The Experimental Ecology of Education

Author(s): Urie Bronfenbrenner

Source: Educational Researcher, Vol. 5, No. 9 (Oct., 1976), pp. 5-15
Published by: American Educational Research Association
Stable URL: http://www.jstor.org/stable/1174755
Accessed: 23-12-2016 23:27 UTC
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.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

American Educational Research Association is collaborating with JSTOR to digitize, preserve and
extend access to Educational Researcher

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

The Experimental Ecology of Education'

URIE BRONFENBRENNER

Cornell University

Ishall begin by stating threeThe

basic
scientific study of both sets
of
nificant
inferences - in short, is
requirements that, in my relations
judg- as they affect learning conmore elegant and constitutes
ment, must be met if we are to
makethe ecology of education and
stitutes
"harder" science than the best posprogress in the scientific study
of
represents
a major and necessary
sible contrived experiment adfocus for educational research.
educational systems and processes.
dressed to the same research ques1. Our researches cannot be re3. The strategy of choice fortion.
investigating person-environment and
istricted to the laboratory; for the
environment-environment relations
The Ecological Structure of the
most part, they must be carried out
in real-life educational settings. is
Asthe ecological experiment, defined Educational Environment
as a systematic contrast between two
will be indicated below, this does
The environment is conceived
or more environmental systems, or topologically as a nested arrangenot mean that laboratory experitheir structural components, with a ment of structures, each contained
ments cannot serve a useful and,
indeed, essential purpose, but theycareful attempt to control for possi- within the next. For the purpose of
must be carried out with explicit bly confounding influences, either describing these successive levels, I
recognition of the delimiting and by random assignment or by match- shall employ a terminology adapted
from Brim (1975).
distorting nature of the laboratorying on subject characteristics or

as a setting and deliberately de- other relevant factors.

tion I regard as unwarranted on

1) A micro-system is an immediate setting containing the learner

(e.g., home, day care center,
classroom, etc.) A setting is defined as a place in which the occupants engage in particular activities in particular roles (e.g.,
parent, teacher, pupil, etc.) for
particular periods of time. The
factors of place, time, activity,
and role constitute the elements

strictly scientific grounds. As I shall

of a setting.

signed to articulate closely with and Henceforth, I shall refer to the

complement companion researches former as a contrived experiment,
carried out in real-life situations.
and the latter as an experiment of
2. Whether and how people learn nature or natural experiment. I dein educational settings is a function liberately eschew the term typically
of sets of forces, or systems, at two employed in the literature of statislevels:
tical design - "quasi-experimental"
a. The first comprises the rela-

tions between the characteristics

of learners and the surroundings

in which they live out their lives
(e.g., home, school, peer group,
work place, neighborhood, community).

b. The second encompasses the

relations and interconnections

that exist between these environments.

-because it suggests a lower level

of methodological rigor, an implica-

endeavor to show, there are many

instances in which a design exploiting an experiment of nature provides a more critical contrast, insures greater objectivity, permits
more precise and theoretically sig-

October

2) The meso-system comprises

the interrelations among the major settings containing the learner
at a particular point in his or her
life. Thus, for an American elementary school child, the meso-

1976

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

system typically encompasses interactions among family, school,

peer group, television; for some
children, it might include as well
church, camp, or work place, although the last would be less
common in the United States
than in some other societies. In

situation implied in our definition of

the ecology of education, draws

heavily on the theories of Kurt
Lewin (1935, 1936, 1948, 1951). Indeed, this paper may be viewed as
an attempt to provide psychological
and sociological substance to Lewin's brilliantly conceived topological

sum, stated succinctly, the meso-territories.

gy of education. To adapt Dearborn's dictum to this domain: if

you wish to understand the relation

between the learner and some aspect
of the environment, try to budge the

one and see what happens to the

other. Implicit in this injunction is

system is the system of micro-sys-

The Experiment as a

tems.

asset that makes its early applica-

tion critical for research in the ecolo-

3) The exo-system is an extensionHeuristic Strategy

of the meso-system embracing the In ecological research, the early
concrete social structures, both
use of the experiment becomes deformal and informal, that imsirable for heuristic purposes; namepinge upon or encompass the imly, to analyze systematically the nature of the relation that exists bemediate settings containing the
tween the learner and the surroundlearner and, thereby, influence

the recognition that the relation be-

tween person and environment has

the properties of a system with a

momentum of its own; the only way

to discover the nature of this inertia,

and its interdependencies, is to try

to disturb the existing balance.
It is from this perspective that the

primary purpose of the ecological

ing milieu. That relation is not an experiment becomes not hypothesiswhat goes on there. These struc-easy phenomenon to recognize. It testing but discovery-the identifitures include the major institu-falls within the purview of those cation of those systems-properties
tions of the society, both deliber-events of which Goethe wrote with
and processes that affect, and are
ately structured and spontaneoushis poet's- prescience: "Was ist das affected by, the behavior and dely evolving, as they operate at theSchwerste van allem? Was dir has
velopment of the learner. Since the
local community level. These en-Leichtests dunket, mit den Augen
environment as here conceived encompass, among others, the worldze sehen, was vor den Augen dir
compasses both immediate and
of work, the neighborhood, massdeigt." (What is the most difficult
larger social contexts, the experimedia, agencies of government of all? That which seems to you the
mental design cannot be simplistic;
(local, state, and national), the
easiest, to see with one's eyes what
it is necessarily complex. And, in
distribution of goods and servis lying before them.)
keeping with its heuristic function,
ices, communication and transIf looking is not enough, whatitismust fulfill more than the usual
portation facilities, and informal
one to do? How can observers
and essential requirement of consocial networks.
quicken their sensitivity to the trolling
critifor possibly confounding
4) Macro-systems are the overcal features of the observed? The
factors. It must perform the more
arching institutions of the culture
answer to this question was givenscientifically fruitful task of providor subculture, such as the ecoing a highly differentiated and
me a quarter of a century ago, long
nomic, social, educational, legal
before I was ready to appreciate it,thereby sensitive grid that makes
and political systems, of which loby my first mentor in graduate
possible more precise detection of
cal micro-, meso-, and exo-sysschool, Walter Fenno Dearborn. In differences and changes in the state
and even determine or delimit

tems are the concrete manifesta-

his quiet, crisp New England ac-and structure of ecological systems.

cent, he remarked: "BronfenbrenThis brings us to yet another and
ner, if you want to understand some-perhaps most important feature of
in structural terms but as carriers
thing, try to change it." And wheth-the ecological experiment that disof information and ideology that,
er one studies change by deliberately tinguishes it from the laboratory
both explicitly and implicitly, enaltering conditions in a contrived ex-prototype: The difference lies not
dow meaning and motivation to
periment or by systematically ex-only in the requirement of a real-life
particular agencies, social netploiting an "experiment of nature,"setting, but in the nature of the unworks, roles, activities, and their
the methodological objective is thederlying research model. In the
interrelations. Whether children,
same: to maximize one's sensitivityclassical psychological experiment,
parents, pupils, teachers, or other
to phenomena through the juxtaantecedent and consequent condipersons directly involved in the
position of contrasts. As with a
tions are couched in terms of varialearning process have any place
vernier, it is only when two similar bles that are conceived as linear, ador priority in these macro-systems
but different systems are put side by ditive, and distinct from each other.
is of especial importance in deterside, that one can begin to see clear- In an ecological model, status and
mining how such persons are
ly the nature of the differences be- change are viewed not solely in
treated and interact with each
tween them. The systematic juxta- terms of different levels of one or
other in different types of educaposition of the similar but different more separate variables, but also,
tional settings.
constitutes the core of the experi- and primarily, as differences in sysEspecially in its formal propermental method and creates its mag- tems or system states. To understand
ties, the foregoing conception of the
nifying power.
what is meant by these constructs,
environment, as well as the dynamBut the strategy of experimenta- we must undertake an analysis of
ic relation between person and
tion has an even more important the properties of systems involving
tions. Such macro-systems are
conceived and examined not only

ER

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

in distress. Similarly enhanced ef-Proposition 3. The criteria for the

ecological validity of a research
environment. These properties,fects
in in a real-life situation were obsetting for particular research
tained
in an ingenious exploitation
turn, must then be incorporated
in
participants are dictated by the
the research models we employ of
foran "experiment of nature" by
characteristics of the larger social
Seaver (1973) on the controversial
investigating the ecology of educaand cultural context from which
tion.
phenomenon of induced teacher exthe participants are drawn. This
pectancies first reported by Rosenis the requirement of contextual
The Micro-system: Properties of
thal and Jacobson (1968).
validity.
the relation between learner and

the Immediate Setting

Real-life vs. Laboratory

Proposition 1. An experiment is
ecologically valid when it is conducted in settings that occur in

the culture or subculture for other

than research purposes, or might

occur if social policies or practices were altered. The requirement of ecological validity applies to all the elements of the
setting; that is, place, time, roles,

and activities.

The foregoing proposition is violated, for example, whenever research participants are placed in a
setting, or asked to engage in a task,
that is alien to the socioeconomic,

Proposition 2. In contrast to the

common experimental practice of
employing only selected aspects

of or analogues to the real-life

situation (i.e., simulation) or introducing extraneous elements, an

But even in the absence of signifi-

cant ecological ambiguity or dis-

sonance, the tradition of the labora-

tory can put blinders on the re-

searcher once he moves out into the

world. Two types of omissions are

ecological experiment entails a

especially noteworthy in this regard.

distortions to a minimum. This is

first is illustrated in the otherwise

determined effort to keep such

the requirement of preserving the

Phenomenological validity. The

ecological integrity of the setting.

exemplary research of Seaver cited

above. Ironically, this to-date most

As in all science, the foregoing re-

definitive study of teacher expectan-

quirement represents an ideal that cy, did not actually investigate

can never be fully achieved, since teacher expectancy at all; it exthe very fact that the setting is be- amined only the presumed effects
ing investigated constitutes an intru- of such expectancy on the performsion. If a "determined effort is

ance of school children. Thus the

can
were never asked any quesethnic, or social milieu from which made," however, the intrusionteachers
espe- that might have shed light on
they come. The experiment must indeed be kept to a minimum, tions
how they viewed their pupils, or
also last long enough to approximate cially is an experiment of nature.
whether they held different attitudes
what happens in real life. Finally,
Contextual validity. The caution
the roles and activities in which the
or expectations toward youngsters
against placing research particiassigned to the experimental vs. the
participants are engaged should be pants in an ecologically ambiguous
control
group. (Also not examined
appropriate to the situation and situation applies not only to the
imin Seaver's study was the role of
have established social meaning formediate setting, but to the larger
time as an element of the setting.
the participants. From this point ofcontext from which the participants
the expectancy effect
view, it is noteworthy that the only are drawn. Specifically, if the Presumably,
locale
condition for which the laboratory into which the participants should
are have been greater when the
age in
interval between sibling pairs
turns out to be an ecologically validplaced, or the roles and activities
wasdo
shorter.) In consequence, as
setting is for studying the behaviorwhich they are asked to engage,
and development of researchers in not occur frequently in theirSeaver
own himself acknowledges, the
their native habitat.
are possibly confounded.
subculture, then, regardless offindings
how
common such experiences may Teacher
be in expectancy may not have
Preserving the integrity of the
been operative at all. Instead, the
the society at large, they become
setting. But what about the possitransmitter
of the message to the
ecologically invalid for the group
in
bility of bringing into the laboratory
younger
child
may have been the
question.
This
is
the
basis
for
the
people and pieces from the oustide
world, or reasonable facsimilies
severe criticism (Labov, 1967; older
Rie- sibling, and perhaps parents
thereof - so-called simulation experiments? Is not this a reasonable road

as well.
gel, 1975; Sroufe, 1970; Tulkin,

to reality? In the last analysis, this

pants'
against studies of social class
andviews of the experimental

The failure to obtain the partici1972) that has been properly levied

is an empirical question and some ethnic differences based on stand-

situation is typical of research in the

area, and represents yet another

ardized psychological tests, not to
transfer to the real-life situation of
mention experimental measures of
the limited perspective of the laboraconservation, performance on the
tory, in this instance, its exclusive
prisoner's dilemma, or other operations borrowed from the laboratory.focus on objective behavior to the
decrease manifestations of comneglect of subject elements-the
Particularly when these procedures
perceptions and feelings of the perare administered in university setpetence and consideration for others

data are becoming available on the

issue. In general, the results indicate
that the strangeness and ambiguity
of the laboratory situation tend to
increase negative feeling states and

(Ross et al, 1975; Lamb 1975b).

That naturalistic situations are also

sons serving as subjects in the

tings, the results may give a misleadexperiment.
ing picture of how effectively the

more likely to evoke constructive very same persons may in fact be The exclusion of the subjective
from the domain of rigorous scienactivity in adults is indicated in functioning in their own milieu.
tific inquiry in all likelihood had its
This consideration brings us to our
Piliavin, Rodin, and Piliavin's
origins in the desire to eliminate the
third proposition.
study (1969) of reactions to persons
October

1976

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

"personal equation" in early studies

in astronomy and physics. The stric-

ture has been seldom challenged in

experimental psychology, probably
because so much of the work has
been done with animals. But once

to attribute meaning to stimuli that

about the nature of the setting, the
roles, activities, and other elements
species. From an ecological perspecpresent, or other environmental feative, this implies that the impacttures
of that in fact account for the ob-

we differ most from subhuman

served differences or similarities. In

the setting cannot be understood

without some information on how

all these cases, the main emphasis is

again, a research model that maythe
besetting, and its various elements, on analyzing the differential characreasonably adequate for the study were
of
perceived by the participants.teristics of the children, not of the
behavior and development in subAccordingly, we come to our fourthsettings in which they are found. As
human species turns out to be insufproposition.
a result, interpretations of environ-

ficient for the human case.

Proposition 4. In contrast to the

The reason for the insufficiency

classical laboratory study, in
was recognized by sociologists
which the data are typically limearlier than by psychologists, since ited to objective measures of the

the former, from the very beginnings subject's performance, an ecologiof their discipline, were more ori- cal experiment cannot be solely
ented toward studying events in the behavioristic; provision must also
real world. It was the Chicago school be made for assessing each parof Cooley (1902), Mead (1934), and ticipant's definition of the situation, how he or she perceives the
in particular Thomas (Thomas,
setting
and its various elements.
1927; Thomas & Thomas, 1928;
This is the requirement of pheThomas & Znaniecki, 1927) that
nomenological analysis.

mental effects are often couched in

what Lewin (1935) called classtheoretical terms; that is, observed

differences in children from one or

another setting (e.g., lower class vs.

middle class, French vs. American,
day care vs. home care) are "explained" simply as attributes of the
context in question. And even when
the environment is described, it is in

terms of a static, self-contained

structure of relations and values that

stressed the importance of the person's subjective view -in Thomas's

makes no allowance for processes of

A number of implications follow interaction through which the belanguage the definition of the situa- from Proposition 4, of which one havior of participants in the systion-as a major determinant of ac- merits explicit mention as a corolla- tem is instigated, sustained, and
tion. Perhaps the only sociological ry.
developed. These deficiencies disproposition that approaches the
close, by default, the defining core
status of an immutable law is
Corollary 4a. Phenomenological
of an ecological approach to educaanalysis is especially important
Thomas's inexorable dictum: "If
tion; namely, its focus upon the dyfor the construct validation of exnamic relations between learners
men define situations as real, they
perimental manipulations and
are real in their consequences"
and their surrounds, with both the
outcomes; that is, an examination
(Thomas & Thomas, 1928, p. 572).
person
and the environment enof whether these elements are perIn psychology, the role of pergaged
in
reciprocal tensions and
ceived by the participants in a
ceived reality in influencing behavior manner consistent with the conactivities, and undergoing progresand development was first empha- ceptual definitions explicit and
sive changes over time.
sized by Lewin in his concept of the implicit in the research design.
It is only recently that investigaThis is the requirement of phe"psychological field" (Lewin, 1935;
tors have begun to employ research
Bronfenbrenner, 1951). The princi- nomenological validity.
models that allow not only for asple was applied to social psychosessing the effects upon children of
logical experiments in a classic
Participants in a vacuum. It is
exposure to different kinds of setpaper by MacLeod (1947). Drawnot only the subjective world that is
tings but also for analyzing the
ing upon the European tradition of excluded from the experimenter's structure and pattern of activity spephenomenological analysis in the
view in the classical laboratory cific to each setting as these affect
psychology of perception (Katz,
model; some features of objective and are affected by the developing
1911, 1930; Koffka, 1935; Kohler,
reality are also ignored. Thus, it is child. A case in point is the ongoing
1929, 1938; Wertheimer, 1911),
not uncommon, especially when the longitudinal study by Cochran
researcher moves out of the laboraMacLeod emphasized the need to
(1975, 1976) of the development of

answer the question: "What is

'there' for the individual . . . . What

is the social structure of the world

tory into the real-life setting, to Swedish children brought up in their

omit description of the setting itself

own homes, in family day care, and

and the people and activities within in group care.2
he is living in?" (p. 204).
it. For example, in many studies of
Unfortunately, these essentiallyalternative structures for child rearProposition 5. In contrast to the
conventional research model, in
theoretical analyses had little im- ing and education such as day care
which scientific attention is fopact upon empirical work, so thatvs. home care, intact vs. fathercused primarily on the behavior
experimental studies in education absent homes, kibbutz vs. family,
of
certain persons, all engaged in
continued to be overwhelmingly beexperimental teaching programs,
the same role and designated as
havioristic and thereby berefit ofand, especially, cross-cultural difexperimental participants, an eco"meaning." This omission is criticalferences in socialization and schoollogical experiment requires equal
in research on human beings, for,ing, data are reported only for outattention to the properties of the
as Mead pointed out (1934, pp. 304come measures (e.g., psychological
setting, in terms of both its physi355), it is precisely in our capacity
tests) with no information provided
cal and its social structure, and to
8

ER

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

the relations that obtain between

the properties of the setting and

the behavior manifested by the
participants. This is the require-

is one that many readers will recog-ment of recognizing the totality

nize and applaud. In the classical of the functional social system in

the setting.
psychological research model,

whether in the laboratory or in the

Beyond the dyad. As we have already noted, the classical psychoonly two parties-an experimenter,
ments in educational research. Havlogical experiment allows for only
identified solely, and apparently
two participants: E and S. In most
ing examined the properties that dis-still acceptably, as E; and, another
real-life settings, there are usually
tinguish the laboratory situation,
ment of setting analysis.

The role of laboratory experi-

field, there were, and often still are,

person equally informatively de-

more than two people acting in more

and its derivatives, from naturalistic scribed as S-the "subject." The
than two roles. Thus, in the home
settings, we are now in a position to
term subject is apt, for it reflects the

there is usually also a father and,

fact that, with few exceptions, the
often, other siblings and adults; in
process operating between E and S
valuable contribution to research
the
day care center, preschool, or
has been viewed as unidirectional;
school, there are other children, as
on the ecology of education. There
the experimenter presents the
well as care-taking adults. Even in
are three major prerequisites that
stimulus, and the subject gives thethose researches that take into acemerge from the preceding discusresponse. Nowadays, we all know

indicate the conditions under which

the former approach can make a

count the activities of more than

sion:

that the process goes both ways. In

two persons in differing roles, the
1) Laboratory experiments both
more formal terms:
behavior of each is usually analyzed
in their design and execution, must
Proposition
6.
In
contrast
to
the
separately and interpreted as an inbe combined and closely articulated
dependent effect. As a case in point,
with parallel researches in real-lifeconventional, unidirectional model typically employed in the lawe may consider recent work on
settings.
boratory, an ecological experifather-infant interaction. For a
2) The results of laboratory exment must allow for reciprocal
comprehensive review of this literaperiments must be interpreted with
processes; that is, not only the
ture,
see Lamb (1975a, 1976a).
due regard to the limitations and
effect of A on B, but also the efIn
more
general terms, the actual
possible distortions revealed by the
fect of B on A. This is the requiresystem operating in a given setting
ment of reciprocity.
foregoing analysis.
often extends beyond a simple dyad
3) Laboratory studies have the
The principle of reciprocity has
to triads, tetrads, etc., and this fact
advantage of being more readily
special significance when applied to must be taken into account. To state
amenable to rigorous experimental
educational research. It means, for
the issue in propositional form:
control. Their reduced ecological
example, that we should look not
validity, however, seriously limits
the generalizability of results to real-

life settings. Accordingly, in educational research, their primary scientific value is for the exploration and

clarification of hypotheses, not for

definitive testing.

only for the influence of the parent

on the development of the child, but

Proposition 8. In contrast to the

conventional dyadic research
model, which is limited to assessing the direct effect of two
agents on each other, the design
of an ecological experiment must

also for the effect of the child on the

development of the parent. One suspects that among the most significant psychological changes that
take place in adulthood are those

take into account the existence in

the setting of systems that include

that occur as a function of the be-

Analysis of the Setting

havior and development of our chil-

as a System

Even when educational research

dren.

taken into account, the requirementsis often disregarded in practice.

concern with mutual accommodation between learner and environ-

ment implies consideration of the

rics at Case Western Reserve Uni-

lations taking place between the

others, operating as a subsystem.
This is the requirement of analyz-

ing social interactions and second-order effects in the N+2 systems, where they are in fact pres-

ent.

conducted by a group of investiga-

tors from the Department of Pediat-

permit the indirect influence of

any one of these on the direct re-

While the thesis that most behav-

is conducted in a real-life setting,

ior in social situations is reciprocal
and all of the critical elements are
is generally accepted in principle, it
of an ecological model are not yetPositive examples are found in a
met so long as these elements are
series of ingenious ecological exexamined only one at a time. The
periments and follow-up studies

three or more elements and hence

To examine this principle in prac-

tice, one may refer to a series of in-

various factors, including the behav-versity (Hales et al., 1976; Kennel

vestigations by Lamb (1976b, 1976c,
ior of the different participants, si-et al., 1974; Klaus et al., 1970, 1972;
1976d), one of the few researchers

multaneously as members of a syson parent-child interaction who has

Ringler et al., 1975).
tem. The next set of propositions
employed a true three-person model
deals with these systems-properties Proposition 7. An ecological ex- in the analysis of his data.3
periment requires recognition of
as they apply to the immediate setThe presence of N+3 systems
the social system actually operating containing the learner.
and
associated higher effects is not
tive in the research setting. This
Reciprocity. It is a sign of some system will typically involve all likely to be recognized unless such
progress that the first systems- the participants, not excluding the possibilities are systematically conproperty to which we call attention experimenter. This is the require- sidered in the research setting. The
October

1976

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

need for such consideration dictates

the following proposition:

Proposition 9. The design and

analysis of an ecological experi-

ment in a setting involving more

than three persons (i.e., an N+3
system) must take into account
all possible subordinate systems
(i.e., dyads, triads, etc.) and the
potential higher-order effects as-

into account temporal and spatial

Barker & Schoggen, 1973; Barker
arrangements, including the ob& Wright, 1954) represents a notajects contained in the space,4 asble exception, although in their repossible indirect influences on so-

search, settings are conceived and

analyzed almost exclusively in beThe analysis of the setting in syshavioral terms with only incidental
cial interaction in the setting.

tems terms. We have now concluded

reference to their social-structural

our analyses of systems properties

properties. Usually we carry out our
in the immediate setting containing
researches either in the laboratory,
the
learner.
As
a
result,
we
are
in
a the home, or the classroom, but
or
sociated with them. This is the
seldom in more than one context
position
to
formulate
a
final
comrequirement of comprehensive
analysis of possible subsystems prehensive proposition regarding thesimultaneously. From a theoretical
viewpoint, we may note here a conand high-order effects within conceptual medel that underlies an
ecological experiment.
tinuity of the traditional research
settings.
Proposition 11. Whereas in a con- paradigm, now across domains; the
Are such complex phenomena

worth investigating? Have we not

come to the point of diminishing returns in examining details of negligible significance for learning and de-

velopment? Paradoxically, from the

perspective of an ecological theory,
the argument runs the other way.
As one moves from a dyad to an
N+2 system, the resulting structure
offers possibilities for greater sta-

bility and power through mutual as-

sistance, complementarity, spelling

each other off, and reinforcement,
provided both directly and indirectly through third parties. The optimal size of systems for various types

of educational tasks, and, more importantly, the optimal structure of

such systems remain empirical ques-

ventional research model anteced-

ent and consequent conditions are

conceptualized in terms of sepa-

restrictive two-person system at the

level of the individual becomes an

analogous person-in-single-context
model at the level of settings. Once
distinct from one another, linear, a second setting is introduced, the
and additive, in ecological experi-system becomes triadic and, accordments both antecedents and coningly, allows for the possibility of
sequences can be conceived as
second-order effects, now at the
rate variables that are treated as

variations in the structure or

level of settings. Such theoretical

state of the setting as a system;

that is, the organization of theenrichment generates an array of

new and provocative research queselements of place, time, roles, and

tions. Not only does it necessarily

activities that define a setting at
a particular point in its developintroduce a comparative perspecment. This general requirement
tive, but it also calls attention to the
for the ecological research model
importance of investigating joint
is referred to as the conceptualieffects and interactions between setzation and analysis of the setting
tings (for example, home and
in systems-terms.

school, family and children's peer

group, the peer group and the
school, etc.), and thereby highlights
tions, but surely worthy of investigathe possibility that events in one
tion. Some evidence that N+2 sysmilieu may influence the child's betems, particularly as they cut across
havior
and development in another.
settings, do have untapped potential
Thus
the
experience of a child in
the
propositions
they
illustrate,
deal
for enhancing educative and developmental processes appear in studies with a restricted segment of the day care, in the classroom, or the
informal peer group, may change
to be cited in support of subsequent environment. In our topological
the
pattern of activities and interacschema
of
the
ecological
field,
they
propositions. But before turning to
remain at the level of the microtion
with parents or siblings in the
these we must take note of yet anhome,
or vice versa, with consesystem,
the
immediate
situation
conother source of higher order effects.
quent
implications
for learning and
taining
the
child,
and
even
there,
The indirect impact of physical
development.
they
deal
with
only
one
setting
at
a
and temporal factors. EnvironIn order to examine the joint efmental influences in educative proc- time. We consider next the implications for our research model of treatfects of exposure to more than one
esses are of course not limited to
ing two or more settings simultane- setting, an ecological research model
human beings. However, in keeping

with the classic two-element re-

Examples of the application of

this principle in concrete research
design are cited in both preceding
and succeeding sections. All of the
foregoing examples, however, and

ously.

must have certain additional proper-

ties presented in the next series of

propositions. We begin with a general principle that outlines the
While learners have been studied
range of phenomena the research
tended to overlook the possible opmodel
must encompass.
in
a
variety
of
environments,
there
eration of higher order effects opare
few
investigations
in
which
the
erating indirectly.

search model, these influences are

The Meso-system:
usually thought of as acting directly
Relations Between Settings
on the learner. As a result, we have

behavior and development of the Proposition 12. In the traditional

Proposition 10. Among the eleresearch model, behavior and desame learners has been examined as
ments of the setting that can invelopment are investigated in one
stigate second-order effects are a function of their exposure to dif- setting at a time without regard
time factors and features of theferent settings. The work of Barker, to possible interdependencies bephysical surroundings. Ecological Schoggen, Wright, and their col- tween settings. An ecological apexperiments must therefore takeleagues (Barker & Bump, 1964;
proach invites consideration of
10

ER

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

the joint impact of two or more

settings or their elements. This is
the requirement, where more than
one setting occurs, of analyzing
interactions between settings.

The next proposition represents

an extension of Proposition 9 beyond a single setting.

Proposition 13. The design of an
ecological experiment involving
the same person in more than one
setting should take into account
the possible subsystems, and associated higher-order effects, that
could exist across settings.

This proposition illuminates the

shortcomings of existing research
bearing on the next transition commonly experienced by young children in Western society: that from
the home to the day care center or
preschool. The work in this area has
recently been reviewed by Bronfenbrenner (1976).
Reciprocity Between Settings

Although no data are presently

available documenting the effects

on the child of reciprocal influences

found that the teenagers' descrip-

tions and evaluations of the behav-

pact of a two-way process taking

place between home and school. The

phenomenon of reciprocity between
settings was first "brought home"
in almost literal fashion to the pres-

ent investigator a decade ago in connection with an ongoing program of

cross-cultural research on patterns
of child rearing. In each of our com-

parative studies from the first conducted in West Germany (Devereux

et al., 1962), to the most recent in
Israel (Devereux et al., 1974), we

ent schools as well). In statistical

ior of their parents varied systemati- terms, the error term employed for

cally with characteristics and activi- testing the main effect for culture
ties of the peer group, drawn pri- would have to be based not on varimarily from among the young per- ance among individuals but among
son's classmates in school. Conclassrooms. This requirement can be
versely, in two companion reaccomplished by conforming to the
searches, Condry and Siman (1974, condition set forth in the following
1976) report that the involvement of proposition.
children in informal peer groups was
less a function of positive attraction
than of perceived inattention and indifference in the home; peer-oriented
youngsters described their parents
as being less affectionate and less
firm in discipline. Taken together,
their findings reveal a reciprocal re-

lation between family and peer

Proposition 15. If the results of

an ecological experiment are to be
generalized to other settings of
the same type, several examples
of each type of setting must be in-

cluded in the research design.

This is the requirement of replication at the level of settings.

group settings. Since such reciprociAdherence to this requirement can

ty between settings is likely to be a have sobering consequences. For exgeneral ecological phenomenon, we ample, Olds (1976)5 in a compretake note of it in the next proposi- hensive experiment on cross-age
tion:
tutoring, provided in his design for
replication at the level of schools.
Proposition 14. The design of an
experiment involving more than
one setting must take into account the possibility of reciprocal
interactions between settings as

systems.

between the family and the day care

center, there is evidence on the im-

classrooms also drawn from differ-

Replication at the Level of Settings

The finding of significant differ-

ences from one classroom or peer

group to the next has important

methodological as well as substantive implications. Specifically, dis-

Under these circumstances, he found

no reliable evidence for the widely

publicized claim (Gartner, Kohler,
& Riessman, 1971) that older children teaching younger children
themselves gain in learning skills.

Olds' results, and his thorough

analysis of earlier work, strongly

suggest that the reported success
was a function of rather specific

situations and circumstances related

to a high frequency and length of

tutoring sessions, the teaching re-

regarding the possibility of such

variation seriously increases the
risk of a Type I error-that is,

sources made available to the tutors

not in fact exist. The nature of this

needs of the tutor and the skills that

by the school, and, most critically, a

claiming a difference when one does close match between the learning

pitfall is best revealed by an ex-he is teaching the younger child.

treme, but probably not wholly hy- Replication of settings as a

pothetical, example. Let us supposemethodological issue. Replication
observed the same seemingly un- that in a particular cross-cultural on a sufficient scale to permit relialikely phenomenon. Our data on
study, say, of cooperative behaviorsble generalization is usually beyond
among ten-year-olds, all of the chil-the means of an individual investichild rearing practices were obdren in each society were selected gator. The inclusion of more than
tained from reports by sixth-graders
from one classroom. Under these
in their school classrooms, and,
one setting in the research design is
within every culture including the circumstances, any significant difnevertheless desirable, again for
United States, we found significant ferences between the two groups
heuristic purposes. Such replication
differences in parental behavior might reflect variation not between
can not only alert the investigator
from one classroom to the next.
to unwarranted conclusions but,
cultures but simply between the
The probable origins of this phe- classrooms (as a function of teacher, more importantly, can illuminate
nomenon have been traced in a docgroup climate, etc.). To determine
relationships existing both within
toral dissertation by Siman (1973)whether two societies actually dif- and between settings that may proon the interaction between familyfered in children's cooperative be- foundly influence the educative
and peer group in the socializationhavior, it would be necessary to
process.
process. In a study of 41 naturally-demonstrate that cultural differThe foregoing line of argument
should not be interpreted to imply
occurring adolescent friendship ences overrode the variation among
classrooms within cultures (with the
that experiments limited to a single
groups in New York City, Siman
October

1976

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

11

cial networks, transportation systems, law enforcement practices,

shopping facilities, means of communication, patterns of recreation
of view of science and of economics.
and social life, and a host of other
There is no advantage in replicating ecological circumstances and events
that determine with whom and how
a mistake that can be recognized
from a single trial. But in carrying the learners spend their time; for
out such ecologically delimited ex- example, the fragmentation of the
periments, it is necessary to heed
extended family, the separation of
two methodological precautions suf- residential and business areas, the
ficiently important to warrant statebreakdown of social networks, the

setting are not worth doing. On the

contrary, given our heuristic *aims,
this is probably the most efficient
way to proceed, both from the point

disappearance of neighborhoods,

ment in a corollary.
Corollary 15a. When replication
of settings is minimal or completely precluded because of lim-

ited resources:

1' The selection of the specific

example(s) within a setting category (e.g., classrooms, schools,

neighborhoods) should involve
careful consideration of the range

of possible choices in order to

maximize what can be learned

from the particular cases chosen.

zoning ordinances, geographic and

social mobility, growth of singleparent families, the abolition of the
apprentice system, consolidated
schools, commuting, the working
mother, the delegation of child care
to specialists and others outside the
home, urban renewal, or the existence and character of an explicit

national policy on children and

families. In sum, here in the third

ple. For the same reason, a de-

liberate effort should be made to

identify specific characteristics of

the examples employed and take

them into account in the inter-

pretation of results.

the process of making human beings

human. It is this heuristic function

that is embodied in our next proposition.

Proposition 16. Research on the
ecology of education requires experiments that go beyond the im-

mediate setting containing the

learner to the examination of

larger contexts, both formal and

informal, that affect events within
the immediate setting.

As already indicated, examples

that meet the foregoing criteria are

difficult to find. We have been able

to discover only a few correlational

findings and fragmentary facts, and
offer three instances: Giovannoni

and Billingsley (1970); Scarr-Salapatek and Williams (1973); Gar-

barino (1976).

circle of our ecological model are Developmental Transitions as

2) In the interpretation of findwhole subcontinents waiting for sciings, it should be explicitly acentific exploration - waiting beknowledged that the observed recause,
to date, there have been very
sults may be specific to features
few investigations of exo-system efof the particular examples employed, and hence are not gen- fects on learning processes.
eralizable to other settings of the One might challenge this assertion
same kind until the findings areon the grounds that studies of social

cross validated in another sam-

searchers to aspects of the larger en-

vironment that may be critical for

class differences provide a massive

body of information about the impact of the larger environment on
learning. Such studies are certainly

Ecological Experiments
A number of such changes have
served as the focus of investigation

in the researches we have already

cited. To mention a few: a mother

is presented with her newborn infant
for the first time, the baby returns
home from the hospital, the child is

enrolled in a day care center, or

promoted to the next grade in

school. It is not difficult to think of

relevant, but they fail to meet a other situations along the same line:
basic requirement of our ecological the arrival of a sibling; the move
model; namely, in educational re-

In the absence of such precausearch, social class is usually treated

tions, there is the danger, in studies

as a variable rather than in systems-

lacking replication at the level of

terms as stipulated by Proposition

11.
settings, of treating as a main effect

a finding that is actually a higherIn fact, the properties of the reorder interaction specific to the parsearch model for investigating rela-

from preschool to school; getting a

new teacher; going to camp; graduations; "dropping out"; finding one's
first job; changing jobs; losing ajob;
marriage; becoming pregnant; having relatives or friends move in (and

out again); buying one's first family

ticular examples of the setting intions at the level of the exo-system
TV set, car, or home; vacations;
cluded in the sample.
are precisely those that have been
travel; moving; divorce; remarriage;
specified in our prior propositions;
changing careers; emigrating; or to
The Exo-system: Learning
the only difference is that these
return to the more universal-beSettings in Context
stipulations are now applied to set- coming sick; going to the hospital;
This section takes us beyond the
tings and systems beyond the im- getting well again; returning to
immediate settings containing the
mediate situation containing the work; and-the final experience to
learner to the level of structures enlearner and have impact on that im- which there are no exceptionscompassing or impinging upon these
mediate situation. In other words, death.
immediate settings. Such exo-sysexo-systems represent sources of
Systems-properties of developtems are both formal and informal:
higher-order effects from more re-mental transitions. We call attenexamples include the nature and re- mote regions of the environment.
tion to this varied array of events
quirements of the parents' work,
From this point of view, exo-sys-in everyday life not for their percharacteristics of the neighborhood, tems do not generate any new func- sonal but for their scientific sighealth and welfare services, govern- tional principles; their place and nificance. For each one constitutes,
ment agencies, the relations between
purpose in our theoretical schema isin effect, a ready-made experiment
school and community, informal so- essentially heuristic: to alert reof nature with a built-in, before-

12

ER

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

after design in which each subject

a brave beginning, a sad ending,as traditional and necessary in our

and an empty middle. Given a perculture. But here once again we are
over, these developmental transispective in which development is
brought to the last level of our ections are sufficiently diverse toinstigated
inand paced primarily byological structure-the ideological
volve every one of the settings events
and
within the organism-that and institutional macro-system. Besystems-properties set forth inis,
our
fore entering this new domain, we
by biological change-the outsixteen propositions thus far. To
come is a segmented science that
sum up our analysis of developmenbegin with, since they all take place abounds with knowledge about the tal transitions as contexts for ecin real-life settings, they are ecologearly years, grows less informative ological research by stating three
ically valid by definition. In terms through middle childhood and
additional propositions. Unlike
of the elements of the setting, they adolescence, and then becomes vir- their predecessors, which spoke
entail changes over time in role
tually silent for decades, until the mainly to theory and method, these
activity, and often place as well
organism begins to decline, when expand upon scope and substance.
(wife to mother, child at home to there is once again a spurt of sciProposition 17. A fruitful context
pupil at school, student to worker, entific activity.
for ecological research in educaetc.). The magnitude of the microDevelopmental experiments in
tion is provided by the developsystem expands and contracts with the ecology of education. To be sure,
mental transitions that typically
marriages, births, graduations, di- a number of events in the life cycle
occur in the life of the learner.
vorces, and deaths. Reciprocal pro- discussed above have been the obThese transitions include changes
cesses, second- and higher-order ef- jects of scientific study. But such
in role and setting as a function
fects are the rule, since a develop- investigations have seldom been
of the learner's maturation or of
mental change in the state and
planned and conducted for the exevents in the life cycle of others
status of one member of the sys- plicit purpose of assessing the imresponsible for his or her care
and education. Such transitions
tem invariably alters the relations pact of the experience upon probetween the others. Since almost
are to be conceived and analyzed
cesses of learning. And even when
as changes in ecological systems
every transition involves more than
this aim has been pursued, the rerather
than solely within indivione setting, these interactive prosearch design has typically been
duals.
cesses occur not only within but
cross-sectional rather than longituProposition 18. Developmental
also across setting boundaries, thus
dinal (as, for example, in most
transitions are not limited to the
involving interactions within higherstudies of home vs. day care). As
early years but recur, in various
order systems. For example, when
a result, the inquiry can shed little forms throughout the life of the
a child enters day care, the pattern
light on the learning process as a
learner. Hence the experimental
of family activities changes; a didevelopmental experience. Also,
ecology of education must incorvorce can alter the child's behavior
whether cross-sectional or longitu- porate a life-span perspective if it
is to do justice to the phenomena
in the classroom; dropping out of
dinal, studies to date, as already
within its purview.
school has reverberations in the
noted, have focused almost exclusserves as his own control. More-

family; and a new job in another

ively on one class of persons des-

Proposition 19. Developmental

transitions invite not only naturtown affects home, school, and

ignated as the experimental subalistic but also contrived experievery other learning environment. jects. The impact of a developmenments-those that introduce inThe last example calls attention tal transition not merely on the
novations in the established seto the fact that developmental tran-learner but on the enduring sysquence and structure of successive settings and events.
sitions often have their origin nottems of which he is a part (e.g.,
in the immediate learning setting,family, peer group, etc.) remains an
but in the larger world of those re-unexplored and scientifically prom- The Macro-system: Experiments
sponsible for the learner's educa-ising terrain for ecological research on Institutions and Ideolgies
in education.
tion and care. To state the case
To formulate our final proposiis the scientific potential of tion, we take cognizance of one
more broadly in terms of our Nor
theoretical framework, what is microdevelopmental transitions limited more delimiting characteristic of
system for one learner (e.g., the to their exploitation as experiments conventional research in education.
of nature. They may offer even The foreshortened perspective was
father mastering a new job) becomes exo-system for another (e.g., greater promise as contexts for con- first brought to my attention by
trived experiments. The researches Professor A. N. Leontiev of the
the child in the classroom).
Finally, developmental transi- of Klaus (1970, 1972) and of Scarr- University of Moscow. At the time,
Salapatek and Williams (1973) rep- a decade ago, I was an exchange
tions provide a structure for conceptualizing the dimension of time in resent cases in point. Instead of scientist at the Academy of Pedtreating the transitions in the con- agogical Sciences. We had been
our ecological model. The almost
ventional course regarded as normal discussing differences in the asexclusive focus, within both developmental and educational psychology, in our society, these investigators sumptions underlying educational
research in the Soviet Union and
on the properties of the individual introduced unorthodox innovations
with little reference to context has
which invite equally radical trans- in the United States. In summing
formation of the educational tranup his views, Professor Leontiev ofgenerated a curiously broken trajectory of theory and research that has sitions that have become regardedfered the following provocative
October

1976

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

13

judgment: "It seems to me that

with new social or educational

American researchers are constant-

they tend to be systems "in abscis-

sion," cut off from each other. Thus

forms as contexts for realizing
ly seeking to explain how the childhuman potential. "After all,"the
we home is separated from the

came to be what he is; we in the

say, "you can't change human school,

nathe peer group from the famture." This precept underlies our
ily, the school from the neighbornot how the child came to be what
national stance on social and educahood, and all of these settings lack
connections with the world of work.
he is, but how he can become whattional policy, and much of our educational science as well.
he not yet is."
Similarly, our research has charIt is obvious that such discourse
acteristically been confined to one
U.S.S.R. are striving to discover

setting at a time-the home, the

is not confined to settings and strucclassroom, the work place, etc.
This dissociation of social strucindeed moved from the mundane
cent of Dearborn's injunction, but

The Transforming Experiment

Leontiev's statement is reminis-

tures on the local scene. We have

tures has been increasing rapidly

structures of a particular communit goes much farther; indeed, in
in recent decades and has been acLeontiev's view, it is revolutionary
ity to the level of macro-systemsthe institutions, and their associated
complished by a parallel deteriorain its implications. Soviet psycholtion of socialization processes and
ogists often speak of what they call
ideologies-that pervade major segoutcomes (Bronfenbrenner, 1974c,
ments of the society or the culture
the "preobrazuyuschchi eksper1975). Hence experiments that uniment," the "transforming exper- as a whole. The implications of this
dertake to reverse the process by
iment." By this term they mean an shift for an ecological research modconstructing and strengthening inexperiment that radically restruc- el concern the nature of the conterconnections between ecological
tures the environment, producing a trasts to be employed in our expersystems offer promise both for scinew configuration that activates iments. It is one thing to compare
ientific understanding and for sopreviously unrealized behavioral the effects on education of systems
cial policy.
potentialties of the subject. Rus- or system elements already present
sian developmental psychologists
within the culture; it is quite another
Notes
have indeed been ingenious in de- to introduce experimental changes
IAn abbreviated version of a paper prethat represent a restructuring of
vising clever experiments that
pared for presentation as the AERA Award

evoked new patterns of response

established institutional forms and

Address at the 1976 Annual Meeting of the

American Educational Research Associaprimarily in the sphere of psycho- values.

motor and perceptual development
With these unorthodox thoughts, tion, San Francisco, Ca., April 19-23, 1976.
The full text of the paper will be published in
(Cole & Maltzman, 1969). But once we arrive at the last and most dethe December 1976 Teachers College Record.
Soviet research moves out of the
manding of our propositions definThe ideas set forth in this paper grow out of
laboratory, the control group dising the nature and scope of ecologwork carried out by the author as a Belding
Fellow of the Foundation for Child Developappears, systematic data yield place
ical experiments in education.

to anecdotal accounts, and the

Proposition 20. Research on the

"transforming experiment" degenecology of education requires exerates into a dutiful demonstration
periments involving the innovaof ideologically prescribed proces- tive restructuring of prevailing
ses and outcomes.
ecological systems in ways that
depart from existing institutional
For rather different reasons,
ideologies and structures by re"transforming experiments" in the
real world are equally rare in Amer- defining goals, roles, and activities, and providing interconnecican educational research. As Leontions between systems previously

ment. Appreciation is expressed to the Foundation and its staff, in particular Orville G.
Brim and Heidi Sigal.

2Cochran's research has been supported

by a grant from the FCD program on the
Ecology of Human Development.

3Lamb's research has been supported by a

grant from the FCD program on the Ecology
of Human Development.

4Although the rapid growth in recent

years in environmental psychology (e.g.

tiev implied, most of our scientificisolated from each other.
Proshansky et al., 1970; Moos, 1976) has led
ventures into social reality perpetuto a proliferation of studies on the impact of
The final phrase of Proposition physical factors on behavior, little of this reate the status quo; to the extent
that we include ecological contexts 20 deserves special comment. In the search has focused on learning processes
and even less on the indirect effects of the
in our research, we select and treat course of two decades of cross-culphysical environment through its impact on
tural research on child rearing and
them as sociological givens rather
those who deal with the learner rather than
than as evolving social systems sus- education the author has been imon the learner himself.
ceptible to significant and novel pressed with a distinctive feature 5Olds' research was supported by a
transformation. Thus we study so- of socialization in American sogrant from the FCD program on the Ecology
of Human Behavior.
cial class differences, ethnic difciety. I first perceived this characferences, rural-urban differences- teristic in terms of our marked segReferences
or, at the next level down, children regation by age (Bronfenbrenner,
Barker, R. G., & Gump, P. V. Big school,
a
from one- vs. two-parent homes, 1970), but have since seen it in
small school. Stanford, Calif.: Stanford Unilarge vs. small families - as if the
broader perspective; namely, comversity Press, 1964.
nature of these structures, and their
pared to those in other modern Barker,
in- R. G., & Schoggen, P. Qualities
of community life. San Francisco: Josseydevelopmental consequences, were dustrialized nations, socialization
Bass, 1973.
eternally fixed and unalterable, ex- systems in the United States are
Barker, R. G., & Wright, H. F. Midwest

cept, perhaps, by violent revolu- more dissociated from each other.

and its children. Evanston, Ill.: Row, Peterson, 1954.
tion. We are loath to experiment In terms of Lewinian theory (1936),
14

ER

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

Brim, O. G. Macro-structural influences

on child development and the need for

childhood social indicators. American Jour-

nal of Orthopsychiatry, 1975, 45, 516-524.

neglect among the poor: A study of parental

adequacy in families of three ethnic groups.
Child Welfare, 1970, 49, 196-204.
Goslin, D. Children in the world of work.

Lewin, K. Field theory in social science.

New York: Harper, 1951.
Maccoby, E. E. Television: Its impact on
school children. Public Opinion Quarterly,

Bronfenbrenner, U. Toward an integrated New Society: The Weekly Review of Social 1951, 15, 423-444.
theory of personality. In R. R. Black & G. Issues, 1971, 2, 409-412.
MacLeod, R. B. The phenomenological
V. Remsey (Eds.), Perception, an approach Hales, D., Kennell, J. H., & Sosa, R. How
approach to social psychology. Psychologito personality. New York: Ronald Press, early is early contact? Defining the limits of cal Review, 1947, 54, 193-210.
1951.
the sensitive period. Report submitted to the
Mead, G. H. Mind, self, and society. ChiBronfenbrenner, U. Two worlds of child- FCD Program on the Ecology of Human
cago: University of Chicago Press, 1934.
hood: U.S. and U.S.S.R. New York: Russell
Moos, R. H. The human context: EnvironDevelopment. New York: Foundation for
Child Development, 1976.
mental determinants of behavior. New York:
Sage Foundation, 1970.
Bronfenbrenner, U. Experimental humanKatz, D. Die Erscheinungsweisen der
Wiley-Interscience, 1976.
Farben. Leipzig: Barth, 1911.
ecology: A reorientation to theory and reOlds, D. Cross-age tutoring and parent
search on socialization. Invited address at
Katz, D. Der Afbau der Farbwelt. Leipzig: involvement. Doctoral dissertation, Cornell
Barth, 1930.
University, 1976.
the meeting of the American Psychological

Piliavin, I. M., Rodin, J., & Piliavin, J. A.

Association, New Orleans, August 1974. (a)Kennell, J. H., Jerauld, R., Wolfe, H.,
Chesler, D., Kreger, N. C., McAlpine, W., Good samaritanism: An underground
Bronfenbrenner, U. Is early intervention
Steffa, M., & Klaus, M. H. Maternal bephenomenon? Journal of Applied Social
effective? A report on longitudinal evaluaPsychology, 1969, 13, 289-299.
havior one year after early and extended
tions of preschool programs (Vol. 2). Washpost-partum contact. Developmental Medi- Proshansky, H. M. et al. Environmental
ington, D.C.: Department of HEW, Office
cine and Child Neurology, 1974, 16, 172-179.psychology: Man and his physical setting.
of Child Development, 1974.
New York: Holt, Rinehart & Winston, 1970.
Klaus, M. H., Jerauld, R., Kreger, N. C.,
Bronfenbrenner, U. The origins of alienaMcAlpine, W., Steffa, M., & Kennell, J. H. Riegel, K. F. Subject-object alienation in
tion. Scientific American, 1974, 231, 53-61.
Maternal attachment: Importance of the first
psychological experiments and testing.
(c)
post-partum days. New England Journal of
Human Development, 1975, 18, 181-193.
Bronfenbrenner, U. The ecology of human
Medicine, 1972, 286, 460-463.
Ringler, N., Kennell, J. H., Jarvella, R.,
development: A research perspective. Invited
Klaus, M. H., Kennell, J. H., Plumb, N.,
Navojosky, B. J., Klaus, M. H. Mother-toMaster Lecture at the meeting of the Ameri& Zuehlke, S. Human maternal behavior at
child speech at two years-effects of incan Psychological Association, Chicago,
the first contact with her young. Pediatrics,
creased early postnatal contact. Journal of
August-September 1975.
Pediatrics, 1975, 86, 141-144.
1970, 46, 187-192.
Cochran, M. The Swedish childrearing
Rosenthal, R., & Jacobson, L. Pygmalion
Koffka, K. Principles of Gestalt psycholostudy: An example of the ecological approach
gy. New York: Harcourt Brace, 1935.
in the classroom; teacher expectation and
to the study of human development. Paper
presented at the International Conference on
Kihler, K. Gestalt psychology. New York: pupils' intellectual development. New York:
Holt, Rinehart and Winston, 1968.
Liveright, 1929.
Ecological Factors in Human Development
Ross, G., Kagan, J., Zelazo, P., & Kotelsponsored by the International Society for
Koihler, K. The place' of value in a world
chuck, M. Separation protest in infants in
offacts. New York: Liveright, 1938.
the Study of Behavioral Development, UniLabov, W. Some sources of reading probhome and laboratory. Developmental Psyversity of Surrey, Guildford, England, July
1975.
chology, 1975, 11, 256-257.
lems for Negro speakers of nonstandard
Scarr-Salapatek, S., & Williams, M. L.
English. In A. Frazier (Ed.), New directions
Cochran, M. Proposal to the FCD Proin elementary English. Champaign, Ill.: Na- The effects of early stimulation on low-birthgram on the Ecology of Human Developweight infants. Child Development, 1973, 44,
ment. Cornell University, Department of tional Council of Teachers of English, 1967.
Lamb, M. E. Fathers: Forgotten contribu- 94-101.
Human Development and Family Studies,
Seaver, W. B. Effects of naturally induced
Ithaca, New York, 1976.
tors to child development. Human Development, 1975, 18, 245-266. (a)
teacher expectancies. Journal of Personality
Cole, M., & Maltzman, I. (Eds.), A handand Social Psychology, 1973, 28, 333-342.
Lamb, M. E. Infants, fathers, and mothbook of contemporary Soviet psychology.
Sherif, M. Experiments in group conflict.
New York: Basic Books, 1969.
ers: Interaction at 8 months of age in the
home and in the laboratory. Paper presented Scientific American, 1956, CXCV(2), 54-58.
Condry, J. C., & Siman, M. A. CharacSherif, M. et al. Intergroup conflict and
to the Eastern Psychological Association,
teristics of peer- and adult-oriented children.
cooperation: The Robbers Cave Experiment.
Journal of Marriage and the Family, 1974, New York, April 1975. (b)
Lamb, M. E. The role of the father: An
Norman: University of Oklahoma Book Ex36, 543-554.
overview. In M. E. Lamb (Ed.), The role of change, 1961.
Condry, J. C., & Siman, M. A. An experiSiman, M. A. Peer group influence during
the father in child development. New York:
mental study of adult vs. peer orientation.
adolescence: A study of 41 naturally-existing
Wiley, 1976. (a)
Unpublished manuscript, Cornell University,
1976.
Lamb, M. E. Interactions between eight- friendship groups. Doctoral dissertation,
month-olds and their fathers and mothers.
Cornell University, 1973.
Cooley, C. H. Human nature and the
In M. E. Lamb (Ed.), The role of the father Sroufe, L. A. A methodological and philosocial order. New York: Scribner's, 1902.
sophical critique of intervention-oriented
in child development. New York: Wiley,
Devereux, E. C., Jr., Bronfenbrenner, U.,
research. Developmental Psychology, 1970,
1976. (b)
Shouval, R., Rodgers, R. R., Kay Venaki,
2, 140-145.
Lamb, M. E. Parent-infant interaction in
S., Kiely, E., & Karson, E. Socialization
Thomas, W. I. The unadjusted girl. Boseight-month-olds. Child Psychiatry, 1976, 7,
practices of parents, teachers, and peers in
ton: Little, Brown, 1927.
in press. (c)
Israel: The kibbutz versus the city. Child
Thomas, W. I., & Thomas, E. S. The
Lamb, M. E. Effects of stress and cohort
Development, 1974, 45, 269-281.
on mother- and father-infant interaction.
child in America. New York: Knopf, 1928.
Devereux, E. C., Jr., Bronfenbrenner, U.,
Developmental Psychology, 1976, 12, in Thomas, W. I. & Znaniecki, F. The Polish
& Suci, G. Patterns of parent behavior in
peasant in Europe and America. New York:
press. (d)
America and West Germany: A cross-nationKnopf, 1927.
Lamb, M. E. The one-year-old's interacal comparison. International Social Science
Tulkin, S. R. An analysis of the concept
tion with its parents. Paper presented at
Journal, 1962, 14, 488-506.
of cultural deprivation. Developmental Psythe meeting of the Eastern Psychological AsGarbarino, J. A preliminary study of some
chology, 1972, 6, 326-339.
sociation. New York, April 1976. (e)
ecological correlates of child abuse: The
Wertheimer, M. Experimentelle Studien
Lewin, K. A dynamic theory of personaliimpact of socioeconomic stress on mothers.
iuber das Sehen von Bewegung. Zeitschrift
ty. New York: McGraw-Hill, 1935.
Child Development, 1976, 47, 178-185.
fur Psychologie, 1911, 61, 161-265.
Lewin, K. Problems of topological psyGartner, A., Kohler, M., & Riessman, F.
Zajonc, R. B. Family configuration and
chology. New York: McGraw-Hill, 1936.
Children teach children. New York: Harper
intelligence. Science, 1976, 192, 227-236.
Lewin, K. Resolving social conflict. New
& Row, 1971.
York: Harper, 1948.
Giovannoni, J., & Billingsley, A. Child

October

1976

This content downloaded from 164.73.224.2 on Fri, 23 Dec 2016 23:27:36 UTC
All use subject to http://about.jstor.org/terms

15