COMMENTARY
Effects of
early events
Stem cell
research guidelines
600
603
LETTERS I BOOKS I POLICY FORUM I EDUCATION FORUM I PERSPECTIVES
Studying Students in Montessori Schools
IN THEIR EDUCATION FORUM “EVALUATING MONTESSORI EDUCATION” (29 SEPT. 2006, P. 1893),
A. Lillard and N. Else-Quest present Montessori education as being superior to other types of
schools and having “remarkable outcomes.” Unfortunately, the analyses backing these value
judgments are plagued by methodological and statistical problems and thus do not provide support for such claims.
To evaluate Montessori education, Lillard and Else-Quest measured the performance of
children from a single Montessori school—a textbook example of pseudo-replication. The
only test performed was thus if this particular
school is a good school or not. But not even on
that level was the comparison entirely valid,
because the pupils in the school were compared
with children mainly attending inner-city public
schools (72% in 5-year-olds; 78% in 12-yearolds)—the category of schools most often subject
to social and economic problems.
Further, the sex ratio in the Montessori sample
was equal or skewed toward girls (50% girls in 5year-olds; 59% girls in 12-year-olds), while the Children using a computer at a Montessori
sex ratio in the inner-city public-school sample school.
was skewed toward boys (60% boys in 5-year-olds;
64% boys in 12-year-olds). Girls in these age groups generally outscore boys (1, 2). However,
not even with the study set up in this way did the Montessori school come out on top.
Pupils in the girl-dominated Montessori sample rather unsurprisingly scored higher on tests
measuring social ability than pupils in the boy-dominated inner-city school sample. But the
study also reported that although 5-year-olds in the Montessori sample did better in tests measuring cognitive/academic measures, no such difference could be found in the 12-year-olds.
The only really noteworthy result in this study is that for a sample consisting of mainly girls
in a particular school in Milwaukee, several years of Montessori education has resulted in no
detectable difference in academic performance in comparison with a group of mainly boys having been subject to schooling in Milwaukee inner-city public schools. Because alternative
schools can be assumed to attract more dedicated teachers and often have superior resources to
public schools (Montessori schools require a high teacher density and a special set of educational materials) and because young girls tend to outscore young boys academically, this is
more a quality statement for Milwaukee public schools than anything else.
To make informed choices about schooling, parents and policy-makers everywhere are in
dire need of proper comparisons between different education systems. Unintended misinformation through poorly performed studies only serves to make the current state of confusion
over the pros and cons of various education systems worse.
PATRIK LINDENFORS
Department of Zoology, Stockholm University, S-106 91 Stockholm, Sweden.
References
1. C. Dwyer, L. Johnson, in Gender and Fair Assessment, W. Willingham, N. Cole, Eds. (Erlbaum, Mahwah, NJ, 1997),
pp. 127–156.
2. D. R. Entwisle, K. L. Alexander, L. S. Olson, Children, Schools, and Inequality (Westview Press, Boulder, CO, 1997).
596
2 FEBRUARY 2007
VOL 315
SCIENCE
Published by AAAS
Downloaded from www.sciencemag.org on May 18, 2012
edited by Etta Kavanagh
IN THEIR EDUCATION FORUM “EVALUATING
Montessori education” (29 Sept., p. 1893), A.
Lillard and N. Else-Quest do not consider that
differential peer influences between their test
and control groups of students may contribute
to the differences they observed. The authors
controlled for parental effects by examining
only students whose parents had entered a
lottery for entry into a Montessori school.
However, the students who were unable to
attend the Montessori school because their
parents “lost” the lottery were dispersed to traditional schools, where they would have been
educated with a majority of peers whose parents did not enter the lottery at all. The differences they found in the academic and behavioral performance of students in Montessori
and traditional schools may not reflect
the superiority of the former educational
approach, but the negative effect of peer relationships in the latter.
PHILLIP MACKINNON
Faculty of Medicine, Nursing and Health Sciences, Monash
University, Clayton, VIC 3800, Australia.
Response
THE MONTESSORI SCHOOL WE STUDIED WAS
an inner-city public school [see our note (3)]
in the same school district as the other 27 city
public schools, was given the same per-pupil
funding as the district gives all its schools, and
received no other funds toward the educational program. Although it is a single school,
the traditional Montessori implementation,
regulated by the Association Montessori
Internationale (AMI), makes it very similar to
other AMI Montessori schools; I would be
wary of assuming similar results at non-AMI
Montessori schools. Still, replication in other
school districts is clearly desirable.
Gender did not contribute significantly to
any of the differences reported in the paper; in
all tests, we found significant differences
favoring the Montessori samples, but analyses by gender on these variables at those ages
revealed no differences between the boys and
girls. In general, gender differences tend to be
small enough that they often do not attain
significance in studies with small samples.
There were gender differences on several
tests at age 12, where gender is unbalanced,
but in the opposite direction for Montessori
www.sciencemag.org
CREDIT: NUBAR ALEXANIAN/CORBIS
LETTERS
Asymmetry from
symmetry
609
610
than the Letter presumes: Montessori boys at
age 12 performed best of all the groups on
those tests. In addition, the study was not set
up to have gender representation differences
across samples; the goal was equal representation, but the returned permission slips dictated the distributions.
Although the 12-year-olds had 6 more
years of Montessori schooling, and thus one
could argue should have shown stronger academic differences than the 5-year-olds, they
also had 6 more years of enculturation in their
low-income communities and homes that
might work against what happens inside the
school walls. Although the current environment emphasizes only academic tests, social
skills were included in this study, on the
grounds that they are extremely important in
life—often, one could argue, more important
than having superior academic skills.
Contrary to Lindenfors’s assumption, traditional Montessori actually has a rather low
teacher density. The AMI standard teacher-child
ratio is 1:28 to 1:35. Regardless, the notion that
teacher-child ratio is related to student achievement is misguided: Relations have sometimes
been found in first grade (1), but not consistently enough to claim it generally true (2).
Also, Montessori is not more expensive
than traditional schooling; once a classroom is
outfitted with materials (at a cost of roughly
$25,000), the costs per year are less (estimated at $800 per year by the North American
Montessori Teachers’ Association) than in
traditional classrooms because there are no
textbooks to replace. The notion that perpupil expenditures have an impact on student
achievement is also not upheld by the data (3).
MacKinnon makes the excellent suggestion that child outcomes at the Montessori
school might be the result of having peers
whose parents entered them in a lottery. One
way to investigate this is to see whether children in control schools that also admit by lottery (thus, their classmates might have listed
that school as their first choice) did better than
children at control schools that did not.
Milwaukee does not make obvious to an outsider which schools admit by lottery, but they
do indicate that Citywide Specialty Schools
are more likely to. The control children in our
study who were at Citywide Specialty Schools
did not as a group perform as well as the children at the Montessori school studied, suggesting that the results were not simply due to
schooling with peers whose parents entered
them in a lottery.
ANGELINE LILLARD1 AND NICOLE ELSE-QUEST2
1Department
of Psychology, University of Virginia,
Charlottesville, VA 22904, USA. 2Department of Psychology,
University of Wisconsin, Madison, WI 53202, USA.
References
1. NICHD Early Childcare Research Network, Dev. Psychol.
40, 651 (2004).
2. R. G. Ehrenberg, D. J. Brewer, A. Gamoran, J. D. Willms,
Psychol. Sci. Public Interest 2, 1 (2001).
3. E. Hanuschek, Educ. Eval. Policy Anal. 19, 141 (1997).
Darwin: Not the First to
Sketch a Tree
IN THE PERSPECTIVE BY A. ROKAS, “GENOMICS
and the tree of life” (29 Sept., p. 1897), the
caption to the accompanying photo identifies Charles Darwin’s 1868 notebook sketch
as “the first known sketch of an evolutionary
tree.” This is mistaken. Nearly 60 years earlier, in 1809, Jean Baptiste Lamarck presented an evolutionary tree of the animals in
Philosophie Zoologique (1). Modern evolutionary biologists would benefit from readTABLE
SHOWING THE ORIGIN OF THE VARIOUS ANIMALS.
Worms.
Infusorians.
Polyps.
Radiarians.
Insects.
Arachnids.
Crustaceans.
Annelids.
Cirrhipedes.
Molluscs.
Fishes.
Reptiles.
Birds.
Monotremes.
Amphibian Mammals.
Cetacean Mammals.
Ungulate Mammals.
Unguiculate Mammals.
Lamarck’s evolutionary tree [redrawn from (1)]
www.sciencemag.org
SCIENCE
Published by AAAS
VOL 315
ing Lamarck; there is much nonsense in his
work, but there are also quite substantial
insights. Lamarck towered over his contemporaries in both the rigor of his evolutionary
thought and the identification of a credible,
coherent, and falsifiable hypothesis for the
mechanism of evolutionary change. That
his mechanism was wrong does not take
anything away from the originality of his
thought (although his failure to test his
hypothesis is not consistent with current scientific practice and is largely responsible for
his current neglect).
MARK WHEELIS
Section of Microbiology, University of California, Davis,
Davis, CA 95616, USA. E-mail: mlwheelis@ucdavis.edu
Reference
1. J. B. Lamarck, Zoological Philosophy: An Exposition with
Regard to the Natural History of Animals, H. Elliot, transl.
(Macmillan, London, 1914), p. 179 [reprinted by the
University of Chicago Press, 1984].
Downloaded from www.sciencemag.org on May 18, 2012
X-ray diffraction
of bismuth
Human Dispersal
into Australasia
WE AGREE WITH P. MELLARS THAT LONGdistance dispersal of prehistoric human populations may involve multiple small founder
effects with consequent loss of genetic diversity—and possibly also cultural traits—with
increasing distance from source (“Going east:
new genetic and archaeological perspectives on
the modern human colonization of Eurasia,”
Special Section: Migration and Dispersal, Review, 11 Aug. 2006, p. 796). However, Australian genetic and stone artifact data are not
consistent with this pattern.
Genetic studies of the mitochondrial DNA
(mtDNA) of indigenous Australians have only
just begun, with fewer than 300 individuals
sampled from less than 5% of the continent’s
landmass (1–3). Even so, observed mtDNA
variation in Australians already includes five
Australia-specific haplogroups, and more if
New Guinea samples are added to represent the
landmass that was Sahul for most of the 50,000
to 60,000 years that humans have occupied the
island continent (4, 5). Larger, continent-wide
samples may well show Australia to have been
as genetically diverse as Europe or Africa.
The earliest Australian stone artifact industries consist of small flakes and flake tools
with some prepared platform cores and occasional radial or centripetal cores, and are
strongly influenced by the form of local stone
materials (6, 7). It is unlikely that these industries carry the sort of cultural information that
Mellars and others assume. His argument
that these are “Mode 4” (Upper Palaeolithic)
industries that devolved in response to local
conditions (8) lacks parsimony, given that
2 FEBRUARY 2007
597
LETTERS
MIKE A. SMITH,1 PAUL S. C. TACON,2
DARREN CURNOE,3 ALAN THORNE4
1National
Museum of Australia, GPO Box 1901, Canberra
ACT 2601, Australia. 2School of Arts, Griffith University, PMB
50, Gold Coast Mail Centre, QLD 9726, Australia. 3School of
Medical Sciences, University of New South Wales, Sydney,
NSW 2052, Australia. 4Archaeology and Natural History,
Research School of Pacific and Asian Studies, The Australian
National University, Canberra, ACT 0200, Australia.
References
1. K. Huoponen, T. G. Schurr, Y.-S. Chen, D. C. Wallace,
Hum. Immunol. 62, 954 (2001).
2. S. M. van Holst Pellekaan, M. Ingman, J, RobertsThomson, R. M. Harding, Am. J. Phys. Anthropol. 131,
282 (2006).
3. J. C. Presser, A. J. Deverell, A. Redd, M. Stoneking, Pap.
Proc. R. Soc. Tasmania 136, 35 (2002).
4. J. Friedlaender et al., Mol. Biol. Evol. 22, 1506 (2005).
5. M. J. Pierson et al., Mol. Biol. Evol. 23, 1966 (2006).
6. R. G. Roberts, R. Jones, M. A. Smith, Nature 345, 153
(1990).
7. W. Shawcross, Archaeol. Oceania 33, 183 (1998).
8. R. Foley, M. M. Lahr, Camb. Archaeol. J. 7, 3 (1997).
9. A. Brumm et al., Nature 441, 624 (2006).
10. P. J. Brantingham, A. I. Krivoshapkin, L. Jinzeng,
Y. Tserendagva, Curr. Anthropol. 42, 735 (2001).
11. M. Otte, A Derevianko, Antiquity 75, 44 (2001).
12. X. Gao, C. J. Norton, Antiquity 76, 397 (2002).
Response
I APPRECIATE SMITH ET AL.’S SUPPORT FOR
my general model of a progressive loss in the
genetic and technological diversity as modern human populations dispersed from their
African homeland to other parts of the world.
However, I find the remainder of their arguments unconvincing.
Despite the high genetic diversity of modern Australian and New Guinea populations,
the current genetic data indicate unambiguously that all of these populations derive ultimately from the two out-of-Africa mtDNA
lineages M and N (in turn derived directly
from the African L3 lineage) and from the
Y chromosome founder lineages C and F
(1–5). Any subsequent diversity in these populations must derive from genetic mutations
that occurred after the original out-of-Africa
dispersal around 50,000 to 60,000 years ago
[(1, 2, 4, 5); my Report].
I am equally unconvinced by their observations on the archaeological data. My own
598
CORRECTIONS AND CLARIFICATIONS
Reports: “Dual infection with HIV and malaria fuels the spread of both diseases in sub-Saharan Africa” by L. J. Abu-Raddad et al.
(8 Dec. 2006, p. 1603). The first sentence of the paper, “In Africa, an estimated 40 million people are infected with HIV,
resulting in an annual mortality of over 3 million (1), while over 500 million clinical Plasmodium falciparum infections occur
every year with more than a million malaria-associated deaths…” is incorrect. These numbers refer to worldwide numbers for
both infections, not just in Africa. The number of HIV-infected persons in Africa is approximately 25 million, and the number
of malaria infections is roughly 350 million.
Perspectives: “How fast does gold trickle out of volcanoes?” by C. A. Heinrich (13 Oct. 2006, p. 263). In line 7 of the first
full paragraph of column 3, “10 to 20 mg” should be “10 to 20 µg.”
TECHNICAL COMMENT ABSTRACTS
COMMENT ON “Rapid Advance of Spring Arrival Dates in Long-Distance
Migratory Birds”
Christiaan Both
Jonzén et al. (Reports, 30 June 2006, p. 1959) proposed that the rapid advance of spring migration dates of longdistance migrants throughout Europe reflects an evolutionary response to climate change. However, most migrants
should not advance their migration time because the phenology of their breeding grounds has not changed. It is more
likely that migration speed has changed in response to improved environmental circumstances.
Full text at www.sciencemag.org/cgi/content/full/315/5812/598b
RESPONSE TO COMMENT ON “Rapid Advance of Spring Arrival Dates in LongDistance Migratory Birds”
Niclas Jonzén, Andreas Lindén, Torbjørn Ergon, Endre Knudsen, Jon Olav Vik, Diego Rubolini,
Dario Piacentini, Christian Brinch, Fernando Spina, Lennart Karlsson, Martin Stervander, Arne
Andersson, Jonas Waldenström, Aleksi Lehikoinen, Erik Edvardsen, Rune Solvang, Nils Chr.
Stenseth
Both’s comment questions our suggestion that the advanced spring arrival time of long-distance migratory birds in
Scandinavia and the Mediterranean may reflect a climate-driven evolutionary change. We present additional arguments
to support our hypothesis but underscore the importance of additional studies involving direct tests of evolutionary
change.
Full text at www.sciencemag.org/cgi/content/full/315/5812/598c
impression is that the earliest Australian
technologies are far too simple, generalized,
and “expedient” (as they seem to accept) to
support any specific technological links with
other technologically simple and expedient
technologies, such as those from Flores and
other earlier Pleistocene sites in southeast
Asia (my Report). I am not aware of any convincing Levallois or other “Mode 3” (Middle
Palaeolithic) technologies in Australia and
see no reason why the Australian technologies should not be viewed as heavily
simplified or “devolved” forms of Upper
Palaeolithic (“Mode 4”) technologies, under
PAUL MELLARS
Letters to the Editor
Letters (~300 words) discuss material published
in Science in the previous 6 months or issues of
general interest. They can be submitted through
the Web (www.submit2science.org) or by regular
mail (1200 New York Ave., NW, Washington, DC
20005, USA). Letters are not acknowledged upon
receipt, nor are authors generally consulted before
publication. Whether published in full or in part,
letters are subject to editing for clarity and space.
2 FEBRUARY 2007
VOL 315
SCIENCE
Published by AAAS
the influence of varying raw material effects
and other purely local economic adaptations
(my Report). I note that they make no reference to the apparent similarities between the
forms of the early Australian “horse-hoof ”
cores and simple forms of single-platform
blade cores (my Report). And I am totally
unconvinced by the arguments for purely
local origins of Upper Palaeolithic/Mode 4
technologies in northeast and Central
Asia [my Report; (6)]. To employ these
data to support some form of multiregional,
as opposed to African, origins for modern
Australian populations would seem to be
poorly founded in either the genetic or archaeological data.
Department of Archaeology, Cambridge University,
Cambridge CB2 3DZ, UK.
References
1. P. Forster, Philos. Trans. R. Soc. London B 359, 255
(2004).
2. P. Forster, personal communication.
3. T. Kivisild et al., Genetics 172, 373 (2006).
4. P. Mellars, Proc. Natl. Acad. Sci. U.S.A. 103, 9381
(2006).
5. M. Richards, H.-J. Bandelt, T. Kivisild, S. Oppenheimer,
Nucleic Acids Mol. Biol. 18, 223 (2006).
6. P. Mellars, Evol. Anthropol. 15, 167 (2006).
www.sciencemag.org
Downloaded from www.sciencemag.org on May 18, 2012
similar assemblages are found in middle Pleistocene and other early contexts in neighboring
parts of Southeast Asia (9). Also, the balance of evidence indicates only a weak relationship between Mode 4 industries and the
spread of modern humans (8). In northeast
and central Asia, these industries appear to
have locally developed (10) and subsequently
diffused into Europe and the Near East, rather
than originating in Africa (11, 12).
Collectively, these data suggest that the
cultural and genetic history of Australasia is
more complex than a single dispersal model
such as “Out-of-Africa 2” allows.