Page 3 - 2011-ciani

different populations have controlled the islands, all attracted
by the island’s strategic position in the Mediterranean. In
the middle ages the archipelago was attacked, raided and
invaded by the Normans, the Angioinis, the Sicilians and the
Hispanics, so that no permanent population was able to
survive. The Tunisian pirate Adir Kadir undertook in 1516
the last significant raid that depleted the whole population
of the Egadi. After that last episode, no other large scale
invasions occurred and, in the second half of 1500, the
Spanish crown and the new owner, the Count Pallavicini,
re-founded the Egadi population by forcing about 70 families
to move to the archipelago from Liguria (homeland of the
Pallavicini) and Sicily and Spain (Gallitto, 2008), giving
them housing, permission to cultivate the islands and work
in the tuna fish farms (Calleri, 2006). Most of the surnames
of these founding families are still present in the island
population as indicated in the parish records. Since that time
no further significant immigration has occurred (Calleri,
2006; Gallitto, 2008; Veronese, 2003). Given that life on the
islands was anything but attractive, the only source of
immigration over the last 400 years was prisoners, exiled
people (from all over Italy and North Africa) and their jail
officers, which were sent to the renowned caved prisons on
the two main islands of the archipelago, Favignana and
Marettimo. This excluded the possibility of auto-selective
immigration based on personality traits. (Calleri, 2006;
Gallitto, 2008; The Italian Heritage, 2009; Veronese, 2003).
Only well after 1960 did a new source of immigration arrive
in the form of the new opportunities offered by tourism,
which attracted businessmen and workers from the tourism
industry.
Emigration, on the other hand, has been very high, over
30% for each generational cohort due to population growth
and agricultural overexploitation (Calleri, 2006). Emigra-
tion was so high that in 1810 the Florio family, the new
owners of the islands, was obliged to transfer three to four
new families to work in the tuna fishing farms. Emigration
continued due to the hard conditions on the islands. The
rocky environment made agriculture extremely hard and
tuna fishing began to decline as a result of the competition
of other developing Mediterranean countries (Gallitto,
2008).
The historical accounts show that the population seldom
exceeded 2000 inhabitants in the whole archipelago. In the
last 400 years, fecundity, as documented in parish records,
averaged between 5 and 7 children per woman, and infant
mortality has never exceeded typical rates for Southern Italy
(Calleri, 2006). The infant mortality rate averaged between
20% and 30% from the 16th to the 19th century, then
progressively decreased to 1% in 1952 and 0.4% in 2009
(Italian National Statistic Institute: ISTAT, 2009). Despite
high fecundity and relatively low mortality, the population
did not grow because there was a continuous outflow of
emigrants, which in the past century averaged about 35% of
each generation. Most emigrants in the last century (the last
four generations) went to California (Monterrey), Argentina,
Australia and Northern Italy (The Italian Heritage, 2009).
Some of them, in late age, returned to the archipelago and
were sampled in this study.
Copyright # 2010 John Wiley & Sons, Ltd.
Personality differences in small island populations 55
Survey subjects
The study subjects were selected using systematic and snowball
sampling (Kish, 1995), by a team of 18 trained volunteer
assistants, who were unaware of the research hypotheses. The
same selection procedure was used for both the island and
the mainland sub-samples. Eight hundred and seventy-eight
subjects were examined, 622 living on the three islands and 106
resident and born in Trapani; 150 subjects were excluded from
the analysis because they lacked one or more of the inclusion
prerequisites: that is having Sicilian origins and having a clearly
identified ancestry (known parents and grandparents origin). In
the sample, there are a number of related individuals due to the
relatively small size of the total population and the relatively
high rate of endogamy, given that our sample includes over 15%
of the Egadi resident population. Particular effort was made
to classifying all possible sub-groups of individuals living in
the island, to be able to discriminate between genetic and
environmental effects. The following analytical classification of
subjects was therefore used (see Figure 1).
Assessment of personality
In order to assess personality traits we used the framework of
the Big Five model of personality (Costa & McCrae, 1992;
Goldberg, 1990; McCrae & Costa, 1999), which is particularly
appropriate for evolutionary personality research (Buss, 1991;
Penke, Denissen, & Miller, 2007b; Nettle & Penke, in press).
We used an adjective-based questionnaire, developed in the
Italian language and widely validated in various Italian
samples; internal reliability and inter-correlation of the
subscales are reported in Perugini (Caprara & Perugini,
1994) and Piconi (Piconi, 1998). This questionnaire consists
of 50 adjectives taken from a pool of adjectives widely used
for personality assessment in Italian subjects (Di Blas &
Perugini, 2002), ten adjectives for each of the five personality
dimensions, five with positive and five with negative polarity.
For each adjective the subject had to rate how much it
describes him on a 7-point scale. Individual scores were
computed as a total of the response values for each factor, after
inverting the scale for the negative items. Questionnaires were
handed out by the assistants in person, self-administered by
the subjects, and immediately recollected.
Statistical analysis
Standardized T-scores (with mean ¼ 50 and standard
deviation ¼ 10) were used to compare personality traits
between populations (Benjamin et al., 1996; Ebstein et al.,
1996; Terracciano et al., 2005; Allik, MoËœttus, Realo, Pullmann,
Trifonova, & McCrae, 2009). The T-scores were computed by
standardizing the raw scores with reference to the distribution
of the Mainlanders control sample, thus allowing a comparison
with the previous work (Camperio Ciani et al., 2007).
Many studies have reported significant correlations
between personality traits and sex, age and educational level
(Costa, Terracciano, & McCrae, 2001; Feingold, 1994;
Goldberg, Sweeney, Merenda, & Hughes, 1998). This finding
was confirmed in our sample by a preliminary stepwise
Eur. J. Pers. 25: 53–64 (2011)
DOI: 10.1002/per