Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
EP 173 340
AUTHOBTITLE
INSTITUTION
SPONS: AGENCYPUB DATECONTRACTNOTEAVAILABLE FROM
FORS PRICEDESCRIPT0ES
DOCUMENT RESUME
SP 014 4 64
Brophy, Jer E.Advances in Tnlchar Effectiwenese Eesearch.Occasional Paper No.' 13.Michigan tat Univ., nAt Lansing. Inst. forResearch on Teaching.National Inst. of Education (DHEw), shington, D.C..
Apr 79400-76=0013:2bp.; Some paqs may not :-..produc 01 arlyInstitute foi Aesoarch en 7achin1, College ofEducation, MichigAn Stat? Uffivrsity, 252 EricksonHall, East Lansing, michilan 43624 (-82.00)
MF01/PCO2 Plus Postag.*Academic Achievem,,,,rt; Basic Skills; BehaviorPatterns; *..-71as5 nag-m an; Class Organization;*Educational ii..ese-arch; *Effctive Te,aching;Elementary 3econdary lAucation; Group Instruction;*T6acher Behavior; *T-laching m.,,thod3
ABSTBACTFollowing a reviw of resarch findings on the
,rFlationship 'between teach: br,'havior and student aohiavement,
current activitis in teach,?: eftectiveass research ars discussed.Two maior trends emerge it recc'nt rese-aroh studies. Existingcorrelational findings- on topic ar.e being integraterland appliedto contexts' beyond 6a sic skil1 !3 instruction in the elementary grades.
clusters at correlational findiaqsar!:t bing asressed for degree ofimplementation by techerS in their daily insrucinal mnthods.Aecont findings indicate that dirct instruction in small groupscoupled with formal classroom organization and management are aseffective in producing sitiSfactcry leanaing resultsat the secondary
level as they in teaching basic skills at tL elementary level.
(JD)
****.,t ************************************************ **
* Reproductions supplied by EDRS are the Lest that can be madefrom original ,iocuirr=.11t.
****************************************,t*********************
Occasional Paper No. 18
ADVANCES
IN TEACHER EFEECTIVENESS RES- RCH
Jere E.
'Pols ish d by
The lost ,; [.nt for Research on Teaching
252 rric:spn HAllhigan Staue Un[yersiv
EnSt 1,ans4ai,,, Michigln 48,824
April 1979
U S DEPARTMENT OF HEALTH.EDUCATION WELFARENATIONAL INSTITUTE OF
ECM t:A TION
THIS DOCUMENT HAS BEEN PEBBO.DUCED ExACTLy AS RECEIVED FROMTHE PFRSoNON OHGANIzATioNAims/0 IT POINTS ot, VI EW OR OpINIONESTATED DO NOT NECESSARILY PEPRE.SENT OFFICIAL NATIONAL INSTITUTE_ OFEDUCATION POSITION OP POLICY
"PERMISSION TO REPRODUCE THISMATERIAL HAS BEEN GRANTED BY
TO THE EDUCATIONAL RESOURCESINFORMATION CENTER (ERIC)."
This work is sponsored in part bv theInstitute for Re erch Lin
lehlug Colic-TO of iducation, Michigr:n StateUniversity. The Institute_
Research On Teaellin is funded primariLyby the Teaching Division of
Nntionni:Institute of Education, United StatesDepartment of Health,
Education, an&Welfaro .The opinionscxpreed in this publication do not
rr.crssari the positin, policy, orendorsement of the National
Institute Education.(Contract No. 400-76-0023)
2
INSTITUTE FOR RESEARCH TEAM
Teache rs hougl r and do it roan , 'Ire the focus of studcurrently
under way at Michigan State Univern[mu-flute fc r ReSearch oaa Teach-
lag (1ST). The IRT, fc nd'd. in rjaar II1976,with a $3.6 million grant from
the National jhstitutc., of I' ducati on, has aaa jor aatojeelsvestigntlng
teacher decision- aki
classroom wn nemont
gong- art,
ti
is i'cssa of reading diagnosis and reme-
in the areas of len-
ading, aid atl motic, teacher education, teacher planning
feats of external pressures on ton(' s' decision, and
cape:
-s' percep-
t affect. Researchers from many different disciplines ca-
1ST re= search. In additicc public school teachers work atIRT
as half-time collaborators in arch, helping to design andplan studies,
collect data, and analyze results. The Institute publishes research reports,
conference proccedi ocensi.Laaafl Nip°. - and e quarterlynewsletter
for practitioners. For more information or to heplaced on the 1ST mailing
li t please write to The IRT Editor, 252 Erickson, MSU, East Lansing, Mich-
igan 48824.
Dl rector: Lee S. Shulman,
Associate Director: Judith E, Lanier
Editor a- Staff:
once W. Lezotte coordinator of Comm n- ons ssemination
Linda Shainway, 1ST _editorJames Kern, editorial assiStantJanice Aehrs, editorial assistant
Advances i cher Effec t vcness Research
Jere E. upiL
Historically, reviewing r ,ear( c,n. the links b, .aching
behavior and student learning has been a frustratingtask. Morsh and
Wilder (Note 1) and Medley and Mitzel (1963) couldfind virtually no
P
clear results to discuss. The situation improvedsomewhat over the next
10 years, so that Rosenshine and Furst ( 9and Dunkin and Biddle (1974)
could point- to weak but consistent findings supportingsuch variables as
an organized, businesslike approaLh _ tteaching, clarity, and enthusiasm.
Still, even where clear trends c___ obviouS, there always seemed tobe
exceptions and a- nt contradictions.Dunkin and Biddle resolved many
of these by taking into account canteat variablessuch as grade level and
subject matter but a great manycontradictions, remained. ..
In the last five or six years, the situation has improved considerably.
The National Institute of Education began fundingexpensive studies, and
important improvements research design began to appear;among these
were rational sampling of teachers(rather than random or convenience
sampling), inclusion of enough teachers to a--11 ow for meaningful statistical
analyses, collection of many hours worth of data perclassroom, develop
ment of multifaceted and sophisticated classroomcoding instruments
that accounted for context and sequence of interactionrather than just
1-This paper was presented as part of a research symposium entitled,
"Progress Report:, Advances During the Past Year in the KnowledgeBase for the
Preparation of Teachers," at the annual meeting of. the AmericanAssociation of
Colleges for Teacher Education, )Chicago;- 1979.Portions of the paper were
presented earlier as part of the keynote speech to the annual meeting-Of
the .Midwest Association of Teachers of EducationalPsychology, Bloomington,
Indiana, ,1978. 42Jere E. Brophy\is a senior researcher in the
Institute for Research on
Teaching andprofessok of teacher education and educational 'psychology.
2
behavioral frequen ies, and con nuration on the individualteacher and
class as the unit of analysis. These studies havefocused on instruction'
in basic skills in primary grade classrooms, usingstandardized 'a ieve-
ment tests ag learning measures.
This effort began with several
conducted at various
emscale field correlational studies
ary grade 1evel6 (Brophy & Evertson, Note 2;
Good & Grouws, Note 3; McDonald & Elias, Note 4; Soar & Soar,1972;
Stallings & Kaskowitz, Note 5; Tikunoff, Berliner, Rist, Note 6).
These studies varied with regard to the types of teachers andstudents
included, the kinds of variables addressed, and methods used, butthdre
was sufficient overlap and replication to prov;hledependable knowledge
about relationships between types of teaching, particularlydirect ins_ uc-
tion and student learning of basic skins ifithe elementary grades (see
reviews by Borich 1977; Medley, 'Vote 7; and Rosenshine, 1976).Since then,
other studies have built on the results ofresearch; the work has been
extended to the junior high and high school levels, and experimental
studies designed to test hypotheses developed from earlier correlational
k have been conducted.
The data frora the correlational studies hang together quitewell
to support Rosenshine's (Note 8) claim that "directinstruction" is
effI7tive for producing student learning of basic skills. Rosenshine
(in press) suggests that for direct instruction to be effective,teachers
must (1) focus on academic goals;)(2) promote extensive contentcoverage
and high Levels of student involvement; (3) selectinstructional goals
and materials and actively monitor student progress; (4) structure
learning activities and include immediate, academically. .orientedfeedbac
and (5) create an environment that is task oriented butrelaxed.
Taken together, the correlational studies provide strong supportfor
the for.owing generalizations:
10 Teachers make a difference.Contrary to the theorizing of
Stephens (1967) and the implications of prOjects like the Colemanreport
(Coleman t al., 1966), which, unfortunatelyanalyzed data only at the
school le,/el, research that analyzes at the teacherlevel makes it clear
that certain teachers elicit much more student learningthan others, and
that their success is tied to consistent differences inteaching behavior
(see the'studies cited on page 1, and also Good, Biddle, &Brophy, 1975;
sand Rakow, Aira an, & Madaus, 1978).
2. Despite the first generalization,there is no support for the
notion of generic teaching skills, if these are defined asthe types
4*
very specific behaviors typically included in performance based teacher
education programs. Few, if any, specific teachingbehaviors are
appropriate in all contexts.
On the other hand, when data c integrated at ahigher level of
generality, several clusters or patterns of behavior emerge that are
consistently related to learning gains.
3. One of these behavior patterns -1-des teacher expectations
and role definitions. Teachers who believe thatinstructing students in
the curricuum is basic to their role, who fully ex2ect toconduct such
instruo t.ion, and who attempt to do so ir, their classrooms are more
successful than teachers who do not. The more effective teachersallocate
more of their time for teaching and actuallyspend more of that time
accordingly.
4. Another basic Cluster includes such variables as classroom
management skills, student engagement/time on task, and studentopportunity
4
to learn materidl. Eftr ctive teachers know howto organize and maintain
a classroom learning environment that mximizes thetime spent on
productive activities ar and minimizes the time lost duringtransiti
periods of confusion, car distuptions that require disciplinaryaction.
5. A third cluster indicates sup t for the variouselements
direct instruction. Studies of general approaches toinstruction con
sistently reveal that students taught with a structuredcurriculum do
better than those taught with ind.vidualized or discovery learning
approaches, and those who receive muo11 instruction directly twin the
teacher do better than those expected to le 1own or from one
another. Teacher talk in the form cif 1.es and demonstrations is
imp.rta t as are the me h--on red mcthn 1.of recitation, drill, and
prance. It appears that most Drms of open education orindividualized
imr.t,-uctionunrealistic expectations about the degree to which
stud , . in the rly grades can manage theiractivities and learn
-tdepc4elently (see studies cited on page 2, also Gage, .1978; Ininan,
INL: _and Stallings & Hentzell, Note 10).
The instruction that seems most efficient is the kind in which the
teacher works w class (or with small groupsin the early
grades), presents information in lectures/demonstrations,and then
follows up with recitations or practice exercises in whichthe students
get opportunities to make responses and receive corrective feedback.The
teacher maintains an academic focus, keeping the studentsinvolved in a
lesson or engaged in seatiork, monitoring their performance,and. providing
individualized feedback. The pace is rapid in thesense-that the class
moves efficienti y through the curriculum as awhole (and through the
successive objectives of any given lesson), but. progressfrom one
objective to the next involves very small, easy steps.Success rates
7
5
in answering teachur questions durint l --ons are high (about75%), and
success rates on assignments designed be done independently are very
high (approaching 100%). (See studies cited on page 1,and. also the
reviews by Borich, 1977; Medley, Note 7; and Rosenshine, Note 8, in
press.)P
These specifics concern -W-,.pp -uction vary somewhat
11 context, particularly grade level and student ability level. In
the primary grades, where the emphasis i.s on mastering the basic skills,
the teaching/learning situation differs from later grades, wherestudents
0are expected to use basic skills to learn other things, and to manage-
their own learning to a greater degree. The early grades appropriaCely
involve more small -group instruction relative to whole-class instruction,
more teacher circulation around the room and initiation ofcontact with
the students who are working on assignments opposed to letting the
students come to the teacher for help), more recitation and drill (but
less genuine discussion), morepraise and affect generally,very low
error rates, and 'a low cognitive level due to the emphasis on repetition,
recitation, and drill, (Higher cognitive level activities seem counter-
productive in the early grades, although they become more important
later.) More specifics about grade leveldifferences can be found in
Eve- son, Anderson, and Brophy, Note 11; Fisher, Filby, Marliave, Cahen,
Dishaw, Moo e and Berliner, Note 12; McDonald and Elias, Note 4; Murnane
and Phillips, Note 13; and Trismen, Waller, aiLd Wilder, Note 14.
Within any given grade level, teachers working with low-ability
students need to move at a slower pace and provide more repetition
and individualized monitoring to make sure that overlearning isattained
before moving on to objectives that assume prior mastery. of Tresent
objectives. They also need to supply gr ter warmtl
6
oura Bement, and
personalized teaching gen -ally but Tess challenge ((although not less than
`iticism (Brophy &
. Eve tson, Note 2, 1976; Program on TeachingEffoLtiveness, Note 15).
the students can handle) and less demand
Current Progress.
Current activities in teacher effectiveness research feature two.
major trends: (1) integration of exi ti.ng correlationalfindings add
probing Of the limits of th
Skills ins -rue
llization tr ntexts beyond basic
ion in the elementary grades; and (2) experimental
studies in which clusters of correlat.ic na.t E tndiugs are brought together
into treatment packages and assessed Inc ee of implementation by
teachers and for success in producing more learning than what is observed
in control groups.
The maturation __ the field can he seen in recentreviews, which read
less like laundry lists of random findings and more like integrated
discussions of organized approaches to instruction. Good (in press) is
publishing an interesting paper that makes many of the same points made
here and some, others that are worth mentioning.
First Good notes that the support for group-based rather than
individualized instruction, and for whole class instruction rather than
small-group instruction (except in the early grades), should renew
educators' appreciation for the advantages of these methods. Group-
based instruction is often maligned by those who favor individualizedand
self-paced instruction, but, like recitation, it survives. Good suggests,
and I concur, that group-basedinstruction survives because it has
important advantages. It is easier to plan and manage,provides more
modeling of correct thinking and responses for slower students, and
avoids the elitism And labeling problems that can crop up whenability
9
7
grouping is used. This does not mean that large -groupinstruction should
be the exclusive method, of course; qt only indica _s that thisapproach
has advantages, is effective, and may be the method of choice for many
goals and contexts.
Good also notes that tradi onai or direct instruction seems
clearly superior to open education for producing mastery of basickills.
-He adds, however; that it may not be the hest appr {each, or even
appropriate, curricular areas that do not involveskill mastery,but
instead seek to promote appreciationgeneral familiarity, enrichment,
or student personal development. Nor is open educationnecessarily
effective here,'either; Good notes that open-education advocates have
put too much stress on things like free choi.ce of tasks or free movement
around the room, which are less vital to real-life application than things
like developing skills for problem solv g and self-evaluation. In any
case, he observes. that some structure is needed for most educational
6
active and that relatively more structureis needed in the early
grades, for low-ability. studehts, and for anxious or dependent students.
Classroom Mane egnt
Recent publications by Brophy and Putnam (1979) and Evertson and
Anderson (Note 16) have elaborated on what constitutes effective classroom
management and how it ,interacts with effective instruction. Brophy and
Putnam review studies on classroom management generally, not just those
that link management with student learning. They note strong support for
most of the variables stressed by Kounin (1970): "withitness," "over-'
lapping," signal continuity and momentum during lessons, and variety and
challenge during seat- ork They also !totthat recent stud {es h ve not
supported Kounin's variables of group aler i nd accountability, which
call for teachers to be random and unitredictah.lc in their questioning,
to call on nonvolunteers frequently, arid to _require studentsto comment
on one another's responses (to make sure that they pay atten nto peers
as well as try the teacher). These group alerting and accountability
techniques either correlate negatively or show curvilinear relations ips
learning gains. Apparently, teachers who do all theother things
that'Kpuninstrb ses, and therefp successful in maximizing student
attention and engagement, should not need to use group alertingand
accountability behaviors very
Thesece ents help reconcile Kounin's findings with the findingsof
Brophy and Eve -son (Note 1976) and Anderson, Evertson, addBrophy
(1979), which indicate that teachers whocalled on students in a
predictable pattern in going around the reading group had more successthan
teachers who were unpredictable. Apparently', any disadvantages thatthis
-technique might involve (e.g., student' who can predict when their turn
will come may pay less attention when it is not their turn) areoutweighed
by the advantages gained: ...the method insures, that(1) everyone.zets
roughly equal opportunities to r&cite and participate in the group4(often,
"random" questioning really means calling on the brighter and more eager
students often and the slow or alienated students rarely);(2) the greater
structure that the technique provides may be helpful toanxious students'
and (3) the automatic determination, turns prevents thedistractions'
involved when students call out answer petition the teacher tocall'
on them.
-"Mithltness, as defined by Kounin(1970), is a strategy wherebyteachers
are conti=nuously aware of all that is going on in the'classroom, and they
communllate this awareness -to their students,'tOverlapping," according to
Kounin, is the ability to do 'several things at once, such asmonitoring the
classroom while teaching a small group. ,
A
Ivcrtson and Anderson (Note 1.6) hNve ex Loring the specifics.'
involved in organ]. zing and managing the classro n and the interar:tions
between management and instruction. During the 1977 -7d' school year,they
.
.Observed 28 third-grade classrooms extensively during the firstthree
weeks of school, and periodically the-,after, gathering information
on what rules and procedures the teachers introduced,dnd low they did so.
During 1978-79, they are ohs rving junior high. lassrooms.
suits from the study strongly support two major generalize-/
,1?
Classroom organization and management sl;rills are intimately
related to instruction skills; that ts, good ihstructor_ tend to be good
managers, and (2) good organiz'ti-n and mane e e t is good instruction,
at least at the third-grade level; that is, successful classroom managers
`spend a great deal of time early in the year conducting semiformal lessons
familiarize students with rules and procedures. This research is
yielding very rich, detailed information about procedures involved in
setting up effective classrooms, and ultimately should be extremely
valuable to teacher educators.
ILTIETIllallatijchool Studies
Several investigators are probing the limits co generalization of the
linkageS between direct instruction and student' learning observed in
basic skill'instruction in the early grades. Recent studies.by Stallings
(Note 17) and Everston et I. (Note 11) indicate that the key to
generalization may not be student age or grade level, but mastery of
basic skills. Stallings, (Note 17) has been studying reading instruction
the junior and senior high school levels. Her findings are very
similar to the findings reviewedearlier for 'basic skills in the early
10
grades:. Growth -in reading skills isassociated with maximizing time on
task, instructing the total. group most of the time,directing questions
to specific students (rather than volunteers).,regularly providing feed-
Amok, controlling negative behavior, encouraging positivebehavior, and'.
using guides and probing qUestions when students do notknow the answer.
Negative indicators include grading papers during theclasS period,
socializing of allowing students to socialize, allowinginterruptions
and intrusions into the class activities, and allowingnegative behavio
McCennell (Note 18) reported the following correlates of student
learning in high school algebra else es: task orientation, clarity,
frequent probing to improve student response, anthus asm,and frequent
teacher talk. Again, these are familiaraspects of the direct instru
tion approach,
Evertson,et al. (Note 11) _report the following correlatesof
student learning in seventh- and eighth -grade math classes:considerable'
class time spent on discussion jecture; and drill, and notjust indivi-
-dualiZed instruction or individual seatwork; taskoriented, bueinesslike
instruction; much teacher time spent actively instructing andinteracting
with .students; greater praise of good contributions.(althoUgh praise was
not frequent in an absolute' sense); good classroommanagement; especially
withitnese; asking of process (thought or:explana ion ) questions aswell
as preduct (fact or memory) questions.
With two exceptions (discussed in, the next paragraph), these findings
replicate what was found in the earlygrades, and suggest that direct
'instructional may b_ the most effective method at any'g adelevel when
mastery Of basic skills is the goal,This . pa "tte rn did not appear, however,
seventh- and ,eighth-grade English classes, where Evertson etal., (Note 11)
obtainedjstrikingly different results. Significant relationships between
1-3
11
classroom process variables and student learning in these classes were
infrequent, and there was little support for the direct instruction model.
Several factors probably explain this finding, but the' ajor one seems
to be that basic skill mastery is not a primarygoal of seventh- and
eighth-grade English classes. The instructional objectives pursued in
these classes are more numerous and variable than in math classes._ Many.
objectives, such as poetry composition, oral dramatization, or literature
appreciation, are not easily or even appropriately pursued with the direct
instruction approach.
One implication, then, of recent work is that the findings concerning
direct instruction do generalize to higher grade levels and different
kinds of students, but only when basic skill mastery is the primary goal.
Not everythinggeneralizes; of cou Positive findings in the Everts on
et al. (Note 11) study for public praise of student contributions and
for asking higher level questions in addition to factual questions are
not usually observed in the early grades. Praise correlates sometimes
positively, sometimes negatively, but usually not at all with learning,
depending on context factors such as student ability, levels, teacher
versus student initiation, and specification and elaboration of the
praise itself. (Praise seems to be generally overrated, although it
s seem impo tant for low ability/anxious/dependent students, provided
=hat it is genuine and deserved and the praiseworthy aspects of the
-,performance are specified.)
Level-of question or cognitive demand usually shawsa negative
correlation (although sometimes a curvilinear relationship) with
learning"in the early grades (Rosenshine & Berliner, 1978; Soar & SOar
Note 19). The implications_ or teaching in the early grades seemto
be:- (1) move in very small steps ancloverteach to the point of over-
12
learning and (2) move at a rapid pace,but do not challenge students
beyond their ability to respond meaningfully.
Several recent studies indicate that the situationis somewhat
different in.the middle and upper grades(Anderson & Scott, 1978; Evertson
et al., Note 11; McDonald & Elias, Note 4;Mu nane & Phillips, Note 13;
Trismen, Waller, & Wilder, Note 14 ). Compared to the elementarygrades,
the later grades. tend to have more large-groupand whole-class activities;
less frequent and less affectively-toned'dyadic teacher-stu&nt interactions;
less recitation and drill and more discussion; morecognitive challenge
and high level cognitive activity;less teachercenteredness and more
student autonomy; more sustained concentrar
a more rapid pace within these acti-'
In the early grades, it is 4mporgunt for theteacher to elicit
(This is-a. major
on academic and
responses from, and provide feedback to each student.
reason why small-group instruction is important atthese grade levels.)
Later, this individualized (within the group continstruction is no
longer necessary, and it becomes moreImportant for the teacher to keep the
whole class together and move along at a good,pace.Basic skills have been
mastered, and learning objectives now involvehigher cognitive activity,
so challenging Students with difficult orcomplex questions is more
app-45priate, Even so, learningshould-be relatively easy -- most questions
should be answered and students should beable to Complete independent
work assignments correctly
-Eliciting student cone but , integrating them into thediscussion,
and praising the more noteworthy ones allbecome useful techniques that
y
correlate positively with leerning gain, .Recent work. has helped clear
up the apparent discrepanciebetween the writings of Flanders (1970) in
this area and some of the data supporti=ng' thedirect ,instruction
I
13
There is continuing and increasing support for the effectiveness in the
Upper grades of certain aspects of what Flanders called "indirect
teaching ": praise, use of student ideas, andhigh frequency of student
talk (if it is focused on academic objectives; non-academic student
talk correlates negatively with learning).
These data must be placed in context, however. It appears that
the:really important determinants of learning at the highergrade levels
are not the things that Flanders clustered under "indirect instruction
but instead are other aspects of teaching that Rosenshine includes under
"direct instruction": frequent lectures, demonstrations,and teacher-led
discussions (Barr & Dreeben, 1977). In the process of using these
techniques, teachers elicit frequent student contributions, which makes
it possible for them to use student ideas and to integrate them into
the discussion, as well as
ideas in the first place 'seems
praising them or inte sting them Into the discussion.
.them. In any case, eliciting the
the, crucial variable her e not
Interactions with Learner Characteristics
Another recent trend is the qualification of general process-product
findingsrby analysis of the data for aptitude- treatment interactions (ATI's
or other interactions betweenlearner characteristics and optimal
instruction. Brophy and Evertson (Note 2,1976), Evertson et al. (Note.11),
Good and Grouws (Note 3); and'the Program on TeacherEffectiveness (Note 15)
all noted somewhat.4,ifferent patterns of optimal instruction for students
who differed in socioeconomic status or ability level. .Other investigators=
have analyzed .interactions between instructional methods and student:
personality characteristics or classroom behavior patterns indetermining
student learning (Bennett, 1976; Cunningham, 1975; Ebmeier &Good; 17:
press, Peterson, 1977; Solomon & Kendall, 1976).These, findings have not
16
14
been well,`integrated:yet, because somewhat different studenttraits have
been used as the basis for classification, but certailvt-ends arealready
evident: direct instruction (and close teachermo monitoring and supervision
-generally) is needed more by students who are anxious and dependent, easily
distracted, low in ability, or low in achievement Motivation. 'Students
with opposite traits can handle more of their learning independently.
expect to see more research on interactions in the nearfuture, followed
by attempts to'integrate the interaction data with maineffects data
in order to make prescriptions about how teachers can optimizethe
tradeoffs that are necessarily involved'in teaching groups of studet
Experimental Studies
The final recent trend discussed here is probably the mostimportant:
the design of experiMental studies to test the causal linkages between
.teacher'behavior and student learning that are implied but not proven by
correlational studies. Obviously, such work needs to bedone if we are to
claim that teacher behavior correlated with student learningactually causes
that learning.
Recently, three major field experiments have been conducted to
follow up on the process- product work reviewed here. Anderson et al.
f.
(1979) pulled together 22 princlples ofsmall -group instruction derived
from earlier work and organized them into a coherent treatment deSigned
for first-grade teacher's to use with their reading groups. Good and
Grouws (in'pres-)incorporated a variety of principles draWn from their-
earlier correlational work into a systematic approach for teaching
mathematics in fourth grade, and tested these principles in an
-,experimental study. Fin'ally, Crawford, Gage, and their-colleagues in the
Program on Teaching Effectiveness (Note 15) aC Stanford pulledtogether.'
l5
a large number of principles drawn from previouswork, by Brophy and
Evertson (Note 2,, 1976), McDonald and Elias (Note 4), Soar and Soar
(1972), and Stallings and Kaskowitz (Note 5), and included them in,
a treatment designed for the third-grade level.
Each of these, studies yielded statistically significantresults
favoring tTeatment teachers over control teachers in producingstudent
learning gains on standardized achievement tests.
Each study also involved-a strong observation component, so the
tea e ,be monitored forthe degree to which they implemented
the tre nt (and control teachers could be monitored for thedegree
to which they spontaneously included treatment behaviorin their
teaching). These observational data yieldedimplementation scores for each
of the teacher behaviors included in the treatment,and these scores
were analyzed to determine whether they showed theexpected relationships
with learning scores,
Not all treatment elements have been implementedproperly, of
course, and not all of those that were implemented haveshown
expected significant relationships with learning scores.However, where
the treatment behaviors were implemented 'sufficiently,and'Where significant
results were obtained, the findings have been overwhelminglypositive,
replicating previous correlational work and providing strangerevidence
of a causal linkage between teacher behavior and studentlearning;
M9st of these findings are prescriptive, althoughy allow for
teacher judgment This can be seen in the followingexamples, drawn frot,
the study by'-Anderson et al.( 1979), all of which werevellimplemehted'
by the treatment group teachers and significantly related tolearning -gains.
16
Once in (reading) group, the children should be seatedwith
their backs to the rest of the class bile the teacher is facing
the class.
2. The introduction to the lessonshould contain an overview of
what is to come to mentally kepare the students for the
presentation.
The teacher should work with one ndividual at- a time in
having the child practice the new skill and apply the new
concept, making sure that everyone is checked and receives.
feedback.
The teacher should use a pattern for selecting children to take
their turns reading in the group or answering questions (such
as going from one end of the group to the other) rather than
calling on them randomly and unpredictably:'
When call-outs occur, the teacher should remind the child that ,
everyone gets a turn and that he/she must wait his/her,turn to
answer.
6. After asking a crestion, theteacher should wait for the child to
respond acid also see that other children wait and do not call
out answers, If the child does not respondwithin a reasonable
title, the teacher. should indicate that some response is
expected by probing.
7. Praise should be used In moderation.The teacher 'shbUld praise
thinking and effort'more than just getting the answer, and
should Make-praise as specific and individual as possible.
Criticism should also be as,specif c.as possible, and'shouldinclude-specification of desirable or correct alternatives.
Similar examples can be found in the other -two experimental studies.
Taken together, these.studies provide an impressive'number ofguidelines
for direct instruction in the early grades the great majority o
which are either overlapping or complementary (but not contradictory):
Thus, in conclusion, I am happy to say that recent studies, linking
teacher behavior to student learning, and especially these experimental
studies, are making significant progress in developing a scientific,
basis lor teacher educati-
1 9
17
Reference Notes
Morsh, J., & Wilder, E. Identifyinginstructor:
review of the uantitative studies 190071952.Project No 7714.
Task No. 77248. Chanute Air ForceBase, Illinois: Air Force
Personnel and TrainingAtesearch Center, 1954.
2. Brophy, J., & Evertson, C. Process-(paint correlations in the Texas--
Teacher Effectiveness Stud:: Final re ortResearch Report no 74-4),1
Austin, Texas: Research and DevelopmentCenter for Teacher Education,
University of Texas, 1974.
Good, T., & Grouws, D. Process- rod_uctrelationships in fourth grade
mathematics classrooms. Final Report ofNational Institute of
Education Gra (NEG-00-30123). Columbia:University of Missouri,
1975.
McDonald, F & Elias, P. The eft.of teacher performance on
learning, _ ginning TeacherEvaluation Study: Phase II, final
report:
Vol. Princeton N.J.:Edu.,-:atiophl 'eating Service, 1976.
5. Stallings, J., & Kaskowitz,D. .Follow Through classroom observation
evaluation, 1972-1973 (SRI Project URU-7370).-Menlojark, Cali
'Sianford Research Institute, 1974,
Tikunoff Berliner, D., & RIst,R. An_ethnographic atudylfth
fort classroms of theTe_?cher
(Technical Report No. 75-10-5). San Francisco:Far West Laboratory
for Educational Research and Development,1975.
7. Medley'
D. Teacher corn =teice and teachreview
P
procelsictresearch. .Washington, D-JC. The American Association,-
of Colleges for Teacher Education,- 1977.
8. Rodenshine, B. New itmla11LapaLaH!EIiopsfrom recent research
21ementaajsheelf Paper presented atthe annual meeting of h
Educational Research Association, 1977.
11.
Classroom actices and basicskill_ Kinde_ erten and third
)Lale. Division of Research,North Carolina State Department of Public
Instruction, 1977.
Stallings, J., & Hentzell, S. Effective. eachhnd-learnin
schools. Paper presented atthe National Conference-on Urban
Education, 1978.
Evertson, C., Anderson, L., & Brophy, J.Texas 4nivalanqq1221alAx:
IiisE.rfpmt21_RLossa-outcomerelationships, Vblume 1 .(Researcb Report
No. 4061). Adstin: Research and. DevelopmentCenterifor Teacher Education,
The Univer&ity of Texas at Austin,1978.
12. ,Fisher, C, Filby, N.,MarliaVe, R., Cahen, Dishaw, M., Moore,
J;,
Berliner, D. Teaching behaviorsacademic. learnin t me and student
achievement. Final report ofPhase II;I -B, Beginning 'Teacher Evaluation
Study. San,FranciscoFar West Laboratory for Educational
Research and
Developmenti, 1978.
0
13. Murnane, R., & Phillips, B. Effective teache±
Who the., are and _what_ they do. Princeton, New
Policy Research, 1978.
18
inner c en;
ersey: athematica
14. Trismen, D., Waller, M., & Wilder, G.C.lass.room observation in a stud
iplfcompenatoryreading_programl. Paper presented at the annual
meeting of the American Educational Research Association, 1977.
15. Program on Teaching Effectiveness.Arieacfriment on cher effectiveness
and arent-assisted instruction inthe third grade. get of five papers
presented at the annual meeting of the American EducationalResearch
Association, 1978.
16. Eyertson, C., & Anderson, L. Theassroom_oi7anization stud- : Interim
pt2Ezt (Report No. 6002). Austin:Research and.Development
Center for Teacher Education, The Universityof Texas at Austin, 1978.
17. Stallings, J.Is,shlag_t2.Airpaons skills in the secondary schools.
Paper presented at the annual meeting of theAmerican Educational
Research ASsociation 1978.
18. McConnell, J. Relationshi.s betweenselected teacher behaviors and
attitudes/achievemen ebra 'classesPaper preSented at the annual
meeting of the American Educational ResearchAssociation,1977.
19. Soar, R.S., & Soar, R.M. Settinvariables, classroom interaction and
multiple outcomes.,. Final report for National Instituteof Education,
Project No. 6-0432. Gainesville:, University ofFlorida, 1978.
19
References
Anderson, L., Evert or, C. & Brophy, J. An expprimentalstudy of
effective torching in first-grade reading groups.
Journal, 19 79, 193-223.
Anderson, L., & Scott, C. The relationship amongteaching methods, student
characteristics and student involvement in learning.Jobrnal of Teacher
Education, 1978, 29(3), 52-57,
.
in classrooms. In L. Shulman (Ed.),
Review of Research in Education (Vol. 5). Itasca, Illinois:Peacock, 1977
Bennett, N. Teachin -les and o ress.Cambridge, Mass. Harvard
University Press, 1976.
Barr, R., & Dreeben, R.. Instruction
Borich, G. (with Fenton, K.).If2tairsiaTtot teach n Conceits and
process. Reading, Mass.: AddisonWesley, 1977.
Brophy, J., & Evertson, C. IL:e2IRkkai_E2!Iteaching: A developmental
2fElpecCive. Boston: Allyn and Bacon, 1976.
BLophy,-J., & Putnam, J. Classroom management in the elementarygrades.
In D. Duke (Ed.), Classroom management. 78th earbook of the National
SocieStyIdofEducation. Chicago University of ChicagoPress, 1979. .An expanded version is available as IRT Research
Series
No. '32. East Lansing, Mi.: Institute for Research onTeaching,
Michigan State University, 1978.
Coleman, J., et al. Lguality of educationalWashington, D.C.:
uperintendent of.Documents, U.S: Government Printing Office, 1966.
Cunningham, W. Theimpact of Student-teacher pairings onteacher effective-
nees. American Edueational Rebearch Journal,1975, 12, 169-189./
Dunkin, & Biddle, B. study ofMolt, Rinehart
- & Winston, 1974.
Ebmeier, H., & Good, T. An invest gation'of the interactive effects among
student types, teacher types, and instruction types on the mathematics
achievement c).f fourth grade students. American Educational Research,
Journal, in press.
Flanders, N. Aaalarjm-b
'Gage N. The-011ege Pre
vier. Readingss.: Addison-Wesley,
c basis of the art of:teachina New York: Teachers
Columbia University, 1978.
Good,.T. Teacher effectiveness in the elementaryschool: What we know about
it now. Jot- aal of Teacher Education, in press.'
Good, T.Rol
Biddle, B., & Brophy, J. Teache d ake a difference. New York:
Rinehart and Winston, 1975.
22
Good, T., & Grouws, D. .The Missouri MatherAn experimental study in fourth grade class oo s. Journal of
Educational o-, in press.
20
iveness Project:
Kounin, J. Discipline rou rfpepent in cl srooms. New York:
Holt, Rinehart and Winston, 1970.
Medley, D., & Mitzel, H. Measuring classroom behaViorJoy systematic ob-
servation. In N. Gage (Ed.), Handbook of research on teachin
Chicago: Rand McNally, 1963.
Peterson, P. Interactive effects of student anxiety, achievement oa-
tion, and teacher behavior on student achievement and attitude.
Journal of Educational Ps cholggyi 1977, 69,,779-792.
Rakow, E., P., & Madaus, G. Assessing school and programeffective-
fleas: Estimating teacher level effects. Journal of Educational
Measurement, 1978, 15, 15-21.
Rosenshine, B. Content, time, anti uctIon.In P. Peterson and
H. Walberg (Eds.), Research on teachin C9
implications. Berkeley, Ca.: McCatchan, in press.
Rosenshine; B. Recent research on teaching behaviors and student achieve-
'went. Journal of Teachei' Education 1976 7 27,'61 -64.
Rosenshine, B., & Berlin'er,- D. Academic -Aga ec time. British Journal
Teacher Education, 1978, -4, 3-16.
Rosenshine, B,, & Furst-, N. The use of direct ';r1)servation to study teaching.
In R. Travers (Ed.). Second handbook of research on teaching.Chicago:
Rand McNally, 1973.
Soar, R.S. 1S Soar, R.M. An empirical analysis of selected Follow Through,
programs: An example of a process approach to evaluation.In I.
Cordon (Ed.), Earl= ,t education. Chicago: National Society
for the` Study of Education, 19772.
Solomoni, D., & Kendall, A. Individual charaterist:cs and children's
forrnance in "opeh" and "traditional" classroom settings'. Journal'
f Educational Psychology, 1976, 68,613-625.
Stephens, J. nerc,cesp.inL. New York: Holt, Rinehart and Winston,1967.
47t
PUBLICATIONS
Of-the
Institute for Research op reachirsNtchigan StateUnieerresty
250f
rebrvary 1, 1979
To order any of ihe following' publicatIons please send check, Insmey order,or prepaid purchase order -- payabie to Michigan State University-- to IRTPublications, 252 erlaSon, MS0. East Lansing, NT 48F114. ,PublicatIon prices in-clude only the-cost of productIor and mal.ling. Michigan residents should add a4 state sales tam to all orders. The Institute for Research on Teach,ng isflinded primarily by the Nation31 Tnstitute of Education, Unito4 Stares Dcpartmentof Health, Educ.7,tien, and Welfare.
Research cries
No. 1 The Inuiry Theory: An informaLion-pr,wci;*:,, Qpproaciv to clin-1-cajprehtem7sol v in ref.e:Ireh and applioa. Vinsemhafer, C .C. Wn1,ner, &A.S. Fl Loin '1978. $2.00.
1.
No 2 -Impact on what?- The ipoi t nr of ront,T07 cover_od, A.C. Poi id. W .1.Schmidt, R.E. Floden, & D.J. Freeman. 19/8. $1.75 ,
No,. 3 Ins_tructlow--:_for nfAinc!, rho AJ 83:2_ill()Lr.in.2,ofcware,pn_c1:,111, as modifjedfor the CDC 6500 Oonlimter.' M.C:arlyn. 1977, $1.75.
No. 4 Traininl manual for Lhe elaifloi.ion of Lhcoont.entof fourth7grado..mathematics. 14.11.'SchmidL, A.C. Porter, R.E. Floden, D.J. 'Freeman.1978. $2.25
No, 55 Flexner, acc,redtation:6 andcvpluation, R.E. Pin-den. 1978: $1.75
No 6 Analoov and crN1t,nitialfing. RFE Flodon.. 1978. $1,25
a. 7 Conceptions of roadln,g: The Rep Test. -M. Johnston. 1978. $1.00
No. 8. Bovarch andAeveleTmeftt,peed5!for the advancement of teacherJ.E. Lanier & R.E, Floden. 1978, $3.00
N(,) 9 On:iHe:coricontnalizatton Of_Minical_problem o1 1i C.C.C C' Wahuer &J.F. VInsonhalor. 1978. $1.50
education.
* No. 10
No. 11
Ayillfeations of the lfiqpirvotheou Lo roaqin _and JearnlnlydisaMiqes_.L. Patriarcn., VanRookol, ,& J.F. Vironhaler. 1979.
3h701110S_SorL111.seovrivo_ rtiosO waftllaj142c1rs.C.M. Clark & R.J. ringor.. T;1978.. $1.00
Si
'24
dds
No. 12
No. 13
No.
a
14
No. 15
No. 16
No. 17
18
No. 19
No. 0
No'. 21
No. 22
No.
N
Research onteaeher . R.J
rlaa _(1,11r.!
IETILD(1(1141vjclual
CLIPIR_Pilot Chsetvat-i: fee -& A. Weinshank.
nw_.r. 1978. $2.25
t)il't :/ 1
The Inc-wiry Theoretic ;Tip).
/11-0°W";11-Vol, 1. J.P. Vinsonhalor, C.C. V
1978. $4.50
rph npd
mn
indman.
Tlieuser's manual for the Ba rmntion:Svo B_NTS
technical manual Vol__. C.- fler, J.F. Vinvooholci', A.S. L stein,
L.S. Shulman. 1978. $17.00
Teaell2IaLf2pLuponsof reading: The ,_volute of a
R. Barr & Duffy,- -.1978. $U'
-A study-. of teqsbEI2)!ntiint: Dcrlc tic?n and mo+lcl t f ITeactive do oiion
making.' R.J. Yinger,. 1978. $4.25
Fieldwork-as halts for .theor buldin_a_in research on tea-h . R.J. Ying,
1978. $2.25
Choice of. a model
$1.50
ch on teacher thinkin C.M. Clark. 1978.
Conceptual issues 10_ content-
1978. $2.25
An econOmeD.. Saks. 1978 $1.75'
v disc cion. D. Frccuuar
on classroom B.W. Brown
DiffILLLyjng, cues for use in stud tencherIEL2iL. c-R.J. Yinger, & S.C. Wildfoug. 1.9x81. $1.75
ruttonomv mllhe control of content A.C. Porter. 1978, '$3.75-24' Teac
No. 25
*No' i 26
No 27
Clark,
Donit_th all measurb_the 'same thin3Conscwimes.df standardizedelection. R.E. Floden, A.C. Portet, w.n. Schmidt, & D.J. Freeman. 1978. $1,50
Critical moments in the teaChina_ o_ mathemUtics. J.C. Shroyer1978.
Char4ctqyistics of the clinical, apabl.em-solv118 model and ltsirelevance
to educa4onnl_ research. .(ForMerly available as an IRT collateral paperL.S. Shn111 J.F. Vinsonhaleri,C.C. Wagner, &. L. Bader.
1978. $1-.25
* NO." 28 The conals:tency_nfreadfn .diaga psis. A. `Wetl.nslrank. 1978.
4 No. 30
by lufluyy Theory_re_sprchh C.
Itatin,7 the_iniohlAmt
1 Tra i s_prcf sts
e
No. ,32,Classr r m mannAmont in the 01o.,.-,ntary
trl
.C. 'utuam. 1916. $3.25
An othnrwruhlejtudy of a toP...11cy'f-:
for curriculum. V.J.19-/8,
The problem of dead_l
witinE S. Florio. 1978. $1.50
No. 35 Mcasurinz content o
*No. 36 The
Sh
-struetion. W.H. hm dt.1978. $1.5
11 or
ship of te.-cher aliena fto schoel workplace eha--
torast3cs
and pare or st et crf teachers.M.J. Vavrus. 1978 $250 ,
No. 37 The rcl-tionshi)between diagnosis`
study. We Lnshttnk.1978. $1.75
=
* No. Teacher judgment of 'child en'sByers. 1979. $2.00
nd -median itt )-eading:
ling p,refer&uces.Evans &
at South Bay_S'chol-. Joyce1979. $2 . ?.5
* . 39 The achinz sev,
* No. 40 Teachers' thoughts while -teacl!faa.B. Joyce 1979. $2
41 Measurin toacltcconceptions about rcading.
G.G. Duffy & W. Meeheny.
_1979. $1.75A
* 111: 42 Studies of,clinical_problatinzbehavior in re adinp diagnosis. D. Gil
J,11. Vinsonhaler, & G.C: Wagner.1979. $2.00'
* No. 43 Study lue /_beliefittetns of teachers and administrators. :
P. Chbick 1979.6=.
No. 44 Teochu_EacRpLions rstudent affect. R.S. Prawat.
1979. $2.50
Ar
Occasional Papers
1 Teachers'- con nd toneut1(A G. B6111,,
number of free copies.
of re dip and the eeachin of roaclin:Blau & A. Reisor. 1977. Limited
No. 2 Teachers And - se'_. ,-------hers:.___ TownrcLnswalALLIILIJ:'
be r. C. Kennedy.
1977. Limited number offv0e coNes.
3 A:causal analypis.of Attitudestownrship training in 4plassroom,
setpInE, J.E. Hunter,It. F. 1461ter & ;
'26Lapis. 1978. $2.00
No. 4 :1170 twich coljoapua In _classr_(
1978.
No. 5 Form and funci:lon mDonahue. 1978. $1.75 4.
Nei 6' Individual school lolldinT. do aoc oon: i ttr di!
Performance'. L. VI. -LezoL & lin !-:!;,.1. I , .(1
No. 7 Besearch on1978. $1.25
No.
chin A dynim :Ire
t de A vlew f rompraetj
Na. 9 RelationShips hctween tosm
Walsh.
u-t:4:11,. MA.
J.E. Lior.
h.S. Shuh 11. 1978. $1.75
A.C. Porter. 1978. $1.00
sd: 1976. L.S. ShLlm 1.No. _10___RsclioloiLyi piathemati.c s educJ it on r
1978. $1.75
No. 11 Science and mathematics, educatiov. .trouse: acid prpsnect.L.S.
Shulman & P Pamir. 1978. $2.00
No 12 Rel a tin ice in cdotat ional reriearch: A:inipaper.
L.S. Shulman. 1978. $1.75
No. 13 Glim.sin dt.scp: line: Toward a in+.,nn ;ri.c modr1 intec!rat fug_ teachers'thouThts and action s. D. Gil & P.S. Heiler. 1978. $1.00
No. 14 Attention and modnlil)2IELQLts in STg: A second look.T.Evnns & J.
Byers. 1978. $1.75
15 1~lere cthnngiap_hy :
F. Nricksc.n. 1979. $ 2.00
in it .use in educational practic
16 On standard of descri Live validity in Studies of eldssr.opm
F. Erickson. $2.00
No. 17 Ch.:1!.1 ksehool, chaa icte_ristics coincicyLLLwijilolarTesin
achievement. . W. B. Brookover, & 1, .14. Lezotte. ,1979. $5.100(Executive Summary $1.*00)
Con e en les
tudent
No. Current lirect .ons in r_esenreh ,on 'tenehin1) ...._mv,ELtmor thecAkuL:larElmL9aIa104:ELN vorcher 17- 1 9, 1= 9 TO_e 1977. $4.25
4 ,
R QL0 of a- nar on field research tncthuds iri n dot ation.- Cusick,1978. $1.5C.
Procgoc____Bms of the Research-. -'reaching 1 ma ics 'Conja1977. 1978. $8.25
Tcachors qtminfur ralcs
oommunitz. Shalaway, J.E
in production and will bo avallablc2 l5y nly 1. 1
exact price and publication dat :!.
for