European Journal of Psychological Assessment, Vol. 20, Issue 1, pp.

2738
J. Rossier et al.: A EJPA
Comp20
arison
(1),
of2004
the NEO
Hogrefe
PI-R
& Huber
and the
Publishers
16 PF 5

The Hierarchical Structures of

the NEO PI-R and the 16 PF 5*
Jrme Rossier1,2, Franz Meyer de Stadelhofen1, and Samuel Berthoud1
1

Institute of Psychology, University of Lausanne, Switzerland

Department of Psychology, University of Fribourg, Switzerland

Keywords: Personality, NEO personality inventory, sixteen personality factors questionnaire,

five-factor personality model, factor structure
Summary: The present study compares the higher-level dimensions and the hierarchical structures of the fifth
edition of the 16 Personality Factors (16 PF 5) with those of the NEO Personality Inventory Revised (NEO
PI-R). Both inventories measure personality according to five higher-level dimensions. These inventories were,
however, constructed according to different methods (bottom-up vs. top-down). Both questionnaires were filled
out by 386 participants. Correlations, regressions, and canonical correlations made it possible to compare the
inventories. As expected, they roughly measure the same aspects of personality. There is a coherent association
among four of the five dimensions measured in the tests. However, Agreeableness, the remaining dimension in
the NEO PI-R, is not represented in the 16 PF 5. Our analyses confirmed the hierarchical structures of both
instruments, but this confirmation was more complete in the case of the NEO PI-R. Indeed, a parallel analysis
indicated that a four-factor solution should be considered in the case of the 16 PF 5. On the other hand, the
five-factor solution of the NEO PI-R was confirmed. The top-down construction of this instrument seems to
make for a more legible structure. Of the two five-dimension constructs, the NEO PI-R, thus, seems the more
reliable. This confirms the relevance of the Five-Factor Model of personality.

Introduction
The Five-Factor Model (FFM) of personality is currently
the most common dimensional approach to personality.
The 16 Personality Factors (16 PF 5) and NEO Personality Inventory Revised (NEO PI-R) are two widely used
personality inventories measuring personality according
to five higher-level dimensions. The higher-level dimensions measured by each inventory are similar (Cattell,
1995, 1996; Conn & Rieke, 1994). Both tests are implementations of hierarchical models of personality based
on the lexical hypothesis. However, these hierarchical
structures differ in that they were constructed according
to different methods (bottom-up vs. top-down). The purpose of this study is to compare the five higher-level
dimensions and the hierarchical structures of both questionnaires.
In the forties, Cattell (1943) used the lexical method
to develop his instrument. The lexical hypothesis postu*

lates that language supplies a valuable sample of behavior descriptions and that the analysis of language makes
it possible to identify personality traits and their organization (Allport & Odbert, 1936; John, Angleitner, &
Ostendorf, 1988). Cattell (1945) started with a list of
adjectives (the 35 markers) administered in the peerrating domain (life-record data, L-data). From the data
he obtained, Cattell extracted 12 factors. Cattells following studies (Cattell, 1947) were concerned with the
replication of these findings. Cattell (1950a) then tried to
measure these 12 factors with a personality questionnaire (self-report-data, Q-data). Cattell (1950a) selected
marker items for factors from personality scales, which
he sampled in his standard list. But Cattell also developed new items, which were supposed to cover the factors detected in the L-data domain (peer-rating domain).
A factor analysis of the item pool (80 items) yielded 19
to 20 extracted factors; according to Cattell, 12 of these
factors showed similarities to the factors from the L-data.
Cattell detected four additional factors with the Q-data

The original data upon which this paper is based are available at http://www.hhpub.com/journals/ejpa
EJPA 20 (1), 2004 Hogrefe & Huber Publishers
DOI: 10.1027//1015-5759.20.1.27

28

J. Rossier et al.: A Comparison of the NEO PI-R and the 16 PF 5

16 P F 5
(Cattell, Cattell, & Cattell, 1993)

NEO P I-R
Figure 1. Hierarchical structures of both inven(Costa & McCrae, 1992) tories.

Higher-level dimensions

5 global scales

5 domains

Lower-level dimensions
or traits

16 primary factors

30 facet scales

Botom-up

medium (self-report domain) labeled Q1, Q2, Q3, and

Q4 (Cattell, 1956b). Later, Cattell (1957) proposed to
group the 16 primary dimensions into global scales (also
called second-order personality factors by Cattell before
the fifth edition of the 16 PF) allowing for the description
of personality structure at a higher level (Cattell, 1996).
At first, Cattells model included eight global scales.
Four of them were largely accepted as being major dimensions of personality (Argentero, 1989; Cattell, Eber,
& Tatsuoka, 1970; Krug & Jones, 1986). In the fifth and
last edition of the 16 PF (16 PF 5), five global scales are
proposed (Cattell & Cattell, 1995).
The 16 PF 5 has a good factorial validity (Saville &
Blinkhorn, 1981; Rolland & Mogenet, 1996). The firstorder structure seems globally confirmed even if there
are some discrepancies in the literature (Eysenck, 1991).
In a recent study, Chernyshenko, Stark, and Chan (2001)
were able to confirm the hierarchical structure of the fifth
edition of the 16 PF. The five-factor structure of the
16 PF (six factors if one includes reasoning) has already
been clearly confirmed by Hofer, Horn, and Eber (1997)
and Ormerod, McKenzie, and Woods (1995) on large
samples. The internal consistency of the 16 PFs scales
is satisfying. It ranges from .57 to .81 (Rolland & Mogenet, 1996). The global scales of the last edition are similar
to the five higher-level dimensions of the FFM of personality (Cattell, 1996; Chernyshenko, Stark, & Chan,
2001). The only exception is the Independence scale,
which is not equivalent to Agreeableness (Cattell, 1996).
Following the publication of the FFM (Digman, 1990;
Peabody & Goldberg, 1989; De Raad, 2000), the interest
in personality psychology and the number of studies
about the structure of personality have considerably increased (Meyer de Stadelhofen, Rossier, Rigozzi, Zimmermann, & Berthoud, submitted; Rossier, Rigozzi, &
Berthoud, 2002; Trull & Geary, 1997). Five main dimensions are supposed to underlie the structure of traits. A
large consensus exists about the FFM, even if some disagreement persists about the exact content of some of the
five dimensions. In particular, there is still some controversy about the Openness dimension (Goldberg, 1992).
The NEO Personality Inventory Revised (NEO PI-R)
(Costa & McCrae, 1992) was specifically designed to
assess the five main dimensions of personality (termed
EJPA 20 (1), 2004 Hogrefe & Huber Publishers

Top-down

domains). Costa and McCrae (1985) started the development of their NEO measure, which included the domains
Neuroticism (N), Extraversion (E) and Openness (O),
based on a cluster analytic approach of personality as
measured by the 16 PF (Costa & McCrae, 1976). These
dimensions were derived from the clusters observed in
data obtained by administering the 16 PF to a sample
divided into three age groups. Later, the domains of
Agreeableness (A) and Conscientiousness (C) were added to the NEO (Costa & McCrae, 1992). The internal
consistencies of these five domains are good and vary
between .87 and .91 (Rossier, Wenger, & Berthoud,
2001). Test-retest reliability is satisfying and varies between .63 and .83. A great number of studies have confirmed the factorial validity of the NEO PI-R (Rolland,
Parker, & Stumpf, 1998).
Byravan and Ramanaiah (1995) investigated the factorial structure underlying the 16 PF 5, the NEO PI, and
the Goldberg Markers. Using a global principal axis
factor analysis with varimax rotation on the five factor
scales of the NEO PI, the five Goldberg factor scales and
the 15 primary scales of the 16 PF 5 (excluding the primary factor Reasoning), they were able to extract five
factors corresponding to the FFM. Each factor correlated
with a specific domain of the NEO PI and with one or
several specific primary factors of the 16 PF. The authors
concluded by saying that the FFM seems to be a comprehensive framework for describing personality and for interpreting different personality systems. Moreover, the
factor loadings presented in this study do not completely
confirm the structure of the global scales of the 16 PF 5.
Indeed, for the factor identified as Extraversion, the loadings were above .40 for the primary scales Social boldness (H), Warmth (A), Liveliness (F), Dominance (E),
Openness to change (Q1), and Abstractedness (M) when
the global scale Extraversion as measured by the 16 PF 5
is actually a linear combination of Warmth (A), Liveliness (F), Social boldness (H), Privateness (N), and Selfreliance (Q2). It should be noted that the number of subjects (n = 188) was quite small in regard of the number
of variables taken into account.
Hierarchical models like the FFM allow an all-encompassing view of personality. This type of structure, in
which higher-level dimensions are made up of lower-

J. Rossier et al.: A Comparison of the NEO PI-R and the 16 PF 5

level dimensions, makes for an easier to interpret and

more legible profile (Cattell, 1956a). The lower-level dimensions are numerous and can be directly observed;
they represent personality traits. The higher-level dimensions are less explicit; they represent the structure of personality (Figure 1). The 16 PF 5 has a bottom-up hierarchy in which the five higher-level dimensions are obtained by combining the 16 primary factors into five
theoretically independent global scales. The fact that
some traits (primary factors) contribute to more than one
global scale can make it difficult to identify the higherlevel dimensions, which then seem somewhat artificial.
Nonetheless, these dimensions are close to the five domains around which the NEO PI-R is constructed (Cattell, 1996). In contrast to the 16 PF 5, the NEO PI-R has
a top-down hierarchy. Costa and McCrae (1985) first
identified five orthogonal higher-level dimensions.
Then, in each domain they defined six lower-level dimensions or facets. Thus, in the NEO PI-R, each trait
belongs to only one higher-level dimension.
As previously stated, the purpose of this study is to
compare the five higher-level dimensions and the hierarchical structures of the two instruments which are both
based mainly on the psycholexical approach. Our hypothesis is that the higher-level dimensions of the two
instruments are similar but that the top-down method
will lead to a more reliable inventory. More precisely, a
top-down method based on the results of prior bottom-up
approaches, as is the case for the NEO PI-R, will favor
the development of reliable instruments. For example, a
top-down approach allows the creation of dimensions
that have the same weight. In order to compare the higher-level dimensions of the 16 PF 5 and of the NEO PI-R,
we used three different methods. We first analyzed correlations and then used both linear regression and canonical correlation to determine to what degree the higherlevel dimensions of one instrument could explain the
higher-level dimensions of the other. Finally, in order to
reveal and compare the hierarchical structures, we conducted principal axis factor analyses on both instruments.

Method
Sample
386 subjects from the general population, 230 woman
and 156 men, participated voluntarily and anonymously
in a study comparing two self-administered personality
inventories. The mean age of this sample was 32.5 with
a standard deviation of 13.4 (minimum 18 and maximum
78). The diversity of our sample was assessed according

29

to Hollands vocational theory (Frew & Shaw, 1999;

Holland, 1973) with 14.8% having a realistic profession
(R), 17.9% an investigative one (I), 4.4% an artistic one
(A), 36.5% a social one (S), 8.5% an enterprising one (E)
and 10.1% a conventional one (C). Our study is in compliance with the ethical code of the Swiss Association of
Psychology (FSP).

Instruments
Sixteen Personality Factors 5th Edition (16 PF 5,
Cattell, Cattell, & Cattell, 1993; Mogenet & Rolland,
1995)
The 16 PF 5 is a self-rating questionnaire of 170 items.
For each question, participants had to choose between
three answers, generally yes, no and ?. The
16 PF 5 measures 16 primary factors: Warmth (A), Reasoning (B), Emotional stability (C), Dominance (E),
Liveliness (F), Rule-consciousness (G), Social boldness
(H), Sensitivity (I), Vigilance (L), Abstractedness (M),
Privateness (N), Apprehension (O), Openness to change
(Q1), Self-reliance (Q2), Perfectionism (Q3), and Tension (Q4). These 16 primary factors can be combined
into five global scales: Extraversion (Ex), Anxiety (An),
Tough-mindedness (Tm), Independence (In), and Selfcontrol (Sc). The Extraversion score (Ex) is a linear combination of the standardized scores for Warmth (A),
Liveliness (F), Social boldness (H), Privateness (N) and
Self-reliance (Q2) (Ex = 4.4 +.3A +.3F +.2H .3N
.3Q2). Anxiety (An) is a linear combination of the standardized scores for Emotional stability (C), Vigilance
(L), Apprehension (O), and Tension (Q4) (An = 1.6 .4C
+.3L +.4O +.4Q4). Tough-mindedness (Tm) is a linear
combination of the standardized scores for Warmth (A),
Sensitivity (I), Abstractedness (M), and Openness to
change (Q1) (Tm = 13.8 .2A .5I .3M .5Q1). Independence (In) is a linear combination of the standardized
scores for Dominance (E), Social boldness (H), Vigilance (L), and Openness to change (Q1) (In = 2.2 +.6E
+.3H +.2L +.3Q1). Self-control (Sc) is a linear combination of the standardized scores for Liveliness (F), Ruleconsciousness (G), Abstractedness (M), and Perfectionism (Q3) (Sc = 3.8 .2F +.4G .3M +.4Q3). One should
note that a primary factor can contribute to more than one
global scale.
NEO PI-R (Costa & McCrae, 1992; Rolland & Petot, 1998)
The NEO PI-R is a self-rating questionnaire of 240
items. Responses are made on a five-point Likert-type
scale, ranging from strongly agree to strongly disagree. The NEO PI-R measures 30 subscales termed
facets by Costa and McCrae (1985): Anxiety (N1), HosEJPA 20 (1), 2004 Hogrefe & Huber Publishers

30

J. Rossier et al.: A Comparison of the NEO PI-R and the 16 PF 5

Table 1. Cronbachs alpha coefficients and number of items of all scales for both inventories in our sample (n = 386).
Scale
16 PF 5
Warmth (A)
Reasoning (B)
Emotional stability (C)
Dominance (E)
Liveliness (F)
Rule-consciousness (G)
Social boldness (H)
Sensitivity (I)
Vigilance (L)
Abstractedness (M)
Privateness (N)
Apprehension (O)
Openness to change (Q1)
Self-reliance (Q2)
Perfectionism (Q3)
Tension (Q4)
Extraversion (Ex)
Anxiety (An)
Tough-mindedness (Tm)
Independence (In)
Self-control (Sc)
NEO PI-R
Neuroticism (N)
Anxiety (N1)
Hostility (N2)
Depression (N3)
Self-consciousness (N4)
Impulsiveness (N5)

No. items

Cronbachs

11
15
10
10
10
11
10
11
10
11
10
10
14
10
10
10
51
40
47
44
42

.56
.58
.74
.66
.65
.73
.84
.72
.74
.77
.81
.73
.65
.68
.79
.71
.86
.85
.74
.78
.85

48
8
8
8
8
8

.92
.83
.76
.81
.63
.64

tility (N2), Depression (N3), Self-consciousness (N4),

Impulsiveness (N5), Vulnerability (N6), Warmth (E1),
Gregariousness (E2), Assertiveness (E3), Activity (E4),
Excitement seeking (E5), Positive emotions (E6), Fantasy (O1), Aesthetics (O2), Feelings (O3), Actions (O4),
Ideas (O5), Values (O6), Trust (A1), Straightforwardness
(A2), Altruism (A3), Compliance (A4), Modesty (A5),
Tender-mindedness (A6), Competence (C1), Order (C2),
Dutifulness (C3), Achievement (C4), Self-discipline
(C5), and Deliberation (C6). These 30 facets are combined into five higher-level personality dimensions
termed domains: Neuroticism (N), Extraversion (E),
Openness to experience (O), Agreeableness (A), and
Conscientiousness (C). Each dimension is made up of six
facets.

Procedure
The anonymous participants were instructed to respond
to both questionnaires successively during the same session (the order of presentation was balanced). After data
capture, participants could ask for a brief summary of
their personality profile. The subjects who didnt comEJPA 20 (1), 2004 Hogrefe & Huber Publishers

Scale

No. items

Cronbachs

Vulnerability (N6)
Extraversion (E) 4
Warmth (E1)
Gregariousness (E2)
Assertiveness (E3)
Activity (E4)
Excitement seeking (E5)
Positive emotions (E6)
Openness (O)
Fantasy (O1)
Aesthetics (O2)
Feelings (O3)
Actions (O4)
Ideas (O5)
Values (O6)
Agreeableness (A)
Trust (A1)
Straightforwardness (A2)
Altruism (A3)
Compliance (A4)
Modesty (A5)
Tender-mindedness (A6)
Conscientiousness (C)
Competence (C1)
Order (C2)
Dutifulness (C3)
Achievement (C4)
Self-discipline (C5)
Deliberation (C6)

8
8
8
8
8
8
8
8
48
8
8
8
8
8
8
48
8
8
8
8
8
8
48
8
8
8
8
8
8

.81
.84
.71
.72
.75
.61
.63
.72
.87
.76
.73
.65
.57
.78
.55
.88
.83
.79
.61
.62
.75
.54
.90
.57
.76
.64
.63
.80
.75

pletely answer the two inventories were removed from

the sample (this concerned very few people as we had
tried to make sure that the subjects were motivated before testing). Therefore there is no missing data in our
sample.

Results
For the 16 PF 5 primary factors, Cronbachs coefficients ranged from .56 to .85 with a median of .72 (Table
1) and for the 16 PF 5 global scales, Cronbach coefficients ranged from .74 to .86 with a median of .85. For
the NEO PI-R facet scales, Cronbach coefficients
ranged from .54 to .83 with a median of .72 and for the
NEO PI-R domains, Cronbach coefficients ranged
from .84 to .92 with a median of .88. These results were
similar to those reported by the authors of these scales
and to those found in other studies (Byravan &
Ramanaiah, 1995; Rolland, Parker, & Stumpf, 1998;
Rossier, Wenger, & Berthoud, 2001).
Correlations between the higher-level dimensions of
the 16 PF 5 (global scales) and those of the NEO PI-R

J. Rossier et al.: A Comparison of the NEO PI-R and the 16 PF 5

31

Table 2. Correlations between the global scales of the 16 PF 5 and the domains of the NEO PI-R.
NEO PI-R

Ex

Neuroticism (N)
Extraversion (E)
Openness to experience (O)
Agreeableness (A)
Conscientiousness (C)

An

.10
.63***
.30***
.14**
.14**

.80***
.23***
.01
.25***
.24***

16 PF 5
Tm
.21***
.13**
.62***
.11*
.32***

In
.19***
.49***
.26***
.34***
.13**

Sc
.12*
.20***
.50***
.07
.66***

Note. Ex = Extraversion; An = Anxiety; Tm = Tough-mindedness; In = Independence; Sc = Self-control. Significance levels: *p < .05, **p <
.01, ***p < .001.
Table 3. Regressions with the stepwise method predicting each higher-level dimension of both inventories.
16 PF 5

Extraversion (Ex)
Extraversion (E)
Openness (O)
Agreeableness (A)
Conscientiousness (C)
R2

.63
.09
.16
.18
.47

.0001
.0199
.0001
.0001
.0001

.79
.09
.08
.65

.0001
.0026
.0120
.0001

.10
.57
.09
.16
.43

.0150
.0001
.0228
.0003
.0001

.0001
.0001
.0001
.0001

Extraversion (Ex)
Anxiety (An)
Tough-mindedness (Tm)
Independence (In)
Self-control (Sc)
R2

.57
.10
.24
.30
.10
.52

.0001
.0051
.0001
.0001
.0096
.0001

Tough-mindedness (Tm)
Self-control (Sc)
R2

.50
.27
.44

.0001
.0001
.0001

.23
.26
.25
.49
.16
.30

.0001
.0001
.0001
.0001
.0014
.0001

.10
.22
.28
.67
.55

.0112
.0001
.0001
.0001
.0001

Agreeableness (A)
.15
.37
.24
.39
.11
.42

.0008
.0001
.0001
.0001
.0088
.0001

Self-control (Sc)
Neuroticism (N)
Extraversion (E)
Openness (O)
Agreeableness (A)
Conscientiousness (C)
R2

.76
.19
.21
.69

Openness (O)

Independence (In)
Neuroticism (N)
Extraversion (E)
Openness (O)
Agreeableness (A)
Conscientiousness (C)
R2

Anxiety (An)
Tough-mindedness (Tm)
Independence (In)
R2
Extraversion (E)

Tough-mindedness (Tm)
Neuroticism (N)
Openness (O)
Agreeableness (A)
Conscientiousness (C)
R2

Neuroticism (N)

Anxiety (An)
Neuroticism (N)
Openness (O)
Agreeableness (A)
R2

NEO PI-R

Extraversion (Ex)
Anxiety (An)
Tough-mindedness (Tm)
Independence (In)
Self-control (Sc)
R2
Conscientiousness (C)

.16
.12
.35
.10
.65
.61

.0001
.0004
.0001
.0033
.0001
.0001

(domains) show a marked similarity between the global

scale Anxiety and the domain Neuroticism (r = .80). A
correspondence can also be observed between the global
scale Extraversion and the domain Extraversion (r =

Extraversion (Ex)
Anxiety (An)
Independence (In)
Self-control (Sc)
R2

.63), as well as between the global scale Self-control and

the domain Conscientiousness (r = .66). There is also a
negative correlation between the global scale Toughmindedness and the domain Openness to experience (r =
EJPA 20 (1), 2004 Hogrefe & Huber Publishers

32

J. Rossier et al.: A Comparison of the NEO PI-R and the 16 PF 5

Table 4. Canonical correlations between the global scales of the 16 PF 5 and the domains of the NEO PI-R.
1
16 PF 5
Extraversion (Ex)
Anxiety (An)
Tough-mindedness (Tm)
Independence (In)
Self-control (Sc)
NEO PI-R
Neuroticism (N)
Extraversion (E)
Openness (O)
Agreeableness (A)
Conscientiousness (C)
Canonical correlation
% of variance

.06
.83
.28
.37
.18

Canonical variables
2
r
3

.19
.89
.28
.37
.18

.30 .61
.14 .15
.17 .64
.01 .31
.75 .91

.90 .97
.18 .43
.03 .12
.12 .06
.12 .48
.84
41.7

.25 .02
.37 .49
.48 .70
.00 .05
.70 .68
.82
34.5

.62). No correspondence, however, can be observed between the global scale Independence and the domain
Agreeableness (Table 2).
Regarding the correlations between the global scales
of the 16 PF 5 and the facets of the NEO PI-R, we find
that the global scale Extraversion correlates with the facets Warmth (r = .60) and Gregariousness (r = .61). The
global scale Anxiety correlates positively with the facets
Anxiety (r = .67), Hostility (r = .68), Depression (r =
.70), Self-consciousness (r = .59), and Vulnerability (r =
.62) as well as negatively with the facet Trust (r = .42).
The global scale Tough-mindedness correlates negatively with the facets Fantasy (r = .49), Aesthetics (r = .50)
and Feelings (r = .47). The global scale Independence
correlates positively with the facet Assertiveness (r =
.65) and negatively with the facets Compliance (r = .44)
and Modesty (r = .41). The global scale Self-control
correlates negatively with the facet Fantasy (r = .56)
and positively with the facets Order (r = .67), Dutifulness (r = .53), Self-discipline (r = .46) and Deliberation
(r = .54).
Regarding correlations between the domains of the
NEO PI-R and the primary factors of the 16 PF 5, we find
that the domain Neuroticism correlates negatively with
the primary factor Emotional stability (r = .75) as well
as positively with the primary factors Apprehension (r =
.61) and Tension (r = .49). The domain Extraversion correlates with Liveliness (r = .60) and Social boldness (r =
.57). The domain Openness to experience correlates with
Abstractedness (r = .44) and Openness to change (r =
.59). The domain Agreeableness correlates negatively
with Vigilance (r = .45), and the domain Conscientiousness correlates negatively with Abstractedness (r = .48)
and positively with Perfectionism (r = .69). Correlations
below .40 are not reported. These results confirm that the
EJPA 20 (1), 2004 Hogrefe & Huber Publishers

.55
.37
.15
.56
.13

.89
.09
.16
.90
.19

.42
.21
.12
.62
.17

.57 .23
.94 .72
.38 .17
.24 .40
.24 .18
.65
12.6

.19
.38
.53
.94
.21
.54
6.9

.08
.16
.29
.83
.09

.75
.58
.61
.48
.15

.06 .34
.10 .03
1.00 .70
.10 .29
.80 .30
.35
.05
.72
.35
.79
.45
4.3

.04
.19
.62
.38
.51

correlations between the global scales of the 16 PF 5 and

the facets of the NEO PI-R, as well as the correlations
between the domains of the NEO PI-R and the primary
factors of the 16 PF 5, form a coherent pattern.
In order to more precisely compare the higher-level
dimensions measured by the two personality inventories
we conducted a series of linear regressions with the stepwise method in order to predict the global scales of the
16 PF 5 from the domains of the NEO PI-R and vice versa
(Table 3). That analysis confirmed the association between the Extraversion scales of both instruments, as well
as between the global scale Anxiety and the Neuroticism
domain, between the global scale Tough-mindedness and
the Openness to experience domain, and, finally, between
the global scale Self-control and the Conscientiousness
domain. We found no single correspondence for the global scale Independence. Nevertheless, the R square for this
scale is also above .40. Doing the regressions the other
way around confirmed all the above relations. The only
domain that was not predicted with an R above .40 was
the Agreeableness scale. This shows that this last dimension is not well measured by the 16 PF 5. Regressions
with the enter method provided very similar R.
Regressions were confirmed by a canonical analysis
on the global scales of the 16 PF 5 and on the domains
of the NEO PI-R. The standardized canonical coefficients and correlations of the higher-level dimension
with each variate are displayed in Table 4. The first canonical variable explains 41.7% of the common variance
and is associated positively with the global scale Anxiety
and with the domain Neuroticism. The second canonical
variable explains 34.5% of the common variance and is
mainly associated with the global scale Self-control and
the domain Conscientiousness. The third canonical variable explains 12.6% of the common variance and is

J. Rossier et al.: A Comparison of the NEO PI-R and the 16 PF 5

33

Table 5. Rotated factor matrix, communalities and correlations between the extracted factors and the five higher-level dimensions after
principal axis analysis of the primary factors of the 16 PF 5 for the four-factor and the five-factor solutions.
16 PF 5

Four-factor solution
h2
I
II

Warmth (A)
Emotional stability (C)
Dominance (E)
Liveliness (F)
Rule-consciousness (G)
Social boldness (H)
Sensitivity (I)
Vigilance (L)
Abstractedness (M)
Privateness (N)
Apprehension (O)
Openness to change (Q1)
Self-reliance (Q2)
Perfectionism (Q3)
Tension (Q4)

.43
.67
.58
.38
.41
.55
.30
.38
.52
.47
.42
.31
.22
.38
.31

Correlations
Extraversion (Ex)
Anxiety (An)
Tough-mindedness (Tm)
Independence (In)
Self-control (Sc)

III

IV

.65
.07
.15
.44
.01
.49
.27
.42
.06
.67
.01
.34
.45
.09
.04

.09
.79
.02
.11
.01
.28
.33
.39
.27
.06
.59
.06
.11
.09
.54

.04
.16
.03
.40
.63
.15
.05
.04
.67
.03
.10
.38
.07
.57
.01

.02
.13
.75
.12
.11
.45
.35
.24
.02
.13
.24
.20
.04
.20
.15

II

III

.93
.18
.53
.31
.14

.07
.94
.26
.02
.02

.19
.03
.52
.20
.96

Five-factor solution
h2
I
II

IV

.09
.78
.00
.09
.01
.28
.34
.43
.27
.06
.60
.07
.10
.08
.52

.58
.05
.11
.23
.02
.38
.28
.56
.09
.66
.02
.34
.35
.09
.01

.06
.17
.03
.34
.63
.11
.06
.11
.65
.04
.09
.41
.01
.57
.00

.06
.12
.77
.10
.12
.47
.31
.24
.03
.19
.23
.24
.04
.18
.16

.24
.05
.01
.73
.07
.29
.04
.12
.08
.13
.05
.06
.36
.07
.07

IV

II

III

IV

.21
.03
.06
.91
.00

.06
.95
.27
.04
.03

.79
.19
.51
.21
.06

.15
.01
.53
.18
.94

.24
.02
.01
.93
.02

.62
.03
.18
.18
.29

mainly associated negatively with the global scale Extraversion and with the domain Extraversion. The fourth
canonical variable explains 6.9% of the common variance and is associated negatively with the global scale
Independence and positively with the domain Agreeableness. The last canonical variable explains 4.3% of the
common variance and is associated negatively with the
global scale Tough-mindedness and positively with the
domain Openness. Together the five canonical variates
extract 71% of the variance of the five 16 PF 5s global
scales and 69% of the variance of the five NEO PI-Rs
domains.
In order to assess the hierarchical structure of the
16 PF 5, we conducted a principal axis factor analysis
with varimax rotation on the primary factors. We took all
the primary factors of the 16 PF 5 into account except the
Reasoning factor (B), which is not usually considered to
measure a personality trait. A parallel analysis (Horn,
1965; OConnor, 2000) with the 95th percentile as the
comparison baseline prescribed extracting four factors
explaining 56.9% of variance. The first six eigenvalues
were: 3.03, 2.25, 1.73, 1.52, 0.97, and 0.87. For the fourfactor solution, the loading matrix (Table 5) shows that
Factor I correlates positively with the primary factor
Warmth and negatively with the primary factor Privateness. This first factor correlates positively with the Extraversion global scale (r = .93) and negatively with the
Tough-mindedness global scale (r = .53). Factor II cor-

.41
.66
.60
.72
.42
.54
.30
.58
.51
.50
.42
.34
.26
.37
.31

III

relates positively with the primary factors Apprehension

and Tension and negatively with the primary factor Emotional stability. This second factor highly correlates with
the global scale Anxiety (r = .94). Factor III correlates
positively with the primary factor Abstractedness and
negatively with the primary factors Rule-consciousness
and Perfectionism. This third factor correlates negatively
with the global scales Self-control (r = .96) and Toughmindedness (r = .52). Factor IV correlates positively
with the primary factor Dominance. This fourth factor
highly correlates with the global scale Independence (r =
.91). A principal axis factor analysis with oblimin rotation give very similar results. Indeed, the correlations
among the oblimin factors range from .20 to .03 with a
mean of .07, indicating that factors are essentially orthogonal. A comparison between the two sets of factors
obtained with varimax and oblimin rotations shows that
there is a one to one correspondence (r > |.99|).
In order to compare the factorial structure with the
hierarchical structure of the inventory, we chose to extract five factors explaining 63.3% of the variance. The
loading matrix (Table 5) shows that Factor I correlates
positively with the primary factor Apprehension and
negatively with the primary factor Emotional stability.
This first factor is highly correlated with the Anxiety
global scale (r = .95). Factor II correlates positively with
the primary factor Warmth and negatively with the primary factor Privateness. This second factor correlates
EJPA 20 (1), 2004 Hogrefe & Huber Publishers

34

J. Rossier et al.: A Comparison of the NEO PI-R and the 16 PF 5

Table 6. Rotated factor matrix, communalities and correlations between the five extracted factors and the five higher-level dimensions
after principal axis analysis of the facets of the NEO PI-R.
NEO PI-R
Anxiety (N1)
Hostility (N2)
Depression (N3)
Self-consciousness (N4)
Impulsiveness (N5)
Vulnerability (N6)
Warmth (E1)
Gregariousness (E2)
Assertiveness (E3)
Activity (E4)
Excitement seeking (E5)
Positive emotions (E6)
Fantasy (O1)
Aesthetics (O2)
Feelings (O3)
Actions (O4)
Ideas (O5)
Values (O6)
Trust (A1)
Straightforwardness (A2)
Altruism (A3)
Compliance (A4)
Modesty (A5)
Tender-mindedness (A6)
Competence (C1)
Order (C2)
Dutifulness (C3)
Achievement (C4)
Self-discipline (C5)
Deliberation (C6)

h2
.73
.67
.74
.55
.47
.66
.70
.30
.49
.31
.36
.45
.60
.40
.52
.30
.49
.31
.51
.46
.52
.57
.38
.36
.49
.51
.59
.46
.66
.49

I
.85
.62
.83
.69
.35
.74
.14
.11
.39
.08
.02
.26
.13
.17
.25
.22
.15
.06
.36
.04
.03
.09
.17
.18
.32
.05
.07
.09
.30
.07

Correlations

Neuroticism (N)
Extraversion (E)
Openness (O)
Agreeableness (A)
Conscientiousness (C)

.95
.29
.04
.07
.18

Five-factor solution
II
III

.00
.08
.20
.16
.45
.31
.14
.15
.24
.36
.19
.01
.39
.03
.07
.18
.10
.18
.02
.09
.18
.00
.13
.01
.61
.69
.71
.61
.75
.60

.00
.52
.04
.13
.27
.03
.27
.02
.44
.25
.46
.01
.09
.03
.16
.01
.10
.18
.52
.67
.48
.75
.54
.40
.01
.04
.23
.26
.00
.24

.06
.00
.09
.08
.14
.00
.22
.03
.12
.05
.09
.22
.64
.59
.55
.42
.67
.49
.16
.01
.21
.04
.19
.26
.10
.18
.13
.01
.06
.08

.08
.03
.11
.17
.24
.10
.73
.51
.27
.33
.32
.58
.10
.13
.35
.20
.06
.01
.28
.04
.46
.00
.04
.32
.03
.04
.11
.08
.04
.25

II

III

IV

.27
.10
.17
.02
.96

.15
.29
.04
.92
.05

.05
.17
.97
.11
.10

.07
.83
.19
.29
.03

positively with the global scale Extraversion (r = .79)

and negatively with the global scale Tough-mindedness
(r = .51). Factor III correlates positively with the primary factors Rule-consciousness and Perfectionism and
negatively with the primary factor Abstractedness. This
third factor correlates positively with the Self-control
(r = .94) and Tough-mindedness global scales (r = .53).
Factor IV correlates positively with the primary factor
Dominance. This fourth factor correlates positively with
the global scale Independence (r = .93). Factor V correlates positively with the primary factor Liveliness. This
fifth factor correlates positively with the global scale Extraversion (r = .62). A principal axis factor analysis with
oblimin rotation gives quite similar results. The correlations among the oblimin factors range from .45 to .20
with a mean of .06. A comparison between the two sets
EJPA 20 (1), 2004 Hogrefe & Huber Publishers

IV

of factors obtained with varimax and oblimin rotations

shows that there is a one to one correspondence (r >
|.85|). It is notable that the fifth factor correlates with the
first factor (r = .42). These principal axis factor analyses only partially corroborate the hierarchical structure
of the French version of the 16 PF 5.
The principal axis factor analyses with varimax rotation and a parallel analysis of the NEO PI-R allowed us
to extract five factors explaining 58.1% of variance. The
first six eigenvalues were: 5.7, 4.0, 3.5, 2.4, 1.7, and 1.1.
The loading matrix (Table 6) shows that Factor I correlates well with most of the facets of the domain Neuroticism, that Factor II correlates well with all the facets of
the domain Conscientiousness, that Factor III correlates
well with most of the facets of the domain Agreeableness, that Factor IV correlates well with most of the fac-

J. Rossier et al.: A Comparison of the NEO PI-R and the 16 PF 5

35

Table 7. Correlations between the four-factor and five-factor solutions obtained from the 16 PF 5 and the five-factor solution obtained
from the NEO PI-R.

NEO PI-R

Four-factor solution
I
II
III

I
II
III
IV
V

.10
.06
.13
.28
.59

.79
.12
.20
.04
.10

.05
.61
.19
.48
.09

IV
.25
.31
.59
.05
.15

16 PF 5
Five-factor solution
I
II
III
.79
.13
.20
.03
.09

.10
.02
.22
.30
.46

.05
.60
.16
.51
.05

IV

.23
.30
.58
.08
.16

.01
.18
.13
.05
.53

the four-factor solution (r = .59) and with the fourth

factor of the five-factor solution (r = .58) obtained from
the 16 PF 5. The fourth factor obtained from the
16 PF 5
NEO PI-R correlates negatively with the third factor of
16 PF 5
Ex
An
Tm
In
Sc
the five-factor solution (r = .51) obtained from the
Extraversion (Ex)
1.00
16 PF 5. The fifth factor obtained from the NEO PI-R
Anxiety (An)
.16** 1.00
correlates positively with the first factor of the four-facTough-mindedness (Tm)
.44*** .10* 1.00
tor solution (r = .59) and with the fifth factor of the fiveIndependence (In)
.40*** .05 .29*** 1.00
Self-control (Sc)
.27*** .03
.46*** .17** 1.00 factor solution (r = .53) obtained from the 16 PF 5. This
comparison corroborates both some similarities and
Note. Significance levels: *p < .05, **p < .01, ***p < .001.
some discrepancies between the two hierarchical structures.
ets of the domain Openness to experience and that Factor
According to Cattell (1950b) the global scales should
V correlates well with most of the facets of the domain
be orthogonal. In our study, however, we found that some
Extraversion. The correlations between our factors and
global scales are intercorrelated (Table 8). Indeed, the
the five domains of the French version of the NEO PI-R
global scale Tough-mindedness correlates with both the
are very high. Factor I correlates with the domain Neuglobal scales Extraversion and Self-control. Concerning
roticism (r = .95), Factor II with the domain Conscienthe NEO PI-R, on the other hand, none of the correlatiousness (r = .96), Factor III with the domain Agreeabletions between domains is superior to .40.
ness (r = .92), Factor IV with the domain Openness to
These results clearly show some correspondence beexperience (r = .97), and Factor V with the domain Extween four of the five higher-level dimensions of both
traversion (r = .83). Thus, the hierarchical structure of
personality inventories. Nevertheless, this corresponthe inventory is confirmed. A principal axis factor analdence is not as strong as we would have expected. In
ysis with oblimin rotation gives very similar results. Inaddition, the hierarchical structure of the 16 PF 5 is not
deed, the correlations among the oblimin factors range
entirely validated, whereas the hierarchical structure of
from .10 to .27 with a mean of .06, indicating that the
NEO PI-R seems very reliable.
factors are essentially orthogonal. A comparison between the two sets of factors obtained with varimax and
oblimin rotations shows that there is a one to one correspondence (r > |.98|).
Discussion
The comparison of the two sets of factors obtained
from the 16 PF 5 with the set of factors obtained from the
Our results confirm the similarity between Cattells globNEO PI-R shows that the first factor obtained from the
al scales (Cattell et al., 1993) and the FFM as measured
by Costa and McCraes NEO PI-R (1985, 1992). These
NEO PI-R correlates positively with the second factor of
the four-factor solution (r = .79) and with the first factor
two tests share four of their five global dimensions. Both
inventories are based mainly on a lexical hypothesis and
of the five-factor solution (r = .79) obtained from the
the NEO was primarily created through a cluster analytic
16 PF 5 (Table 7). The second factor obtained from the
NEO PI-R correlates negatively with the third factor of
approach of the 16 PF. Based on this evidence, the discrepancies between the two tests can be considered imthe four-factor solution (r = .61) and positively with the
third factor of the five-factor solution (r = .60) obtained
portant. This can be due either to theoretical differences
from the 16 PF 5. The third factor obtained from the
or to methodological or psychometrical differences. We
consider that the different construction methods used for
NEO PI-R correlates negatively with the fourth factor of
Table 8. Inter-correlations between the global scales of the
16 PF 5.

EJPA 20 (1), 2004 Hogrefe & Huber Publishers

36

J. Rossier et al.: A Comparison of the NEO PI-R and the 16 PF 5

the two questionnaires (bottom-up versus top-down)

might partly explain these results.
Concerning the comparison between the higher-level
dimensions of the NEO PI-R and those of the 16 PF 5,
the results show an adequate internal consistency for all
global dimensions of both inventories. However, two
global scales of the 16 PF 5, Tough-mindedness and Independence, have an internal consistency below .80.
Correlations between both inventories show that there is
a good correspondence between the Neuroticism domain
and the global scale Anxiety; as well as slightly weaker
correspondences between the two Extraversion scales,
between Openness to experience and Tough-mindedness, and between Conscientiousness and Self-control.
Nevertheless, these scales are indeed close from a semantic point of view. However, concerning the last pair
of higher-level dimensions, correlations between both
inventories show that the domain Agreeableness of the
NEO PI-R has no correspondent in Cattells questionnaire. This domain only slightly negatively correlates
with the primary factor Vigilance. The linear regressions
as well as the canonical correlations confirm that the
Agreeableness domain is not represented in the 16 PF 5.
Furthermore, five of the facets of the NEO PI-R do not
correlate well with any of the factors of the 16 PF 5 (no
correlation above .40). Nevertheless, there is an adequate
correspondence between four of the higher-level dimensions of the two inventories. This was confirmed by the
canonical correlation analysis that indicated that the
overlap between the two sets of global dimensions is
important. This shows that both questionnaires have a
good convergent and predictive validity. It is interesting
to note that our results did not confirm the conclusion
that the concept of Openness to experience is not represented in the 16 PF 5 (McKenzie, Tindell, & French,
1997). This is not surprising, considering that the Openness to experience domain correlates well with the Openness to change primary factor that participates in the
Tough-mindedness global scale. The correlations we obtained between the NEO PI-Rs domains and the
16 PF 5s primary scales do not reproduce the rotated
factor loadings obtained by Byravan and Ramanaiah
(1995) even though the factors they extracted were closely associated to the five NEO PI-Rs domains.
The validity of the hierarchical structure of the
16 PF 5 was confirmed for four of the global scales by
our principal axis factor analyses. Indeed, the parallel
analysis we conducted indicated extraction of four factors. None of these factors was closely associated to the
global scale Tough-mindedness. This was also the case
with the five-factor solution. That scale always loaded on
two factors. This is certainly due to the relatively high
correlations between Tough-mindedness and Extraversion and between Tough-mindedness and Self-control. It
EJPA 20 (1), 2004 Hogrefe & Huber Publishers

is notable that this state of affairs was not really improved by a principal axis factor analysis with oblimin
rotation. Our results do not agree with the observations
of Hofer, Horn, and Eber (1997) and Chernyshenko,
Stark, and Chan (2001) who were able to identify a stable
and solid five-factor structure. The difference might be
due to the sample size (n > 10,000) or to the translation.
On the other hand, the hierarchical structure of the
NEO PI-R has been confirmed for all five dimensions.
The parallel analysis confirmed this number of dimensions. Thus, the structure of the French version of the
16 PF 5 seems less robust than the structure of the French
version of the NEO PI-R. The correlations between the
factors obtained from the principal axis analyses of the
NEO PI-R and of the 16 PF 5 confirm the correspondence we observed for four of the five global personality
dimensions. We also conducted confirmatory factor
analyses (CFA) but we decided not to present these results because CFA models with large degrees of freedom
may easily lead to a rejection of the model as a result of
statistical artifacts stemming from overly high power because of the size of the model (McCrae, Zonderman,
Costa, Bond, & Paunonen, 1996; Raykov, 1998).
The higher-level dimensions of the two inventories
were constructed very differently. The global scales of
the 16 PF 5 are linear combinations of the standardized
scores on the primary factors, each made up of several
items. Thus, the global scales are made of fractions of
traits or primary factors. A particular primary factor can
contribute to none, or to one, or more, global scales. The
idea was to create a small number of orthogonal higherlevel dimensions based on dependent primary factors
(Cattell, 1996). Furthermore, the instruction manual for
the French version suggests using the coefficients obtained for the English version. This seems inappropriate
knowing that translation can sensibly transform the
meaning of an item. Our data might contribute to the
calculation of more appropriate coefficients for this
French version. The fact that coefficients were not recalculated for the French version might account for the high
intercorrelation between the global scales of the 16 PF 5.
The correlations we obtained correspond to those reported by Mogenet and Rolland (1995). Moreover, to calculate the scores for the global scales on the basis of the
standardized scores on the primary factors implies a loss
of information that could also account for this intercorrelation. Each domain of the NEO PI-R is made up of 48
items and each item belongs to only one dimension. Each
dimension is made up of six subscales. Thus, the 48 items
are divided into six groups of eight items each measuring
a specific trait or facet. This legible structure lends itself
well to a validation based on a factorial approach.
Cattells theoretical structure in terms of factors and
Costa and McCraes more recent theoretical structure in

J. Rossier et al.: A Comparison of the NEO PI-R and the 16 PF 5

terms of domains show how these authors have built their

inventories. Cattell (1943) was particularly interested in
describing concrete behaviors or traits (the lower part of
the hierarchical structure). Only later did he try to combine traits into global scales to describe the structure of
personality. For Cattell, the hierarchical structure was an
a posteriori way of looking at personality. Costa and McCrae (1985, 1992) first tried to define dimensions and
were only later able to identify subscales. They followed
the top-down method. The hierarchical structure was, for
them, an a priori way of describing personality (Borkenau & Ostendorf, 1990; Johnson, 2000). This method
made it possible to elaborate an instrument that has a
particularly legible and solid structure.

Conclusion
Both inventories roughly measure the same aspects of
personality. The structure of the 16 PF 5 seems less reliable than the structure of the NEO PI-R. This might be
partly due to the method of construction (bottom-up or
top-down). It is interesting to note that the top-down hierarchy of the NEO PI-R is based on the bottom-up hierarchy of the 16 PF. The very legible structure of
NEO PI-R makes the interpretation of a profile much
easier than does the oblique structure of the 16 PF 5. One
advantage of the 16 PF 5 is that the response styles can
be easily assessed (social desirability, defensiveness, and
acquiescence). To conclude, the NEO PI-R seems particularly reliable in assessing the five global dimensions of
personality. Moreover, the top-down method makes it
possible to construct inventories that are remarkably legible and reliable.
Acknowledgments
We thank Les ditions du Center de Psychologie Appliques (ECPA) and particularly Professor Jean-Pierre
Rolland from the University of Paris X who allowed us
to use the French translation of the NEO PI-R. We also
thank Dr. J.-P. Antonietti and Dr. A. Berchtold for their
support for the data analysis.

References
Allport, G.W., & Odbert, H.S. (1936). Trait names: A psycho-lexical study. Psychological Monographs, 47, 1171.
Argentero, P. (1989). Second-order factor structure of Cattells 16
Personality Factor Questionnaire. Perceptual and Motor Skills,
68, 10431047.
Borkenau, P., & Ostendorf, F. (1990). Comparing exploratory and

37

confirmatory factor analysis: A study on the 5-factor model of

personality. Personality and Individual Differences, 11,
515524.
Byravan, A., & Ramanaiah, N.V. (1995). Structure of the 16 PF
fifth edition from the perspective of the five-factor model. Psychological Reports, 76, 555560.
Cattell, H.E.P. (1995). Some comments on a factor analysis of the
16PF and the NEO Personality Inventory-Revised. Psychological Reports, 77, 13071311.
Cattell, H.E.P. (1996). The original Big Five: A historical perspective. Revue Europenne de Psychologie Applique, 46, 514.
Cattell, R.B. (1943). The description of personality: Basic traits
resolved into clusters. Journal of Abnormal and Social Psychology, 38, 476506.
Cattell, R.B. (1945). The description of personality: Principles and
findings in a factor analysis. American Journal of Psychology,
58, 6990.
Cattell, R.B. (1947). Confirmation and classification of primary
personality factors. Psychometrika, 12, 197220.
Cattell, R.B. (1950a). The main questionnaire factors in questionnaire self-estimate material. Journal of Social Psychology, 31,
338.
Cattell, R.B. (1950b). Personality. New York: McGraw Hill.
Cattell, R.B. (1956a). Second-order personality factors in the
questionnaire realm. Journal of Consulting Psychology, 20,
411418.
Cattell, R.B. (1956b). Validation and intensification of the sixteen
personality factor questionnaire. Journal of Clinical Psychology, 12, 204214.
Cattell, R.B. (1957). Personality and motivation structure and
measurement. New York: World Book.
Cattell, R.B., Cattell, A.K., & Cattell H.E. (1993). Sixteen Personality Factor Questionnaire, Fifth Edition. Champaign, IL: Institute for Personality and Ability Testing.
Cattell, R.B., & Cattell, H.E.P. (1995). Personality structure and
the new fifth edition of the 16PF. Educational and Psychological Measurement, 55, 926937.
Cattell, R.B., Eber, H.W., & Tatsuoka, M.M. (1970). The handbook for the Sixteen Personality Factor Questionnaire. Champaign, IL: Institute for Personality and Ability Testing.
Chernyshenko, O.S., Stark, S., & Chan, K.Y. (2001). Investigating
the hierarchical factor structure of the fifth edition of the 16PF:
An application of the Schmid-Leiman orthogonalization procedure. Educational and Psychological Measurement, 2, 290
302.
Conn, S.R., & Rieke, M.L. (Eds.). (1994). The 16PF Fifth Edition
technical manual. Champaign, IL: Institute for Personality and
Ability Testing, Inc.
Costa, P.T., & McCrae, R.R. (1976). Age differences in personality
structure: A cluster analytic approach. Journal of Gerontology,
31, 564570.
Costa, P.T., & McCrae, R.R. (1985). The NEO Personality Inventory manual. Odessa, FL: Psychological Assessment Resources.
Costa, P.T., & McCrae, R.R. (1992). Revised NEO Personality
Inventory (NEO PI-R) and NEO Five-Factor Inventory (NEOFFI) professional manual. Odessa, FL: Psychological Assessment Resources.
De Raad, B. (2000). The Big Five personality factors; The psycholexical approach to personality. Gttingen: Hogrefe & Huber.
Digman, J.M. (1990). Personality structure: Emergence of the
five-factor model. Annual Review of Psychology, 41, 417440.
Eysenck, H.J. (1991). Dimensions of personality: 16, 5, or 3?
EJPA 20 (1), 2004 Hogrefe & Huber Publishers

38

J. Rossier et al.: A Comparison of the NEO PI-R and the 16 PF 5

Criteria for a taxonomic paradigm. Personality and Individual

Differences, 12, 773790.
Frew, E.A., & Shaw, R.N. (1999). The relationship between personality, gender, and tourism behavior. Tourism Management,
20, 193202.
Goldberg, L.R. (1992). The development of markers for the BigFive factor structure. Psychological Assessment, 4, 2642.
Hofer, S.M., Horn, J.L., & Eber, H.W. (1997). A robust five-factor
structure of the 16PF: Strong evidence from independent rotation and confirmatory factorial invariance procedures. Personality and Individual Differences, 23, 247269.
Holland, J.L. (1973), Making vocational choices: A theory of careers. Englewood Cliffs, N.J.: Prentice-Hall.
Horn, J.L. (1965). An empirical comparison of methods for estimating factor scores. Educational & Psychological Measurement, 25, 313322.
John, O.P., Angleitner, A., & Ostendorf, F. (1988). The lexical
approach to personality: A historical review of trait taxonomic
research. European Journal of Personality, 2, 171203.
Johnson, J.A. (2000). Predicting observers ratings of the Big Five
from the CPI, HPI, and NEO-PI-R: A comparative validity
study. European Journal of Personality, 14, 119.
Krug, S.E., & Johns, E.F. (1986). A large-scale cross-validation of
second-order personality structure defines by the 16 PF. Psychological Reports, 59, 683693.
McCrae, R.R., Zonderman, A.B., Costa, P.T., Bond M.H., &
Paunonen, S.V. (1996). Evaluating replicability of factors in
the Revised NEO Personality Inventory: Confirmatory factor
analysis versus procrustes rotation. Journal of Personality and
Social Psychology, 70, 552566.
McKenzie, J., Tindell, G., & French, J. (1997). The great triumvirate: Agreement between lexically and psycho-physiologically
based models of personality. Personality and Individual Differences, 22, 269277.
Meyer de Stadelhofen, F., Rossier, J., Rigozzi, C., Zimmermann,
G., & Berthoud, S. (submitted). Validation de la version franaise de linventaire Rationnel-Experientiel (REI) et application au tabagisme [Validation of a French version of the Rational-Experiential Inventory and its application to the study of
tobacco smoking].
Mogenet, J.-L., & Rolland, J.-P. (1995). 16 PF 5 de R.B. Cattell.
Paris: ditions du Center de Psychologie Appliques.
OConnor, B.P. (2000). SPSS and SAS programs for determining
the number of components using parallel analysis and Velicers
MAP test. Behavior Research Methods, Instruments, & Computers, 32, 396402.
Ormerod, M.B., McKenzie, J., & Woods, A. (1995). Final report
on research relating to the concept of five separate dimensions
of personality or six including intelligence. Personality and
Individual Differences, 18, 451461.
Peabody, D., & Goldberg, L.R. (1989). Some determinants of fac-

EJPA 20 (1), 2004 Hogrefe & Huber Publishers

tor structures from personality-trait descriptors. Journal of

Personality and Social Psychology, 57, 552567.
Raykov, T. (1998). On the use of confirmatory factor analysis in
personality research. Personality & Individual Differences, 24,
291293.
Rolland, J.-P., & Mogenet, J.-L. (1996). lments de validit des
dimensions primaires de ladaptation franaise de linventaire
16 PF 5 [Evidence of the primary dimensions of the 16PF5
Franch form]. Revue Europenne de Psychologie Applique,
46, 2530.
Rolland, J.-P., Parker, W.D., & Stumpf, H. (1998). A psychometric
examination of the French translations of the NEO-PI-R and
NEO-FFI. Journal of Personality Assessment, 71, 269291.
Rolland, J.-P., & Petot, J.M. (1998). NEO PI-R, Inventaire de Personnalit-Rvis [Manual of the NEO PI-R, Franch adaptation]
Paris: ditions du Center de Psychologie Applique.
Rossier, J., Rigozzi, C., & Berthoud, S. (2002). Validation de la
version franaise de lchelle de contrle de Levenson (IPC),
influence de variables dmographiques et de la personnalit
[Validation of the French translation of Levensons locus of
control scale (IPC), influence of demographic variables and
personality]. Annales Mdico-Psychologiques, 160, 138148.
Rossier, J., Wenger, S., & Berthoud, S. (2001). Validation de la
version franaise du NEO PI-R et influence de lge, du sexe
et de la profession [Internal validation of the French version of
the NEO PI-R and impact of age, gender, and profession]. Psychologie et Psychomtrie, 22, 5982.
Saville, P., & Blinkhorn, S. (1981). Reliability, homogeneity and
the construct validity of Cattells 16PF. Personality & Individual Differences, 2, 325333.
Schmit, M.J., & Ryan, A.M. (1993). The Big Five in personnel
selection: Factor structure in applicant and nonapplicant populations. Journal of Applied Psychology, 78, 966974.
Trull, T.J., & Geary, D.C. (1997). Comparison of the Big-Five
Factor structure across samples of Chinese and American
adults. Journal of Personality Assessment, 69, 324341.

Jrme Rossier
Institute of Psychology
University of Lausanne
BFSH 2 Dorigny
CH-1015 Lausanne
Switzerland
Tel. +41 21 692 3272
Fax +41 21 692 3265
E-mail Jerome.Rossier@ip.unil.ch