Professional Documents
Culture Documents
Analysis of The Descriptive and Psychometric Characteristics of The Internal Structure of The Ribs in Spanish
Analysis of The Descriptive and Psychometric Characteristics of The Internal Structure of The Ribs in Spanish
To cite this article: Verónica López-Fernández, César Merino-Soto, María Luisa Maldonado
Fruto & César Augusto Orozco Garavito (2019) Analysis of the Descriptive and Psychometric
Characteristics of the Internal Structure of the Ribs in Spanish, Creativity Research Journal, 31:2,
229-235, DOI: 10.1080/10400419.2019.1577123
César Merino-Soto
Universidad de San Martín de Porres
This study analyzed the internal structure of the Runco Ideational Behavior Scale
(RIBS) in a sample of 116 Spanish-speaking individuals. The characteristics of the
items and the structural relationship with the construct were analyzed. The results
supported a 2-factor model and showed good discrimination. The reliability of the
scores was satisfactory for group and individual descriptive purposes. The application
of hierarchical modeling seemed to justify a second order dimension, but showed little
variance. The results obtained are discussed and an interpretation of the findings is
proposed. The RIBS can be an adequate and relevant instrument in Spanish sample to
measure creativity.
Creativity is certainly an important variable in different fields of itself. There are several tools for assessing creativity. In
life (Runco, Plucker, & Lim, 2000-2001). It could be defined as Spain, there are specific instruments such as the CREA test
a capacity that can be developed and potentialized, which allows (Corbalán et al., 2003), which is a cognitive measure of
obtaining new, original, and valid products (Runco & Jaeger, creativity that assesses the subject’s ability to generate
2012). For Pizarro, Detweiler-Bedell, and Bloom (2006), crea- questions in the theoretical context of searching and pro-
tivity is linked to functions of flexibility and innovation pro- blem-solving. Another relevant instrument in this popula-
cesses. This is the reason why creative thinking stands out for tion is the Test de Marín (1995), which seeks an overall
originality, expressiveness, and flexibility, and is characterized view of the state of development of its creative potential-
by finding novel ideas facing different problem situations. ities and consists of a verbal part and a graphic part with
To enhance this ability, it is necessary that a starting two exercises. In addition, another instrument of great
point of such a construct can be evaluated. However, the importance is the Prueba de Imaginación Creativa (PIC)
evaluation of creativity is complex because the construct is test (Artola, Ancillo, Barraca, Mosteiro, & Pina, 2004) that
measures divergent thinking through three verbal and one
graphic game. Finally, the Test de Creatividad
Infantil (TCI) (Romo, Alfonso Benlliure, & Sánchez Ruíz,
Address correspondence to Verónica López-Fernández, Profesora
e Investigadora de la Universidad Internacional de La Rioja (UNIR),
2008) is an instrument that seeks the evaluation of culture
Avda. La Paz 137, Logroño, La Rioja 26006, España. E-mail: veronica. fair creative thinking through a preliminary drawing. These
lopez@unir.net performance tests would be based on more objective
230 LÓPEZ-FERNÁNDEZ ET AL.
asymmetry and kurtosis statistics (Fisher, 1970), which could be a possible acceptance that the departure from uni-
served as inputs to evaluate statistical normality (Jarque & variate normality may not be strong. The multivariate tests
Bera, 1987). (Small, 1980) for asymmetry Q1 = 61.68, gl = 23, p < 0.01)
Prior to the main analyzes, some data quality controls and normality (VQ3 = 151.88, gl = 46, p < 0.01) indicated that
were performed regarding response rate and response pat- the nonmultivariate normality of the items. On the other hand,
terns. A complete questionnaire response rate of 93.1% was the dip test indicated that in all items, except item 10, the
obtained; 6 (5.2%) and 2 (1.7%) subjects did not respond to unimodality could not be accepted. Therefore, the items
1 or 2 items respectively. Consequently, given this low obtained more than one modal categories of response.
percentage of lost values in the 9 subjects, their absent
responses were replaced by the modal response in the
Structural analysis
corresponding item. About the acquiescent response,
a subject responded with option 5 to the 22 items of the Regarding the structural analysis, a semiconfirmatory
RIBS. Its removal did not produce any appreciable impact, approach was applied, using the Procrustes rotation of the
so it was assumed that a subject could express such inten- data matrix toward a target matrix containing the hypothe-
sity of response in the construct and it was maintained in sized location of the items in their factors (Ferrando, Varea,
the database. No other suspicious patterns were detected & Lorenzo, 1999; Laher, 2010; McCrae, Zonderman, Bond,
(extreme response, central response, etc.). & Paunonen, 1996). This target matrix was specified as
About the univariate item analysis, the distributional follows: Items of factor 1 were allowed to load on it,
results indicated, in general, mild or moderate negative or without some information about its magnitude, and its
positive asymmetric distributions, and mainly negative kurto- load was set to zero in factor 2; this same specification
sis, indicating that the orientation of the responses is not the was applied for factor 2. To evaluate the adjustment in this
same (see Table 1). The combined effect of asymmetry and approach, the coefficient of congruence Π (Burt, 1948;
distortions on kurtosis did not have a serious enough impact to Tucker, 1951) applied to each item (Πi), to the factors
reject the hypothesis of distributional normality of the items, (Πf) and to the total solution (ΠT); was used; it quantifies
except in two items, 10 and 14. Compared with the distribu- the degree of similarity between two vectors of factorial
tion 2 (gl = 2), the JB coefficient for both items showed that load. This comparison is made between the empirically
normality could be not be accepted. In the rest of the items, it extracted factor loads and the hypothetical specification of
TABLE 1
Descriptive information for items of RIBS
Note. M = median. Mo = mode. SD = standard deviation. Min = minimum. Max = maximum. IOV = ordinal variation index. Dip = unimodality test. g1
= asimetria. g2 = kurtosis. JB = omnibus normality test.
232 LÓPEZ-FERNÁNDEZ ET AL.
the items in their expected factors (target matrix). The force to the general factor because there they all had high
adjustment criterion is Π ≥ 0.90 (Lorenzo-Seva & Ten factor loads. The factor 1 items maintained their heteroge-
Berge, 2006; McCrae et al., 1996). The adjustment was neity in the magnitude of their factorial loads. Except for
also evaluated by adding a measure of simplicity of the three items of F1, it is also observed that the loads in Fg
factor solution for the items, factors, and total solution were above 0.30 constant criterion to assess the factor loads
(Fleming, 2003; Fleming & Merino, 2005). Finally, to (Wolff & Preising, 2005), several were close to this limit,
examine the feasibility of the interpretation of a higher- and that the factorial strength of the Items varied according
order dimension, the Schmid and Leiman (1957) transfor- to the first order factor. The total variance explained by Fg
mation was applied from the exploratory framework, to was 43.5%; the F1 and F2 factors accounted for 39.6%
differentiate the variance of the first-order factors (factors (32.5% and 7.1%, respectively), both not substantially
1 and 2) against a common or general dimension. The different.
greater the proportion of variance explained by the com- The internal consistency reliability (α) for F1 and F2
mon general factor, the interpretation of a single general was 0.91 and 0.83, respectively. The differences between
dimension could be ensured (Wolff & Preising, 2005). these reliabilities were statistically significant: 2 (gl:
The Bartlett statistic (2, gl: 253 = 1569.5, p < 0.01) and the 1) = 13.09, p < 0.001 (Lautenschlager & Meade, 2008).
Kaiser-Meyer-Olkin test (KMO = 0.885) verified that the cor- Using the loads of the first factorial analysis, using the
relation matrix was not identity. Two solutions were extracted, coefficient ω (McDonald, 1999), the reliability was 0.85
one of two factors and one of a factor. The 1-factor solution and 0.83. The discrepancies found in F1 appear to reflect
yielded RMSR = 0.111; the two-factor model, RMSR = 0.066; correlated errors between some items (Raykov, 2001).
compared to the criterion derived from the standard error of Comparing the value of α in men and women, F1 was
correlation (.093; Kelley, 1935), the two-factor model seems to equal for both groups (0.92), and in F2, it was higher in
be more representative of the data structure. women (0.84 vs. 0.80). However, these differences were
The results of the factorial analysis (Table 2) showed the not statistically significant, W = 1.17, F (51, 24) = 0.577,
following: F2 factor loads were > 0.65 (except item 11); F1 p = 0.33 (Merino-Soto, 2016).
showed seven items at or below 0.40; even its item 1 was
completely divergent because it loaded on the other factor. TABLE 2
The evaluation of the adjustment using the coefficient of Factor analysis results
congruence (Π) for these factors (ΠF) was acceptable for
F2, but unacceptable for F1. The impact of items with low Análisis Factorial Transformación Schmid-
Semiconfirmatorio Leiman
factor load, most related to their low congruence coefficients
(Πi), is the main source of the low F1 adjustment. The F1 F2 Πi IFS F1 F2 Fg
adjustment of the total solution of two factors (ΠT = 0.886)
Run1 −0.014 0.646 −0.022 0.999 −0.011 0.309 0.558
was also below the critical criterion (Π < 0.90), indicating Run2 0.734 −0.033 0.999 0.996 0.577 −0.016 0.424
the questionable general adjustment. Run3 0.391 0.066 0.986 0.945 0.307 0.032 0.299
On the other hand, the load of the items in the noncorre- Run4 0.822 −0.197 0.973 0.891 0.647 −0.094 0.335
sponding factor was essentially close to zero, the degree of Run5 0.717 −0.097 0.991 0.964 0.564 −0.046 0.358
factorial simplicity of each item (IFS) confirmed that. Table Run6 0.567 0.070 0.992 0.970 0.446 0.034 0.412
Run7 0.270 0.370 0.589 0.305 0.212 0.177 0.492
2 also shows these results, observing a reduced factorial Run8 0.862 −0.150 0.985 0.941 0.678 −0.072 0.400
complexity of the items (around IFS = 0.90), except for Run9 0.820 −0.228 0.964 0.856 0.645 −0.109 0.306
items 7, 12, and 13, whose complexity was excessively Run10 0.436 −0.031 0.998 0.990 0.343 −0.015 0.242
high. The degree of factorial simplicity for the solution of Run12 0.259 0.386 0.556 0.379 0.203 0.185 0.499
F1 (0.96), F2 (0.91), and total (0.95) was high. The correla- Run13 0.422 0.360 0.761 0.158 0.332 0.172 0.577
Run19 0.380 0.058 0.988 0.954 0.299 0.028 0.286
tion between the factors (0.56) was moderately high, which Run20 0.632 0.107 0.986 0.944 0.497 0.051 0.485
indicates that there was a significant amount of linear depen- Run21 0.709 0.101 0.990 0.960 0.557 0.048 0.526
dence between them. To evaluate the dependence of these Run22 0.785 −0.076 0.995 0.981 0.617 −0.036 0.418
factors in relation to the interpretation of a single factor, the Run23 0.677 0.098 0.990 0.959 0.532 0.047 0.504
results of the Schmid Leiman (SL) transformation are evi- Run11 −0.094 0.440 0.978 0.913 −0.074 0.211 0.329
Run14 0.018 0.749 1.000 0.999 0.014 0.358 0.669
denced. The application of the SL transformation showed Run15 0.011 0.707 1.000 1.000 0.009 0.338 0.628
that the introduction of the general factor (Fg) reduced the Run16 0.072 0.665 0.994 0.977 0.057 0.318 0.628
variance of the items in their factors, obtained in the first Run17 0.090 0.684 0.991 0.966 0.071 0.327 0.656
analysis. On the other hand, except for two items, the rest Run18 −0.175 0.748 0.974 0.896 −0.138 0.358 0.549
may still be representative of their factors but the strength of ΠF 0.844 0.915
this relationship varies substantially in factor 2 (F2). Indeed, Note. Π: Coefficients of congruence for the items (Πi) y factors (ΠF).
the strength of their items was reduced and they gave more IFS - index of factorial simplicity. Fg = General factor
CREATIVITY RESEARCH JOURNAL 233
research. It is necessary to continue researching the proper- Hartigan, J. A., & Hartigan, P. M. (1985). The dip test of unimodality.
ties of this instrument to evaluate creativity, for its important Annals of Statistics, 13, 70–84. doi:10.1214/aos/1176346577
Jarque, C. M., & Bera, A. K. (1987). A test for normality of observations
implications in the lives of people and in the advancement of
and regression residuals. International Statistical Review, 55, 163–172.
societies within an intercultural framework. doi:10.2307/1403192
Kālis, E., & Roķe, L. (2011). Adaptation of runco ideational behavior
scale in Latvia. Journal of Pedagogy and Psychology Signum Temporis,
ACKNOWLEDGMENTS 4(1), 36–45. doi:10.1037/e578442014-162
Kelley, T. L. (1935). Essential traits of mental life. Harvard studies in
education (Vol. 26, pp. 146). Cambridge, MA: Harvard University
We thank all the people who have collaborated in this Press.
research, especially Manuela Mena, María Luisa Maldonado Laher, S. (2010). Using exploratory factor analysis in personality research:
and César Augusto Orozco for their support in the translation. Best-practice recommendations. SA Journal of Industrial Psychology,
36. doi:10.4102/sajip.v36i1.873
Lautenschlager, G. J., & Meade, A. W. (2008). AlphaTest: A windows pro-
DISCLOSURE STATEMENT gram for tests of hypotheses about coefficient alpha. Applied Psychological
Measurement, 23, 502–503. doi:10.1177/0146621607312307
López, V. (2015). Importancia de la valoración de la creatividad desde las
No potential conflict of interest was reported by the authors. bases neuropsicológicas [Importance of the assessment of creativity
from the neuropsychological bases]. In P. Martín-Lobo & E. Vergara-
Moragues (Eds.), Procesos e instrumentos de evaluación
ORCID neuropsicológica educativa (pp. 140–161). Madrid, España: Ministerio
de Educación, Cultura y Deporte.
Verónica López-Fernández http://orcid.org/0000-0003- Lorenzo-Seva, U., & Ten Berge, J. M. F. (2006). Tucker’s Congruence
Coefficient as a meaningful index of factor similarity. Methodology, 2,
0483-5884 57–64. doi:10.1027/1614-2241.2.2.57
César Merino-Soto http://orcid.org/0000-0002-1407- Marín, R. (1995). La creatividad: Diagnóstico, evaluación e investigación.
8306 Madrid, España: Universidad Nacional de Educación a Distancia,
UNED.
McCrae, R. R., Zonderman, A. B., Bond, M. H., & Paunonen, S. V.
(1996). Evaluating replicability of factors in the Revised NEO
REFERENCES Personality Inventory: Confirmatory factor analysis versus Procrustes
rotation. Journal of Social and Personality Psychology, 70, 552–566.
Artola, T., Ancillo, I., Barraca, J., Mosteiro, P., & Pina, J. (2004). PIC, doi:10.1037/0022-3514.70.3.552
Prueba de Imaginación Creativa. Madrid, España: TEA Ediciones. McDonald, R. P. (1999). Test theory: A unified approach. Mahwah, NJ:
Berry, K. J., & Mielke, P. W., Jr. (1994). A test of significance for the Lawrence Erlbaum Associates.
index of ordinal variation. Perceptual and Motor Skills, 79, 1291–1295. Merino, C., & Grimaldo, M. (2010). Complejidad factorial de la permisi-
Burt, C. (1948). The factorial study of temperament traits. British Journal of vidad moral hacia las conductas morales controvertidas [Factorial
Psychology, Statistical Section, 1, 178–203. doi:10.1111/j.2044- Complexity of Moral Permissiveness toward controversial moral
8317.1948.tb00236.x behaviors]. Interdisciplinaria, 27(2), 297–314.
Chen, K. T. (2015). Assessing a Chinese version of the Runco Ideational Merino-Soto, C. (2016). Diferencias entre coeficientes alfa de Cronbach,
Behavior Scale. Social Behavior and Personality: an International con muestras y partes pequeñas: Un programa VB [Differences between
Journal, 43(7), 1111–1122. doi: 10.2224/sbp.2015.43.7.1111 Cronbach’s alpha, in Little samples and parts: A VB program]. Anales
Corbalán, F. J., Martínez, F., Alonso, C., Donolo, D., Tejerina, M., & de Psicología, 32(2), 587–588. doi:10.6018/analesps.32.2.203841
Limiñana, R. M. (2003). CREA. Inteligencia Creativa. Una medida Navarro, J., Merino, C., Dominquez, S., & Fleming, J. (2016). Importancia de la
cognitiva de la creatividad. Madrid, España: TEA Ediciones. evaluación de la simplicidad factorial: Re-análisis a Zicado, Palma y Garrido
Ferrando, P., Varea, M., & Lorenzo, U. (1999). Evaluación psicométrica del (2012) [carta al Editor] [Importance of the evaluation of the factorial simpli-
cuestionario de ansiedad y rendimiento (CAR) en una muestra de escolares city: Re-analysis to Zicado, Palma and Garrido (2012)]. Revista
[Psychometric evaluation of the anxiety and performance questionnaire Latinoamericana de Ciencias Sociales, niñez y Juventud, 1, 1682–1684.
(CAR) in a sample of schoolchildren]. Psicothema, 11, 225–236. Pizarro, D. A., Detweiler-Bedell, B., & Bloom, P. (2006). The creativity of
Ferrando, P. J., & Anguiano-Carrasco, C. (2010). El análisis factorial everyday moral reasoning: empathy, disgust and moral persuasion. In
como técnica de investigación en psicología [Factor analysis as J. C. Kaufman & J. Baer (Eds.), Creativity and reason in cognitive develop-
a research technique in psychology]. Papeles del Psicólogo, 31, 18–33. ment (pp. 81–98). Cambridge, England: Cambridge University Press.
Fisher, R. A. (1970). Statistical methods for research workers (14th ed.). Raykov, T. (2001). Bias of coefficient alpha for fixed congeneric measures
Edinburgh, Scotland: Oliver & Boyd. with correlated errors. Applied Psychological Measurement, 25, 69–76.
Fleming, J. S. (2003). Computing measures of simplicity of fit for loadings doi:10.1177/01466216010251005
in factor-analytically derived scales. Behavior Research Methods, Romo, M., Alfonso Benlliure, V., & Sánchez Ruíz, M. J. (2008). TCI: Test
Instruments & Computers, 35(4), 520–524. doi:10.3758/bf03195531 de Creatividad Infantil. Madrid, España: TEA Ediciones.
Fleming, J. S., & Merino, C. (2005). Medidas de simplicidad y de ajuste Runco, M. A., & Jaeger, G. J. (2012). The standard definition of creativity.
factorial: Un enfoque para la construcción y revisión de escalas derivadas Creativity Research Journal, 24(1), 92–96. doi:10.1080/
factorialmente [Measures of simplicity and factorial adjustment: An 10400419.2012.650092
approach to the construction and revision of factorially derived scales]. Runco, M. A., Plucker, J., & Lim, W. (2000-2001). Development and
Revista de Psicología de la Pontificia Universidad Católica del Perú, 23 psychometric integrity of a measure of ideational behavior. Creativity
(2), 252–266. Research Journal, 13, 393–400. doi:10.1207/s15326934crj1334_16
CREATIVITY RESEARCH JOURNAL 235
Schmid, J., & Leiman, J. (1957). The development of hierarchical factor 12. Realizaría un curso que estuviera basado en ideas
solutions. Psychometrika, 22(1), 53–61. doi:10.1007/bf02289209 originales.
Small, N. J. H. (1980). Marginal skewness and kurtosis in testing multi-
variate normality. Applied Statistics, 29, 85–87. doi:10.2307/2346414
13. Soy capaz de pasarme horas pensando detenida-
Tucker, L. R. (1951). A method for synthesis of factor analysis studies. mente en una idea.
Personnel Research Section Report, 984. Washington, DC: Department 19. Procuro ejercitar mi mente analizando las cosas en
of the Army. profundidad.
Wolff, H. G., & Preising, K. (2005). Exploring item and higher order 20. Soy capaz de dar con respuestas a problemas para
factor structure with the Schmid-Leiman solution: Syntax codes for
SPSS and SAS. Behavior Research Methods, Instruments &
los que todavía no se han encontrado soluciones.
Computers, 37, 48–58. doi:10.3758/bf03206397 21. Soy capaz de combinar ideas de formas que a otros
jamás se les había ocurrido.
22. Mis amigos me piden que les ayude a pensar ideas
APPENDIX y soluciones.
23. Se me ocurren ideas para generar inventos o mejorar
Escala de evaluación del comportamiento ideacional de cosas.
Runco (The Runco Ideational Behavior Scale, RIBS)
Instructions in Spanish:
A continuación se muestran afirmaciones de las que has
de responder en una escala de 1 a 5. Siendo 1 que nunca; 2 Segundo factor
pocas veces; 3 algunas veces; 4 muchas veces y 5 siempre
o casi siempre
11. La gente suele considerar mis ideas poco prácticas
Primer factor e incluso alocadas.
14. A veces me intereso tanto por una nueva idea que me
1. Se me ocurren muchas ideas alocadas. olvido de mis obligaciones.
2. Suelo tener ideas con más frecuencia que la mayoría 14. A veces me intereso tanto por una nueva idea que me
de la gente. olvido de mis obligaciones.
3. Normalmente me entusiasmo con las ideas que tengo. 14. A veces me intereso tanto por una nueva idea que me
4. Se me ocurren muchas ideas o soluciones a problemas. olvido de mis obligaciones.
5. Doy con ideas y soluciones que a la mayoría de la 16. Cuando estoy escribiendo un artículo o hablando en
gente nunca antes se le han ocurrido. público, muchas veces me cuesta ceñirme a un solo
6. Me gusta jugar a generar ideas por pura diversión. tema como consecuencia de todas las ideas que se
7. Considero que es importante ser capaz de pensar en me van ocurriendo.
alternativas alocadas y disparatadas. 17. En muchas ocasiones siento cómo una idea me con-
8. Me considero una persona con gran capacidad para duce a otras ideas que, a su vez, me conducen a otras
generar nuevas ideas. y al final me encuentro con una idea que no sé muy
9. Siempre he sido un pensador activo – siempre se me bien de dónde procede.
ocurren ideas. 18.Mucha gente podría pensar que soy algo despistado/a
10. Me gusta tener libertad y margen para decidir sobre o distraído/a porque a veces pienso en muchas cosas
las cosas que hago. a la vez.