Exploring the Cross-Cultural Generalizability of the Five-Factor Model of Personality

Abstract

In this study, the authors examine the generalizability of the Five-Factor Model of personality to the Romanian population by describing the translation and validation of the Romanian version of the Revised NEO Personality Inventory (NEO PI-R). Using data from five samples and multiple sources (self-reports, other-reports, and school records) and over two thousand participants, they examined the internal consistency, test−retest reliability, factor structure, self−other agreement, and correlations with age, gender, education, and academic performance. Construct-related validity evidence was obtained by examining the correlations of the NEO PI-R with the Big Five Questionnaire and Big Five Adjectives. The results suggest that the Five-Factor Model generalizes to the Romanian cultural context and that the Romanian NEO PI-R has sound psychometric properties comparable with normative samples in America and elsewhere.

Keywords

personality five-factor model NEO PI-R Romania psychometrics

Although not without its critics (e.g., Block, 1995), the Five-Factor Model (FFM) of personality is the most established and validated model of personality (Caspi, Roberts, & Shiner, 2005; John, Naumann, & Soto, 2008; John & Srivastava, 1999). The five factors are Neuroticism (N), Extraversion (E), Openness to Experience (O), Agreeableness (A), and Conscientiousness (C). The FFM structure provides a comprehensive mapping of personality traits (Goldberg, 1993; John & Srivastava, 1999; Ozer & Reise, 1994). The five factors have been shown to predict a variety of important outcomes in clinical, organizational, and research settings (e.g., Ozer & Benet-Martinez, 2006; Roberts, Kuncel, Shiner, Caspi, & Goldberg, 2007).

The goal of the current article is to add to the personality literature by examining the FFM in an understudied culture: Romania. We describe the comprehensive development and validation of the Romanian version of the Revised NEO Personality Inventory (NEO PI-R). The NEO PI-R was developed by Costa and McCrae (1992) to operationalize the FFM (Digman, 1990). The NEO PI-R is considered the most prominent FFM questionnaire (Hofstee, 2003). For each of the five broad personality domains, the NEO PI-R contains six facets measured by eight-item scales. Compared with broad markers of the domains of the FFM (e.g., Goldberg, 1992; John, Donahue, & Kentle, 1991), the facet approach allows a more precise examination of individual differences within domains (Piedmont, 1998). The FFM structure has been replicated across cultures and languages using both self and observer ratings (McCrae & Terracciano, 2005; Roland, 2002). However, Romanian samples are not included in any of the major cross-cultural studies focused on examining the FFM structure (e.g., McCrae & Terracciano, 2005; Roland, 2002). Given that the FFM has been proposed as a universal feature with a stable structure across languages and cultures (Allik & McCrae, 2004; McCrae & Costa, 1997), it is important to evaluate it in a wide number of cultures.

A Brief Note on Romania and Psychology in Romania

Romania has a population of around 22 million and is located in South-Eastern Europe with borders along the Ukraine, Moldova, Bulgaria, Serbia, and Hungary. Romania is a Latin nation of Roman heritage. The Romanian language is phonetically similar to Italian. Romania is also the only Latin country in which the Orthodox faith is embraced by a large majority of the population. Romania was under communist rule until 1989 and is now transitioning to a democracy and a market economy. During the communist regime, psychology was outlawed as an independent science and was forced to survive in departments of philosophy and educational sciences (for more on the history of psychology in Romania, see Iliescu, Ispas, & Ilie, 2007a, 2007b). As a consequence, education and research in psychology were negatively impacted. Although psychology departments were reinstated following December 1989, the publication of psychological research originating in Romania is still quite limited in international journals (David, Moore, & Domuta, 2002). One possible reason is the lack of validated psychological tests and questionnaires. Therefore, there is a great need to translate and develop psychological tests and questionnaires for use in applied and research settings in Romania. In addition, a new NEO PI-R translation would allow the inclusion of Romanian samples in large-scale cross-cultural studies.

Translating and Adapting Psychological Tests

Several psychometric considerations are used to establish the measurement utility of a test translation (e.g., Paunonen & Ashton, 1998; Piedmont & Chae, 1997). First, the factor structure of the facets should be consistent with their rational placement and should replicate the factor structure from the original normative sample. Second, each scale should be internally consistent and have adequate test−retest reliabilities. Third, evidence for construct validity should be provided. We examine the correlations between NEO PI-R and the Big Five Questionnaire (BFQ; Caprara, Barbaranelli, & Borgogni, 2005) and Big Five Adjectives (BFA; Caprara, Barbaranelli, &Steca, 2002). Fourth, correlations with outside variables should be examined. Based on previous research, we examined the correlations between the NEO PI-R and demographic variables (i.e., participant age, sex, and education) and using a predictive design, the correlations between the NEO PI-R and academic performance.

In addition to exploring the FFM structure in a new culture, our study makes three further contributions to the personality literature. First, we examined the associations between NEO PI-R and the BFQ and BFA. Second, although the NEO PI-R is generally accepted as the canonical FFM, a common criticism of FFM in general, and the NEO PI-R in particular, is the failure to show good fit when robust approaches to construct validity such as confirmatory factor analysis (CFA) are applied (Church & Burke, 1994; McCrae, Zonderman, Costa, Bond, & Paunonen, 1996). This is linked to the ongoing debate regarding the higher-order structure and the number of facets of the FFM (e.g., Ashton, Lee, Goldberg, & de Vries, 2009; Perugini & Gallucci, 1997; Rushton & Irwing, 2009). We employ a newly proposed method, exploratory structural equation modeling (ESEM; Asparouhov & Muthén, 2009), to examine the factor structure of the NEO PI-R. ESEM is a more flexible, less restrictive approach that allows both non-zero loadings of items and facets on non-target factors (which is common in personality data), and the assessment of fit using traditional SEM indices (Asparouhov & Muthén, 2009; Furnham, Guenole, Levine, & Chamorro-Premuzic, 2013; Marsh et al., 2010). Third, we integrate the results from Romania in the wider cross-cultural personality literature by comparing them with NEO PI-R data from other cultures (e.g., Allik & McCrae, 2004; Allik et al., 2009; Costa, Terracciano, & McCrae, 2001)

Method

Participants and Procedure¹

Sample 1

The sample consisted of 2,200 participants (50% female) with an age range between 14 and 62 (M = 30.30, SD = 11.61). All participants in this sample and the other four samples were Caucasian. In Romania, there are no Human Subject Boards to approve research studies; still, common ethical practices were observed: The interviewers were trained in ethical data collection; participation was voluntary and anonymous; all participants signed an informed consent form and were offered the option to receive a one-page summary of their results. The participants were recruited by professional interviewers, most of them with degrees in psychology. The interviewers were extensively trained and had at least 2 years of experience in field data collection. The participants filled out the NEO PI-R on the spot or the interviewers stopped by at a later time to pick up the filled out questionnaires.

Sampling was done based on a stratified sampling procedure, the strata being built on the eight geographically distinct regions in Romania, from both rural and urban areas, with each stratum sampled as an independent sub-population (random sampling). Most participants were administered the NEO PI-R individually, but for 211 cases, two members of the household were administered the measure at the same time. A total of 240 questionnaires were removed from the analysis due to excessive missing data (more than 40 unanswered items) and unusual patterns in their responses (e.g., answering strongly agree or strongly disagree to every item).

Sample 2

In order to assess the test-retest reliability of the Romanian NEO PI-R, we recruited 126 graduate students (57 women) from a Romanian university. Their age ranged between 19 and 26 (M = 22.06, SD = 2.41). They completed the NEO PI-R twice, two weeks apart. The participants were recruited from four different graduate programs during class time; they completed the questionnaires in groups of 5-20 and were compensated with course credit points.

Sample 3

To assess self−other agreement, we recruited 88 students (47 women) from a Romanian university. Their age ranged between 19 and 25 (M = 20.26, SD = 0.93). In addition to completing self-reports of the NEO PI-R, they were also rated by a parent (40 mothers and 48 fathers, the parents’ age ranged from 39 to 62, M = 47.41, SD = 5.10). The Form R of the NEO PI-R is the observer-evaluation form and has exactly the same items as Form S (self-report), but worded in such a way as to facilitate assessment of another person than the respondent (items are written in the third person). The questionnaire was accompanied by a letter from the researcher, asking the observer to evaluate his or her child (the actual participants in the study) and emphasizing the importance of accurate ratings, of privacy in completing the task and urging him or her not to confer in any way with the target person when completing the assessment. The parents returned the questionnaires directly to the researchers.

Sample 4

To obtain evidence for the construct-related validity of the Romanian NEO PI-R, we recruited 323 participants (187 women) from a community sample. The participants’ age ranged between 18 and 62 (M = 34.37, SD = 10.80). They completed the Romanian NEO PI-R and Romanian versions of the BFQ and BFA.

Sample 5

To examine the criterion-related validity of the Romanian NEO PI-R, we collected data from 122 undergraduate students (62 women, mean age 24.40, SD = 4.74) from a Romanian university. The participants took the NEO PI-R at the beginning of the fall semester, and gave us permission to collect their academic performance data from school records at the end of the spring semester, approximately 9 months later.

Instruments

NEO PI-R

The NEO PI-R (Costa & McCrae, 1992) is a 240-item questionnaire specifically designed to measure the five domains of the FFM of personality. Each of the five domains is measured by 48 items, grouped in six 8-item facet scales. The NEO PI-R was designed for use with individuals over the age of 18. A 5-point Likert response scale was used ranging from 0 (strongly disagree) to 4 (strongly agree).

BFQ

The BFQ (Caprara et al., 2005) is a 132-item questionnaire that covers the five major domains: Energy, Amicability, Conscientiousness, Emotional Stability, and Openness. Each of the five domains is measured with 24 items and is further split into two facet scales with 12 items each. A 12-item Lie scale is also included. A 5-point Likert response scale is used ranging from 1 (strongly disagree) to 5 (strongly agree).

BFA

The BFA (Caprara, Barbaranelli, & Steca, 2002) is a 135-item adjective list. The BFA covers the same domains as BFQ with each domain split into two facet scales. Each facet scale contains 10 adjectives; each domain scale contain the 20 adjectives of its respective facets and 5 supplementary adjectives. A 10-item Lie scale is also included. A 7-point Likert response scale is used ranging from 1 (extremely uncharacteristic for me) to 7 (extremely characteristic for me).

Translation Procedure

The cultural adaptation process of the NEO PI-R in Romania followed closely the guidelines for test translation and adaptation issued by the International Testing Commission (ITC; 2008). Multiple rounds of translation and back translation were conducted using professional translators and bilingual PhD-level psychologists experienced in test translation. The back-translations were reviewed by one of the NEO PI-R authors (Robert McCrae), who gave extensive feedback. After each translation, we collected data from pilot samples (three samples, different from the samples included in the current study, with a total number of participants of around 950) and examined internal consistency estimates, item-total correlations, and factor structures. Similar procedures were followed for the BFQ and BFA (Caprara, Barbaranelli, Borgogni, et al., 2008; Caprara, Barbaranelli, Steca, et al., 2008)

Results²

Normative Data

Table 1 presents the descriptive statistics for the normative sample, both combined and separated by gender. At the item level, the means ranged from 1.19 to 3.16 and only five items had means higher than 3. Only four items had a standard deviation lower than .80 (range = 0.74 to 1.21) indicating that restriction of range was not a problem.

Table 1.

Descriptive Statistics, Gender Differences, Internal Consistencies, Test−Retest Reliabilities, and Self−Other Agreement.

	Men		Women		Combined
NEO-PI R scale	M	SD	M	SD	M	SD	S	K	d	α	Test−retest	Self−other
Domains
N	78.67	18.83	90.97	20.41	84.82	20.57	.04	−.07	.63	.91	.79	.46
E	109.57	16.17	108.15	16.97	108.86	16.59	.01	.26	−.09	.85	.83	.42
O	104.66	12.85	110.65	16.68	107.65	15.18	.40	.56	.40	.83	.83	.44
A	111.12	16.83	113.76	17.55	112.44	17.24	−.05	.14	.15	.85	.78	.44
C	121.65	20.58	118.84	20.13	120.25	20.40	−.12	.05	−.14	.90	.73	.50
N facets
N1: Anxiety	14.40	4.44	16.57	5.08	15.49	4.89	.07	.11	.45	.77	.82	.55
N2: Angry Hostility	13.06	4.39	15.21	4.60	14.14	4.62	.12	−.21	.48	.72	.77	.38
N3: Depression	12.71	4.72	14.94	5.41	13.83	5.20	.28	−.03	.44	.78	.77	.56
N4: Self-Consciousness	14.53	3.96	16.77	4.46	15.65	4.37	.09	.25	.53	.68	.71	.28
N5: Impulsiveness	14.04	4.23	15.31	4.48	14.68	4.40	.07	−.06	.29	.68	.74	.43
N6: Vulnerability	9.92	4.59	12.16	4.60	11.04	4.73	.34	.24	.49	.74	.75	.44
E facets
E1: Warmth	21.49	4.05	21.50	4.23	21.49	4.14	−.35	.28	.00	.71	.81	.53
E2: Gregariousness	18.03	4.73	17.42	4.85	17.73	4.80	−.09	.01	−.13	.68	.72	.46
E3: Assertiveness	16.90	4.13	15.65	4.62	16.27	4.43	.05	−.17	−.29	.74	.79	.39
E4: Activity	17.77	3.91	18.65	4.21	18.21	4.09	.09	.17	.22	.63	.75	.45
E5: Excitement-Seeking	16.71	4.52	16.07	4.67	16.39	4.60	.14	.06	−.14	.70	.67	.16
E6: Positive Emotions	18.68	3.99	18.86	4.58	18.77	4.29	−.14	.28	.04	.73	.78	.56
O facets
O1: Fantasy	15.98	3.91	17.27	4.58	16.63	4.31	.33	.12	.15	.74	.69	.48
O2: Aesthetics	17.27	4.61	19.44	5.15	18.36	5.01	−.11	−.02	.44	.75	.81	.35
O3: Feelings	18.42	3.65	20.54	4.08	19.48	4.01	−.04	−.15	.55	.63	.81	.56
O4: Actions	15.09	3.48	15.85	3.54	15.47	3.53	.08	.07	.22	.55	.81	.13
O5: Ideas	18.97	4.43	18.41	4.89	18.69	4.67	.11	.05	−.12	.74	.71	.50
O6: Values	18.93	3.73	19.13	3.69	19.03	3.71	−.01	.06	.05	.63	.73	.43
A facets
A1: Trust	18.22	4.49	18.09	4.82	18.16	4.66	−.05	−.09	−.03	.75	.79	.48
A2: Straightforwardness	18.78	5.00	19.91	5.33	19.35	5.20	−.04	−.31	.22	.70	.74	.46
A3: Altruism	20.76	4.53	21.69	4.45	21.22	4.52	−.43	.12	.21	.71	.80	.53
A4: Compliance	17.05	4.18	16.60	4.29	16.83	4.24	−.08	.06	−.11	.57	.79	.54
A5: Modesty	16.79	4.37	17.14	4.57	16.96	4.48	−.11	−.31	.08	.72	.77	.29
A6: Tender-Mindedness	19.52	3.99	20.33	4.02	19.92	4.02	−.17	.10	.20	.57	.77	.39
C facets
C1: Competence	20.69	4.31	19.65	3.93	20.17	4.16	−.11	.21	−.25	.66	.82	.44
C2: Order	19.31	4.46	19.04	4.57	19.18	4.52	−.08	−.03	−.06	.64	.91	.47
C3: Dutifulness	21.99	4.59	22.58	4.58	22.28	4.59	−.35	.42	.13	.69	.89	.51
C4: Achievement Striving	20.05	4.30	20.13	4.32	20.09	4.31	−.15	−.02	.02	.65	.89	.64
C5: Self-Discipline	20.78	4.55	20.22	5.13	20.50	4.85	−.20	−.10	−.12	.72	.88	.49
C6: Deliberation	18.82	4.48	17.22	5.01	18.02	4.82	−.22	.19	−.34	.71	.93	.39

Note. n = 2,200, data from Sample 1. S = skewness (SE =.05); K = kurtosis (SE = .10); d = Cohen’s d. Positive values for d indicate higher scores for women. NEO PI-R = Revised NEO Personality Inventory; N = Neuroticism; E = Extraversion; O = Openness to Experience; A = Agreeableness; C = Conscientiousness.

Skewness and kurtosis values are presented in Table 1.To assess the normality of the distribution, we ran the Kolmogorov-Smirnov (K-S) test at both domain and facet level. The results of the K-S tests were statistically significant indicating departures from normality. However, this is not surprising given that the K-S test is very sensitive to sample size (Field, 2009). We supplemented the K-S test with a visual inspection of the Q-Q plots. The plots indicate no problems with normality at the domain level, and only minor departures at the facet level with a slightly positively skewed distribution for N3 (Depression) and N6 (Vulnerability) and a slightly negatively skewed distribution for C3 (Dutifulness).

Age and Gender Differences

Gender differences at the domain and facet level are also presented in Table 1. Using the entire Sample 1, at the domain level, women scored higher than men on N (d = .63), O (d = .40), and A (d = .15), while men scored slightly higher than women on E (d = −.09) and C (d = −.14). At the facet level, women scored higher than men on all the N facets (d range between .29 for Impulsiveness to .53 for Self-Consciousness). For E, there were mixed findings, with women scoring higher on Activity and men scoring higher on Assertiveness, Excitement-seeking, and Gregariousness. For O, women scored higher than men on all facets except Ideas. On A, women scored higher on Straightforwardness, Altruism, Tender-Mindedness, and Modesty, and there were no differences on Trust. For C, there were mixed findings, with men scoring higher overall than women mostly due to differences in Deliberation and Competence.

Table 2 presents the correlations between the NEO PI-R domains and demographic variables, and the domain partial correlations while controlling for age and gender. Age was negatively related to N (r = −.17, p< .01) and with E (r = −.08, p< .01) and positively related to A (r = .36, p< .01) and C (r = .30, p< .01).

Table 2.

Correlations Between NEO PI-R Domains and Demographic Variables.

	N	E	O	A	C	Age	Sex	Education
N	—	−.37 (−.39)	−.05* (−.11**)	−.29 (−.29)	−.55 (−.54)	−.17**	.30**	−.15**
E		—	.40 (.42)	.11 (.16)	.31 (.35)	−.08**	−.04*	.07**
O			—	.12 (.12)	.14 (.17)	−.02	.20**	.21**
A				—	.45 (.39)	.36**	.08**	.12**
C					—	.30**	−.07**	.24**
Age						—	−.03	.47**
Sex							—	−.09**
Education								—

Note. Gender is coded 1 = male and 2 = female. n = 2200, data from Sample 1. Values in parentheses are partial correlations controlling for age and gender. NEO PI-R = Revised NEO Personality Inventory; N = Neuroticism; E = Extraversion; O = Openness to Experience; A = Agreeableness; C = Conscientiousness.

p< .05. **p< .01.

Factor Analyses

The 30 facets of the Romanian NEO PI-R were analyzed using principal components analysis. Table 3 shows the results of the varimax-rotated five-factor solution. The five-factor solution explained 52.80% of the variance, at similar levels with other European translations (e.g., Gorostiaga et al., 2011; Rolland. Parker, & Stumpf, 1998; Terracciano, 2003) and the 56.9% reported by McCrae and Terracciano (2005) in a combined sample from 50 countries. The N, O, and C factors had their highest loadings on their expected factors. E and A had a total of three facets with their highest loadings on different factors, E3: Assertiveness on the A factor, E4: Activity, and A3: Altruism on the C factor. E and A are usually the factors that are not easily identified in NEO PI-R translations (Terracciano, 2003).

Table 3.

Factor loadings of the Romanian version of the NEO PI-R.

	Varimax-rotated component					Procrustes-rotated component
	I	II	III	IV	V	N	E	O	A	C	Facet congruence
N1	−.09	.78	.06	−.07	.01	.78	−.03	−.05	.07	−.09	.99
N2	−.14	.63	−.37	−.23	.02	.66	−.05	−.02	−.40	−.11	.99
N3	−.30	.75	.05	−.13	−.03	.75	−.10	−.09	.02	−.30	.99
N4	−.06	.71	.09	−.14	−.01	.71	−.11	−.06	.07	−.06	.99
N5	−.30	.48	−.39	.11	.12	.49	.26	.07	−.33	−.29	.98
N6	−.49	.60	.04	−.19	−.09	.61	−.18	−.13	−.03	−.48	.98
E1	.27	−.09	.07	.71	.15	−.14	.66	.11	.32	.21	.99
E2	.06	−.24	−.03	.68	.06	−.29	.64	.03	.19	.01	.97
E3	.22	−.39	−.48	.32	.06	−.40	.45	.07	−.35	.22	.96
E4	.44	−.05	−.30	.35	.17	−.06	.46	.15	−.13	.43	.97
E5	−.14	−.01	−.44	.40	.14	−.02	.52	.11	−.29	−.15	.99
E6	.06	−.23	−.25	.50	.33	−.24	.56	.31	−.07	.03	.89
O1	−.16	.12	−.08	.06	.67	.17	.13	.65	−.04	−.17	.96
O2	.19	.13	.07	.18	.56	.15	.20	.53	.16	.16	.95
O3	.19	.30	−.12	.27	.57	.32	.36	.52	.02	.17	.99
O4	−.28	−.11	−.17	.21	.45	−.09	.26	.44	−.11	−.29	.86
O5	.24	−.18	−.03	.09	.52	−.15	.13	.53	.02	.23	.94
O6	.15	−.25	.11	−.12	.63	−.19	−.11	.66	.08	.15	.81
A1	.09	−.23	.36	.58	.04	−.29	.41	.02	.52	.03	.94
A2	.38	−.10	.59	.02	.12	−.11	−.16	.13	.59	.34	.92
A3	.49	.07	.39	.42	.22	.04	.31	.19	.55	.43	.89
A4	.12	−.21	.72	.14	−.08	−.25	−.12	−.07	.72	.07	.98
A5	−.11	.13	.57	−.01	−.06	.11	−.20	−.07	.53	−.14	.96
A6	.25	.20	.45	.41	.06	.15	.27	.02	.59	.19	.93
C1	.66	−.32	−.02	.16	.10	−.32	.18	.12	.07	.65	.99
C2	.69	−.08	.06	−.01	.02	−.08	.00	.03	.10	.68	.94
C3	.78	−.11	.24	.11	.11	−.12	.06	.12	.32	.75	.98
C4	.71	−.08	−.06	.17	.14	−.08	.22	.14	.05	.70	.97
C5	.75	−.34	.10	.09	.03	−.35	.07	.06	.16	.74	.97
C6	.54	−.28	.25	−.01	−.31	−.31	−.11	−.28	.25	.53	.91

Note. n = 2,200, data from Sample 1. Loadings > |.40| are in bold. NEO PI-R = Revised NEO Personality Inventory; N = Neuroticism; E = Extraversion; O = Openness to Experience; A = Agreeableness; C = Conscientiousness.

In line with previous studies examining the factor structure of translated versions of the NEO PI-R (e.g., Rolland et al., 1998; Terracciano, 2003), we also conducted a Procrustes rotation (McCrae et al., 1996) in which the varimax solution was rotated to a target matrix based on the U.S. normative sample (Costa & McCrae, 1992). The results are shown in Table 3. Every facet had its highest loading on its intended factor. The factor congruence coefficients were .98 for N; .96 for E, A, and C; and .90 for O. The facet congruence coefficients are presented in the last column of Table 3. With the exception of O6: Values, all the facet-level congruence coefficients were higher than .89 with 26 facets out of 30 higher than .94. The Romanian NEO PI-R closely resembles the factor structure of the American normative structure.

Using data from Sample 1, the results of traditional CFA show a poor fit to the data: χ²(320) = 5,720.57, p< .001, root mean square error of approximation (RMSEA) = 0.088, 90% confidence interval (CI) for RMSEA = [0.086, 0.090], comparative fit index (CFI) = 0.77. However, the results of the ESEM analyses implemented using Mplus (Muthen & Muthen, 2012) show a marked improvement of the fit indices to acceptable levels: χ²(295) = 1,873.123, p< .001, RMSEA = .049, 90% CI for RMSEA = [0.047, 0.051], CFI = 0.92. The standardized factor loadings are presented in Table 4.

Table 4.

Loading Estimates and Corresponding Standard Errors From the Standardized ESEM.

Facet	Est.	SE	Est.	SE	Est.	SE	Est.	SE	Est.	SE
N1	.81	.11	.04	.11	−.03	.12	.02	.09	.01	.10
N2	.46	.14	.04	.17	.14	.27	−.45	.13	.00	.08
N3	.81	.13	.00	.13	−.04	.12	.00	.08	−.20	.14
N4	.59	.10	−.09	.12	.01	.11	−.05	.10	.01	.09
N5	.31	.13	.24	.18	.21	.26	−.13	.15	−.21	.14
N6	.52	.11	−.07	.12	−.02	.11	−.03	.08	−.43	.12
E1	.03	.05	.38	.16	.05	.10	.49	.13	.17	.13
E2	−.11	.15	.51	.17	−.04	.10	.52	.17	−.01	.07
E3	−.32	.13	.44	.17	.02	.15	−.06	.12	.18	.14
E4	−.04	.11	.22	.21	.16	.17	−.04	.10	.36	.12
E5	−.02	.10	.45	.16	.12	.21	.03	.12	−.09	.14
E6	−.22	.13	.29	.19	.31	.15	.15	.13	.00	.09
O1	−.01	.06	−.09	.17	.60	.13	−.02	.10	−.17	.29
O2	.12	.15	.00	.14	.42	.19	.14	.16	.16	.27
O3	.19	.12	.08	.16	.44	.13	.05	.10	.16	.24
O4	−.12	.11	.07	.14	.32	.12	.02	.08	−.19	.15
O5	−.15	.14	.03	.14	.32	.19	.04	.13	.19	.20
O6	−.20	.12	−.20	.14	.38	.17	−.01	.07	.07	.19
A1	−.12	.13	.11	.18	.01	.11	.58	.13	−.01	.11
A2	−.05	.09	−.50	.19	.12	.26	.29	.20	.33	.21
A3	.18	.12	−.01	.08	.13	.17	.48	.12	.39	.18
A4	−.05	.11	−.26	.21	−.14	.29	.51	.14	.00	.06
A5	.16	.14	−.27	.18	−.07	.18	.28	.15	−.11	.14
A6	.21	.11	−.04	.12	.04	.12	.39	.12	.15	.15
C1	−.19	.11	.07	.12	−.01	.10	.02	.08	.55	.10
C2	.03	.10	−.01	.10	−.12	.17	−.03	.10	.64	.12
C3	.02	.07	−.16	.14	.02	.14	.06	.11	.78	.11
C4	.00	.08	.10	.15	.02	.11	−.01	.09	.63	.10
C5	−.22	.11	−.04	.11	−.11	.17	−.02	.08	.79	.12
C6	−.08	.15	−.01	.14	−.48	.21	.14	.14	.53	.21

Note. n= 2,200, data from Sample 1. Loadings > |.40| are in bold. ESEM = exploratory structural equation modeling; N = Neuroticism; E = Extraversion; O = Openness to Experience; A = Agreeableness; C = Conscientiousness.

Reliability

The internal consistency reliabilities for the domains and facets of the Romanian NEO PI-R are presented in Table 1. At the domain level, Cronbach’s α ranged between .83 and .91. At the facet level, Cronbach’s α ranged between .55 for O4 and .78 for N3 with a median of .71. All these values are comparable with previous translations of the NEO PI-R (e.g., Terracciano, 2003). Test−retest reliability coefficients, computed with data from Sample 2, are also presented in Table 1 and range between .67 for E5 and .91 for C2 with a median of .79. Facets with relatively low internal consistencies (such as the C facets) have large test−retest reliabilities. Item-total correlations at the domain level range between .28 for O and .39 for N. At the facet level, item-total correlations ranged between .26 for O4: Actions and .48 for N3: Depression with a mean of .38 (Mdn = .41, SD = .05).

Self−Other Agreement

Using participants from Sample 3, we examined the self−other agreement. The results are shown in the last column of Table 1. At the domain level, the correlations range from .42 for E to .50 for C with a mean of .45, a value that compares favorably to the .43 mean value obtained between self and peer ratings in the United States (Costa & McCrae, 1992). At the facet level, the correlations range from .13 for O4: Actions to .64 for C4: Achievement striving with a median of .46 (compared to the .34 obtained from data reported in the U.S. manual).

NEO PI-R and BFQ/BFA

The construct validity of the Romanian NEO PI-R was assessed by examining the correlations between the NEO PI-R domains and the BFQ and BFA domains using data from Sample 4. The results are presented in Table 5. The corresponding correlations (e.g., E from NEO PI-R and Energy from BFQ/BFA) were all above .54; all statistically significant at p< .001, providing evidence for the construct validity of the Romanian NEO PI-R.

Table 5.

Correlations Between NEO PI-R and BFQ/BFA.

	Energy	Friendliness	Conscientiousness	Emotional stability	Openness
	BFQ/BFA	BFQ/BFA	BFQ/BFA	BFQ/BFA	BFQ/BFA
	(.86, .92)	(.79, .84)	(.84, .90)	(.90, .91)	(.80, .85)
NEO PI-R N (.90)	−.11/−.06	−.08/−.10	−.17/−.19	−.54*/−.57*	−.17*/−.15
NEO PI-R E (.90)	.54*/.56*	.20*/.16	.03/.10	.12/.13	.19/.18
NEO PI-R O (.84)	.29*/.32*	.18/.17	.20/.18	.16/.22**	.57*/.56*
NEO PI-R A (.85)	.10/.13*	.59*/.63*	.11/.20***	.17/.16	.09/.15**
NEO PI-R C (.91)	.06/.17**	.06/.15**	.63*/.62*	.26*/.30*	.21*/.25*

Note.n = 323. Reliabilities (coefficient alpha) are presented in parentheses. Correlations in bold are convergent validity coefficients. Neuroticism is called Emotional Stability in BFQ/BFA hence the negative correlations with NEO PI-R domain N. NEO PI-R = Revised NEO Personality Inventory; BFQ = Big Five Questionnaire; BFA = Big Five Adjectives; N = Neuroticism; E = Extraversion; O = Openness to Experience; A = Agreeableness; C = Conscientiousness.

p< .05. **p< .01. ***p< .001.

Correlations With Education and Academic Performance

In the Romanian normative sample, educational attainment—measured on a 1 to 5 scale with 1 = basic level (less than 9 years of education) to 5 = graduate education—had the strongest correlations with C (r = .24, p< .01) and O (r = .21, p< .01), similar to previous research (Costa & McCrae, 1992; Terracciano, 2003).

Using data from Sample 5, we found that C (r = .28, p< .01), N (r = −.19, p< .05) and E (r = .18, p< .05) had statistically significant correlations with academic performance measured 9 months later from school records (on a 1-10 scale, M = 7.65, SD = 1.05).

Comparisons With Other Cultures

To integrate the Romanian data into the wider picture of cross-cultural studies using the NEO PI-R (Allik & McCrae, 2004; Allik et al., 2009; Costa et al., 2001; McCrae et al., 1999), we conducted several exploratory analyses using the adult part of Sample 1.

In terms of mean differences, compared to the U.S. normative sample (Costa & McCrae, 1992), Romanians scored higher on N (d = .15) and C (d = .05), lower on O (d = .11) and A (d = .54), and virtually the same on E (d = 0). To compare the correlation matrix obtained for the Romanian sample with the one for the U.S. sample (Costa & McCrae, 1992, pp. 100-101), we used Kendall’s tau-b rank-order correlation coefficient (τ). For the full 30 × 30 facet intercorrelation table τ = .77, p< .001, indicating an acceptable fit of the two correlation structures. We also integrated the Romanian data in Allik and McCrae’s (2004) multidimensional scaling plot. According to these analyses, Romania fits in the upper-left quadrant (high N, low E) close to the Russian, Spain, and Portuguese samples.

The correlations between age and the five domains replicate the general pattern found in other cultures (e.g., Costa & McCrae, 1992; McCrae et al., 1999; Terracciano, 2003) with declines for N, E, and O and increases for A and C. Age correlated negatively with N (r = −.07, p< .01), E (r = −.17, p< .001), and O (r = −.15, p< .001) and positively with A (r = .25, p< .001) and C (r = .15, p< .001).

The biggest differences between the Romanians and the other cultures were found for gender. While women in general tend to score higher on N across cultures, we found higher differences in our Romanian adult sample across all N facets, even for N2 Angry Hostility where we found a larger gender difference (d = .54) compared with U.S. (d = .09; Costa & McCrae, 1992) and combined data from 25 other cultures (d = .19; Costa et al., 2001). Exaggerated differences were also found on the O facets. For O1: Fantasy, the Romanian data show women scoring higher than men (d = .35), which is higher than the differences found in other cultures (d = .06), and the opposite of U.S. data where men scored higher (d = −.16). Another difference emerges on facets where women traditionally score higher than men such as E1: Warmth and the A facets. Although women universally tend to score higher on all of the A facets, this was not the case in the adult Romanian sample where women scored lower on A4: Compliance (d = −.16) and there were no gender differences on A1: Trust and A5: Modesty. A full comparison with U.S. and data from 25 cultures is presented in Table 6.

Table 6.

Gender Differences (d) on the NEO PI-R Facets in Romania, United States, and 25 Other Cultures.

	Romanian Adults	U.S. Adults	Other cultures
N1	0.42	0.40	0.43
N2	0.54	0.09	0.19
N3	0.47	0.24	0.29
N4	0.50	0.30	0.23
N5	0.27	0.23	0.11
N6	0.53	0.44	0.36
Average N	0.46	0.28	0.27
E1	0.05	0.33	0.23
E2	−0.13	0.21	0.14
E3	−0.33	−0.19	−0.27
E4	0.22	0.11	0.11
E5	−0.09	−0.31	−0.38
E6	0.07	0.29	0.16
Average E	−0.04	0.07	0
O1	0.35	−0.16	0.06
O2	0.52	0.34	0.35
O3	0.66	0.28	0.31
O4	0.25	0.19	0.17
O5	−0.12	−0.32	−0.16
O6	0.05	−0.07	0.01
Average O	0.29	0.04	0.12
A1	−0.05	0.19	0.17
A2	0.19	0.43	0.32
A3	0.22	0.43	0.25
A4	−0.16	0.38	0.17
A5	0	0.38	0.22
A6	0.16	0.31	0.28
Average A	0.06	0.35	0.24
C1	−0.27	−0.20	−0.10
C2	−0.12	0.05	0.10
C3	0.17	0	0.13
C4	0.04	0.08	−0.04
C5	−0.09	−0.02	0.04
C6	−0.37	−0.12	−0.06
Average C	−0.11	−0.04	0.01

Note. Positive values indicate higher scores for women. Romanian data are based on the adult part of Sample 1 (n = 1,650). U.S. data (n = 1,000) are based on Costa and McCrae (1992) as reported in Costa, Terracciano, and McCrae (2001). Other cultures data (n = 10,690) are based on 25 cultures as reported in Costa et al. (2001), Table 2, p. 327. NEO PI-R = Revised NEO Personality Inventory; N = Neuroticism; E = Extraversion; O = Openness to Experience; A = Agreeableness; C = Conscientiousness.

Exploratory Analyses: The NEO PI-R in Participants Under 21 Years Old

Although the NEO PI-R was designed for use with adult population, part of Sample 1 consisted of participants under 21 years old. We conducted exploratory analyses for the 14- to 17-years-old group (n = 178) and the 18- to 20-years-old group (n = 372). For the 14- to 17-years group, at the domain level, coefficient alpha ranged between .77 and .88; and between .59 and .81 at the facet level (M = 67.47, Mdn = 65.5, SD = 6.07). For the 18- to 20-years group, alpha at the domain level ranged between .81 and .87, and between .48 and .74 at the facet level (M = 64.43, Mdn = 66, SD = 7.58). We also conducted a Procrustes rotation factor analysis for the under 21-years-old group (n = 550). The factor congruence coefficients at the domain level range are .90 for O, .95 for E, and .97 for N, A, and C. At the facet level, only three factor congruence coefficients are below .91: .86 for O6, .87 for O4, and .89 for A5.

Exploratory Analyses: Inter-Regional Differences

With the exception of Allik et al. (2009), personality research has mostly ignored potential within country differences. Using the adult part (over 21 years old) of Sample 1 (n = 1,650), we conducted ANOVAs to compare the eight geographically distinct areas on the NEO Domain and facets. None of the differences were statistically significant. We also compared rural (N = 719) and urban residents (N = 931). The only differences were at facet level with urban residents scoring higher on N4: Self-Consciousness (d = .11) and O3: Feelings (d = .10) and lower on C2: Order (d = −.10).

Discussion

Summary of Findings

The results of our study show that the FFM generalizes to the Romanian cultural context. Factor analyses showed that the factor structure that was found in the American normative sample replicated in our sample in which the Romanian version of the NEO PI-R was used. Internal consistency, test−retest reliabilities, and self−other agreement were all comparable to and in some cases better than the values found in other studies using translated versions of the NEO PI-R. The intercorrelations among the factor domains are also in line with previous research. Evidence for the construct validity was presented by examining the correlations between corresponding domains from the Romanian NEO PI-R and the BFQ and BFA. In terms of criterion validity, we found that C was the best predictor of educational attainment and academic performance.

Implications for the Cross-Cultural Adaptation of FFM Inventories

When compared to the results reported for other cultural adaptations of the NEO PI-R, these results have some particularities. Based on the domain scale reliabilities, N is the most easily identified factor in the Romanian context, with a reliability of .91 and with facet reliabilities in the range .68 to .77. The factor congruence coefficient also outlines N (.98) as the factor which is most similar to the original U.S. N. Openness to experience is identified in the Romanian version of the NEO PI-R as the weakest factor. It has the weakest reliability (.83 for the domain scale), the smallest mean item-total correlation (.28), and has facet scales with reliabilities as low as .55 (O4: Actions). Also, the factor congruence coefficient for O is the weakest of all five factors (.90). E, A, and C are placed between these two extremes, with reliabilities in the mid .80 interval and factor congruence coefficients of .96 to .98, showing a good fit with the original model.

These results are in line with the results reported by Rolland et al. (1998) for the French adaptation of the NEO PI-R, and by Gorostiaga et al. (2011) for the Spanish and Basque adaptation. Other research, summarized by Terracciano (2003), has shown that E and A are usually the factors that are not easily identified in NEO PI-R translations when using Varimax rotations; however, this was not the case for the Romanian version of the inventory.

When integrating the Romanian data into the wider cross-cultural personality literature, we found both similarities (age trends) and differences (gender). However, we urge caution in the interpretation of these differences given that the Romanian data are based on a national representative sample which is not the case for most of the other cultures data from Costa et al. (2001).

Strengths and Limitations

Several strengths of the current study should be outlined. First, we used data collected from five different samples, one of them a national representative sample. A second strength resides in the fact that the study does not only report on the construct validity of the NEO PI-R but also on its criterion validity, as related to age, gender, education, and academic performance. Third, we contribute to the literature by examining the confirmatory factor structure of NEO PI-R using ESEM. This is the first translation of the NEO PI-R and only the second study following Furnham et al. (2013) that employed ESEM to test the factor structure of the NEO PI-R.

Also, three limitations of the current study ought to be mentioned. First, some of the samples included are small (e.g., Sample 3 included only 88 participants, and Sample 5 included only 122 participants). This limitation is in many aspects counterbalanced by the large volume of the first sample, which is a national representative sample of 2,200 participants. Second, some of the samples used include student participants. While this is warranted by the scope of the study for Sample 5, where we examined academic performance as a criterion, Samples 2 and 3 also include students. Third, Samples 1 to 3 were previously reported in the Romanian manual for NEO PI-R (Iliescu et al., 2008). However, the Romanian manual has limited availability and the fact that it is written in Romanian limits its usefulness to international researchers.

Directions for Future Research

While our findings are encouraging, there are areas that require further research. Chief among these are to evaluate the applied utility and validity of the Romanian NEO. Doing so might entail investigating samples from clinical and organizational populations. For example, the finding that conscientiousness is a significant predictor of job performance (Barrick, Mount, & Judge, 2001) could be investigated in Romanian work organizations using the version of the NEO PI-R that was validated in the current study. Also, additional research is needed using the NEO-Five-Factor Inventory (NEO-FFI), the short version of the NEO PI-R. The NEO-FFI is a 60-item version of the NEO PI-R that can be used in situations where less respondent time is available. By sampling the 60 NEO-FFI items out of the 240 item pool of the NEO PI-R on the data from the first sample, convergence of NEO-FFI scales with NEO PI-R factor scales is large: .93, .90, .94, .88, and .89 for N, E, O, A, and C, respectively. Still, due to its simpler structure and potential usage in clinical and occupational screening settings (due to the small number of items), the NEO-FFI should be investigated separately in the Romanian context. The incremental validity of the Romanian NEO PI-R over other individual difference constructs when predicting various clinical and organizational criteria should also be examined. Future research should also examine test−retest reliability and self−other agreement in younger samples.

The evidence presented here suggests that the FFM structure generalizes to the Romanian cultural context and that the NEO PI-R, a valid psychological measure of personality, can be used to research and better understand the role of personality in this underrepresented country and to facilitate cross-cultural research.

Footnotes

Acknowledgements

We thank Prof. Jüri Allik and two anonymous reviewers for their helpful comments and suggestions. We gratefully acknowledge the help and guidance received from Dr. Robert R. McCrae throughout the translation and adaptation process.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Dragos Iliescu is the managing director of Testcentral, the Romanian publisher of the NEO PI-R; the company derives income from selling the Romanian version of the test.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Allik

McCrae

R. R.

(2004). Toward a geography of personality traits: Patterns of profiles across 36 cultures. Journal of Cross-Cultural Psychology, 35, 13-28.

Allik

Mõttus

Realo

Pullmann

Trifonova

McCrae

R. R.

, & 56 Members of the Russian Character and Personality Survey. (2009). Personality traits of Russians from the observer’s perspective. European Journal of Personality, 23, 567-588.

Ashton

M. C.

Lee

Goldberg

L. R.

de Vries

R. E.

(2009). Higher order factors of personality: Do they exist? Personality and Social Psychology Review, 13, 79-91.

Asparouhov

Muthén

(2009). Exploratory structural equation modeling. Structural Equation Modeling, 16, 397-438.

Barrick

M. R.

Mount

M. K.

Judge

T. A.

(2001). Personality and performance at the beginning of the new millennium: What do we know and where do we go next? International Journal of Selection and Assessment, 9, 9-30.

Block

(1995). A contrarian view of the five-factor approach to personality description. Psychological Bulletin, 117, 187-215.

Caprara

G. V.

Barbaranelli

Borgogni

(2005). BFQ–Big Five Questionnaire: Manuale (2nd ed.)[Manual for BFQ-Big Five Questionnaire].Firenze, Italy: O.S. Organizzazioni Speciali.

Caprara

G. V.

Barbaranelli

Borgogni

Pitariu

Vercellino

Iliescu

(2008). Technical manual for the Romanian version of the BFQ Questionnaire. Cluj-Napoca, Romania: Odiseea.

Caprara

G. V.

Barbaranelli

Steca

(2002). BFA–Big Five Adjectives: Manuale[Manual for the BFA-Big Five Adjectives]. Firenze, Italy: O.S. Organizzazioni Speciali.

10.

Caprara

G. V.

Barbaranelli

Steca

Pitariu

Vercellino

Iliescu

(2008). Technical manual for the Romanian version of the Big Five Adjectives. Cluj-Napoca, Romania: Odiseea.

11.

Caspi

Roberts

B. W.

Shiner

R. L.

(2005). Personality development: Stability and change. Annual Review of Psychology, 56, 453-484.

12.

Church

A. T.

Burke

P. J.

(1994). Exploratory and confirmatory tests of the Big Five and Tellegen’s three- and four-dimensional models. Journal of Personality and Social Psychology, 66, 93-114.

13.

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.

14.

Costa

P. T.

Jr. Terracciano

McCrae

R. R.

(2001). Gender differences in personality traits across cultures: Robust and surprisingfindings. Journal of Personality and Social Psychology, 81, 322–331.

15.

David

Moore

Domuta

(2002). Romanian psychology on the international psychological scene: A preliminary critical and empirical approach. European Psychologist, 7, 153-160.

16.

Digman

J. M.

(1990). Personality structure: Emergence of the five-factor model. Annual Review of Psychology, 41, 417-440.

17.

Field

(2009). Discovering statistics using SPSS (3rd ed.). Los Angeles, CA: SAGE.

18.

Furnham

Guenole

Levine

S. Z.

Chamorro-Premuzic

(2013). The NEO Personality Inventory–Revised: Factor structure and gender invariance from exploratory structural equation modeling analyses in a high-stakes setting. Assessment, 20, 14-23.

19.

Goldberg

L. R.

(1992). The development of markers for the Big-Five factor structure. Psychological Assessment, 4, 26-42.

20.

Goldberg

L. R.

(1993). The structure of phenotypic personality traits. American Psychologist, 48, 26-34.

21.

Gorostiaga

Balluerka

Alonso-Arbiol

Haranburu

(2011). Validation of the Basque Revised NEO Personality Inventory (NEO PI-R). European Journal of PsychologicalAssessment, 27, 193-205.

22.

Hofstee

W. K. B.

(2003). Structures of personality traits. In Weiner

I. B.

(Series Ed.), & Millon

Lerner

(Vol. Eds.), Handbook of psychology: Vol. 5.Personality and social psychology (pp. 231-254). Hoboken, NJ: Wiley.

23.

Iliescu

Ispas

Ilie

(2007a). Psychology in Romania. The Psychologist, 20, 34-35.

24.

Iliescu

Ispas

Ilie

(2007b). Industrial-organizational psychology in Romania. The Industrial-Organizational Psychologist, 45, 71-76.

25.

Iliescu

Minulescu

Nedelcea

Ispas

(2008).Technical manual for the NEO PI-R. Cluj-Napoca: Sinapsis.

26.

International Test Commission. (2008). International Test Commission Guidelines for Translating and Adapting Tests. London, England: The British Psychological Society.

27.

John

O. P.

Donahue

E. M.

Kentle

R. L.

(1991). The Big Five Inventory—Versions 4a and 54. Berkeley: Institute of Personality and Social Research, University of California, Berkeley.

28.

John

O. P.

Naumann

L. P.

Soto

C. J.

(2008). Paradigm shift to the integrative Big-Five trait taxonomy: History, measurement, and conceptual issues. In John

O. P.

Robins

R. W.

Pervin

L. A.

(Eds.), Handbook of personality: Theory and research (pp. 114-158). New York, NY: Guilford Press.

29.

John

O. P.

Srivastava

(1999). The Big Five trait taxonomy: History, measurement, and theoretical perspectives. In Pervin

L. A.

John

O. P.

(Eds.), Handbook of personality: Theory and research (2nd ed., pp. 102-138). New York, NY: Guilford Press.

30.

Marsh

H. W.

Lüdtke

Muthén

Asparouhov

Morin

A. J. S.

Trautwein

Nagengast

(2010). A new look at the Big-Five factor structure through exploratory structural equation modeling. Psychological Assessment, 22, 471-491.

31.

McCrae

R. R.

Costa

P. T.

Jr. (1997). Personality trait structure as a human universal. American Psychologist, 52, 509-516.

32.

McCrae

R. R.

Costa

P. T.

Jr. Lima

M. P.

Simoes

Ostendorf

Angleitner

. . . Piedmont

R. L.

(1999). Age differences in personality across the adult lifespan: Parallels in five cultures. Developmental Psychology, 35, 466-477.

33.

McCrae

R. R.

Terracciano

(2005). Universal features of personality traits from the observer’s perspective: Data from 50 cultures. Journal of Personality and Social Psychology, 88, 547-561.

34.

McCrae

R. R.

Zonderman

A. B.

Costa

P. T.

Jr. 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, 552-566.

35.

Muthén

L. K.

Muthén

B. O.

(2012). Mplus user’s guide (7th ed.). Los Angeles, CA: Author.

36.

Ozer

D. J.

Benet-Martinez

(2006). Personality and the prediction of consequential outcomes. Annual Review of Psychology, 57, 401-421.

37.

Ozer

D. J.

Reise

S. P.

(1994). Personality assessment. Annual Review of Psychology, 45, 357-388.

38.

Paunonen

S. V.

Ashton

M. C.

(1998).The structured assessment of personality across cultures. Journal of Cross-Cultural Psychology, 29,150–170.

39.

Perugini

Gallucci

(1997). A hierarchical faceted model of the Big Five. European Journal of Personality, 11, 279-301.

40.

Piedmont

R. L.

(1998). The Revised NEO Personality Inventory: Clinical and research applications. New York, NY: Plenum Press.

41.

Piedmont

R. L.

Chae

J. H.

(1997). Cross-cultural generalizability of the five-factor model of personality: Development and validation of the NEO PI-R for Koreans. Journal of Cross-Cultural Psychology, 28, 131-155.

42.

Roberts

B. W.

Kuncel

N. R.

Shiner

Caspi

Goldberg

L. R.

(2007). The power of personality: The comparative validity of personality traits, socioeconomic status, and cognitive ability for predicting important life outcomes. Perspectives on Psychological Science, 2, 313-345.

43.

Rolland

J.P.

(2002). Cross-Cultural Generalizability of the Five- Factor Model of Personality. In McCrae

R.R.

Allik

(Eds.), The Five-factor model of personality across cultures (pp. 7–28). New York: Kluver/Plenum.

44.

Rolland

J. P.

Parker

W. D.

Stumpf

(1998). A psychometric examination of the French translations of the NEO-PI-R and NEO-FFI. Journal of Personality Assessment, 71, 269-291.

45.

Rushton

J. P.

Irwing

(2009). A General Factor of Person-ality (GFP) from the Multidimentional Personality Questionnaire. Personality and Individual Differences, 47, 571-576.

46.

Terracciano

(2003). The Italian version of the NEO PI-R: Conceptual and empirical support for the use of targeted rotation. Personality and Individual Differences, 35, 1859-1872.

Exploring the Cross-Cultural Generalizability of the Five-Factor Model of Personality

Abstract

Keywords

A Brief Note on Romania and Psychology in Romania

Translating and Adapting Psychological Tests

Method

Participants and Procedure 1

Sample 1

Sample 2

Sample 3

Sample 4

Sample 5

Instruments

NEO PI-R

BFQ

BFA

Translation Procedure

Results 2

Normative Data

Age and Gender Differences

Factor Analyses

Reliability

Self−Other Agreement

NEO PI-R and BFQ/BFA

Correlations With Education and Academic Performance

Comparisons With Other Cultures

Exploratory Analyses: The NEO PI-R in Participants Under 21 Years Old

Exploratory Analyses: Inter-Regional Differences

Discussion

Summary of Findings

Implications for the Cross-Cultural Adaptation of FFM Inventories

Strengths and Limitations

Directions for Future Research

Footnotes

Acknowledgements

Declaration of Conflicting Interests

Funding

Notes

References

Participants and Procedure¹

Results²