ADVENTURE AS A STIMULUS FOR COGNITIVE DEVELOPMENT EDWARD H. CORNELL, DIANNE C. HADLEY , TREENA M. STERLING , MELANIE A. CHAN, AND P ATRICIA BOECHLER University of Alberta Abstract As illustrated in two studies of the development of children’s independent way ¢nding, the happenstance ofadventure provides natural opportunities to shape analytical and strategic thinking. Although they had not been instructed, when walking to the limits of their home range, 12-year-olds more than 8-year-olds selectively attended to environmental features with good landmark qualities. There was evidence of learning: older chil- dren who had recently experienced the requirements of leading the way to and from a distant site increasingly noted landmarks in the skyline and landmarks near intersections. Universal adventures of childhood may be especi all y importan t for the dev el opme nt of sensi tivi ty to con textu al feed bac k and £ex ibil ity in ac hie ving goals. # 200 1 Academic Press Introduction It of ten happens that there is a di screpan cy be- tween what parents think their children are doing and wha t in fact their children are doin g. Pare nts ’ naivete ¤ about their children’s activities may be evi- dent in several important domains, such as indepen- dent travel from home, viewing of violent television programming, and sexual behavior. One of the more serious implications of parental naivete ¤ is that chil- dren are participating in potentially harmful activ- ities wit ho ut guid anc e. Parents may caution or fo rbid and ther eaf ter assume that thei r ch il d is av oiding an u nacc eptab le acti vity . The younger child who nev ertheless tries for bidden acti vities ma y be learning ways from peers or by the consequences ex- perienced as a result of his or her own independent attempts. The observations that we report here sug- gest that children’s self-directed adventures are im- portant opp ortu nities for wha t Sieg ler (1996) has chara cterized as the natural sel ecti on of cogni tiv e strategies. Strategy development In an earlier analysis, Siegler and Jenkins (1989) de- ¢ned cognit iv e stra tegies as non-o bl ig at ory and goal-directed sequences of acti vities. They di¡eren- tiated strategies from in variant procedural solu- tio ns, which children ma y repres en t as the onl y wa y to achie ve a goal. They also di¡er ent iat e stra te- gies from plans, which are considered to be volun- tar y and deli ber ate . Th us , Si eg le r and Jenkins ‘de¢ne str at egi es as di¡eri ng from procedures in that strategies necessarily involve choice, and as dif- fering from plans in that the choice process is not necessarily conscious’ (p. 12). As we shall illustrate, the se dist inc ti ons seem espe cia lly appro pria te to chara cteriz e ho w chil dren begin ¢nding their wa y in new territory. The developmen t of spa tial cog nit ion in large- scale en vironments is class ically des cribed as an age-stage sequence (Piaget & Inhelder, 1967; Piaget, et al ., 1960; Har t & Moore, 1973; Si ege l & White, 1975). Prior to puber ty , children’s spat ial prob lem solving is thought to be constrained by limited abil- iti es to reason and the predominan ce of landmar k and rou te-based forms of repre sen tati on. Siegler’s (1996) theory of strategy development suggests a dif- ferent descri pti on. The theory begins wit h the as- sumpti on tha t child ren have a v ariet y of cognit iv e strategies available and further assumes that there is creation , deconstructi on, and selecti on among strategies when new problems are confronted. Chil- dren’s initial attempts at solving a problem result in fe edb ac k, ou tcomes indi ca ti ng whi ch stra tegy proved most e¡ective in particular situations. Chil- dren may experiment with partial or ine⁄cient stra- tegi es, sometimes regressing to cl umsy methods. However , in ge ne ra l, ch ildren are assumed to Journal of Environmental Psychology (2001) 21, 219^231 0272 - 4944/01/030219 + 13$30. 00/0 # 2001 Academic Press doi:10.1006/jevp.2001.0207, available online at http://www.idealibrary.com on
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
EDWARD H. CORNELL, DIANNE C. HADLEY, TREENA M. STERLING, MELANIE A. CHAN, AND PATRICIA BOECHLER
University of Alberta
Abstract
As illustrated in two studies of the development of children’s independent way ¢nding, the happenstance of adventure provides natural opportunities to shape analytical and strategic thinking. Although they had not
been instructed, when walking to the limits of their home range, 12-year-olds more than 8-year-olds selectively
attended to environmental features with good landmark qualities. There was evidence of learning: older chil-
dren who had recently experienced the requirements of leading the way to and from a distant site increasingly
noted landmarks in the skyline and landmarks near intersections. Universal adventures of childhood may be
especially important for the development of sensitivity to contextual feedback and £exibility in achieving
goals. # 2001 Academic Press
Introduction
It often happens that there is a discrepancy be-
tween what parents think their children are doing
and what in fact their children are doing. Parents’naivete ¤ about their children’s activities may be evi-
dent in several important domains, such as indepen-
dent travel from home, viewing of violent television
programming, and sexual behavior. One of the more
serious implications of parental naivete ¤ is that chil-
dren are participating in potentially harmful activ-
ities without guidance. Parents may caution or
forbid and thereafter assume that their child is
avoiding an unacceptable activity. The younger child
who nevertheless tries forbidden activities may be
learning ways from peers or by the consequences ex-
perienced as a result of his or her own independentattempts. The observations that we report here sug-
gest that children’s self-directed adventures are im-
portant opportunities for what Siegler (1996) has
characterized as the natural selection of cognitive
strategies.
Strategy development
In an earlier analysis, Siegler and Jenkins (1989) de-
¢ned cognitive strategies as non-obligatory and
goal-directed sequences of activities. They di¡eren-
tiated strategies from invariant procedural solu-
tions, which children may represent as the only
way to achieve a goal. They also di¡erentiate strate-
gies from plans, which are considered to be volun-
tary and deliberate. Thus, Siegler and Jenkins
‘de¢ne strategies as di¡ering from procedures in
that strategies necessarily involve choice, and as dif-fering from plans in that the choice process is not
necessarily conscious’ (p. 12). As we shall illustrate,
et al ., 1960; Hart & Moore, 1973; Siegel & White,
1975). Prior to puberty, children’s spatial problem
solving is thought to be constrained by limited abil-
ities to reason and the predominance of landmarkand route-based forms of representation. Siegler’s
(1996) theory of strategy development suggests a dif-
ferent description. The theory begins with the as-
sumption that children have a variety of cognitive
strategies available and further assumes that there
is creation, deconstruction, and selection among
strategies when new problems are confronted. Chil-
dren’s initial attempts at solving a problem result in
feedback, outcomes indicating which strategy
proved most e¡ective in particular situations. Chil-
dren may experiment with partial or ine⁄cient stra-
tegies, sometimes regressing to clumsy methods.
However, in general, children are assumed to
Journal of Environmental Psychology (2001) 21, 219^231 0272 - 4944/01/030219 + 13$30.00/0# 2001 Academic Pressdoi:10.1006/jevp.2001.0207, available online at http://www.idealibrary.com on
year-olds. Each age group had an equal distribution
of boys and girls, mean ages 6 : l (range 5:9^6:6)
and 12:2 (range 11 : l0^12:5). The families resided in
middle- to upper-class suburban neighborhoods of
FIGURE 1. A child’s home range is represented on a cadastral map. The solid line between the child’s home (H: back door) and intendeddestination (ID: soccer ¢eld) is the crow’s £ight measure of the farthest distance travelled. The dashed line illustrates the actual path
walked by the child to reach the intended destination. The remaining solid line completes a wedge used to estimate the dispersion of the child’s travel.
Because of the di¡erence in scale of travel by the
two age groups, a logarithmic transformation was
applied to the crow’s £ight distance measures to as-
sess whether there was a discrepancy between the
actual travel by the child and the distance of the
location reported by parents to be their child’s
farthest destination. A 262 (Age group6Gender)
multivariate analysis of variance (MANOVA) of the
two distance measures indicated the interaction il-
lustrated in Figure 2, F (1,34)= 6Á25, p50
Á02. The lo-
cations of distant destinations known by parents of
12-year-olds averaged only 30 m more crow’s £ight
distance from home than the destinations they
walked to, but there was on average a 262 m under-
estimate of 6-year-old’s actual travel.
When asked to lead us to a place they knew they
could get to, 6 -year-olds seemed to lead the assis-
tant along direct routes. Route maps indicated few
turns and destinations that could be seen early in
the walk. To estimate the challenges of di¡erent
route choices, we created an index of extra travel,
measured as the actual distance travelled by thechild divided by the distance of the shortest possi-
ble route to the destination. Values over 1Á0 indicate
unnecessary travel. We found that, leading the walk
from home to destination, the mean value for the 12-
year-olds was 1Á24, indicating that they travelled
more than one ¢fth as far as they needed to,
whereas the mean extra distance index for the 6 -
year-olds on the outgoing walk was 1Á06 (S.D. = 0Á29
and 0Á19, for the older and younger groups, respec-
tively). A 262 (Age group6Gender) analysis of var-
iance (ANOVA) of the extra distance index
indicated the age e¡ect was reliable, F (1,35) = 7Á92,
p50Á01.
All children successfully reached their chosen
destination.When asked at their destinations if they
could ¢nd a di¡erent way home, 16 12-year-olds
elected to try, but only 11 6-year-olds did so,w2 = 3Á7, df = 1, p = 0Á05. Two 6-year-olds considered a
return on the parallel sidewalk on the opposite side
of the street to be a new route. The di¡erence be-
tween the age groups in extra distance travelled
during the walk from destination to home was not
reliable, F (1,35) = 0Á239. The mean extra distance in-
dex of 12-year-old children was 1Á23, whereas that of
6 -year-old children increased to 1Á17 (S.D. = 0Á25 and
0Á44 for the older and younger groups, respectively).
Scanning behavior
On average, the 12 -year-olds were recorded to be
scanning on M = 19 per cent (S.D. = 15) of their
M = 108min (S.D. = 56) walks, which was less than
the M =32 per cent (S.D.=18) of minute-sampled ob-
servations of the M =36min (S.D.=16) duration walks
of the 6-year-olds, as indicated by a 262 (Age
group6Gender) ANOVA, F (1,35) = 4Á94, p50Á05.
However, 12-year-olds were looking from side-to-side
M =88 per cent (S.D.=24) of the time they were re-
corded to be scanning, whereas 6-year-olds showed
M = 75 per cent (S.D. = 19) side-to-side scanning
F (1,32) = 2Á93, p50Á10, a marginal di¡erence re£ect-
ing younger children’s tendencies to look downwardat features of the path. The result is corroborated in
the study that follows. One 6 -year-old boy in the
present study volunteered that directing attention
downward was an important technique for him: ‘I
just know how to get there by looking at the ground.
All I need to look at is the ground’. Note that a
downward pattern of scanning would not help to
register landmarks that could be used if the boy
stepped o¡ path during his return.
Discussion
These results describe natural opportunities for
children to try way ¢nding strategies. Both the
duration of excursions and extraneous travel on the
way to their destination increased from early to
middle childhood. Parental emphasis on safety in-
structions left open the possibility that children
were learning to selectively attend to landmarks
during independent excursions or with peers.
Although some were cautious about attempting
new routes, 6 -year-old children were travelling be-
yond the limits their parents expected for them.
Their path choices for these excursions typically in-
volved linear extensions along established routes.
FIGURE 2. The discrepancy between parents’ reports of their chil-
dren’s travel and 6 -year-old children’s actual travel from urbanand suburban homes. Each bar represents the mean distance todestinations of 18 participants.
£ight distance traveled indicated no reliable e¡ects,
all F s 51Á6. When the two selected walks were com-
bined, the mean crow’s £ight distance from home to
destinations was 1116 and 1045 m, (S.D. = 324 and 732)
for the 12- and 8-year-olds, respectively. Hence, dif-
ferences in way ¢nding performance cannot be read-
ily attributed to di¡erences in the distance that the
children had to walk.
Route e⁄ciency. When asked prior to the walk, 8-
year-olds readily accepted the challenge to lead the
way to distant destinations that they had infre-
quently visited, but subsequently found it di⁄cult
to ¢nd their way to these sites. Indeed, two mea-
sures of route e⁄ciency indicated reliable di¡er-
ences between the 8-year-olds and 12-year-olds. The
paths taken by the younger children were more dis-
perse and they tended to walk farther than the least
distance route to the destination.
We estimated the dispersion of the children’s
paths as an angle. After the paths taken by thechild were drawn on a survey map, the paths were
bracketed within a wedge of a circle centered on the
child’s home (see Figure 1). The rationale for this
measure is that e⁄cient travel to a destination
should not involve excessive lateral displacements
from the crow’s £ight line between the origin of tra-
vel and the destination. Typically, displacements are
inevitable in cities because of the layout of blocks.
Regardless, the mean size of the smallest wedge to
contain the dispersion of the 12 -year-olds was 808
(S.D. = 26) and the mean dispersion of 6 -year-olds’
paths was 1048 (S.D. =26). A26262 (Age Group6Gender6Name Landmarks Condition) ANOVA of
the angle of dispersion indicated only a main e¡ect
of age group, F (1,31) = 6Á96, p50Á02.
The children’s route e⁄ciency was also indexed by
extra travel, measured as the actual distance tra-
velled by the child divided by the distance of the
shortest possible route. A 26262 (Age Group6
Gender6Name Landmarks Condition) ANOVA of
the extra distance index indicated only a main ef-
fect of age group, F (1,31) = 8Á79, p50Á01. Over all
outgoing and incoming walks, the mean extra dis-
tance index for the 12-year-olds was 1Á14 (S.D. = 0Á11),
whereas the mean index for the 8-year-olds was 1Á49
(S.D. = 0Á41), indicating that they travelled almost half
again as far as they needed to.
To isolate the source of the age di¡erence in way
¢nding performance, the extra distance index was
calculated separately for travel from home to desti-
nation and from destination to home for both the
route reversal and new route procedures.
A 26262 ANOVA was conducted with age group
as a between-subjects variable and two within-sub- jects variables, direction of travel (from or toward
home) and requirements for returning (old or new
paths). The main e¡ect of age group was repeated,
F (l,30) = 10Á07, p50Á01, and Figure 3 illustrates a 3-
way interaction, F (1,30) = 4Á03, p = 0Á5: The younger
children especially had di⁄culty returning home
when asked to use a new route. Interestingly, even
when asked to return by the same paths they had
used to reach their destination, 7 of the 16 younger
children initiated shortcuts. Three of the shortcuts
generally followed a line of sight between segments
of the previously walked path; four shortcuts re-quired an inferred connection.
Scanning. Two measures of selective attention were
recorded during these walks. Because our earlier
observations indicated that time-sampling was not
necessary, all episodes of side-to-side or up-down
head movements of more than 158 were recorded.
A 26262 (Age Group6Gender6Name Landmarks
Condition) ANOVA of the number of scans indicated
no reliable e¡ects, all F s51Á2, indicating that age
groups did not di¡er in frequency of scanning
when travelling to relatively unfamiliar destina-
tions. The mean number of scanning episodes
FIGURE 3. The crossover illustrates that younger children tendedto wander when asked to take a new route home from a distantdestination but successfully attempted shortcuts when asked toreturn by familiar routes. The extra distance index is the dis-
tance in meters of the route used by the child divided by the dis-tance in meters of the shortest possible route. Each pointrepresents the index for 16 children.
progressed down the alley, she said ‘Oh, no! That’s
badöThere’s recycle boxes everywhere!’
E¡ects of experience. An important implication of
our hypothesis about the natural selection of strate-
gies is that experiences on the ¢rst walk should af-
fect measures of attention on the second walk. The
e¡ects of experience were assessed with 262
ANOVAs, with age group as a between-subjects vari-
able and ¢rst and second walk as levels in a within-
subjects variable. Note that subjects would not en-
ter into analyses if proportional data could not be
calculated for either walk.
There were no obvious changes in proportion of
horizontal scanning, although the main e¡ect of
age group was sustained, F (l, 30) = 4Á16, p = 0Á05. In
addition, there were two interactions indicating re-
liable age-related shifts in the qualities of land-
marks selected on the ¢rst and second walks. The
¢rst interaction involved the proportion of land-
marks named near intersections, F (1,19) = 22Á30,
p50Á01. During the second walk, 12 -year-olds se-
lected a greater proportion of landmarks near
FIGURE 4. Columns indicate the proportion of the total landmarks named by 16 children that were characterized by permanence, un-iqueness, proximity to intersections, or visibility in the distance. The dashed lines represent estimates of the baseline frequency of thefour characteristics.
intersections (M = 83%, S.D. = 13) than did 8-year-
olds (M = 47%, S.D=26), whereas there was no reliable
di¡erence between the age groups on the ¢rst walk
(M = 54% and 62% S.D. = 22 and 24, respectively, forthe older and younger children). The second interac-
tion involved the proportion of distant landmarks
named, F (1,19) = 5Á78, p50Á05. During the second
walk, 12 -year-olds selected a greater proportion of
distant landmarks (M = 49%, S.D. = 26) than did 8-
year-olds (M = 21%, S.D. = 14), whereas there was no
reliable di¡erence between the age groups on the
¢rst walk (M = 44% and 43% S.D. = 27 and 21, respec-
tively, for the older and younger children). Hence,
the older children chose more appropriate land-
marks after their experiences leading the way to
and from a distant destination. As explained byone 12 -year-old girl, ‘I can remember that this is
the corner I turn at to go home because that is my
school and that was my classroom door that I went
in every day’.
Diverse learning. Although we found that older
children more than younger children selectively at-
tended to objects with good landmark qualities, re-
cordings taken during the walks indicated that
attentive strategies were only one of a variety of
cognitive developments that would allow for e⁄-
cient way ¢nding. The development of a local knowl-
edge base is indicated here:
SG, an 11-year-old girl: ‘I know my way because of the bus route signs.’Research Assistant: ‘How do you know that thesesigns wont lead you onto a di¡erent bus route?’SG: ‘Because this is the only bus route around here’.
The ability to translate routes into con¢gurational
knowledge could also help with way ¢nding deci-
sions:
NK, an 8-year-old boy: ‘Hah! This street is kindalike a ‘U’, isn’t it.Look!öit goes back to the school we saw before.
That for sure is the school we saw’.
Finally, at least one child was beginning to use
calculations based on conventions of the urban grid
system:
LN, an 11-year-old girl: ‘‘Hmmm, 143rd street, and weneed to be on 146th street, so we can go along for afew more blocks. . .’’
These observations remind us that, although selec-
tive attention to landmarks is fundamental, there
are multiple solutions to the problem of human na-
vigation. A repertoire of these solutions seems to
develop as outcomes of adventure.
General Discussion
The development of home range
Interestingly, the extent of travel we observed in our
¢rst study was substantially greater than what has
been estimated from structured interviews with
children. For example, Matthews (1987) recorded
that 6-year-olds in the suburbs of Coventry,
England, named places they could travel to alone
that were 100 m from their home, and the children
reported that they had been to places with older
children that were 290 m from their home. Our ob-
servations indicate 3^4 times more actual travel. In
addition, Matthews and others (Coates &
Bussard, 1974; Hart, 1979; Payne & Jones, 1977) havenoted gender di¡erences in home range, with par-
ents reporting more constraints on their daughters
and some girls themselves reporting closer range of
experience than similar primary-school-age boys. We
did not discover gender di¡erences in any of the
measures observed during actual travel in both stu-
dies reported here. Moreover, the distances we re-
corded and lack of gender di¡erences are
consistent with the results of an earlier observa-
tional study with more age groups and larger sam-
ple sizes (Cornell & Heth, 1996).
There are several possible explanations for these
di¡erent results. Our observations were done in theneighborhoods, following behind children. It may be
that our participants selected unusually distant des-
tinations because they were con¢dent of the accom-
paniment by an adult. Or, it may be that young
children respond di¡erently as leaders in an out-
door activity than they do when they are inter-
viewed by adults. Finally, there are signi¢cant
cultural and cohort di¡erences between the chil-
dren who have participated in studies of home
range. The general ¢nding is that both boys and
girls are extending their activities into their neigh-
borhood so that by middle childhood all childrentravel well beyond the territory visible from their
home.
Our second study indicated that, by at least eight
years of age, children know the features of their
neighborhood that are distinctive. Their naming of
unique landmarks was impressive, but perhaps in
part the younger children were attracted by the sal-
ience of objects rather than to features that could
be linked to way ¢nding decisions. This interpreta-
tion is consistent with the ¢nding that 8 -year-old
Barker, R. G. (1979). The in£uence of frontier environmentson behavior. In J. Ste¡en (Ed.), The American Westpp. 61^93. Norman, OK: University of OklahomaPress.
Berry, J. W. (1966). Temne and Eskimo perceptual skills.International Journal of Psychology, 1, 207^229.
Coates, G. & Bussard, E. (1974). Patterns of children’s spa-tial behavior in a moderate-density housing develop-ment. In R. Moore (Ed.), Childhood City, Man-
Environment Interactions, pp. 131^141. Milwaukee, WI:EDRA.Cornell, E. H. & Heth, C. D. (1996). Distance traveled dur-
ing urban and suburban walks led by 3- to 12-year-olds. Response! The Journal of the National Association
for Search and Rescue, 15, 6^9.Cornell, E. H., Heth, C. D. & Broda, L. (1989). Children’s
way ¢nding: Response to instructions to use environ-mental landmarks. Developmental Psychology, 25,755^764.
Cornell, E. H., Heth, C. D. & Skoczylas, M. J. (1999). Thenature and use of route expectancies following inci-dental learning. Journal of Environmental Psychology,19, 209^229.
Dawson, J. (1967). Cultural and physiological in£uences
upon spatial-perceptual processes in West Africa.International Journal of Psychology, 2, 171^185.Golledge, R. G. (1995). Primitives of spatial knowledge. In
T. Nyerges, D. Mark, R. Laurini, & M. Egenhofer(Eds), Cognitive aspects of human-computer interaction
for geographic information systems. Series D: Behavioral and social sciences, pp. 29^44. Boston, MA: KluwerAcademic Publishers.
Golledge, R. G., Smith, T. R., Pellegrino, J. W., Doherty, S.& Marshall, S. P. (1985). A conceptual model and em-pirical analysis of children’s acquisition of spatialknowledge. Journal of Environmental Psychology, 5,125^152.
Harper, L. & Sanders, K. (1975). Preschool children’s use of space: Sex di¡erences in outdoor play. Developmental
Psychology, 11, 119^120.Hart, R. (1979). Children’s Experience of Place. New York,NY: Irvington.
Hart, R. A. & Moore, G. T. (1973). The development of spa-tial cognition: A review. In R. Downs & D. Stea (Eds)Image and Environment, pp. 246^288. Chicago, IL:Aldine.
Heth, C. D., Cornell, E. H. & Alberts D. M. (1997). Di¡er-ential use of landmarks by 8- and 12-year-old childrenduring route reversal navigation. Journal of Environ-mental Psychology, 17, 199^213.
Hill, K. (1998). The psychology of lost. In K. Hill (Ed.), LostPerson Behavior pp. 1^16. Ohaws, ON: NationalSecretariat for Search and Rescue.
Huttenlocher, J. & Newcombe, N. (1984). The child’s repre-
sentation of information about location. In C. Sophian(Ed.), The Origin of Cognitive Skills, pp. 81^111.Hillsdale, NJ: Erlbaum.
Kirasic, K. C. & Mathes, E. A. (1990). E¡ects of di¡erentmeans for conveying environmental information on el-derly adults’ spatial cognition and behavior. Environ-ment & Behavior, 22, 591^607.
Landy, D. (1965). Tropical Childhood. New York, NY:Harper & Row.
Matthews, M. H. (1987). Gender, home range, and environ-
mental cognition. Transactions of the Institute of British Geographers (New Series), 12, 43^56.Montello, D. R. (1998). A new framework for understand-
ing the acquisition of spatial knowledge in large-scaleenvironments. In R. Golledge & M. Egenhofer (Eds),Spatial and Temporal Reasoning in Geographic Informa-tion Systems. New York: Oxford University Press.
Moore, R. & Young, D. (1978). Childhood outdoors: Towarda social ecology of the landscape. In I. Altman & J.Wohlwill (Eds), Human Behavior and Environment:Vol. 3. Children and the Environment, pp. 83^130. NewYork: Plenum Press.
Munroe, R. L. & Munroe, R. H. (1971). E¡ect of environ-mental experience on spatial ability in East Africansociety. Journal of Social Psychology, 83, 15^22.
Payne, R. J. & Jones, D. R. W. (1977). Children’s urban land-scapes in Huntington Hills, Calgary. In P. Sued¢eldand J. Russel (Eds), The Behavioral Basis of Design.Book 2 pp. 306^312. Stroudsburg, PA: Dowden,Hutchinson & Ross.
Piaget, J. & Inhelder, B. (1967). The Child’s Conception of Space. New York: Norton.
Piaget, J., Inhelder, B. & Szeminska, A. (1960). The Child’sConception of Geometry. New York: Basic Books.
Resnick, L. B. (1976). Task analysis in instructional design:Some cases from mathematics. In D. Klahr (Ed.),Cognition and Instruction, pp. 51^80. Hillsdale, NJ:Erlbaum.
Rittle-Johnson, B. & Siegler, R. S. (1999). Learning tospell: Variability, choice, and change in children’s
strategy use. Child Development, 70, 332^348.Rogo¡, B., Gauvain, M. & Gardner, W. (1987). The develop-ment of children’s skills in adjusting plans to circum-stances. In S. Friedman, E. Scholnick & R. Cocking(Eds), Blueprints for Thinking: The Role of Planning inCognitive Development, pp. 303^320. New York, NY:Cambridge University Press.
Siegel, A. W. & White, S. H. (1975). The development of spa-tial representations of large-scale environments. InH. Reese (Ed.), Advances in Child Development and Be-havior, Vol. 10 , pp. 9^55. New York: Academic Press.
Siegler, R. S. (1996). Emerging Minds: The Process of Change in Children’s Thinking . New York: OxfordUniversity Press.
Siegler, R. S. & Crowley, K. (1991). The microgenetic meth-
od: A direct means for studying cognitive develop-ment. American Psychologist, 46, 606^620.
Siegler, R. S. & Jenkins, E. (1989). How Children DiscoverNew Strategies. Hillsdale, NJ: Erlbaum.
Van Vliet, W. (1983). Exploring the fourth environment: Anexamination of the home range of city and suburbanteenagers. Environment & Behavior, 15, 567^588.