ArticlePDF Available

Abstract and Figures

The organization of knowledge according to relations between concepts is crucially important for many cognitive processes, and its emergence during childhood is a key focus of cognitive development research. Prior evidence about the role of learning and experience in the development of knowledge organization primarily comes from studies investigating differences between preexisting, naturally occurring groups (e.g., children from rural vs. urban settings, children who own a pet vs. children who do not) and a handful of studies on the effects of researcher-developed educational interventions. However, we know little about whether knowledge organization can be relatively rapidly molded by shorter-term real-world learning experiences (e.g., on a timescale of days vs. years or months). The current study investigated whether naturalistic learning experiences can drive rapid measurable changes in knowledge organization in children by investigating the effects of a week-long zoo summer camp (compared with a control school-based camp) on the degree to which 4- to 9-year-old children's knowledge about animals was organized according to taxonomic relations. Although there were no differences in taxonomic organization between the zoo camp and the school-based camp at pretest, only children who participated in the zoo camp showed increases in taxonomic organization at posttest. Moreover, analyses of changes in taxonomic organization in zoo camp children suggested that these changes were primarily driven by improvements in the degree to which children differentiated between taxonomic categories. These findings provide novel evidence that naturalistic experiences can drive rapid changes in knowledge organization.
Content may be subject to copyright.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 1
Rapid, Experience-Related Changes in the Organization of Children’s Semantic Knowledge
Layla Unger1
Anna V. Fisher2
1Department of Psychology, Ohio State University, Columbus OH
2Department of Psychology, Carnegie Mellon University, Pittsburgh PA
Acknowledgements
This work was supported by a Graduate Training Grant awarded to Carnegie Mellon
University by the Department of Education, Institute of Education Sciences (R305B040063) and
by the James S. McDonnell Foundation 21st Century Science Initiative in Understanding Human
Cognition Scholar Award (220020401) to the second author. We thank the Pittsburgh Zoo, St.
Edmund’s Academy, and participating families for their participation in this research, as well as
Anna Vande Velde, Kristen Boyle, Camille Warner, Ashli-Ann Douglas, and Lindsay Gorby, for
their contributions to this research. Additionally, we thank Vladimir Sloutsky for input on an
earlier version of this paper.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 2
Abstract
The organization of knowledge according to relations between concepts is crucially important for
many cognitive processes, and its emergence during childhood is a key focus of cognitive
development research. Prior evidence about the role of learning and experience in the development
of knowledge organization primarily comes from studies investigating differences between pre-
existing, naturally-occurring groups (such as children from rural vs. urban settings, or children
who own a pet vs. children who don’t), and a handful of studies on the effects of researcher-
developed educational interventions. However, we know little about whether knowledge
organization can be relatively rapidly molded by shorter-term real-world learning experiences
(e.g., on a time-scale of days vs. years or months). The present study investigated whether
naturalistic learning experiences can drive rapid, measurable changes in knowledge organization
in children by investigating the effects of a week-long Zoo summer camp (compared to a control
school-based camp) on the degree to which 4- to 9-year-old children’s knowledge about animals
was organized according to taxonomic relations. Although there were no differences in taxonomic
organization between the Zoo and the school-based camp at pre-test, only children who
participated in the Zoo camp showed increases in taxonomic organization at post-test. Moreover,
analyses of changes in taxonomic organization in Zoo camp children suggested that these changes
were primarily driven by improvements in the degree to which children differentiated between
taxonomic categories. These findings provide novel evidence that naturalistic experiences can
drive rapid changes in knowledge organization.
Keywords: Cognitive Development; Semantic Knowledge; Semantic Development; Knowledge
Organization; Concept Learning
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 3
Our knowledge about the world is often characterized not as an encyclopedia, with
independent entries for information about separate topics, but rather as an interconnected
network organized according to relational links between concepts (Cree & Armstrong, 2012;
Jones, Willits, Dennis, & Jones, 2015; McClelland & Rogers, 2003). Organization is a key facet
of semantic knowledge, because it plays a pervasive role in many cognitive processes, including
knowledge retrieval (Bower, Clark, Lesgold, & Winzenz, 1969; Jimura, Hirose, Wada et al.,
2016), reasoning (DiSessa, 1982; Heit, 2000), and the acquisition of new knowledge (Bein,
Livneh, Reggev et al., 2015; Luiten, Ames, & Ackerson, 1980; Tse, Langston, Kakeyama et al.,
2007). Therefore, a key facet of understanding cognition is understanding how semantic
knowledge organization emerges with learning and development.
A large body of prior research has investigated the emergence of semantic knowledge
organization by evaluating differences in relational knowledge between pre-existing groups, such
as children who have grown up in different communities (e.g., urbal vs. rural settings) or
children of different ages, (Bauer & Mandler, 1989; Coley, 2012; Fisher, 2011; Gobbo & Chi,
1986; Inagaki, 1990; Lucariello, Kyratzis, & Nelson, 1992; Nelson & Nelson, 1990; Smiley &
Brown, 1979; Tversky, 1985; Unger, Fisher, Nugent, Ventura, & MacLellan, 2016; Walsh,
Richardson, & Faulkner, 1993; Waxman & Namy, 1997). Evidence of group differences from
these studies suggests that long-term exposure to different learning experiences acquired by
different groups of children is associated with different patterns of knowledge organization
development. Implied by these findings is the possibility that change in knowledge organization
is prompted by individual learning experiences, and that differences in the accumulation of these
learning experiences leads to differences in knowledge organization between groups of children
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 4
(Fisher, Godwin, & Matlen, 2015; Waxman & Medin, 2007). However, few studies have
investigated the role of learning experiences in knowledge organization directly by moving
beyond evaluating differences between pre-existing groups of individuals, and studying changes
in knowledge organization prompted by a specific learning experience over relatively brief time
scales (Alexander, Johnson, Leibham, & DeBauge, 2005; DeMarieDreblow, 1991). In addition
to being few in number, these studies have investigated the effects of interventions designed
specifically by researchers to prompt knowledge organization change, and therefore are not
directly informative about whether brief real-world experiences that children naturalistically
acquire can drive rapid knowledge organization change.
In what follows, we first review evidence about the emergence of knowledge
organization from research that has measured differences in knowledge organization between
naturally occurring groups of children, then review evidence from the handful of studies that
have investigated knowledge organization changes prompted by brief learning experiences. We
then highlight the limitations of the extant studies in illuminating whether brief, naturalistic
learning experiences can drive rapid changes in knowledge organization, and present a study
designed to address this gap.
Comparisons of Knowledge Organization in Pre-Existing Groups
A large body of prior research has investigated the emergence of knowledge
organization by comparing knowledge organization differences between pre-existing occurring
groups of children, including children of different ages (Bauer & Mandler, 1989; Fisher, 2011;
Lucariello et al., 1992; Nelson & Nelson, 1990; Smiley & Brown, 1979; Tversky, 1985; Unger et
al., 2016; Walsh et al., 1993; Waxman & Namy, 1997), and children with different long-term
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 5
histories of learning experiences (Coley, 2012; Gobbo & Chi, 1986; Inagaki, 1990). Evidence
from this research suggests that learning over time leads to changes in knowledge organization.
For example, many studies have investigated cross-sectional age differences in
knowledge for and ability to reason on the basis of a variety of relations, including perceptual
similarity, co-occurrence or interaction in the environment (e.g., schematic or thematic relations,
such as dog and bone), and membership in the same stable taxonomic category (e.g., the
membership of ostrich and robin in the category of “birds” regardless of their perceptual
differences and lack of environmental co-occurrence). According to the results of some studies,
the nature of the relations that organize knowledge changes with age (Blaye, Bernard-Peyron,
Paour, & Bonthoux, 2006; Lucariello et al., 1992; Unger et al., 2016), such that knowledge of
relations that are more easily observable in the environment (e.g., perceptual similarity and
environmental co-occurrence) is evident from an earlier age than knowledge of less readily
observable relations (e.g., membership in the same taxonomic category). According to the results
of other studies (Bauer & Mandler, 1989; Nguyen, 2007; Waxman & Namy, 1997), children
readily learn both easily observable relations and less observable taxonomic relations from an
early age (possibly with the aid of innate biases towards utilizing taxonomically-relevant
information in the environment, such as labels for taxonomic categories). Beyond whether
knowledge of less readily observable taxonomic relations emerges early or later in development,
some limits on taxonomic relation knowledge may commonly persist into adulthood.
Specifically, results from a number of studies (Trowbridge & Mintzes, 1985, 1988; Yen, Yao, &
Chiu, 2004) suggest that substantial minorities of students from elementary school through
college possess misconceptions about the taxonomic category membership of items that are more
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 6
perceptually similar to or co-occur more often with members of a different taxonomic category
(e.g., grouping whales with fish rather than mammals, or turtle with amphibians rather than
reptiles).
Although the results of these cross-sectional studies suggest that organization of
knowledge changes with age, these studies do not shed light on how such changes may come
about in the course of development and learning. This indirect nature of research on the role of
learning experiences in the development of knowledge organization also characterizes research
that has investigated whether the organization of knowledge in a given domain, such as
knowledge about animals, varies across children with different background experiences in that
domain. For example, comparisons between children who were experts versus novices in the
domain of dinosaurs (Gobbo & Chi, 1986) have yielded correlational evidence that expertise is
associated with the organization of knowledge in a domain according to more stable, meaningful
relations, such as the organization of knowledge about dinosaurs based on taxonomic category
membership rather than visual features. Similarly, comparisons of children raised in rural versus
urban settings (Coley, 2012) have provided complementary evidence that the former group of
children has more experiences with animals and shows a more sophisticated pattern of reasoning
about animals based on meaningful relations at a younger age (e.g., reasoning about insides
based on biological taxonomic relatedness, and reasoning about diseases based on ecological
interaction). Finally, similar differences in reasoning about animals have been documented
between children who have experience raising a pet versus those who do not (Inagaki, 1990).
Taken together, findings from these studies suggest that long-term exposure to different
learning experiences that is acquired by different groups of children is often associated with
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 7
different patterns of knowledge organization. Implied by these findings is the possibility that
change in knowledge organization is prompted by individual learning experiences, and that
differences in the accumulation of these learning experiences leads to differences in knowledge
organization between groups of children. Specifically, findings from the studies described above
may reflect the cumulative effect of changes in knowledge organization prompted by many
learning experiences in a domain, such as the effects on the organization of knowledge about
animals from interactions with a pet or farm animals (Waxman & Medin, 2007) over a time-scale
that may range from weeks to years. Yet, this research has not directly probed or manipulated
experiences that presumably drive changes in semantic organization. Therefore, the timescale of
changes in knowledge organization remains unknown it is unclear whether experiences must
accumulate over months or years in order to produce detectable changes in knowledge
organization, or whether changes in knowledge organization can be more rapidly prompted by
learning experiences over shorter timescales, such as weeks or days. As reviewed in the
following section, only a handful of prior studies have investigated these questions using
researcher-designed educational interventions.
Effects of Brief Learning Experiences
A small number of prior studies have investigated whether brief educational interventions
can drive changes in knowledge organization. For example, in DeMarie-Dreblow’s (1991)
studies, 8-11 year old children (Study 1) and undergraduate students (Study 2) completed a series
of knowledge and memory assessments before and/or after watching five short researcher-
developed educational videotapes that provided information about the characteristics of several
individual birds, and their membership in one of five bird categories (e.g., birds of prey). Of
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 8
particular interest here is the finding that the ability to group birds into categories was improved
following the educational intervention in both children and adults. Similarly, in Alexander and
colleagues’ (2005) study, kindergarten children completed a series of knowledge and reasoning
assessments over the course of a three-week researcher-developed curriculum, in which children
completed activities designed to teach them about several dinosaur species and their membership
in one of five dinosaur “families” (e.g., sauropods). As in the study conducted by DeMarie-
Dreblow, children in this study showed improvements in their ability to organize the dinosaur
species into their corresponding families over the course of participation in the curriculum.
Taken together, the results of these studies provide evidence that even relatively brief
learning experiences can drive changes in the organization of knowledge in a domain. However,
these results come from studies in which children were explicitly taught the same organization of
items into categories that was assessed, using interventions designed by researchers for this
purpose. Therefore, beyond expanding upon this small body of prior research, it is important to
investigate whether naturalistic learning experiences can prompt changes in knowledge
organization, and do so over relatively brief timescales. However, prior studies that investigated
the effects of naturalistic learning experiences such as a field trip to a museum or zoo (Bexell,
Jarrett, & Ping, 2013; Farmer, Knapp, & Benton, 2007; Gottfried, 1980; Prokop, Tuncer, &
Kvasničák, 2007; Randler, Baumgärtner, Eisele, & Kienzle, 2007; Randler, Kummer, &
Wilhelm, 2012; Stronck, 1983) largely focused on evaluating learning of specific facts (e.g.,
“Kangaroo rats have giant feet”, (p. 172, Gottfried, 1980) rather than changes in in the way that
knowledge is organized according to relations among different concepts.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 9
The distinction between learning of isolated facts on the one hand, and relations among
concepts on the other hand, is a critical one. Specifically, individual units of information can be
acquired piecemeal (e.g., through rote memorization) without changing overall knowledge
organization, or the many cognitive processes that are affected by knowledge organization (such
as retrieval of knowledge from memory and reasoning) (Ambrose, Bridges, DiPietro, Lovett, &
Norman, 2010; Bein et al., 2015; Jimura et al., 2016; Tse et al., 2007). However, the extent to
which brief, naturalistic learning experiences can promote changes in knowledge organization
has remained largely unexplored in the literature.
Present Study
The aim of the present study was to investigate whether measurable, rapid changes in the
organization of knowledge about animals transpires given the acquisition of naturalistic learning
experiences in this domain over a short timescale of days. To accomplish this aim, we compared
changes in the organization of knowledge about animals that transpired in 4- to 9-year old children
over the course of a week of summer camp that did or did not include learning experiences with
animals.
1
Specifically, in this quasi-experimental study, we measured the effects of a week-long,
zoo-based summer camp versus a control, school-affiliated summer camp on the degree to which
children’s knowledge about a set of animals was organized according to taxonomic relations.
Critically, as we describe below, taxonomic organization was equivalent in the two groups of
children at the beginning of the week of summer camp, such that rapid effects of learning
1
According to a 2015 survey conducted by the American Camp Association (https://www.acacamps.org/press-
room/aca-facts-trends), there are approximately 14,000 summer camps in the US, many of which (approximately
12%) provide nature/environment education programs like the one studied here.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 10
experiences on knowledge organization could manifest as a greater increase in taxonomic
organization at the end of the week for children who attended the zoo versus the control summer
camp.
Overall, the present study extends prior research in the following ways. First, in contrast to
prior studies, we directly investigated whether naturalistic learning experiences in a domain
prompt relatively rapid changes in knowledge organization. Second, our approach allowed us to
more thoroughly describe the learning experiences of interest than was possible in prior studies
involving pre-existing group differences. However, it should be noted that we did not aim to
evaluate whether educational programming at the zoo has unique pedagogical features that are
particularly effective for fostering learning about animals; therefore, similar to prior studies, we
did not specifically investigate which pedagogical components of the zoo camp learning
experiences may have played a causal role in prompting changes in knowledge organization.
Methods
Participants
Participants were 4- to 9-year-old children who were enrolled in the Zoo or the Control
Camp located in the same Northeastern US city. The initial sample included 34 Zoo Camp (20
females) and 32 Control Camp (17 females) children. Of this sample, data from six Zoo Camp
children were lost due to experimenter error (see the Results section below). Additionally, data
from one of the two outcome measures from one child in the Control Camp were also lost due to
an experimenter error (data from the other outcome measure were included in the analyses).
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 11
Overall, the final sample consisted of 28 children in the Zoo camp and 32 children in the Control
Camp.
Critically, although random assignment to a camp was not possible, children enrolled in
the two camps performed equivalently on measures of taxonomic knowledge at pre-test (see the
Results section below). The two camp samples were approximately matched for age (Zoo Camp:
Mage=6.89 years, SD=1.43; Control Camp: Mage=6.29 years, SD=1.21; Welch’s t(51.28)=1.73,
p=.098) and number of children in each of three age groups spanning the total age range (i.e., 4-5,
6-7, and 8-9; see below for further details) (p=.15, Fisher’s exact test). In the Zoo camp group,
78% (n=22) of participants were Caucasian; 7.14% (n=2) were of more than one race/ethnicity,
3.5% (n=1) were African American, 3.5% (n=1) were Asian/Pacific Islander, 3.5% (n=1) reported
“Other”, and 3.5% (n=1) did not respond. In the Control camp, 69% (n=22) of the participants
were Caucasian, 19% (n=6) were Asian or Pacific Islander, 6% (n=2) were Black, and 6% (n=2)
did not provide the race and ethnicity information.
Study Sites
Both the Zoo and Control summer camps were paid admission programs. Both camps
provided an option of full-day (i.e., 9 am 4 pm) attendance or half day (i.e., 9 am noon)
attendance. The sample included children who attended full-day as well as children who attended
half-day, although we do not have information about the scheduled attendance for each participant.
Anecdotally, younger children were more likely to attend half-day and older children were more
likely to attend full-day. The costs of both programs are comparable to each other. Specifically, at
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 12
present
2
, the cost of a week-long camp at the Zoo is $215 for half-day attendance and $315 for
full-day attendance (with a discount of $30 and $50, respectively, available to Zoo members). At
the Control camp, at present, the cost of a week-long camp is $150 for half-day attendance and
$300 for full-day attendance.
Both summer programs were located in the same city, within four miles of each other in
family-friendly neighborhoods, with similar rates of families with children (25.4% in the
neighborhood where the Control camp was located, and 22.0% in the neighborhood where the Zoo
camp was located). Overall, the neighborhood where the Control camp was located was less
racially diverse and more affluent than the neighborhood where the Zoo camp was located.
Specifically, according to data from the US Census Bureau (https://statisticalatlas.com/United-
States/Overview), the racial and ethnic make-up of the neighborhoods is as follows: 74.8% White,
5.6% Black, 4% Hispanic, and 12% Asian (for the neighborhood where the Control camp was
located); and 66.6% White, 22.6% Black, 2.4% Hispanic, and 5.9% Asian (for the neighborhood
where the Zoo camp was located). According to the same source, the median household income is
as follows: $87.5K in the neighborhood where the Control camp was located, and $56.2K in the
neighborhood where the Zoo camp was located.
Zoo Camp. Broadly, zoo camp consisted of lessons, interactions with preserved and live
animals, tours of the zoo, games, and craft sessions for children of all ages. Each day, activities
were designed to help children learn about a specific theme, such as “creatures of the night”. Each
year, the zoo camp organizers arrange a different set of themes into curricula for different age
2
The price information was retrieved online on April 10, 2018. We did not collect price information during the
years when we collected data (2015-2017).
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 13
groups that span the age range of children who attend the camp. The age groups spanning our
sample were 4-5, 6-7, and 8-9 years of age. The themes taught in the curricula for these age groups
over the two summers
3
during which we conducted this study are listed in Table 1. It is worth
noting that the majority of themes were not oriented towards explicitly teaching children biological
taxonomic relations, with the exception of the themes for children in the 8 to 9 age group in Year
2 and one instance of a “Reptiles” theme for children in the 4 to 5 age group in Year 2.
To provide a representative example of how these themes shaped the activities in which
children took part, the activities incorporated into the “Extreme Families” theme were as follows.
Children took part in two lessons: 1) A lesson about “Extreme Parents” that engage in behaviors
to protect their offspring (e.g., octopus, chicken, penguin), with contrasting examples of animals
3
One additional child in the 4-5 age group was tested in 2017. The themes for this age group in this year were the
same as those used in 2015.
Table 1
Curriculum themes at the Zoo camp for each age group during the study.
Year 1
Year 2
4-5 group
Domestic Animals
Super Senses
Tropical Treasures
Savanna Survival
Animal Locomotion
Reptiles
Aquatic Animal Diets
Savanna Animal Patterns
6-7 group
Animal Babies
How Animals Learn
Animal Families
Aquarium Animals
Rainforest
African Savanna
Ocean
Islands
8-9 group
Extreme Families
Extreme Senses
Extreme Animal Architects
Animal All-Stars
Mammals
Birds
Reptiles
Amphibians
Note. Because our pre- and post-test sessions took place on Monday and Friday
mornings, respectively, Friday themes are not included here.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 14
that do not (e.g., sharks), and 2) A lesson about the benefits to animals that live in family groups
(e.g., skink and elephant). During the lessons, counselors showed children live specimens of some
of the animals discussed (e.g., a live skink), and other animal-related specimens relevant to the
lessons (e.g., shark egg cases). Following each lesson, children visited a subset of the animals
described at the zoo, while a counsellor highlighted the relevance of the animal to the lesson (e.g.,
when visiting a penguin, a counsellor described how a penguin parent cares for its chick). Children
additionally completed crafts projects (e.g., made a shark tooth necklace) and played games (e.g.,
octopus tag) that were related to some of the animals about which they had learned, but were not
designed to convey further pedagogical information about the topic of the lessons. Finally, the
activities for the 4-5 and 6-7 age groups often included a counsellor reading aloud a picture book
relevant to the day’s theme.
Activities on all days of zoo camp followed this format of presenting lessons on a topic
with specimens, visiting animals in the zoo mentioned in the lessons while providing further
information relevant to the topic, and engaging in crafts projects/games that are thematically
related to animals mentioned in the lessons but that do not provide further pedagogical content.
However, the amount of information presented in lessons, and the ratio of time spent on lessons
versus animal visits and craft projects/games, increased over the 4-5, 6-7, and 7-8 age groups.
Control Camp. The Control Camp was a summer program offered through a private non-
denominational Pre-K through Grade 8 school (although summer camp participants did not have
to be students at the school). Importantly, the Control camp did not provide immersive experiences
with animals, but offered a variety of other recreational and learning opportunities. At the Control
camp, children engaged in outdoor play, water play, storytime, dance, crafts, games, and cooking.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 15
Additionally, children went on a field trip each week (e.g., to a library or a baseball game), but did
not visit the Zoo during the summer when testing took place.
Design
The study was a quasi-experiment in which children recruited from Zoo and Control camps
participated in both Pre- and Post-Test sessions at the beginning and end of a week of camp. To
ensure sufficient number of participants in the Zoo camp condition, data in this condition were
collected across multiple summers and collapsed across years for analysis.
Stimuli
The animal stimuli were selected from the zoo camp curricula, such that knowledge in a
given age group was assessed for animals about which Zoo Camp children in that age group
learned. Accordingly, we developed a separate stimulus set for each age group in each year (see
Supplemental Materials). Each set consisted of 15 animals, with an equal number of items in each
of three biological taxonomic categories: Mammals, birds, and reptiles. To represent the animal
stimuli, we used line drawings chosen to minimize perceptual similarity between animals in the
same taxonomic category (see examples in Figure 1A-B, and full set in Appendix A).
Materials and Procedures
Participants completed pre- and post-test sessions on Monday and Friday morning of the
same week that took place during a “before care” period prior to the start of camp activities. In all
years, these sessions included a Spatial Arrangement Method (SpAM) task (Goldstone, 1994), and
in Year 2, children additionally completed a Match-to-Sample task (Figure 1). These tasks have
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 16
complementary advantages: The match-to-sample task provides a straightforward assessment of
taxonomic reasoning that is well-established in developmental research (Fisher, 2011; Smiley &
Brown, 1979; Waxman & Namy, 1997), whereas SpAM yields a more graded measure of
taxonomic relations knowledge and the degree to which it changes with training. A recent
longitudinal study (Fisher, Godwin, Matlen, & Unger, 2014) provided evidence that the two
measures converge on the same underlying construct.
SpAM Task. Participants were seated at a game board consisting of a 10x10 grid, and were
told that they were going to play a game in which their job was to help a fictional character, Zibbo,
organize his favorite animals on the board (Figure 1-A). The experimenter then showed
participants a stimulus sheet that depicted all 15 animal stimuli selected for the child’s age group,
named each animal, and removed the sheet from view. Therefore, children were made aware of
the full set of animals they would judge prior to beginning of the task. Next, the experimenter told
the participant that they would organize the animals using cards that each depicted an animal on
the game board, such that “animals that are the same kind of animal go close together, and different
kinds of animals go farther apart”. The experimenter placed one of the 15 animal cards on a central
game board square, then named and presented each of the remaining cards for the participant one-
by-one. Cards were presented in a pre-determined, pseudo-random order in which no more than
two animals from the same taxonomic group appeared consecutively. Participants were allowed to
move cards that they had placed earlier in the task. At the conclusion of the task, the experimenter
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 17
photographed the board to record the locations of the cards. Due to experimenter error
4
, data from
this task were lost for six participants at the Zoo camp and one participant at the Control camp.
Match-to-Sample Task. The animal stimuli for a given age group were arranged into six
triads that each consisted of a Target, a Taxonomic Match that belonged to the same taxonomic
category as the Target and a Lure that belonged to a different category (Figure 1-B). Of the six
triads, two had mammal Targets, two had bird Targets, and two had reptile Targets (see
Supplemental Materials). Triads were designed to eliminate non-taxonomic cues to Taxonomic
Matches, such as visual similarity, shared habitat, or shared mode of locomotion. For example, a
triad might consist of a type of flightless bird such as penguin as the Target, a bird capable of flight
such as owl as the Taxonomic Match, and a non-bird such as polar bear as the Lure.
4
On the digital camera that was used, the button for taking photos, when gently rather than firmly pressed, would
not record a photo, but would refocus in a way that gave some experimenters the impression that a photo had been
taken.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 18
For each triad, the experimenter asked participants to choose whether the Taxonomic
Match or the Lure was “same kind of animal” as the Target. The experimenter pointed to and
labeled the animals while providing these instructions (e.g., “Which one is the same kind of animal
as the penguin, the owl or the polar bear?).
Figure 1. Schematic depiction of the SpAM task (stimulus sheet, cards, and game board) (Panel
A) and the match-to-sample task (Panel B).
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 19
Results
Of the Zoo Camp sample, children in all years (N=28 following exclusion of six children’s
data due to experimenter error) completed the SpAM task, whereas only children in Year 2 (N=16)
completed the Match-to-Sample task. All 32 children in the Control Camp sample were tested
during Year 2, and completed both tasks (although one participant’s SpAM data were lost due to
experimenter error).
Scoring
SpAM Task. The photographs taken following each arrangement trial were scored by
treating the 10x10 board as a coordinate plane, identifying the coordinates of each card, and
calculating the Euclidean distance between the coordinates. The range of possible distances was 1
(adjacent cards) to 12.73 (cards on diagonally opposite corners of the board). These distances were
taken as a measure of the degree to which participants judged a given pair of animals to be the
“same kind of animal”, where shorter distances indicate stronger judgments that the pair are of the
same kind.
Due to the fact that the set of items was composed of 5 animals from 3 taxonomic categories
(Mammal, Bird, and Reptile), the pair distances for a given participant at Pre- or Post-Test
consisted of 30 distances between pairs of animals that were both from the same taxonomic
category (i.e., 10 distances for each of the 3 taxonomic categories), and 75 distances between pairs
of animals from different taxonomic categories. To capture the degree to which participants placed
pairs of animals from the same taxonomic category closer together than pairs of animals from
different taxonomic categories, we used these distance data to calculate a Taxonomic Score for
each pair of animals from the same taxonomic category. To calculate a Taxonomic Score for a
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 20
given same taxonomic category pair (e.g., Duck-Owl) for a given participant in a given session,
we took the distance between this pair for this participant in this session, and subtracted it from
the mean distance between all pairs of animals from different categories (e.g., Alligator-Sheep)
measured for the same participant in the same session. Because each set of animals contained 30
same taxonomic category pairs, this calculation yielded 30 Taxonomic Scores for each participant
in each session. Taxonomic Scores greater than zero indicate that members of the same taxonomic
category were judged to be more closely related than members of different categories, such that
larger Taxonomic Scores reflected stronger judgments that taxonomically related versus unrelated
animals were of the “same kind”. Taxonomic Scores equal to zero indicate that members of the
same taxonomic category were judged as no more strongly related than members of different
categories, and Taxonomic Scores below zero indicate that participants judged members of
different categories as more strongly related than members of the same taxonomic category.
Match-to-Sample Task. For this task, we calculated an Accuracy measure for each
participant in which we scored their responses for each of the six triads as either 1 (chose the
Taxonomic Match) or 0 (chose the Distractor).
Pre-Test Performance
We first assessed whether participants in both camps performed comparably on the two
measures of taxonomic relations knowledge at Pre-Test. Note that the range of Taxonomic Scores
on the SpAM task at Pre-Test was -6.92 to 6.31 (chance=0), and the range of mean Accuracy
scores on the Match-to-Sample task was .17 to .83 (chance=.5). To conduct this comparison, we
calculated each participant’s mean SpAM Taxonomic Score and mean Match-to-Sample Accuracy
at Pre-Test, and assessed differences between the two camps using Welch’s t-tests. The results of
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 21
comparisons for both measures yielded no significant differences between camps (Taxonomic
Score: Mzoo=0.76, SDzoo=1.30, Mcontrol=0.60, SDcontrol=1.20, t(55.23)=.49, p=.62; Accuracy:
Mzoo=0.54, SDzoo=0.18, Mcontrol=0.58, SDcontrol=0.17, t(28.86)=.68, p=.50).
Correspondence between Outcome Measures
Prior evidence indicates that children’s responses on the Match-to-Sample task are related
to performance on the SpAM task, suggesting that both paradigms capture aspects of relational
knowledge (Fisher et al., 2014). To test whether this was the case in the present study, we
calculated the overall mean of each participant’s Taxonomic Score and Accuracy at each testing
session
5
, and assessed whether these two measures were correlated. We found an overall
significant positive correlation between mean Taxonomic Score and mean Accuracy (r=.33,
p=.002; Figure 2). Moreover, this significant correlation occurred for both Pre- and Post-Test
sessions for Zoo camp participants and for the Post-Test session for Control camp participants (all
ps < .01) (but was not significant at Pre-Test for Control camp participants; p=.12). These results
add to the small body of prior evidence that Match-to-Sample and SpAM paradigms measure
aspects of the same underlying body of relational knowledge.
5
Note that this averaging procedure results in a range of Taxonomic Scores and Accuracies that is narrower than the
full range of raw values.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 22
Pre- to Post-Test Change in Zoo versus Control Camp
These analyses measured the effects of Zoo Camp versus Control Camp on changes from
Pre- to Post-Test in taxonomically organized knowledge: I.e., the degree to which participants’
SpAM Taxonomic Scores indicated that they made “same kind” judgments based on taxonomic
relations, and the degree to which participants chose the Taxonomic Match on the Match-to-
Sample Task. For each of these outcome measures, our primary analyses consisted of mixed effects
regression models with fixed effects of Camp (Zoo versus Control) and Test Session (Pre-Test
versus Post-Test)
6
, and random effects of participant and item. The mixed effects model was linear
for the Taxonomic Score outcome measure, and logistic for the Accuracy outcome measure.
Therefore, we report F-statistics for the results of analyses of Taxonomic Score, and 2-statistics
6
Note: Age was not included in these analyses, because it may have been correlated with other factors, including the
amount/type of information conveyed in Zoo camp lessons, and whether children attended camp on a full- or half-
day basis. We therefore present exploratory analyses of Age in a separate, exploratory section.
Figure 2. Correspondence between each participant’s overall SpAM Taxonomic
Score and Match-to-Sample Accuracy.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 23
for the results of analyses of Accuracy. These analyses were conducted in the R environment (R
Development Core Team, 2008) using functions from the lme4 package (Bates, Maechler, Bolker,
& Walker, 2015) for mixed-model linear and logistic regression, and the car package (Fox &
Weisberg, 2011) for calculating test statistics and significance values for main effects and
interactions from mixed-model regression.
SpAM Taxonomic Score. The mixed effects model for Taxonomic Score revealed a main
effect of Test Session (F(1,3450)=10.34, p=.001), which was qualified by an interaction with
Camp (F(1,3450)=14.41, p=.0001). The interaction between Test Session and Camp indicates
differences between Pre- and Post-Test performance that vary between the Zoo and Control camps.
To investigate this interaction, separately for each camp, we generated mixed effects models with
a fixed effect of Test Session, and random effects of participant and item. These models revealed
a main effect of Test Session in the Zoo camp (F(1,1622)=25.39, p<.0001), such that Taxonomic
Scores were greater at Post-Test than Pre-Test, but no effect of Test Session in the Control Camp
(F(1,1799)=.08, p=.78) (see Figure 3 panel A).
Match-to-Sample Accuracy. The mixed effects model for Accuracy revealed no main
effects, but as in the model for Taxonomic Score, Test Session significantly interacted with Camp
(
2(1)=6.33, p=.012). As we did to investigate the Test Session by Camp interaction for
Taxonomic score, we generated mixed effects models for the Accuracy outcome measure
separately for each camp comprised of a fixed effect of Test Session, and random effects of
participant and item. These models also revealed a main effect of Test Session in the Zoo Camp
(
2(1)= 7.43, p=.006), such that Accuracies were greater at Post-Test than Pre-Test, but no effect
of Test Session in the Control Camp (
2(1)= 0.18, p=.67) (see Figure 3 panel B).
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 24
Taxonomic Improvement from Pre- to Post-Test in Zoo Camp: What Changed?
The preceding analyses indicate that the degree of taxonomic organization in the animal
domain increased during a week of learning about animals Zoo camp, but not during a week of
camp that did not involve learning experiences in this domain. To further investigate this effect,
we investigated two facets of the change in taxonomic organization: 1) Whether changes consisted
of treating animals from the same taxonomic category as more strongly related, treating animals
from different taxonomic categories as less strongly related, or both, and 2) Whether changes
varied across the three taxonomic categories investigated in this study (i.e., mammals, birds, and
reptiles).
Same versus Different Taxonomic Category. It is not possible to disentangle Pre- to
Post-Test increases in treating animals from the same taxonomic category as related from
A B
Figure 3. Pre- to Post-Test changes in SpAM Taxonomic Scores (panel A) and Match-to-
Sample Accuracies (panel B) for participants in the Control and Zoo camps.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 25
decreases in treating animals from different taxonomic categories as related based on performance
on the Match-to-Sample task, because this task yields only a single measure of whether or not
participants chose the taxonomic match in each triad. However, these possibilities can be
disentangled based on performance on the SpAM task. Specifically, it is possible to measure, for
each participant in each testing session, the distances between all pairs of animals from the same
taxonomic category, and all pairs of animals from different categories. Increases in treating
animals from the same taxonomic category as related would manifest as shorter distances between
same taxonomic category pairs at Post- versus Pre-Test, and decreases in treating animals from
different taxonomic categories as related would manifest as longer distances between different
taxonomic category pairs.
To investigate these possibilities, we took distances between all pairs for each Zoo camp
participant at each testing session as our outcome measure, and predicted these distances using a
linear mixed effects model in which Testing Session and Category Membership (Same Category
versus Different Category) were fixed effects, and participant and item were random effects. This
model revealed main effects of Testing Session (F(1,202)=8.26, p=.004)) and Category
Membership (F(1,202)=247.13 , p<.0001) that were qualified by an interaction with each other
(F(1,202)=14.41, p=.0001). To investigate the source of this interaction, we generated follow-on
models separately for each Category Membership condition that included Testing Session as a
fixed effect. From Pre- to Post-Test, these models revealed a significant increase in Different
Category distances (F(1,144)=20.04, p<.0001), and no significant decrease in Same Category
distances (F(1,58)=2.68, p=.11) (although there was a numeric trend in this direction; Figure 4).
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 26
These results therefore suggest that learning experiences at Zoo camp served to improve
differentiation between taxonomic categories.
Mammals, Birds, and Reptiles. In these analyses, we tested whether the improvement in
taxonomic organization from Pre- to Post-Test in Zoo camp participants varied across the three
taxonomic categories investigated in this study. It was possible to conduct these analyses using
data from both SpAM and Match-to-Sample tasks, because our primary outcome measures for
both tasks (i.e., Taxonomic Score and Accuracy, respectively) included an equal number of
observations in each taxonomic category.
We generated mixed effects models for each outcome measure for Zoo camp participants
in which Testing Session and Taxonomic Category were fixed effects, and participant and item
were random effects. As in our primary analyses of Pre- to Post-Test changes in the Zoo and
Figure 4. Pre- to Post-Test changes in SpAM task distances between pairs of
animals from the same taxonomic category, versus pairs of animals from
different taxonomic categories.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 27
Control camps, these models revealed a significant main effect of Testing Session for both
outcome measures (both ps < .01). This effect was not significantly qualified by an interaction
with Taxonomic Category (although the interaction was marginally significant for SpAM task
Taxonomic Scores; p=.09). This outcome suggests that the improvement from Pre- to Post-Test
occurred across all three taxonomic categories included in this study (for additional exploratory
analyses of patterns across individual items, see Appendix B).
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 28
The main effect of Taxonomic Category was not significant for the analysis of SpAM task
Taxonomic Scores (p=.14), but was significant in the analysis of Match-to-Sample Accuracies
(p=.003). As shown in Figure 5-A, the relative overall performances across the three taxonomic
categories varied across the two measures, which may be associated with the somewhat different
demands of these tasks (e.g., in the Match-to-Sample task, a target item is evaluated against only
A
B
Figure 5. A) Pre- to Post-Test changes for Zoo camp participants in SpAM task Taxonomic
Score (left) and Match-to-Sample task Accuracy (right), plotted separately for each taxonomic
category. B) Same Pre- to Post-Test scores plotted for Control camp participants.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 29
one taxonomic match and one distractor, whereas in the SpAM task, each item is evaluated against
all other items). In spite of such variability, overall, these analyses indicate that Zoo camp
experience influenced taxonomic organization across the three taxonomic categories investigated
in this study. In contrast, no such improvement was observed in Control camp participants on
either measure of semantic organization for any of the taxonomic categories (ps > .49 for the effect
of Testing Session for all categories; see Figure 5-B).
Variation in Taxonomic Improvement with Age and Prior Knowledge: Exploratory Analyses
It is possible that the degree to which the immersive experiences with animals acquired at
zoo camp prompted changes in taxonomic organization varied with children’s ages or extent of
prior knowledge in the animal domain (e.g., older or more knowledgeable children learned more
readily). A rigorous analysis of relationships between taxonomic organization improvement and
age is not possible in the present study, because age varied with the content of the lessons
children experienced at camp, and may have varied with the likelihood that children attended a
full- versus half-day of camp. Similarly, although we can take Pre-Test performance as a
measure of prior knowledge, prior knowledge may have varied with age (indeed, Age was
significantly correlated with Pre-Test SpAM performance (r=.24, p<.001), though not with Pre-
Test Match-to-Sample performance (p=.53)). Therefore, we instead present exploratory analyses
of the relationship between improvement in taxonomic organization in zoo camp children and
age and prior knowledge, with the caveat that these factors may have been confounded with the
amount of time children spent at camp and the specific content that children experienced at
camp.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 30
To conduct these exploratory analyses, we calculated Learning Gain scores for both the
Match-to-Sample and SpAM tasks by subtracting each zoo camp child’s overall Pre-Test score
from their overall Post-Test score. We then measured the correlation between children’s
Learning Gain scores and age: Age was marginally correlated with Learning Gain scores on both
Match-to-Sample task (r = .45, p =.06) and SpAM task (r = .35, p = .06). In contrast, prior
knowledge (as measured from Pre-Test performance) was not correlated with Learning Gain
scores on either measure (both ps > .14). Therefore, the results of these exploratory analyses may
be taken as suggestive that the degree of improvement in taxonomic organization prompted by
zoo camp experiences may have been greater for older children than younger children. However,
caution is warranted in interpreting these findings because the correlations between Learning
Gain scores and age were only marginally significant in this sample, and may have been
confounded with other factors discussed above.
Discussion
The purpose of the present study was to investigate whether the organization of knowledge
in the animal domain is subject to rapid, learning-driven changes. To accomplish this goal, we
assessed changes in the taxonomic organization of children’s knowledge about sets of animals that
transpired between the beginning and end of the end of a four-day period that either did or did not
include learning experiences in the animal domain. Learning experiences in the animal domain
consisted of attendance at a zoo summer camp (versus a non-zoo control camp).
Across two complementary measures, children who attended zoo camp showed significant
increases in taxonomic organization from the beginning to the end of the camp week. In contrast,
there was no detectable change in taxonomic organization over the same time period for children
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 31
who attended a summer camp that did not include learning experiences in the animal domain.
Critically, this difference between children who attended the two camps transpired in spite of no
significant differences in performance on the taxonomic organization measures at the beginning
of the camp week. Additionally, the results of follow-up analyses on the change in taxonomic
organization in children who attended zoo camp suggested that these changes were primarily
associated with improvements in the degree to which children treated members of different
taxonomic categories as different, rather than the degree to which they treated members of the
same taxonomic category as related. These results indicate that naturalistic learning experiences
in a domain can prompt rapid changes in semantic knowledge organization. Additionally, the
results of exploratory analyses provided some suggestion that the degree to which these experience
prompt changes may increase with age (although these effects reached only marginal significance
and may have been confounded with other factors).
These findings are consistent with evidence from prior studies, which have shown that
knowledge organization differs across pre-existing groups of children with different long-term
learning experience histories, such as children who have grown up in different communities
(Coley, 2012; Gobbo & Chi, 1986; Inagaki, 1990). Moreover, the present findings expand upon
these prior studies by providing evidence that, given two groups of children from the same (urban)
setting and with comparable initial performance on measures of knowledge organization, the
accumulation of learning experiences over a brief period of time in one group of children prompted
rapid changes in knowledge organization.
The results of the present study also expand upon prior research that has investigated the
influence of brief learning experiences. First, these results both add to evidence from the small
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 32
number of prior studies that have shown knowledge organization change prompted by a brief
intervention, and expand upon this evidence by demonstrating knowledge organization change
following a real-world, naturalistic learning experience. Second, these results also expand upon
evidence about the effects of similar real-world, naturalistic learning experiences, such as a field
trip to a museum or zoo (Bexell et al., 2013; Farmer et al., 2007; Gottfried, 1980; Prokop et al.,
2007; Randler et al., 2007; Randler et al., 2012; Stronck, 1983). Specifically, this body of research
has provided evidence that such experiences can prompt the acquisition of knowledge for
individual facts or principles. The present study further demonstrates that these learning
experiences can go beyond fostering knowledge of isolated items of information by fostering
changes in the way that knowledge in a domain is organized into an interconnected network of
concepts.
Limitations
Although the absence of pre-test differences between the children in two camps on the
assessments of taxonomic organization supports the conclusion that learning experiences in the
animal domain acquired at Zoo camp prompted the observed improvements in taxonomic
organization, a direct investigation of the causal influence of learning experiences on knowledge
organization was not possible in the present study due to its quasi-experimental design. Such a
direct investigation could be accomplished using a design in which participants are randomly
assigned to conditions in which they either do or do not acquire learning experiences in a domain,
or acquire learning experiences in one of two different domains.
The results of the present study also only provide a suggestion that the magnitude of the
effects of learning experiences on knowledge organization may increase with age, but do not
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 33
rigorously test this possibility. A rigorous investigation of this influence was not possible,
because age in the present study may have been confounded with both whether children attended
a full- versus a half-day of camp, and for Zoo camp participants, with the amount/type of
information conveyed during lessons. Developmental changes in the readiness with which
knowledge organization is shaped by experience would represent an important avenue for future
research. However, any such research would need to carefully disentangle effects of age from the
effects of other factors that may vary with age, such as prior knowledge in a domain.
Open Questions
The present demonstration of rapid, experience-related change in knowledge organization
highlights multiple key questions for future research. First, does the readiness with which
knowledge organization changes given learning experiences vary over development? With age
comes many changes that may improve a child’s receptiveness to learning experiences. For
example, development is accompanied by improvements in executive functions such as working
memory and selective attention capacity (e.g., Fisher, Thiessen, Godwin, Kloos, & Dickerson,
2013; Fry & Hale, 2000), which may help children take in the information presented during a
learning experience. Moreover, improvements in prior knowledge in a domain that may come with
age (Unger et al., 2016) may improve the integration of new information into semantic memory
(Kaefer, Neuman, & Pinkham, 2015; Shing & Brod, 2016). As described earlier in this section,
the results of the present study provided some suggestion that the effects of learning experiences
at Zoo camp on knowledge organization increased with age. However, a rigorous examination of
this question remains open for future research.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 34
Second, do changes to knowledge organization prompted by a given learning experience
generalize to other concepts in the same domain? In the present study, we investigated changes in
the organization of knowledge about animals that were part of the learning curricula that children
at zoo camp experienced. It is possible that learning experiences at Zoo camp only fostered changes
in relations between the animal concepts children encountered. Alternately, learning experiences
at Zoo camp might have fostered changes that could generalize to other animal concepts, such as
highlighting the salience of taxonomically-relevant features.
Third, the present results, like the results of prior research that has compared knowledge
organization across groups of children with different long-term experience histories, do not
illuminate what specific input and mechanisms drive knowledge organization change. Some
possible candidates come from research on pedagogical approaches that help learners group
together or differentiate between concepts, such as structural alignment, in which instruction
highlights the underlying relations between features that different concepts do or do not share
(Jee, Uttal, Gentner et al., 2013; Kurtz, Miao, & Gentner, 2001). Aside from explicit instruction,
knowledge organization may also be shaped by sensitivities to regularities in the environment.
Evidence for this possibility primarily comes from computational modeling research involving
the creation of simulated semantic networks that form representations of concepts and their
interrelations from input available in the environment, such as the features of entities (Kemp &
Tenenbaum, 2008; McClelland & Rogers, 2003), or the visual or linguistic contexts in which
entities or their labels occur (Frermann & Lapata, 2015; Huebner & Willits, 2018; Sadeghi,
McClelland, & Hoffman, 2015). The success with which such models have captured human
semantic behavior (Cree & Armstrong, 2012; Jones et al., 2015) suggests that they represent
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 35
viable hypotheses about the sources of input and mechanisms that contribute to semantic
organization. Empirically testing these hypotheses by investigating whether and how the
organization of children’s semantic knowledge changes when exposed to these sources of input
would be a fruitful path for future research.
Conclusion
This study reports the first investigation of changes in semantic knowledge organization
as a result of a relatively brief naturalistic learning experience (i.e., a week-long summer camp at
a local Zoo). We found that the taxonomic organization of children’s knowledge about animals
changed measurably over the course of a short, four-day period when it contained learning
experiences in that domain acquired at a Zoo summer camp. In contrast, children who performed
similarly on assessments of taxonomic organization at the beginning of this period, but who
attended a Control camp and did not acquire learning experiences in the animal domain, showed
no evidence of increased taxonomic organization by the end of the period. These results
therefore provide novel evidence of rapid, learning-related changes in knowledge organization.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 36
References
https://statisticalatlas.com/United-States/Overview
ACA Facts and Trends. Retrieved from https://www.acacamps.org/press-room/aca-facts-trends
Alexander, J. M., Johnson, K. E., Leibham, M. E., & DeBauge, C. (2005). Constructing domain-
specific knowledge in kindergarten: Relations among knowledge, intelligence, and
strategic performance. Learning and Individual Differences, 15, 35-52.
Ambrose, S. A., Bridges, M. W., DiPietro, M., Lovett, M. C., & Norman, M. K. (2010). How
learning works: Seven research-based principles for smart teaching: John Wiley & Sons.
Bates, D., Maechler, M., Bolker, B., & Walker, S. (2015). Fitting Linear Mixed-Effects Models
Using lme4. Journal of Statistical Software, 67, 1-48.
Bauer, P. J., & Mandler, J. M. (1989). Taxonomies and triads: Conceptual organization in one-to
two-year-olds. Cognitive Psychology, 21, 156-184.
Bein, O., Livneh, N., Reggev, N., Gilead, M., Goshen-Gottstein, Y., & Maril, A. (2015).
Delineating the effect of semantic congruency on episodic memory: The role of
integration and relatedness. PloS ONE, 10, 1-24.
Bexell, S. M., Jarrett, O. S., & Ping, X. (2013). The effects of a summer camp program in China
on children's knowledge, attitudes, and behaviors toward animals: A model for
conservation education. Visitor Studies, 16, 59-81.
Blaye, A., Bernard-Peyron, V., Paour, J.-L., & Bonthoux, F. (2006). Categorical flexibility in
children: Distinguishing response flexibility from conceptual flexibility. European
Journal of Developmental Psychology, 3, 163-188.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 37
Bower, G. H., Clark, M. C., Lesgold, A. M., & Winzenz, D. (1969). Hierarchical retrieval
schemes in recall of categorized word lists. Journal of Verbal Learning and Verbal
Behavior, 8, 323-343.
Coley, J. D. (2012). Where the wild things are: Informal experience and ecological reasoning.
Child Development, 83, 992-1006.
Cree, G. S., & Armstrong, B. C. (2012). Computational models of semantic memory The
Cambridge Handbook of Psycholinguistics (pp. 259-282). Cambridge: Cambridge
University Press.
DeMarieDreblow, D. (1991). Relation between knowledge and memory: A reminder that
correlation does not imply causality. Child Development, 62, 484-498.
DiSessa, A. A. (1982). Unlearning Aristotelian physics: A study of knowledge-based learning.
Cognitive Science, 6, 37-75.
Farmer, J., Knapp, D., & Benton, G. M. (2007). An elementary school environmental education
field trip: Long-term effects on ecological and environmental knowledge and attitude
development. The Journal of Environmental Education, 38, 33-42.
Fisher, A., Thiessen, E., Godwin, K., Kloos, H., & Dickerson, J. (2013). Assessing selective
sustained attention in 3-to 5-year-old children: Evidence from a new paradigm. Journal
of Experimental Child Psychology, 114, 275-294.
Fisher, A. V. (2011). Processing of perceptual information is more robust than processing of
conceptual information in preschool-age children: Evidence from costs of switching.
Cognition, 119, 253-264.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 38
Fisher, A. V., Godwin, K. E., & Matlen, B. J. (2015). Development of inductive generalization
with familiar categories. Psychonomic Bulletin & Review, 22, 1149-1173.
Fisher, A. V., Godwin, K. E., Matlen, B. J., & Unger, L. (2014). Development of Category
Based Induction and Semantic Knowledge. Child Development, 86, 4862.
Fox, J., & Weisberg, S. (2011). An R Companion to Applied Regression (Second Edition ed.).
Thousand Oaks, CA: Sage.
Frermann, L., & Lapata, M. (2015). Incremental Bayesian Category Learning From Natural
Language. Cognitive Science, 40, 13331381.
Fry, A. F., & Hale, S. (2000). Relationships among processing speed, working memory, and
fluid intelligence in children. Biological psychology, 54, 1-34.
Gobbo, C., & Chi, M. (1986). How knowledge is structured and used by expert and novice
children. Cognitive Development, 1, 221-237.
Goldstone, R. (1994). An efficient method for obtaining similarity data. Behavior Research
Methods, Instruments, & Computers, 26, 381-386.
Gottfried, J. (1980). Do children learn on school field trips? Curator: The Museum Journal, 23,
165-174.
Heit, E. (2000). Properties of inductive reasoning. Psychonomic Bulletin & Review, 7, 569-592.
Huebner, P. A., & Willits, J. A. (2018). Structured semantic knowledge can emerge
automatically from predicting word sequences in child-directed speech. Frontiers in
Psychology, 9.
Inagaki, K. (1990). The effects of raising animals on children's biological knowledge. British
Journal of Developmental Psychology, 8, 119-129.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 39
Jee, B. D., Uttal, D. H., Gentner, D., Manduca, C., Shipley, T. F., & Sageman, B. (2013).
Finding faults: analogical comparison supports spatial concept learning in geoscience.
Cognitive Processing, 14, 175-187.
Jimura, K., Hirose, S., Wada, H., Yoshizawa, Y., Imai, Y., Akahane, M., . . . Konishi, S. (2016).
Relatedness-dependent rapid development of brain activity in anterior temporal cortex
during pair-association retrieval. Neuroscience Letters, 627, 24-29.
Jones, M. N., Willits, J., Dennis, S., & Jones, M. (2015). Models of semantic memory. In J.
Busemeyer & J. Townsend (Eds.), Oxford Handbook of Mathematical and
Computational Psychology (pp. 232-254). New York, NY: Oxford University Press.
Kaefer, T., Neuman, S. B., & Pinkham, A. M. (2015). Pre-existing background knowledge
influences socioeconomic differences in preschoolers’ word learning and comprehension.
Reading Psychology, 36, 203-231.
Kemp, C., & Tenenbaum, J. B. (2008). The discovery of structural form. Proceedings of the
National Academy of Sciences, 105, 10687-10692.
Kurtz, K. J., Miao, C.-H., & Gentner, D. (2001). Learning by analogical bootstrapping. The
Journal of the Learning Sciences, 10, 417-446.
Lucariello, J., Kyratzis, A., & Nelson, K. (1992). Taxonomic knowledge: What kind and when?
Child Development, 63, 978-998.
Luiten, J., Ames, W., & Ackerson, G. (1980). A meta-analysis of the effects of advance
organizers on learning and retention. American Educational Research Journal, 17, 211-
218.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 40
McClelland, J. L., & Rogers, T. T. (2003). The parallel distributed processing approach to
semantic cognition. Nature Reviews Neuroscience, 4, 310-322.
Nelson, K., & Nelson, A. P. (1990). Category production in response to script and category cues
by kindergarten and second-grade children. Journal of Applied Developmental
Psychology, 11, 431-446.
Nguyen, S. P. (2007). Cross-classification and category representation in children's concepts.
Developmental Psychology, 43, 719-731.
Prokop, P., Tuncer, G., & Kvasničák, R. (2007). Short-term effects of field programme on
students’ knowledge and attitude toward biology: a Slovak experience. Journal of
Science Education and Technology, 16, 247-255.
Randler, C., Baumgärtner, S., Eisele, H., & Kienzle, W. (2007). Learning at workstations in the
zoo: A controlled evaluation of cognitive and affective outcomes. Visitor Studies, 10,
205-216.
Randler, C., Kummer, B., & Wilhelm, C. (2012). Adolescent learning in the zoo: Embedding a
non-formal learning environment to teach formal aspects of vertebrate biology. Journal
of Science Education and Technology, 21, 384-391.
Sadeghi, Z., McClelland, J. L., & Hoffman, P. (2015). You shall know an object by the company
it keeps: An investigation of semantic representations derived from object co-occurrence
in visual scenes. Neuropsychologia, 76, 52-61.
Shing, Y. L., & Brod, G. (2016). Effects of prior knowledge on memory: Implications for
education. Mind, Brain, and Education, 10, 153-161.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 41
Smiley, S. S., & Brown, A. L. (1979). Conceptual preference for thematic or taxonomic
relations: A nonmonotonic age trend from preschool to old age. Journal of Experimental
Child Psychology, 28, 249-257.
Stronck, D. R. (1983). The comparative effects of different museum tours on children's attitudes
and learning. Journal of Research in Science Teaching, 20, 283-290.
R Development Core Team. (2008). R: A language and environment for statistical computing.
Vienna, Austria: R Foundation for Statistical Computing.
Trowbridge, J. E., & Mintzes, J. J. (1985). Students' alternative conceptions of animals and
animal classification. School Science and Mathematics, 85, 304-316.
Trowbridge, J. E., & Mintzes, J. J. (1988). Alternative conceptions in animal classification: A
crossage study. Journal of Research in Science Teaching, 25, 547-571.
Tse, D., Langston, R. F., Kakeyama, M., Bethus, I., Spooner, P. A., Wood, E. R., . . . Morris, R.
G. (2007). Schemas and memory consolidation. Science, 316, 76-82.
Tversky, B. (1985). Development of taxonomic organization of named and pictured categories.
Developmental Psychology, 21, 1111-1119.
Unger, L., Fisher, A. V., Nugent, R., Ventura, S. L., & MacLellan, C. J. (2016). Developmental
changes in semantic knowledge organization. Journal of Experimental Child Psychology,
146, 202-222.
Walsh, M., Richardson, K., & Faulkner, D. (1993). Perceptual, thematic and taxonomic relations
in children’s mental representations: Responses to triads. European Journal of
Psychology of Education, 8, 85-102.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 42
Waxman, S., & Medin, D. (2007). Experience and cultural models matter: Placing firm limits on
childhood anthropocentrism. Human Development, 50, 23-30.
Waxman, S. R., & Namy, L. L. (1997). Challenging the notion of a thematic preference in young
children. Developmental Psychology, 33, 555-567.
Yen, C.-F., Yao, T.-W., & Chiu, Y.-C. (2004). Alternative conceptions in animal classification
focusing on amphibians and reptiles: A cross-age study. International Journal of Science
and Mathematics Education, 2, 159-174.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 43
Appendix A
Items used in SpAM and Match-to-Sample Tasks
Table A1
Animal Stimuli used for 4-5 year olds in each year, selected from the animals that appeared in Zoo camp curricula for this age group.
Year 1
Year 2
Mammals
Birds
Reptiles
Mammals
Birds
Reptiles
Giraffe
Gorilla
Elephant
Sheep
Sloth
Duck
Ostrich
Owl
Toucan
Vulture
Alligator
Gecko
Milk
Snake
Python
Tortoise
Bat
Beaver
Hedgehog
Elephant
Rabbit
Duck
Ostrich
Owl
Penguin
Toucan
Alligator
Pancake
Tortoise
Python
Sea
Turtle
Skink
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 44
Table A2
Animal Stimuli used for 6-7 year olds in each year, selected from the animals that appeared in Zoo camp curricula for this age group.
Year 1
Year 2
Mammals
Birds
Reptiles
Mammals
Birds
Reptiles
Bat
Kangaroo
Orangutan
Sea Otter
Zebra
Duck
Kestrel
Macaw
Owl
Penguin
Alligator
Gecko
Komodo
Dragon
Python
Sea
Turtle
Gorilla
Fennec
Fox
Sea Otter
Sloth
Walrus
Macaw
Ostrich
Penguin
Toucan
Vulture
Boa
Pancake
Tortoise
Python
Sea
Turtle
Skink
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 45
Table A3
Animal Stimuli used for 8-9 year olds in each year, selected from the animals that appeared in Zoo camp curricula for this age group.
Year 1
Year 2
Mammals
Birds
Reptiles
Mammals
Birds
Reptiles
Bat
Beaver
Fennec
Fox
Elephant
Polar
Bear
Chicken
Kestrel
Ostrich
Penguin
Swift
Alligator
Horned
Lizard
Python
Sea
Turtle
Skink
Bat
Fennec
Fox
Polar
Bear
Sea Lion
Sea Otter
Kiwi
Macaw
Ostrich
Owl
Penguin
Alligator
Glass
Lizard
Komodo
Dragon
Pancake
Tortoise
Skink
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 46
Table A4
Match-to-Sample trials used in each age group in Year 2.
Target
Match
Distractor
4-5 Years
Hedgehog
Sea Turtle
Rabbit
Owl
Alligator
Duck
Bat
Python
Elephant
Ostrich
Sea Turtle
Owl
Penguin
Beaver
Pancake Tortoise
Hedgehog
Toucan
Beaver
6-7 Years
Gorilla
Sea Turtle
Ostrich
Pancake Tortoise
Penguin
Sea Otter
Walrus
Boa
Toucan
Python
Macaw
Sloth
Vulture
Macaw
Fennec Fox
Sloth
Fennec Fox
Skink
8-9 Years
Bat
Alligator
Sea Lion
Penguin
Pancake Tortoise
Kiwi
Otter
Glass Lizard
Bat
Owl
Glass Lizard
Macaw
Macaw
Ostrich
Komodo Dragon
Polar Bear
Sea Lion
Fennec Fox
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 47
Appendix B
Variation in Taxonomic Improvement Across Items: Exploratory Analyses
An investigation of patterns of Taxonomic improvement in Zoo Camp children’s
knowledge of specific items is of interest for determining whether the overall improvement in
taxonomic organization observed in these children was indicative of a general shift towards
taxonomic organization, or the clarification of the taxonomic category membership of only a
handful of items. However, item effects are difficult to investigate in this study because items
varied across age groups and years (such that when an item was only used with children of a
given age group, its effects may be confounded with age). Therefore, we present here an
exploratory, qualitative analysis of patterns of taxonomic improvement across items.
To explore item effects in SpAM task data, for each item, we calculated the average
improvement in Taxonomic Score from Pre- to Post-test across all pairs that included the item.
Similarly, to explore item effects in Match-to-Sample task data, we calculated the average
improvement in accuracy from Pre- to Post-test for each Target item. Graphs of these data can be
found below. Inspection of these data reveals that, according to SpAM Taxonomic Scores,
improvement occurred for the majority of items. Inspection of Match-to-Sample data reveals
improved accuracies for only approximately half of the Target items (though it should be noted
that this includes items from all three taxonomic categories used in the study). It is likely that
inspection of the SpAM data reveals a more complete picture of improvements, because it is
designed to provide a more sensitive, graded measure of taxonomic organization (in contrast
with the Match-to-Sample task, in which participants are only presented with a binary decision
on which they can be scored as correct or incorrect).
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 48
Figure C1. Average changes in SpAM Taxonomic Score (top) and Match-to-Sample Accuracy
for each item, color-coded according to indicate consistency across tasks.
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 49
Several caveats should be taken with this analysis about why individual items may
sometimes show different patterns of change in the Match-to-Sample versus SpAM tasks (as
identified in the graphs above). First, whereas the SpAM task used the full set of items for a
given age group in a given year, only a subset of items appeared as Targets in the Match-to-
Sample task. This means that, in some cases, data for an item in the Match-to-Sample task come
from only a subset of the participants who judged the same item in the SpAM task. For example,
as shown in Appendix A, “Gorilla” was included in the full set of items judged in the SpAM task
for both 4-5 year-old children in Year 1 and 6-7 year-old children in Year 2, but was only a
Target in the Match-to-Sample task in the latter group. This means that the SpAM data for
Gorilla include children from an additional, younger age group in comparison to the Match-to-
Sample task data, and therefore may contribute to the fact that Gorilla appeared to improve more
in the Match-to-Sample task than the SpAM task.
Second, most of the items that showed inconsistent patterns across the tasks (7 our of 8
items) are the ones that had relatively smaller improvements in the SpAM task, but no
improvement in the Match-to-Sample task. As noted in above, this could be due in part to the
fact that the Match-to-Sample task provides a binary, all-or-nothing measure, whereas the SpAM
task is more graded and nuanced. Therefore, the SpAM task might be sensitive to smaller
improvements in taxonomic knowledge that do not manifest in the Match-to-Sample task.
Finally, the Match-to-Sample and SpAM tasks place somewhat different contextual
demands. For example, when Gorilla appeared in both tasks, it was judged only against Walrus
(category match) and Vulture (distractor) in the Match-to-Sample task, whereas it was judged
against all other mammals and non-mammals in the SpAM task. Learning in Zoo Camp might
have affected knowledge about how an item used as a Target in the Match-to-Sample task is
Running Head: RAPID KNOWLEDGE ORGANIZATION CHANGE 50
related to its category match vs. its distractor relatively more or less than it affected knowledge
about the overall relatedness between that item and a variety of same- versus different-category
items.
These caveats notwithstanding, the results of this exploratory qualitative item analysis
indicate that the overall improvement in taxonomic organization observed in this study was not
driven by a few items, but instead occurred for a substantial subset of items that children were
tested on.
... In addition, two recent studies using the spatial arrangement method have reported pretest-toposttest changes in the arrangements produced by the same children, with those changes being specific to children's learning experiences in a week-long summer camp at a zoo (Unger & Fisher, 2019) and a botanical garden (Vales, States, & Fisher, 2020). Specifically, children's arrangements reflected increased differentiation for the domain they experienced during the summer camp-with no such differences being observed in a control group (Unger & Fisher, 2019) or in the same children for a nonexperienced domain (Vales et al., 2020). ...
... In addition, two recent studies using the spatial arrangement method have reported pretest-toposttest changes in the arrangements produced by the same children, with those changes being specific to children's learning experiences in a week-long summer camp at a zoo (Unger & Fisher, 2019) and a botanical garden (Vales, States, & Fisher, 2020). Specifically, children's arrangements reflected increased differentiation for the domain they experienced during the summer camp-with no such differences being observed in a control group (Unger & Fisher, 2019) or in the same children for a nonexperienced domain (Vales et al., 2020). ...
... The current results are in line with a number of recent findings documenting the role of repeated experiences with items of a domain in increasing both across-domain differentiation (Vales et al., 2020) and within-domain differentiation (Badger & Shapiro, 2019;Unger & Fisher, 2019;Vales et al., 2020) and add to this literature by examining this hypothesis in a cross-sectional sample. These results are also broadly consistent with recent work suggesting that semantic networks become increasingly connected over the life span (Dubossarsky, De Deyne, & Hills, 2017), suggesting that experience-based learning likely continues to shape semantic structure over the life span. ...
Article
Full-text available
Organized semantic representations encoding across- and within-domain distinctions are a hallmark of mature cognition, and understanding how they change with experience and learning is a key endeavor in developmental science. Existing computational modeling studies provide a mechanistic framework for understanding how structured semantic representations emerge as a result of development and learning. However, their predictions remain largely untested in young children, with the existing evidence providing only indirect tests of these predictions. Across two experiments, we provide the first direct examination of a key prediction derived from these computational models—that early in development, broad across-domain distinctions should generally be more strongly represented relative to finer-grained within-domain distinctions. The results support this hypothesis, being consistent with the exploitation of patterns of covariation among entities as a mechanism supporting the acquisition of structured semantic representations.
... A smaller body of research has also shown effects of brief learning experiences on children's semantic networks. For example, participating in enrichment learning activities at a zoo-whether a week-long program (Unger & Fisher, 2019) or a single session (Badger & Shapiro, 2019)-lead to pre-to posttest changes in children's grouping of animals in biologically meaningful ways. Although broadly consistent with the predictions from computational modeling studies, by only examining changes in a single domain, the existing studies have not examined across-domain differentiation as a direct result of experience. ...
... To test the hypothesis above, we examined changes in semantic structure in preschool-and kindergarten-aged children enrolled in summer camps at a botanical garden; these young children are unlikely to have highly differentiated representations of biological categories targeted by the camps (Hatano et al., 1993;Unger et al., 2016). Unlike prior studies that recruited separate training and control groups (Badger & Shapiro, 2019;Unger & Fisher, 2019)-leaving open the possibility that differences between groups were driven by factors other than the learning experience-all participants were children whose parents enrolled them in a program at the same botanical garden. Also unlike prior studies which examined changes in a single domain, we measured changes in within-and across-domain differentiation of two biological domains in children who completed one of two summer camps. ...
... The final sample included data from 29 children (19 girls and 10 boys; M = 4.5 years, SD = .6). This sample size is comparable to Unger and Fisher (2019), who examined changes in semantic structure in children attending a zoo camp. Data from all children who completed both the pre-and the posttest were combined for analyses because (a) the programs' objectives and activities were identical across the 2 years, (b) the same educator lead all activities, and (c) the hypotheses were tested using a within-subjects design. ...
Article
Full-text available
Organized semantic networks reflecting distinctions within and across domains of knowledge are critical for higher‐level cognition. Thus, understanding how semantic structure changes with experience is a fundamental question in developmental science. This study probed changes in semantic structure in 4–6 year‐old children (N = 29) as a result of participating in an enrichment program at a local botanical garden. This study presents the first direct evidence that (a) the accumulation of experience with items in a domain promoted increases in both within‐ and across‐domain semantic differentiation, and that (b) this experience‐driven semantic differentiation generalized to nonexperienced items. These findings have implications for understanding the role of experience in building semantic networks, and for conceptualizing the contribution of enrichment experiences to academic success.
... Una categoría es un conjunto de objetos o entidades que comparten un núcleo esencial o son similares en las propiedades perceptivas, biológicas o funcionales que poseen. Los objetos pueden ser considerados como miembros de múltiples categorías: un perro puede ser vinculado con un gato porque ambos son animales (categorización taxonómica) o con un hueso porque le gusta morderlo, enterrarlo y desenterrarlo (categorización temática) (Blaye & Jacques, 2009); (Favarotto, García Coni, Magani, & Vivas, 2014); (Unger & Fisher, 2019). ...
... Tradicionalmente, se ha puesto el foco en la categorización o en las relaciones taxonómicas, que se caracterizan por las propiedades observables e inobservables que comparten los objetos y que forman una clase (e.g., animal o medios de transporte). Sin embargo, también existe la categorización puramente perceptiva, regida por los atributos físicos de los objetos y la temática, que se conforma a partir de relaciones de contigüidad e incluye elementos heterogéneos que pertenecen a un mismo evento (Estes, Golonka, & Jones, 2011); ; (Lewis, Poeppel, & Murphy, 2015); (Sadeghi, McClelland, & Hoffman, 2015); (Unger, Fisher, Nugent, Ventura, & MacLellan, 2016); (Unger & Fisher, 2019). ...
Article
Full-text available
El estudio de la organización del conocimiento en la memoria semántica suscita gran interés en Psicología Cognitiva y Neuropsicología. El conocimiento semántico está representado por conceptos que comparten características y forman una jerarquía inclusiva -organización taxonómica-, o que se vinculan en tiempo y espacio -organización temática o situacional-. Se considera que a lo largo del desarrollo cambia la preferencia por estos tipos de organización, pero son pocos los estudios que comparan las organizaciones conceptuales de niños en edad escolar, adultos jóvenes y mayores, y sus resultados son divergentes. Asimismo, la organización conceptual también varía en función del dominio al que pertenece el concepto (vivo vs no vivo). Por lo tanto, el objetivo general de este estudio fue estudiar qué tipos de organización conceptual empleaban esos grupos en una tarea de producción de atributos para conceptos de seres vivos y no vivos. Los resultados indican que la producción de atributos taxonómicos fue significativamente mayor para los adultos jóvenes que para los adultos mayores y los niños, en tanto la producción taxonómica de estos dos últimos grupos fue pareja. En cuanto a la producción temática, fue alta y homogénea en los tres grupos de edad. Por último, para el dominio de los seres no vivos los atributos resultaron en su mayoría temáticos y perceptivos, y para el dominio de los seres vivos, mayormente perceptivos.
... To do so, we adapted the spatial arrangement method (SpAM) developed by Goldstone (1994), in which participants freely sort images according to the extent to which they perceive stimuli as semantically related without imposing the use or primacy of any specific dimension, category, or label. SpAM has been used with both adults and children (Coburn et al., 2019;Hout & Goldinger, 2016;Koch et al., 2020;Richie et al., 2020;Unger et al., 2016;Vales, Stevens, et al., 2020), validated alongside more traditional pairwise similarity judgment tasks in both adults (Hout et al., 2013) and children (Unger et al., 2016), and shown to demonstrate external validity, capturing experiencedriven changes in children's semantic knowledge in domains such as plants, animals, foods, and tools (Unger & Fisher, 2019;Vales, States, et al., 2020). Furthermore, this task uses graded similarity judgments (i.e., the distance between images) to assess children's emotion knowledge, rather than labeling particular sorting strategies as right or wrong, which allows us to better characterize patterns of change across development. ...
Article
Full-text available
The present study examined how children spontaneously represent facial cues associated with emotion. 106 three‐ to six‐year‐old children (48 male, 58 female; 9.4% Asian, 84.0% White, 6.6% more than one race) and 40 adults (10 male, 30 female; 10% Hispanic, 30% Asian, 2.5% Black, 57.5% White) were recruited from a Midwestern city (2019–2020), and sorted emotion cues in a spatial arrangement method that assesses emotion knowledge without reliance on emotion vocabulary. Using supervised and unsupervised analyses, the study found evidence for continuities and gradual changes in children's emotion knowledge compared to adults. Emotion knowledge develops through an incremental learning process in which children change their representations using combinations of factors—particularly valence—that are weighted differently across development.
... In this sample, rural children were more likely to report unstructured activities, but rural, suburban and urban children did not differ in their engagement in structured activities (Coley, 2012). Indeed, children who attend these types of formal educational experiences (i.e., immersive zoo camps) demonstrate marked improvements in their ability to organize animals into taxonomic groups (Unger and Fisher, 2019). ...
Article
Full-text available
Living things can be classified in many ways, such as taxonomic similarity (lions and lynx), or shared ecological habitat (ducks and turtles). The present studies used card-sorting and triad tasks to explore developmental and experiential changes in conceptual flexibility–the ability to switch between taxonomic and ecological construals of living things–as well as two processes underlying conceptual flexibility: salience (i.e., the ease with which relations come to mind outside of contextual influences) and availability (i.e., the presence of relations in one’s mental space) of taxonomic and ecological relations. We were also interested in the extent to which salience and availability of taxonomic and ecological relations predicted inductive inferences. Participants were 452 six to ten-year-olds from urban, suburban, and rural communities in New England. Across two studies, taxonomic relations were overwhelmingly more salient than ecological relations, although salience of ecological relations was higher among children from rural environments (Study 1) and those who engaged in unstructured exploration of nature (Study 2). Availability of ecological relations, as well as conceptual flexibility, increased with age, and was higher among children living in more rural environments. Notably, salience, but not availability, of ecological relations predicted ecological inferences. These findings suggest that taxonomic categories (i.e., groups that share both perceptual similarities and rich underlying structure) are a salient way to organize intuitive biological knowledge and that, critically, environmental richness and relevant experience contribute to the salience and availability of ecological knowledge, and thereby, conceptual flexibility in biological thinking. More generally, they highlight important linkages between domain-specific knowledge and domain-general cognitive abilities.
Article
Knowledge integration is a crucial part of learning as concepts are built over time and modality. In mathematics, rational numbers are a particularly difficult concept that requires integration across notations to understand the quantity represented by fractions and decimals (e.g., ½ and 0.5). We investigated how 5- to 14-year-old children conceptualize rational numbers, whether it differs across development (Study 1), and whether the conceptualization is related to math achievement (Study 2). A novel Numerical SpAM task was created to document children’s unprompted rational number conceptualization. We found that most children organized rational numbers either by notation or quantity, with quantity becoming more common after age 10. Quantitative conceptualization predicted higher math achievement compared to notation, regardless of age. Moreover, conceptualizing rational numbers by quantity across distinct notations (e.g., ¹/2 and 0.5) was related to children’s math achievement. Implications for numerical understanding as well as general cognition are discussed.
Article
Children face multiple challenges in constructing an organized understanding of the animal domain, and parent-child conversations are a potential source of relevant information, both explicit and implicit. To understand these contributions, this study examined 41 parent-child dyads (child age range: 3–7 years) as they visited a virtual zoo displaying videos of a wide range of animals (humans, chimps, rhinos, beavers, owls, fish, and bees). Conversations were coded for mention of biological, psychological, and perceptual properties; gendered and neuter pronouns; comparisons between animals; and generic utterances. Conversations demonstrated principled variation in organizing the domain, using three subgroups: humans, non-human mammals, and fish and bees. This organization differed substantially from the responses participants provided in a subsequent biological property projection task. These findings inform the role of parent-child conversations in the development of the animal domain, and have implications for downstream consequences on beliefs regarding the natural world.
Article
Full-text available
Remote data collection procedures can strengthen developmental science by addressing current limitations to in-person data collection and helping recruit more diverse and larger samples of participants. Thus, remote data collection opens an opportunity for more equitable and more replicable developmental science. However, it remains an open question whether remote data collection procedures with children participants produce results comparable to those obtained using in-person data collection. This knowledge is critical to integrate results across studies using different data collection procedures. We developed novel web-based versions of two tasks that have been used in prior work with 4-6-year-old children and recruited children who were participating in a virtual enrichment program. We report the first successful remote replication of two key experimental effects that speak to the emergence of structured semantic representations (N = 52) and their role in inferential reasoning (N = 40). We discuss the implications of these findings for using remote data collection with children participants, for maintaining research collaborations with community settings, and for strengthening methodological practices in developmental science.
Article
Full-text available
Previous research has suggested that distributional learning mechanisms may contribute to the acquisition of semantic knowledge. However, distributional learning mechanisms, statistical learning, and contemporary “deep learning” approaches have been criticized for being incapable of learning the kind of abstract and structured knowledge that many think is required for acquisition of semantic knowledge. In this paper, we show that recurrent neural networks, trained on noisy naturalistic speech to children, do in fact learn what appears to be abstract and structured knowledge. We trained two types of recurrent neural networks (Simple Recurrent Network, and Long Short-Term Memory) to predict word sequences in a 5-million-word corpus of speech directed to children ages 0–3 years old, and assessed what semantic knowledge they acquired. We found that learned internal representations are encoding various abstract grammatical and semantic features that are useful for predicting word sequences. Assessing the organization of semantic knowledge in terms of the similarity structure, we found evidence of emergent categorical and hierarchical structure in both models. We found that the Long Short-term Memory (LSTM) and SRN are both learning very similar kinds of representations, but the LSTM achieved higher levels of performance on a quantitative evaluation. We also trained a non-recurrent neural network, Skip-gram, on the same input to compare our results to the state-of-the-art in machine learning. We found that Skip-gram achieves relatively similar performance to the LSTM, but is representing words more in terms of thematic compared to taxonomic relations, and we provide reasons why this might be the case. Our findings show that a learning system that derives abstract, distributed representations for the purpose of predicting sequential dependencies in naturalistic language may provide insight into emergence of many properties of the developing semantic system.
Article
Full-text available
Semantic knowledge is a crucial aspect of higher cognition. Theoretical accounts of semantic knowledge posit that relations between concepts provide organizational structure that converts information known about individual entities into an interconnected network in which concepts can be linked by many types of relations (e.g., taxonomic, thematic). The goal of the current research was to address several methodological shortcomings of prior studies on the development of semantic organization, by using a variant of the spatial arrangement method (SpAM) to collect graded judgments of relatedness for a set of entities that can be cross-classified into either taxonomic or thematic groups. In Experiment 1, we used the cross-classify SpAM (CC–SpAM) to obtain graded relatedness judgments and derive a representation of developmental changes in the organization of semantic knowledge. In Experiment 2, we validated the findings of Experiment 1 by using a more traditional pairwise similarity judgment paradigm. Across both experiments, we found that an early recognition of links between entities that are both taxonomically and thematically related preceded an increasing recognition of links based on a single type of relation. The utility of CC–SpAM for evaluating theoretical accounts of semantic development is discussed.
Article
Full-text available
Inductive generalization is ubiquitous in human cognition. In the developmental literature, two different theoretical accounts of this important process have been proposed: a naïve theory account and a similarity-based account. However, a number of recent findings cannot be explained within the existing theoretical accounts. We describe a revised version of the similarity-based account of inductive generalization with familiar categories. We tested the novel predictions of this account in two reported studies with 4-year-old children (N = 57). The reported studies include the first short-term longitudinal investigation of the development of children's induction with familiar categories, and it is the first study to explore the role of individual differences in semantic organization, general intelligence, working memory, and inhibition in children's induction.
Article
Full-text available
An influential position in lexical semantics holds that semantic representations for words can be derived through analysis of patterns of lexical co-occurrence in large language corpora. Frith (1957) famously summarised this principle as "you shall know a word by the company it keeps". We explored whether the same principle could be applied to non-verbal patterns of object co-occurrence in natural scenes. We performed latent semantic analysis (LSA) on a set of photographed scenes in which all of the objects present had been manually labelled. This resulted in a representation of objects in a high-dimensional space in which similarity between two objects indicated the degree to which they appeared in similar scenes. These representations revealed similarities among objects belonging to the same taxonomic category (e.g., vehicles, items of clothing) as well as cross-category associations (e.g., between fruits and kitchen utensils). We also compared representations generated from this scene dataset with two established methods for elucidating semantic representations: (a) a published database of semantic features generated verbally by participants and (b) LSA applied to a linguistic corpus in the usual fashion. Statistical comparisons of the three methods indicated significant association between the structures revealed by each method, with the scene dataset displaying greater convergence with feature-based representations than did LSA applied to linguistic data. The results indicate that information about the conceptual significance of objects can be extracted from their patterns of co-occurrence in natural environments, opening the possibility for such data to be incorporated into existing models of conceptual representation.
Chapter
Full-text available
Meaning is a fundamental component of nearly all aspects of human cognition, but formal models of semantic memory have classically lagged behind many other areas of cognition. However, computational models of semantic memory have seen a surge progress in the last two decades, advancing our knowledge of how meaning is constructed from experience, how knowledge is represented and used, and what processes are likely to be culprit in disorders characterized by semantic impairment. This chapter provides an overview of several recent clusters of models and trends in the literature, including modern connectionist and distributional models of semantic memory, and contemporary advances in grounding semantic models with perceptual information and models of compositional semantics. Several common lessons have emerged from both the connectionist and distributional literatures, and we attempt to synthesize these themes to better focus future developments in semantic modeling.
Article
Full-text available
Maximum likelihood or restricted maximum likelihood (REML) estimates of the parameters in linear mixed-effects models can be determined using the lmer function in the lme4 package for R. As for most model-fitting functions in R, the model is described in an lmer call by a formula, in this case including both fixed- and random-effects terms. The formula and data together determine a numerical representation of the model from which the profiled deviance or the profiled REML criterion can be evaluated as a function of some of the model parameters. The appropriate criterion is optimized, using one of the constrained optimization functions in R, to provide the parameter estimates. We describe the structure of the model, the steps in evaluating the profiled deviance or REML criterion, and the structure of classes or types that represents such a model. Sufficient detail is included to allow specialization of these structures by users who wish to write functions to fit specialized linear mixed models, such as models incorporating pedigrees or smoothing splines, that are not easily expressible in the formula language used by lmer.
Article
The encoding, consolidation, and retrieval of events and facts form the basis for acquiring new skills and knowledge. Prior knowledge can enhance those memory processes considerably and thus foster knowledge acquisition. But prior knowledge can also hinder knowledge acquisition, in particular when the to-be-learned information is inconsistent with the presuppositions of the learner. Therefore, taking students' prior knowledge into account and knowing about the way it affects memory processes is important for optimization of students' learning. Recent behavioral and neuroimaging experiments have shed new light on the neural mechanisms through which prior knowledge affects memory. However, relatively little is known about developmental differences in the ability to make efficient use of one's knowledge base for memory purposes. In this article, we review and integrate recent empirical evidence from developmental psychology and cognitive neuroscience about the effects of prior knowledge on memory processes. In particular, this may entail an extended shift from processing in the medial temporal lobes of the brain toward processing in the neocortex. Such findings have implications for students as developing individuals. Therefore, we highlight recent insights from cognitive neuroscience that call for further investigation in educational settings, discussing to what extent these novel insights may inform teaching in the classroom.
Article
Models of category learning have been extensively studied in cognitive science and primarily tested on perceptual abstractions or artificial stimuli. In this paper, we focus on categories acquired from natural language stimuli, that is, words (e.g., chair is a member of the furniture category). We present a Bayesian model that, unlike previous work, learns both categories and their features in a single process. We model category induction as two interrelated subproblems: (a) the acquisition of features that discriminate among categories, and (b) the grouping of concepts into categories based on those features. Our model learns categories incrementally using particle filters, a sequential Monte Carlo method commonly used for approximate probabilistic inference that sequentially integrates newly observed data and can be viewed as a plausible mechanism for human learning. Experimental results show that our incremental learner obtains meaningful categories which yield a closer fit to behavioral data compared to related models while at the same time acquiring features which characterize the learned categories. (An earlier version of this work was published in Frermann and Lapata .).
Article
The goal of the current study is to explore the influence of knowledge on socioeconomic discrepancies in word learning and comprehension. After establishing socioeconomic differences in background knowledge (Study 1), the authors presented children with a storybook that incorporates this knowledge (Study 2). Results indicated that middle-income children learned significantly more words and comprehended the story better than lower-income children. By contrast, Study 3 presented children with a novel category and found that children performed equally in their word learning and comprehension. This suggests that socioeconomic differences in vocabulary and comprehension skills may be partially explained by differences in extant knowledge.
Article
Category-based induction is a hallmark of mature cognition; however, little is known about its origins. This study evaluated the hypothesis that category-based induction is related to semantic development. Computational studies suggest that early on there is little differentiation among concepts, but learning and development lead to increased differentiation based on taxonomic relatedness. This study reports findings from a new task aimed to (a) examine this putative increase in semantic differentiation and (b) test whether individual differences in semantic differentiation are related to category-based induction in 4- to 7-year-old children (N = 85). The results provide the first empirical evidence of an age-related increase in differentiation of representations of animal concepts and suggest that category-based induction is related to increased semantic differentiation.