
pp. 452±501. San Diego, CA: Academic Press.
Alexander JM, Carr M and Schwanenflugel PJ (1995) Pressley M, Borkowski J and Schneider W (1989) Good
Development of metacognition in gifted children: information processing: what is it and what education
directions for future research. Developmental Review can do to promote it. Journal of Experimental Child
15: 1±37. Psychology 43: 194±211.
Ames C and Archer J (1988) Achievement in the Pressley M and Ghatala ES (1990) Self-regulated learning:
classroom: student learning strategies and monitoring learning from text. Educational Psychologist
motivational processes. Journal of Educational Psychology 25: 19±33.
80: 260±267. Pressley M and Wharton-McDonald R (1997) Skilled
Bandura A (1997) Self-Efficacy: The Exercise of Control. comprehension and its development through
New York, NY: Freeman. instruction. School Psychology Review 26: 448±466.
Bouffard T, Boisvert J, Vezeau C and Larouche C (1995) Rosenshine B, Meister C and Chapman S (1996) Teaching
The impact of goal orientation on self-regulation and students to generate questions: a review of the
performance among college students. British Journal of intervention studies. Review of Educational Research 66:
Educational Psychology 65: 317±329. 181±221.
Bruning R, Schraw G and Ronning R (1999) Cognitive Schraw G (1994) The effect of meta-cognitive knowledge
Psychology and Instruction, 3rd edn. Upper Saddle River, on local and global monitoring. Contemporary
NJ: Merrill. Educational Psychology 19: 143±154.
 Self-regulated Learning 1073

Schraw G and Moshman D (1995) Metacognitive Arts, Sciences, Sports, and Games, pp. 1±50. Mahwah, NJ:
theories. Educational Psychology Review 7: 351±371. Lawrence Erlbaum.
Schunk D (1996) Goal and self-evaluative influences Ericsson KA and Kintsch W (1995) Long-term working
during children's cognitive skill learning. American memory. Psychological Review 102: 211±245.
Educational Research Journal 33: 359±382. Hofer B, Yu S and Pintrich P (1998) Teaching college
Weiner B (1986) An Attributional Theory of Motivation and students to be self-regulated. In: Schunk DH and
Emotion. New York, NY: Springer-Verlag. Zimmerman BJ (eds) Self-regulated Learning: From
Winne P and Perry N (2000) Measuring self-regulated Teaching to Self-reflective Practice, pp. 57±85. New York,
learning. In: Boekaerts M, Pintrich P and Zeidner M NY: Guilford.
(eds) Handbook of Self-Regulation, pp. 531±566. San Diego, Kintsch W (1998) Comprehension: A Paradigm for Cognition.
CA: Academic Press. Cambridge, UK: Cambridge University Press.
Zimmerman B (2000) Attaining self-regulated learning: a Schunk DH and Ertmer PA (2000) Self-regulation and
social-cognitive perspective. In: Boekaerts M, Pintrich P academic learning: self-efficacy enhancing
and Zeidner M (eds) Handbook of Self-Regulation, interventions. In: Boekaerts M, Pintrich P and Zeidner
pp. 13±39. San Diego, CA: Academic Press. M (eds) Handbook of Self-regulation, pp. 631±649. San
Zimmerman B and Kitsantas A (1999) Acquiring writing Diego, CA: Academic Press.
revision skills: shifting from process to outcome self- Schunk DH and Zimmerman BJ (1998) Conclusions and
regulatory goals. Journal of Educational Psychology 91: future directions for academic interventions. In: Schunk
241±250. DH and Zimmerman BJ (eds) Self-regulated Learning:
From Teaching to Self-reflective Practice, pp. 225±235. New
York, NY: Guilford.
Further Reading
Zeidner M, Boekaerts M and Pintrich P (2000) Self-
Brown R and Pressley M (1994) Self-regulated reading regulation: directions and challenges for future
and getting meaning from text: the transactional research. In: Boekaerts M, Pintrich P and Zeidner M
strategies instruction model and its ongoing validation. (eds) Handbook of Self-regulation, pp. 750±768. San Diego,
In: Schunk DH and Zimmermann BJ (eds) Self- CA: Academic Press.
regulation of Learning and Performance: Issues and Zimmerman B (1998) Academic studying and the
Educational Implications, pp. 155±180. Hillsdale, NJ: development of personal skill: a self-regulatory
Lawrence Erlbaum. perspective. Educational Psychologist 33: 73±86.
Ericsson KA (1996) The acquisition of expert
performance. In: Ericsson KA (ed.) The Road to
Excellence: The Acquisition of Expert Performance in the

Self-stimulation
See Brain Self-stimulation; Reward, Brain Mechanisms of
______________________________________________________________________________

Semantic Analysis
See Latent Semantic Analysis; Syntax and Semantics: Formal Approaches
______________________________________________________________________________

View publication stats