Content uploaded by Douglas Kauffman
Author content
All content in this area was uploaded by Douglas Kauffman on Mar 26, 2019
Content may be subject to copyright.
Intermediate articleSelf-regulated Learning
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
Self-regulated learning relates to our ability to
understand and control our learning environments.
DEFINITIONS AND BACKGROUND
Self-regulated learning refers to our ability to
understand and control our learning environ-
ments. To do so, we must set goals, select strategies
that help us achieve those goals, implement those
strategies, and monitor our progress towards
our goals (Schunk, 1996). Few students are fully
self-regulated; however, those with better self-
regulatory skills typically learn more with less
effort and report higher levels of academic satisfac-
tion (Pintrich, 2000; Zimmerman, 2000).
Self-regulated learning theory is a relatively
recent development in cognitive psychology, with
its origins dating back to the social-cognitive learn-
ing theory of Albert Bandura (1997). At the heart of
Bandura's theory is the idea of reciprocal determin-
ism, which suggests that learning is the result of
personal, environmental, and behavioral factors.
Personal factors include a learner's beliefs and atti-
tudes that affect learning and behavior. Environ-
mental factors include things such as the quality of
instruction, teacher feedback, access to informa-
tion, and help from peers and parents. Behavioral
factors include the effects of prior performance.
Reciprocal determinism states that each of these
three factors affects the other two factors.
Since the 1980s, researchers have applied Ban-
dura's (1997) social-cognitive theory to many set-
tings, including school learning. These attempts
led to the development of self-regulated learning
theory, which states that learning is governed by
a variety of interacting cognitive, meta-cognitive,
and motivational components (Butler and Winne,
1995; Zimmerman, 2000). Social-cognitive ap-
proaches to self-regulated learning postulate that
individuals learn to become self-regulated by
advancing through four levels of development, in-
cluding observational, imitative, self-controlled,
and self-regulated levels (Schunk, 1996; Zimmer-
man, 2000). Learning at the observational level
focuses on modeling, whereas learning at the imi-
tative level focuses on social guidance and feed-
back. Both of these levels emphasize a reliance on
external social factors. In contrast, as students de-
velop, they rely more and more on internal, self-
regulatory skills. Thus, at the self-controlled level,
students construct internal standards for accept-
able performance and become self-reinforcing
via positive self-talk and feedback. At the self-
regulatory level, individuals possess strong self-
efficacy beliefs, as well as a large repertoire of cog-
nitive strategies, that enable them to self-regulate
their learning.
Contemporary self-regulated learning theory
differs in two important ways from Bandura's
social-cognitive learning theory. One is the em-
phasis that self-regulated learning theory places
on the construction and management of cognitive
strategies to control one's academic learning, other-
wise known as meta-cognitive control. A second
change is the inclusion of broad motivational con-
structs, such as causal attributions, goal orienta-
tions, and intrinsic motivation, that extend far
beyond the boundaries of social-cognitive motiv-
ational constructs such as self-efficacy.
The remainder of this article is divided into five
sections. The first of these describes three main
components of self-regulated learning, including
cognitive, meta-cognitive, and motivational com-
ponents. Section two summarizes three different
process models of self-regulated learning based
on the work of Michael Pressley, Philip Winne,
and Barry Zimmerman. Section three summarizes
recent research on strategy instruction and
CONTENTS
Definitions and background
Components of self-regulated learning
Processes and models
Self-regulation and strategies
Comprehension monitoring
Summary
Self-regulated Learning 1063
provides a seven-step guide to implementing strat-
egy instruction. Section four reviews recent re-
search on comprehension monitoring. Section five
summarizes our main themes.
COMPONENTS OF SELF-REGULATED
LEARNING
Experts agree that self-regulated learning includes
three main components, including cognition,meta-
cognition, and motivation. Cognition includes skills
necessary to encode, memorize, and recall infor-
mation. Meta-cognition includes skills that enable
learners to understand and monitor cognitive pro-
cesses. Motivation includes beliefs and attitudes
that affect the use and development of cognitive
and meta-cognitive skills. Each of these three
main components is necessary for self-regulation.
Those who possess cognitive skills, but are unmoti-
vated to use them, for example, do not achieve
at the same level of performance as individuals
who possess skills and are motivated to use them
(Zimmerman, 2000). Similarly, those who are mo-
tivated, but do not possess the necessary cognitive
and meta-cognitive skills, often fail to achieve high
levels of self-regulation.
The three main components of self-regulation
can be further subdivided into the subcomponents
shown in Figure 1. We describe each of these com-
ponents below, as well as several finer-grained
subcomponents.
Cognition
The cognitive component includes encoding,organ-
ization,elaboration, and inferencing subcomponents.
Encoding refers to our ability to process infor-
mation currently in working memory in order
to store it in long-term memory. Working memory
is thought to be temporary and limited in
capacity, whereas long-term memory is thought
to be permanent and unlimited in capacity
(Neath, 1998).
Organization refers to how information is sorted
and arranged in long-term memory. Most experts
agree that information in long-term memory is or-
ganized into knowledge structures called schemata
and scripts (Neath, 1998). A schema is an organized
body of declarative knowledge (i.e. stateable fac-
tual or conceptual knowledge). One example is a
politics schema in which government is divided
into three branches (i.e. executive, judicial, and le-
gislative) headed by the President, Supreme Court,
and Congress respectively. Schemata are crucial to
self-regulated learning because they enable us to
organize, store, and recall large amounts of infor-
mation very quickly. Indeed, without schemata, the
automated processing on which expertise depends
would be impossible. Similarly, a script is an or-
ganized body of procedural knowledge that en-
ables us to perform a task or skill automatically.
Examples include scripts for getting dressed,
ordering food at restaurants, riding a bicycle, and
most other routine procedural activities. Collect-
ively, schemata and scripts enable us to organize
and access huge amounts of information in
memory quickly.
Elaboration refers to our ability to embellish new
information by linking it to information in long-
term memory. Elaboration can occur at a shallow
or deep level. Shallow elaboration is often ref-
erred to as maintenance rehearsal. For example, stu-
dents may memorize the great lakes by repeating
them over and over. In contrast, students could
engage in several types of elaborative rehearsal such
as creating an acronym (e.g. HOMES: Huron, On-
tario, Michigan, Erie, and Superior), or constructing
a mental image based on a map. Students could
encode this information even more deeply by asso-
ciating different lakes with different colors. For
example, Lake Ontario could be remembered as
orange, while Lake Michigan is remembered
as magenta.
Inferencing refers to our ability to infer new
information from existing knowledge and informa-
tion. Inferencing is crucial to self-regulated learn-
ing because it enables us to go beyond what we
know, constructing what we need to know to per-
form at a higher level of proficiency. The role of
inference generation has been studied extensively
by cognitive psychologists, particularly as it per-
tains to reading. Researchers know that self-regu-
lated readers combine the facts and main ideas in a
text into themes that are never stated explicitly, but
are essential for comprehension.
Self-regulated learning
Cognition Meta-cognition Motivation
Examples: Rehearsal
Organization
Elaboration
Inferencing
Knowledge
of cognition
Regulation
of cognition
Self-efficacy
Attribution
Goal
orientation
Intrinsic
motivation
Figure 1. Components of self-regulated learning.
1064 Self-regulated Learning
Meta-cognition
Meta-cognition includes two main subcomponents
generally referred to as knowledge of cognition and
regulation of cognition (Schraw and Moshman, 1995).
Knowledge of cognition refers to what we know
about our cognition, and usually includes three
subcomponents. The first, declarative knowledge,
includes knowledge about ourselves as learners
and what factors influence our performance. For
example, most adult learners know the limitations
of their memory system and can plan accordingly
based on this knowledge. Procedural knowledge,in
contrast, refers to knowledge about strategies and
other procedures. For instance, most adults possess
a basic repertoire of useful strategies such as note-
taking, slowing down for important information,
skimming unimportant information, using mne-
monics, summarizing main ideas, and periodic
self-testing. Finally, conditional knowledge includes
knowledge of why and when to use a particular
strategy. Individuals with a high degree of condi-
tional knowledge are better able to assess the
demands of a specific learning situation and, in
turn, select strategies that are most appropriate for
that situation.
Research suggests that knowledge of cognition is
late developing and explicit (Alexander et al., 1995).
Adults tend to have more knowledge about their
own cognition and are better able to describe that
knowledge than children and adolescents. How-
ever, many adults cannot explain their expert
knowledge and performance, and often fail to
spontaneously transfer domain-specific knowledge
to a new setting. This suggests that meta-cognitive
knowledge need not be explicit to be useful and, in
fact, may be implicit in some situations (Butler and
Winne, 1995).
Regulation of cognition typically includes at least
three components, planning, monitoring, and
evaluation (Schraw and Moshman, 1995). Planning
involves the selection of appropriate strategies and
the allocation of resources. Planning includes goal
setting, activating relevant background knowledge,
and budgeting time. Previous research suggests
that experts are more self-regulated compared to
novices largely due to effective planning, particu-
larly global planning that occurs prior to beginning
a task. Monitoring includes self-testing skills neces-
sary to control learning. Research indicates that
adults monitor at both the local (i.e. an individual
test item) and global levels (i.e. all items on a test).
Research also suggests that even skilled adult
learners are poor monitors under certain conditions
(Pressley and Ghatala, 1990). Evaluation refers to
appraising the products and regulatory processes
of one's learning. Typical examples include re-
evaluating one's goals, revising predictions, and
consolidating intellectual gains.
Experts believe that self-regulatory processes, in-
cluding planning, monitoring, and evaluation, may
not be conscious or explicit in many learning situ-
ations. One reason is that many of these processes
are highly automated, at least among adults. A
second reason is that some of these processes may
develop without any conscious reflection and
therefore are difficult to report to others.
Motivation
The motivation component shown in Figure 1 in-
cludes four important subcomponents, consisting
of self-efficacy,attributions,goal orientations, and in-
trinsic motivation.
Self-efficacy: refers to the degree to which an
individual is confident that he or she can perform
a specific task or accomplish a specific goal
(Bandura, 1997). Self-efficacy is extremely import-
ant for self-regulated learning because it affects the
extent to which learners engage and persist at chal-
lenging tasks. Previous research indicates that stu-
dents with higher self-efficacy are more likely to
engage in a difficult task and more likely to persist
at a task even in the face of initial failures compared
to low-efficacy students (Pajares, 1996). Higher
levels of self-efficacy are related positively to
school achievement and self-esteem. The trends
observed with respect to student self-efficacy also
generalize to teachers and even schools. Teachers
with higher levels of teaching self-efficacy, for
example, set higher goals and standards, give
more autonomy to students, and help students
reach higher levels of achievement than do teachers
with lower levels of self-efficacy (Goddard et al.,
2000).
Self-efficacy is affected by a number of variables,
but especially vicarious learning and modeling.
Vicarious learning occurs when individuals learn
by observing others perform a skill or discuss
a topic. Vicarious learning is advantageous to
learners because they are not expected to perform
the task, and therefore experience less anxiety, and
can also focus all of their resources on observing
experts. Modeling occurs when learners learn inten-
tionally from other individuals such as teachers
and students. Modeling typically includes the
teacher breaking a complex task into manageable
parts and asking students to demonstrate each
part separately in sequence. Bandura (1997) pro-
posed that modeling is effective because it raises
Self-regulated Learning 1065
expectations that a new skill can be acquired, in
addition to providing a great deal of knowledge
about the skill. Peer models are usually the most
effective because they are most similar to the
learner. Indeed, students are most likely to in-
creases their own self-efficacy when observing a
model of similar ability level perform the skill
(Schunk, 1996).
There are two main ways to increase students'
self-efficacy. One is to use both expert (e.g. teacher)
and non-expert (e.g. student peers) models. Re-
search demonstrates that models improve cogni-
tive skills and self-efficacy. The second is to
provide as much informational feedback to stu-
dents as possible. Feedback should indicate not
only whether the skill was performed acceptably,
but provide as much information as possible about
how to improve subsequent performance. Given
detailed informational feedback, performance and
self-efficacy can increase even after students ex-
perience initial difficulty performing a skill.
Attributions: refer to causal explanations of
events that happen in our lives. For example, two
students may do poorly on a test. One student may
attribute her poor performance to bad luck, while
the other student attributes her poor performance
to lack of effort. These attributions provide very
different explanations of the same event. Attribu-
tion theory states that it is not an event per se that
affects us, but our interpretation of that event
(Graham and Weiner, 1996; Weiner, 1986).
Weiner (1986) proposed that attributions vary
along three dimensions. The first is locus of control,
which defines the cause of an outcome as either
internal or external to the individual. Mood and
emotions are examples of internal causes, whereas
teachers are external causes. A second dimension is
stability, which pertains to whether an attributional
cause is permanent or temporary. Ability is stable,
whereas effort tends to be less stable. A third di-
mension is controllability, which refers to whether
an event is under the student's control or is uncon-
trollable. Controllable causes of academic success
include effort and strategy use, whereas uncontrol-
lable causes include luck and task difficulty.
Researchers have considered the separate effects
of locus of control, stability, and controllability;
however, of greater importance is how the three
dimensions contribute simultaneously. Internal,
controllable, stable causes such as effort promote
positive academic responses, whereas external, un-
controllable, unstable causes such as luck produce
frustration or undermine academic confidence.
Weiner (1986) reported that internal, controllable
causes, such as strategy use, promote positive
affective responses, whereas internal, uncontrol-
lable causes such as ability may create negative
emotions such as shame and guilt.
Fortunately, students may be helped to change
negative attributional responses through observa-
tion and training. A review of the attributional
retraining literature found that the majority of at-
tribution retraining programs are successful. Suc-
cessful programs included the following three
components: (1) individuals are taught to identify
desirable behaviors such as effort and strategy use,
(2) attributions that support positive behaviors are
evaluated, (3) favorable attributional responses
are rewarded. Overall, the attributional retrain-
ing literature provides evidence that individuals
can learn to make more adaptive attributional re-
sponses that improve motivation and achievement-
related behaviors such as effort, help seeking, and
persistence.
Goal orientations: refer to beliefs about ability
and how those beliefs affect learning. Dweck and
Leggett (1988) proposed that learners adopt either
performance goals or learning goals based on personal
beliefs about the stability of intelligence. Students
who believe that intelligence is fixed and un-
changeable adopt performance goals, in which
they seek to prove their competence in academic
settings. Those who believe that intelligence is mal-
leable and changeable adopt learning goals, in
which they seek to improve their competence. A
number of studies suggest that students who
adopt learning goals are more adaptive and satis-
fied than students who adopt performance goals.
Learning-oriented students typically achieve more
because they seek challenge, persist, use strategies,
attribute success to effort, and demonstrate positive
responses to periodic failure. In contrast, perform-
ance-oriented students often adopt maladaptive re-
sponse patterns characterized by avoidance of
challenge, quitting after initial failure, use of in-
appropriate strategies, helplessness, and attribut-
ing success to uncontrollable causes such as
ability and luck (Ames and Archer, 1988).
Learning- and performance-oriented students
differ with respect to academic self-efficacy and
self-regulated learning (Midgley et al., 1995). Schunk
(1996) reported that students with learning goals
report higher levels of self-efficacy which, in turn,
is related to higher levels of academic achievement.
Bouffard et al. (1995) found that college students
who reported strong learning goals also attained
the highest level of academic self-regulation.
Learning-oriented students also appear to
have better relationships with their teachers. In
fact, in a study by Ames and Archer (1988),
1066 Self-regulated Learning
learning-oriented students considered teachers to
be more important than effort, ability, or strategy
use. Surprisingly, however, learning-oriented stu-
dents did not attribute their failure to teachers,
whereas performance-oriented students did!
Bruning et al. (1999) have suggested a number of
ways to foster adaptive goals. One is to promote a
flexible attitude about the role of ability. Students
should be encouraged to make the most of their
existing ability rather than focus on how much
ability they have compared to other students.
Second, teachers and parents should concentrate
on rewarding effort. Third, teachers should stress
that mistakes are a normal part of learning and are
best dealt with by persistence, help seeking, and
strategy use.
Intrinsic motivation: refers to behaviors that are
engaged in for their own sake (Deci and Ryan,
2000). When an individual is intrinsically motiv-
ated, tasks are performed for internal reasons
such as joy and satisfaction, rather than for external
reasons such as rewards, obligation, or threat. Ex-
trinsic motivation refers to behaviors that are
performed to achieve some externally prized con-
sequence, not out of interest or personal desire for
mastery. Studies reveal that performing a task be-
cause of intrinsic motivation results in satisfaction
and a desire to perform the task again. In contrast,
performing a task due to extrinsic motivation may
lead to indifference or displeasure, and may de-
crease the desire to perform the task again.
Summary
Self-regulated learning refers to learners' abilities
to understand and control their learning environ-
ments. Self-regulated learning involves a combin-
ation of cognitive strategy use, meta-cognitive
processing, and motivational beliefs. Cognitive
strategies take the form of encoding, organization,
elaboration, and inference-making. Meta-cognitive
processing refers to knowledge and control of cog-
nitive skills, and usually involves planning, moni-
toring, and evaluating. Finally, the motivational
component refers to students' beliefs in their cap-
acity to learn. Motivation takes many forms includ-
ing self-efficacy, attributions, goal orientation, and
intrinsic motivation.
PROCESSES AND MODELS
Self-regulated learning is a relatively new field of
study in cognitive psychology. Different theorists
view self-regulated learning in different ways.
Most experts agree that self-regulated learning
includes the three main components described
above (i.e. cognition, meta-cognition, and motiv-
ation). However, experts differ in terms of the rela-
tive contribution of each of these three components.
In this section, we describe three overlapping yet
distinct models of self-regulated learning. The first
model is based on the work of Michael Pressley
and colleagues and is known as the Good Informa-
tion Processor model (Pressley et al., 1989). This
model places special emphasis on the role of cogni-
tive strategies. The second model is based on the
work of Philip Winne and colleagues and is known
as the Self-regulated Learning model. This model
emphasizes the interactive relationships among
cognitive, meta-cognitive, and motivational com-
ponents, although it differs from the good infor-
mation processor model in that it more strongly
emphasizes the role of meta-cognition, and espe-
cially the role of monitoring and feedback. The
third model is based on the work of Barry Zimmer-
man and is often referred to as the Phases of Self-
regulation model. Though quite similar to Winne's
self-regulated learning model, the phases of
self-regulation model differs in that it has fewer
phases and emphasizes the role of personal
volition.
Pressley's Good Information Processor
Model
The good information processor model was de-
veloped initially to explain effective strategy use.
The model includes five main characteristics: (1) a
broad repertoire of strategies, (2) meta-cognitive
knowledge about why, when, and where to use
strategies, (3) a broad knowledge base that is rele-
vant to the task at hand, (4) the ability to eliminate
unwanted distractions, (5) automaticity in the four
components mentioned previously.
Regarding the first of these characteristics, Press-
ley et al. (1989) distinguished between two different
types of strategies. The first of these include
domain-specific strategies that are appropriate
only for a specific task such as solving a quadratic
equation. The second type is a higher-order strat-
egy, which is used to control other lower-level
strategies. One example of a higher-order strategy
is sequencing the use of several domain-specific
strategies while reading; tactics such as skim-
ming before reading, drawing conclusions, then
reviewing. Using higher-order strategies to orches-
trate lower-level strategies is crucial to one's ability
to self-regulate.
The second characteristic corresponds closely to
what experts call conditional knowledge; that is,
Self-regulated Learning 1067
knowledge about when, why, and where to use
strategies in an optimal fashion. Conditional know-
ledge is important because knowing how to do
something is of little practical use unless one also
knows when to do it. For example, one can study
the correct information for a test but do so quite
poorly. Knowing how to study is at least as import-
ant as knowing what to study.
The third characteristic is a broad knowledge
base. Researchers agree that learning is extremely
difficult and time-consuming without supporting
knowledge already in long-term memory. Indeed,
a number of studies report that average-ability stu-
dents with high levels of knowledge about a topic
generally outperform higher-ability students with
low levels of background knowledge (Bruning et al.,
1999). In addition, background knowledge is re-
lated to the effective use of memory resources and
the ability to construct integrated internal represen-
tations of a task.
The fourth characteristic is the ability to elimin-
ate unwanted distractions. Pressley et al. (1989)
refer to this as action control. Students with action
control are able to motivate themselves in several
ways. One is to allocate effort to the task and persist
when the task is difficult. A second is to attribute
their success to controllable causes such as effort
and strategy use. A third is to tune out unwanted
distractions.
The fifth characteristic is automaticity of the four
previous characteristics. Automaticity refers to being
able to perform a task or retrieve information from
memory with little conscious effort. Automaticity is
important for two interrelated reasons. First, auto-
mating lower-level cognitive skills conserves our
resources for higher-level cognitive tasks that are
less likely to be fully automated. Second, because
fewer resources are consumed, students have more
resources available to engage in complex informa-
tion processing. Automaticity is one of the key
components of self-regulation because our effort
can be devoted to planning and monitoring the
outcome of our performance, rather than perform-
ing the task.
Collectively, the five skills characteristics of good
information processors enable students to self-
regulate learning with a great deal of efficiency.
Needless to say, each of the five components is
necessary and must work in synchrony with the
others. For this reason, Pressley and colleagues
(Pressley and Wharton-McDonald, 1997) suggest
teaching the five components described above in
an integrated fashion in which all components are
addressed simultaneously.
Winne's Self-Regulated Learning Model
Winne's Self-Regulated Learning model makes
three broad assumptions about self-regulated
learning. First, unlike the Good Information
Processor model, it emphasizes the sequence of
self-regulated learning over the five individual
components described by Pressley and colleagues.
Second, Winne conceptualizes self-regulated learn-
ing as the ability to bridge the gap between setting
and achieving learning goals (Winne and Perry,
2000). Third, he highlights the importance of self-
generated feedback as a mechanism that supports
self-regulated learning. Winne and colleagues em-
phasize four stages in the self-regulated learning
process, including (1) defining the task, (2) plan-
ning and goal setting, (3) enacting tactics, (4)
adapting meta-cognition (Butler and Winne, 1995;
Winne and Perry, 2000).
Phase one of the self-regulated learning model
consists of defining the task. This phase can be
broken down into two major subcomponents, task
conditions and cognitive conditions that support
self-regulated learning. Task conditions refer to
factors external to the learner such as time, instruc-
tional cues, and availability of resources that affect
the learner's ability to perform a task successfully.
Cognitive conditions refer to factors internal to the
learner that affect performance. Examples include
personal beliefs such as self-efficacy and attribu-
tions, domain knowledge, knowledge of the task,
conditional knowledge, and personal motivational
factors such as goals and intrinsic motivation.
According to Winne, both task and cognitive con-
ditions conjointly influence the learner's ability to
evaluate the task and formulate outcome expect-
ations. After defining the task, the self-regulated
learner must plan and set goals.
Phase two consists of planning and goal setting.
In this phase, learners evaluate task and cognitive
conditions information in order to establish their
main goals. Learners may have multiple goals, each
with its own standard for performance. Learners
set these standards based on their knowledge about
the task domain, automaticity performing the task,
and an assessment of how well they can monitor
their performance.
Phase three consists of selecting and coordinat-
ing a wide variety of cognitive learning strategies
based on the goals and standards that have been set
previously. These include information search strat-
egies such as retrieving information from long-
term memory, information management strategies
such as identifying important information and
1068 Self-regulated Learning
summarizing, and help-seeking strategies such as
working in groups or asking peers or teachers for
help. The purpose of enacting strategies is to pro-
duce cognitive products such as organized infor-
mation in memory or written reports. In turn,
products can be evaluated against the goals and
standards set in phase two.
Phase four consists of using meta-cognitive
knowledge, particularly monitoring, to evaluate
one's performance. Winne and Perry (2000) distin-
guished between meta-cognitive knowledge and
meta-cognitive monitoring. Meta-cognitive know-
ledge includes conditional knowledge about
cognitive strategies, knowledge about the task,
knowledge about one's current knowledge base,
and one's own interests. Meta-cognitive monitor-
ing includes judgments of one's available re-
sources, assessing the relative difficulty of the
task, evaluating current performance, and generat-
ing feedback to correct comprehension errors.
Meta-cognitive knowledge and monitoring are
used to assess the fit between students' initial
goals and final performance. The extent to which
a disparity exists necessitates a return to phase one
of the cycle to eliminate this disparity. This self-
regulated learning process continues until per-
formance matches one's learning goals.
Zimmerman's Phases of Self-
Regulation Model
Zimmerman and colleagues proposed a cyclical
model of self-regulated learning that consists of
three distinct phases: (1) forethought, (2) perform-
ance control, (3) self-reflection (Zimmerman and
Kitsantas, 1999; Zimmerman, 2000). Zimmerman's
model is similar to Winne's sequential self-regu-
lated learning model in many respects, yet differs
in three ways. First, motivational factors such as
self-efficacy play a more influential role during
the forethought phase. Second, self-affect plays a
more influential role during the self-reflection
phase. Third monitoring plays a smaller role in
Zimmerman's model than in the Pressley and
Winne models.
The forethought phase of Zimmerman's model
includes two main components consisting of task
analysis and motivational beliefs. Task analysis vari-
ables are similar to those described by Winne and
Perry (2000), and include planning and goal setting.
However, Zimmerman includes more motivational
variables, emphasizing the role of self-efficacy, goal
orientations, and intrinsic motivation.
The performance phase likewise includes two
components consisting of self-control and self-
observation. Self-control includes attention focusing
strategies that enable students to tune out un-
wanted distractions. This component also includes
the use of a wide variety of study strategies to
control learning. In addition, students may also
use what Zimmerman refers to as self-instruction,
in which individuals vocalize instructions to them-
selves explaining how to perform a task or monitor
their comprehension. The self-observation com-
ponent includes a variety of record keeping activ-
ities in which learners keep track of their cognitive
progress and emotional reactions.
The self-reflection phase includes both self-
judgments and self-reactions. Self-judgments include
monitoring one's cognitive performance, evaluat-
ing affective reactions to performance, and making
appropriate causal attributions. During this phase,
learners monitor whether they have met their pre-
viously established learning goals. Progress is
evaluated on a number of dimensions, including
whether basic goals have been mastered, how well
they have performed relative to others, and how
well they have performed compared to their previ-
ous performance. Self-reactions pertain mainly to
judgments of their affective engagement. Zimmer-
man (2000) argues that task satisfaction depends on
one's ability to meet previously established goals
and standards.
Summary
The three models of self-regulation described
above agree on the need for cognitive, meta-cogni-
tive and motivational components. They differ in
the extent to which these main components inter-
act. Pressley's Good Information Processor Model
focuses more on the componential makeup of
self-regulated learning, while the other models
emphasize the sequential (Winne) or the cyclical
(Zimmerman) nature of the self-regulated learning
process. In Winne's model, the process of bridging
the gap between setting and achieving goals, as
well as the importance of self-generated feedback
is crucial. In contrast to Winne, Zimmerman em-
phasizes the importance of motivational variables,
such as self-efficacy as well as suggesting that self-
reactions play an integral part in the learner's con-
tinued regulatory processes.
SELF-REGULATION AND STRATEGIES
Research on Strategy Instruction
Research on strategy instruction has boomed since
the early 1980s. Unfortunately, implementing a
Self-regulated Learning 1069
strategy training program is time-consuming and
expensive; thus, most interventions have focused
on the effectiveness of only one or two strategies.
Researchers have conducted meta-analyses to better
understand the effectiveness of strategy instruc-
tion. A meta-analysis is a procedure that aggregates
similar studies to determine their overall effective-
ness. Two analyses by Hattie et al. (1996) and
Rosenshine et al. (1996) supported the following
four claims:
1. Strategy instruction typically is moderately to highly
successful.
2. Strategy instruction appears to be most helpful for
younger and under-achieving students.
3. Programs that combine several interrelated strategies
are more effective than those that include only one
strategy. An interrelated repertoire of four or five
strategies seems optimal (Pressley and Wharton-
McDonald, 1997).
4. Strategy interventions are more effective when they
teach conditional knowledge.
Strategy research has also addressed whether
strategy instruction is more effective in teacher-
centered versus student-centered classrooms. Nei-
ther type of setting appears to increase the
effectiveness of interventions, although it should
be noted that few studies have compared different
instructional approaches directly.
Another important question addressed by strat-
egy instruction research is what strategies should
be taught to students. Experts generally agree that a
limited number of general strategies are most ef-
fective (e.g. 4±8). Based on their analysis, Hattie
et al. (1996) suggested the following set of strat-
egies: self-checking, creating a good study environ-
ment, planning and goal setting, reviewing,
summarizing, and seeking teacher and peer assist-
ance. Similarly, in a comprehensive review of the
strategy instruction literature, Dole et al. (1991) rec-
ommended a similar set of five core learning strat-
egies that included determining what is important
to learn, summarizing, drawing inferences, gener-
ating questions before and during studying, and
monitoring one's comprehension.
How to Teach Strategies
Strategy instruction should be an integral part
of every class (Pressley and Wharton-McDonald,
1997). Prior to strategy instruction, teachers should
help students understand the value of strategies
and decide which strategies to teach their students.
Most experts recommend an instructional sequence
that stretches from 10 to 20 weeks in which strat-
egies are introduced, modeled, practiced, and
finally automated. The following seven-step se-
quence is typical of effective strategy instruction
programs:
1. Discuss and explain the value of strategies. Strategies
increase efficiency, save time, and enhance deeper
processing.
2. Introduce a limited number of strategies. Most pro-
grams recommend four or five general strategies such
as summarizing and comprehension monitoring.
3. Practice each strategy over an extended period of time
until it becomes automated.
4. Model strategies extensively so students acquire not
only the strategy, but conditional knowledge about
how, when, and where to use the strategy.
5. Provide feedback to students about strategy use. This
information helps them evaluate the effectiveness of
strategies and monitor their comprehension more ef-
fectively.
6. Promote transfer by encouraging students to use strat-
egies in new settings or by adapting them to new
tasks. Previous research suggests that strategies
learned in one setting do not transfer unless students
are instructed to use them in different settings.
7. Encourage reflection on strategy use. Students who
reflect on strategy use acquire more meta-cognitive
knowledge and are more apt to use strategies in a
flexible way to self-regulate. One way students can
self-reflect is through journals. A second way is by
comparing the advantages and disadvantages of dif-
ferent strategies with peers.
Overall, research suggests that effective strategy
use is critical for self-regulated learning. Effective
strategy use is accomplished most efficiently
through the extended instruction, modeling and
practice of a small repertoire of general strategies
such as planning, inferencing, and monitoring. In
addition to a repertoire of strategies, increasing
self-regulated learning is dependent on the
learner's ability to monitor comprehension.
COMPREHENSION MONITORING
Comprehension monitoring refers to evaluating the
ongoing state of one's understanding. Monitoring
takes place during or after a learning activity and
provides information about the effectiveness of that
activity. Monitoring is important because it pro-
vides self-generated feedback to the learner. With-
out accurate monitoring, efficient control of one's
performance is impossible.
Monitoring studies typically require individuals
to make subjective judgments of learning or test
performance during or after an initial study
phase. Four types of judgments have been used in
the adult monitoring literature, including ease of
learning (i.e. judgments of encoding difficulty),
1070 Self-regulated Learning
judgments of learning (i.e. the degree to which infor-
mation was learned during the study phase), feeling
of knowing (i.e. the degree to which one has access to
previously learned information in memory), and
performance judgments (i.e. assessments of perform-
ance accuracy).
Studies measuring these four types of judgments
indicate that adults monitor their learning and
performance with a moderate degree of success,
although results vary from study to study. Surpris-
ingly, monitoring proficiency does not appear to be
related strongly to relevant domain knowledge
or academic achievement (Pressley and Ghatala,
1990). These conclusions have been supported in
the children's monitoring literature as well, al-
though there is considerable debate regarding
whether children monitor as accurately as adults
(Alexander et al., 1995).
These studies indicate that there are three spe-
cific factors that affect monitoring proficiency.
First, situational constraints affect estimates of
monitoring proficiency. One constraint is the
point in the learning-test sequence in which moni-
toring judgments are made. A number of studies
indicate that calibration of comprehension (i.e. the
correlation between pre-test judgments and actual
test performance) is often quite poor, with most
studies reporting correlations in the 0.00 to 0.25
range (Pressley and Ghatala, 1990). In contrast,
calibration of performance (i.e. the correlation be-
tween post-test judgments and actual test perform-
ance) appears to be much better in both children
and adults, often ranging from 0.30 to 0.50 (Pressley
and Ghatala, 1990).
Second, specific testing conditions affect moni-
toring proficiency. For example, calibration of com-
prehension can be improved under the following
circumstances: (1) when adjunct questions similar
to post-test questions are provided during study,
(2) when periodic feedback is provided to test
takers, (3) when expert knowledge about the to-
be-learned material is minimized, (4) when test
takers generated missing text information. Sur-
prisingly, calibration of comprehension does not
appear to improve when learners are specifically
requested to monitor their comprehension or when
they are given the opportunity to re-study the to-
be-learned materials, or when they are given prac-
tice questions prior to study.
Like calibration of comprehension, calibration of
performance improved under a number of testing
conditions, especially when adjunct questions were
provided during the study phase, when test takers
received external incentives to improve monitoring
accuracy, and when test takers received recall
rather than recognition tests. Calibration of per-
formance was also related to level of test perform-
ance. In addition, individuals monitored with less
bias when judging their performance on easy rather
than more difficult items.
Third, feedback, incentives, practice, and
training positively affect monitoring proficiency.
Schraw (1994) reported that pre-experimental esti-
mates of monitoring proficiency were related to
both local (i.e. the accuracy of item-specific per-
formance judgments made during testing) and
global (i.e. judgments of overall performance
made after testing) monitoring accuracy. The ac-
curacy of local monitoring was correlated posi-
tively to the accuracy of global monitoring. In
addition, the change in monitoring accuracy be-
tween local and global monitoring improved sig-
nificantly among good monitors, but did not
improve among poor monitors.
Monitoring training also improves performance.
Delclos and Harrington (1991) examined fifth- and
sixth-graders' ability to solve computer problems
after assignment to one of three conditions. The
first group received specific problem-solving
training, the second received problem-solving plus
self-monitoring training and practice, while the
third received no training. The monitored prob-
lem-solving group solved more of the difficult
problems than either of the remaining groups and
took less time to do so. The group receiving prob-
lem-solving and monitoring training also solved
complex problems faster than the control group.
The monitoring research summarized above
leads to a number of conclusions. Overall, adults
monitor their performance with a moderate degree
of accuracy. Monitoring accuracy improves as tests
become easier and more factual. Second, monitor-
ing proficiency appears to be independent of intel-
lectual ability (Alexander et al., 1995) and academic
achievement (Pressley and Ghatala, 1990). Third,
monitoring proficiency may be independent or
even negatively related to domain knowledge, in-
dependent of ease of comprehension judgments,
but correlated with other types of meta-cognitive
knowledge. Fourth, one's ability to monitor one's
performance may improve with practice (Delclos
and Harrington, 1991).
SUMMARY
Self-regulated learning theory evolved from Ban-
dura's (1997) social-cognitive learning theory. In
2002 self-regulated learning theory focuses on the
transition from social to self-directed learning pro-
cesses. Several main themes emerge from this
Self-regulated Learning 1071
research. The first is that self-regulated learners
rely on an integrated repertoire of cognitive, meta-
cognitive, and motivational skills. Second, self-
regulated learners use these skills to plan, set
goals, implement and monitor strategy use, and
evaluate their learning goals. Third, self-regulated
learners use a wide variety of strategies in flexible
ways, augmenting these strategies with a variety
of adaptive motivational beliefs such as high self-
efficacy, attributions to internal controllable causes
of academic success, learning goals, and intrinsic
motivation.
A review of the strategy instruction literature
suggests that a repertoire of four or five general
strategies such as summarizing and comprehen-
sion monitoring skills can be taught to students of
all ages. Strategy instruction is most effective when
it extends over a 10- to 20-week period, includes
extensive modeling and feedback, teaches condi-
tional knowledge necessary for effective strategy
use, and helps students recognize the cognitive
and motivational benefits of strategy use. In add-
ition to strategies, self-regulated learners monitor
their comprehension and debug learning problems
when they occur. Research suggests that monitor-
ing is unrelated to ability and improves with
practice.
Overall, there is strong agreement that self-regu-
lated learning is necessary for academic success
and is attributable to meta-cognitive knowledge
and a repertoire of learning strategies, rather than
ability per se. These skills are learned through ob-
servation and modeling, feedback from others, and
consistent practice. Models also provide important
motivational support.
References
Alexander JM, Carr M and Schwanenflugel PJ (1995)
Development of metacognition in gifted children:
directions for future research. Developmental Review
15: 1±37.
Ames C and Archer J (1988) Achievement in the
classroom: student learning strategies and
motivational processes. Journal of Educational Psychology
80: 260±267.
Bandura A (1997) Self-Efficacy: The Exercise of Control.
New York, NY: Freeman.
Bouffard T, Boisvert J, Vezeau C and Larouche C (1995)
The impact of goal orientation on self-regulation and
performance among college students. British Journal of
Educational Psychology 65: 317±329.
Bruning R, Schraw G and Ronning R (1999) Cognitive
Psychology and Instruction, 3rd edn. Upper Saddle River,
NJ: Merrill.
Butler DL and Winne PH (1995) Feedback and self-
regulated learning: a theoretical synthesis. Review of
Educational Research 65: 245±281.
Deci EL and Ryan RM (2000) Self-determination theory
and the facilitation of intrinsic motivation, social
development, and well being.
Delclos VR and Harrington C (1991) Effects of strategy
monitoring and proactive instruction on children's
problem-solving performance. Journal of Educational
Psychology 83: 35±42.
Dole JA, Duffy GG, Roehler LR and Pearson PD (1991)
Moving from the old to the new: research on reading
comprehension instruction. Review of Educational
Research 61: 239±264.
Dweck CS and Leggett ES (1988) A social-cognitive
approach to motivation and personality. Psychological
Review 95: 256±273.
Goddard RD, Hoy WK and Woolfolk Hoy A (2000)
Collective teacher efficacy: its meaning, measure and
impact on student achievement. American Educational
Research Journal 37: 479±508.
Graham S and Weiner B (1996) Theories and principles of
motivation. In: Berliner DC and Calfee RC (eds) The
Handbook of Educational Psychology, pp. 63±84. New
York, NY: Macmillan.
Hattie J, Biggs J and Purdie N (1996) Effects of learning
skills interventions on student learning: a meta-
analysis. Review of Educational Research 66: 99±136.
Midgley C, Anderman EM and Hicks L (1995)
Differences between elementary and middle school
teachers and students: a goal theory approach. Journal of
Early Adolescence 15: 90±113.
Neath I (1998) Human Memory: An Introduction to
Research, Data, and Theory. Pacific Grove, CA:
Brooks-Cole Publishing.
Pajares F (1996) Self-efficacy beliefs in academic settings.
Review of Educational Research 66: 543±578.
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,
pp. 452±501. San Diego, CA: Academic Press.
Pressley M, Borkowski J and Schneider W (1989) Good
information processing: what is it and what education
can do to promote it. Journal of Experimental Child
Psychology 43: 194±211.
Pressley M and Ghatala ES (1990) Self-regulated learning:
monitoring learning from text. Educational Psychologist
25: 19±33.
Pressley M and Wharton-McDonald R (1997) Skilled
comprehension and its development through
instruction. School Psychology Review 26: 448±466.
Rosenshine B, Meister C and Chapman S (1996) Teaching
students to generate questions: a review of the
intervention studies. Review of Educational Research 66:
181±221.
Schraw G (1994) The effect of meta-cognitive knowledge
on local and global monitoring. Contemporary
Educational Psychology 19: 143±154.
1072 Self-regulated Learning
Schraw G and Moshman D (1995) Metacognitive
theories. Educational Psychology Review 7: 351±371.
Schunk D (1996) Goal and self-evaluative influences
during children's cognitive skill learning. American
Educational Research Journal 33: 359±382.
Weiner B (1986) An Attributional Theory of Motivation and
Emotion. New York, NY: Springer-Verlag.
Winne P and Perry N (2000) Measuring self-regulated
learning. In: Boekaerts M, Pintrich P and Zeidner M
(eds) Handbook of Self-Regulation, pp. 531±566. San Diego,
CA: Academic Press.
Zimmerman B (2000) Attaining self-regulated learning: a
social-cognitive perspective. In: Boekaerts M, Pintrich P
and Zeidner M (eds) Handbook of Self-Regulation,
pp. 13±39. San Diego, CA: Academic Press.
Zimmerman B and Kitsantas A (1999) Acquiring writing
revision skills: shifting from process to outcome self-
regulatory goals. Journal of Educational Psychology 91:
241±250.
Further Reading
Brown R and Pressley M (1994) Self-regulated reading
and getting meaning from text: the transactional
strategies instruction model and its ongoing validation.
In: Schunk DH and Zimmermann BJ (eds) Self-
regulation of Learning and Performance: Issues and
Educational Implications, pp. 155±180. Hillsdale, NJ:
Lawrence Erlbaum.
Ericsson KA (1996) The acquisition of expert
performance. In: Ericsson KA (ed.) The Road to
Excellence: The Acquisition of Expert Performance in the
Arts, Sciences, Sports, and Games, pp. 1±50. Mahwah, NJ:
Lawrence Erlbaum.
Ericsson KA and Kintsch W (1995) Long-term working
memory. Psychological Review 102: 211±245.
Hofer B, Yu S and Pintrich P (1998) Teaching college
students to be self-regulated. In: Schunk DH and
Zimmerman BJ (eds) Self-regulated Learning: From
Teaching to Self-reflective Practice, pp. 57±85. New York,
NY: Guilford.
Kintsch W (1998) Comprehension: A Paradigm for Cognition.
Cambridge, UK: Cambridge University Press.
Schunk DH and Ertmer PA (2000) Self-regulation and
academic learning: self-efficacy enhancing
interventions. In: Boekaerts M, Pintrich P and Zeidner
M (eds) Handbook of Self-regulation, pp. 631±649. San
Diego, CA: Academic Press.
Schunk DH and Zimmerman BJ (1998) Conclusions and
future directions for academic interventions. In: Schunk
DH and Zimmerman BJ (eds) Self-regulated Learning:
From Teaching to Self-reflective Practice, pp. 225±235. New
York, NY: Guilford.
Zeidner M, Boekaerts M and Pintrich P (2000) Self-
regulation: directions and challenges for future
research. In: Boekaerts M, Pintrich P and Zeidner M
(eds) Handbook of Self-regulation, pp. 750±768. San Diego,
CA: Academic Press.
Zimmerman B (1998) Academic studying and the
development of personal skill: a self-regulatory
perspective. Educational Psychologist 33: 73±86.
Self-stimulation
See Brain Self-stimulation;Reward, Brain Mechanisms of
______________________________________________________________________________
Semantic Analysis
See Latent Semantic Analysis;Syntax and Semantics: Formal Approaches
______________________________________________________________________________
Self-regulated Learning 1073
