International Journal of Psychology and

Psychological Therapy
ISSN: 1577-7057
riptp@ual.es
Universidad de Almería
España

Manga, Dionisio; Ramos, Francisco; Morán, Consuelo

The spanish norms of the NEO five-factor inventory: new data and analyses for its improvement
International Journal of Psychology and Psychological Therapy, vol. 4, núm. 3, november, 2004, pp.
639-648
Universidad de Almería
Almería, España

Available in: http://www.redalyc.org/articulo.oa?id=56040311

How to cite
Complete issue
Scientific Information System
More information about this article Network of Scientific Journals from Latin America, the Caribbean, Spain and Portugal
Journal's homepage in redalyc.org Non-profit academic project, developed under the open access initiative
International Journal of Psychology and Psychological Therapy 2004, Vol. 4, Nº 3, pp. 639-648

The Spanish Norms of the NEO Five-Factor Inventory:

New Data and Analyses for its Improvement
Dionisio Manga*1, Francisco Ramos2, and Consuelo Morán1
1 2
Universidad de León, España; Universidad de Salamanca, España

ABSTRACT

This study examined the psychometric properties of the Spanish version of the
NEO Five-Factor Inventory (NEO-FFI) in a large sample of 1136 adults (444 men
and 692 women). Participants who completed the NEO-FFI were university students
(65%), professionals of education and health (27%), and a small group of older
people (8%). Item-level analyses provided reliability measures and factor structure
of the NEO-FFI. The N, E, and C scales showed higher internal consistency and
homogeneity than A and O. Some weaker items, particularly in the O scale, were
found. Overall, its reliability and factor structure were in line with results reported
in other studies; in addition, the analyses provided similar means of scales and
intercorrelations between the five factors. The present study found that normative
data of Neuroticism and Conscientiousness scales can be improved in the Spanish
edition of the NEO-FFI. The discussion includes the utility of improved norms for
the profile analysis, suggesting its applicability as a dimensional measure for
personality disorders.
Keywords: Five-Factor model, Item analysis, Norms, Personality, NEO-FFI.

RESUMEN

Hacia una mejora de los datos normativos en la adaptación española del NEO-
FFI. En una amplia muestra de 1136 sujetos adultos (444 varones y 692 mujeres)
se han estudiado las propiedades psicométricas de la versión española del Inventa-
rio abreviado del NEO-PI-R (el NEO-FFI). Los participantes en el estudio que
completaron el NEO-FFI fueron estudiantes universitarios (65%), profesionales
sanitarios y de la educación (27%) y un pequeño grupo de personas mayores (8%).
Mediante análisis de ítems se obtuvieron medidas de la fiabilidad y estructura
factorial del NEO-FFI. Las escalas de Neuroticismo (N), Extraversión (E) y Res-
ponsabilidad (C) fueron superiores a las de Amabilidad (A) y Apertura (O) en
consistencia interna y homogeneidad. Se encontraron algunos ítems más débiles, en
particular de la escala O. En conjunto, la fiabilidad y estructura factorial del NEO-
FFI son coincidentes con las halladas por los estudios con la versión en inglés, con
similares medias de las escalas y correlaciones entre los cinco factores. Según este

* Corresponding author: Dionisio Manga. Universidad de León. Departamento de Filosofía y Ciencias de la Educación.
Campus de Vegazana, s/n. 24071 León, España. Fax: 34-987-291-135. E-mail: dfcdmr@unileon.es. Acknowledgements:
Preparation of this article was supported in part by a Grant from the Universidad de León (ULE-2001-06 B) to the
first author. We thank a dedicated group of undergraduates for collaboration with data collection and data entry, and
the assistance of Jaclyne Manzy in preparing the manuscript.
640 MANGA, RAMOS, AND MORÁN

estudio, los datos normativos de las escalas N y C de la adaptación española pueden

mejorarse. En la discusión se tiene en cuenta la utilidad de los datos normativos del
NEO-FFI mejorados para el análisis de perfiles, así como se sugiere su aplicabilidad
como medida dimensional de los trastornos de personalidad
Palabras clave: Modelo de los Cinco Factores, Análisis de ítems, Datos
normativos, Personalidad, NEO-FFI.

The Five-Factor model consists of hierarchical trait organization and comprises

five basic personality dimensions or factors. These factors are often termed the “Big
Five” and represent a general consensus in differential psychology. During the past
decades, Costa and McCrae (1985) and McCrae and Costa (1997) have proposed this
model as general framework for studying the different traits of normal personality into
lexical research. The five factors are named Neuroticism (N), Extraversion (E), Openness
(O), Agreeableness (A), and Conscientiousness (C). The Big Five model has derived
mainly from the lexical approach to the study of personality (John, 1990; McCrae &
John, 1992). The lexical approach to the taxonomy of personality traits has been followed
by psychometric studies, many of which have given support to the Bige Five model
(McCrae & Costa, 1987; McCrae, 1989). Indeed, as can be seen in Goldberg (1993),
there are two five-factor models, one developed by Costa and McCrae (1985) and
operationalized in the NEO Personality Inventory (NEO-PI), and a second model associated
with studies based on the lexical hypothesis. Between the two versions of the five-
factor model there are high similarities and agreement (Golberg, 1993; Saucier & Golberg,
1998).
The NEO Personality Inventory-Revised (NEO-PI-R) of Costa and McCrae (1992)
is a self-report inventory, one of the most widely used measures of the Five-Factor
model. The NEO-PI-R comprises 240 elements or items. The participants respond by
marking whether they are strongly agree, agree, neutral, disagree or strongly disagree
with a given proposition about themselves. The scores of items are summed to provide
an overall measure of the five factors. Every factor comprises six facets or scales for
more specific characteristics of personality. For example, the N factor refers to a tendency
to experience tension, hostility, depression, social-anxiety, impulsivity, and stress
vulnerability, all of which are facets of neuroticism.
Costa and McCrae also developed a short form of the NEO-PI, that is the NEO-
FFI (NEO Five Factor Inventory, Costa & McCrae, 1992). The NEO-FFI comprises
only 60 items derived from a factor analysis on scores of the NEO-PI. The NEO-FFI
was developed to provide a concise measure of the five basic personality factors and
also uses a five-point Likert response format. The NEO-FFI has been translated into
several different languages and shown validity and utility in a number of different
contexts, according to McCrae and Costa (2004).
“The psychometry behind the items and factor structure of the NEO-FFI also
appear more ambiguous than one would perhaps desire” (Egan, Deary & Austin, 2000).
Some studies have provided recently item-level analyses of the English version of
NEO-FFI (e.g., Egan et al., 2000; Holden & Fekken, 1994; McCrae & Costa, 2004).

© Intern. Jour. Psych. Psychol. Ther. http://www.ijpsy.com

 THE SPANISH NORMS OF THE NEO FIVE-FACTOR INVENTORY 641

Generally, some items in the O and A scales have low loading in their correspondent
factor and result deviated from the norms presented in the manual. So, for example, the
obtained scores on the sample of the female Canadian student by Holden and Fekken
(1994). This weakness of some items and scales is recognized in the contemplated
revision of the NEO-FFI by McCrae and Costa (2004).
Recently, McCrae and Costa (2004) have proposed a revision of this short form
of NEO-PI-R. They propose a contemplated revision of the NEO-FFI by replacing 14
of their 60 items. Different selection of elements was performed in the Spanish NEO-
FFI normalization than in the English original version. By this reason, both versions do
not have item-by-item coincidence. The Spanish NEO-FFI resulted from selected items
with higher loadings on factors, according to data of the Spanish NEO-PI-R adaptation
(see Costa & McCrae, 1999). This selection of 60 items already included 10 of the 14
items proposed for replacement by McCrae and Costa (2004).
The current study sought to examine the psychometric properties of the Spanish
version of the NEO-FFI to ascertein whether similar results could be observed in other
studies with the English version. The NEO-FFI was given to 1136 Spanish participants
as part of three independent research studies. Data from these studies were used together
and subjected to item-level analyses. The main purpose of this study was the improvement
of existent norms (Costa & McCrae, 1999) in Spanish version of NEO-FFI.

METHOD

Participants

The sample consisted of 1136 subjects between the ages of 18 and 81 (M= 39.8,
SD= 12.7, median= 26). Subjects who completed the NEO-FFI were 733 (65%) students
of several careers in Unniversidad de León (España), 312 (27%) professionals mainly
of education and health, and a small group of 91 (8%) old persons attending courses
in the named “University of Experience”. The majority of the present sample were
university students who ranged in age from 18 to 34 years, a subsample of professionals
who ranged from 23 to 60 years, and a small subsample of older persons between 55
and 75 years of age. The full sample comprises only 444 men because between the 693
university students women largely exceeded the men: among university students more
than two-thirds were girls.

Test and procedure

The NEO-FFI (NEO Five Factor Inventory, Costa & McCrae, 1992) was
administered to all subjects of the research. The NEO-FFI comprises 60 items derived
from a factor analysis on scores of the Spanish NEO-PI-R translation. The instructions
indicated the response type on the Likert five-point scale. The participants responded
by marking on each of 60 items whether they are strongly agree, agree, neutral, disagree
or strongly disagree with a given proposition about themselves. The scores of 12 items
are summed to provide an overall measure of every factor. The NEO-FFI was administered

http://www.ijpsy.com © Intern. Jour. Psych. Psychol. Ther.

642 MANGA, RAMOS, AND MORÁN

to university students in the classroom (in small groups) and was corrected by the same
students. The NEO-FFI was offered to professionals to complete in their homes by
means of trained collaborators. Several older persons completed the questionnaire, not
in the classroom, but in their homes, as well. The correction of responses of both
groups was carried out by trained collaborators.

RESULTS

Normative data and item-level-analysis of the Spanish NEO-FFI.

Table 1 presents a summary of factor means, standard deviations, reliabilities

(alpha coeficients) and mean inter-item correlation (mic) as an estimate of scale
homogeneity for the full sample. As shown in Table 1, all scales had acceptable reliabilities
and homogeneity, particularly the N, E and C scales, and the A scale resulting with the
lower reliability and homogeneity. Table 1 presents a comparison of males and females
scores on the Spanish NEO-FFI scales.
The gender differences are highly significant for N and A, with men being lower
than women. There was no significant difference for E, O and C, but it is a clear trend

Table 1. Means, standard deviations, internal consistency (Cronbach’s alpha

coeficients), and mean inter-item correlation (mic) for a large Spanish sample
tested using the NEO-FFI (n= 1136), and divided by sex (raw scores).

All subjects (n = 1136) Men (n = 444) Women (n = 692) t p<

Mean SD Alpha Mic Mean SD Mean SD
N 20.53 7.46 0.82 0.28 18.88 7.13 21.58 7.48 -6.03 .001
E 31.72 6.74 0.81 0.27 31.27 6.86 32.05 6.64 -1.91 ns
O 28.27 6.73 0.76 0.22 27.81 6.99 28.58 6.55 -1.88 ns
A 29.95 5.94 0.71 0.17 28.76 6.05 30.72 5.76 -5.49 .001
C 30.49 6.55 0.81 0.28 30.06 6.69 30.82 6.50 -1.89 ns

Table 2. Normogram to convert adult raw scores on the NEO-FFI to equivalent T-

scores and Percentile-scores based on the new Spanish norms.

Pc All subjects (n = 1136) Males (n = 444) Females (n = 692) T

N E O A C N E O A C N E O A C
99 40 45 44 43 45 37 45 44 43 45 41 46 44 43 45 73
98 38 44 42 41 44 36 44 43 41 44 38 45 42 42 44 71
95 34 42 40 40 41 32 42 39 38 40 34 43 40 40 41 66
85 29 39 35 36 37 26 39 35 35 37 30 39 35 37 37 60
65 23 35 31 32 33 21 34 31 31 33 24 35 31 33 34 54
50 20 32 28 30 31 18 32 28 29 30 21 32 29 31 31 50
30 16 28 25 27 27 15 28 24 26 27 17 29 25 28 28 44
15 13 25 21 24 24 12 25 20 22 23 14 25 22 25 24 40
5 9 20 17 20 19 8 19 17 19 19 10 21 18 21 20 34
3 8 19 16 18 17 6 17 15 17 17 9 20 16 19 17 31
1 6 13 12 15 14 5 13 12 15 14 7 15 14 16 14 27

© Intern. Jour. Psych. Psychol. Ther. http://www.ijpsy.com

 THE SPANISH NORMS OF THE NEO FIVE-FACTOR INVENTORY 643

in the same direction, it is, with women being higher than men. Given that significant
gender differences were found across some factors, separate means and standard deviations
were calculated for total sample, for male participants, and for female participants.
These initial standardization data are presented in Table 1.
The percentiles and T-score norms for total sample, and also separate norms for
men and women, are presented in Table 2 (the raw scores were converted to T-values,
with a mean of 50 and an SD of 10). As in the original Spanish version of NEO-FFI,
it is not standard but empirical normalization.
Table 3 presents a comparison between mean scores (when T-score= 50) for men
and women of UK, US, and Spain. According to the data, the means of the present
research (New) show considerable difference with the means of Spanish norms (Ma-
nual) particularly for N and C factors, being men and women in Manual lower for N
and higher for C. In fact, there is great similarity between our norms (New) and all
measures in English version of NEO-FFI, particularly between scores of Spanish norms
(New) and scores of British norms (see Egan et al., 2000). Nevertheless, all Spanish
scores (of the Manual and New) for E appear over those of English version (British and
American).

Varimax rotation of the Spanish NEO-FFI items.

To determine its underlying factor structure, the Spanish NEO-FFI scores of

item-level (60 items) were subjected to a principal components analysis with varimax
rotation. Fourteen factors initially were extracted with eigenvalues greater than 1, which
explained a total of 54.09% of the observed variance in the NEO-FFI. The five-factor
solution suggested by a scree test accounted for approximately 35% of the total variance.
For the NEO-FFI analysis, the first five eigenvalues (with percentage of explained
variance) were N 7.10 (11.83), E 4.34 (7.23), C 3.63 (6.05), A 3.03 (5.05), and O 2.82
(4.69).
Table 4 provides the item composition and their loadings on any of the five
factors. As can be seen in Table 4, only N56, O28, O33, O48 and A59 shared higher
loadings in other factors (for example, the three O items said share higher loadings in
E factor). Factors N, E and C had all items loading highly and positively on the factors,

Table 3. Comparison of British, American, and Spanish (Manual and New) raw scores
for the NEO-FFI when T-score= 50.
UK US Spanish Spanish
(Manual) (New)
Men Women Men Women Men Women Men Women
N 19 20 18 20 14 16 19 21
E 27 27 27 27 32 33 31 32
O 26 26 27 27 29 30 28 28
A 30 30 32 34 32 34 29 31
C 32 32 34 35 36 36 30 31

http://www.ijpsy.com © Intern. Jour. Psych. Psychol. Ther.

644 MANGA, RAMOS, AND MORÁN

Table 4. Factor-analysis of the five varimax factors extracted from the 60 items of the NEO-
FFI in a sample of 1136 Spanish subjects (the parentheses indicate the number of items in
the NEO-PI-R, and the bold numbers are those with higher loading in the other scale).
Factor 1 Factor 2 Factor 3 Factor 4 Factor 5
N 1 (136) .62
N 6 (61) .44
N 11 (211) .53
N 16 (41) .68
N 21 (191) .68
N 26 (86) .61
N 31 (91) .66
N 36 (186) .59
N 41 (236) .48
N 46 (71) .60
N 51 (201) .40 -.33
N 56 (156) .27 .32
E 2 (177) .52
E 7 (122) .52
E 12 (217) .54
E 17 (147) .54
E 22 (37) .57
E 27 (137) -.41 .47
E 32 (227) .41 .41
E 37 (102) .33
E 42 (7) .52
E 47 (117) .54
E 52 (142) .55
E 57 (32) .47
O 3 (188) .68
O 8 (128) .66
O 13 (233) .50
O 18 (98) .60
O 23 (53) .68
O 28 (63) .34 .31
O 33 (123) .33 .27
O 38 (173) .63
O 43 (113) .57
O 48 (133) .35 .30
O 53 (108) .39
O 58 (43) .37
A 4 (184) .57
A 9 (189) .56
A 14 (159) .55
A 19 (169) .49
A 24 (139) .53
A 29 (154) .52
A 34 (74) .50
A 39 (214) .33
A 44 (114) .32
A 49 (94) .45
A 54 (49) .40
A 59 (119) .34 .32
C 5 (130) .56
C 10 (50) .65
C 15 (110) .62
C 20 (235) .65
C 25 (120) .51
C 30 (195) .59
C 35 (200) .61
C 40 (215) .55
C 45 (85) .63
C 50 (90) -.41 .43
C 55 (205) .44
C 60 (95) .49

© Intern. Jour. Psych. Psychol. Ther. http://www.ijpsy.com

 THE SPANISH NORMS OF THE NEO FIVE-FACTOR INVENTORY 645

with the exception of only one item (N56) which had higher loading in A factor. With
this criterion in the item-level factor analysis, the Manual of Spanish NEO-FFI edition
shows in Table A4 similar higher loadings of five items in the other five factors, being
again three items of the O factor which share higher loadings in the E and N factors.
In fact, factor 3 contains nine of 12 items, as in the present analysis; but only the O33
or O7 item (133 in the NEO-PI-R) shares similar loadings in both analyses. Also in the
two Spanish samples the lower percentage of variance explained by five factors proceeds
from the O dimension. Factor 2 contains the E27 or E7 item (137 in the NEO-PI-R)
which has a negative loading in N dimension, as in the Spanish NEO-FFI edition.
The first five extracted factors with eigenvalues over 1 are ordered for the NEO-
FFI differently in dependence of the actual study. So, in the Spanish edition (Costa and
McCrae, 1999) the Factor 1 corresponds to C because the higher percentage of explained
variance, being N Factor 2 and E Factor 3. Instead, our analysis shows N as Factor 1,
E as Factor 2, C as Factor 3, A as Factor 4, and O as Factor 5; the same order appeared
in McCrae and Costa (2004), except C being the second Factor. In the study of Egan
et al. (2000), Factor 1 “is clearly and unequivocally N, with all items loading highly
and positively on the factor” (p. 911), and “N, A, and C dimensions represent coherent
separate traits, while E and O were more problematic” (p. 912). On the contrary to
other studies already quoted, our analysis also shows that N and E are the two more
robust dimensions measured with the Spanish NEO-FFI; of 35% of variance explained
by the first five factors (N-E-C-A-O) with item-level analysis, the 19.06% is due to N
(11.83%) and E (7.23%).
To test the stability of the factor solution, additional principal components analyses
with varimax rotations were conducted on the following subsamples: randomly selected
50% of the total sample, men-women samples, and younger-older age group samples
(ages 18-24 and 25-80). Across all subsamples, the item content of the factors changed
only minimally (the same weaker items appeared across the subsamples), and the five
factors remained equivalent when compared to the factor analysis results for the overall
sample.
DISCUSSION

The NEO-FFI was designed as a brief instrument that would provide reasonable
estimates of the five factors of personality, perhaps mainly of use in exploratory research
in a variety of contexts and cultures, as affirmed by McCrae and Costa (2004). The aim
of this study has been to improve the psychometric properties of the Spanish NEO-FFI
version (Costa & McCrae, 1999) identifying weaker items and dimensions. In the
present study some weak items were found, two of which (123 and 133) were also
detected by McCrae and Costa (2004). See, for example, E32 which loads equally in
O scale in Table 3; this item (227 of NEO-PI-R: “I am a very active person”) was given
also as weak in other studies (e.g., Egan et al., 2000; McCrae & Costa, 2004).
Overall, our results with item factor analyses suggested that the most consistently
problematic items were from the O and A scales. Also, that these two scales had lower
reliability than the N, E and C scales, in particular the lesser reliability was that of the
A scale. It is also worth noting that the O personality dimension shared higher loadings

http://www.ijpsy.com © Intern. Jour. Psych. Psychol. Ther.

646 MANGA, RAMOS, AND MORÁN

of several items with E scale. For example, the item 123 of the NEO-PI-R appeared as
a weak item of the O scale across different studies. Instead, the other nine items (186,
7, 32, 133, 189, 169, 139, 184, and 85) selected for replacement from the NEO-PI-R
for contemplated revision of original English NEO-FFI (see McCrae & Costa, 2004)
were inicially included in Spanish edition of the NEO-FFI.
McCrae and Costa (2004) said that the new version is not appreciably better than
the old except in the case of A, in which 5 of 12 items were changed. They think that
the internal consistency estimates are all acceptable, being the estimates 0.70-0.75 for
O scale and 0.72-0.69 for A scale in their two samples of NEO-FFI. According to our
analyses (see Table 1), both O and A scales have similar estimates (0.76 and 0.71,
respectively), being the mean inter-item correlation (mic) of A scale (0.17) out of
optimal level of homogeneity, which is in the 0.20 to 0.40 for personality scales (Briggs
& Cheek, 1986).
The item factor analyses in this study with Spanish version supports the idea
proposed by McCrae and Costa (1997) on the universality of personality traits. Also the
intercorrelations (by means of the Pearson’s r) between NEO-FFI scale scores were
similar to those of other studies (e.g., Egan et al., 2000; McCrae & Costa, 2004), with
most correlations that were significant due to the large samples analysed, with the
exception of O which was uncorrelated with N and C dimensions. The scales are
correlated, specially N with lower E, lower A, and lower C; E is associated with higher
levels of C and O. The means for the five factors of the Spanish version of the NEO-
FFI (Costa & McCrae, 1999) are very different from those found in this study for N
and C scales, being our data similar to the findings of the NEO-FFI in English language.
With respect to reliability, although internal consistency and homogeneity estimates
were all acceptable, results suggested high levels of internal consistency and homogeneity
for N and E dimensions, with C next to them. The similarity of reliability with English
studies gives to these three dimensions the needed stability for future practical applications,
as also for research. Nevertheless, in existing Spanish norms there is recognition that
samples influenced the low scores of N scale, being also the samples responsible for
the high scores of the C scale. We think that the normative data of the present study
will be a contribution to improve the norms of the Spanish NEO-FFI (see Costa &
McCrae, 1999), specially those of the N and C scales showing a satisfactory reliability.
The results of the present study have implications for both research and practice.
Research on the relations between personality traits and job performance is now of
absolute crucial importance for the optimal deployment of human resources (Goldberg,
1993). For example, in the personnel selection those measures associated with C are
most likely to be valid predictors for all jobs. The links between the personality profiles
and the job requirements could be better understood with the improved NEO-FFI of the
Big Five model.
If personality factors are reasonably invariant across ages including adolescence
(McCrae et al., 2002), we think that the improved Spanish NEO-FFI should prove
useful as a acceptable instrument for research into the five-factor framework from age
12 to age 18. There was agreement in the classroom (university students of personality
psychology) about the weaker items of the Spanish version of NEO-FFI for adolescents,

© Intern. Jour. Psych. Psychol. Ther. http://www.ijpsy.com

 THE SPANISH NORMS OF THE NEO FIVE-FACTOR INVENTORY 647

being necessary the change for some items as 123, 133, 227, 119, 63, and 156 (numbers
in the NEO-PI-R).
Some clinical implications may be considered about the applicability of NEO-
FFI framework for personality disorders, that is, if these could be better understood
with a dimensional model than with a categorical system (see Butcher & Rouse, 1996).
For example, Schmitz, Hartkamp, Baldini, Roolnik, and Tress (2001) intended to
discriminate patients with personality disorders from those with other mental disorders
using the NEO-FFI measures.
Finally, our data provides strong evidence that a good instrument, like the Spanish
NEO-FFI, can be improved, adding thus more clinical utility because of its applicability
as a dimensional measure for personality disorders, and for a better profile analysis
particularly in personnel selection.

REFERENCES

Briggs, S.R., & Cheek, J.M. (1986). The role of factor analysis in the development and evaluation of
personality scales. Journal of Personality, 54, 106-148.
Butcher, J.N., & Rouse, S.V. (1996). Personality: Individual differences and clinical assessment. Annual
Review of Psychology, 47, 87-111.
Costa, P.T. Jr., & McCrae, R.R. (1985). The NEO Personality Inventory manual. Odessa, FL:
Psychological Assessment Resources.
Costa, P.T. Jr., & McCrae, R.R. (1992). Revised NEO Personality Inventory (NEO-PI-R) and NEO
Five-Factor Inventory (NEO-FFI): professional manual. Odessa, FL: Psychological Assessment
Resources.
Costa, P.T. Jr., & McCrae, R.R. (1999). NEO PI-R, Inventario de Personalidad NEO Revisado. NEO-
FFI, Inventario NEO reducido de Cinco Factores [Revised NEO Personality Inventory (NEO-
PI-R) and NEO Five-Factor Inventory (NEO-FFI)]. Manual. Madrid: TEA Ediciones.
Digman, J.M. (1990). Personality structure: emergence of the Five Factor Model. Annual Review of
Psychology, 41, 417-440.
Egan, V., Deary, I., & Austin, E. (2000). The NEO-FFI: emerging British norms and an item-level
analysis suggest N, A and C are more reliable than O and E. Personality and Individual
Differences, 29, 907-920.
Goldberg, L.R. (1993). The structure of phenotypic personality traits. American Psychologist, 48, 26-
34.
Holden, R.R., & Fekken, G.C. (1994). The NEO Five-Factor Inventory in a Canadian context:
psychometric properties for a sample of university women. Personality and Individual
Differences, 17, 441-444.
John, O.P. (1990). The “Big Five” factor taxonomy: Dimensions of personality in natural language and
in questionnaires. In L. A. Pervin (Ed.), Handbook of personality: theory and research (pp. 66-
100). New York: Guilford.
McCrae, R.R. (1989). Why I advocate the five.factor model: Joint analyses of the NEO-PI and other
instruments. In D.M. Buss and N. Cantor (Eds.), Personality psychology: Recent trends and

http://www.ijpsy.com © Intern. Jour. Psych. Psychol. Ther.

648 MANGA, RAMOS, AND MORÁN

emerging directions (pp. 237-245). New York: Springer.

McCrae, R.R., & Costa, P.T.Jr. (1987). Validation of the five-factor model of personality across
instruments and observers. Journal of Personality and Social Psychology, 52, 81-90.
McCrae, R.R., & Costa, P.T.Jr. (1997). Personality trait structure as a human universal. American
Psychologist, 52, 509-516.
McCrae, R.R., & Costa, P.T.Jr. (2004). A contemplated revision of the NEO Five-Factor Inventory.
Personality and Individual Differences, 36, 587-596.
McCrae, R.R., Costa, P.T.Jr., Terracciano, A., Parker, W.D., Mills, C.J., De Fruyt, F., & Mervielde, I.
(2002). Personality trait development from age 12 to age 18: Longitudinal, cross-sectional,
and cross-cultural analyses. Journal of Personality and Social Psychology, 83, 1456-1468.
McCrae, R.R., & John, O.P. (1992). An introduction to the Five Factor model and its applications.
Journal of Personality, 60, 175-215.
Saucier, G., & Goldberg, L.R. (1998). What is beyond the Big Five? Journal of Personality, 66, 495-
524.
Schmitz, N., Hartkamp, N., Baldini, C., Roolnik, J., & Tress, W. (2001). Psychometric properties of
the German version of the NEO-FFI in psychosomatic outpatients. Personality and Individual
Differences, 21, 713-722.

Received September 1, 2004

Final acceptance October 26, 2004

© Intern. Jour. Psych. Psychol. Ther. http://www.ijpsy.com