Psicologia: Reflexão e Crítica

Validity evidence of two short scales measuring the big five personality factors.
--Manuscript Draft--

Manuscript Number:

Full Title: Validity evidence of two short scales measuring the big five personality factors.

Article Type: Research

Section/Category: Psychological assessment

Funding Information:

Abstract: The objective of this study was to obtain evidence of the convergent and factor validity
of the Reduced Scale of Big Five Personality Factors (ER5FP), with 20 items, and of
the Reduced Inventory of Big Five Personality Factors (IGFP-5R), with 32 items. The
two Brazilian scales were administered to 554 participants aged 16-69 years (M=30.6,
SD=8.6). The measurement model of each instrument was tested using Confirmatory
Factor Analysis. Both scales showed an adequate adjustment of the measurement
model to the data (RMSEA<.06; SRMR<.06) after excluding a number of items.
Correlations between the factors of both instruments were estimated. Strong evidence
of the convergent validity was found for the factors Extraversion, Neuroticism and
Openness to Experience (correlations between .60 and .80); for the factors
Agreeableness and Conscientiousness this evidence was less strong (correlations
between .43 and .48).

Corresponding Author: Alexandre José de Souza Peres Peres, Ph.D.

National Institute for Educational Research (Inep, Brazil)
BRAZIL

Corresponding Author Secondary

Information:

Corresponding Author's Institution: National Institute for Educational Research (Inep, Brazil)

Corresponding Author's Secondary

Institution:

First Author: Jacob Arie laros, PhD

First Author Secondary Information:

Order of Authors: Jacob Arie laros, PhD

Alexandre José de Souza Peres Peres, Ph.D.

Josemberg Moura de Andrade, PhD

Maria Fabiana Damásio Passos, PhD

Order of Authors Secondary Information:

Suggested Reviewers: Cristiano Mauro Assis Gomes, PhD

Professor, Hospital das Clinicas da Universidade Federal de Minas Gerais
cristianomaurogomes@gmail.com
The potential reviewer investigates the subject of the submitted manuscript.

Ricardo Primi, PhD

Professor, Universidade Sao Francisco
rprimi@mac.com
The potential reviewer investigates the subject of the submitted manuscript.

Carlos Henrique Sancineto da Silva Nunes, PhD

Professor, Universidade Federal de Santa Catarina
carloshnunes@mac.com
The potential reviewer investigates the subject of the submitted manuscript.

Opposed Reviewers:

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Cover Letter Click here to download Cover Letter Cover letter.docx

1 Title:

2 Validity evidence of two short scales measuring the big five personality factors

4
5

10

11

12 Authors:

13 Jacob Arie Laros1

14 Alexandre José de Souza Peres2

15 Josemberg Moura de Andrade3

16 Maria Fabiana Damásio Passos4

17

18

1
Universidade de Brasília, Instituto de Psicologia, Departamento de Psicologia Social e do Trabalho, Campus
Universitário Darcy Ribeiro, Plano Piloto, Brasília, DF, Brasil. CEP: 70910900. E-mail: jalaros@gmail.com
2
Corresponding author. Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira (Inep). Setor
de Indústrias Gráficas (SIG), Quadra 04, Lote 32. Brasília, DF, Brasil. CEP: 70610-908. E-mail:
alexandre.peres@gmail.com
3
Universidade de Brasília, Instituto de Psicologia, Departamento de Psicologia Social e do Trabalho, Campus
Universitário Darcy Ribeiro. Plano Piloto, Brasília, DF, Brasil. CEP: 70910900. E-mail:
josemberg.andrade@gmail.com
4
Fundação Oswaldo Cruz. Avenida L3 Norte, Campus Universitário Darcy Ribeiro - Gleba A, SG 10, Brasília,
DF, Brasil. CEP 70.904-970. E-mail: damasio.fabiana@gmail.com
 2

19 Declarations

20

21 Competing interests

22 The author(s) declare(s) that they have no competing interests.

23

24 Funding

25 No funding was received.

26

27 Availability of data and materials

28 Databanks can be requested to the first author by the e-mail address listed in the
29 contact details.

30 Authors' contributions

31 JL has made substantial contributions to the conception and design of the study,
32 analysis and interpretation of the data, drafting and revising the manuscript, and given final
33 approval of the version to be published. AP has made substantial contributions to
34 interpretation of the data, drafting and revising the manuscript, and given final approval of the
35 version to be published. JA has made substantial contributions to interpretation of the data,
36 revising the manuscript, and given final approval of the version to be published. MP has made
37 substantial contributions to the conception and design, acquisition of the data, analysis and
38 interpretation of the data, and given final approval of the version to be published.

39 Authors' information

40 JAL is a professor at the Institute of Psychology of the University of Brasília, Brazil.

41 AJSP is a researcher on educational assessment at the National Institute for Educational

42 Research (Inep), Brazil.

43 JMA is a professor at the Institute of Psychology of the University of Brasília, Brazil.

44 MFDP is analyst on health management at the Fundação Oswaldo Cruz (Fiocruz), Brazil.
Blinded Manuscript Click here to download Blinded Manuscript Blinded
manuscript.docx
Click here to view linked References
1

1 1 Title:
2
3
4 2 Validity evidence of two short scales measuring the big five personality factors
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 2

1 3 Abstract
2
3
4 4 The objective of this study was to obtain evidence of the convergent and factor validity of the
5
6 5 Reduced Scale of Big Five Personality Factors (ER5FP), with 20 items, and of the Reduced
7
8
9 6 Inventory of Big Five Personality Factors (IGFP-5R), with 32 items. The two Brazilian scales
10
11 7 were administered to 554 participants aged 16-69 years (M=30.6, SD=8.6). The measurement
12
13
14
8 model of each instrument was tested using Confirmatory Factor Analysis. Both scales showed
15
16 9 an adequate adjustment of the measurement model to the data (RMSEA<.06; SRMR<.06) after
17
18 10 excluding a number of items. Correlations between the factors of both instruments were
19
20
21 11 estimated. Strong evidence of the convergent validity was found for the factors Extraversion,
22
23 12 Neuroticism and Openness to Experience (correlations between .60 and .80); for the factors
24
25
26 13 Agreeableness and Conscientiousness this evidence was less strong (correlations between .43
27
28 14 and .48).
29
30
31 15 Keywords: convergent validity; big five personality factors; reduced scales.
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 3

1 16 The five-factor model (FFM), also known as Big Five, is currently one of the most
2
3
4 17 influential and investigated models used in the personality research field (De Raad & Mlacic,
5
6 18 2015; McCrae, 2011). In Brazil, although the number of studies published using the FFM as
7
8
9 19 reference is relatively small in comparison to other countries (Silva & Nakano, 2011), it can
10
11 20 be stated that the development of instruments based on this model is increasing. Among the
12
13
14
21 psychological instruments approved by the Federal Council of Psychology (SATEPSI, 2017)
15
16 22 and available on the Brazilian market, we can cite: the NEO Personality Inventory - Revised
17
18 23 (NEO-PI-R [Flores-Mendoza, 2007]); the Bateria Fatorial de Personalidade (Factorial Battery
19
20
21 24 of Personality – BFP [Nunes, Hutz & Nunes 2010]); the Escala Fatorial de Neuroticismo
22
23 25 (Factorial Scale of Neuroticism [Hutz & Nunes, 2001]); the Escala Fatorial de Extroversão
24
25
26 26 (Factorial Extraversion Scale [Nunes & Hutz, 2007a]); and the Escala Fatorial de Socialização
27
28 27 (Factorial Scale of Socialization [Nunes & Hutz, 2007b]). In addition to these, there are a
29
30
31
28 number of instruments under development, either for research purposes or for future
32
33 29 commercialization (e.g., Andrade, 2008; Carvalho, Nunes, Primi, & Nunes, 2012; Hauck
34
35 30 Filho, Machado, Teixeira, & Bandeira, 2012; Gomes & Golino, 2012; Hutz, Silveira, Serra,
36
37
38 31 Anton, & Wieczorek, 1998; Natividade & Hutz, 2015; Passos & Laros, 2015; Primi, Santos,
39
40 32 John, & De Fruyt, 2016; Vasconcelos, 2005; Vasconcellos & Hutz, 2008).
41
42
43 33 Since the instruments for personality assessment tend to be extensive, usually consisting
44
45 34 of more than 100 items, one of the research objectives in this area is the development of
46
47
48 35 reduced scales (Carvalho et al., 2012; Hauck Filho et al., 2012; Natividade & Hutz, 2015;
49
50 36 Passos & Laros, 2015). The literature shows evidence that a large number of items in an
51
52
53
37 instrument can be a source of measurement error. For example, very long instruments can
54
55 38 induce discouragement, fatigue, and inattention in the participants. Furthermore, a large
56
57 39 number of items may difficult to operationalize the joint administration of two or more
58
59
60
61
62
63
64
65
 4

1 40 instruments. Therefore, scales with relatively few items can potentially minimize such
2
3
4 41 problems and be useful in different testing settings (e.g., surveys and screenings), provided
5
6 42 that they have acceptable reliability indices and satisfactory evidence of validity.
7
8
9 43 One possible disadvantage of using reduced scales concerns the relatively high amount
10
11 44 of measurement errors due to the fact that the number of items of a scale is negatively related
12
13
14
45 to the amount of measurement errors. Another drawback of using reduced scales is related to
15
16 46 the possible occurrence of increasing Type I and Type II errors (Credé, Harms, Niehorster, &
17
18 47 Gaye-Valentine, 2012). However, the literature already presented evidence that reduced
19
20
21 48 instruments can be a reliable alternative for personality assessment (Donnellan, Oswald,
22
23 49 Baird, & Lucas, 2006; Laverdiere et al., 2010; Rammsted, 2007). With this motivation, some
24
25
26 50 reduced scales for personality assessment based on the FFM have been proposed in recent
27
28 51 years in Brazil.
29
30
31
52 Andrade (2008), for example, developed the Inventário Reduzido dos Cinco Grandes
32
33 53 Fatores de Personalidade (Reduced Inventory of the Five Great Personality Factors - IGFP-
34
35 54 5R), based on the translation and adaptation for Brazil of the Spanish version of the Big Five
36
37
38 55 Inventory (Benet-Martinez & John, 1998). Originally composed of 44 items, the inventory
39
40 56 was administered to a sample of 5,247 subjects from all Brazilian geographic regions.
41
42
43 57 Confirmatory factor analysis demonstrated the adequacy of a model with 32 items, five
44
45 58 correlated factors with reliability coefficients ranging from .66 to .76. By analyzing the items
46
47
48 59 using Item Response Theory (IRT), the author certified that the instrument had adequate
49
50 60 psychometric parameters, even considering that the items did not cover the entire latent trait
51
52
53
61 continuum. Among other conclusions, Andrade (2008) pointed out the need to carry out
54
55 62 studies investigating the evidence of the convergent validity between the IGFP-5R and other
56
57 63 FFM-based instruments, such as NEO-PI-R and NEO-FFI-R.
58
59
60
61
62
63
64
65
 5

1 64 Hauck Filho et al. (2012) proposed the Escala de Marcadores Reduzidos (Reduced
2
3
4 65 Marker Scale - EMR), based on adjective markers for personality assessment identified by
5
6 66 Hutz et al. (1998). Analyzing data from a sample of 674 university students, a five-factor
7
8
9 67 structure was found. Each factor was composed of five items, with reliability coefficients of
10
11 68 factor scores ranging from .61 to .80. In another study with the EMR on a sample of 208
12
13
14
69 adolescents (Hauck Filho, Teixeira, Machado, & Bandeira, 2012), factor analysis also
15
16 70 indicated a five-factor structure with each factor containing 4 items. The reliability
17
18 71 coefficients of the factor scores ranged from .55 to .80. Subsequently, Machado, Hauck Filho,
19
20
21 72 Teixeira, and Bandeira (2014) analyzed the responses of 887 university students to the EMR.
22
23 73 Factor analysis pointed to a structure with five oblique factors, and IRT analysis indicated a
24
25
26 74 good adjustment of the measurement model, although with a concentration of the items in a
27
28 75 restricted range of the latent trait continuum, as occurred in the study conducted by Andrade
29
30
31
76 (2008). Finally, Pariz, Haddad, and Machado (2016) gathered evidence of convergent and
32
33 77 criterion validity regarding the EMR. The results indicated statistically significant correlations
34
35 78 between the EMR and the BFP (Nunes et al., 2010), with the exception of the Socialization
36
37
38 79 (i.e., Agreeableness) and Openness to Experience factors.
39
40 80 Another example is the study conducted by Carvalho et al. (2012) regarding the
41
42
43 81 translation and adaptation to Brazil of the Ten Item Personality Inventory (TIPI), developed
44
45 82 by Gosling, Rentfrow, and Swann (2003). According to Carvalho et al., TIPI is one of the
46
47
48 83 most cited reduced scales based on the FFM. The results of data analysis of a sample of 404
49
50 84 Brazilian adolescents indicated a structure of three factors, but with low reliability indices
51
52
53
85 ranging from .41 to .63. Among other notes, the authors indicated that before discarding the
54
55 86 use of the TIPI in the Brazilian cultural context, it is necessary to gather evidence of validity
56
57 87 based on external variables, including other instruments constructed based on the FFM.
58
59
60
61
62
63
64
65
 6

1 88 Natividade and Hutz (2015) elaborated the Escala Reduzida de Descritores de

2
3
4 89 Personalidade (Reduced Scale of Personality Descriptors - Red5), composed of 20 items. In
5
6 90 one study, analyses of a sample of 1,889 adults resulted in an orthogonal structure with five
7
8
9 91 principal components with reliability coefficients ranging from .59 to .84. In a second study,
10
11 92 with a sample of 512 adults, confirmatory factor analysis corroborated the structure obtained
12
13
14
93 in the first study. There was also evidence of convergent validity between the Red5 and the
15
16 94 BFP (Nunes et al., 2015).
17
18 95 Finally, Passos and Laros (2015) proposed a 47 items semantic differential scale for
19
20
21 96 assess the FFM, called Escala Reduzida de Cinco Grandes Fatores de Personalidade (Reduced
22
23 97 Scale of Five Great Personality Factors - ER5FP). Exploratory and confirmatory factor
24
25
26 98 analyzes with data from a sample of 365 undergraduate students indicated a five-factor
27
28 99 structure with 20 items. The reliability coefficients ranged from .71 to .85. The authors
29
30
31
100 concluded that the scale was adequate as a measure of the five factors, considering the sample
32
33 101 analyzed. As a research agenda, the authors highlighted the objective of increasing and
34
35 102 diversifying the sample (e.g., the inclusion of participants with other levels of education and
36
37
38 103 from all Brazilian geographic regions) and carrying out convergent and criterion validity
39
40 104 studies.
41
42
43 105 In order to contribute to the development of reduced instruments for the assessment of
44
45 106 personality based on the FFM and given continuity to the studies of Passos and Laros (2015)
46
47
48 107 and Andrade (2008), the present study has as overall objective obtaining evidence of
49
50 108 convergent validity between the ER5FP and the IGFP-5R. Moreover, this study aimed to test
51
52
53
109 the factorial validity of the two instruments, analyzing the theoretically expected
54
55 110 measurement model for the two scales (i.e., five correlated factors). Finally, this study also
56
57
58
59
60
61
62
63
64
65
 7

1 111 had the objective of analyzing the role of covariates in the description of the factors that
2
3
4 112 compose the scales.
5
6 113 The covariates gender, age, and level of education were used to carry out comparisons
7
8
9 114 between groups. The covariant sex was chosen based on studies that indicate that there are
10
11 115 differences between men and women regarding personality traits. This is what Schmitt, Realo,
12
13
14
116 Voracek, and Allik (2008) concluded in their study with data from 55 nations. The findings
15
16 117 indicated that in more prosperous and egalitarian societies, the personality traits of men and
17
18 118 women tend to be less similar. For example, of the 55 studied nations women presented
19
20
21 119 higher rates of Neuroticism in the 49 more prosperous nations. The effect of age on
22
23 120 personality traits has also been investigated in several studies (Donnellan & Lucas, 2008;
24
25
26 121 Rammsted, 2007; Srivastava, John, Gosling, & Potter, 2003). In the study of the development
27
28 122 of a reduced version of the Big Five Inventory, the BFI-10, Rammsted (2007) concluded that
29
30
31
123 younger participants tended to present higher scores on Extraversion, while the older ones had
32
33 124 the highest scores on Agreeableness and Conscientiousness. Regarding the effect of level of
34
35 125 education, the findings in the studies of Rammsted (2007) indicated that people with higher
36
37
38 126 levels of education tend to present higher scores on Openness to Experience. In addition to the
39
40 127 covariates cited above, we also analyzed variables informing marital status and if the
41
42
43 128 participant is a parent in order to verify if there is any association of both with the personality
44
45 129 traits assessed.
46
47
48 130
49
50 131 Method
51
52
53
132 Participants
54
55 133 The sample of this study is composed of 554 subjects (Table 1), mostly female (58.9%)
56
57 134 and with a mean age of 30.6 years (SD = 8.61). Regarding the level of education, 1.5%
58
59
60
61
62
63
64
65
 8

1 135 reported having started or finished the primary education level, 29.7% the upper secondary,
2
3
4 136 51.4% the first stage of tertiary education, and 23.3% graduate school (i.e., lato sensu
5
6 137 specialization, master or doctoral degrees). The data were collected in the Distrito Federal and
7
8
9 138 in the Metropolitan Region of Salvador, Bahia. An item of the questionnaire asking about the
10
11 139 region of origin of the respondents, revealed that 47.5% of the participants were originally
12
13
14
140 from the Center-West region, 39.2% from the Northeast, and 13.3% from other geographical
15
16 141 regions of the country.
17
18 142
19
20
21 143 Table 1
22
23 144 Sociodemographic information of the participants (N = 554)
24
25
26 Variable Response category n %
27 Male 226 40.80
28 Sex Female 324 58.50
29
Not informed 4 .70
30
31 < 20 41 7.40
32 21-25 122 22.00
33 26-30 164 29.60
34 Age (years)
31-35 92 16.60
35
36 >36 134 24.20
37 Not informed 1 .20
38 Married 289 52.20
39
40 Single 210 37.90
Marital status
41 Other status 55 9.90
42 Not informed
43 Primary 7 1.20
44
45 Upper secondary 131 23.60
46 Level of education Tertiary education (first stage) 282 50.40
47 Graduate school* 128 23.10
48 Not informed 5 .90
49
50 North 17 3.10
51 Northeast 217 39.20
52 Originary geographic region Center-West 263 47.50
53
54 Southeast 52 9.40
55 South 4 .90
56 145 Note: *lato sensu specialization, master or doctoral degree
57
58
59
60
61
62
63
64
65
 9

1 146 Instruments
2
3
4 147 Escala Reduzida de Cinco Grandes Fatores de Personalidade (ER5FP).
5
6 148 The ER5FP (Passos & Laros, 2015) is a semantic differential rating scale composed of
7
8
9 149 20 pairs of bipolar adjectives, with 7-point Likert type response scale. In the study of Passos
10
11 150 and Laros, satisfactory coefficients of reliability of the factor scores were found (Extraversion
12
13
14
151 = .85, Conscientiousness = .78, Agreeableness = .81, Neuroticism = .80, and Openness to
15
16 152 Experience = .71).
17
18 153 Inventário Reduzido dos Cinco Grandes Fatores de Personalidade (IGFP-5R).
19
20
21 154 The IGFP-5R is an instrument based on the Big Five Inventory (BFI) that was
22
23 155 submitted to a validation process for the Brazilian context by Andrade (2008). The IGFP-5R
24
25
26 156 is composed of 32 items, with a 5-point Likert type response scale. The coefficients of
27
28 157 reliability of the factor scores found in the Andrade study were considered acceptable
29
30
31
158 (Extraversion = .76, Conscientiousness = .66, Agreeableness = 0.74, Neuroticism = 0.75, and
32
33 159 Openness to Experience= .68).
34
35
160 Procedures
36
37
38 161 Data Collection
39
40 162 The researchers followed the guidelines of the Resolution 510/2016 of the Conselho
41
42
43 163 Nacional de Saúde (Brazilian National Council of Health) regarding scientific research
44
45 164 involving human beings. One of the authors administered the instruments in one public and
46
47
48 165 two private educational institutions. Prior to the basic instructions, the participants were
49
50 166 informed about the research objectives and the confidentiality concerning their identity and
51
52
53
167 signed the Free and Informed Consent Term. The participants individually answered the
54
55 168 instruments, with completion time ranging from 10 to 30 minutes.
56
57 169
58
59
60
61
62
63
64
65
 10

1 170 Data analysis

2
3
4 171 The data analysis occurred in four stages. The first stage involved cleaning the database
5
6 172 and performing exploratory analyses. Participants who did not respond to one of the
7
8
9 173 instruments or did not report sociodemographic information were withdrawn, as well as those
10
11 174 considered outliers based on the Mahalanobis distance. As a result, 9.2% of the cases were
12
13
14
175 excluded, leaving 554 valid cases remaining. Then, the missing data were imputed using the
15
16 176 linear trend at point technique of SPSS 18.0. After the exploratory analyses, the assumptions
17
18 177 of Confirmatory Factor Analysis (CFA) for the data were checked. First, the assumption of
19
20
21 178 univariate normality was evaluated through the inspection of the asymmetry and kurtosis of
22
23 179 the items. An item was considered non-normally distributed if its asymmetry was greater than
24
25
26 180 1.0 and its kurtosis was superior to 2.0 (Miles & Shevlin, 2001; Osborne, 2002). None of the
27
28 181 items of the two personality instruments showed univariate non-normality. Afterwards, the
29
30
31
182 multivariate normality of the items was verified using the multivariate kurtosis coefficient of
32
33 183 Mardia (Byrne, 2016).
34
35 184 In the second stage, the CFA of the two instruments was conducted through Structural
36
37
38 185 Equation Modeling with Maximum Likelihood estimation. The measurement model defined
39
40 186 for the two scales aimed the identification of five correlated factors: Extraversion,
41
42
43 187 Conscientiousness, Agreeableness, Neuroticism, and Openness to Experience. Based on the
44
45 188 modification indices, changes were made in the initial models of the two scales. Modification
46
47
48 189 indices indicating a substantial impact on the improvement of fit suggested including a
49
50 190 correlation between the error terms of several items related to the same factor. Byrne (2016)
51
52
53
191 states that the suggestion of including a correlation between two-item errors in the model
54
55 192 reveals, in most cases, that the items are very similar in terms of content. In these cases, it is
56
57 193 suggested to exclude one of the two items, since the second item does not add information.
58
59
60
61
62
63
64
65
 11

1 194 Thus, the item with the lowest factor loading was excluded of the item pairs for which the
2
3
4 195 modification index suggested correlation between the errors. In the CFA analyses, the
5
6 196 recommendations of Byrne (2016), Kline (2016), and Weston, Gore, Chan, and Catalano
7
8
9 197 (2008) were followed.
10
11 198 In the third stage, Guttman’s coefficient lambda 2 (λ2) was calculated to evaluate the
12
13
14
199 internal consistency of the scores of each factor. This coefficient was adopted because it is
15
16 200 more appropriate in studies that have used instruments with reduced number of items for each
17
18 201 latent component (Sijtsma, 2012). In sequence, the raw convergent validity coefficients and
19
20
21 202 the ones corrected for attenuation were calculated. Pearson's correlation was used to obtain
22
23 203 evidence of the convergent validity of the two personality instruments. To correct the raw
24
25
26 204 coefficients for the effect of measurement errors, we used the dual correction formula for
27
28 𝑟𝑥𝑦
205 attenuation (Osborne, 2013), which is as follows: 𝑟𝑥 ′ 𝑦 ′ = ⁄ 𝑟 𝑟 . In this formula, 𝑟𝑥 ′ 𝑦 ′
29 √ 𝑥𝑥 𝑦𝑦
30
31
32 206 is the validity coefficient corrected for the measurement error in both the test and the
33
34 207 criterion, 𝑟𝑥𝑦 is the observed correlation, 𝑟𝑥𝑥 is the reliability of the test, and 𝑟𝑦𝑦 is the
35
36
37 208 reliability of the criterion.
38
39 209 The last stage was the analysis of differences between groups on the five factor scores
40
41
42 210 of each instrument. The groups were selected based on the following covariates: gender (male
43
44 211 vs. female), age (less than 30 years old vs. more than 30), being a parent, marital status, and
45
46
47 212 level of education. The statistical package SPPS 18.0 was used in the exploratory and
48
49 213 descriptive analyzes and for testing the differences between means. The SPSS Amos 18.0
50
51
52
214 package was employed for the verification of the univariate and multivariate normality, for
53
54 215 the design of the factor models, and for the CFA.
55
56 216
57
58
59 217
60
61
62
63
64
65
 12

1 218 Results
2
3
4 219 Model fitting
5
6 220 As can be seen in Table 2, the initial ER5FP model has an RMSEA value greater than
7
8
9 221 .06 and an SRMR value greater than .08, which is not considered a good fit. According to the
10
11 222 recommendations of Weston et al. (2008), the minimum criteria for assuring the quality of a
12
13
14
223 model fit are RMSEA ≤ .06 and SRMR ≤ .08. The CFI and GFI values of the initial ER5FP
15
16 224 model also indicated an unsatisfactory fit since they fell below the .95 criterion. To improve
17
18 225 the quality of model fitting, alterations in the model were made based on the modification
19
20
21 226 indices suggested by the Amos package. Following the procedure described in the method
22
23 227 section, five of the 20 items were excluded from the initial ER5FP measurement model. The
24
25
26 228 fit of the final model with 15 items was satisfactory, with RMSEA = .059, SRMR = .054, and
27
28 229 CFI and GFI around .95. Thus, all adjustment indices indicated that the final ER5FP
29
30
31
230 measurement model is appropriate for the data.
32
33 231 The original model of measurement of the IGFP-5R was also not satisfactory, with
34
35 232 RMSEA and SRMR values above the criterion value and with IFC and GFI values below the
36
37
38 233 criterion value (Table 2). Changes were made based on the modification indices to obtain a
39
40 234 good fit. Sixteen of the 32 items of the original measurement model were excluded. The fit of
41
42
43 235 the final measurement model was good, with RMSEA = .038, SRMR = .043, CFI = .95, and
44
45 236 GFI = .94.
46
47
48 237
49
50 238
51
52
53
239
54
55 240
56
57 241
58
59
60
61
62
63
64
65
 13

1 242 Table 2
2
3
4 243 Fit indices of the initial and final measurement models of the scales
5
6
Number of
7 Scale Model ² ² / DF CFI TLI RMSEA (IC 90%) SRMR
8 items
9
10 Initial 20 710.91 4.43 .83 .80 .079 (.073 - .085) .088
11 ER5FP
12 Final 15 233.57 2.92 .93 .91 .059 (.050 - .068) .054
13
14 Initial 32 1618.73 3.57 .66 .63 .068 (.065 - .072) .078
IGFP-5R
15
16 Final 16 169.61 1.80 .95 .94 .038 (.029 - .047) .043
17 Note: ² = chi-square; DF = degrees of freedom; CFI = Comparative Fix Index; GFI = Goodness of Fit Index;
18 RMSEA = Root Mean Square Error of Approximation; SRMR = Standardized Root Mean Square Residual.
19
20 244
21
22
245 Reliability coefficients, factor loadings, and correlations between factors
23
24
25 246 Table 3 displays the factor loadings, the communalities, the correlations between the
26
27 247 factors, the item-rest (𝑟𝑖𝑟 ) correlations, the mean correlation between the items, and the
28
29
30 248 reliability coefficients of the scores on the factors regarding the ER5FP. It is noteworthy that,
31
32 249 even with a smaller number of items, the reliability of the factor scores is reasonable, except
33
34
35 250 for the last factor, Openness to Experience, which showed a reliability coefficient of .58. The
36
37 251 reliability coefficients of the other factors ranged between .67 and .79, with the greater value
38
39
40
252 referent to the Extraversion factor, which also presented the highest mean correlation among
41
42 253 its items (r =.56).
43
44
45
254 Regarding the correlation between the factors, it is observed that Openness to
46
47 255 Experience presented positive and high correlations with Conscientiousness (r=.86) and
48
49 256 Agreeableness (r =.66). Neuroticism had a negative correlation with Conscientiousness (r =-
50
51
52 257 .27) and Agreeableness (r =-.26). The values found have similarities with the results of the
53
54 258 study conducted by Herzberg and Brahler (2006) on the evaluation of reduced scales.
55
56
57 259 Analyzing the 16-Adjective Measure, the authors also found negative correlations between
58
59
60
61
62
63
64
65
 14

1 260 Neuroticism and Conscientiousness (r =-.09), and positive correlations between Openness and
2
3
4 261 Agreeableness (r =.23).
5
6 262 Table 4 displays the same information for the IGFP-5R. The reliability coefficients of
7
8
9 263 the factors are reasonable, ranging from .65 to .72. The correlations between the items of each
10
11 264 factor ranged from .34 to .46, with the highest value related to the Neuroticism factor.
12
13
14 265 Concerning the correlations between the factors, there is a positive correlation between
15
16 266 Agreeableness and Openness (r =.33) and a negative correlation between Neuroticism and
17
18
19
267 Conscientiousness (r =-.44). This information is consonant with the results regarding the
20
21 268 ER5FP and the findings of Herzberg and Bahler (2006).
22
23 269
24
25
26 270
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 15

1 271 Table 3
2
3
4 272 ER5FP: Reliability coefficients (Guttman’s λ2), factor loadings (FL), communalities (h2), item-
5 273 rest correlations (rir), and correlations between factors (N=554)
6 274
7
Factor 1 - Extraversion (λ2 = .79) FL h2 rir
8
9 Item 1. Extrovertido (extraverted) / Tímido (timid) .69 .48 .60
10
Item 5. Comunicativo (communicative) / Calado (silent) .87 .76 .69
11
12 Item 6. Sociável (sociable) / Reservado (reserved) .69 .48 .60
13
Mean .75 .57 .69
14
15 Mean correlation between the items = .56
16
Fator 2 - Conscientiousness (λ2 = .67) FL h2 rir
17
18 Item 2. Persistente (persistent) / Desistente (quitting) .64 .41 .49
19
Item 10. Motivado (motivated) / Desmotivado (unmotivated) .77 .59 .54
20
21 Item 12. Obstinado (obstinate) / Inconstante (fickle) .51 .26 .41
22
Mean .64 .42 .48
23
24 Mean correlation between the items = .40
25
Fator 3 - Agreeableness (λ2 = .69) FL h2 rir
26
27 Item 4. Simpático (sympathetic) / Antipático (unpleasant) .65 .42 .52
28
Item 7. Amoroso (amorous) / Indiferente (indifferent) .62 .38 .47
29
30 Item 14. Gentil (kind) / Rude .70 .49 .53
31
Mean .66 .43 .51
32
33 Mean correlation between the items = .43
34
Fator 4 – Neuroticism (λ2 = .74) FL h2 rir
35
36 Item 3. Calmo (calm) / Nervoso (nervous) .69 .48 .58
37
Item 13. Paciente (patient) / Impaciente (impatient) .77 .59 .58
38
39 Item 18. Tranquilo (tranquil) / Ansioso (anxious) .63 .40 .54
40
Mean .70 .49 .57
41
42 Mean correlation between the items = .49
43
Fator 5 – Openness to Experience (λ2 = .58) FL h2 rir
44
45 Item 8. Criativo (creative) / Prosaico (prosaic) .57 .32 .42
46
Item 9. Entusiasta (passionate) / Apático (apathetic) .59 .35 .40
47
48 Item 15. Autêntico (authentic) / Simulado (feigned .50 .25 .32
49
Mean .55 .31 .38
50
51 Mean correlation between the items = .30
52
53 Correlations between the factors: F1-F2 = .40; F1-F3 = .25; F1-F4 = -.02; F1-F5 = .47; F2-F3 = .63; F2-
54 F4 = -.27; F2-F5 = .86; F3-F4 = -.26; F3-F5 = .66; F4-F5 = -.10.
55
56
275
57
58
59
60
61
62
63
64
65
 16

1 276 Table 4
2
3
4 277 IGFP-5R: Reliability coefficients (Guttman’s λ2), factor loadings (FL), communalities (h2),
5 278 item-rest correlations (rir), and correlations between factors (N=554)
6 279
7
Factor 1 - Extraversion (λ2 = .65) FL h2 rir
8
9 Item 12. É reservado (Is reserved) .50 .25 .41
10
11 Item 16. É, às vezes, tímido, inibido (Sometimes is timid, inhibited) .64 .41 .46
12 Item 42. Tende a ser quieto, calado (Tend to be quiet, silent) .70 .49 .50
13
14 Mean .61 .38 .46
15 Mean correlation between the items = .37
16
17 Factor 2 – Conscientiousness (λ2 = .68) FL h2 rir
18 Item 17. Pode ser um tanto descuidado (Can be somewhat careless) .61 .37 .50
19
20 Item 19. Tende a ser preguiçoso (Tend to be lazy) .57 .32 .43
21 Item 22. É facilmente distraído (Is easily distracted) .57 .32 .44
22
23 Item 38. Tende a ser desorganizado (Tend to be disorganized) .61 .37 .47
24 Mean .59 .35 .46
25
26 Mean correlation between the items = .34
27 Factor 3 – Agreeableness (λ2 = .66) FL h2 rir
28
29 Item 8. Gosta de cooperar com os outros (Enjoys to cooperate with others) .61 .37 .48
30
Item 15. É prestativo e ajuda os outros (Is cordial and helps others) .76 .58 .55
31
32 Item 18. É amável, tem consideração pelos outros (Is agreeable, cares
.60 .36 .48
33 about the others)
34 Mean .66 .44 .50
35
36 Mean correlation between the items = .43
37 Factor 4 – Neuroticism (λ2 = .72) FL h2 rir
38 Item 10. É temperamental, muda de humor facilmente (Is temperamental,
39 .54 .29 .46
changes mood easily)
40
41 Item 34. Fica tenso com frequência (Get tense often) .70 .49 .55
42 Item 36. Fica nervoso facilmente (Get nervous easily) .80 .64 .61
43
44 Mean .68 .47 .54
45 Mean correlation between the items = .46
46
47 Factor 5 – Openness to Experience (λ2 = .67) FL h2 rir
48 Item 9. É original, tem sempre novas ideias (Is original, allways has new
49 .69 .48 .51
ideas)
50 Item 11. É inventivo, criativo (Is inventing, creative) .81 .66 .57
51
52 Item 39. Gosta de refletir, brincar com as ideias (Enjoys thinking, playing
.43 .18 .37
53 with ideas)
54 Mean .64 .44 .48
55
Mean correlation between the items = .43
56
57 Correlations between the factors: F1-F2 = .25; F1-F3 = .03*; F1-F4 = .18; F1-F5 = .11*; F2-F3 = .26; F2-F4
58 = -.44; F2-F5 = .19; F3-F4 = -.23; F3-F5 = .33; F4-F5 = -.12*
59 *non-significant correlations, p<.05.
60
61
62
63
64
65
 17

1 280 Table 5 shows the mean, standard deviation, and variance of the scales. For the ER5FP,
2
3
4 281 a comparison was made between the means found in the construction study (Passos & Laros,
5
6 282 2015) and those found in this validity study. For the IGFP-5R, data is presented only for this
7
8
9 283 study. Regarding the distribution of the factor scores, the Agreeableness factor showed the
10
11 284 highest mean both for the IGFP-5R (M = 4.48; SD = .61) and the ER5FP (M = 5.71; SD = 1.05).
12
13
14
285 In comparison with the ER5FP development study, the variance of factor scores in this study is
15
16 286 lower, especially for Openness and Extraversion (28% and 18% less variance, respectively).
17
18 287
19
20
21 288 Table 5
22
23 289 Statistics of the factor scores of the scales
24
25 ER5FP IGFP-5R
26 Development Convergent validity Convergent validity
27 studya study study
28 (N=365) (N = 554)b (N = 554)
29 M SD Var
Factor M SD Var M SD Var
30
31 Extraversão 4.48 1.71 2.91 4.55 1.54 2.38 2.61 1.01 1.02
32 Conscientiousness 5.61 1.12 1.26 5.34 1.08 1.17 3.37 0.95 0.90
33 Agreeableness 5.84 1.07 1.14 5.71 1.05 1.11 4.48 0.61 0.38
34
Neuroticism 3.95 1.64 2.68 3.81 1.60 2.57 2.74 1.13 1.27
35
36 Openness to
5.33 1.13 1.27 5.40 0.96 0.92 3.90 0.75 0.57
37 Experience
38 Legend: M (mean); SD (standard-deviation); Var (variance).
39 a. Passos e Laros (2015).
40 b. the factor scores were computed using the same items.
41 290
42
43 291 Testing differences between groups
44
45
46 292 Differences on the five factor scores of the ER5FP and the IGFP-5R were analyzed
47
48 293 between groups of participants based on the following covariates: sex (men vs. women), age
49
50
51 294 (less than 30 years old vs. more than 30), being a parent or not, marital status (single vs.
52
53 295 married), and level of education. Concerning the comparisons of means using the ER5FP
54
55
56
296 data, with regard to sex, it was observed that women (M = 3.96) tended to present higher
57
58 297 levels of Neuroticism than men (M = 3.59), p < .01. When evaluating the covariate age, older
59
60
61
62
63
64
65
 18

1 298 people (M1) present higher scores than the younger ones (M0) in Openness (M1 = 5.52, M0 =
2
3
4 299 5.27, p < .05) and Agreeableness (M1 = 5.79, M0 = 5.60, p < .05). As for being a parent (M1)
5
6 300 or not (M0), the results indicated that the participants who had children tended to be more
7
8
9 301 extraverted (M1 = 4.70, M0 = 4.42, p < .05), agreeable (M1 = 5.85, M0 = 5.59, p < .01), and
10
11 302 open to new experiences (M1 = 5.52, M0 = 5.30, p < .01). Finally, although the difference
12
13
14
303 was not large, married participants (M1) had higher scores than the single ones (M0) in
15
16 304 Agreeableness (M1 = 5.82, M0 = 5.60, p < .01). Using the ER5FP data, no significant
17
18 305 differences were found for the level of education, considering the various levels.
19
20
21 306 The comparisons between groups using the factor scores of the IGFP-5R produced the
22
23 307 following results. There were higher rates of Neuroticism among women (M = 2.90) than
24
25
26 308 among men (M = 2.51), p < .01. Although the differences were not large, women (M = 4.53)
27
28 309 also presented higher scores on the Agreeableness factor (M = 4.41), p < .05. The results also
29
30
31
310 revealed that males (M = 4.00) tended to be more open to new experiences than females (M =
32
33 311 3.82), p < .05. In the evaluation of age differences, older people (M1) tended to show higher
34
35 312 indices than the younger ones (M0) in Conscientiousness (M1 = 3.48, M0 = 3.25, p < .01),
36
37
38 313 Agreeableness (M1 = 4.54, M0 = 4.42, p < .05), and Neuroticism (M1 = 2.63, M0 = 2.85, p <
39
40 314 .05). When assessing differences in the groups based on the covariate being a parent, it was
41
42
43 315 observed that the participants who had children (M1) presented higher averages for the
44
45 316 Agreeableness factor (M1 = 4.56, M0 = 4.42, p < .05). Regarding the other covariates, marital
46
47
48 317 status and level of education, the averages found for the groups did not reveal statistically
49
50 318 significant differences.
51
52
53 319 Convergent validity evidences
54
55
320 The correlation coefficients corrected for attenuation between the factors of the ER5FP
56
57
58 321 and the factors of the IGFP5-R are shown in the diagonal of Table 6, together with the raw
59
60
61
62
63
64
65
 19

1 322 coefficients, which are presented between parentheses. The corrected correlation coefficients
2
3
4 323 can be interpreted as evidence of convergent validity between the instruments. The lower and
5
6 324 upper triangles represent evidence of discriminant validity. As can be observed, the values of
7
8
9 325 the corrected coefficients of validity ranged from .43 to .80. According to Urbina (2014),
10
11 326 coefficients of validity between .20 and .30 are the common ones and values greater or equal
12
13
14
327 .40 can be interpreted as satisfactory coefficients.
15
16 328
17
18 329 Table 6
19
20
21 330 Rawa and corrected for attenuation correlation coefficients between factors of the ER5FP
22 331 and IGFP-5R scales (N=554)
23 332
24
25 IGFP-5R
26 Factors
27 EX NE AG CO AG
28
29 EX .80 (.57) -.10 (-.07) - .23 (-.14) -.20 (-.13) -.02(-.01)
30
31 NE -.07 (-.05) .60 (.44) -.17 (-.11) -.39 (-.27) -.28 (-.20)
32
33 ER5FP OP .27 (.20) -.09 (-.06) .74 (.46) .42 (.28) .29 (.20)
34
35 CO .11 (.08) -.13 (-.09) .35 (.21) .43 (.29) .24 (.15)
36
37 AG .22 (.16) -.10 (-.07) .39 (.25) .31 (.21) .48 (.33)
38
39 333 Note: EX (Extraversion), NE (Neuroticism), OP (Openness to Experience), CO (Conscientiousness), and AG
40 334 (Agreeableness).
41 335 a. the raw correlation coefficients are presented between parenthesis.
42 336
43
44
45
337 The highest coefficient of convergent validity was found for Extraversion (r =.80) and
46
47 338 the lowest for Conscientiousness (r =.43). The correlations between different factors also
48
49 339 merit attention, such as between Openness to Experience (IGFP-5R) and Conscientiousness
50
51
52 340 (ER5FP), r=.42, and between Openness to Experience (ER5FP) and Agreeableness (IGFP-
53
54 341 5R), r =.39. The finding that the Extraversion factor have higher coefficient of convergent
55
56
57 342 validity may be due to the evidence that this factor tends to be identified in several cultures, as
58
59
60
61
62
63
64
65
 20

1 343 indicated by Yamagata et al. (2006). The Openness to Experience and Neuroticism factors
2
3
4 344 presented a convergent validity coefficient of respectively .74 and .60. These two factors are
5
6 345 also generally identified in other cultures.
7
8
9 346 Discussion and Conclusions
10
11 347 Both reduced scales of personality presented adequate psychometric quality in this
12
13
14
348 study, as shown by a good model fit in confirmatory factor analysis and adequate reliability
15
16 349 coefficients of the factors, ranging from .58 (Openness to Experience) to .79 (Extraversion) in
17
18 350 the ER5FP, and from .65 (Extraversion) to .72 (Neuroticism) in the IGFP-5R. The factors
19
20
21 351 Extraversion and Neuroticism in both instruments presented the highest reliability coefficients
22
23 352 and inter-item correlations. This fact may be due to the quality and clarity of the definition of
24
25
26 353 the factors in question, which may have allowed greater coverage of the content of each factor
27
28 354 (Frazier, Naugle, & Harggety, 2006).
29
30
31
355 The reliability indices found are compatible with the values reported in national and
32
33 356 international studies based on the FFM model. However, the high correlation between the
34
35 357 factors, both for the ER5FP and for the IGFP-5R, may represent the presence of a hierarchical
36
37
38 358 and oblique model for the instruments, as discussed by Hauck Filho et al. (2012).
39
40 359 The means of the ER5FP factor scores identified in the development study and in the
41
42
43 360 current study resemble the findings of Donnellan et al. (2006). The data for the
44
45 361 Conscientiousness and Openness factors approximate the results presented by the authors
46
47
48 362 regarding the TIPI (Ten Item Personality Inventory) data, which also have a response scale of
49
50 363 7 points.
51
52
53
364 In the comparison between groups, similarities were found for the data obtained with
54
55 365 the two instruments. In both scales, women presented higher averages for Neuroticism.
56
57 366 Although they presented approximated means regarding Agreeableness in the ER5FP, women
58
59
60
61
62
63
64
65
 21

1 367 were shown to be more agreeable than the men in the IGFP-5R. Regarding the covariate being
2
3
4 368 a parent, there is congruence between the results regarding the two scales for three factors:
5
6 369 Agreeableness, Openness, and Extraversion. However, only the Agreeableness factor
7
8
9 370 presented significant differences between groups’ means in the two scales, denoting that the
10
11 371 participants who have children tended to be more agreeable. Concerning the covariate age, the
12
13
14
372 results revealed significant differences for Agreeableness in the two scales, indicating that the
15
16 373 older people tend to be more agreeable. The study also indicated that older people presented
17
18 374 higher scores on Openness, considering the IGFP-5R data. No significant differences were
19
20
21 375 identified in the ER5FP data.
22
23 376 Still regarding the comparisons of age groups, the analysis of the Conscientiousness
24
25
26 377 factor is also worth highlighting. The data indicated that older people tend to have higher
27
28 378 scores on this factor. However, the differences considering the ER5FP data were not
29
30
31
379 significant, p<.07. These results were similar to the findings of Donnellan and Lucas (2008)
32
33 380 and Terraciano, McCrae, Brant, and Costa (2005), who pointed out a similarity between the
34
35 381 averages presented by adolescents and the older ones, over 60 years old. According to the
36
37
38 382 authors, the means of these age groups tend to be lower than those evidenced by participants
39
40 383 between 20 and 50 years old.
41
42
43 384 Notwithstanding the positive results of this study it is important to point out its
44
45 385 limitations. The first one refers to the sample. Because it was carried out basically in the
46
47
48 386 Distrito Federal and in the Metropolitan Region of Salvador, Bahia, the study did not present
49
50 387 a representative sample of the entire Brazilian population. In addition, since data collection
51
52
53
388 was performed in educational institutions, there was little variability in the response pattern,
54
55 389 which is not the most recommended in studies that seek to evaluate the psychometric quality
56
57 390 of instruments. Another limitation refers to the low level of discrimination between some of
58
59
60
61
62
63
64
65
 22

1 391 the five personality factors. In the data from the ER5FP, for example, the high and positive
2
3
4 392 correlation between Conscientiousness and Openness to Experience (r=.86) means that the
5
6 393 discriminant validity of these two factors is less than desirable. Even when working with the
7
8
9 394 hypothesis of an oblique factor structure, it is expected that the correlations between the
10
11 395 factors would be smaller.
12
13
14
396 According to Byrne (2016), after making modifications to the original model, the
15
16 397 researcher stops working in a confirmatory mode and starts working in an exploratory
17
18 398 fashion. Thus, the results obtained in this study with the Structural Equation Modeling need to
19
20
21 399 be checked in future studies. As a research agenda, it is suggested that further studies should
22
23 400 be carried out with representative samples, so as to cover the entire national territory.
24
25
26 401 Investigating convergent validity is an essential step in the development of
27
28 402 psychometric instruments. Thus, this study makes a contribution for the advancement of the
29
30
31 403 research with the FFM, as well for the development of measurement instruments. In addition
32
33 404 to investigating the psychometric properties of the instruments, it is believed that this study
34
35
36
405 may also encourage further research on the relationship between personality and issues related
37
38 406 to social context. Finally, the results of this study and those from the works of Passos and
39
40 407 Laros (2015) and Andrade (2008) provided evidence that ER5FP and IGFP-5R are good
41
42
43 408 alternatives for use in future research involving personality assessment within the theoretical
44
45 409 framework of the five-factor model.
46
47
48 410
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 23

1 411 References
2
3
4 412 Andrade, J. M. (2008). Evidências de validade do inventário dos cinco grandes fatores de
5
6 413 personalidade para o Brasil. Tese de Doutorado, Universidade de Brasília, Brasília.
7
8 414 Benet-Martinez, V. & John, O. P. (1998). Los cinco grandes across cultures and ethnic groups:
9 415 Multitrait multimethod analyses of the big five in Spanish and English. Journal of
10
11 416 Personality and Social Psychology, 75(3), 729-750.
12
13 417 Byrne, B. M. (2016). Structural Equation Modeling with AMOS: Basic concepts, applications
14
15 418 and programming (3rd ed.). New York: Routledge.
16
17 419 Canivez, G. L., Watkins, M. W., & Dombrowski, S. C. (2016). Structural validity of the
18 420 Wechsler Intelligence Scale for Children - Fifth edition: Confirmatory factor analyses with
19
20 421 the 16 primary and secondary subtests. Psychological Assessment, 29(4), 458-472.
21
22 422 Carvalho, L. de F., Nunes, M. F. O., Primi, R., & Nunes, C. H. S. da S. (2012). Evidências
23
24 423 desfavoráveis para avaliação da personalidade com um instrumento de 10 itens. Paidéia,
25 424 22, 63-71.
26
27 425 Conselho Federal de Psicologia. (2017, August 21). SATEPSI - Sistema de Avaliação de
28
29 426 Testes Psicológicos. ReIRTeved from http://satepsi.cfp.org.br/listaTeste.cfm?status=1
30
31 427 Credé, M., Harms, P., Niehorster, S., & Gaye-Valentine, A. (2012). An evaluation of the
32
33 428 consequences of using short measures of the Big Five personality traits. Journal of
34 429 Personality and Social Psychology, 102, 874-888. http://dx.doi.org/10.1037/a0027403
35
36 430 De Raad, B., & Mlacic, B. (2015). The lexical foundation of the big five factor model. In T.
37
38 431 Widiger (Ed.), The Oxford handbook of the five factor model. doi:
39
40 432 10.1093/oxfordhb/9780199352487.013.12.
41
42 433 Donnellan, M. B., & Lucas, R. E. (2008). Age differences in the Big Five across the life span:
43 434 Evidence from two national samples. Psychology and Aging, 23, 558-562.
44
45 435 Flores-Mendoza, C. (2007). Manual do NEO-PI-R e NEO-FFI. São Paulo: Vetor Editora.
46
47
48
436 Hauck Filho, N., Machado, W. L., Teixeira, M. A. P., & Bandeira, D. R. (2012). Evidências
49 437 de validade de marcadores reduzidos para a avaliação da personalidade no modelo dos
50
51 438 cinco grandes fatores. Psicologia: Teoria e Pesquisa, 28, 417-423.
52
53 439 Hauck Filho, N., Teixeira, M. A. P., Machado, W. L., & Bandeira, D. R. (2012). Marcadores
54 440 reduzidos para a avaliação da personalidade em adolescentes. Psico-USF, 17(2), 256-261.
55
56
57
441 Hutz, C. S., & Nunes, C. H. S. (2001). Escala Fatorial de Neuroticismo. São Paulo: Casa do
58 442 Psicólogo.
59
60
61
62
63
64
65
 24

1
2 443 Hutz, C. S., Nunes, C. H, Silveira, A. D., Serra, J., Anton, M., & Wieczorek, L. S. (1998). O
3 444 desenvolvimento de marcadores para a avaliação da personalidade no modelo dos cinco
4
5 445 grandes fatores. Psicologia: Reflexão e Crítica, 11(2), 395-411. doi: 10.1590/S0102-
6 446 79721998000200015.
7
8
9
447 Machado, W. L., Hauck Filho, N., Teixeira, M. A. P., & Bandeira, D. R. (2014). Análise de
10 448 Teoria de Resposta ao Item de marcadores reduzidos da personalidade. Psico, 45(4), 551-
11
12 449 558. doi: 10.15448/1980-8623.2014.4.13138.
13
14 450 McCrae, R. R. (2011). Personality theories for the 21st century. Teaching of Psychology,
15 451 38(3), 209-214. doi: 10.1177/0098628311411785
16
17
18
452 Miles, J., & Shevlin, M. (2001). Applying regression & correlation. A guide for students and
19 453 researchers. London: Sage Publications.
20
21 454 Natividade, J. C., & Hutz, C. S. (2015). Escala reduzida de descritores dos cinco grandes fatores
22
23 455 de personalidade: prós e contras. Psico, 46(1), 79-89. doi: 10.15448/1980-
24 456 8623.2015.1.16901
25
26
27
457 Nunes, C. H. S., & Hutz, C. S. (2007a). Escala Fatorial de Extroversão. São Paulo: Casa do
28 458 Psicólogo.
29
30 459 Nunes, C. H. S., & Hutz, C. S. (2007b). Escala Fatorial de Socialização. São Paulo: Casa do
31
32 460 Psicólogo.
33
34 461 Nunes, C. H. S., Hutz, C. S., & Nunes, M. (2010). Bateria Fatorial de Personalidade (BFP):
35
36
462 Manual Técnico. São Paulo: Casa do Psicólogo.
37
38 463 Osborne, J. W. (2002). Notes on the use of data transformations. Practical Assessment,
39 464 Research & Evaluation, 8(6), 1-8.
40
41 465 Osborne, J. W. (2013). Best practices in data cleaning: A complete guide to everything you
42
43 466 need to do before and after collecting your data. Thousand Oaks, Estados Unidos: Sage.
44
45 467 Passos, M. F., & Laros, J. A. (2014). O modelo dos cinco grandes fatores: revisão de
46
47 468 literatura. Peritia – Revista Portuguesa de Psicologia, 21, 13-21.
48
49 469 Passos, M. F. D., & Laros, J. A. (2015). Construção de uma escala reduzida de Cinco Grandes
50 470 Fatores de personalidade. Avaliação Psicológica, 14(1), 115-123. doi:
51
52 471 10.15689/ap.2015.1401.13.
53
54 472 Pariz, J., Haddad, E., & Machado, W. L. (2016). Convergent and criterion-based validity for a
55
56 473 brief scale of the five-factor model. Avaliação Psicológica, 15(3), 346-351. doi:
57 474 10.15689/ap.2016.1503.07.
58
59
60
61
62
63
64
65
 25

1
2 475 Primi, R., Santos, D., John, O. P. & De Fruyt, F. (2016). The development of a nationwide
3 476 inventory assessing social and emotional skills in Brazilian youth. European Journal of
4
5 477 Psychological Assessment, 32, 5–16. doi: 10.1027/1015-5759/a000343.
6
7 478 Schmitt, D., Realo, A., Voracek, M., & Allik, J. (2008). Why can’t man be more like a
8
9
479 woman? Sex differences in big five personality traits across 55 cultures. Journal of
10 480 Personality and Social Psychology, 94, 168-172.
11
12 481 Sijtsma, K. (2012). Future of psychometrics: Ask what psychometrics can do for psychology.
13
14 482 Psychometrika, 77, 4-20.
15
16 483 Silva, I. B., & Nakano, T. C. (2011). Modelo dos cinco grandes fatores da personalidade:
17
18
484 análise de pesquisas. Avaliação Psicológica, 10(1), 51-62.
19
20 485 Srivastava, S., John, O. Gosling, S., & Potter, J. (2003). Development of personality in early
21 486 and middle adulthood: Set like plaster or persistent change? Journal of Personality and
22
23 487 Social Psychology, 84(5), 1041-1053.
24
25 488 Tellegen, P. J., & Laros J. A. (2014). SON-R 6-40. Snijders-Oomen Non-verbal intelligence
26
27
489 test. Volume I: Research report. Hogrefe: Göttingen, Germany.
28
29 490 Terraciano, A., McCrae, R. R., Brant, L. J., & Costa, P. T., Jr. (2005). Hierarchical linear
30 491 modeling analyses of the NEO-PI-R scales in the Baltimore longitudinal study of aging.
31
32 492 Psychology and Aging, 20, 493–506.
33
34 493 Urbina, S. (2014). Essentials of psychological testing. Hoboken, New Jersey: Wiley.
35
36 494 Vasconcelos, T. S. de (2005). O inventário fatorial dos cinco fatores de personalidade no
37
38 495 ambiente de trabalho. Tese de Doutorado. Universidade de Brasília, Brasília.
39
40 496 Vasconcellos, S. J. S., & Hutz, C. S. (2008). Construção e validação de uma escala de
41 497 abertura à experiência. Avaliação Psicológica, 7(2), 135-141.
42
43
44
498 Weston, R., & Gore, P. A., Chan, F., & Catalano, D. (2008). An introduction to using
45 499 structural equation models in rehabilitation psychology. Rehabilitation Psychology, 53,
46
47 500 340-356.
48
49 501 Yamagata, S., Suzuki, A., Ando, J., Ono, Y., Kijima, N., Yoshimura, K., ..., & Jang, K. L.
50 502 (2006). Is the genetic structure of human personality universal? A cross-cultural twin study
51
52 503 from North America, Europe and Asia. Journal of Personality and Social Psychology, 90,
53
54
504 987-998.
55
56 505
57
58
59
60
61
62
63
64
65
 26

1 506 Declarations
2
3
4 507 List of abbreviations
5
6 508  ²: chi-square
7
509  AG: Agreeableness
8
9 510  BFI: Big Five Inventory
10 511  BFP: Bateria Fatorial da Personalidade
11
12 512  CFI: Comparative Fix Index
13 513  CO: Conscientiousness
14
15 514  DF: degrees of freedom
16 515  EMR: Escala de Marcadores Reduzidos
17
18 516  ER5FP: Reduced Scale of Big Five Personality Factors
19 517  EX: Extraversion
20
21 518  FFM: five-factor model
22 519  GFI: Goodness of Fit Index
23
24 520  IGFP-5R: Reduced Inventory of Big Five Personality Factors
25 521  NE: Neuroticism
26
27 522  NEO-PI-R: NEO Personality Inventory - Revised
28 523  OP: Openness to Experience
29
30 524  Red5: Escala Reduzida de Descritores de Personalidade
31 525  RMSEA: Root Mean Square Error of Approximation
32
33 526  SRMR: Standardized Root Mean Square Residual
34 527  TIPI: Ten Item Personality Inventory
35
36
528  λ2: Guttman’s coefficient lambda 2
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65