IQ (The Intelligence Quotient)

Full text

I
IQ
Louis D. Matzel and Bruno Sauce
Department of Psychology, Rutgers University,
Piscataway, NJ, USA
Introduction
Intelligence is the ability to think rationally, learn
effectively, understand complex ideas, and adapt
to the environment. Accordingly, intelligence is
best seen as a general ability that can influence
performance on a wide range of cognitive tasks.
IQ (the intelligence quotient) is the quantification
of an individual’s intelligence relative to peers of a
similar age. IQ is one of the most heritable psy-
chological traits, and an individual’s score on a
modern IQ test is a good predictor of many life
outcomes, including educational and career suc-
cess, health, longevity, and even happiness
(Gottfredson 1998). Like humans, several species
of animals express a “general cognitive ability”
that influences performance on broad and diverse
cognitive tasks, and moreover, animals exhibit a
wide range of individual variations in this ability.
Intelligence and Intelligence Testing (IQ) in
Humans
It has long been recognized that intelligence
varies across individuals. Colloquially, we refer
to someone as “brilliant”or comment that our dog
is a “little dull.”While it is easy (and common) to
make these kind of characterizations, it has histor-
ically been difficult to formulate a definition of
this trait. In 1995, a committee of the American
Psychological Association stated that “Individ-
uals differ from one another in their ability to
understand complex ideas, to adapt effectively to
the environment, to learn from experience, to
engage in various forms of reasoning, to over-
come obstacles by taking thought. Concepts of
‘intelligence’are attempts to clarify and organize
this complex set of phenomena”(Neisser et al.
1996). In an article in the Wall Street Journal
(December 13, 1994) signed by 52 intelligence
researchers, it was asserted that intelligence was
“a very general mental capability that, among
other things, involves the ability to reason, plan,
solve problems, think abstractly, comprehend
complex ideas, learn quickly and learn from expe-
rience. It reflects a broader and deeper capability
for comprehending our surroundings.”
The above definitions are simultaneously
vague and broad. Although provided by experts
on intelligence, they differ little (if only in form)
from colloquial descriptions of the trait that one
might hear from a random sample of college
undergraduates. While it has been more than
100 years since Spearman (1904) formally
described the concept of “general intelligence”
(also called “g”), we still struggle with its defini-
tion, but nevertheless, we recognize it and we
make inferences about its consequences. In this
regard, the quantification of intelligence is best
relegated to performance on psychometric tests.
#Springer International Publishing AG 2017
J. Vonk, T.K. Shackelford (eds.), Encyclopedia of Animal Cognition and Behavior,
https://doi.org/10.1007/978-3-319-47829-6_1080-1

The rationale for most psychometric tests is
roughly based on Spearman’s early observation
that performance on a wide range of cognitive
tasks is positively correlated (i.e., if you perform
well on one, you tend to perform well on others)
and, as such, can be reduced to a single index of
aggregate performance across a battery of diverse
tests. In fact, psychometric tests (e.g., the
Stanford-Binet, the Wechsler or WAIS, and the
Raven’s Progressive Matrixes or RPM) do differ
in their content and structure. For instance, the
Stanford-Binet includes questions that are cultur-
ally relevant and thus is best suited to predict
performance in a particular culture’s school sys-
tem. The WAIS is less culturally biased but, like
the Stanford-Binet, includes categories of ques-
tions that are presumed to reflect domains of abil-
ities (verbal comprehension, working memory,
perceptual reasoning, processing speed). An indi-
vidual’s performance on tests within a particular
domain (e.g., reasoning) tends to be highly corre-
lated, while performance on tests across domains
(e.g., a reasoning task and a spatial task) is usually
less correlated. Nevertheless, positive correlations
are observed between performance on all tests in
the battery. This is in line with the conclusion that
all cognitive abilities are regulated (to varying
degrees) by one general factor, or Spearman’s
“g,”while other specific abilities might influence
performance within a particular domain.These
kinds of observations have led to the development
of hierarchical models regarding the structure of
intelligence, where ginfluences domains of spe-
cific abilities, which influence tasks within those
domains. An illustration of a hierarchical model is
provided in Fig. 1.
Since many studies on intelligence use factors
analyses, a brief explanation of this technique is
warranted. Briefly, a factor analysis is a statistical
method which reduces a large number of correla-
tions into as few explanatory factors as possible.
If, for example, all of the correlations across sev-
eral tests of cognitive ability are strongly positive,
the factor analysis recognizes that a common
source of variance contributed to performance on
all tasks, and this would be described as a general
factor. In reality, the outcome of such an analysis
can be much more complicated, and of course we
might be interested in large numbers of cognitive
tasks, some of which represent clusters of what are
presumed to be specialized abilities. In these
cases, the factor analysis might extract a general
factor, as well as secondary factors, which explain
relationships between only subsets of the tasks
being considered. Of course, if no single source
of variance was common to all tasks, a factor
analysis might reveal no common factor at all.
When factor analyses are performed on human
intelligence test data (such as from the WAIS), it
is typical to find a general factor (i.e., general
intelligence) as well as secondary factors that
describe specific cognitive domains (e.g., spatial
abilities; see Fig. 1).
Remember that the Stanford-Binet and the
WAIS include tests of many different abilities,
and an individual’s aggregate performance across
all of these tests is used to estimate that individ-
ual’s intelligence. In contrast, the RPM is an intel-
ligence test that is based exclusively on only one
ability and, accordingly, includes only progres-
sively difficult tests of perceptual (analogical) rea-
soning. This test structure is based on an
assumption that reasoning is representative of
the core ability that regulates all intelligence
(Raven et al. 1998). Because of its format, the
RPM requires no knowledge of culture or
language.
Unlike a qualitative description of intelligence,
the IQ score is a quotient, that is, it is an individ-
ual’s score on a standardized test relative to that
individual’s age-matched peers. It is true that an
individual’s IQ score will tend to remain stable
across the lifespan, i.e., the IQ scores of a group of
8-year-olds will be highly correlated with their
scores at 90 years of age (r=~.80). This does
not mean that individual’s raw cognitive ability is
the same across the lifespan. For example, were
we to administer an RPM to one individual at 8,
25, 50, and 90 years of age, the number of correct
answers would be about the same at 8 and 90 years
of age, while at 25, the individual would answer at
least twice as many questions correctly (with the
50-year-old somewhere in between). So why do
we say that an individual’s IQ remains constant
across the lifespan? Because IQ is approximately
unchanging relative to persons of a similar age,
2 IQ

i.e., a person who is smarter than most of his/her
peers at 8 years of age will be smarter than his/her
peers at 50 and 90 years of age (Deary 2014),
despite the inevitable truth that our cognitive abil-
ities decline with age.
Regarding the nature of intelligence or IQ,
many persons will incorrectly assume that high
intelligence is necessarily reflected in a high level
of knowledge. In fact, high intelligence promotes
the ease with which we acquire knowledge, but
intelligence itself is independent of knowledge.
Why then do some IQ tests (such as the
Stanford-Binet) have components that test knowl-
edge? Simply because all other things being
equal, a smarter individual is likely to acquire
more knowledge. Learning is easier for that indi-
vidual than it might be to someone of lesser intel-
ligence. In this regard, scholastic aptitude tests
such as the SAT are often a good approximation
of intelligence as measured on a knowledge-free
test such as the RPM (r=.5–.6). However,
knowledge and intelligence need not always be
related. For instance, an individual with innately
high intelligence might (through some act of fate)
live in an impoverished environment where the
opportunities to acquire knowledge are severely
limited. This is exactly why an IQ test such as the
RPM has no measures of knowledge (only per-
ceptual reasoning) and is considered by many to
be a more pure measure of innate ability.
Given the different content and structure of
psychometric intelligence tests, it might be sur-
prising to find that individuals’scores on these
tests are strongly correlated (rs will typically
range from 0.8 to 0.9). Even more surprising is
the popular assertion (sometimes even by some
with advanced degrees in psychology) that “IQ
tests measure nothing of functional significance.”
Standardized intelligence tests first received wide-
spread recognition owing to the US government’s
use of a modified version of the early Stanford-
Binet to determine assignments of new recruits in
World War I. These assignments were highly
effective (relative to the previous practice of
assignments based on patronage or chance) and
are widely regarded as having contributed to the
USA’s success in WWI. Since that time, we have
collected a wide array of data regarding the pre-
dictive capacity of IQ tests. For instance, a child’s
IQ score is highly predictive of obvious outcomes
g
Processing
Speed Domain
Memory
Domain
Reasoning
Domain
?
Domain
Comprehension
Domain
reasoning
tasks
speed
tasks
memory
tasks
spatial
tasks
?
tasks
Level 3:
General ability
Level 2:
Domains of ability
Level 1:
Specific tests
IQ, Fig. 1 The hierarchical model of intelligence. Level
1represents specific tests that are emblematic of various
domains of cognitive ability. Some potential domains are
illustrated in Level 2. The number and content of these
domains is a matter of some debate, although there is wide
agreement on the existence of the four domains that are
illustrated. The fifth domain (?) acknowledges that other
domains may exist. People who perform well on tasks from
one domain tend to perform well on tasks from other
domains. This suggests the existence of a general influence
on cognitive abilities, represented in Level 3. This general
influence is commonly referred to as general intelligence or
simply “intelligence.”This model does not require only
one type of intelligence. Rather, it assumes that a general
ability influence other more domain-specific abilities.
IQ 3

such as educational and career success, as well as
lifetime income. But IQ test performance predicts
many less obvious outcomes such as the distance
one will travel from his/her place of birth, the
likelihood of incarceration, the likelihood of
drug addiction, the age of death, incidence of
type II diabetes, ratings of happiness, and even
your spouse’sincome and IQ (for a comprehen-
sive review of the predictive capacity of the IQ
test, see Gottfredson 1998). IQ scores are even
inversely related to the likelihood that an individ-
ual will murder their spouse! To quote
Gottfredson (1998, page 24), “No matter their
form or content, tests of mental skills invariably
point to the existence of a global factor that per-
meates all aspects of cognition. This factor seems
to have considerable influence on a person’s prac-
tical quality of life. Intelligence as measured by IQ
tests is the single most effective predictor known
of individual performance at school and on the
job”as well as many other aspects of well-being.
Thus, far from being a “social construct”with no
functional significance, the modern IQ test is a
highly effective (and widely used) diagnostic
and predictive tool.
Intelligence in Nonhuman Animals
Although studies of individual differences in ani-
mal intelligence had been frequent early in the
twentieth century (Thorndike 1911,1935; Tolman
1924; Tryon 1940), the emergent focus on exper-
imental (rather than correlational) studies tended
to limit the interest in this topic in the later part of
that century. However, during the past two
decades, interest in individual differences in ani-
mal intelligence has seen a dramatic reemergence.
As discussed above, contemporary definitions of
intelligence tend to be vague, broad, and, to some
degree, a matter of debate (Sternberg 1985). Nev-
ertheless, psychometric tests of intelligence do
appear to characterize a trait captured in both
colloquial and empirical definitions of intelli-
gence, i.e., the ability to understand, learn, and
reason. To explore a trait analogous to intelligence
in nonhuman animals, researchers have developed
tests to characterize a similar set of skills, most
notably in mice and monkeys.
Genetically heterogeneous mice (i.e., mice
with genetic variability that translates into mea-
surable individual differences) have been tested
on large batteries of cognitive tasks to determine
the existence of a general cognitive ability in mice
analogous to IQ. In one such study (Kolata et al.
2008), 241 mice were tested on seven cognitive
tasks, which included tests of working memory
capacity, associative learning, operant learning,
and spatial learning abilities. Using factor analy-
sis, it was observed that a general factor
influenced performance in these mice and this
factor accounted for 38% of the variance across
tasks. This is comparable to what is known from
tests of humans’abilities, where it is believed that
general intelligence accounts for 40–50% of the
variance in performance across a broad array of
cognitive tests. In addition, a domain-specific fac-
tor was found to account for the performance of
mice on a subset of tasks that shared a dependence
on spatial processing. These results provide evi-
dence for a general learning/cognitive factor in
genetically heterogeneous mice. Furthermore
(and similar to human cognitive performance),
these results suggest a hierarchical structure (see
Fig. 1) of cognitive abilities in mice, where a
general factor influences performance on sub-
domains of abilities. Importantly, mice also
exhibited considerable variability in their general
cognitive performance. In fact, the general abili-
ties of mice were normally distributed, such that
most mice expressed average abilities, while some
were “bright”(performing well on all tasks),
while some were “dull”(performing poorly on
all tasks).
As described above, reasoning is considered to
be a hallmark of intelligence and is considered by
some to be the general factor that underlies varia-
tions in intelligence. It has previously been
established that humans are capable of “fast map-
ping”(Carey and Bartlett 1978), a process whereby
a new concept can be acquired based on a logical
inference, corresponding with Aristotle’sdescrip-
tion of deductive reasoning. Fast mapping is
believed to play a critical role in the extraordinarily
rapid acquisition of information during early
human development and explains (in part) the pro-
digious rate at which children gain vocabulary. For
4 IQ

example, when faced with a group of familiar items
described by familiar words, an infant will quickly
associate an unfamiliar word with a novel item
added to the set of familiar items, and this associ-
ation requires no overt “pairing”of the novel word
and its corresponding novel item.
Fast mapping based on responses to human
language has also been demonstrated in dogs
(Tomasello and Kaminski 2004; Pilley and Reid
2011), where border collies can successfully find a
novel object when commanded (with a novel
word) to retrieve that novel object from within a
large set of familiar objects. Using a similar strat-
egy, fast mapping has been assessed in mice,
although the task was not based on responses to
language. Mice were first trained to associate pairs
of objects, where, upon exposure to a sample
object, the correct choice of a target object earned
the mouse a food reward. Following training, the
mice could successfully use the sample object to
guide its choice of a target object out of a set of
familiar objects. (This type of performance is
emblematic of “paired associate learning.”)To
test “fast mapping,”the animal was then presented
with a novel sample object and allowed to choose
a target object from a set containing several famil-
iar objects and one novel object. If the mice were
capable of fast mapping (inference by exclusion),
they should choose the novel target object
(in response to the novel sample) since all other
objects in the set had a previously established
meaning. Mice perform quite well in this task,
choosing the novel object at an average rate far
better than chance. However, not all mice perform
similarly, and while some exhibit perfect perfor-
mance, some consistently make incorrect choices.
The likelihood of a mouse’s success in this fast
mapping task is correlated with their performance
on other more elemental cognitive tasks (e.g.,
associative learning, spatial learning, operant
learning), suggesting that as in humans, this
form of reasoning ability is related to more general
cognitive abilities (Wass et al. 2012).
General cognitive abilities of mice have also
been described by Galsworthy et al. (2002), who
compared the performance of 40 genetically het-
erogeneous mice across a battery of cognitive
tests distinct from those reported in the studies
described above. All measures of cognitive per-
formance loaded positively on a principal compo-
nent that accounted for 31% of the variance across
mice, again suggesting the presence of a common
influence on performance on all tasks. In addition,
Galsworthy et al. calculated the heritability of this
general cognitive ability in mice. (This was
accomplished through a classic sibling analysis,
which assesses the degree of relatedness between
siblings on some variable of interest, in this case
general cognitive ability.) The heritability of the
general cognitive ability of mice was estimated at
approximately 0.4 (on a scale of 0–1), suggesting
a moderate genetic contribution to the expression
of this trait. These results of Galsworthy et al. are
quite informative. They indicate that the “intelli-
gence”of mice is moderately heritable, at a level
that is comparable to what is observed among
teenage humans. Note that the heritability of
human intelligence actually increases across the
lifespan, reaching a plateau of approximately
.80 at 50 years of age. This increase in heritability
is presumed to reflect the interactions of genes
with the environment, where persons of similar
IQ become even more similar as they gravitate to
similar cognitive challenges. Unlike typical
humans, laboratory mice are maintained in a
behaviorally sterile and homogeneous environ-
ment. Consequently, these mice cannot select the
environments or challenges that might maximize
cognitive differences, thus constraining the gene-
environment interaction.
In addition to rodents, individual differences in
a general cognitive ability have been observed in
several species of nonhuman primates. While
most studies of nonhuman primates have been
designed to compare differences in intelligence
between species (leading to a popular hypothesis
that brain size is related to intelligence; Burkart
et al. 2016), at least one study was designed
explicitly to assess individual differences in the
expression of a general cognitive influence within
a single species. Banerjee et al. (2009) adminis-
tered a large and diverse battery of cognitive tests
to 22 tamarin monkeys (Saguinus oedipus). The
cognitive tasks covered a wide range of cognitive
skills and domains, including occluded reach,
targeted reach (reward retrieval from a moving
IQ 5

pendulum), adaptation to an observed change in
reward location (a measure of executive control),
reversal learning, novel object recognition,
numerical discrimination, acoustic habituation,
object tracking (an index of attention), social
tracking (gaze at a conspecific), hidden reward
retrieval after various delays, and a food retrieval
puzzle (which was asserted to tax reasoning).
Banerjee et al. observed positive correlations in
the monkeys’performance across all tasks. Using
a type of factor analysis, all tasks loaded posi-
tively on a common factor. The weight of these
loadings (an index of the degree to which a vari-
able is impacted by that factor) could be described
as “weak”to “moderate.”Expectedly, the tasks
with the least obvious cognitive demands
(targeted reach and social tracking) loaded most
weakly. In total, these results provide evidence for
individual differences in the expression of a gen-
eral cognitive ability among tamarins, and more-
over, that the general factor’sinfluence is directly
related to the level of the cognitive demand.
What is the Latent Factor that Regulates
Intelligence?
Many factors, such as speed of processing or brain
size, have been suggested to underlie variations in
intelligence. However, correlational analyses
have typically found only weak relationships
between these factors and intelligence. Two clear
exceptions should be noted. Both reasoning abil-
ity and working memory capacity are strongly
predictive of IQ (and as noted previously, the
RPM intelligence test is based solely on perfor-
mance on analogical reasoning tasks). Although it
was once commonly asserted that reasoning abil-
ity was the latent factor which regulated individ-
ual differences in intelligence, it has been more
recently hypothesized that working memory may
serve such a function. In his classic textbook on
intelligence, Mackintosh describes the full ratio-
nale for this hypothesis and points out that it is
easy to surmise the way that working memory
could influence reasoning, as will be seen below,
while it is more difficult to imagine the opposite
being true (Mackintosh 1998).
Since their inception, intelligence test batteries
commonly included tests of simple memory span
(e.g., the number of items from a briefly studied
list that an individual can correctly recall). Some-
what surprisingly though, this seemingly elemen-
tal ability has only a weak relationship to general
intelligence. In 1980, an important observation by
Daneman and Carpenter (1980) shed light on the
relationship between memory and intelligence.
Daneman and Carpenter found that simply
remembering a list of words was only weakly
related to general intelligence (in this case, esti-
mated through reading comprehension). In con-
trast, if the same words appeared at the end of
sentences, the ability to remember those words
was strongly correlated with general intelligence.
This led to the hypothesis that simple retention
had only a small (if any) role in the regulation of
intelligence, while “working memory capacity”
had a more central role.
While short-term memory simply holds infor-
mation, the working memory system is one which
stores information while manipulating and utiliz-
ing that information (often during conditions of
high interference) for a particular goal. Working
memory is employed for most cognitive tasks. For
instance, your ability to read and comprehend this
paragraph requires that your remember words,
synthesize the meaning of strings of words, and
try to extract the overall message embedded in
those strings of words. Obviously, your memory
and manipulation of words and thoughts can
become confused depending on the content of
the paragraph. A similar rationale for the imple-
mentation of working memory can be applied to
virtually any task; imagine doing a mental math
problem or solving a spatial puzzle. In this regard,
an analogical reasoning problem (such as might
appear on the RPM test of intelligence) requires
the individual to hold potential solutions in mem-
ory, compare the utility of those solutions, revise
the solutions, and store the revised solutions in
temporary memory. But while analogical reason-
ing depends on the efficient application of work-
ing memory, it is not clear that the application of
working memory has any dependence on reason-
ing abilities. It is this ubiquitous demand for
working memory that has led to the assertion
that working memory may be the basis for the
overall performance on an intelligence test.
6 IQ

Since the original report of Daneman and Carpen-
ter, many studies have found evidence for the
relationship of working memory capacity to gen-
eral intelligence (for a brief review, see Engle
2002).
Unlike human research, only limited work has
been done to assess the relationship between
working memory and intelligence in nonhuman
animals. Some studies have found a relationship
between working memory and intelligence in
mice, but such correlations cannot be assumed to
reflect a cause-and-effect relationship. The direc-
tion of cause between working memory and intel-
ligence cannot be determined, and moreover, both
traits might be influenced by a third, hidden var-
iable. It should be noted that the same difficulties
exist when interpreting this relationship in
humans. However, in both humans and mice, a
causal relationship between working memory and
intelligence has been explored. For instance,
Jaeggi et al. (2008) exposed humans to intensive
working memory training by having them per-
form a “dual n-back”task for several weeks. The
dual n-back task requires the subject to simulta-
neously monitor a stream of visual and auditory
cues (a sequence of visual locations and a
sequence of auditory letters). The subject’s task
is to identify matches that occur in each stream of
information (e.g., auditory “B”matches auditory
“B,”or upper right grid location matches upper
right grid location) that occur a specific number of
places back in the stream of information, e.g.,
2-back, 3-back, and 4-back. Humans typically
find this task to be extremely difficult (and even
stressful), and the larger the n-back requirement
(e.g., 4-back rather than 2-back), the more diffi-
cult the task becomes. This task is considered to
tax working memory capacity, and humans will
typically improve across days of training; they
may initially find 2-back to be very difficult but
might eventually master 6-back. Jaeggi et al.
observed that several weeks of such training
improved working memory and had positive
(although small) effects on intelligence test per-
formance. This suggests that working memory
has a direct causal influence on an individual’s
intelligence.
The work by Jaeggi et al. (2008)isbyno
means conclusive. While it has been replicated
several times, others have failed to replicate
these results, often after extensive attempts to do
so (Redick et al. 2012; Shipstead et al. 2012).
Relatedly, commercial “brain training”devices
based on working memory training have been
widely criticized as ineffective (Simons et al.
2016). Although this controversy has not been
resolved, it is clear that training working memory
in humans is complicated by the fact that humans
regularly engage in the use of working memory
outside of the laboratory (e.g., your comprehen-
sion of this paragraph), and so any working mem-
ory training that occurs in the laboratory is small
in comparison. To this end, it might be useful to
consider the effects of working memory training
on laboratory animals that live in sterile cognitive
environments. Light et al. (2010) developed a task
to train working memory in mice. In this task, the
mice were required to perform simultaneously in
two mazes, and each maze required the animals to
keep track of eight locations. Since the locations
were marked by a common set of visual cues, the
mice become very confused (presumably owing
to an overload of working memory). Like n-back
training, mice get better at these mazes over a
period of weeks and, when later tested, exhibit
improvements in working memory. Likewise,
they exhibit improvements in general cognitive
performance, suggesting that the efficacy of work-
ing memory can under certain circumstances have
a direct causal impact on a mouse’s intelligence.
In the case of both humans and mice, these studies
of the impact of working memory training on
intelligence provide further evidence that intelli-
gence is malleable. That is, although intelligence
is heritable, genes interact with environmental
experience to regulate an individual’s IQ.
Space does not permit a detailed explanation of
the neuroanatomical systems that contribute to the
expression of intelligence or working memory.
However, these kinds of analyses also suggest
that these abilities are strongly related. Brain
areas that are active during tests of general intel-
ligence overlap considerably with brain areas
active during performance of a working memory
task (Jung and Haier 2007), and the same brain
IQ 7

areas have been implicated in the processing of
working memory in both monkeys (Konecky et al.
2017; Riley and Constantinidis 2015) and rodents
(Wass et al. 2013). In total, and although this issue
is far from resolved, our current state of under-
standing suggests that variations in working mem-
ory capacity contribute directly (at least in part) to
variations in intelligence.
Conclusion
Humans and nonhuman animals exhibit individ-
ual differences in their ability to “reason, plan,
solve problems, think abstractly, comprehend
complex ideas, learn quickly and learn from expe-
rience”(Neisser et al. 1996). This complex of
abilities is referred to as intelligence. In both
humans and animals, this trait can be assessed
through batteries of cognitive tests, and in
humans, these tests give rise to an intelligence
quotient (an “IQ score”) which quantifies an indi-
vidual’s performance relative to those of a similar
age. Studies in nonhuman animals, most remark-
ably in primates and mice, have utilized diverse
batteries of cognitive tests to measure something
analogous to IQ. The intelligence in these animals
varies among individuals and seems to be corre-
lated with processes such as reasoning and work-
ing memory. Recent research in both humans and
mice suggest that working memory training might
make causal contributions to the improvement of
IQ. Those findings have not only theoretical
implications concerning the structure and neuro-
biological insanitation of intelligence, but it also
opens up opportunities for future practical
applications.
Cross-References
▶Analogical Reasoning
▶Behavioral Genetics
▶Behavioral Variation
▶Brain Size
▶Deductive Reasoning
▶Genetic Variation
▶Heredity
▶Heritability of Behavior
▶Inductive Reasoning
▶Intelligence
▶Learning
▶Raven Scales
▶Working Memory
References
Banerjee, K., Chabris, C. F., Johnson, V. E., Lee, J. J.,
Tsao, F., & Hauser, M. D. (2009). General intelligence
in another primate: Individual differences across cog-
nitive task performance in a new world monkey
(Saguinus oedipus). PloS One, 4(6), e5883.
Burkart, J. M., Schubiger, M. N., & van Schaik, C. P.
(2016). The evolution of general intelligence. Behav-
ioral and Brain Sciences,1–65. doi:10.1017/
S0140525X16000959.
Carey, S., & Bartlett, E. (1978). Acquiring a single new
word. Proceedings of the Stanford Child Language
Conference, 15, 17–29.
Daneman, M., & Carpenter, P. A. (1980). Individual dif-
ferences in working memory and reading. Journal of
Verbal Learning and Verbal Behavior, 19, 450–466.
Deary, I. J. (2014). The stability of intelligence from child-
hood to old age. Psychological Science, 23, 239–245.
Engle, R. W. (2002). Working memory capacity as execu-
tive attention. Current Directions in Psychological Sci-
ence, 11(1), 19–23. doi:10.1111/1467-8721.00160.
Galsworthy, M. J., Paya-Cano, J. L., Monleón, S., &
Plomin, R. (2002). Evidence for general cognitive abil-
ity (g) in heterogeneous stock mice and an analysis of
potential confounds. Genes, Brain, & Behavior, 1(2),
88–95.
Gottfredson, L. S. (1998). The general intelligence factor.
Scientific American Presents,9,24–30.
Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Perrig, W. J.
(2008). Improving fluid intelligence with training on
working memory. Proceedings National Academy of
Sciences U.S.A, 105(19), 6829–6833.
Jung, R. E., & Haier, R. J. (2007). The Parieto-frontal
integration theory (P-FIT) of intelligence: Converging
neuroimaging evidence. The Behavioral and Brain Sci-
ences, 30(2), 135–154.
Kolata, S., Light, K., & Matzel, L. D. (2008). Domain-
specific and domain-general learning factors are
expressed in genetically heterogeneous CD-1 mice.
Intelligence, 36, 619–629.
Konecky, R. O., Smith, M. A., & Olson, C. R. (2017).
Monkey prefrontal neurons during sternberg task per-
formance: Full contents of working memory or most
recent item? Journal of Neurophysiology. doi:10.1152/
jn.00541.2016. pii 00541 02016.
Light, K., Kolata, S., Wass, C., Denman-Brice, A.,
Zagalsky, R., & Matzel, L. D. (2010). Working memory
training promotes general cognitive abilities in
8 IQ

genetically heterogeneous mice. Current Biology, 20,
777–782.
Mackintosh, N. J. (1998). IQ and human intelligence.
Oxford: Oxford University Press.
Neisser, U., Boodoo, G., Bouchard Jr, T. J., Boykin, A. W.,
Brody, N., Ceci, S. J., & Urbina, S. (1996). Intelli-
gence: Knowns and unknowns. American Psycholo-
gist, 51(2), 77–101. doi:10.1037/0003-066X.51.2.77.
Pilley, J. W., & Reid, A. K. (2011). Border collie compre-
hends object names as verbal referents. Behavioural
Processes, 86(2), 184–195.
Raven, J. C., Raven, J. E., & Court, J. H. (1998). Progres-
sive matrices. Oxford: Oxford Psychologists Press.
Redick, T. S., Shipstead, Z., Harrison, T. L., Hicks, K. L.,
Fried, D. E., Hambrick, D. Z.,  & Engle, R. W.
(2012). No evidence of intelligence improvement
after working memory training: A randomized,
placebo-controlled study. Journal of Experimental Psy-
chology General, 142, 359–379.
Riley, M. R., & Constantinidis, C. (2015). Role of prefron-
tal persistent activity in working memory. Frontiers in
Systems Neuroscience, 9, 181. doi:10.3389/fnsys.2015.
00181.
Shipstead, Z., Redick, T. S., & Engle, R. W. (2012). Is
working memory training effective? Psychological
Bulletin, 138(4), 628–654.
Simons, D. J., Boot, W. R., Charness, N., Gathercole, S. E.,
Chabris, C. F., Hambrick, D. Z., & Stine-Morrow,
E. A. L. (2016). Do “brain-training”programs work?
Psychological Science in the Public Interest, 17(3),
103–186. doi:10.1177/1529100616661983.
Spearman, C. (1904). General intelligence, objectively
determined and measured. American Journal of Psy-
chology, 15, 201–293.
Sternberg, R. J. (1985). Human Intelligence: The Model Is
the Message. Science, 230(4730), 1111–1118.
Thorndike, E. L. (1911). Animal intelligence: experimental
studies. New York: Macmillan.
Thorndike, E. L. (1935). Organization of behavior in the
albino rat. Psychological Monographs, 17,1–70.
Tolman, E. C. (1924). The inheritance of maze-learning
ability in rats. Journal of Comparative Psychology,
4(1), 1–18. doi:10.1037/h0071979.
Tomasello, M., & Kaminski, J. (2004). Like infant, like
dog. Science, 325, 1213–1214.
Tryon, R. C. (1940). Genetic differences in maze-learning
abilities in rats. Yearbook of the National Society for
Studies in Education, 39,111–119.
Wass, C., Denman-Brice, A., Rios, C., Light, K. R.,
Kolata, S., Smith, A. M., & Matzel, L. D. (2012).
Covariation of learning and “reasoning”abilities in
mice: Evolutionary conservation of the operations of
intelligence. Journal of Experimental Psychology. Ani-
mal Behavior Process, 38(2), 109–124.
Wass, C., Pizzo, A., Sauce, B., Kawasumi, Y., Sturzoiu, T.,
Ree, F.,  & Matzel, L. D. (2013). Dopamine D1
sensitivity in the prefrontal cortex predicts general cog-
nitive abilities and is modulated by working memory
training. Learning and Memory, 20(11), 617–627.
IQ 9