03 Chapter1b

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

1.

2 DEVELOPMENTAL BIG FIVE RESEARCH

The lexical approach to personality is the most prominent approach in today’s personality

psychology. It is a challenging task to test its applicability to all relevant fields of personality

and applied psychology and to widen its range of validity.

1.2.1 The lexical approach to personality

There are several introductions available for readers to the Big Five approach to personality,

so that here only a short overview of the main characteristics and the development of the

model are provided (see De Raad, 2000; De Raad & Perugini, 2002; John & Srivastava,

1999).

The lexical approach to personality claims that the personality relevant characteristics

of individuals become encoded into the language and so can be systematically analyzed and

taxonomized. The basic elements of personality taxonomy are traits, though research has been

extended to situations (Ten Berge & De Raad, 2002), to instincts (De Raad & Doddema,

1999) and to the nonverbal domain (Paunonen, Ashton & Jackson, 2001).

The most frequently cited definition of the lexical approach is conveyed in the lexical

hypothesis by Goldberg (1982):

Those individual differences that are the most significant in the daily transactions

of persons with each other will eventually become encoded into their language.

The more important such a difference is, the more people will notice it and wish to

talk of it, with the result that eventually they will invent a word for it (p. 204).

Pioneers of the lexical research are Allport and Odbert (1936) who constructed an

alphabetic list of traits with around 4500 stable personality characteristics. This list was

ultimately reduced to 35 traits. Ratings on those traits were factor analyzed by Cattell (1943,

1945, 1947) who first arrived to as many factors as 9-12, according to different samples and

methods. The researchers who reanalyzed his data, could not however replicate the number of

factors, and reported only five robust personality factors instead (Borgotta 1964; Digman &

26
Takemoto-Chock, 1981; Fiske, 1949; Norman, 1963; Tupes & Christal, 1961). These five

factors, Extraversion, Agreeableness, Conscientiousness, Emotional Stability and Intellect,

were named the “Big Five” (Goldberg, 1990).

The lexical Big Five factors now look back on a long tradition, and are widely

administered and researched. There is an ongoing discussion on their exact number (cf.

Ashton et al. 2004; Peabody & De Raad, 2004) or on the exact meaning of the factors (Ashton

& Lee, 2001; John & Srivastava, 1999; Peabody & De Raad, 2002), but a general consensus

is reached about their fundamental importance in personality.

Together with the development of the factor analytic methods, the research of the

lexical domain has been increased and extended. The research of the fundamental personality

factors as observed in everyday language have become more and more popular in personality

psychology and today it is one of the true leading approaches in personality and applied

personality psychology.

1.2.2 The Big Five model in developmental research

The psycho-lexical approach to personality yielded the “Big Five” personality factors that are

considered to represent the major domains of personality description and that are embraced by

a growing number of researchers. The main interest in the lexically oriented personality

investigations is focused on adult personality characteristics and on the stability and

applicability of the five factors across languages, cultures, methods, and applied fields.

Although the Big-Five dimensions have mainly been studied in adult samples, Big Five

ratings have also been obtained in samples of children and young adolescents, though less

extensively. Digman (1963), for example, started the first lexically oriented research with

major interest in child personality structure through judgments of behavioral characteristics.

His main goal was to test the complexity of personality in childhood and to search for the

fundamental dimensions of personality at an early age. Digman’s (1963, 1965, 1972) early

work was inspired and influenced by Cattel’s personality investigations. He was also looking

27
for parallels between his own and Cattel’s adult data and used the advanced technology of

computers as soon as they became available. Maybe because of his focus on complexity and

his aim to search for the developmental antecedents of Cattell’s personality dimensions, he

first suggested, “seven or eight factors would be an expected value for the number of factors

to be found in the domain of child behavior ratings” (Digman, 1972, p. 588).

Later, Digman and Takemoto-Chock (1981), Digman and Inouye (1986) and Digman

(1989) did several lexically oriented studies in developmental personality using traits and

behavioral characteristics, and finally reported five recurring factors, which they described as

equivalents of the adult Big Five factors. Digman (1994) states, that whereas the five-factor

model clearly reappears in child personality, more complex systems do not. Digman and

Shmelyov (1996) extended the investigations to the Russian language and had 480 Russian

school children rated by their teachers on three sources of scales (temperament, personality,

and education). They found high similarities with the traditional Big Five structure. This

supports the international validity of teacher-ratings and it provides information on the

comprehensiveness of the Big-Five domains across culture and measures.

Digman (1997) reanalyzed fourteen Big Five studies, among them four with children

and one with adolescents, and came to the conclusion that both in the developmental and adult

samples, two higher order factors (metatraits) may be distinguished: Factor α and β (Digman,

1994). Factor α was interpreted as a socialization factor relying on the Big Five factors

Agreeableness, Conscientiousness and Emotionality. He interpreted metatrait β in terms of

personal growth versus personal construction. This higher order factor captured Extraversion

and Intellect. He assumed that child, adolescent and adult studies do imply the presence of the

higher order factors and “these constructs furnish links between the atheoretical Big Five

model and traditional contemporary theories of personality” (Digman, 1997, p. 1253).

Later, Goldberg (2001) also reanalyzed “one of the world’s richest collection of

teachers descriptions of elementary-school children” (Goldberg, 2001, p. 699), namely

28
Digman’s six data-pools from the years between 1959 and 1967. Goldberg (2001) concluded

that in all six samples of children no other broad domains than the Big Five factors appear,

and so provided significant evidence for the Big Five relevance of teacher based personality

assessment in childhood.

Hampson et al. (2001) relied on the developmental data, and searched for former

participants who - in the meantime - had already reached their late adulthood. They collected

personality relevant data from as much as 60 percent of the original sample with the goal of

establishing possible links between the child and adult personality.

The classification into the Big Five factors of teacher’s assessment of traits in children

aged 4-12 was described by Mervielde (1994) in his study on the relevance of the Big Five in

childhood. In another research group, the validity of the Big Five factors on the basis of

teachers’ ratings of children‘s personality was studied (Mervielde, Buyst & De Fruyt, 1995).

Both studies yielded a factor structure of personality characteristics fairly well corresponding

to the Big Five factors, especially for the ages of 7-12 years. Mervielde

and De Fruyt (2000) investigated the relevance of the Big Five model for the age group of 9

to 10 year olds. All of the five factors could not be fully recovered. Instead, their study

revealed a clearly interpretable three factorial structure with an Intellect-Conscientiousness,

Extraversion-Emotional Stability, and Agreeableness factor. They attributed the results to a

lesser degree of differentiation at younger ages that relies on the limited cognitive abilities of

children on one hand, and to highly evaluative judgments typical for the age group, on the

other.

John, Caspi, Robins, Moffitt and Stouthamer-Loeber (1994) searched for the Big

Five in young adolescent boys (aged 12 and 13 years old) and used ratings by the mothers.

They developed Big Five relevant scales on the basis of the 100 items of the California Child

Q-set (CCQ; Block & Bock, 1980). Using a set of scales largely based on the CCQ items,

they concluded on seven factors, a ‘Little-Five’ structure, fairly equivalent to the adult Big

29
Five factors, plus two other factors, respectively labeled Irritability (i.e., “He whines and

pouts often”) and Positive Activity (for example: “He is physically active”). They argued, that

the two additional factors are “relatively independent personality dimensions in early

adolescence and that they may eventually merge with Extraversion and Neuroticism,

respectively, to form a single super ordinate dimension in adulthood” (John et al, 1994,

p.173).

It is interesting to note, that this seven-factor solution showed some striking

similarities to the Big Seven factor model (Almagor, Tellegen & Waller, 1995; Tellegen &

Waller, 1987), that was found cross-culturally stable (Benet & Waller, 1995; Waller, 1999).

Irritability shared a lot with Negative Emotionality (“nervous, moody, feeling hurt”) and

Positive Activity with Positive Emotionality (“sociable, animated, energetic”). Still, John et

al. (1994), in spite of their arguments in favor of these two additional factors, recommended

the use of only the traditional Big Five factors for further research purposes until the

acceptance of the two additional dimensions is proven through independent research. They

suggested this in favor of a “conceptual continuity” (John et al. 1994, p. 174) in

developmental and adult personality research. Nevertheless, they also suggested, that the two

additional factors could be added to any instrument as separate scales. The two additional

factors could not be replicated entirely in other investigations. In a longitudinal developmental

study Lamb, Chuang, Wessels, Broberg and Hwang (2002) failed to replicate Irritability and

Positive Activity as independent factors.

1.2.3 Free developmental personality descriptions and the Big Five

The following developmental studies, which provided comprehensive and international data

of child and young adolescent personality, contributed substantially to child personality and

temperament research and yielded new aspects for research in developmental psychology.

Research projects that aimed at defining the five dimensions in children’s personality

were conducted on the basis of a ‘lexicon’ of free parental descriptions of children (Buyst, De

30
Fruyt & Mervielde, 1994; Havill, Allen, Halverson & Kohnstamm, 1994; Kohnstamm,

Mervielde, Besevegis & Halverson, 1995; Slotboom, Elphink & Kohstamm, 1996). These

studies came up with very similar concepts to the adult Big Five dimensions. Central

questions to Kohnstamm et al. (1998) were which antecedents personality and temperament

have, at what age these individual differences emerge, how universal these dimensions in

childhood are, and how early personality characteristics can be interpreted in terms of the

domains of the Big Five factors (Kohnstamm, Halverson, Mervielde & Havill, 1998).

Kohnstamm et al.’s Big Five oriented study was based on free descriptions collected

in seven countries to provide a comprehensive pool of descriptors. Over two thousand

children between the ages of 2 and 12 were described in this research and over two thousand

mothers and fathers provided personality relevant data about their children (see Slotboom &

Elphick, 1998). There was a remarkable variation in the average number of descriptors used

by parents: it varied between 37 for Germany and 11 for the U.S.A. These variations were

considered to be partly cultural and partly situational as the interview settings varied across

the countries (Kohnstamm, Halverson, Mervielde & Havill, 1998). In regards to the relevance,

Mervielde (1998) reported that 68 percent of the free descriptors used by parents could be

sorted into the categories of the Big Five.

Since this international study concentrated on parental description of children and

descriptions of personality by the parents in all age groups, it does not answer the question

whether children and young adolescents themselves can fill out Big Five instruments.

1.2.4 The stability of personality through adolescence

According to Erikson (1968), crucial changes occur in adolescence during the development of

identity. The choice of an occupation, the sexual orientation, and the adaptation of a value

system are important markers of this life-segment. Blasi and Milton (1991) investigated the

subjective experience of changes in self in adolescence and found that self-identity differs

drastically between early and middle adolescence: older adolescents have a more emotional

31
and more important relation to themselves and are committed to being loyal to themselves.

The question is whether these developmental changes influence personality as conceptualized

by the Big Five dimensions.

Changes in the Big Five factors through adolescence were investigated by McCrae, et

al. (2002) by obtaining mean level changes in the Big Five dimensions. McCrae et al. (2002)

found that three of the five factors remain stable through adolescence, namely Extraversion,

Agreeableness and Conscientiousness. Changes were reported in Openness to Experience,

showing an increase in Neuroticism in the group of girls.

Two-year test-retest reliabilities were reported by Pullmann, Kiik and Allik (2004)

who administered the NEO-FFI in an Estonian sample of adolescents. Pullmann et al. (2004)

obtained stabilities of the personality dimensions in the group of 16-18 year old adolescents

(from .62 to .73). In the youngest age group of the twelve year olds the test re-test reliabilities

were overall lower (from .48 to .57). In a nine-year longitudinal study, Asendorpf and Van

Aken (2003) investigated Big Five relevant personality judgments and behavioral

observations of aggressiveness, inhibition, and cognitive aspects such as cognitive self-esteem

and school achievement correlates and found a consistent relationship between these variables

through childhood. A relative stability of the Big Five factors could be observed, especially at

later adolescence, which spoke in favor of early formation of basic personality characteristics.

Nevertheless, recent results reported changes in the Big Five characteristics when a larger life

spectrum was investigated later in life (Helson, Kwan, John & Jones, 2002; Srivastava, John,

Gosling & Potter, 2003).

When looking for links between childhood and adulthood, it seems inevitable that

some continuity exists between the different developmental stages. The real challenge for

future research is, therefore, not only in the investigation of the origins and reasons of this

stability and robustness in personality, but also the investigation of the changing

developmental aspects.

32
In conclusion, in order to be able to provide more comprehensive results on the

stability and change of personality over the life-span development, replications of former

results are needed together with more developmentally oriented personality research that

covers a larger life span from childhood to adulthood, that relies on different methods, and

that provides results based on both observers and self-reports.

1.2.5 Recent developmental issues in Big Five - developmental Big Five measures

According to Mervielde and De Fruyt (2002), there are only a few developmentally oriented

Big Five inventories for children and adolescents. There were a few attempts to apply an adult

personality measure at the end of the ‘90s, for example, by Parker and Stumpf (1998), or Roth

(2002), who used the NEO-FFI and by De Fruyt, Mervielde, Hoekstra and Rolland (2000),

who used the NEO-PI-R with adolescents. These applications were performed with

reasonable success in recapturing the original questionnaire domains. In applied research,

mostly the NEO-PI-R was administered to subjects as young as 11 years old (for example,

Gullone & Moore, 2000; Hrebickova, Cermak & Osecka, 2000).

To fill the gap between Big Five and developmental research Mervielde and De Fruyt

(1999) developed the Hierarchical Personality Inventory for Children (HiPIC, Mervielde &

De Fruyt, 1999), which they based on the free personality descriptions described earlier. This

measure comprises the five Big Five relevant dimensions Extraversion, Benevolence,

Conscientiousness, Emotional Stability and Imagination. It includes 18 hierarchical facets and

can be administered to the age group of 6 to 12 years old. The authors suggested also using it

for self-ratings of adolescents, although the HiPIC is primarily an observer inventory. The

inventory consists of 144 behavior-oriented items (like “want to shine at everything”) and

takes, according to the authors, 15-20 minutes to fill out (Mervielde & De Fruyt, 2002).

Barbaranelli, Caprara, Rabasca and Pastorelli (2003) constructed another Big Five

measure for adolescents, which they called the Big-Five Questionnaire - Children version

(BFQ-C). This measure was developed for 9-13 year old children and adolescents and can be

33
administered as an observer measure for teachers and parents but can also be used for self-

ratings. It consists of 65 items that are distributed equally (13 each) among the following Big

Five domains: Energy/Extraversion, Agreeableness, Conscientiousness, Emotional Instability

and Intellect/Openness (cf. Barabaranelli & Caprara, 2002). In their study, a moderately

significant relation between observer-ratings and self-ratings was found, but two factors

differentiated across the raters’ samples, namely Conscientiousness and Intellect/Openness.

While Van Lieshout and Haselager (1994) argued that teachers are better raters than mothers,

Barbaranelli et al. (2003) found ratings by the mothers more informative. Finally, the authors

concluded that, in spite of the incongruence in the different factor solutions, this measure is a

good candidate for applied settings as well, such as in counseling or in the educational field

(Barbaranelli & Caprara, 2002; Barbarabelli et al. 2003).

A less known questionnaire for self-ratings for adolescents was developed by Tatum

(2000). This measure, the Adolescent Big Five Inventory (ABFI), consists of 85 items and

includes items for measuring social desirability. The author found significant relations to the

teachers' ratings in three factors, namely Extraversion, Openness to Experience and

Agreeableness. Lounsbury, Tatum, Gibson, Park, Sundstrom and Wilburn reported to have

developed the Adolescent Personal Style Inventory (APSI, 2003). This measure also aims to

measure the Big Five domains, and is reported to be robust and reliable in adolescence.

In conclusion, in order to be able to anchor the Big Five dimensions in a

developmental setting, not only the developmental antecedents of the Big Five, but also their

relation to other aspects that have been linked to adolescent personality, should be

investigated (Robins, John & Caspi, 1994).

34
1.3 CONCLUSIONS AND RESEARCH QUESTIONS

This present work aims to contribute to the investigation of the Big Five dimensions, and their

relation to health relevant aspects of adolescent behavior, like the appearance of

psychosomatic problems or the engagement in substance use. The above-presented scientific

guidelines provide a firm basis for the following empirical chapters. While doing so, the

following five research questions are put into the focus of the investigation:

1. How do Big Five dimensions develop during young adolescence?

2. Can reliable self-ratings be obtained in early adolescence?

3. How do personality characteristics and the success of a school oriented primary

prevention correspond?

4. Can future health problems (for example, psychosomatic complaints) be predicted on

the basis of personality structure in early adolescence?

5. Do personality variables play a predicting role in legal drug prevention (for example,

in smoking and alcohol consumption)?

35

You might also like