The Power of Feedback (Review of Educational Research, Vol. 77, Issue 1) (2007)

The Power of Feedback
Author(s): John Hattie and Helen Timperley

Reviewed work(s):
Source: Review of Educational Research, Vol. 77, No. 1 (Mar., 2007), pp. 81-112
Published by: American Educational Research Association
Stable URL: http://www.jstor.org/stable/4624888 .
Accessed: 12/03/2013 15:55
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
American Educational Research Association is collaborating with JSTOR to digitize, preserve and extend
access to Review of Educational Research.
http://www.jstor.org
This content downloaded on Tue, 12 Mar 2013 15:55:12 PM

All use subject to JSTOR Terms and Conditions
Reviewof Educational
Research
March2007,Vol.77,No. 1,pp. 81-112
DOI:10.3102/003465430298487
The Power of Feedback
John Hattieand Helen Timperley

University of Auckland
Feedbackis one of the mostpowerfulinfluenceson learningand achieve-

ment, but this impactcan be eitherpositive or negative.Its power is fre-
quentlymentionedin articlesaboutlearningand teaching,butsurprisingly
few recentstudieshave systematicallyinvestigatedits meaning.Thisarticle
providesa conceptualanalysisoffeedbackand reviewstheevidencerelated
to its impacton learningandachievement.Thisevidenceshowsthatalthough
feedbackis amongthe majorinfluences,the typeoffeedbackand the way it
is givencan be differentiallyeffective.A modeloffeedbackis thenproposed
thatidentifiestheparticularpropertiesand circumstancesthatmakeit effec-
tive, and some typicallythornyissues are discussed,includingthe timingof
feedbackandtheeffectsofpositiveandnegativefeedback.Finally,thisanaly-
sis is usedto suggestwaysin whichfeedbackcan be usedto enhanceits effec-
tivenessin classrooms.
KEYWORDS:
feedback, assessment,studentand teacherlearning.
Although it is often mentionedin articlesaboutlearningand teaching, surpris-

ingly few recent studieshave systematicallyinvestigatedthe meaningof feedback
in classrooms. In this article,we begin with a conceptualanalysis of the meaning
of feedback and a synthesis of the evidence related to the power of feedback to
improveteaching and learning.We then propose a model of feedbackthat is used
to identify the circumstances under which feedback has the greatest impact.
Specifically, the researchevidence relatedto the differenttypes of feedback and
their effectiveness in terms of promotingstudentlearning are discussed, the dif-
ferent ways students deal with feedback are described, and the relationship
between assessment and feedback is provided. Finally, the model, togetherwith
the evidence underpinningit, is used to show how feedbackcan be used to enhance
classroom learningand teaching.
The Meaningof Feedback
In this review, feedbackis conceptualizedas informationprovidedby an agent
(e.g., teacher,peer, book, parent,self, experience)regardingaspects of one's per-
formance or understanding.A teacheror parentcan provide corrective informa-
tion, a peer can provide an alternativestrategy,a book can provide informationto
clarify ideas, a parentcan provide encouragement,and a learnercan look up the
answerto evaluatethe correctnessof a response.Feedbackthusis a "consequence"
of performance.
81

Hattie & Timperley
To assist in understandingthe purpose,effects, and types of feedback,it is use-
ful to considera continuumof instructionand feedback.At one end of the contin-
uum is a clear distinctionbetween providinginstructionand providingfeedback.
However, when feedbackis combined with more a correctionalreview, the feed-
backandinstructionbecome intertwineduntil"theprocessitself takeson the forms
of new instruction,ratherthan informing the student solely about correctness"
(Kulhavy, 1977, p. 212). To take on this instructionalpurpose,feedbackneeds to
provideinformationspecificallyrelatingto the task or process of learningthatfills
a gap between what is understoodand what is aimed to be understood(Sadler,
1989), andit can do this in a numberof differentways. These may be throughaffec-
tive processes, such as increasedeffort, motivation,or engagement.Alternatively,
the gap may be reducedthrougha numberof differentcognitive processes,includ-
ing restructuringunderstandings,confirming to studentsthat they are correct or
incorrect,indicatingthatmoreinformationis availableor needed,pointingto direc-
tions studentscould pursue,and/orindicatingalternativestrategiesto understand
particularinformation.Winne andButler(1994) providedan excellent summaryin
theirclaim that"feedbackis informationwith which a learnercan confirm,add to,
overwrite,tune, or restructureinformationin memory,whetherthatinformationis
domainknowledge,meta-cognitiveknowledge,beliefs aboutself andtasks,or cog-
nitive tactics and strategies"(p. 5740).
Feedbackhas no effect in a vacuum;to be powerfulin its effect, theremustbe a
learningcontextto whichfeedbackis addressed.It is butpartof the teachingprocess
and is thatwhich happenssecond-after a studenthas respondedto initial instruc-
tion-when informationis providedregardingsome aspect(s)of the student'stask
performance.It is most powerfulwhen it addressesfaultyinterpretations, not a total
lack of understanding.Underthe lattercircumstance,it may even be threateningto
a student:"If the materialstudied is unfamiliaror abstruse,providing feedback
shouldhave little effect on criterionperformance,since thereis no way to relatethe
new informationto what is alreadyknown"(Kulhavy, 1977, p. 220).
The focus of this article on feedback as informationabout the content and/or
understandingof the constructionsthatstudentshave madefromthe learningexpe-
rience is not the same as a behavioristinput-outputmodel. Contraryto the behav-
iorists' argument,Kulhavy(1977) demonstratedthatfeedbackis not necessarilya
reinforcer,because feedbackcan be accepted,modified, or rejected.Feedbackby
itself may not have the power to initiate furtheraction. In addition,it is the case
thatfeedbackis not only given by teachers,students,peers, and so on, but can also
be soughtby students,peers, and so on, and detectedby a learnerwithoutit being
intentionallysought.
The Effectiveness of Feedback
The first question to ask is, How effective is feedback?We answer this ques-
tion by referringto the usual effects of schooling on studentachievementandthen
comparingthem with the evidence related to feedback. Hattie (1999) reporteda
synthesis of over 500 meta-analyses,involving 450,000 effect sizes from 180,000
studies, representingapproximately20 to 30 million students, on various influ-
ences on studentachievement.This analysis includedmore than 100 factorsinflu-
encing educational achievement and covered various aspects of those typically
82

TABLE1
Summaryof effect sizesfrom 12 meta-analysesassessing the influencesoffeedback
Numberof Effect
Study Context effects size
Skiba,Casey,and Forspecialeducation
Center(1985-1986) students 35 1.24
Lysakowskiand
Walberg(1982) Cues,correctivefeedback 54 1.13
Walberg(1982) Cues,motivational
influences,andreinforcement 19 0.81
Tenenbaum and Cues,participation,
Goldring(1989) reinforcement,feedback,
andcorrectives 15 0.74
Rummeland
Feinberg(1988) Extrinsicfeedbackrewards 45 0.60
YeanyandMiller(1983) Diagnosticfeedbackin
science 49 0.52
KlugerandDe
Nisi (1996) Feedback 470 0.38
L'Hommedieu, Menges,
andBrinko(1990) Fromstudentratings 28 0.34
Moin(1986) Feedback 0.29
Bangert-Drowns,
Kulik,Kulik,and
Morgan(1991) Fromtesting 40 0.28
KulikandKulik(1988) Immediateversusdelayed 53 0.28
Getsie,Langer,and
Glass(1985) Rewardsandpunishments 89 0.14
Wilkinson(1981) Teacherpraise 14 0.12
identified, such as attributesof schools, homes, students,teachers,and curricula.

The averageor typical effect of schooling was 0.40 (SE = 0.05), and this provided
a benchmarkfigure or "standard"from which to judge the various influences on
achievement,such as thatof feedback.
At least 12 previousmeta-analyseshave includedspecific informationon feed-
back in classrooms(Table 1). These meta-analysesincluded 196 studiesand6,972
effect sizes. The averageeffect size was 0.79 (twice the averageeffect). To place
this averageof 0.79 into perspective,it fell in the top 5 to 10 highest influences on
achievementin Hattie's (1999) synthesis, along with directinstruction(0.93), rec-
iprocalteaching(0.86), students'priorcognitive ability(0.71), andalso can be con-
trastedwith otherinfluences such as acceleration(0.47), socioeconomic influences
(0.44), homework(0.41), the use of calculators(0.24), reducingclass size (0.12),
and retentionback 1 year (-0.12). Clearly,feedbackcan be powerful.
The effect sizes reportedin the feedback meta-analyses,however, show con-
siderable variability, indicating that some types of feedback are more powerful
83

TABLE2
Summaryof effect sizes relating tofeedback effects
Numberof Numberof Numberof Effect
Variable meta-analyses studies effects size
Cues 3 89 129 1.10
Feedback 74 4,157 5,755 0.95
Reinforcement 1 19 19 0.94
Videooraudiofeedback 1 91 715 0.64
Computer-assisted
instructional
feedback 4 161 129 0.52
Goalsandfeedback 8 640 121 0.46
Studentevaluationfeedback 3 100 61 0.42
Corrective
feedback 25 1,149 1,040 0.37
Delayedversusimmediate 5 178 83 0.34
Reward 3 223 508 0.31
Immediate versusdelayed 8 398 167 0.24
Punishment 1 89 210 0.20
Praise 11 388 4,410 0.14
Programmed instruction 1 40 23 -0.04
than others. Those studies showing the highest effect sizes involved students
receiving informationfeedback about a task and how to do it more effectively.
Lower effect sizes were relatedto praise,rewards,and punishment.
A more detailed synthesis of 74 meta-analysesin Hattie's (1999) databasethat
included some informationabout feedback (across more than 7,000 studies and
13,370 effect sizes, includingthose in Table 2) demonstratedthat the most effec-
tive forms of feedback provide cues or reinforcementto learners;are in the form
of video-, audio-, or computer-assistedinstructionalfeedback; and/or relate to
goals. Programmedinstruction,praise,punishment,andextrinsicrewardswere the
least effective for enhancingachievement(Table 3). Indeed,it is doubtfulwhether
rewardsshould be thoughtof as feedback at all. Deci, Koestner,and Ryan (1999)
described tangible rewards(stickers, awards, etc.) as contingencies to activities
ratherthan feedback because they contain such little task information.In their
meta-analysisof the effects of feedbackon motivation,these authorsfound a neg-
ative correlationbetweenextrinsicrewardsandtaskperformance(-0.34). Tangible
rewardssignificantlyunderminedintrinsicmotivation,particularlyfor interesting
tasks (-0.68) comparedwith uninterestingtasks (0.18). In addition,when the feed-
back was administeredin a controlling manner (e.g., saying that students per-
formed as they "should"have performed),the effects were even worse (-0.78).
Thus, Deci et al. concluded that extrinsic rewardsare typically negative because
they "underminepeople's takingresponsibilityfor motivatingor regulatingthem-
selves" (p. 659). Rather,they are a controllingstrategythat often leads to greater
surveillance,evaluation,and competition,all of which have been found to under-
mine enhancedengagementand regulation(Deci & Ryan, 1985).
The most systematicstudyaddressingtheeffects of varioustypesof feedbackwas
conductedby Kluger and DeNisi (1996). Their meta-analysisincluded studies of
84

TABLE3
Summaryof effect sizes relating to types offeedback
Moderator Numberof effects Effectsize
Correctfeedback
'Tiscorrect 114 0.43
'Tisincorrect 197 0.25
Taskfeedbackaboutchangesfrom
previoustrials
Yes 50 0.55
No 380 0.28
Taskfeedbackdesignedto
discouragethestudent
Yes 49 -0.14
No 388 0.33
Praisefeedbackaboutthetask
Yes 80 0.09
No 358 0.34
Feedbackprovidedfroma computer
Yes 87 0.41
No 337 0.23
Numberof timesfeedbackwas
provided
Lots 97 0.32
Little 171 0.39
Taskcomplexity
Verycomplex 107 0.03
Notcomplex 114 0.55
Goalsetting
Difficultgoals 37 0.51
Easy,do yourbestgoals 373 0.30
Threatto self-esteem
Muchthreat 102 0.08
Littlethreat 170 0.47
Source.KlugerandDeNisi(1996).
feedbackinterventionsthatwere not confoundedwith othermanipulations,included

at least a controlgroup,measuredperformance,andincludedat least 10 participants.
Manyof theirstudieswerenotclassroombased.Fromthe 131 studies,theyestimated
470 effect sizes on the basis of 12,652 participantsand23,663 observations(reflect-
ing multipleobservationsper participant).The averageeffect size was 0.38 (SE =
0.09), and 32% of the effects were negative (Table 3). Over all comparisons,it
appearsthatthe powerof feedbackis influencedby the directionof the feedbackrel-
ative to performanceon a task. Specifically,feedbackis moreeffective when it pro-
vides informationon correctratherthanincorrectresponsesand when it builds on
changesfromprevioustrails.The impactof feedbackwas also influencedby the dif-
ficulty of goals andtasks.It appearsto have the most impactwhen goals arespecific
85

Hattie& Timperley
and challengingbut task complexityis low. Praisefor task performanceappearsto
be ineffective, which is hardly surprisingbecause it contains such little learning-
relatedinformation.It appearsto be more effective when there are perceivedlow
ratherthanhigh levels of threatto self-esteem,presumablybecauselow-threatcon-
ditionsallow attentionto be paidto the feedback.
Given these mixed effects of feedback, we devote the remainderof this article
to identifying the conditions that maximize the positive effects on learning. A
model of feedback is used as a frameworkto understandwhy particularkinds of
feedback promotelearningeffectively and why some othersdo not.
A Modelof Feedback
Figure 1 presentsa frameworkin which feedbackcan be considered.The claim
is made thatthe main purposeof feedbackis to reducediscrepanciesbetween cur-
rent understandingsand performanceand a goal. Strategiesstudentsand teachers
use to reduce this discrepancymay be more or less effective in enhancinglearn-
ing, so it is importantto understandthe circumstancesthatresultin the differential
outcomes. Effective feedback must answer three major questions asked by a
teacherand/orby a student:Where am I going? (Whatare the goals?), How am I
going? (Whatprogressis being madetowardthe goal?), andWhereto next? (What
activities need to be undertakento make betterprogress?)These questions corre-
spondto notions of feed up, feed back, andfeed forward.How effectively answers
to these questionsserve to reducethe gap is partlydependenton the level at which
the feedback operates. These include the level of task performance,the level of
process of understandinghow to do a task,the regulatoryor metacognitiveprocess
level, and/or the self or personal level (unrelatedto the specifics of the task).
Feedbackhas differingeffects across these levels.
A key themearisingfromthis review of the literatureis the importanceof ensur-
ing thatfeedbackis targetedat studentsat the appropriatelevel, because some feed-
back is effective in reducingthe discrepancybetween currentunderstandingsand
what is desired,and some is ineffective. These issues areexploredin greaterdepth
as each aspect of the model is furtherexplored.
How FeedbackWorks:Reducingthe DiscrepancyBetweenCurrentand
DesiredUnderstanding
There are many possible ways for studentsto reduce the gap between current
anddesiredunderstandingsin responseto feedback,andthey arenot always effec-
tive in enhancing learning. Those likely to be effective include the following.
Students can increase their effort, particularlywhen the effort leads to tackling
more challengingtasks or appreciatinghigherqualityexperiencesratherthanjust
doing "more."We are more likely to increase effort when the intended goal "is
clear,when high commitmentis securedfor it, andwhen belief in eventualsuccess
is high" (Kluger & DeNisi, 1996, p. 260). Students may also develop effective
errordetection skills, which lead to their own self-feedback aimed at reachinga
goal. Such errordetectioncan be very powerful,providedstudentshave some mod-
icum of knowledge and understandingabout the task on which to strategize and
regulate.In addition,studentscan seek betterstrategiesto complete the task or be
taughtthem, or they can obtainmore informationfrom which they can then solve
problemsor use their self-regulatoryproficiencies.
86

Purpose
Toreducediscrepancies
betweencurrent anda desiredgoal
understandings/performance
The discrepancy can be reduced by:

Students
* Increased
effortandemployment of moreeffectivestrategiesOR
* Abandoning, or lowering
blurring, thegoals
Teachers
*
* Providingappropriate challenging
andspecificroals
Assistingstudentsto reachthemthrougheffectivelearning
strategiesandfeedback
Effective
feedbackanswersthreequestions
WhereamIgoing? (thegoals) Feed Up
HowamI going. Feed Back
Wheretonext? Feed Forward
Each feedback question works at four levels:
Task level Process level level Self level

Self-regulation
Howwelltasksare Themainprocessneeded Self-monitoring, Personalevaluations and
understood/performed to understand/perform and
directing, affect(usuallypositive)
tasks of actions
regulating aboutthelearner
FIGURE1. A modeloffeedbackto enhancelearning.
Some strategiesto reduce the gap are less productive.Studentsmay abandon

goals andthuseliminateany gap, andthis often leads to nonengagementin the pur-
suit of furthergoals (Bandura, 1982; Mikulincer, 1988; Steinberg, 1996). They
may choose to blurthe goals, combiningthem with so many othersthatafterper-
forming,they can pick and choose those goals they attainedand ignore the others.
Alternatively,studentscan change the standardby setting less challenging goals,
acceptingperformancefar below theircapabilitiesas satisfactory.
There are also multiple ways teachers can assist in reducing the gap between
actual performanceand desired goal attainment.These include providing appro-
priatechallenging and specific goals. Specific goals are more effective than gen-
eral or nonspecific ones, primarilybecause they focus students' attention, and
feedbackcan be more directed(Locke & Latham,1984). The goals and associated
feedback are also more likely to include informationaboutthe criteriafor success
in attainingthem than more generalgoals.
Teachers can also assist by clarifying goals, enhancing commitment or
increasedeffort to reachingthem throughfeedback. Goals can also be made more
manageable by narrowingthe range of reasonable hypotheses (Sweller, 1990).
More generally, teachers can create a learning environment in which students
develop self-regulationand errordetection skills (Hattie,Biggs, & Purdie, 1996).
How feedback contributes to these processes depends largely on the focus of
feedback and the level to which it is directed. In the next section, we develop a
frameworkto assist in identifying the circumstanceslikely to result in the more
productive outcomes.
87

Addressingthe ThreeFeedbackQuestions
Effective teachingnot only involves impartinginformationandunderstandings
to students(or providingconstructivetasks, environments,and learning)but also
involves assessing and evaluatingstudents'understandingof this information,so
that the next teaching act can be matchedto the presentunderstandingof the stu-
dents.This "secondpart"is the feedbackpart,andit relatesto the threemajorques-
tions identified in Figure 1: Where am I going? How am I going? and Where to
next?These threequestionsaddressthe dimensionsof feed up, feed back, andfeed
forward.An ideal learningenvironmentor experienceoccurs when both teachers
and studentsseek answersto each of these questions.Too often, teacherslimit stu-
dents' opportunitiesto receive informationabouttheir performancein relationto
any of these questionsby assumingthatresponsibilityfor the studentsandnot con-
sideringthe learningpossibilities for themselves.
WhereAm I Going?
A criticalaspectof feedbackis the informationgiven to studentsandtheirteach-
ers aboutthe attainmentof learninggoals relatedto the taskor performance.These
goals can be wide rangingandincludeitems such as singing a song, runninga race,
noting beauty in a painting,sandinga piece of wood, or riding a bicycle. Judging
the success of goal attainmentmay occuron manydimensions.Thejudgmentsmay
be direct, such as "passing a test" or "completingan assignment";comparative,
such as "doingbetterthan Mary"or "doingbetterthan last time";social, such as
"notgetting a detention"or "seekingteacherapproval";engagementrelated,such
as "singinga song"or "runninga race";or automaticandtriggeredoutside of spe-
cific awareness, such as "doing well on a task" or "seeking more challenging
tasks."On this last type, Bargh, Gollwitzer, Lee-Chai, Barndollar,and Trotschel
(2001) demonstratedthat goals can promote goal-directedaction (e.g., achieve-
ment or cooperationon tasks), producepersistenceat task performancein the face
of obstacles, and favor the resumptionof disruptedtasks even in the presence of
more attractivealternatives.As Black and Wiliam (1998) concluded, "the provi-
sion of challenging assignments and extensive feedback lead to greaterstudent
engagementand higher achievement"(p. 13)
Goals may relateto specific attainmentsor understandingsor to differingqual-
ities of experience,andthey typicallyinvolve two dimensions:challengeandcom-
mitment. Challenging goals relate to feedback in two major ways. First, they
inform individuals
as to whattypeorlevel of performance is to be attainedso thattheycandirect
andevaluatetheiractionsandeffortsaccordingly.Feedbackallows themto
set reasonablegoals andto tracktheirperformancein relationto theirgoals
so that adjustmentsin effort, direction,and even strategycan be made as
needed.(Locke& Latham,1990,p. 23)
These levels of attainmentwe have termed"successcriteria,"andgoals without
clarityas to when andhow a student(andteacher)would know they were success-
ful are often too vague to serve the purposeof enhancinglearning.Second, feed-
back allows students(and/ortheirteachers)to set furtherappropriatelychallenging
88

ThePowerof Feedback
goals as the previousones areattained,thus establishingthe conditionsfor ongoing
learning.
The relationshipbetween feedback and goal-relatedchallenge is complex. If
feedback does not lead to reducingthe discrepancybetween currentunderstand-
ings and goals, studentsare likely to close the gap by overstatingtheircurrentsta-
tus or claiming various attributionsthat reduce effort and engagement.Feedback
cannotlead to a reductionin this discrepancyif the goal is poorly defined, because
the gap betweencurrentlearningandintendedlearningis unlikelyto be sufficiently
clear for studentsto see a need to reduce it (Earley, Northcraft,Lee, & Lituchy,
1990; Erez, 1977; Frost & Mahoney, 1976). An additionalproblemoccurs when
feedback is not directedtowardthe attainmentof a goal. Too often, the feedback
given is unrelatedto achieving success on critical dimensions of the goal. For
example, students are given feedback on presentation,spelling, and quantity in
writingwhen the criteriafor success require,say, "creatingmood in a story."Such
feedback is not effective in reducing the gap relatingto the intentionof creating
mood (Clarke,Timperley,& Hattie, 2003; Timperley & Parr,2005).When goals
have appropriatechallenge andteachersand studentsarecommittedto these goals,
a clearerunderstandingof the criteriafor success is likely to be shared.
Goals are more effective when studentssharea commitmentto attainingthem,
because they are more likely to seek and receive feedback (Locke & Latham,
1990). Teachersandparentsoften assumethatstudentssharea commitmentto aca-
demic goals, whereasthe reality is thatdeveloping this sharedcommitmentneeds
to be nurturedand built. Commitmentcan be induced by authorityfigures; peer
groups;competition;role models; public statementsabout intentions,incentives,
andrewards;punishment;and generalvalence and instrumentality(e.g., Bandura,
1986; Carroll,Houghton,Durkin,& Hattie,2001; Hollenbeck,Klein, O'Leary,&
Wright, 1989; Latham & Lee, 1986; Lee, Locke, & Latham, 1989; Locke &
Latham,1984). For example,Earleyand Kanfer(1985) showed thatmodelingcan
be effective by having participantswatch a film of either a high-performingstu-
dent or a low-performingstudentcompleting a class-schedulingtask. Following
this, studentswere assigned or encouragedto set difficult goals. Those who had
observed the high-performingstudentin the video had significantly higher com-
mitmentlevels thanthose who had observedthe low-performingrole model.
How Am I Going?
Answering this question involves a teacher (or peer, task, or self) providing
information relative to a task or performance goal, often in relation to some
expected standard,to prior performance,and/or to success or failure on a spe-
cific part of the task. This aspect of feedback could be termed the feed-back
dimension. Feedbackis effective when it consists of informationaboutprogress,
and/or about how to proceed. Students often seek informationabout "how they
are going," althoughthey may not always welcome the answers.Too often, atten-
tion to this question leads to assessment or testing, whereas this is not the fun-
damental conception underlyingthis question. "Tests"are but one method used
by teachers and studentsto addressthis question and, as discussed below, often
fail to convey feedback information that helps teachers and their students to
know how they are going.
89

Whereto Next?
Instructionoften is sequential,with teachers providing information,tasks, or
learningintentions;studentsattemptingtasks;and some subsequentconsequence.
Too often,the consequenceis moreinformation,moretasks,andmoreexpectations;
studentsthus learn that the answer to "Whereto next?"is "more."The power of
feedback,however, can be used to specifically addressthis questionby providing
information that leads to greater possibilities for learning. These may include
enhancedchallenges,moreself-regulationoverthe learningprocess,greaterfluency
andautomaticity,more strategiesandprocessesto workon the tasks,deeperunder-
standing,andmoreinformationaboutwhatis andwhatis not understood.This feed-
forwardquestioncan have some of the most powerfulimpactson learning.
Integratingthe ThreeQuestions
Ratherthan the above three questions working in isolation at each of the four
levels, they typically work together.Feedbackrelatingto "How am I going?"has
the power to lead to undertakingfurthertasksor "Whereto next?"relativeto a goal
"Wheream I going?"As Sadler (1989) convincingly argued,it is closing the gap
between where studentsareandwherethey areaimingto be thatleads to the power
of feedback.
The Focusof Feedback:The Four Levels

The focus of feedback is critically important,and in this article,we claim that
there are four majorlevels and that the level at which feedback is directedinflu-
ences its effectiveness. First, feedback can be about a task or product, such as
whetherwork is corrector incorrect.This level of feedbackmay includedirections
to acquiremore, different, or correct information,such as "You need to include
moreaboutthe Treatyof Versailles."Second, feedbackcan be aimedat the process
used to createa productor complete a task. This kind of feedbackis more directly
aimed at the processing of information,or learning processes requiringunder-
standing or completing the task. For example, a teacher or peer may say to a
learner,"You need to edit this piece of writingby attendingto the descriptorsyou
have used so the readeris able to understandthe nuances of your meaning,"or
"Thispage may make more sense if you use the strategieswe talkedaboutearlier."
Third,feedback to studentscan be focused at the self-regulationlevel, including
greaterskill in self-evaluationor confidenceto engage furtheron a task.Forexam-
ple, "You alreadyknow the key featuresof the opening of an argument.Check to
see whetheryou have incorporatedthem in your first paragraph."Such feedback
can have majorinfluences on self-efficacy, self-regulatoryproficiencies,and self-
beliefs about students as learners, such that the students are encouraged or
informedhow to betterand more effortlessly continue on the task. Fourth,feed-
back can be personalin the sense that it is directedto the "self,"which, we argue
below, is too often unrelatedto performanceon the task. Examples of such feed-
back include "You are a great student"and "That'san intelligent response, well
done."
Thus,thereis a distinctionbetweenfeedbackaboutthe task (FT), aboutthe pro-
cessing of the task (FP), aboutself-regulation(FR), and aboutthe self as a person
(FS). We argue that FS is the least effective, FR and FP are powerful in terms of
90

ThePowerof Feedback
deep processing and masteryof tasks, and FT is powerful when the task informa-
tion subsequentlyis useful for improving strategyprocessing or enhancing self-
regulation(which it too rarelydoes).
FeedbackAbout the Task
This level includes feedback about how well a task is being accomplished or
performed,such as distinguishingcorrectfrom incorrectanswers,acquiringmore
or differentinformation,andbuildingmore surfaceknowledge. This type of feed-
back is most common and is often called corrective feedback or knowledge of
results, and it can relateto correctness,neatness,behavior,or some othercriterion
related to task accomplishment.About 90% of teachers' questions (sometimes
written but typically verbal) in classrooms are aimed at this informationlevel
(Airasian, 1997). Teacherscommonly mix correctivefeedback with information
at the self level, which dilutes the power of the FT (e.g., "Good boy, that is cor-
rect";see Bennett & Kell, 1989). By itself, correctivefeedback can be powerful.
From various meta-analyses,Lysakowski and Walberg(1982) reportedan effect
size of 1.13, Walberg(1982) reported0.82, and Tenenbaumand Goldring(1989)
reported0.74, all of which are substantialeffects. Having correctinformationis a
pedestal on which the processing and self-regulationis effectively built.
FT is more powerful when it is about faulty interpretations,not lack of infor-
mation. If studentslack necessaryknowledge, furtherinstructionis more power-
ful thanfeedbackinformation.One of the problemswith feedbackat the task level
is that it often does not generalizeto othertasks. Thompson(1998), for example,
demonstratedthat improvementwas specific to the questions for which feedback
was providedand was not used to answerotherquestions.
Feedbackaimed to move studentsfrom task to processing and then from pro-
cessing to regulationis most effective. Too muchfeedbackwithina level may even
detractfrom performance.For example, FT that provides very specific informa-
tion about the correctnessof the minutiaeof tasks and is not also directedto the
processing requiredto complete the task can direct attentionbelow the level nec-
essary for high-level performanceand thus interfere with task accomplishment
(Kluger & DeNisi, 1996). In the case of a rotatinga "turtle"in a LOGO simula-
tion, Simmons and Cope (1993) found thatwhen FT was too specific, the students
were unable to build up an estimate of rotationby successive increments.When
the immediatefeedbackspecific to each performancewas reduced,strategieswere
promotedthat involved more of the problem elements representinghigher level
responses.Hence, too muchfeedbackonly at the task level may encouragestudents
to focus on the immediategoal and not the strategiesto attainthe goal. It can lead
to more trial-and-errorstrategies and less cognitive effort to develop informal
hypotheses aboutthe relationshipbetween the instructions,the feedback, and the
intendedlearning.
Winne andButler(1994) arguedthatthe benefitsof FT dependheavily on learn-
ers' (a) being attentiveto the varyingimportanceof the feedbackinformationdur-
ing studyof the task, (b) havingaccuratememoriesof those featureswhen outcome
feedback is provided at the task's conclusion, and (c) being sufficiently strategic
to generateeffective internalfeedbackaboutpredictivevalidities (e.g., Which fac-
torsboost my performance?).It is likely thatfeedbackat this tasklevel is most ben-
eficial when it helps studentsrejecterroneoushypotheses and provides cues as to
91

Hattie & Timperley
directionsfor searchingand strategizing.Such cues can sensitize studentsto the
competence or strategy informationin a task or situation (Harackiewicz, 1979;
Harackiewicz,Mabderlink,& Sansone, 1984).
Feedbackrelatingspecifically to the task can be conceived along a numberof
dimensions,such as high to low complexity,individualor groupperformance,and
writtenor numericnotations.Simple more than complex task performancebene-
fits from FT (Balzer, Doherty, & O'Connor, 1989). Similarly, simple ratherthan
complex FT tends to be more effective. Kulhavy,White, Topp, Chan,and Adams
(1985) provided students with reading passages and multiple-choice items with
increasinglycomplex feedbackprovided.Firstthey were given the correctanswer,
andthen they discussed the four incorrectresponses.Each sentenceof the passage
was subsequentlyreadandused to explain why one of the errorchoices was incor-
rect, andfinally the relevantsection of the passagein which the correctanswerwas
identified.The less complex feedbackthatprovidedthe correctanswerresultedin
higher levels of subsequenttask performancethan the more complex versions of
the feedback, for which the effect was small. It may be, the authorsconjectured,
that providing additional information about the incorrect responses actually
increased the likelihood that the error was remembered by the learner.
Alternatively, it may be that studentsprocessed extra feedback informationat a
more surfacelevel, because they did not perceive it as being directlyrelatedto the
issue of identifying a correctresponse. However, these results were mediatedby
the readers'confidence in their responses. Those with high response confidence,
who had little troubleunderstandingor interactingwith the material,were more
likely to make efficient use of the feedbackwhateverits complexity.
Feedback,particularlyat this task level, can be delivered and received in both
individualand groupsituations.When deliveredin groups,the feedbackmessages
may be confoundedby the perceptionsof relevance to oneself or to other group
members.For example, a studentmay interpretthe feedback as pertainingto him
or her or may interpretit as relatingto the groupas a whole or to otherindividuals
in the group. In these lattertwo situations,it is likely either to be diluted or to be
perceivedas irrelevantto the individualstudent'sperformance(Nadler,1979). The
effectiveness of FT in these situationsdepends very much on students' commit-
ment and involvement in the task and on their notions about whetherit relates to
theirperformance.
The effectiveness of marks or written comments has also been investigated.
There is considerableevidence that providingwrittencomments (specific FT) is
more effective than providinggrades (Black & Wiliam, 1998; Crooks, 1988). In
one of the earlyandinfluentialstudies,Page (1958) foundthatfeedbackin the form
of shortwrittencommentsratherthangradesalone significantlyimprovedthe test
performance of students in 74 classrooms (see also Cardelle & Corno, 1981;
Elawar & Corno, 1985; McLaughlin, 1974). R. Butler (1987) demonstratedthat
gradescan increaseinvolvement,but they do not affect performance(relativeto a
no-FT condition). She also showed (R. Butler, 1988) that feedbackthroughcom-
ments alone led to learninggains, whereasmarksalone or commentsaccompanied
by marksor giving praisedid not. She claimed thatsuch resultscalled in question
the whole classroom cultureof marks,grades, gold stars,merit awards,competi-
tion ratherthan personalimprovement.As will become a theme later in this arti-
cle, feedback thatmixes FS with FT is less effective thanFT by itself.
92

FeedbackAbout the Processing of the Task
FP is more specific to the processes underlyingtasks or relatingand extending
tasks. Such feedback concerns information about relations in the environment,
relations perceived by a person, and relations between the environmentand the
person's perceptions (Balzer et al., 1989). A surface understandingof learning
involves the acquisition, storing, reproduction,and use of knowledge and thus
relates more to FT. A deep understandingof learninginvolves the constructionof
meaning (understanding) and relates more to the relationships, cognitive
processes, and transference to other more difficult or untried tasks (Marton,
Dall'Alba, & Beaty, 1993; Purdie,Hattie,& Douglas, 1996; Sdilj6,1979; Watkins
& Regmi, 1992; Watkins,Regmi, & Astilla, 1991).
A majortype of FP relates to students' strategiesfor errordetection,thus pro-
viding oneself with feedback. Such errors may indicate failure and a need to
restrategize,to choose differentstrategies,to be more effective in applyingstrate-
gies, and/orto seek help. Whetherstudentsengage in errorcorrectionstrategies
following errordetection depends on their motivation to continue to pursue the
goal or to reduce the gap between currentknowledge and the goal. Carverand
Scheier (1981, 1982, 1990) arguedthatwhen a studentencountersan impediment
(such as detecting an error)while pursuinga goal, a reassessmentof the situation
is triggered.In consideringtheirreassessments,studentsestimatehow probableit
is thatthey can achieve theirgoals if they invest furthereffort, modify theirplans,
or both.
Feedback informationabout the processes underlyinga task also can act as a
cueing mechanismand lead to more effective informationsearch and use of task
strategies.Cues are most useful when they assist studentsin rejecting erroneous
hypotheses and provide direction for searching and strategizing (Earley, 1988;
Harackiewicz, 1979; Harackiewiczet al., 1984; Wood & Bandura, 1987). This
type of feedback is akin to D. L. Butler and Winne's (1995) notion of task valid-
ity feedback,which bringsto a learner'sattention"therelationshipbetween a cue,
such as the presence and use of an advance organizer,and the probabilityof suc-
cessful performance"(p. 262).
Feedbackat the process level appearsto be more effective thanat the task level
for enhancing deeper learning (e.g., Balzer et al., 1989). Earley et al. (1990)
claimed that "usingprocess feedback with goal setting appearsto be a direct and
powerful way of shaping an individual's task strategy,and using outcome feed-
back is a muchless efficient way of shapingstrategy"(p. 103). It needs to be noted,
however, that there can be a powerful interactiveeffect between feedback aimed
at improvingthe strategiesand processes and feedbackaimed at the more surface
task information. The latter can assist in improving task confidence and self-
efficacy, which in turnprovidesresourcesfor more effective and innovativeinfor-
mation and strategysearching(Earleyet al., 1990).
FeedbackAbout Self-Regulation
Self-regulationinvolves an interplaybetween commitment,control, and confi-
dence. It addressesthe way studentsmonitor,direct, and regulateactions toward
the learning goal. It implies autonomy, self-control, self-direction, and self-
discipline.Suchregulationinvolves "self-generatedthoughts,feelings, andactions
93

Hattie & Timperley
that are planned and cyclically adapted to the attainment of personal goals"
(Zimmerman,2000, p. 14) and can lead to seeking, accepting, and accommodat-
ing feedbackinformation.
Thereareat least six majoraspectsof FR thatmediatethe effectiveness of feed-
back, discussed in more detailbelow. By way of overview, these include the capa-
bility to createinternalfeedbackand to self-assess, the willingness to invest effort
into seeking and dealing with feedback information,the degree of confidence or
certaintyin the correctnessof the response, the attributionsabout success or fail-
ure, and the level of proficiency at seeking help.
Effective learnerscreateinternalfeedbackandcognitive routineswhile they are
engaged in academictasks. D. L. Butlerand Winne (1995) arguedthatfor
all self-regulatedactivities,feedbackis aninherentcatalyst.As learnersmon-
itortheirengagementwith tasks,internalfeedbackis generatedby the mon-
itoring process. That feedbackdescribesthe natureof outcomes and the
qualitiesof the cognitiveprocessesthatled to those states.We hypothesize
thatmoreeffectivelearnersdevelopidiosyncraticcognitiveroutinesfor cre-
atinginternalfeedbackwhiletheyareengagedwithacademictasks.(p. 245)
Less effective learnershave minimalself-regulationstrategies,andthey depend
much more on externalfactors(such as the teacheror the task) for feedback.They
rarelyseek or incorporatefeedbackin ways thatwill enhancetheirfuturelearning
or self-regulationstrategies.
Self-assessmentis a self-regulatoryproficiencythatis powerfulin selecting and
interpretinginformation in ways that provide feedback. There are two major
aspects of self-assessment: self-appraisal and self-management (Paris &
Winograd, 1990). Self-appraisalrelates to students' facility to review and evalu-
ate their abilities, knowledge states, and cognitive strategiesthrougha variety of
self-monitoringprocesses. Self-managementis the monitoringand regulatingof
students'ongoing behaviorthroughplanning,correctingmistakes, and using fix-
up strategies.When studentshave the metacognitiveskills of self-assessment,they
can evaluatetheirlevels of understanding,theireffort and strategiesused on tasks,
their attributions and opinions of others about their performance, and their
improvementin relationto theirgoals and expectations.They can also assess their
performancerelativeto others' goals and the global aspects of theirperformance.
As studentsbecome more experiencedat self-assessment,multipledimensionsof
performancecan be assessed (Paris & Cunningham,1996). Most important,stu-
dents know how and when to seek and receive feedbackfrom others.
Students' willingness to invest effort in seeking and dealing with feedback
informationrelates to the transactioncosts invoked at the self-regulatorylevel.
These transactioncosts include effort costs (the effort necessary for feedback
search), face costs (the evaluative effects of others on the individualfor seeking
feedback),andinferencecosts (the implicationsof inferentialerrorsresultingfrom
inaccurately interpretingfeedback; Ashford & Cummings, 1983; de Luque &
Sommer,2000). The benefit incurredto offset these costs is a reductionin the gap
between currentand desiredor expected performance.It is the existence of evalu-
ative uncertaintythat makes seeking feedback worth incurringthe related costs
(Trope, 1975, 1980). When the cost/benefit ratio becomes prohibitive,however,
people refrainfrom seeking feedback.
94

ThePowerof Feedback
Decisions about whether to seek feedback are not only governed by this
cost/benefit ratio. In general, feedback is psychologically reassuring,and people
like to obtain feedback about their performanceeven if it has no impact on their
performance(Ashford& Cummings, 1983, p. 277). It is important,therefore,not
to confuse feelings that feedback is desirablewith the question of whetherfeed-
back benefits performance.
The degree of confidence thatstudentshave in the correctnessof responsescan
affect receptivityto and seeking of feedback.KulhavyandStock (1989) noted that
if confidence or response certaintyis high and the response turnsout to be a cor-
rect one, little attentionis paid to the feedback. Feedbackhas its greatest effect
when a learnerexpects a response to be correct and it turnsout to be wrong. As
Kulhavyand Stock noted, "highconfidence errorsare the point at which feedback
shouldplay its greatestcorrectiverole, simply because the person studiesthe item
longerin an attemptto correctthe misconception"(p. 225). Conversely,if response
certaintyis low andthe responseturnsout to be wrong,feedbackis largelyignored.
In these circumstances,low confidence places "a studentin a position requiring
associative strategiesratherthan the integrationof new informationinto existing
structures.Under this condition, feedback should have minimal effect regardless
of whetheror not the responseis the correctone" (Kulhavy, 1977, p. 226). Further
instructionand informationare more effective thanfeedbackin this situation.
Feelings of self-efficacy are importantmediatorsin feedback situations.From
theirmajorreview, KlugerandDeNisi (1996) concludedthatfeedbackis effective
to the degree to which it directsinformationto enhancedself-efficacy andto more
effective self-regulation,such thatattentionis directedback to the task and causes
studentsto invest more effort or commitmentto the task. These authorsclaimed
that such feedback is likely "to yield impressive gains in performance,possibly
exceeding 1 sd" (p. 278).
Students'attributionsaboutsuccess or failurecan often have more impactthan
the realityof thatsuccess or failure.Therecan be deleteriouseffects on feelings of
self-efficacy and performancewhen studentsare unable to relate the feedback to
the cause of their poor performance.Unclear evaluative feedback, which fails to
clearly specify the groundson which studentshave met with achievementsuccess
or otherwise, is likely to exacerbatenegative outcomes, engenderuncertainself-
images, and lead to poor performance(Thompson, 1997, 1998, 1999; Thompson
& Richardson,2001). On the flip side, undeserved success feedback increases
outcome uncertainty and can lead to increases in self-handicapping strategies
(Smith, Snyder,& Handelsman,1982). As BerglasandJones (1978) claimed, self-
handicappingstems from a capricious, chaotic feedback reinforcementhistory,
suggestingthat"itis not thattheirhistoriesarepocketedwith repeatedfailure;they
have been amply rewarded,but in ways and on occasions that leave them deeply
uncertainaboutwhat the rewardwas for"(p. 407).
There is considerable evidence that feedback that attributesperformanceto
effort or ability increasesengagementand performanceon tasks (Craven,Marsh,
& Debus, 1991; Dohrn & Bryan, 1994). The impact of feedback about effort or
ability, however, may depend on circumstances. Schunk and Rice (1991), for
example, highlighted the need to explore such feedback over extended periods,
because studentsmay responddifferentlyto the feedback dependingon the stage
of task accomplishment.Effortfeedbackappearsto be crediblein the early stages
95

Hattie& Timperley
of learning,when studentsneed to expendeffort to succeed. As skills develop, and
success should requireless effort, ability feedback may become more credible.
Ability feedback,however, may detractfrom the learningfocus of goals. Mueller
and Dweck (1998), for example, conducteda series of six studieswith elementary
studentsand found thatstudentsprovidedwith ability feedbackwere more perfor-
mance than learningorientedand reportedpoorerperformanceand lower enjoy-
ment of tasks aftera failure.
Seeking help is a learnerproficiency,and many types of help-seekingbehavior
can be consideredaspects of self-regulation.A majordistinctionis made between
instrumentalhelp seeking (askingfor hintsratherthananswers)andexecutive help
seeking (asking for answers or direct help that avoids time or work; Nelson-Le
Gall, 1981, 1985; Ryan & Pintrich,1977). Higherlevels of instrumentalhelp seek-
ing lead to feedback at the self-regulationlevels, whereasexecutive help seeking
is more likely to relateto the task level and sometimesthe processinglevel. When
consideringhow to develop instrumentalhelp-seekingbehavior,it is importantto
keep in mind it is mediatedby emotionalfactors.Many studentsdo not seek help
because of perceived threatsto self-esteem or social embarrassment(Karabenick
& Knapp, 1991; Newman & Schwager, 1993).
FeedbackAbout the Self as a Person
We include a final level of feedbacknot because it is effective but because it is
often presentin class situationsandtoo often used insteadof FT, FP, or FR (Bond,
Smith, Baker, & Hattie, 2000). Personalfeedback, such as "Goodgirl" or "Great
effort," typically expresses positive (and sometimes negative) evaluations and
affect aboutthe student(Brophy, 1981). It usuallycontainslittle task-relatedinfor-
mationand is rarelyconvertedinto more engagement,commitmentto the learning
goals, enhanced self-efficacy, or understandingabout the task. FS can have an
impact on learningonly if it leads to changes in students' effort, engagement,or
feelings of efficacy in relation to the learning or to the strategiesthey use when
attemptingto understandtasks.The effects at the self level aretoo diluted,too often
uninformativeaboutperformingthe task, andtoo influencedby students'self-con-
cept to be effective. The informationhas too little value to resultin learninggains.
Praise addressedto studentsis unlikely to be effective, because it carrieslittle
information that provides answers to any of the three questions and too often
deflects attentionfromthe task.Variousmeta-analyseshave demonstratedits inef-
fectiveness. Wilkinson (1981) completed a meta-analysison teacher praise and
concluded that it bears little, if any, relationshipto studentachievement(overall
effect = 0.12). Klugerand DeNisi (1998) also reporteda similarlylow effect size
for praise (0.09) and found that no praise has a greaterimpact on achievement
(0.34).
It is important,however, to distinguish between praise that directs attention
away from the task to the self (because such praise has low informationvalue to
achievement and learning) and praise directed to the effort, self-regulation,
engagement, or processes relating to the task and its performance(e.g., "You're
really greatbecause you have diligently completedthis task by applyingthis con-
cept"). This lattertype of praisecan assist in enhancingself-efficacy and thus can
be converted by studentsback into impact on the task, and hence the effects are
96

ThePowerof Feedback
much greater.It seems likely, from reading these meta-analyses, however, that
reviewers do not always distinguishbetween praise as a reinforceror reward(for
which it has zero to limited effect on achievement) and praise accompaniedby
informationabout the processes or performance(which has more, but still lim-
ited, effect).
These concerns aboutpraise are not the same as claiming that studentsdo not
like to be praised;they do. Sharp(1985) reportedthat 26% of the adolescent stu-
dents in his samplepreferredto be praisedloudly andpublicly when they achieved
on an academic task, 64% preferredto be praisedquietly and privately,and only
10%preferredteachersto say nothingat all. Burnett(2002) andElwell andTiberio
(1994) reporteda similar percentageamong elementarystudents and found that
studentspreferredpraise for trying hardratherthan for having high ability (espe-
cially when the praise was public) and for achievementratherthan for behavior.
On the otherhand,praisedeliveredpublicly by a teachercan be perceivedas pun-
ishing by some studentsif deliveredin the presence of a peer groupthat does not
esteem school achievementas valuable(Brophy,1981;Carrollet al., 2001; Carroll,
Durkin,Hattie,& Houghton, 1997; White & Jones, 2000).
Praisemay be counterproductiveand have negative consequenceson students'
self-evaluations of their ability. Meyer, Bachman, Hempelmann, Ploger, and
Spiller (1979) andMeyer (1982) conducteda series of studiesrelatedto this issue.
These authorsdemonstratedthat older studentsperceived praise after success or
neutralfeedbackafterfailureas an indicationthatthe teacherperceivedtheirabil-
ity to be low. When given criticismafterfailureandneutralfeedbackaftersuccess,
they perceivedthatthe teacherhadestimatedtheirabilityto be high andtheireffort
low. The same effects were not evident for younger students,however, who per-
ceived praise after success as an indicationof high ability and criticism afterfail-
ure as a sign of low ability.
Partof the reason for the unpredictabilityof praise is that studentsoften adopt
reputationallenses to seek or evaluatefeedbackinformationaimed at the self level
(e.g., "Iwantto be seen as a good student,""Ido not want to be seen as a good stu-
dent"). Students do a lot of "in the head" comparisons (Goethals, Messick, &
Allison, 1991), and it is likely that such comparisons are selected, interpreted,
and/orbiased. Strengthsand positive performancesare seen as unique and self-
created,whereasweakness andnegativeperformancesare seen as common in oth-
ers and possibly caused by others (Campbell, 1986; Goethals, 1986; Klein, 2001;
Suls & Wan, 1987). Such reputationallenses and biases, unless they lead to more
investmentin the task or to the use of betterstrategiesto accomplishthe task, are
of low effectiveness (Carrollet al., 2001).
One of the difficulties with these in-the-headcomparisonsoccurs because they
are rarelytied to specific tasks but rathertend to be more generalized at the self
level, and thus they can be difficult for teachersto change (Craven, 1997; Hattie,
1992), althoughthey help explain why feedback directedat the self level is usu-
ally dissipatedandineffective in increasingengagementor understandingof tasks.
Typically, these strategies have a negative effect on learning (Hattie & Marsh,
1995) because they include or lead to self-handicapping,learnedhopelessness, or
social comparison.The relatedfeedbackitself is usually discountedor dismissed,
and goals of low challenge are adopted.
97

Usingthe FourLevelsand ThreeQuestionsto Untangle
SomeThornyFeedbackIssues
The model presented in Figure 1 is used to address four commonly debated
issues aboutfeedback:the timing of feedback,the effects of positive and negative
feedback, the optimal classroom use of feedback, and the role of assessment in
feedback.
The Timingof Feedback
Therehas been much researchon the timing of feedback,particularlycontrast-
ing immediateanddelayedfeedback.Most of this researchhas been accomplished
withoutrecognitionof the variousfeedbacklevels. For example, immediateerror
correction during task acquisition (FT) can result in faster rates of acquisition,
whereas immediateerrorcorrectionduringfluency building can detractfrom the
learningof automaticityand the associated strategiesof learning(FP). Similarly,
in theirmeta-analysisof 53 studies,KulikandKulik(1988) reportedthatat the task
level (i.e., testing situations), some delay is beneficial (0.36), but at the process
level (i.e., engaging in processing classroom activities), immediate feedback is
beneficial (0.28) (see also Bangert-Drowns, Kulik, Kulik, & Morgan, 1991;
Brackbill,Blobitt,Davlin, & Wagner,1963; Schroth& Lund, 1993; Sturges,1972,
1978; Swindell & Walls, 1993).
Another example demonstratingthat the effects of immediate feedback are
likely to be morepowerfulfor FT anddelayedfeedbackmorepowerfulfor FP was
provided by Clariana,Wagner, and Roher Murphy(2000). They found that the
effectiveness of delayed comparedwith immediatefeedback varied as a function
of the difficulty of items in their test of informationtaughtin a series of lessons.
The effect sizes from delayed feedback were -0.06 for easy items, 0.35 for
midrangeitems, and 1.17 for difficult items. These authorssuggestedthatdifficult
items are more likely to involve greaterdegrees of processing aboutthe task, and
delayed feedback provides the opportunityto do this, whereas easy items do not
requirethis processing and so delay is both unnecessaryand undesirable.
The Effects of Positive and Negative Feedback
Kluger and DeNisi (1996) noted that both positive and negative feedback can
have beneficialeffects on learning,andthe argumentpresentedin this articleis that
the untanglingof these effects dependsmore on the level at which the feedbackis
aimed and processed than on whetherit is positive or negative. Specifically, neg-
ative feedback is more powerful at the self level, and both types can be effective
as FT, but thereare differentialeffects relatingto commitment,masteryor perfor-
mance orientation,and self-efficacy at the FR level.
At the self level (FS), it has alreadybeen noted that no praiseis more effective
thanpraiseif accompaniedby FT. Furthermore,thereis much evidence to suggest
that negative feedback or disconfirmationcan be more potent than positive feed-
back or confirmationat this self level (Brockner,1979; Brunit,Huguet,& Monteil,
2000; Campbell& Fairey, 1985; Hattie, 1992; Janoff-Bulman& Brickman,1982;
Kinch, 1963, 1968;Okun& Sasfy, 1977;Shrauger& Sorman,1977). Swann(1985)
and Swann and Hill (1982) found that individualswill go to greatlengths to con-
firm their self-perceptionsby attendingmost closely to feedbackinformationthat
98

ThePowerof Feedback
fits theirview of the self and by tryingto arrangetheirenvironmentto acquirefur-
ther self-confirming evidence. Individualsalso tend to reject or ignore negative
accounts of their behaviorthat differ from their own (Greenwald,1980; Markus,
1977; Tesser & Campbell,1983) or invoke an externalframeof reference(Marsh,
1987, 1990).
At the self-regulationlevel, the commitmentto goals is a majormediatorof the
effectiveness of positive andnegativefeedback.Van-DijkandKluger(2000, 2001)
demonstratedthatpositive feedbackincreasesmotivationrelativeto negativefeed-
back for a task thatpeople "wantto do" and decreasesmotivationrelative to neg-
ative feedback for a task that people "haveto do." Thus, when we are committed
to a goal, we are more likely to learnas a function of positive feedback,but when
we undertakea task that we are not committedto (and hence have to do), we are
more likely to learn as a function of negative feedback (we need to be driven, in
the older motivation terminology). It is likely, however, that this effect is short
lived in thatit may lead to futuretask avoidancebehavior.
In circumstancesin which studentsare committed to the goals, feedback can
trigger
an internalcomparisonprocess,which determineshow individualsreactto
feedback.Upon receivingnegativefeedback,individualsbecomemoredis-
satisfiedwiththeirpreviousperformancelevel, set higherperformancegoals
for theirfutureperformance,and performat a higherlevel thanthose who
receive positive feedbackor no feedbackat all. (Podsakoff& Farh, 1989,
p. 62)
Positive feedback,however,can increasethe likelihoodthatstudentswill return
to or persistin an activity and self-reporthigherinterestin the activity (Deci et al.,
1999).
There is also an interactioneffect at this FR level between positive and nega-
tive feedback and the self-efficacy of students. Swann, Pelham, and Chidester
(1988) found that for highly self-efficacious students,feedback about initial suc-
cess may signify a talent or potential ability, which leads to bettercoping in the
face of disconfirmationfeedback. They relatedthe feedback to positive verifica-
tions of themselves as learners.As a consequence of disconfirmationfeedback,
highly self-efficacious people make more optimistic predictionsabout their per-
formance after initial failure than after initial success, and they seek specifically
unfavorablefeedbackto excel at the tasks.
For the low self-efficacious students, positive feedback about initial success
may confirmthatthey have deficiencies thatneed to be remedied,which can lead
to a varietyof reactions.One reactionmay be furtherengagementto remedythese
"deficiencies"to reach a passable level of performance,which would affordpro-
tection againstfailure.Alternatively,these studentsmay avoid tasks and feedback
following initial success, because such success signifies that they have already
reachedan adequatelevel of performance,and furthertests merely runthe risk of
disconfirmingthe (sometimes hardgained) favorableoutcome.
Disconfirmatoryfeedbackcan also have a negative impacton subsequentmoti-
vationandperformancefor low self-efficaciousstudents(Brockner,Derr,& Laing,
1987; Moreland& Sweeney, 1984). Kernis,Brockner,and Frankel(1989) argued
that low self-efficacious people are more likely to react to negative feedback by
99

Hattie& Timperley
experiencingnegative affect, exhibitingless motivationon a subsequenttask, and
attributingthe feedbackless to effort and more to ability.
At the task level (FT), it has alreadybeen noted how powerfulcorrectivefeed-
back is for enhancing learning, particularlywhen learning new skills or tasks.
Disconfirmationwith correctiveinformationcan be effective, but disconfirmation
withoutthis informationis of little use because it providesno informationregard-
ing what to do or how to respond next time (Breakwell, 1983; Weiner, 1974a,
1974b, 1977). It is acknowledgedthatFT can be ignoredby studentsif it is poorly
presentedor if the students'knowledge is insufficientto accommodateadditional
feedback information.Howie, Sy, Ford, and Vicente (2000) found that it was the
poor presentation(or lack of informationvalue in the feedback) ratherthan stu-
dents' faulty knowledge that more often explained the low power of some feed-
back information.
Feedback and Classrooms
This feedback model highlights the demands on teachers if they are to teach
effectively. First, they need to undertakeeffective instruction.To reiterate,feed-
back is what happenssecond, andto make the feedbackeffective, teachersneed to
make appropriatejudgmentsaboutwhen, how, andat whatlevel to provideappro-
priatefeedback and to which of the threequestionsit shouldbe addressed.
It is difficult to documentthe frequencyof feedback in classrooms, except to
note that it is low. Bond et al. (2000) intensively documentedthe daily life of 65
teachers(half who had passed nationalboardcertificationand half who had not).
Althoughfeedbackwas one of the variablesthatmost discriminatedbetweenthose
who did and did not pass certificationas "accomplished"teachers,the frequency
of FT was low in the classrooms of both groups(the most common form of feed-
back was praise).
When feedback is given, it is likely to be self related(FS) or at best corrective
task related(FT) and to be influencedby perceptionsof students'need. Teachers
give "poor"studentsmore praise (FS), and the little FR providedis typically neg-
ative (Blote, 1995). Teacherfeedbackto boys is more relatedto a lack of effort or
poor behavior, and feedback to girls is more about ability attributions(Dweck,
Davidson, Nelson, & Enna, 1978).
Feedbackis not only differentiallygiven but also differentiallyreceived (Diehl
& Sterman,1995; Paich & Sterman,1993; Sterman,1989). De Luqueand Sommer
(2000) found that studentsfrom collectivist cultures(e.g., Confucian-basedAsia,
South Pacific nations) preferred indirect and implicit feedback, more group-
focused feedback,andno self-level feedback.Studentsfrom individualistcultures
(e.g., the United States) preferredmore direct feedback particularlyrelated to
effort, were more likely to use directinquiryto seek feedback,and preferredmore
individualfocused self-relatedfeedback.
The climate of the classroomis critical,particularlyif disconfirmationand cor-
rective feedback at any level is to be welcomed and used by the students (and
teachers).Errorsand disconfirmationare most powerfulin climates in which they
are seen as leading to futurelearning,particularlyrelatingto processingand regu-
lation. Studentengagementin learningis likely to be constrainedby the evaluative
dimensions of classroom lessons because there is personal risk involved in
respondingpublicly and failing. Too often, the level of risk is determinedby the
100

ThePowerof Feedback
likelihood that a studentcan supply an answer and by the accountabilityclimate
set up by the teacherand otherstudents(Alton-Lee & Nuthall, 1990, 1998; Doyle,
1983). Typically, students respond only when they are fairly sure that they can
respond correctly,which often indicates they have alreadylearnedthe answer to
the questionbeing asked. Errors,and learningfrom them, are rarelywelcomed.
Simply providingmore feedback is not the answer, because it is necessary to
consider the natureof the feedback,the timing, and how a student"receives"this
feedback (or, better,actively seeks the feedback). As alreadynoted, studentscan
bias and select feedback information.The ways and mannerin which individuals
interpretfeedbackinformationis the key to developing positive and valuablecon-
cepts of self-efficacy about learning, which in turns leads to furtherlearning.
Teachers need to view feedback from the perspectiveof the individualsengaged
in the learningandbecome proactivein providinginformationaddressingthe three
feedback questions and developing ways for students to ask these questions of
themselves. Students,too often, view feedback as the responsibilityof someone
else, usually teachers,whose job it is to providefeedbackinformationby deciding
for the studentshow well they are going, what the goals are, and what to do next.
Feedback and Assessment
Therearemajorimplicationsfromthis review of feedbackfor assessmentin the
classroom. Assessment can be considered to be activities that provide teachers
and/orstudentswith feedbackinformationrelatingto one or moreof the threefeed-
back questions (at the FT, FP, or FR level). Such a definitionplaces emphasison
devising assessment tasks that provide informationand interpretationsabout the
discrepancybetween currentstatusand the learninggoals at any of the three lev-
els: abouttasks, aboutthe processes or strategiesto understandthe tasks, andabout
the regulation,engagement,and confidence to become more committedto learn.
This contrasts with the more usual definition of assessment, an activity used to
assess students'levels of proficiency.This usual definitionplaces more emphasis
on the adequacyof scores (and less on the interpretationof these scores). Crooks
(1988) andBlack andWiliam (1998) demonstratedthereis little evidence thatsuch
classroomtestinghas assistedin the learningprocess.Black andWiliam,for exam-
ple, reviewed 578 publicationsrelatingto the role of assessment in learning and
concluded thatclassroom assessment
typicallyencouragessuperficialandrotelearning,concentratingon recallof
isolateddetails,usuallyitemsof knowledgewhichpupilssoonforget... teach-
ers do not generallyreviewthe assessmentquestionsthattheyuse anddo not
discussthemcriticallywithpeers,so thereis littlereflectionon whatis being
assessed.(p. 17)
Too often, the power of assessment feedback is aimed to "drive" students
toward (often unspecified) goals or to "do more"or "do better."Studentsreceive
little feedback informationin these instances, primarilybecause the assessment
feedback does not addressthe three majorquestions, and rarely does such feed-
back enhance the processes (FP) and metacognitive attributes(FR) of the task.
Furthermore,teachers too often see assessment feedback as making statements
about students,not abouttheirteaching (Timperley& Wiseman, 2002). Thus, the
benefits of feedbackin the classroomfrom such testing are often diluted.
101

Hattie& Timperley
Therearemanyways in which teacherscan deliverfeedbackto studentsandfor
studentsto receive feedbackfrom teachers,peers, and othersources.The implica-
tion is not thatwe shouldautomaticallyuse moretests (Bangert-Drowns,Kulik, &
Kulik, 1991). Rather,for students, it means gaining informationabout how and
whatthey understandandmisunderstand,findingdirectionsandstrategiesthatthey
must take to improve,and seeking assistanceto understandthe goals of the learn-
ing. For teachers,it means devising activities and questionsthatprovidefeedback
to them about the effectiveness of their teaching, particularlyso they know what
to do next. Assessments can performall these feedback functions, but too often,
they are devoid of effective feedbackto studentsor to teachers.
Conclusions
Feedbackis informationprovidedby an agent (e.g., teacher,peer,book, parent,
experience) regardingaspects of one's performanceor understanding.It occurs
typically afterinstructionthatseeks to provideknowledge and skills or to develop
particularattitudes.The model proposedin this articleidentifies threemajorfeed-
back questions: Where am I going? How am I going? and Where to next? The
answersto these questionsenhancelearningwhen there is a discrepancybetween
whatis understoodandwhatis aimedto be understood.It can increaseeffort,moti-
vation,or engagementto reducethis discrepancy,and/orit can increasecue search-
ing and task processes thatlead to understanding(thusreducingthis discrepancy).
Feedbackis among the most critical influences on studentlearning.A majoraim
of the educativeprocess is to assist in identifyingthese gaps ("How am I going?"
relative to "Wheream I going?") and to provideremediationin the form of alter-
native or other steps ("Whereto next?").
The model discriminatesbetween four levels of feedback: the task, the pro-
cessing, the regulatory,and the self levels. Effective feedbackat the task, process,
and self-regulatorylevels is interrelated.FT is more powerfulwhen it resultsfrom
faulty interpretations,not a lack of understanding.It is most effective when it aids
in building cues and informationregardingerroneoushypothesis and ideas and
then leads to the development of more effective and efficient strategiesfor pro-
cessing and understandingthe material.Feedbackat the process level is most ben-
eficial when it helps students reject erroneoushypotheses and provides cues to
directionsfor searchingand strategizing.Such cues sensitize studentsto the com-
petence or strategyinformationin a task or situation.Ideally, it moves from the
task to the processes or understandingsnecessary to learn the task to regulation
aboutcontinuingbeyondthe taskto morechallengingtasksandgoals. This process
results in higher confidence and greaterinvestmentof effort. This flow typically
occurs as studentsgain greaterfluency and mastery.Feedbackthatattendsto self-
regulation is powerful to the degree that it leads to furtherengagement with or
investing furthereffort into the task, to enhancedself-efficacy, and to attributions
that the feedback is deserved and earned.When feedback draws attentionto the
regulatory processes needed to engage with a task, learners' beliefs about the
importanceof effort and their conceptions of learningcan be importantmodera-
tors in the learningprocess.
Feedback at the self or personal level (usually praise), on the other hand, is
rarelyeffective. Praiseis rarelydirectedat addressingthe threefeedbackquestions
and so is ineffective in enhancinglearning.When feedbackdrawsattentionto the
102

ThePowerof Feedback
self, studentstry to avoid the risks involved in tackling challenging assignments,
to minimize effort, andhave a high fear of failure(Black & Wiliam, 1998) to min-
imize the risk to the self.
The three feedback questions are certainly not linearly interpretedor imple-
mented, and the boundariesbetween them are fuzzy. Although it is importantto
know aboutgoals, learningexperiencesdo not necessarilybegin by asking "What
are the goals?"because these can be discovered(usuallyin more specific ways) as
we undertakeparticulartasks. Goals can be many and sometimes competing, and
much of the learningthataccruescan lead to creatingoptions to achieve the goals,
weighing the pros and cons of options, considering the likelihood that a given
course of action will lead to the goals, and learningaboutand evaluatingthe con-
sequences of achieving the goals. Thus, goals may be constantlyat issue, and the
feedback about "How am I going?" can help in these evolving goal-relatedcon-
siderations.Similarly,the answerto "Whereto next?"may be nowhere,if the goal
is unchanging,the "outcome"is furtherengagementwith the same or similartasks,
or the studentbelieves that the answer is "whereverthe teacher tells me to go."
Such reactionstypicallyindicatelow self-regulationor overly dominantclassroom
regimes. The answerto "Whereto next?"needs to be more directedto the refine-
ment and seeking of more challenginggoals, because these have the highest like-
lihood of leading to greaterachievement.
It should be clear thatprovidingand receiving feedbackrequiresmuch skill by
studentsand teachers.The model advancedin this article does not merely invoke
a stimulus-and-responseroutine but requires high proficiency in developing a
classroomclimate, the abilityto deal with the complexities of multiplejudgments,
and deep understandingsof the subject matterto be ready to provide feedback
about tasks or the relationships between ideas, willingness to encourage self-
regulation, and having exquisite timing to provide feedback before frustration
takesover. To be able to devote time andthoughtsto feedbackis aidedwhen teach-
ers automatemany other tasks in the classroom and provide rich learningoppor-
tunities for all studentsand thus have the time and resourcesto be responsive to
feedback (Hattie& Jaeger, 1998).
The model firmly identifies thatfeedback involves both the giving and receiv-
ing (by teachersand/orby students),andtherecan be gulfs between these. Students
constructtheir worlds of learning and classrooms, and it is a major argumentof
this article that it is crucial for teachersto understandand appreciatethat provid-
ing feedbackis only a partof the equation.Similarly,some tasks more thanothers
can lead to moreeffective feedbackby teachers,students,or both. Learningcan be
enhancedto the degree thatstudentssharethe challenginggoals of learning,adopt
self-assessmentand evaluationstrategies,and develop errordetectionprocedures
and heightened self-efficacy to tackle more challenging tasks leading to mastery
and understandingof lessons. Students'self strategiesandhelp seeking can medi-
ate whetherthese effects occur. Studentswho wish to confirmpositive self-belief
ratherthan focus on learninggoals are more likely to adopt or seek feedback that
maximizespositive self-evaluationsand/orminimizesnegative self-evaluations.A
number of self strategies were identified that inhibit the effects of feedback on
learning,and it is only when studentsare groundedin and committedto the goals
of learning and when the feedback is relatedto accomplishmentsof the learning
thatfeedbackis effective (Crocker& Wolfe, 2001). A majortask for teachersand
103

Hattie& Timperley
parentsis to make academic goals salient for all students,because studentswho
are preparedto question or reflect on what they know and understandare more
likely to seek confirmatoryand/or disconfirmatoryfeedback that allows for the
best opportunitiesfor learning.
Feedback,however, is not "the answer";rather,it is but one powerful answer.
With inefficient learners,it is betterfor a teacherto provide elaborationsthrough
instructionthanto providefeedbackon poorly understoodconcepts. If feedbackis
directedat the right level, it can assist studentsto comprehend,engage, or develop
effective strategiesto process the informationintendedto be learned.To be effec-
tive, feedbackneeds to be clear,purposeful,meaningful,and compatiblewith stu-
dents' priorknowledge andto providelogical connections.It also needs to prompt
active informationprocessing on the partof learners,have low task complexity,
relate to specific and clear goals, and provide little threatto the person at the self
level. The major discriminator is whether it is clearly directed to the task,
processes, and/orregulationand not to the self level. These conditions highlight
the importance of classroom climates that foster peer and self-assessment and
allow for learningfrom mistakes.
There are majorimplicationsfor the design of assessments.Too often, assess-
ments are used to providesnapshotsof learningratherthanprovidinginformation
that can be used by studentsor their teachersto addressthe three feedback ques-
tions. Certainly,a critical conclusion is that teachersneed to seek and learn from
feedback (such as from students'responses to tests) as much as do students,and
only when assessmentprovides such learningis it of value to either.Most current
assessments provide minimal feedback, too often because they rely on recall and
are used as externalaccountabilitythermometersratherthan as feedback devices
that are integralto the teaching and learningprocess. It is the feedback informa-
tion and interpretationsfrom assessments,not the numbersor grades,that matter.
In too many cases, testing is used as the measure to judge whether change has
occurredratherthan as a mechanismto furtherenhance and consolidate learning
by teachersor students.The costs of these thermometer-related accountabilitytests
are high, and the feedbackreturnsare minimal (Shepardet al., 1996).
On the other hand, when feedback is combined with effective instructionin
classrooms, it can be very powerful in enhancinglearning.As Klugerand DeNisi
(1996) noted, a feedbackinterventionprovidedfor a familiartask,containingcues
thatsupportlearning,attractingattentionto feedback-standarddiscrepanciesat the
task level, and void of cues thatdirectattentionto the self is likely to yield impres-
sive gains in students' performance.It is importantto note, however, that under
particularcircumstances,instructionis moreeffective thanfeedback.Feedbackcan
only build on something;it is of little use when there is no initial learningor sur-
face information.Feedbackis what happens second, is one of the most powerful
influences on learning,too rarelyoccurs, andneeds to be more fully researchedby
qualitativelyandquantitativelyinvestigatinghow feedbackworksin the classroom
and learningprocess.
References
Airasian,P. W. (1997). Classroomassessment(3rded.). New York:McGraw-Hill.
Alton-Lee,A., & Nuthall,G. (1990). Researchon teachingandlearning:Thirtyyears
of change.ElementarySchoolJournal,90(5), 547-570.
104

ThePowerof Feedback
Alton-Lee,A., & Nuthall,G. (1998).Inclusiveinstructionaldesign:Theoreticalprin-
ciples emergingfrom the Understanding Learningand TeachingProject(Reportto
the Ministry of Education,UnderstandingLearningand Teaching Project 3).
Wellington,New Zealand:Ministryof Education,ResearchDivision.
Ashford, S. J., & Cummings,L. L. (1983). Feedbackas an individualresource:
Personalstrategiesof creatinginformation.OrganizationalBehaviorand Human
Performance,32, 370-398.
Balzer,W. K., Doherty,M. E., & O'Connor,R., Jr.(1989).Effectsof cognitivefeed-
backon performance.PsychologicalBulletin,106(3),410--433.
Bandura,A. (1982).Self efficacymechanismin humanagency.AmericanPsychologist,
37, 122-147.
Bandura,A. (1986). Socialfoundationsof thoughtand action:A social cognitivethe-
ory. EnglewoodCliffs, NJ:PrenticeHall.
Bangert-Drowns, R. L., Kulik,J. A., & Kulik,C. C. (1991). Effectsof frequentclass-
roomtesting.Journalof EducationalResearch,85(2), 89-99.
Bangert-Drowns, R. L., Kulik,C. L., Kulik,J.A., & Morgan,M. T. (1991).Theinstruc-
tionaleffect of feedbackin test-likeevents.Reviewof EducationalResearch,61,
213-237.
Bargh,J. A., Gollwitzer,P. M., Lee-Chai,A., Barndollar,K., & Trotschel,R. (2001).
The automatedwill: Nonconsciousactivationand pursuitof behavioralgoals.
Journalof Personalityand SocialPsychology,81(6), 1014-1027.
Bennett,N., & Kell, J. (1989).A good start?Fouryear olds in infantschools.Oxford,
UK:Blackwell.
Berglas,S., & Jones,E. (1978).Drugchoiceas a self-handicapping strategyin response
to noncontingent success.Journalof PersonalityandSocialPsychology,36, 405-417.
Black, P., & Wiliam,D. (1998). Assessmentand classroomlearning.Assessmentin
Education,5(1), 7-75.
Blote,A. W. (1995). Students'self-conceptin relationto perceiveddifferentialteacher
treatment.Learning& Instruction,5(3), 221-236.
Bond,L., Smith,R., Baker,W. K., & Hattie,J. A. (2000) Certificationsystemof the
NationalBoardfor ProfessionalTeachingStandards:A constructand consequen-
tial validity study. Washingtonl,DC: National Board for ProfessionalTeaching
Standards.
Brackbill,Y., Blobitt, W. E., Davlin, D., & Wagner,J. E. (1963). Amplitudeof
responseandthedelay-retention effect.Journalof Experimental Psychology,66(1),
57-64.
Breakwell,G. M. (1983). Formulationsand searchers.In G. M. Breakwell(Ed.),
Threatenedidentities(pp. 3-26). Chichester,UK:Wiley.
Brockner,J. (1979).Theeffectsof self-esteem,success-failure,andself-consciousness
on taskperformance. Journalof PersonalityandSocialPsychology,37, 1732-1741.
Brockner,J., Derr,W. R., & Laing,W. N. (1987). Self-esteemandreactionsto nega-
tive feedback:Towardsgreatergeneralizability. Journalof Researchin Personality,
21, 318-334.
Brophy, J. (1981). Teacherpraise: A functionalanalysis. Review of Educational
Research,51, 5-32.
Brunit,S., Huguet,P., & Monteil,J. M. (2000).Performance feedbackandself-focused
attentionin the classroom:Whenpast and presentinteract.Social Psychologyof
Education,3, 277-293.
Burnett,P. C. (2002). Teacherpraiseand feedbackand students'perceptionsof the
classroomenvironment.EducationalPsychology,22(1), 1-16.
105

Hattie& Timperley
Butler,D. L., & Winne,P. H. (1995). Feedbackand self-regulatedlearning:A theo-
reticalsynthesis.Reviewof EducationalResearch,65(3), 245-274.
Butler,R. (1987). Task-involvingandego-involvingpropertiesof evaluation:Effects
of differentfeedbackconditionson motivationalperceptions,interestandperfor-
mance.Journalof EducationalPsychology,79,474-482.
Butler,R. (1988).Enhancingandundermining intrinsicmotivation:Theeffectsof task-
involvingandego-involvingevaluationon interestandperformance. BritishJournal
of EducationalPsychology,58, 1-14.
Campbell,J. D. (1986). Similarityanduniqueness:The effectsof attributetype, rele-
vance,andindividualdifferencesin self-esteemanddepression. Journalof Personality
andSocialPsychology,50, 281-294.
Campbell,J.D., & Fairey,P. J. (1985).Effectsof self-esteem,hypothetical
explanations,
andverbalization of expectancieson futureperformance. Journalof Personalityand
SocialPsychology,48, 1097-1111.
Cardelle,M., & Corno,L. (1981). Effectson secondlanguagelearningof variationsin
writtenfeedbackon homeworkassignments.TESOLQuarterly,15(3), 251-261.
Carroll,A., Durkin,K., Hattie,J., & Houghton,S. (1997). Goalsettingamongadoles-
cents: A comparisonof delinquent,at-risk, and not at-risk youth. Journal of
EducationalPsychology,89, 441-450.
Carroll,A., Houghton,S., Durkin,K., & Hattie,J. (2001).Reputationenhancinggoals:
Integratingreputationenhancementand goal settingtheoryas an explanationof
delinquentinvolvement.In F. Columbus(Ed.), Advancesin psychologyresearch
(Vol. 4, pp. 101-129). New York:Nova Science.
Carver,C. S., & Scheier,M. F. (1981).Attentionand self regulation:A controltheory
to humanbehavior.New York:Springer-Verlag.
Carver,C. S., & Scheier,M. F. (1982).Controltheory:A usefulconceptualframework
for personality-social,clinical,andhealthpsychology.PsychologicalBulletin,92,
111-135.
Carver,C. S., & Scheier,M. F. (1990). Originsandfunctionof positiveandnegative
affect:A control-processview. PsychologicalReview,97, 19-35.
Clariana,R. B., Wagner,D., & RoherMurphy,L. C. (2000).Applyinga connectionist
descriptionof feedbacktiming.EducationalTechnologyResearchandDevelopment,
48(3), 5-21.
Clarke,S., Timperley,H., & Hattie,J. A. (2003). Assessingformativeassessment.
Auckland,New Zealand:HodderMoa Beckett.
Craven,R. G. (1997). Enhancingacademicself-concept:A large-scalelongitudinal
studyin aneducationalsetting.DissertationAbstractsInternational, A (Humanities
and Social Sciences),58(5-A), 1577.
Craven,R. G., Marsh,H. W., & Debus,R. L. (1991).Effectsof internallyfocusedfeed-
backandattributional feedbackon enhancementof academicself-concept.Journal
Crocker,J., & Wolfe,C. T. (2001).Contingenciesof self-worth.PsychologicalReview,
108(3), 593-623.
Crooks,T. J. (1988). The impact of classroomevaluationon students.Review of
EducationalResearch,5, 438-481.
Deci, E. L., Koestner,R., & Ryan,M. R. (1999).A meta-analytic reviewof experiments
examiningthe effects of extrinsicrewardson intrinsicmotivation.Psychological
Bulletin,125, 627-668.
Deci, E. L., & Ryan,R. (1985). Intrinsicmotivationand self-determination in human
behavior.New York:Plenum.
106

ThePowerof Feedback
de Luque,M. F., & Sommer,S. M. (2000).Theimpactof cultureon feed-back-seeking
behavior:An integratedmodelandpropositions.Academyof ManagementReview,
25(4), 829-849.
Diehl, E., & Sterman,J. D. (1995). Effectsof feedbackcomplexityon dynamicdeci-
sion making. OrganizationalBehavior and Human Decision Processes, 62,
198-215.
Dohrn,E., & Bryan,T. (1994).Attribution instruction.TeachingExceptionalChildren,
26(4), 61-63.
Doyle, W. (1983). Academicwork.Reviewof EducationalResearch,53, 159-199.
Dweck,C. S., Davidson,W., Nelson,S., & Enna,B. (1978). Sex differencesin learned
helplessness:II. The contingenciesof evaluativefeedbackin the classroomandIII.
An experimentalanalysis.DevelopmentalPsychology,14(3), 268-276.
Earley,P. C. (1988). Computer-generated performancefeedbackin the magazine-
subscriptionindustry.OrganizationalBehaviorand HumanDecision Processes,
41, 50-64.
Earley,P. C., & Kanfer,R. (1985). Theinfluenceof componentparticipationandrole
models on goal acceptance,goal satisfaction,and performance.Organizational
Behavior& HumanDecisionProcesses,36(3), 378-390.
Earley,P. C., Northcraft,G. B., Lee, C., & Lituchy,T. R. (1990). Impactof process
andoutcomefeedbackon the relationof goal settingto taskperformance. Academy
of ManagementJournal,33(1), 87-105.
Elawar,M. C., & Corno,L. (1985). A factorialexperimentin teachers'writtenfeed-
back on studenthomework:Changingteacherbehavioura little ratherthana lot.
Journalof EducationalPsychology,77, 162-173.
Elwell, W. C., & Tiberio,J. (1994). Teacherpraise:Whatstudentswant.Journalof
InstructionalPsychology,21(4), 322-328.
Erez,M. (1977). Feedback:A necessaryconditionfor the goal setting-performance
relationship.Journalof AppliedPsychology,62, 624-627.
Frost,P. J., & Mahoney,T. A. (1976). Goal settingandthe taskprocess:An interac-
tive influence on individualperformance.OrganizationalBehaviorand Human
Performance,16, 250-279.
Getsie,R. L., Langer,P., & Glass,G. V. (1985). Meta-analysisof the effects of type
and combinationof feedback on children's discriminationlearning.Review of
EducationalResearch,55(1), 9-22.
Goethals,G. R. (1986). Fabricatingandignoringsocialreality:Self-servingestimates
of consensus.In J. M. Olson,C. P. Hermann,& M. P. Zanna(Eds.),Relativedepri-
vation and social comparison:The OntarioSymposium(Vol. 4, pp. 135-157).
Hillsdale,NJ:LawrenceErlbaum.
Goethals,G. R., Messick,D. M., & Allison,S. T. (1991).Theuniquenessbias:Studies
of constructivesocial comparison.In J. Suls & T. A. Wills (Eds.),Social compari-
son research: Contemporarytheoryand research (pp. 149-176). Hillsdale,NJ:
LawrenceErlbaum.
Greenwald,A. G. (1980). The totalitarianego: Fabricationandrevision of personal
history.AmericanPsychologist,35, 603-618.
Harackiewicz,J. M. (1979). Theeffectsof rewardcontingencyandperformancefeed-
back on intrinsicmotivation.Journalof Personality& Social Psychology,37(8),
1352-1363.
Harackiewicz,J. M., Mabderlink,G., & Sansone,C. (1984). Rewardingpinballwiz-
ardry:effectsof evaluationandcue valueon intrinsicinterest.Journalof Personality
and Social Psychology,47, 287-300.
107

Hattie& Timperley
Hattie, J., & Jaeger,R. (1998). Assessment and classroomlearning:A deductive
approach.Assessmentin Education,5(1), 111-122.
Hattie,J. A. (1992). Self-concept.Hillsdale,NJ:LawrenceErlbaum.
Hattie, J. A. (1999, June.).Influenceson studentlearning (Inauguralprofessorial
address,Universityof Auckland,New Zealand).Retrievedfrom http://www.arts
.auckland.ac.nz/staff/index.cfm?P=8650
Hattie,J. A., Biggs, J., & Purdie,N. (1996). Effectsof learningskills interventionon
studentlearning:A meta-analysis.Reviewof Researchin Education,66, 99-136.
Hattie,J. A., & Marsh,H. W. (1995). Futureresearchin self-concept.In B. Bracken
(Ed.),Handbookon self-concept(pp.421-463). Hillsdale,NJ:LawrenceErlbaum.
Hollenbeck,J. R., Klein,H. J., O'Leary,A. M., & Wright,P. M. (1989). Investigation
of the constructvalidityof a self-reportmeasureof goal commitment.Journalof
AppliedPsychology,74, 951-956.
Howie,E., Sy, S., Ford,L., & Vicente,K. J. (2000).Human-computer interfacedesign
can reducemisperceptionsof feedback.SystemDynamicsReview,16(3), 151-171.
Janoff-Bulman, R., & Brickman,P. (1982). Expectationsandwhatpeoplelearnfrom
failure.In N. T. Feather(Ed.),Expectationsand actions (pp. 207-237). Hillsdale,
NJ:LawrenceErlbaum.
Karabenick,S. A., & Knapp,J. R. (1991).Relationshipof academichelpseekingto the
use of learningstrategiesand otherinstrumentalachievementbehaviorin college
students.Journalof EducationalPsychology,83, 221-230.
Kernis,M. H., Brockner,J., & Frankel,B. S. (1989). Self-esteemandreactionsto fail-
ure:The mediatingrole of overgeneralization. Journalof Personalityand Social
Psychology,57, 707-714.
Kinch, J. W. (1963). A formalizedtheory of the self-image.AmericanJournal of
Sociology,68, 481-486.
Kinch,J. W. (1968).Experimentson factorsrelatedto self-conceptchange.Journalof
Social Psychology,74, 251-258.
Klein,W. M. (2001).Posthoc constructionof self-performance andotherperformance
in self-servingsocial comparison.Societyfor Personalityand Social Psychology,
27(6), 744-754.
Kluger,A. N., & DeNisi, A. (1996). The effects of feedbackinterventionson perfor-
mance:A historicalreview,a meta-analysis,and a preliminaryfeedbackinterven-
tion theory.PsychologicalBulletin,119(2), 254-284.
Kluger,A. N., & DeNisi, A. (1998). Feedbackinterventions: Towardsthe understand-
ing of a double-edgesword.CurrentDirectionsin PsychologicalScience,7, 67-72.
Kulhavy, R. W. (1977). Feedbackin written instruction.Review of Educational
Research,47(1), 211-232.
Kulhavy,R. W., & Stock,W. A. (1989). Feedbackin writteninstruction:Theplaceof
responsecertitude.EducationalPsychologyReview,1(4), 279-308.
Kulhavy,R. W., White,M. T., Topp,B. W., Chan,A. L., & Adams,J. (1985).Feedback
complexityandcorrectiveefficiency. Contemporary EducationalPsychology,10,
285-291.
Kulik,J. A., & Kulik,C. C. (1988).Timingof feedbackandverballearning.Reviewof
EducationalResearch,58(1), 79-97.
L'Hommedieu,R., Menges,R. J., & Brinko,K. T. (1990), Methodologicalexplana-
tionsforthemodesteffectsof feedbackfromstudentratings.Journalof Educational
Psychology,82(2), 232-241.
108

ThePowerof Feedback
Latham,G. P., & Lee, T. W. (1986). Goal-setting.In E. A. Locke(Ed.), Generalizing
from laboratorytofield settings(pp. 101-117). Lexington,MA: LexingtonBooks.
Lee, T. W., Locke,E. A., & Latham,G. P. (1989). Goal settingtheoryandjob perfor-
mance.In L. A. Pervin(Ed.), Goal conceptsin personalityand social psychology
(pp. 291-321). Hillsdale,NJ:LawrenceErlbaum.
Locke, E. A., & Latham,G. P. (1984). Goal setting:A motivationaltechniquethat
works.EnglewoodCliffs, NJ:PrenticeHall.
Locke,E. A., & Latham,G. P. (1990).A theoryof goal settingand taskperformance.
EnglewoodCliffs, NJ:PrenticeHall.
Lysakowski,R. S., & Walberg,H. J. (1982).Instructional effectsof cues, participation,
andcorrectivefeedback:A quantitativesynthesis.AmericanEducationalResearch
Journal,19, 559-578.
Markus,H. (1977). Self-schemataandprocessinginformationaboutthe self. Journal
of Personalityand SocialPsychology,35, 63-78.
Marsh,H. W. (1987).Thebig-fish-little-pond effecton academicself-concept.Journal
Marsh,H. W. (1990). Theinfluenceof internalandexternalframesof referenceon the
formationof mathandEnglishself-concepts.Journalof EducationalPsychology,
82, 107-116.
Marton,F., Dall'Alba,G., & Beaty,E. (1993). Conceptionsof learning.International
Journalof EducationalResearch,19(3), 277-300.
McLaughlin,T. F. (1974). Effectsof writtenfeedbackin readingon behaviorallydis-
orderedstudents.Journalof EducationalResearch,85(5), 312-316.
Meyer,W. (1982). Indirectcommunicationaboutperceivedabilityestimates.Journal
Meyer,W., Bachmann,U., Hempelmann,M., Ploger,F., & Spiller,H. (1979). The
informationalvalue of evaluationbehavior:Influencesof praiseandblamein per-
ceptionsof ability.Journalof EducationalPsychology,71, 259-268.
Mikulincer,M. (1988). Reactanceandhelplessnessfollowingexposureto unsolvable
problems:The effects of attributionalstyle. Journal of Personalityand Social
Moin,A. K. (1986).Relativeeffectivenessof varioustechniquesof calculusinstruction:
A meta-analysis.Unpublisheddoctoraldissertation,Departmentof Mathematics,
Universityof Syracuse,Syracuse,New York.
Moreland,R. L., & Sweeney,P. D. (1984). Self-expectanciesandreactionsto evalua-
tions of personalperformance.Journalof Personality,52, 156-176.
Mueller,C. M., & Dweck, C. S. (1998). Praisefor intelligencecan underminechil-
dren'smotivationandperformance.Journalof Personalityand Social Psychology,
75(1), 33-52.
Nadler,D. (1979).Theeffectsof feedbackon taskgroupbehavior:A reviewof theexper-
imentalresearch.Organizational BehaviorandHumanPerformance,23, 309-338.
Nelson-Le Gall, S. (1981). Help-seeking:An understudiedproblem-solvingskill in
children.DevelopmentalReview,1, 224-226.
Nelson-LeGall, S. (1985). Help-seekingbehaviorin learning.AmericanEducational
ResearchAssociation,12, 55-90.
Newman,R. S., & Schwager,M. T. (1993). Students'perceptionsof the teacherand
classmatesin relationto reportedhelp seeking in mathclass. ElementarySchool
Journal,94, 3-17.
109

Hattie& Timperley
Okun,M. A., & Sasfy,J. H. (1977). Adolescence,the self-concept,andformalopera-
tions.Adolescence,12, 373-379.
Page,E. B. (1958).Teachercommentsandstudentperformance: A seventy-fourclass-
room experimentin school motivation.Journal of EducationalPsychology,49,
173-181.
Paich,M., & Sterman,J. D. (1993). Boom, bust andfailuresto learnin experimental
markets.ManagementScience,39, 1439-1458.
Paris, S. G, & Cunningham,A. E. (1996). Childrenbecoming students.In D. C.
Berliner& R. C. Calfee(Eds.),Handbookof educationalpsychology(pp. 117-147).
New York:Macmillan.
Paris, S. G., & Winograd,P. (1990). Promotingmetacognitionand motivationof
exceptionalchildren.Rase:Remedial& SpecialEducation,11(6), 7-15.
Podsakoff,P. M., & Farh,J. L. (1989).Effectsof feedbacksign andcredibilityon goal
setting and task performance.OrganizationalBehavior and Human Decision
Processes,44, 45-67.
Purdie,N., Hattie,J. A., & Douglas,G. (1996). Studentconceptionsof learningand
theiruse of self-regulatedlearningstrategies:A cross-culturalcomparison.Journal
Rummel,A., & Feinberg,R. (1988). Cognitiveevaluationtheory:A meta-analytic
reviewof the literature.SocialBehaviorand Personality,16(2), 147-164.
Ryan,A. M., & Pintrich,P. R. (1977). "ShouldI askforhelp?"Therole of motivation
and attitudesin adolescents'help seeking in mathclass. Journalof Educational
Sadler, R. (1989). Formativeassessmentand the design of instructionalsystems.
InstructionalScience,18, 119-144.
Silj6, R. (1979).Learningin the learner'sperspective-I. Somecommonsenseconcep-
tions(ReportNo. 76). Gothenburg, Sweden:Universityof Gothenburg, Department
of Education.
Schroth,M. L., & Lund,E. (1993). Role of delay of feedbackon subsequentpattern
recognitiontransfertasks.Contemporary EducationalPsychology,18, 15-22.
Schunk,D. H., & Rice,J.M. (1991).Learninggoalsandprogressfeedbackduringread-
ing comprehensioninstruction.Journalof ReadingBehavior,23, 351-364.
Sharp,P. (1985). Behaviourmodificationin the secondaryschool:A surveyof stu-
dents' attitudesto rewardsand praise.BehavioralApproacheswith Children,9,
109-112.
Shepard,L. A., Flexer,R. J., Hiebert,E. J., Marion,S. F., Mayfield,V., & Weston,T. J.
(1996).Effectsof introducingclassroomperformanceassessmentson studentlearn-
ing. EducationalMeasurementIssuesand Practice,15, 7-18.
Shrauger,J. S., & Sorman,P. (1977). Self-evaluations,initialsuccessandfailure,and
improvementas determinantsof persistence.Journalof Consultingand Clinical
Simmons,M., & Cope,P. (1993). Angle androtation:Effectsof feedbackon thequal-
ity of learning.EducationalStudiesin Mathematics,21, 375-382.
Skiba,R., Casey,A., & Center,B. A. (1985-1986).Nonaversiveproceduresin thetreat-
mentof classroombehaviorproblems.Journalof SpecialEducation,19, 459-481.
Smith,T., Snyder,C., & Handelsman,M. (1982). On the self-servingfunctionof an
academicwooden leg: Test anxiety as a self-handicappingstrategy.Journal of
Personalityand Social Psychology,42, 314-321.
Steinberg,L. (1996). Beyondthe classroom:Whyschool reformhasfailed and what
parentsneed to do. New York:Touchstone.
110

ThePowerof Feedback
Sterman,J. D. (1989). Misperceptionsof feedback in dynamic decision making.
OrganizationalBehaviorandHumanDecisionProcesses,43, 301-335.
Sturges,P. T. (1972). Informationdelay andretention:Effectof informationin feed-
backandtests.Journalof EducationalPsychology,63(1), 32-43.
Sturges,P. T. (1978). Delay of informativefeedbackin computer-assistedtesting.
Journalof EducationalPsychology,70(3), 378-387.
Suls,J., & Wan,C. K. (1987). In searchof thefalse-uniquenessphenomenon:Fearand
estimatesof social consensus.Journalof Personalityand Social Psychology,59,
229-241.
Swann,W. B. (1985). The self as architectof socialreality.In B. Schlenker(Ed.),The
self and social life (pp. 100-125). New York:McGraw-Hill.
Swann,W. B., & Hill, C. A. (1982). Whenour identitiesaremistaken:Reaffirming
self-conceptionsthroughsocial interaction.Journal of Personality and Social
Swann,W. B., Pelham,B. W., & Chidester,T. (1988).Changethroughparadox:Using
self-verificationto alterbeliefs.Journalof Personalityand Social Psychology,54,
268-273.
Sweller,J. (1990).Cognitiveprocessesandinstructionprocedures.AustralianJournal
of Education,34(2), 125-130.
Swindell,L. K., & Walls,W. F. (1993). Responseconfidenceandthe delayretention
effect. Contemporary EducationalPsychology,18, 363-375.
Tenenbaum,G., & Goldring,E. (1989). A meta-analysisof the effect of enhanced
instruction:Cues,participation, reinforcementandfeedbackandcorrectiveson motor
skill learning.Journalof ResearchandDevelopmentin Education,22, 53-64.
Tesser,A., & Campbell,J. (1983). Self-definitionandself-evaluationmaintenance.In
J. Suls & A. Greenwald(Eds.),Socialpsychologicalperspectiveson the self (Vol.
2, pp. 1-31). Hillsdale,NJ:LawrenceErlbaum.
Thompson,T. (1997). Do we need to trainteachershow to administerpraise?Self-
worththeorysays we do. Learningand Instruction,28, 49-64.
Thompson,T. (1998). Metamemoryaccuracy:Effectsof feedbackandthe stabilityof
individualdifferences.AmericanJournalof Psychology,111(1), 33-42.
Thompson,T. (1999). Underachievingto protect self-worth:Theoryresearchand
interventions.Avebury,UK:Aldershot.
Thompson,T., & Richardson,A. (2001). Self-handicappingstatus, claimed self-
handicapsandreducedpracticeeffortfollowingsuccessandfailurefeedback.British
Journalof EducationalPsychology,71, 151-170.
Timperley,H., & Parr,J. (2005).Literacyprofessionaldevelopment project.Wellington:
New ZealandMinistryof Education.
Timperley,H. S., & Wiseman,J. (2002). Thesustainabilityof professionaldevelop-
mentin literacy.Wellington:New ZealandMinistryof Education.
Trope,Y. (1975). Seekinginformationaboutone's own abilityas a determinantof
choice amongtasks.Journalof Personalityand Psychology,32, 1004-1013.
Trope,Y. (1980). Self-assessment,self-enhancement andtaskperformance. Journalof
ExperimentalSocial Psychology,16, 116-129.
Van-Dijk, D., & Kluger, A. N. (2000, April). Positive (negative) feedback:
Encouragementor discouragement?RetrievedSeptember2001 fromhttp://www/
huji.ac.illunew/main.html
Van-Dijk, D., & Kluger, A.N. (2001). Goal orientation versus self-regulation:
Differentlabelsor differentconstructs?Paperpresentedat the 16thannualconven-
tion of the Societyfor IndustrialandOrganizational Psychology,SanDiego, CA.
111

Hattie& Timperley
Walberg,H. J. (1982). Whatmakes schooling effective? ContemporaryEducation
Review,1, 1-34.
Watkins,D., & Regmi,M. (1992). Howuniversalarestudentconceptionsof learning?
A Nepaleseinvestigation.Psychologia,35, 101-110.
Watkins,D., Regmi, M., & Astilla, E. (1991). The Asian learner-as-a-rote-learner
stereotype:Mythor reality?EducationalPsychology,11, 21-34.
Weiner,B. (Ed.).(1974a).Achievement motivationandattributiontheory.Morristown,
NJ:GeneralLearningPress.
Weiner,B. (1974b). An attributionalinterpretationof expectancy-valuetheory.In
B. Weiner(Ed.), Cognitiveviews of humanmotivation(pp. 51-70). New York:
AcademicPress.
Weiner,B. (1977). An attributionalmodelfor educationalpsychology.In L. Shulman
(Ed.),Reviewof researchin education(Vol 4., pp. 179-209). Itasca,IL:Peacock.
White,K. J., & Jones,K. (2000). Effectsof teacherfeedbackon the reputationsand
peerperceptionsof childrenwithbehaviorproblems.Journalof ExperimentalChild
Wilkinson,S. S. (1981). The relationshipof teacherpraiseand studentachievement:
A meta-analysisof selectedresearch.DissertationAbstractsInternational,
41(9-A),
3998.
Winne,P. H., & Butler,D. L. (1994). Studentcognitionin learningfromteaching.In
T. Husen& T. Postlewaite(Eds.),Internationalencyclopaediaof education(2nd
ed., pp. 5738-5745). Oxford,UK: Pergamon.
Wood,R. E., & Bandura,A. (1987).Impactof conceptionsofabilityon self-regulatory
mechanismsand complex decision-making(Working Paper Series, 87-019).
Sydney: University of New South Wales, Australian Graduate School of
Management.
instructionon sci-
Yeany,R. H., & Miller,P. A. (1983).Effectsof diagnostic/remedial
ence learning:A meta-analysis.Journalof Researchin ScienceTeaching,20, 19-26.
Zimmerman,B. J. (2000). Attainingself-regulation:A socialcognitiveperspective.In
M. Boekaerts& P. R. Pintrich,(Eds.),Handbookof self-regulation(pp. 13-39). San
Diego, CA:AcademicPress.
Authors
JOHNHATTIE of Education
is aProfessor of Auckland,
attheUniversity PrivateBag92019,
Facultyof Education,Auckland, His
NewZealand1142;e-mail:j.hattie@auckland.ac.nz.
interestsareassessment,modelsof teachingandlearning,andstatistics.He andHelen
Timperleyco-authored withShirleyClarke,UnlockingFormative Assessment(Hodder-
Moa,Auckland, NZ,2003).
HELENTIMPERLEY attheUniversity
of Education
is a Professor of Auckland,PrivateBag
of
92019,Faculty Education, Auckland,New Zealand 1142;e-mail: h.timperley@auckland
.ac.nz.Herresearch
investigateshowto promote atalllevelsof theeducation
learning sys-
tem.SheandJohnHattieco-authored withShirleyClarke,UnlockingFormativeAssessment
(Hodder-Moa, Auckland, NZ,2003).
112


The Power of Feedback (Review of Educational Research, Vol. 77, Issue 1) (2007)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Power of Feedback (Review of Educational Research, Vol. 77, Issue 1) (2007)

Uploaded by

Copyright:

Available Formats

The Power of Feedback

Author(s): John Hattie and Helen Timperley

This content downloaded on Tue, 12 Mar 2013 15:55:12 PM