ArticlePDF Available

Problems in deriving Italian regional differences in intelligence from 2009 PISA data

Authors:

Figures

Content may be subject to copyright.
Running head: NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES
The Version of Scholarly Record of this Article is published in INTELLIGENCE (2013),
available online at: http://dx.doi.org/10.1016/j.intell.2012.10.004. Note that this article may not
exactly replicate the final version published in INTELLIGENCE.
Cornoldi, C., Giofrè, D., & Martini, A. (2013). Problems in deriving Italian regional differences in
intelligence from 2009 PISA data. Intelligence, 41, 2533. doi:10.1016/j.intell.2012.10.004
Problems in deriving Italian Regional Differences in Intelligence
from 2009 PISA data
Cesare Cornoldi*, David Giofrè*, & Angela Martini**
* Department of General Psychology, University of Padova, Italy;
** INVALSI, Frascati-Roma, Italy
Correspondence concerning this article should be addressed to:
David Giofrè,
Department of General Psychology,
University of Padova,
via Venezia, 8, 35131, Padova, Italy.
E-mail: david.giofre@gmail.com
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 2
Abstract
Recent results of international assessment programs (e.g., PISA) have shown a large
difference in high school students performance between northern and southern Italy. On this
basis, it has been argued that the discrepancy reflects differences in average intelligence of
the inhabitants of regions and is associated with genetic factors (Lynn, 2010a; 2012). This
paper provides evidence in contrast to this conclusion by arguing that the use of PISA data to
make inferences about regional differences in intelligence is questionable, and in any case,
both PISA and other recent surveys on achievement of North and South Italy students offer
some results that do not support Lynn's conclusions. In particular, a 2006-2009 PISA data
comparison shows a relevant decrease in the North-South difference in only three years,
particularly evident in the case of a single region (Apulia). Other large surveys (including
INVALSI-2011) offer different results; age differences suggest that schooling could have an
important role.
Keywords: International Assessment Programs, intelligence, Educational
achievement, IQ regional differences, Group differences,
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 3
1. Introduction
Even though cognitive ability and academic achievement are distinct constructs and
specific cognitive factors are important to explain specific aspects of achievementnot only
the general factor (Kaufman, Reynolds, Liu, Kaufman, & McGrew, 2012)it is
unquestionable that measures of reading comprehension and mathematical achievement offer
good approximations of the individual's intelligence levels. In fact, the linguistic, reasoning,
working memory and attentional processes that underlie reading and mathematical operations
also underlie intellectual functioning (Deary, Strand, Smith, & Fernandes, 2007; Hunt, 2011).
The relationship is also supported by empirical evidence: Studies have found a good
correlation between achievement tests (like SAT and ACT) and a g-factor measure, and these
results are consistent because correlations are high (typically between .6 and .7) (Coyle &
Pillow, 2008; Frey & Detterman, 2004; Koenig, Frey, & Detterman, 2008). Therefore, using
achievement measures to derive IQ estimations is appropriate. As a consequence, some
researchers have studied regional differences in IQ by taking advantage of the outcomes of
the international assessment projects that have administered the same achievement tests in
different countries (Rinderman, 2007).
Along this line of research, the comparison of the IQ of youngsters living in northern
versus southern Italy has been seriously studied by international scholars, and the results have
also been discussed in the popular Italian media. In particular, an influential and discussed
study by Lynn (2010a) examined achievement scores obtained by southern and northern Italy
students in the PISA2006 (Project for the International Assessment of Achievement) of
students aged 15 (OECD, 2007) and associated the low scores obtained by southern Italy
students with low intelligence levels. The study produced a series of other studies offering
opposing arguments. In particular, Cornoldi, Belacchi, Giofrè, Martini, and Tressoldi (2010)
reconsidered the results of the PISA2006 survey, which had been the basis for Lynn's
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 4
conclusion, and other achievement studies and argued that North-South differences were not
as clear as Lynn assumed (2010a). Beraldo (2010) raised methodological concerns while
Felice and Giugliano (2011) stressed the relevance of socio-cultural factors. However, Lynn
disputed the points raised by these studies (2010b; 2012). In particular, Lynn (2012)
examined the achievement data obtained in the most recent PISA survey (OECD, 2010a) and
offered counterarguments in favour of his thesis. In a latter paper, in agreement with the large
body of evidence (e.g., Dick et al., 2007) showing the genetic bases of intelligence, Lynn also
considered genetic differences between people living in northern versus southern Italy,
further stressing the assumptions that there are strong differences in intelligence between
them and that these differences are inherited. The issue was also examined by Templer
(2012) who offered important data showing that both biological and social variables
differentiating Northern and Southern Italy may explain the differences found in
achievement. In the meantime, other papers were published on these issues. D'Amico,
Cardaci, Di Nuovo, and Naglieri (2012) showed that regional differences may disappear
using other intelligence testing procedures, and Robinson, Saggino, and Tommasi (2011), on
the basis of different sources of information (obtained from INVALSI; Istituto Nazionale per
la VALutazione del Sistema di Istruzione e di Formazione; National Institute for the
Assessment of the Instruction System), showed that the achievement of southern Italy
students may even be higher than that of northern Italy students.
In sum, the case of regional differences in Italy offers elements for the general
discussion on ethnic differences in intelligence and its heritability versus modifiability by
education. In fact, according to some authors (e.g., Ceci, 1991; Ceci & Williams, 1997),
education and other environmental factors have substantial effects on IQ and academic
achievement, and increments in school attendance convey significant increments in
intelligence. For example, a recent study indicates that two extra years of schooling beyond
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 5
seventh grade have relevant effects on IQ above and beyond the Flynn effect, and the effect is
substantial for students who are 19 years old (Brinch & Galloway, 2012). Nevertheless, since
the appearance (1966) of the famous Coleman report, other researchers emphasized the role
of IQ in self-selection into educational levels and provided support for the limited
malleability of IQ by schooling and/or training (Herrnstein & Murray, 1994). Similarly, Lynn
(2010a; 2012) argued that people from southern Italy have lower incomes and school levels
because they are less intelligent and thus are less able to create favourable socioeconomic
conditions for themselves.
At that point, the different theses could seem unfalsifiable and further studies
comparing North and South Italy unproductive. Nevertheless, we think that reconsidering this
point may have general implications for the debate on ethnic differences in intelligence
(Hunt, 2011) and on the use of international data on achievement and thus can take advantage
of the specific Italian case, for which more than a single source of evidence is available. In
this paper, on the basis of the Italian data, we will show that i) it is risky to use PISA data to
make inferences about the population's intelligence; ii) PISA 2009 data, if deeply analyzed
and compared with the PISA 2006 data, offers a different picture than that derived by an
overall North-South comparison; and iii) the outcomes from different sources of information
about the achievement of Italian children offer different descriptions of the competencies of
northern and southern Italy students.
2. Limitations of the PISA Data for the International Debate on Intelligence
The PISA project is designed to evaluate education systems by testing the skills and
knowledge of 15-year-old students in participating countries/economies. It has been argued
that these measures are reliable and a good proxy of intelligence (e.g., Rindermann, 2007;
2008). Therefore, the use of Pisa data may be ambiguous because it may be made both for
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 6
assessing the efficiency of teaching and for deriving general ability measures. However, it
must be taken into account that PISA studies originated for the need of educational
assessment across countries and there is only clear evidence supporting this use. In fact,
evidence supports the use of PISA in the context of national comparisons. For example, the
results of PISA are highly correlated with the results of other achievement examinations (e.g.,
Trends in International Mathematics and Science Study [TIMSS], or Progress in International
Reading Literacy Study [PIRLS]) (INVALSI, 2008a, 2008b).
PISA results also correlate with measures of intelligence (Lynn & Meisenberg, 2010;
Rindermann, 2007). However, this evidence is open to criticisms. For example, according to
Wicherts and Wilhelm (2007), this conclusion was based on aggregated-level analyses of
correlations between means and cannot necessarily be interpreted at the level of individuals.
In fact, in the case of PISA, data were collected to obtain information not about individual
intellectual abilities but about groups. Furthermore data concerned academic achievement
measures that, in a homogeneous population, may be highly related with ability measures, but
in different populations and school systems may reflect educational systems results, which, in
the case of disadvantaged systems, may be substantially improved, even of 1 standard
deviation (Clarke, Snowling, Truelove, & Hulme, 2010) when appropriate teaching is
introduced. The same goals reported in PISA documents specify that PISA is mainly intended
to measure a contingent and modifiable efficiency of school systems: The design of PISA
does not just allow for a comparison of the relative standing of countries in terms of their
learning outcomes; it also enables each country to monitor changes in those outcomes over
time. Such changes indicate how successful education systems have been in developing the
knowledge and skills of 15-year-olds. (OECDb, 2010, p.13)
The fact that the main goal of PISA is to assess the efficiency of the school system,
not to make comparisons across individuals, is confirmed by the decision that participants
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 7
must receive different tests. This is justified on the basis of the item response theory, but it
makes comparisons difficult.
The outcomes of different programs assessing achievement seem only moderately
correlated, and the correlations may be lower when intelligence and achievement scores are
correlated (Baumert, Lüdtke, Trautwein, & Brunner, 2009; Kaufman et al., 2012). Therefore,
examining the sources used for deriving the intelligence scoreswhich were correlated with
achievementis crucial. To our knowledge, these scores were mainly taken from Lynn-
Vanhanen's database (2006), which offers useful preliminary information but also has many
limitations (Hunt, 2011; Wicherts, Dolan, & van Der Maas, 2010).
As we have already argued, if PISA mainly assesses the quality of two school systems
and the quality is dramatically different, then this difference may create confusion in the
consideration of individual rather than school outcomes. Obviously, it can be argued (Lynn,
2010a) that the quality of the school system is related to the wealth of a region and that they
are both the consequence rather than the cause of differences in achievement and intelligence.
However, the opposite explanation is also legitimate. It can be argued that better school
systems produce higher achievement levels because they usually provide a more favourable
environment for fully achieving the students' potential. In fact, the quality of the school
system (e.g., quality of teaching) has an important impact on academic success and academic
achievement (Chetty et al., 2010; Rindermann & Ceci, 2009). In fact, some achievement
effects can be attributed to factors other than intellectual gains, as stressed by Felice and
Giugliano (2011) in their examination of the differences between North and South Italy.
Checchi and Jappelli (2004), for instance, reported a quality score of public schools by
regions (both as perceived by parents and as measured by indicators of school resources).
They found a substantial regional variation and that the school quality is considerably lower
in the South. By using aggregate indicators, they determined that the public schools (primary,
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 8
lower secondary, and upper secondary) in the South are lower in quality. For example, they
reported that the proportion of students in double shifts due to school congestion is virtually
nil in northern Italy, but may range between 6% and 14% in southern areas.
Recent data, independent of data based on achievement scores, confirms that northern
Italy has a better functioning scholastic system (and in general has better public services, e.g.,
a better health care system) compared to southern Italy (Agasisti & Cordero, 2010). In
particular, the difference among Italian regions in educational resources is relevant: Some
regions suffer a shortage in resource quality (e.g., Sardinia) while others report values well
above the OECD average (e.g., Lombardy) with a difference that can reach the value of .68
standardized points in the case of the WLE index (SCMATEDU from the PISA
Questionnaire), which described the quality of educational resources. This index explained
the 9.53% of the variance in the Science performance of Italian students (Agasisti, 2011;
Agasisti & Cordero, 2010).
3. General Increase in Achievement of southern Italy Students
The observation of rapid changes in achievement and intelligence may help to clarify
the impact of the quality of the school system on the level of achievement. On the basis of the
Flynn effect (2009a; 2009b) and the assumption of a genetic basis of intelligence, one could
predict that passing from one PISA administration to the following one, students' intellectual
performances would slightly increase (but as the scores are standardized, they remain
identical), and regional differences would be substantially preserved or would require long
periods of time for minor changes. On the contrary, if we assume that results in achievement
reflect the contingent and modifiable quality of a school system, we can explain rapid
changes in achievement outcomes.
If we compare the scores obtained in different assessment procedures, we can see that
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 9
the size of the gap between different populations may be reduced very rapidly. For example,
Nisbett et al. (2012) stated that the gap between whites and blacks diminished by 0.33 SD in
recent years (but see, for a different conclusion, Rushton and Jensen, 2005). Similarly, the
gap between northern and southern Italy children has become substantially smaller in the last
few years (Table 1). In the last decade, the public national Institute INVALSI has moved
from rough assessment procedures (which had offered the excessively positive description of
southern Italy schools, used by Robinson et al. (2011) to more systematic and better
controlled studies that showed the poor performances of southern Italian regions and
motivated some of them to invest more resources in the education, and the growth in PISA
performance can be attributed to this. Concerning the variation in performance from PISA
2006 to PISA 2009, Table 1 clearly shows that the southern Italy performance in PISA
increased whereas the northern Italy performance remained similar.
Table 1 about here
This pattern is consistent across all the PISA areas (i.e., reading, math, and science)
and is significant. Dividing the difference between the performances of different PISA
administrations by 100 (the population's standard deviation) produces a measure of the
growth expressed in SDs. The overall improvement in the last years for the south and islands
is substantial. Data collected by OECD in 2009 (OECD, 2010b) allows a comparison for the
different regions of Italy for PISA 2006 and 2009. As result, it is possible to establish the
change in performance of the Italian regions participating in both PISA 2006 and PISA 2009.
Table 1 presents a summary of the changes for the different areas of Italy. As can be seen, the
Northwest also changed, but the change was more dramatic in South Italy. We derived a
comparison from the original data that focused on the contrast between the most northern and
most southern Italy regions for which both 2006 and 2009 data were available by treating
different regions as subjects. For northern Italy, we included seven regions (Friuli, Trentino,
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 10
Lombardy, Veneto, Piedmont, Emilia-Romagna, and Liguria); for southern Italy, we included
five regions (Basilicata, Campania, Apulia, Sardinia, and Sicily). The area of Bolzano was
excluded even if it further supported the hypothesis of a reduction of the North-South gap
because its substantial drop is due to the local professional schools (mainly attended by low
achievement students) not being considered in the 2006 survey.
The pattern of results is robust across different observations. All the effects are
significant, and the effect size is high (Cohen, 1988). In fact, 2 × 2 ANOVAs (year [Pisa
2006 and Pisa 2009]) × geographical area [North, South]) on the scores showed a significant
interaction between year and geographical area, with large effect sizes and important
decreases in the differences in score , either for the overall score, ηp²= .45 (Figure 1), or for
the specific scores in reading, ηp²= .54, mathematics, ηp²= .44, and science, ηp²= .37. The
decreases in the score differences between North and South, computed on the basis of the
National 2009 standard deviations (100), were of .22, .18, .21, and .19 SDs respectively for
the overall score, reading, mathematics and science. Furthermore, Bonferroni's post-hoc
comparisons showed for the northern Italy regions no significant difference between PISA
2006 and Pisa 2009 in the four scores; on the contrary, for the southern Italy regions, there
was a significant improvement in all the four scores of .27, .21, .30, and .21 SDs respectively
(p ranging between .036 for Science and .003 for Mathematics).
Figure 1 about here
Using the PISA databank, we could also explore whether relevant school factors
changed in different ways between North and South Italy between 2006 and 2009. To this
purpose, we considered the following indexes: SC14Q04=Shortage qualified teachers;
SC14Q07=Shortage science lab equipment; SC14Q08=Shortage instruct material;
SC14Q09=Shortage computers; SC14Q10=Shortage Internet; SC14Q11=Shortage computer
software; SC14Q12=Shortage library materials; SC14Q13=Shortage audio-visual;
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 11
IRATCOMP= index of availability of computers; and SCMATEDU=Quality of educational
resources. However, we did not find in this short period changes in the differences between
North and South that could explain the change in the achievement differences but rather only
some paradoxical effects due to increases in the complaints about the availability of
educational resources by southern schools that could be also interpreted as signals of an
increased sensibility to the importance of them.
The presence of non-native students may affect the performance (typically of about 8
points in the PISA Italian sample), and this could have lowered the scores of northern Italy
where migrants are more frequent and also may have been partly responsible for the
variations of PISA scores in the last years. However, using the PISA databank, we could
conclude that this effect should have been similar in 2006 and 2009 as the percentages of
native students did not strongly vary. In fact, in 2006, the percentages of native students of
the PISA sample were 92.13 for the North and 95.61at the South. In 2009, the percentages
only changed slightly and were 90.53 and 97.58, respectively, a variation that could only
explain a variation of less than .3 points in the PISA scores (the migrants who moved to the
North could be a particularly intelligent group [Lynn, 2006]).
The improvement of southern Italy students observed by PISA 2009 is not isolated as
it had already been anticipated by another international survey concerning literacy (i.e.,
Progress in International Reading Literacy Study (PIRLS); Mullis, Martin, Kennedy, & Foy,
2007) (Table 2) comparing fourth grade students in 2001 and 2006, respectively (INVALSI,
2008a, 2008b). The comparison shows large differences between northern and southern Italy
regions (about 27 points) in 2001, and smaller differences in 2006 (about 9 points).
Nevertheless, the reduction of the North-South Italy gap was not evident in another
assessment that considered participants within an age range similar to PISA (Trends in
International Mathematics and Science Study (TIMSS); Martin, et al., 2008; Mullis et al.,
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 12
2008). Table 3 offers an overall view of the standardized differences observed between
northern and southern Italy students, showing that the variations in the different studies are so
high to legitimate radically different conclusions, suggesting that further evidence is needed
to reach unquestionable conclusions.
Tables 2 and 3 about here
4. The Case of Apulia
Apulia is an example of how contextual factors may dramatically affect PISA
outcomes. In the Apulia region, there was an enormous growth in the performance in the last
PISA survey (Table 4). The overall change from PISA 2006 to PISA 2009 for the Apulia is
impressive (about 48 points), and if the PISA can be used as a measure of intelligence, it is
equivalent to about 7 IQ points. Moreover, Apulia's performance is now more similar to the
northern Italy regions than to the southern ones (Table 4). This finding cannot be explained
on a genetic basis in such a short period of time nor as a simple case of regression towards
the mean, as it was specific of Apulia and also partly predictable. However the result cannot
be explained with the improvements in the quality of the school system, which could not
easily produce such substantial changes in such a short period.
Table 4 about here
In our view, the impressive improvement of Apulia can be attributed to a specific
program that the region initiated using the European Union's Social Fund for the development
of poor areas and the improvement of the quality of the schools. Four Italian regions obtained
funds (Campania, Calabria, Apulia, and Sicily), and in the case of Apulia, a large portion was
spent on improving achievement and the ability to take achievement tests (by training
teachers, providing additional resources, and sensitizing on the importance of national
examinations) (Rubinacci, 2011). In fact, in 2006, Lucrezia Stellacci was nominated as the
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 13
new Director for the Regional School System (Ufficio Scolastico Regionale) of Apulia and
decided to use the European funds to improve not only the regional school system (quality of
teaching, buildings, labs) but also to expand assessment practices in Apulia schools and
develop test taking skills in its students (Rubinacci, 2011). The dramatic improvement in the
achievement of Apulia students seems therefore to be due to improvements both in the school
system and in the test taking skills (Martini, 2011).
The importance of test taking skills and the difficulties that southern Italy students
have with group administration of written tests is further supported by the curious
observation of Cornoldi et al. (2010) that some regional differences currently present in the
results of group assessments may disappear if assessment is individual and interactive.
Consequently, we have re-examined the mean scores reported by Cornoldi et al. (2010) for
the reading and mathematics tests administered in groups versus individually. We have
transformed them to z-scores based on normative data and then computed the difference in
the mean z-score obtained by North and South Italy. We have found that a difference between
North and South in the group average scores present both in 9th- and 10th-graders (.32 and
.50, respectively) completely disappears in the case of the average scores obtained at the
individual testing (.02 and -.08, respectively). It is true that group testing concerned tasks
(reading comprehension and mathematical reasoning) that have a higher relationship to IQ
than do the tasks that were individually administered (reading decoding and calculation),
butbased on the assumption of strong regional differencesthe differences had to be
reduced and not eliminated.
5. Outcomes of the 2011 INVALSI Survey
The data collected by the official public Italian Institute charged with gathering data
on achievement of Italian students INVALSI (Istitituto Nazionale per la VALutazione del
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 14
Sistema di Istruzione e di Formazione; National Institute for the Assessment of the
Instruction System) appear particularly critical and authoritative and are typically used also in
international studies. For example, Lynn (2012) relied on recent INVALSI (2011) data to find
further evidence on Italian regional differences.
INVALSI data is consistent and reliable and open to international scholars. The 2011
survey included about 40,000 students for each grade that was involved (2nd, 5th, 6th, 8th,
and 10th) who were representative of the Italian population. Moreover, data is corrected to
control for cheating, and an external INVALSI examiner is present during the examination
(with the exception of the 8th grade, in which the test is taken in the context of a diploma
examination). Considering the final report for 2011 data (INVALSI, 2011), Lynn (2012)
found further support for the North-South differences, showing that there is a significant
difference between northern and southern Italy regions. However, the official INVALSI
report only offers mean values and interval confidences. Therefore, due to the large sample
size, it is possible that a significant effect is detected even if it is very small and does not
reflect large differences between groups (Cohen, 1988). For this reason, we took advantage of
the original INVALSI databank. We computed the SDs and were able to calculate the size of
the difference (expressed in terms of mean differences by the Italian standard deviations)
(Table 5; we excluded the 8th grade because there was no an external examiner during the
evaluation). From Table 5, we can see that if we compare the mean scores of northern and
southern Italy regions, North-South differences are relatively small and lower than in the
PISA scores, ranging between -0.02 and 0.32 (M = 0.15).
Table 5 about here
6. Age Changes
Despite the fact that intuitively experience and schooling should have had a greater
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 15
impact after many years of exposure than after fewer years and then have a higher influence
on older children, inherited differences emerge more clearly in older children than in younger
children (Petrill et al., 2004). Therefore, the hypothesis of an inherited nature of Italian
regional differences finds further support from the observation that achievement scores
obtained by southern Italy children decrease with development. In fact, international
assessment projects have even reported no regional differences in Italy for 4th graders, as
occurred for PIRLS (INVALSI, 2008a), but substantial differences for older children.
Furthermore, according to TIMSS data (INVALSI, 2008b), the southern Italy regions have a
drop in performance of about 40 points (roughly .4 SD) from the 4th to the 8th grade. A
similar pattern is apparently present for the recent INVALSI survey (Table 5).
But there are some inconsistencies in the data. First, if the drop in the achievement
test was due to genetic factors, then the drop observed by the TIMSS project would be
limited to South Italy only. Instead, there was also a drop in performance in the North in
some achievement tasksabout 20 points for math and 30 points for science (INVALSI,
2008b). Second, data on the increased North-South Italy gap is partly contradictory.
INVALSI (2011) data show a large difference in the 10th grade, but the size of the difference
is also very large between the 5th and 6th grades (Table 5). This result is inconsistent with the
assumption of a gradual increase with age in the role of genetic factors but is consistent with
the shared opinion that in Italy, especially South Italy, the quality of the secondary school for
the 6th to 8th grades is poor (Ferrer-Esteban, 2011). In fact, in Italy, the transition from the
5th to the 6th grade corresponds to a dramatic change in the type of school, and the new
educational system is poor in many respects (e.g., teachers without any specific pedagogical
and psychological training, higher emphasis on the learning of notions than on general
abilities). This is in line with the fact that Italian teachers of this school system (6th to 8th
grades) are characterized by a low self-perceived competence and by a high mean age
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 16
(around 52), the highest in the OCSE countries (Ferrer-Esteban, 2011).
7. Discussion
The hypothesis that relevant differences in intelligence between different populations
can be found and that these differences are inherited has a long history in psychology and has
found fresh support on the basis of international assessment projects that have administered
exactly the same procedures to different populations. In particular, the PISA program has
offered an impressive amount of data that can be used to respond to crucial questions about
human abilities. Lynns (2010a; 2012) focus on the comparison between North and South
Italy had the advantages of comparing two populations that were administered exactly the
same tasks in the same language and with the same school legislation. The two populations
are apparently significantly different not only on a historical but also on a biological basis as
South Italy is close to African communities and immigrations and North Italy has close
exchanges with Central Europe. Therefore, the comparison between North and South Italy
represents not only an important issue per se but also a crucial case for examining the more
general issue on the existence of ethnic differences in intelligence and the possibility of using
international assessment projects to assess not only the quality of school systems as the
explicit goals of these projects state, but also intelligence.
In this paper, we have offered arguments in favour of two main points. First, the use
of PISA data to make inferences about regional differences in achievement and intelligence
levels raises a series of problems. Second, even if we use PISA data, the North-South Italy
differences are not as clear as they may seem at first glance. Concerning the first point, in our
opinion, PISA data must be used cautiously because it is collected in a particular way that
permits the assessment of how students of different school systems are able to perform on
achievement tests but does not offer direct and reliable measures of the abilities of single
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 17
individuals. The cases of Apulia and of group versus individual testing show that particular
conditions may affect the levels of performance. The partially different outcomes of different
assessment projects (ranging from no difference in individual assessment to a difference of
more than half a standardized point, roughly corresponding to a difference in IQ of 78
points in the PISA scores) further support the hypothesis that PISA data is also affected by
contingent factors. Furthermore, even when basing conclusions on PISA scores, the
deductions that we can derive do not favor strong genetic differences between North and
South Italy.
With the present paper we did not intend to question the general issue of heritability
of intelligence, as, in our view, the evidence supporting the genetic bases of intelligence is
robust and unquestionable, but we wanted to put in evidence the fact that measures related
with intelligence may be affected by a series of other factors. As Hunt (2012) commented,
national indicators of intelligence are markers of national differences in the ability to use the
cognitive artifacts (i.e., physical instruments or styles of reasoning that amplify our ability to
think); further, variations in national capabilities to use cognitive artifacts can be attributed to
differences in the extent to which different nations provide techniques and institutions for the
development of individual cognition. In particular, our criticism concerns the conclusion that
regional differences in academic measures related with intelligence are due to genetic factors.
Further, a genetic hypothesis cannot explain why in a limited number of years, the North
South difference dropped substantially. A genetic hypothesis may also have difficulty in
explaining some variations in age in achievement differences, which are impressively related
to the shift from primary to secondary school, a type of school system that in Italy appears
particularly weak.
We are aware that the nature versus nurture controversy in intelligence cannot find a
complete solution because the arguments in favor of a position can be reversed and used in
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 18
favor of the opposite position, and the present evidence is not decisive against Lynn's theory.
However, we think that the mediating role of the environment should be better emphasized
when examining Italian regional differences, as also suggested for other populations (Barsky,
Semin, & Malykh, 2011; Molenaar, van der Sluis, Boomsma, & Dolan, 2011; Rodic et al.,
2011). Therefore, to examine Italian regional differences in intelligence, if they really exist, it
should be necessary to make assumptions concerning the degree to which there are additional
achievement specific population differences that may bias the estimate of the mean IQs.
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 19
References
Agasisti, T. (2011). The efficiency of Italian secondary schools and the potential role of
competition: a data envelopment analysis using OECDPISA2006 data. Education
Economics, (931809330), 1-25. doi:10.1080/09645292.2010.511840
Agasisti, T. & Cordero, J. M. (2010). Which factors affect students’ performances?Exploring
differences among Regions in Italy and Spain by using PISA2006 data and multilevel
modelling. Paper prepared for the First Lisbon Research Workshop on Economics and
Econometrics of Education. Lisbon, Portugal.
Barsky, P., Semin, A., &Malykh, S. (2011, December). The heritability of verbal and
nonverbal intelligence in Russian adolescents. Paper presented at the 12th annual
conference of ISIR, Limassol, Cyprus.
Baumert, J., Lüdtke, O., Trautwein, U., & Brunner, M. (2009). Large-scale student
assessment studies measure the results of processes of knowledge acquisition: Evidence
in support of the distinction between intelligence and student achievement?. Educational
Research, 4, 165-176. doi:10.1016/j.edurev.2009.04.002
Beraldo, S. (2010). Do differences in IQ predict Italian northsouth differences in income? A
methodological critique to Lynn. Intelligence, 38, 456-461.
doi:10.1016/j.intell.2010.06.007
Brinch, C. N., & Galloway, T. A. (2012). Schooling in adolescence raises IQ scores.
Proceedings of the National Academy of Sciences of the United States of America, 109,
425-30. doi:10.1073/pnas.1106077109
Ceci, S. J. (1991). How much does schooling influence general intelligence and its cognitive
components? A reassessment of the evidence. Developmental Psychology, 27, 703-722.
doi:10.1037/0012-1649.27.5.703
Ceci, S. J., & Williams, W. M. (1997). Schooling, intelligence, and income. American
Psychologist, 52, 1051-1058. doi:10.1037/0003-066X.52.10.1051
Checchi, D., & Jappelli, T. (2004). School Choice and Quality. CEPR Discussion Paper No.
4748. Retrieved from: http://ssrn.com/abstract=654564
Chetty, R., Friedman, J. N., Hilger, N., Saez, E., Schanzenbach, D. H., &Yagan, D. (2010).
How does your kindergarten classroom affect your earnings? Evidence from Project
Star. Cambridge, MA: National Bureau of Economic Research.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale,
NJ: Lawrence ErlbaumAssociates
Coleman, J. S., Campbell, E. Q., Hobson, C. J., McPartland, F., Mood, A. M., Weinfeld, F.
D., et al. (1966). Equality of educational opportunity. Washington, DC: U.S.
Government Printing Office
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 20
Cornoldi, C., Belacchi, C., Giofrè, D., Martini, A., & Tressoldi, P. (2010). The mean
Southern Italian children IQ is not particularly low: A reply to R. Lynn (2010).
Intelligence, 38, 462-470. doi:10.1016/j.intell.2010.06.003
Coyle, T. R., &Pillow, D. R. (2008). SAT and ACT predict college GPA after removing g.
Intelligence, 36, 719-729. doi:10.1016/j.intell.2008.05.001
D'Amico, A., Cardaci, M., Di Nuovo, S., & Naglieri, J. A. (2012). Differences in
achievement not in intelligence in the north and south of Italy: Comments on. Learning
and Individual Differences, 22, 128-132. doi:10.1016/j.lindif.2011.11.011
Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence and educational
achievement. Intelligence, 35, 13-21. doi:10.1016/j.intell.2006.02.001
Dick, D. M., Aliev, F., Kramer, J., Wang, J. C., Hinrichs, A., Bertelsen, S., Kuperman, S., et
al. (2007). Association of CHRM2 with IQ: converging evidence for a gene influencing
intelligence. Behavior Genetics, 37, 265-72. doi:10.1007/s10519-006-9131-2
Felice, E., & Giugliano, F. (2011). Myth and reality: A response to Lynn on the determinants
of Italy’s North–South imbalances. Intelligence, 39, 1-6.
doi:10.1016/j.intell.2010.09.004
Ferrer-Esteban, G. (2011). Beyond the traditional territorial divide in the Italian Education
System. Effects of system management factors on performance in lower secondary
school. Retrieved from: www.fga.it
Flynn, J. R. (2009a). Requiem for nutrition as the cause of IQ gains: Raven’s gains in Britain
19382008. Economics and Human Biology, 7, 1827. doi:10.1016/j.ehb.2009.01.009
Flynn, J. R. (2009b). The WAIS-III and WAIS-IV: Daubert motions favor the certainly false
over the approximately true. Applied Neuropsychology, 16, 17.
doi:10.1080/09084280902864360
Frey, M. C., & Detterman, D. K. (2004). Scholastic Assessment or g? The relationship
between the Scholastic Assessment Test and general cognitive ability. Psychological
science, 15, 373-8. doi:10.1111/j.0956-7976.2004.00687.x
Herrnstein, R. J., & Murray, C. (1994). The Bell Curve: Intelligence and Class Structure in
American Life. New York, NY: Free Press.
Hunt, E. (2011). Human Intelligence. New York, NY: Cambridge University Press.
Hunt, E. (2012). What makes nations intelligent? Perspectives on Psychological Science, 7,
284306, http://dx.doi.org/10.1177/1745691612442905.
INVALSI (2008a). Ricerca Internazionale IEA PIRLS 2006. La lettura nella scuola primaria.
Rapporto Nazionale. [PIRLS 2006. Litteracy in primaryeducation. National repeport].
Roma, Italy: Armando
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 21
INVALSI (2008b). Rapporto Italia TIMSS 2007. Prime valutazioni sugli apprendimenti degli
studenti italiani. [First assessments on Italian students learning]. Retrieved from
www.invalsi.it/ric-int/timss2007/rapporti.php
INVALSI (2011). Servizio Nazionale Di Valutazione 2010-11 [National evaluation service
2010-11]. Retrieved from www.invalsi.it/invalsi/istituto.php?page=rapporti.
Kaufman, S. B., Reynolds, M. R., Liu, X., Kaufman, A. S., & McGrew, K. S. (2012). Are
cognitive g and academic achievement g one and the same g? An exploration on the
WoodcockJohnson and Kaufman tests. Intelligence. 40, 123-138.
doi:10.1016/j.intell.2012.01.009
Koenig, K., Frey, M., &Detterman, D. (2008). ACT and general cognitive ability.
Intelligence, 36, 153-160. doi:10.1016/j.intell.2007.03.005
Lynn, R. (2006). Race Differences in Intelligence: An Evolutionary Analysis. Augusta, GA,
US: Washington Summit Publishers
Lynn, R. (2010a). In Italy, northsouth differences in IQ predict differences in income,
education, infant mortality, stature, and literacy. Intelligence, 38, 93-100.
doi:10.1016/j.intell.2009.07.004
Lynn, R. (2010b). IQ differences between the north and south of Italy: A reply to Beraldo and
Cornoldi, Belacchi, Giofre, Martini, and Tressoldi. Intelligence, 38, 1-5.
doi:10.1016/j.intell.2010.07.001
Lynn, R. (2012). Intelligence IQs in Italy are higher in the north: A reply to Felice and
Giugliano. Intelligence, 40, 255-259. doi:10.1016/j.intell.2012.02.007
Lynn, R., & Meisenberg, G. (2010). National IQs calculated and validated for 108 nations.
Intelligence, 38, 353-360. doi:10.1016/j.intell.2010.04.007
Lynn, R., & Vanhanen, T. (2006). IQ and global inequality. Augusta, GA: Washington
Summit Publishers.
Martin, M.O., Mullis, I.V.S., Foy, P., Olson, J.F., Erberber, E., Preuschoff, C. et al. (2008a).
TIMSS 2007 International Science Report: Findings from IEA’s Trends in International
Mathematics and Science Study at the Fourth and Eighth Grades. Chestnut Hill, MA:
TIMSS & PIRLS International Study Center, Boston College.
Martini, A. (2011). OCSE-PISA 2009: è vero che stiamo migliorando? [OCSE-PISA: is it
true that we are improving?]. Rivistadell’Istruzione, 2, 63-70.
Molenaar, D., van der Sluis, S., Boomsma, D. I., & Dolan, C. V. (2011, December).
Genotype by environment interactionsi in cognitive ability tested in 14 different studies.
Paper presented at the 12th annual conference of ISIR, Limassol, Cyprus.
Mullis, I. V. S., Martin, M. O., Foy, P. Olson, J. F., Preuschoff, C., Erberber, E., et al. (2008).
TIMSS 2007 International Mathematics Report: Findings from IEA’s Trends in
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 22
International Mathematics and Science Study at the Fourth and Eighth Grades.
Chestnut Hill, MA: TIMSS & PIRLS International Study Center, Boston College.
Mullis, I. V. S., Martin, M. O., Kennedy, A. M. & Foy P. (2007). IEA's Progress in
International Reading Literacy Study in Primary School in 40 Countries. Chestnut Hill,
MA: TIMSS & PIRLS International Study Center, Boston College.
Nisbett, R. E., Aronson, J., Blair, C., Dickens, W., Flynn, J., Halpern, D. F., & Turkheimer,
E. (2012). Intelligence: New findings and theoretical developments. The American
psychologist, 67, 130-59. doi:10.1037/a0026699
OECD (Organization for Economic Cooperation & Development). (2007). PISA 2006
competencies for tomorrow's world. Paris (France): OECD.
OECD (Organization for Economic Cooperation & Development) (2010a).PISA 2009
Results: What Students Know and Can Do Student Performance in Reading,
Mathematics and Science (Volume I). doi:10.1787/9789264091450-en
OECD (Organization for Economic Cooperation & Development) (2010b).PISA 2009
Results: Learning Trends: Changes in Student Performance Since 2000 (Volume
V).doi:10.1787/9789264091580-en
Petrill, S. A., Lipton, P. A., Hewitt, J. K., Plomin, R., Cherny, S.S., Corley, R., & DeFries,
J.C. (2004). Genetic and environmental contributions to general cognitive ability
through the first 16 years of life. Developmental Psychology, 40, 805-812.
Rindermann, H. (2007). The g-factor of international cognitive ability comparisons: the
homogeneity of results in PISA, TIMSS, PIRLS and IQ-tests across nations. European
Journal of Personality, 21, 667-706. doi:10.1002/per.634
Rindermann, H. (2008). Relevance of education and intelligence at the national level for the
economic welfare of people. Intelligence, 36, 127-142. doi:10.1016/j.intell.2007.02.002
Rindermann, H., & Ceci, S. J. (2009). Educational Policy and Country Outcomes in
International Cognitive Competence Studies. Perspectives on Psychological Science, 4,
551-577. doi:10.1111/j.1745-6924.2009.01165.x
Robinson, D., Saggino, A., & Tommasi, M. (2011). The case against Lynn’s doctrine that
population IQ determines levels of socio-economic development and public health
status. Journal of Public Mental Health, 10, 178-189. doi:10.1108/17465721111175056
Rodic, M., Tikhomirova, T., Zohu, X., Wei, W., Malykh, S., Ismatulina, V., Sabirova, E.,
Tosto, M., &Kovas, Y. (2011, December). What really counts?: A cross-cultural
perspective on cognitive underpinnings of individual differences in mathematics. Paper
presented at the 12th annual conference of ISIR, Limassol, Cyprus.
Rubinacci, A. (2011). La Puglia brillante esempio di governo [Apulia an example of good
governance].Tuttoscuola, n. 509, 15-24.
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 23
Rushton, J. P., & Jensen, A. R. (2005). Thirty years of research on race differences in
cognitive ability. Psychology, Public Policy, and Law, 11, 235294. doi:10.1037/1076-
8971.11.2.235
Rushton, J. P., & Jensen, A. R. (2010). Race and IQ: A theory-based review of the research in
Richard Nisbett's intelligence and how to get it. The Open Psychology Journal, 3, 935,
http://dx.doi.org/10.2174/1874350101003010009
Templer, D.I. (2012), Biological correlates of northernsouthern Italy differences in IQ,
Intelligence ,doi:10.1016/j.intell.2012.04.005
Wicherts, J. M., Dolan, C. V., & van DerMaas, H. L. J. (2010). A systematic literature review
of the average IQ of sub-Saharan Africans. Intelligence, 38, 1-20.
doi:10.1016/j.intell.2009.05.002
Wicherts, J.M., & Wilhelm, O. (2007). What is the national g-factor? European Journal of
Personality, 21, 763-765.
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 24
Figure 1. Mean PISA scores obtained in 2006 and 2009 by northern and southern Italian
Regions.
Note. Data are provided by INVALSI. The overall score is calculated by the arithmetic mean
of reading, math, and science. North Italy: Trento, Lombardy, Veneto, Piedmont, Emilia-
Romagna, Liguria; South Italy: Basilicata, Campania, Puglia, Sardinia, Sicily.
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 25
Table 1
Mean PISA 2006 and 2009scores in Italy for macro-area (INVALSI elaborations) and
specific and overall mean differences (MD).
Reading
Math
Science
Macro
Area
Pisa
2006
Pisa
2009
MD
Pisa
2006
Pisa
2009
MD
Pisa
2006
Pisa
2009
MD
North
West
494
(4.7)
511
(3.9)
17
(7.3)
487
(4.3)
507
(4.0)
20
(6.0)
501
(4.1)
516
(4.0)
15
(6.2)
North
East
506
(3.2)
504
(2.8)
-2
(5.9)
505
(3.1)
507
(2.9)
1
(4.5)
520
(2.8)
515
(2.8)
-5
(4.7)
Center
482
(8.9)
488
(2.6)
5
(10.1)
467
(8.1)
483
(3.2)
16
(8.8)
486
(8.0)
491
(3.0)
5
(8.9)
South
443
(3.8)
468
(3.9)
26
(6.8)
440
(5.2)
465
(4.8)
25
(7.2)
448
(3.7)
466
(4.2)
19
(6.2)
South and
islands
425
(6.9)
456
(4.8)
30
(9.3)
417
(5.2)
451
(5.1)
34
(7.4)
432
(4.6)
454
(4.8)
22
(7.1)
Italy
469
(2.4)
486
(1.6)
18
(5.0)
462
(2.3)
483
(1.9)
21
(3.2)
475
(2.0)
489
(1.8)
13
(3.7)
Note. Standard error in parentheses. MD = Mean difference PISA 20092006. North West:
Val DAosta, Piedmont, Liguria, and Lombardy; North East: Bolzano, Trento, Veneto, Friuli,
Emilia-Romagna; Central Italy: Tuscany, Umbria, Marche, Lazio; South: Abruzzi, Molise,
Campania, and Apulia; South and Islands: Basilicata, Calabria, Sicilia, Sardinia.
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 26
Table 2
Mean PIRLS scores and score change between 2001 and 2006.
Macro Area
Mean 2001 (SE)
Mean 2006 (SE)
MD
North West
560.22 (4.13)
555.48 (4.20)
-4.74
North East
546.19 (6.24)
555.44 (6.96)
9.25
Center
548.31 (3.89)
557.49 (4.80)
9.18
South
527.83 (5.27)
545.97 (7.23)
18.14
South and Islands
525.38 (4.47)
546.13 (7.36)
20.65
Italy
541 (2.40)
551 (2.90)
11 (3.8)
Note. INVALSI elaborations. Standard error in parentheses. MD = Mean difference PIRLS
20062001. North West: Val DAosta, Piedmont, Liguria, and Lombardy; North East:
Bolzano, Trento, Veneto, Friuli, Emilia-Romagna; Central Italy: Tuscany, Umbria, Marche,
Lazio; South: Abruzzi, Molise, Campania, and Apulia; South and Islands: Basilicata,
Calabria, Sicily, Sardinia.
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 27
Table 3
Overall view of the differences (in standard deviations) in achievement between northern and southern Italy in the international surveys.
Reading
Math
Science
Study
‘00
‘01
‘03
‘06
‘07
‘09
‘11
‘00
‘01
‘03
‘06
‘07
‘09
‘11
‘00
‘01
‘03
‘06
‘07
‘09
‘11
PISA
15 years old
0.67
0.75
0.66
0.46
0.68
0.85
0.68
0.49
0.74
0.90
0.71
0.56
PIRLS
4th grade
0.27
0.09
TIMSS
4th grade
0.22
0.21
0.38
0.29
8th grade
0.42
0.42
0.57
0.49
INVALSI
2nd grade
0.21
0.02
5th grade
0.11
0.19
6th grade
0.29
0.45
10th grade
0.39
0.42
Note. The mean difference was calculated by averaging the means of North- West/East together then subtracting the averaged means of the
South and South and Islands macro-area. The mean difference was then divided by 100 in the case of PISA, PIRLS, and TIMSS, and by the
National standard deviation in the case of INVALSI.
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 28
Table 4
Mean PISA scores change between 2006 and 2009 in Apulia (INVALSI elaborations) and
specific and overall mean differences (MD).
Reading
Math
Science
Macro
Area
Pisa
2006
Pisa
2009
MD
Pisa
2006
Pisa
2009
MD
Pisa
2006
Pisa
2009
MD
Apulia
440
(6.7)
489
(5.0)
49
(9.3)
435
(4.8)
488
(6.9)
53
(8.6)
447
(4.3)
490
(6.3)
43
(8.0)
Note. INVALSI elaborations. Standard error in parentheses. MD = Mean difference PISA
20092006.
NORTH VS. SOUTH ITALY AND ACHIEVEMENT MEASURES 29
Table 5
Mean scores obtained by children of different areas of Italy at the 2011 assessment of language (READ) and mathematical (math) achievement
by Invalsi (Sds in parentheses) and mean differences in Sds between north and south regions.
2nd grade
5th grade
6th grade
10th grade
READ
MATH
READ
MATH
READ
MATH
READ
MATH
North West
70.8
(18.41)
60.6
(17.04)
73.6
(13.03)
69.6
(15.83)
64.7
(16.33)
49.9
(18.48)
73.3
(14.29)
51.4
(17.37)
North East
70.3
(18.54)
60.0
(17.49)
73.3
(13.46)
69.9
(15.82)
63.9
(16.85)
50.8
(18.83)
73.0
(15.05)
52.3
(16.93)
Center
70.8
(18.15)
60.9
(17.74)
74.3
(12.81)
69.0
(15.79)
64.2
(16.56)
48.0
(18.35)
68.9
(16.71)
46.6
(17.39)
South
67.8
(19.56)
60.7
(19.22)
72.8
(14.35)
67.6
(18.07)
60.8
(18.01)
43.3
(18.22)
68.5
(16.27)
46.3
(18.50)
South and Islands
65.4
(20.84)
59.3
(19.91)
71.2
(14.77)
65.4
(18.06)
57.7
(18.27)
40.5
(17.76)
65.3
(16.98)
42.5
(16.66)
Italy
69.2
(19.17)
60.3
(18.24)
73.1
(13.71)
68.4
(16.80)
62.4
(17.38)
46.6
(18.74)
69.8
(16.14)
47.9
(17.80)
North-South
Differences in SD
0.13
-0.03
0.05
0.15
0.21
0.32
0.33
0.32
Note. READ = Reading. North West: Val DAosta, Piedmont, Liguria, and Lombardy; North East: Bolzano, Trento, Veneto, Friuli, Emilia-
Romagna; Central Italy: Tuscany, Umbria, Marche, Lazio; South: Abruzzi, Molise, Campania, and Apulia; South and Islands: Basilicata,
Calabria, Sicily, Sardinia. North: Val DAosta, Piedmont, Liguria, Lombardy, Trento, Veneto, Friuli, Emilia-Romagna, Tuscany, Umbria, and
Marche; South: Lazio, Abruzzi, Molise, Campania, Apulia, Basilicata, Calabria, Sicily, Sardinia.
... Lynn's thesis on north-south disparities in Italy raised much criticism. It has been observed, in particular, that PISA tests measure scholastic achievements, not general intelligence, and, furthermore, that the north-south disparities in achievements and in socioeconomic development are due to historical and economic factors (Beraldo, 2010;Cornoldi et al. 2010Cornoldi et al. , 2013Felice and Giugliano, 2011;Daniele and Malanima, 2011;D'Amico et al. 2012;Daniele, 2015). ...
... The strong negative relationship between relative poverty rates and mean test scores in mathematics across Italian and Spanish regions, and the fact that this relationship can be found across regions of other countries, has implications for the thesis according to which regional inequalities in school achievements in Italy and Spain are due to genetic differences in the populations' IQ (Lynn, 2010;2012a;Piffer and Lynn, 2014). As mentioned, with reference to Italy, this thesis has already been criticized (Beraldo, 2010;Cornoldi et al. 2010Cornoldi et al. , 2013Feliceand Giugliano, 2011;D'Amico et al. 2012;Daniele, 2015). In addition to the previous criticisms, further considerations can be made. ...
Article
Full-text available
The relationship between PISA 2012 maths test scores and relative poverty was tested in a sample of 35 Italian and Spanish regions, together with a larger sample that included Australian, Belgian, and Canadian regions. The correlation between mean scores in mathematics, adjusted for students' socioeconomic and cultural backgrounds, and poverty rates is ‐−0.84 for the Italian and Spanish sample, and −0.68 for the complete sample. In the regressions, the effect of relative poverty on mean scores in mathematics is highly significant (p < 0.01), robust to different specifications, and independent from students' backgrounds and regional development levels. It is proposed that disparities in average scores in mathematics across regions depend on the shares of low-performing students which, in turn, depend on the degree of relative poverty within regions. The implications for the thesis according to which, in Italy and Spain, regional disparities in educational achievements reflect genetic differences in the IQ of populations are discussed.
... INVALSI, an Italian institution, uses larger samples which additionally include children. The northern intellectual advantage is less dramatic in INVALSI scores than in PISA scores (Cornoldi et al., 2010;Cornoldi, Giofré & Martini, 2013) or is not seen at all in the first years of school (Daniele & Malanima, 2011;Robinson, Saggino & Tommasi, 2011). On the other hand, significant latitude-IQ-GDP correlations have emerged in the Raven standardization in Italy (Cornoldi et al., 2010) -objected by D' Amico et al. (2012) -and in other tests (e.g., Piffer & Lynn, 2014). ...
... Lynn (2012) reported Italian correlations between latitude, percentage of blond hair, and frequency of the xR1a allele (a marker for European Mesolithic populations) and E1b1b allele (a marker for North African ancestry), as well as between the genetic markers and PISA scores. On the other hand, the north-south differences in wealth would only have emerged in the process of Italian industrialization (Daniele & Malanima, 2014), and a trend toward northsouth convergence of PISA scores over the years has been reported (Cornoldi et al., 2013;Daniele, 2015). (3) Causality was debated, too; government investment in school infrastructure and the quality of education, greater in northern than southern Italy, may have enhanced IQ differentially (Cornoldi et al., 2010;D'Amico et al., 2012;Daniele, 2015). ...
Article
Full-text available
Lynn’s (2010) attribution of decreasing north-to-south PISA scores and GDP in Italy to inherited intelligence caused debate, but cognitive-capitalism’s tenet that regional/national intelligence, inherited or not, determines provincial/country wealth remains unchallenged. We introduce a new theory into the debate and test a UV radiation → intelligence → socioeconomic latent variable (SELV) model vis-à-vis a radiation → SELV → intelligence one. Only the latter is found well-adjusted in Italy, more so when intelligence is assessed in adolescents than children. Since hair and eye color vary with latitude in Italy and UV radiation could be a proxy for ancestry, the models are retested among White students in the USA. Similar results observed suggest that, in both countries, increased UV radiation impairs socio-economic development, perhaps by reducing industriousness through oxidative stress. The ensuing less-developed social environments exert negative influences on individuals’ cognition. Yet, ancestry appears to add latitudinal socioeconomic variance in Italy. A third study in Brazil, in turn, shows that cognitive effects of UV radiation through SELV are clearly distinguishable from the cognitive effects of race. The results suggest that the postulate that intelligence causes wealth needs revision. Key Words: UV radiation; Intelligence; Socioeconomic latent variable; Italy; USA; Brazil
... In fact, the way that Italian geographical areas are organized into regions offers an appropriate unit, as Italian regions have separate governments, reflect different historical and cultural backgrounds, were historically part of different states up until about 160 years ago, and currently have different dialects (e.g., Neapolitan, which was the official language of the Kingdom of the Two Sicilies), cuisines, artistic tradition and other differences. Analyses carried out in several studies stressed the importance of examining the achievement gap at the regional level rather than at the nationwide level or macro-areas (e.g., Cornoldi, Belacchi, Giofrè, Martini, & Tressoldi, 2010;Cornoldi, Giofrè, & Martini, 2013;Lynn, 2010). ...
Preprint
Full-text available
Whether males outperform females in mathematics is still debated. Such a gender gap varies across countries, but the determinants of the differences are unclear and could be produced by heterogeneity in the instructional systems or cultures and may vary across school grades. To clarify this issue, we took advantage of the INVALSI dataset, that offered over 13 million observations covering one single instructional system (i.e., the Italian system) in grades 2, 5, and 8, in the period 2010-2018. Results showed that males outperformed females in mathematics (and vice versa in reading), with gaps widening from the 2nd through to the 8th grade. The gender gap in mathematics was larger in the richer northern Italian regions (also characterized by greater gender equality) than in southern regions. This was not explained by average performance or fully accounted for by economic factors. No such north-south difference of the gap emerged in reading. Results are discussed with reference to the literature showing that the gender gap varies across world regions.
... In Italy, all children follow the same national curriculum for language and mathematics, so this survey offers reliable and comparable data for the entire Italian student population. As Italy's population is highly heterogeneous in many other respects (see Cornoldi et al., 2013), it also can be considered as representative of the European population. Complete information on the INVALSI assessments, including the test materials, can be obtained only for some years and only up until 2017 (the last year when identical tests were administered to the whole of the population involved in the survey). ...
Article
Whether intellectually gifted children have a greater emotional response when tested is still unclear. This may be due to the marked heterogeneity of this particular population, and the fact that most studies lack the power to reduce the noise associated with this heterogeneity. The present study examined the relationship between performance and emotional response in 468,423 Italian fifth-graders taking a national test on mathematics and language. Analyses were performed using statistical models with polynomial terms. Special attention was paid to estimating the mean emotional response of the children who were gifted (1.5-2.5 standard deviations above the mean) or highly gifted (more than 2.5 standard deviations above the mean). The results showed that, although a lower emotional response correlated with a higher achievement, this relationship is nonlinear, and the estimates for gifted and highly gifted children were virtually the same. Girls showed a greater emotional response than boys on all levels of performance. The theoretical and practical implications of these findings are discussed.
... In fact, the way that Italian geographical areas are organized into regions offers an appropriate unit, as Italian regions have separate governments, reflect different historical and cultural backgrounds, were historically part of different states up until about 160 years ago, and currently have different dialects (e.g., Neapolitan, which was the official language of the Kingdom of the Two Sicilies), cuisines, artistic tradition and other differences. Analyses carried out in several studies stressed the importance of examining the achievement gap at the regional level rather than at the nationwide level or macro-areas (e.g., Cornoldi, Belacchi, Giofrè, Martini, & Tressoldi, 2010;Cornoldi, Giofrè, & Martini, 2013;Lynn, 2010). ...
Article
Full-text available
Whether males outperform females in mathematics is still debated. Such a gender gap varies across countries, but the determinants of the differences are unclear and could be produced by heterogeneity in the instructional systems or cultures and may vary across school grades. To clarify this issue, we took advantage of the INVALSI dataset, that offered over 13 million observations covering one single instructional system (i.e., the Italian system) in grades 2, 5, and 8, in the period 2010–2018. Results showed that males outperformed females in mathematics (and vice versa in reading), with gaps widening from the 2nd through to the 8th grade. The gender gap in mathematics was larger in the richer northern Italian regions (also characterized by greater gender equality) than in southern regions. This was not explained by average performance or fully accounted for by economic factors. No such north-south difference of the gap emerged in reading. Results are discussed with reference to the literature showing that the gender gap varies across world regions.
... The present results need to be confirmed by further research, partly because of the present study's several limitations. In particular, in order to obtain a sample representative of the Italian school population, we involved schools in different geographical regions and social areas, and the role of these variables should be further explored in a larger sample because of the observation of Cornoldi, Belacchi, Giofr e, Martini, and Tressoldi (2010) that regional differences in Italy affect reading comprehension, but not text reading speed and accuracy (see also Cornoldi, Giofr e, & Martini, 2013). The investigation should also be extended to other measures of reading, including speeded text comprehension, word reading decoding, and other reading-related abilities. ...
Article
Background: Reading can be assessed using different materials, including non-words and texts. Unlike reading words or non-words, reading a text may be supported by reading comprehension, and the extent of this support could change with the amount of schooling. Aim: The present study aimed to examine how reading decoding in a shallow orthography like Italian changed with years of schooling, depending on the type of material and the contribution of non-word reading and reading comprehension to text reading speed. Methods: Six hundred and forty two typically developing Italian students from 8 to 16 years old were involved. They were administered grade-appropriate tasks assessing text reading speed, non-word reading speed, and reading comprehension. Results: The results showed that, although the two reading speed measures correlated closely, non-word reading speed improved only slightly with age, while the increase in text reading speed was steeper. Reading comprehension was a significant direct predictor of text reading speed after controlling for non-word reading speed. Importantly, however, while the difference in reading speed between non-words and text widened with schooling, the role of reading comprehension declined significantly, the ΔR2 dropping from .10 in primary school to just .01 in high school. Conclusions: These findings and their implications are discussed in the light of the relationship between reading comprehension and reading speed in a language with a shallow orthography across school grades.
... This thesis was criticized by several Italian researchers who argued that PISA scores are a measure of educational attainment rather than of intelligence and proposed economic, educational, and social explanations for the north-south differences in IQs and PISA scores (Beraldo, 2010;Cornoldi, Belacchi, Giofre, Martini, & Tressoldi, 2010;Daniele & Malanima, 2011;Felice & Giugliano, 2011;D'Amico, Cardaci, Di Nuovo, & Naglieri, 2012;Cornoldi, Giofrè, & Martini, 2013;Daniele, 2013;and Daniele, 2015). ...
Article
Differences in intelligence have previously been found to be related to a wide range of inter-individual and international social outcomes. There is evidence indicating that intelligence differences are also related to different regional outcomes within nations. A quantitative and narrative review is provided for twenty-two countries (number of regions in parentheses): Argentina (24 to 437), Brazil (27 to 31), British Isles (12 to 392), to 79), Spain (15 to 48), Switzerland (47), Turkey (12), the USA (30 to 3100), and Vietnam (61). Between regions, intelligence is significantly associated with a wide range of economic, social, and demographic phenomena, including income (r unweighted = .56), educational attainment (r unweighted = .59), health (r unweighted = .49), general socioeconomic status (r unweighted = .55), and negatively with fertility (r unweighted = −.51) and crime (r unweighted = −.20). Proposed causal models for these differences are noted. It is concluded that regional differences in intelligence within nations warrant further focus; methodological concerns that need to be addressed in future research are detailed.
... Questo dato evidenzia che una differenza esiste ma non spiega il perché di questo scarto. Una possibilità è che i bambini inglesi abbiano punteggi più alti nei test d'intelligenza rispetto a quelli italiani (ossia siano più intelligenti); questo risultato, se vero, confermerebbe l'ipotesi secondo cui gli individui delle regioni del meridione italiano hanno un QI inferiore rispetto alla media nazionale (si veda per una revisione Cornoldi, Giofrè, & Martini, 2013), essendo il campione di standardizzazione della WISC-IV rappresentativo di tali regioni (principalmente Lazio e Campania). I risultati delle curve di crescita, però, smentiscono questo dato: i gruppi in realtà risultano piuttosto simili se confrontati in determinate fasce d'età e le differenze si riscontrano soprattutto nelle curve ipotetiche di crescita; le norme UK e quelle italiane sono state probabilmente ottenute tramite interpolazione lineare dei dati ma se i parametri utilizzati per il calcolo delle norme cambiano, allora le norme saranno necessariamente differenti. ...
Article
Full-text available
The WISC-IV is the most popular psychological battery in the world. However, not all subtests remain the same in the various adaptations of the scale. In fact, only the subtests included in the perceptual reasoning and in the processing speed indexes remain virtually the same across various adaptations of the scale. In this paper we compared Italian with United Kingdom norms in these two indices. Results showed that differences in the processing speed index were small and that they were larger in the perceptual reasoning index. Further, differences between the two standardizations, were not homogeneous, but concentrated in specific age groups. The clinical and theoretical implications of the study will be discussed.
Article
This paper analyses data on average IQ and four measures of political attitudes at both the regional level (n = 11) and the local authority level in Britain (n = 372). At the regional level, average IQ is positively associated with right-wing economic attitudes and trust in experts, but is not significantly associated with liberal social attitudes or intention to vote Remain in the EU referendum. At the local authority level, average IQ is positively associated with all four measures of political attitudes, although the associations with right-wing economic attitudes and trust in experts are stronger than the associations with liberal social attitudes and intention to vote Remain in the EU referendum. In multivariate models, average SES is a better and more robust predictor of political attitudes than average IQ.
Article
Cross-regional correlations between average IQ and socio-economic development have been reported for many different countries. This paper analyses data on average IQ and a range of socio-economic variables at the local authority level in the UK. Local authorities are administrative bodies in local government; there are over 400 in the UK, and they contain anywhere from tens of thousands to more than a million people. The paper finds that local authority IQ is positively related to indicators of health, socio-economic status and tertiary industrial activity; and is negatively related to indicators of disability, unemployment and single parenthood. A general socio-economic factor is correlated with local authority IQ at r = .56. This correlation increases to r = .65 when correcting for measurement error in the estimates of IQ.
Article
In Project STAR, 11,571 students in Tennessee and their teachers were randomly assigned to classrooms within their schools from kindergarten to third grade. This article evaluates the long-term impacts of STAR by linking the experimental data to administrative records. We first demonstrate that kindergarten test scores are highly correlated with outcomes such as earnings at age 27, college attendance, home ownership, and retirement savings. We then document four sets of experimental impacts. First, students in small classes are significantly more likely to attend college and exhibit improvements on other outcomes. Class size does not have a significant effect on earnings at age 27, but this effect is imprecisely estimated. Second, students who had a more experienced teacher in kindergarten have higher earnings. Third, an analysis of variance reveals significant classroom effects on earnings. Students who were randomly assigned to higher quality classrooms in grades K-3-as measured by classmates' end-of-class test scores-have higher earnings, college attendance rates, and other outcomes. Finally, the effects of class quality fade out on test scores in later grades, but gains in noncognitive measures persist.
Article
This is a review of the relationship between schooling, IQ, and the cognitive processes presumed to underpin IQ. The data suggest that much of the causal pathway between IQ and schooling points in the direction of the importance of the quantity of schooling one attains (highest grade successfully completed). Schooling fosters the development of cognitive processes that underpin performance on most IQ tests. In western nations, schooling conveys this influence on IQ and cognition through practices that appear unrelated to systematic variation in quality of schools. If correct, this could have implications for the meaning one attaches to IQ in screening and prediction as well as for efforts to influence the development of IQ through changes in schooling practices.