SHORT TERM MEMORY, FLUID INTELLIGENCE AND BRAIN NERVE SPEED IN A POPULATION WITH VARIABLE CULTURAL EXPOSURE RUNNING TITLE: Ache cognition KEYWORDS: short-term memory, fluid intelligence, central nerve conduction velocity, acculturation, Ache John Wagner* Robert Walker ([email protected]) Kim Hill ([email protected]) Draft #2 for Human Nature 22 pages of text 3 tables 5 figures DO NOT CITE IN ANY CONTEXT WITHOUT PERMISSION OF THE AUTHORS *Address all correspondence to: J. Wagner, Department of Anthropology, University of New Mexico. Albuquerque, New Mexico 87131. phone: 505 277-1628 email: [email protected]webpage: http://www.unm.edu/~wagner
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
SHORT TERM MEMORY, FLUID INTELLIGENCE AND BRAIN NERVE SPEED IN A POPULATION WITH VARIABLE CULTURAL EXPOSURE
(Raven et al. 1998). Our version of SS test is very similar to the Corsi Blocks tapping
task and is considered an excellent measure of short-term memory (STM). Raven�s
Progressive Matrices is the most widely used psychometric in the world and is widely
considered to be the most reliable indicator of Gf.
Short-latency potentials like P100 are only indicative of early sensory processing and
should not be confused with later evoked potentials that occur with subsequent or higher
cortical processing steps in response to a cognitive task, which have been variously but
not consistently correlated to psychometric intelligence (e.g., Caryl 1994). No prior
significant correlations between P100 and Gf have been reported save for the previously
mentioned studies that use P100 to calculate CNCV. We also use P100 values from
pattern-reversal VEP to compute CNCV and test for age-controlled correlations among
all variables.
MATERIALS AND METHODS
Study group
The Northern Ache were nomadic hunter-gatherers without horticulture before first
peaceful contact in the 1970s (Hill and Hurtado 1996; Clastres 1998). There is no
evidence of friendly relations between the Aché and any other ethnic population in
10
Paraguay until the 1960s and 1970s, when various groups made contact with outsiders.
The Aché currently live in five major mission/ reservation settlements with a total
population of about 1,000 individuals. Data for this study were collected in the
communities of Arroyo Bandera and Kuetuvy, cultural groups that lived in isolation until
the mid to late 1970s (Hill and Hurtado 1996). The Aché now have a mixed economy
with some communities heavily dependent on cultigens, farm animals, and wage labor,
while others are still partially dependent on hunting and gathering in the nearby
Mbaracayu Natural Reserve. Game animals comprise up to 80% of the Ache diet in the
forest (Kaplan et al. 2000) but the settlement diet is based on a staple of sweet manioc
planted in slash-and-burn fields.
The Aché continue to suffer considerable mortality and morbidity as a result of
disease exposure and particularly tuberculosis (Hurtado et al. 2003). Health conditions
have likely improved after settlement�evidenced by improved growth patterns in the
younger generation�with increased access to medicine and doctors.
Because of the acculturation process, the sample includes an older generation who
lived most of their lives as hunter-gatherers and a younger generation who have grown up
on reservation settlements. Following contact, the Aché have undertaken formal
classroom instruction with school days lasting about 3-4 hours and punctuated by long
seasonal vacations. All individuals under age 25 have attended some school but no one
over age 40 has ever been to school. A genealogical approach with interview-generated
age ranks was used to age all individuals in the population born before fieldwork
commenced in the late 1970s (see Hill and Hurtado 1996), while ages for individuals
born during the fieldwork period are exact to the day.
11
Visual evoked potentials (VEP) and Central Nerve Conduction Velocity (CNCV)
Most people residing in the two communities were recruited to participate in
February-March 2005 by J.W. and R.W. Results from 161 individuals are included in
these analyses (86 males, age range 5-75, mean = 28.71, std. = 18.62; 75 females, age
range 3-73, mean = 26.76, std. = 17.477). Participants had basic anthropometrics
collected including head length measured with calipers to the nearest millimeter from the
nasion to the inion. Subjects were then fitted with an electrode cap (Electro-Cap
International, Eaton, OH) made of an elastic spandex-type material with pure tin recessed
electrodes attached to the fabric and arrayed in the international 10-20 system. Recessed
electrodes place a conductive layer between the scalp and the metal conductor and have
the advantage of reducing movement artifacts (Cooper et al. 1980). Electrodes were filled
with conductive gel using a blunted needle syringe at locations F3, F4, C3, C4, P3, P4,
O1, O2, T4, T5, FZ with an average reference and a ground electrode positioned slightly
below FZ. Impedances were always held below 10 kilo-ohms and typically below 5 kilo-
ohms. Following collection of resting electroencephalography (EEG) pattern and flash
VEP were collected. However, for our purposes here we only report the P100 latency of
the pattern-reversal VEP to conform to prior studies that calculate CNCV. Of the various
measurable components of VEP, P100 is the most informative due to its stability and low
variability (Chiappa 1990).
Electrophysiological signals were amplified with Neuron-spectrum-3 and analyzed
with software Version 1.4.5.28 (Neurosoft Corporation, Ivanovo, Russia). Bandwidth
was set from 0.5 to 35 Hz with a quantization frequency of 200 Hz. EEG signals for the
VEP test were collected using the registration montage described above. The software
12
allows for post-test reconfiguration of the montage such that a VEP montage was created
with active electrodes at O1 (left occipital) and O2 (right occipital) with a reference
electrode at Fz (frontal sagittal) to better isolate occipital cortex recordings (Chiappa
1990). The pattern-reversal checkerboard was presented on a 15-inch flat-screen monitor
(LiquidVideo E15LCD1). Viewing distance was 1 m and luminance remained constant at
30 cd/m2. Field size was 18° x 13.5° with individual checks subtending an arc of 60�.
Participants were directed to focus on a red dot at the center of the checkerboard pattern,
which reversed at a rate of 1 Hz (two reversals per second) for 100 reversals. Trials were
run in duplicate and P100 was taken as the average of O1 and O2 for both trials although
these values rarely differed by more than a few milliseconds. The analysis epoch was 400
ms.
Data were inspected offline and artifacts from eye blinks, which were clearly visible
as muscle artifacts in the F3 and F4 electrodes, were excluded from the analyses as were
any other obvious signs of major electrical interference. A mean of about 80 pattern
reversals were used for averaging. It was discovered that even removing over half of the
trials had little effect on the VEP waveform and P100 latency such that we are confident
that the P100 values are robust. Nine subjects were removed from the analyses due to
uncertainty in identifying P100 waveforms. For example, one older subject (75) had
obviously occluded vision and no discernable P100 peak. Two others had obvious
indications of mental or physical disability. For the remainder of subjects, VEP
waveforms were identified for the characteristic P100 peak and marked to the nearest
millisecond as shown in Figure 1.
13
Figure 1. Representative VEP waveforms recorded from left and right occipital (O1 and O2) and referenced to Fz. By moving the line cursor to the apex of the major peak (P100) latency is recorded to the nearest millisecond. The earlier peak (N75) occurs about 75ms before the P100. X-axis is measured in milliseconds and vertical axis in microvolts.
Psychometrics: Raven's Colored Progressive Matrices and Forward Spatial Span
All testing sessions began with casual conversation and joking for several minutes.
Participants understood that they were being tested for mental ability and were typically a
bit nervous and competitive. The Raven's Colored Progressive Matrices (CPM) were
administered by K.H. alone with a subject in a small room. CPM consists of a set of 36
schematic colored figures in which one is asked to find a rule connecting a set of figures
and to complete the set according to the rule. Each figure has a piece missing and
respondents are prompted to choose one of six choices below each figure to complete the
pattern (Raven et al 1998). It was clearly explained by Hill, who has over 30 years
experience working with this populations and speaks the native language fluently, that
the subject should look at the pattern and figure out which piece fits into the missing
space correctly. The first two items on the test are extremely easy such that if a subject
missed these items it was assumed they did not understand the task correctly. If there
were errors on either of these two items, the above procedure was repeated until the
10uV
14
subjects picked answered correctly. Respondents were prodded by asking such questions
as, �do you think that piece really fits the pattern? Look, can you see how it is different
and doesn�t make the pattern look right? Pick another one that fits the pattern.,� and so
on. The patterns and rules become increasingly complex throughout the 36-item test and
no time limit was given. Prepubescent subjects � or younger than about 12 years of age �
were not included.
Short-term memory (STM) was assessed with a simple forward spatial span (SS) task,
which asks participants to replicate, following the experimenter, an increasingly long
sequence of spatial locations. Locations were 9 circles (4cm diameter) drawn randomly
on a cardboard plate (30cm diameter). Figure 2 provides an approximate pictorial
representation of the task. A semi-random (non-consecutive) sequence was committed to
memory by J.W. who began the test by pointing at the first circle and then prompted the
participant to mimic his motion. J.W. then repeated pointing to the first location and then
added another location and prompted the subject to again mimic this motion. Subjects
typically caught on very fast as to what was expected both very young and old. It was
assumed that any subject who could not replicate the first 3 locations did not understand
the task and instructions were repeated until it was clear they comprehended the task. The
spatial sequence was then lengthened by one location per trial until the subject failed to
correctly repeat a sequence. At this point, the same sequence was repeated and if subjects
were able to correctly repeat the sequence on this attempt the trial would continue with
another location added. However, if they incorrectly repeated the sequence the second
time their score was recorded as the last number of correctly repeated spatial locations.
15
Figure 2. The forward Spatial Span (SS) task. The experimenter points to the first location (1) and prompts the subject to mimic this motion before again pointing to the first location followed by a second location (2). The sequence is lengthened according to this fashion one location per trial until the participant fails to mimic two consecutive sequences.
Analytical methods and strategy
We rely on non-parametric curve fitting (i.e., least ordinary weighted sum of squares
or LOWESS) plots of performance by age to display subtle changes with age that may
not be captured in a parametric model. Individual scores that control for age are taken as
the residual off of the LOWESS fit for that particular age. We present data together and
by sex given evidence for sex differences in P100 latency and CNCV (Reed et al. 2004b)
that arises from anatomical size differences and hormonal influences (Celesia et al.
1987). Sex differences in SS and CPM are believed to be minimal (Court 1983; Hester et
al. 2004) with a slight male advantage sometimes reported (Lynn and Irwing 2004). We
also compare and contrast trends in psychometrics from other populations across
adulthood ages because this age range encompasses the potential acculturation effect on
psychometric performance previously mentioned. Given previous findings of significant
correlations between short-term memory tasks and Gf, and between Gf and CNCV, one-
tailed significance tests are justified.
16
RESULTS
Descriptive statistics for the sample are provided in Table 1. P100 and head length,
and hence CNCV, all share the same age ranges, means and standard deviations. The
mean age of respondents for the SS task is lower than for CPM because subjects under
the age of 12 were not administered the CPM.
Table 1. Descriptive statistics of the sample and tests. Significance indicators in the left mean column test for sex differences (Independent samples t-test, 2-tailed), * p < 0.05, ** p < 0.01
Test Age N Min Max Mean S.D Range (yr) Mean S.D. P100 (ms) 161 99 149 116.4 7.18 3-75 28.2 17.72
Visual Evoked Potentials (VEP) and Central Nerve Conduction Velocity (CNCV)
Figure 2 indicates that P100 latency and CNCV show developmental changes up until
late adolescence or early adulthood (~18 years of age) and then senesce slightly across
middle adulthood before rapidly senescing after age 60. While the over 60 years old
17
sample is very small (n=4), the marked senescence pattern is a likely accurate for several
reasons. First, the few subjects in this age range with identifiable P100 values exhibited
slow latencies. Second, a 75-year old subject was excluded due to an undecipherable
P100 peak�attributable to obvious visual deterioration. And lastly, the number of
individuals over age 60 is limited, suggesting that strong senescence is operating in this
population at older adult ages.
(a) (b) Figure 3. a) P100 (in ms) across the lifespan for males and females. Note that the Y-axis is inverted to give a more intuitive feel for ontogenetic patterns. b) Central Nerve Conduction Velocity (CNCV), computed as head length divided by P100 latency across the lifespan for males and females. Data are fit with a LOWESS in both figures.
Psychometrics
Figure 4a presents cross-sectional age trends spatial span (SS) and has the same
general shape as P100 latency and CNCV although SS scores appear to decline rapidly
across adulthood and particularly in comparison to a westernized population (Figure 3b).
A significant sex difference also exists in SS between ages 15 and 25 with males
outperforming females (independent samples T-test, p = 0.006, t = -2.937).
! Female" Male
0 10 20 30 40 50 60 70 80
Age
100
110
120
130
140
150
P100
(ms)
!
!
!
!
!
!
!
!
!
!!
!
!!
!!
!!
!
!!!
!!!
!
!
!
!
!!
!
!
!!!
!
!
!
!!!
!!!!
!
!!
!
!!
!
!!
!
!
!
!!
!
!
!
!
!
!
!
!
!
! !
!
!!!!
"
"
"
"
"
"
"
"
"
"
"
"
""
"
"
"""
"
"
"
"
""""
"
"
"
"
"
"
"
""
"
""
""
"
"
""
"
"
""
"""
"
"
"""
"
"
"
"
"
"
"
"
"
"
"
"
"
"
""
"
"
"
" "
"
"
"
"
"
"
"
"
! Female# Male
0 10 20 30 40 50 60 70 80
Age
1.20
1.40
1.60
1.80
CN
CV
(m/s
)
!
!!
!
!
!
!
!
!!
!!
!
!
!
!!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!!
!!!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!!
!
!
!!!
#
#
#
#
#
##
#
#
#
#
#
##
#
#
#
#
#
#
#
##
##
###
##
#
#
##
###
##
#
#
#
#
#
#
#
#
#
#
###
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
#
#
18
(a) (b) Figure 4. a) Forward spatial span scores across the lifespan with LOWESS fit curves for each sex. b) Forward spatial span scores across the lifespan for Ache and Australian samples (data adapted from Hester et al. 2004, reproduced with permission).
Figure 4a presents cross-sectional age trends in CPM and the expected pattern of
general developmental and aging trends is observed. However, a highly significant sex
difference emerges between ages 15 and 30 with males again outperforming females
(independent samples T-test, p = 0.000, t = -4.054) and scores again decline rapidly
across adulthood. The rate of this decline appears to be somewhat faster than observed in
an industrial society (Figure 4b).
4
5
6
7
8
9
10
15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Age
Spa
tial S
pan
AustralianAche
Age
80757065605550454035302520151050
Spa
tial s
pan
15
14
13
12
11
10
9
8
7
6
5
4
3
2
10
Male
Female
19
Figure 5. a) Raven's Colored Progressive Matrices (CPM) scores across the lifespan with LOWESS fit curves for each sex. b) CPM scores across the lifespan for Ache and Italian sample (data adapted from Measso et al. 1993).
Table 3 presents Pearson correlations for test variables taken as residuals off of the
LOWESS best fit line thereby controlling for age. The extremely high correlations
between P100 and CNCV result from P100 being in the denominator for the calculation
of CNVC. CPM correlates with SS (stronger in males) and with CNCV (stronger in
females). Correcting P100 for head length, or dividing head length by P100, strengthens
the correlation of SS and CPM with CNCV except for male SS where there is virtually no
effect.
0
5
10
15
20
25
30
35
20 25 30 35 40 45 50 55 60 65 70 75 80
Age
CPM
AcheItalian
Age
70656055504540353025201510
CPM
35
30
25
20
15
10
5
0
male
female
20
Table 3. Pearson correlation matrices. Lower left halves control for age by using the residuals off the LOWESS fit by age. Upper right halves indicate the number of valid (listwise) cases for a given correlation. * Correlation significant at the 0.05 level, ** correlation significant at the 0.01 level (1-tailed). Significance values provided if correlation was stronger than 0.2 or p-value less than 0.2. TOTAL P100 CNCV Spatial Span CPMP100 161 156 80 CNCV -0.756** 156 79 Spatial Span -0.064 0.076 77 CPM -0.024 0.199* 0.190* p = 0.040 p = 0.049 FEMALES P100 CNCV Spatial Span CPMP100 75 70 33 CNCV -0.827** 70 32 Spatial Span 0.005 -0.061 29 CPM -0.076 0.239 0.099 p = 0.094 p = .304 MALES P100 CNCV Spatial Span CPMP100 86 86 47 CNCV -0.769** 86 47 Spatial Span -0.141 0.139 48 p = 0.097 p = 0.101 CPM -0.072 0.157 0.218 p = 0.146 p = 0.068
DISCUSSION
Assessing cognition within a traditional or non-Western population presents a unique
set of challenges. We found correlations between measures of STM and Gf, and between
Gf and CNCV, collected among a group of traditional living forager-farmers. These
findings are consistent with previous studies (Conway et al. 2002; Engle et al. 1999; Reed
21
et al. 2004a) and are perhaps more impressive considering the nature of the sample, that
is, a traditional population, relatively small N distributed across much of the lifespan, and
controlling for age indirectly through residual analyses. Considered in this light, these
findings lend powerful support to the conceptual and empirical bases for the
interrelationships among STM, Gf and nerve speed.
Accounting for the pattern of Aché psychometric scores across the lifespan is open to
interpretation but requires consideration of the effects of variable cultural exposure in
combination perhaps with changing nutritional and disease patterns. Ache scores on both
tests decline much more rapidly across middle adulthood in comparison to trends from
other industrialized populations; SS scores are comparable at young ages before declining
rapidly and CPM scores are significantly lower at all ages.
The effects of schooling on psychometric test performance should not be
underestimated given the clear evidence, although the effect appears to be stronger for
verbal versus nonverbal tasks (Cahan and Cohen 1987). The abilities encouraged by
schooling relate directly to the functional conceptions of executive function, WMC and
Gf. In general, performance on Gf tasks depends on controlling attention, inhibiting
irrelevant information, and the speed and fidelity of information processing (Engle 2002;
Jensen 1998). To this list we would add that performance on Gf tasks also depends on
prior experience with testing materials and tasks. A cultural exposure or experience effect
is bound to be stronger when it encompasses pre-pubertal periods of higher neural
plasticity but probably operates over the entire lifecourse in terms of continual learning
and cognitive approaches to problem solving. Given that the Ache were nomadic hunter-
gatherers until the mid 1970s, it is perhaps not surprising that subsequent exposure to
22
western culture and moderate schooling could produce significantly improved
performance on psychometric tasks within the more recent generation.
At the same time, some of the cohort difference in test scores could be attributable to
steadily improving living conditions following settlement in the 1970s including
increased access to medicine, healthcare, and more consistent and higher caloric intake
(Hill and Hurtado 1996). Enduring worse living conditions throughout pre-contact life
would now be realized as more severe senescence in the older generation.
Future Directions and Conclusion
Our interest in pursuing this research was not solely guided by a desire to delineate
individual differences in intelligence per se and, based on previous studies, we expected
to encounter some degree of cultural dissonance with respect to psychometric test
dynamics. However, we also endeavored to collect objective physiological measures of
cognitive integrity with which to evaluate overall developmental and aging trends in the
population as a whole. Consider that all of the measures used here (P100/CNCV, SS,
CPM) exhibit expected age-related changes across the lifespan coincident with
development and senescence. However, the psychometric trends for the Aché are
noticeably different from that found in other populations. Combining the above
approaches could potentially offer a method for evaluating the validity of various
psychometrics within any given population.
Consider for example that aging trends in Gf tasks closely approximate those for
information processing speed in a large German sample (Li et al. 2004). Whether a
common underlying source of variation � perhaps attentional focus or neuronal speed
transmission � accounts for this pattern would not be particularly important. All that
23
would be required is for a predictable relationship to exist between the tests of interest,
and that the information be collected in a population where acculturation and schooling
differences are not a factor. If we assume that psychometrics are more prone to cultural
modification than physiological variables, it might then be possible to apply these tests in
a traditional population to compare the expected with the observed trend and evaluate the
age-specific validity of the psychometric task. We note that by recording short-latency
potentials (<200ms) while passively viewing a monitor, VEP demands minimal effort
and cooperation from participants, and thereby avoids most of the confounding effects of
acculturation, motivation and familiarity with test materials. Unfortunately, VEP results
cannot be formally compared across studies due to poorly understood variation arising
among laboratories, requiring the establishment of within-laboratory reference values
(Odom et al. 2004). Developing VEP or CNCV norms in association with Gf measures in
a population where acculturation or schooling differences are not a concern (or even
controlling for IQ) seems to be a reasonable approach for evaluating the differential and
age-specific validity of psychometrics within a given population.
The inability to formally compare VEP across studies limits our ability to characterize
the senescent VEP/CNCV pattern in the Aché. On the one hand, the large check sizes
used in this study (60�) and clear age decline in P100 across adulthood suggest marked
senescence. On the other hand, the relatively low luminance (30 cd/m2) of the
checkerboard stimulus � also presented on a flat screen monitor which has not to our
knowledge been performed before � would mitigate in the opposite direction by
exaggerating senescent trends and is likely contributing to the relatively longer P100
latencies recorded here.
24
There are limitations to this data and particularly its cross-sectional nature. Although
words such as �development", "senescence" and �change� are used, these should be
interpreted with caution. Also, forms of dementia, such as Alzheimer�s disease (AD),
were not gauged in this study and could possibly be contributing to the rapid deterioration
of cognitive skills as evidenced by psychometrics. Mean P100 latencies are normal in AD
patients indicating that optical pathways are spared (Martinelli et al. 1996). However,
from ethnographic experience we doubt dementia is highly prevalent among the Ache
and point out that age-specific mortality acting against cognitively impaired individuals
(Deary and Der 2005) could be particularly strong in relatively high senescence, high
mortality populations such as the Aché.
In conclusion, we found modest but compelling correlations among measures of
STM, Gf and brain nerve speed that accord with an emerging picture of the organization
of mental abilities (Geary 2005). Effects of task relevance, cultural exposure, and
schooling should be considered when applying western-derived psychometric in non-
western populations. The application of psychophysiological techniques such as VEP
appears useful for estimating neural integrity and may assist in evaluating the cross-
cultural efficacy of psychometric tasks. We encourage further investigations of this topic.
25
ACKNOWLEDGMENTS
A Leakey Foundation General Research Grant to JW funded this research.
Comments from several reviewers were instrumental for improving the focus of this
paper.
26
REFERENCES Allison, T., C. C. Wood, and W. R. Goff
1983 Brain stem auditory, pattern-reversal visual, and short-latency somatosensory
evoked potentials: latencies in relation to age, sex, and brain and body size.
Electroencephalography and Clinical Neurophysiology 55: 619-36.
Ardila, Alfredo, and Sonia Moreno
2001 Neuropsychological test performance in Aruaco Indians: An exploratory study.
Journal of the International Neuropsychological Society 7: 510-15.
Baddeley, Alan. D., and G. Hitch
1974 Working memory. In: G.A. Bower (ed.), The psychology of learning and
motivation; pp. 47-89. New York: Academic Press
Brecelj, J.
2003 From immature to mature pattern ERG and VEP. Documenta Ophthalmologica
107: 215-24.
Burns, Nicholas R.
1999 Biological Correlates of IQ Scores do Not Necessarily Mean That G Exists.
Psycoloquy 10.
Cahan, Sorel, and Nora Cohen
1989 Age versus schooling effects on intelligence development. Child Development 60:
1239-49.
Carroll, J. B.
1993 Human cognitive abilities: A survey of factor-analytic studies. New York:
Cambridge University Press
27
Caryl, Peter G.
1994 Early event-related potentials correlate with inspection time and intelligence.
Intelligence 18: 15-46.
Ceci, S. (1991).
1991 How much does schooling influence general intelligence and its cognitive
components? A reassessment of the evidence. Developmental Psychology. 27:
703-22.
Celesia, G. G., D. Kaufman, and S. Cone
1987 Effects of age and sex on pattern electroretinograms and visual evoked potentials.
Electroencephalography and Clinical Neurophysiology 68: 161-71.
Chiappa, K. H.
1990 Evoked potentials in clinical medicine. New York: Raven Press
Clastres, P.
1998 Chronicle of the Guayaki Indians. Cambridge: The MIT Press
Conway, Andrew R.A., Nelson Cowan, Michael F. Bunting, David J. Therriault, and
Scott R.B. Minkoff
2002 A latent variable analysis of working memory capacity, short-term memory
capacity, processing speed, and general f luid intelligence. Intelligence 30: 163-
83.
Cooper, R., J. W. Osselton, and John Crossley Shaw
1980 EEG Technology. Boston: Butterworths
Court, John H.
1983 Sex differences in performance on Raven's Progressive Matrices: A review. Alberta
28
Journal of Educational Research 29: 54-74.
Cowan, N.
1995 Attention and memory: an integrated framework. Oxford: Oxford University Press
Deary, Ian J., and Geoff Der
2005 Reaction time explains IQ�s association with death. Psychological Science 16: 64-9.
Engle, Randall W.
2002 Working memory capacity as executive attention. Current Directions in
Psychological Science 11: 19-23.
Engle, Randall W., Stephen W. Tuholski, J.E. Laughlin, and A.R.A. Conway
1999 Working memory, short-term memory and general fluid intelligence: A latent
variable approach. Journal of Experimental Psychology / General 130: 169-183.
Fiorentini, Adriana, Vittorio Porciatti, M. Concetta Morrone, and David C. Burr
1996 Visual ageing: unspecific decline of the responses to luminance and colour. Vision
Research 36: 3557-66.
Fotiou, Fotis, Konstantinos N. Fountoulakis, Apostolos Iacovides, and George Kaprinis
2003 Pattern-reversed visual evoked potentials in subtypes of major depression.
Psychiatry Research 118: 259-71.
Geary, David C.
2005 The Origin of Mind: Evolution of Brain, Cognition, and General Intelligence.
Washington, D.C.: American Psychological Association
Haier, R. J., K. H. Nuechterlein, E. Hazlett, J. C. Wu, and J. Paek
1988 Cortical glucose metabolic rate correlates of abstract reasoning and attention studies
with positron emission tomography. Intelligence 12: 199-217.
29
Hester, Robert L., Glynda J. Kinsella, and Ben Ong
2004 Effect of age on forward and backward span tasks. Journal of the International
Neuropsychological Society 10: 475-81.
Hill, K., and M. Hurtado
1996 Ache Life History: The Ecology and Demography of a Foraging People. New York:
Aldine de Gruyter, Inc.
Horn, J. L., and R.B. Cattell
1966 Refinement and test of the theory of fluid and crystallized intelligence. Journal of
Educational Psychology 57: 253-70.
Hurtado, A. Magdalena , Kim R. Hill, Wilhelm Rosenblatt, Jacquelyn Bender, and Tom
Scharmen
2003 Longitudinal Study of Tuberculosis Outcomes Among Immunologically Naive
Ache´ Natives of Paraguay. American Journal of Physical Anthropology 121:
134-50.
Jensen, A. R.
1980 Bias in mental testing. New York: Free Press
�
1998 The g factor. Westport: Praeger
Jensen, Arthur. R.
1999 The G Factor: the Science of Mental Ability. Psycoloquy 10.
Kandel, E. R., S. A. Siegelbaum, and J. L. Schwartz
1991 Synaptic transmission. In: E. R. Kandel, J. L. Schwarts and T. M. Jessel (eds.),
Principles of neural science; pp. 123-34. New Jersey: Prentice Hall
30
Kane, Michael J., David Z. Hambrick, Stephen W. Tuholski, Oliver Wilhelm, Tabitha W.
Payne, and Randall W. Engle
2004 The generality of working memory capacity: a latent-variable approach to verbal
and visuospatial memory span and reasoning. Journal of Experimental
Psychology / General 133: 189-217.
Kaplan, H., K. Hill, J. Lancaster, and A. M. Hurtado
2000 A theory of human life history evolution: Diet, intelligence, and longevity.
Evolutionary Anthropology 9: 156-184.
Kyllonen, P. C.
1996 Is working memory capacity Spearman�s g? In: I. Dennis and P. Tapsfield (eds.),
Human abilities: their nature and measurement; pp. 49-75. Mahwah, NJ: Erlbaum
Li, Shu-Chen, Ulman Lindenberger, Bernhard Hommel, Gisa Aschersleben, Wolfgang
Prinz, and Paul B. Baltes
2004 Transformations in the couplings among intellectual abilities and constituent
cognitive processes across the life span. Psychological Science 15: 155-63.
Lynn, Richard, and Paul Irwing
2004 Sex differences on the progressive matrices: A meta-analysis. Intelligence 32: 481-
98.
Martinelli, V., T. Locatelli, G. Comi, C. Lia , M. Alberoni, S. Bressi, M. Rovaris, M.
Franceschi, and N. Canal
1996 Pattern visual evoked potential mapping in Alzheimer's disease: correlations with
visuospatial impairment. Dementia 7: 63-8.
Measso, G., G. Zappala, F. Cavarzeran, T.H. Crook, L. Romani, F.J. Pirozzolo, F.
31
Grigoletto, L.A. Amaducci, D. Massari, and B.D. Lebowitz
1993 Raven's colored progressive matrices: a normative study of a random sample of
healthy adults. Acta Neurol Scand 88: 70-4.
Mitchell, K. W., J. W. Howe, and S. R. Spencer
1987 Visual evoked potentials in the older population: age and gender effects. Clinical
Physics and Physiological Measurement 8: 317-24.
Neisser, Ulric, Gwyneth Boodoo, Thomas J. Bouchard, A. Wade Boykin, Nathan Brody,
Stephen J. Ceci, Diane F. Halpern, John C. Loehlin, Robert Perloff, Robert J. Sternberg,
and Susana Urbina
1996 Intelligence: Knowns and unknowns. American Psychologist 51: 77-101.
Odom, J. Vernon, Michael Bach, Colin Barber, Mitchell Brigell, Michael F. Marmor,
Alma Patrizia Tormene, Graham E. Holder, and Vaegan
2004 Visual evoked potentials standard (2004). Documenta Ophthalmologica 108: 115-
23.
Porciatti, V., D. C. Burr, M. C. Morrone, and A. Fiorentini
1992 The effects of aging on the pattern electroretinogram and visual evoked potential in
humans. Vision Research 32: 1199-209.
Raven, J.C.
2000 The Raven's Progressive Matrices: change and stability over culture and time.
Cognitive Psychology 41: 1-48.
Raven, J.C., John H. Court, and J. Raven
1998 Manual for the Coloured Progressive Matrices (1998 edition). Oxford: Oxford