Documentos de Académico
Documentos de Profesional
Documentos de Cultura
143
Jesús-Miguel MUÑOZ-CANTERO y Luisa LOSADA-PUENTE
the evaluation of this construct has under- ing to the presence or absence of neurologi-
gone notable advances in our country due to cal development disorders. In conclusion, the
the acceptance of international theoretical ARC-INICO scale provides an important ba-
revista española de pedagogía
models and the design of specific instruments sis for decisions making regarding the design
for our context. The ARC-INICO scale (Ver- of care programs, through the development
dugo et al., 2014) assesses four characteris- of resources, guidelines and strategies, and
tics of self-determined behavior in teenagers: provides information for the differential pro-
autonomy, self-regulation, empowerment and vision of said resources and supports.
self-concept. This structure is based on the
Wehmeyer’s Functional Model (1999, 2003). Keywords: self-determination, students,
It has only been validated with Spanish stu- questionnaire, test reliability, test validity.
144
Validación del constructo de autodeterminación a través de la escala ARC-INICO para adolescentes
duales, a fin de establecer planes de acción en una escala de tipo Likert de 3 puntos,
individualizados y grupales. en el caso de la sección de autonomía y de
4 puntos, en el resto de las secciones. Se
revista española de pedagogía
rianza escalar (supone, además, varianzas res de asimetría z(G1) de las parcelas son
iguales para los errores). Para estimar el superiores e inferiores a ± 1.96, excep-
ajuste de los datos, se utilizaron varios ín- to en la parcela P1_4, donde z(G1) = .29
revista española de pedagogía
dices que permiten seleccionar, de entre indica una distribución simétrica. Por lo
los dos modelos considerados, aquel que tanto, la hipótesis nula según la cual la
tiene una menor discrepancia con respecto distribución es simétrica se rechaza en
al modelo verdadero. Estos índices son: el casi todos los casos. Además, el cálculo
Criterio de Información de Akaike (AIC) y de la curtosis indica que se incumple la
el Índice de Ajuste Comparativo (CFI) de hipótesis nula según la cual la distribu-
Bentler. ción es mesocurtica (z[G2] > ± 1.96) en
la mayoría de los casos, excepto en las
Los análisis fueron apoyados por el parcelas P2_1 (z [G2] = 1 587), P2_3 (z
Paquete Estadístico de IBM para Cien- [G2] = −1 067), P3_2 (z [G2] = −1,442),
cias Sociales (IBM SPSS) versión 23.0, y P3_3 (z [G2] = 1.712) y P4_1 (z
en el caso de las operaciones de Análisis [G2] = 1.376). En el conjunto de contras-
Factorial Confirmatorio (AFC) se llevaron te de asimetría y curtosis no se cumple
a cabo utilizando el programa IBM SPSS el supuesto de normalidad univarian-
AMOS 23.0. te, ya que en todos los casos k2 > 5.98.
148
Tabla 1. Análisis de la unidimensionalidad y contrastes univariados de normalidad de la Escala ARC-INICO.
% Asimetría Curtosos Contraste K-S
va- α
Valores
Parce- rianza estan-
Secciones propios
las expli- dariza-
a z ET z (G1) z ET z (G2) K2 K-Sª p
cada do
b
P1_1 4.95 82.442 .957 .28 .052 5.46 −.43 .104 −4.16 47.16 .337 .00
P1_2 4.45 55.637 .884 −.79 .052 −15.23 −.40 .104 −3.87 246.92 .179 .00
Autonomía
P1_3 4.27 71.179 .917 −.18 .052 −3.40 −.87 .104 −8.35 81.24 .188 .00
P1_4 3.72 74.476 .913 .02 .052 0.29 .30 .104 2.86 8.24 .244 .00
P2_1 4.81 96.146 .990 −.23 .052 −4.48 .17 .104 1.59 22.60 .319 .00
Autorregu-
P2_2 3.84 96.006 .986 −.39 .052 −7.52 .22 .104 2.09 60.89 .288 .00
lación
P2_3 2.45 59.861 .857 −.32 .052 −6.15 −.11 .104 −1.07 39.01 .294 .00
P3_1 3.87 77.439 .911 −.48 .052 −9.31 .26 .104 2.53 93.04 .147 .00
Empodera-
P3_2 3.64 90.924 .967 −.27 .052 −5.21 −.15 .104 −1.44 29.24 .253 .00
miento
P3_3 2.88 98.648 .978 −.44 .052 −8.46 .18 .104 1.71 74.54 .311 .00
P4_1 3.79 94.781 .972 −.25 .052 −4.87 .14 .104 1.38 25.04 .307 .00
Autocono-
P4_2 3.45 86.288 .947 −.98 .052 −18.92 .71 .104 6.83 404.69 .204 .00
cimiento
P4_3 2.95 73.803 .881 −.71 .052 −13.67 .52 .104 5.01 212.05 .174 .00
Nota:
a. Valores propios: reflejan, en orden decreciente, la cantidad de variación del conjunto de variables que explica el factor, o lo que es lo mismo,
la cantidad de información que aporta cada variable al factor, siendo la suma de todos los autovalores igual al número de variables que han
sido introducidas en el análisis.
b. % Varianza explicada: porcentaje de varianza es representado por el conjunto de ítems incluidos.
Siglas y Estadísticos: α − índice Alpha de Cronbach; Contraste K-S: Contraste Prueba de Kolmogorov-Smirnov, para el cálculo de la norma-
lidad de la distribución; z − Valor estadístico que hace referencia a la desviación estándar obtenida en una variable, respecto del valor de la
media; ET: Error Típico, que indica las oscilaciones existentes respecto al valor Z.
Fuente: Elaboración propia.
Validación del constructo de autodeterminación a través de la escala ARC-INICO para adolescentes
149
revista española de pedagogía
revista española de pedagogía
150
año 77, nº 272, enero-abril 2019, 143-162
P1_1 1
P1_2 .399** 1
P3_3 .303** .466** .385** .205** .300** .376** .358** .349** .359** 1
P4_1 .202** .393** .318** .194** .352** .387** .311** .387** .315** .320** 1
Jesús-Miguel MUÑOZ-CANTERO y Luisa LOSADA-PUENTE
P4_2 .137** .157** .110** .129** .228** .233** .227** .274** .186** .098** .224** 1
P4_3 −.012 .074** .054* .081** .153** .144** .067** .224** .141** .096** .241** .188** 1
Nota: * p < .05 ** p < .001. Este valor hace referencia a la probabilidad asociada al estadístico, que informa de la aceptación de la hipótesis por la
que se considera que existen correlaciones estadísticamente significativas, teniendo un nivel de confianza asociado de 95 %.
Fuente: Elaboración propia.
Validación del constructo de autodeterminación a través de la escala ARC-INICO para adolescentes
3.2.1. Modelo 1: Modelo Unidimensional de 0), con errores de predicción (e) que
El primer modelo examinado supone varían de .01 a .60, de modo que los co-
la existencia de un único factor, que ex- eficientes de correlación al cuadrado (r2)
plica la covariación de todos los ítems de pueden variar de .99 y .40. Casi todos los
la prueba. Corresponde a una concepción r2 son superiores a .5, a excepción de P1_2
teórica unidimensional de la autodetermi- y P1_3. Estos resultados indican que la
nación. Dada la inexistencia de evidencias proporción de varianza de las variables
empíricas que apoyen la unidimensionali- observadas que puede explicarse por el
dad de este constructo, este modelo se usa factor latente (autodeterminación) se
como un elemento de contraste con respec- aproxima al valor apropiado, siempre que
to a las posibles estimaciones multidimen- se eliminen las variables que están ale-
sionales. jadas de los valores óptimos. Las cargas
factoriales oscilan entre .07 y .78, siendo
El Gráfico 1 muestra la solución es- seis de ellas inferiores a .6. Esto podría
tandarizada inicial para el modelo 1. indicar que el modelo de factor único no
Todos los coeficientes son significativos es suficiente para reproducir la matriz de
(valores t significativamente diferentes covarianza original.
Al analizar el ajuste del modelo a los los índices RMR y RMSEA (RMR = .648;
datos empíricos, cuyos datos se presen- RMSEA = .057), existiendo además, un
tan en la Tabla 3, se observan índices error de especificación (p = .007). Fren-
de ajuste global deficientes (χ2 528.929; te a ello, existen algunos índices de ajus-
χ2 / gl = 8.145; p < .000), y un ajuste par- te parcial con resultados satisfactorios
cial mejorable en su evaluación mediante (GFI = .958; AGFI = .941).
152
Tabla 3. Estadísticos de bondad de ajuste.
Ajuste parcial
Ajuste absoluto
.057 .262
Modelo 1 528.929 65 .000 8.137 .648 .958 .941 .007 580.93 755.27
[.052 − .061] [.230 − .297]
.049 .199
Modelo 2 420.1 60 .000 6.334 .553 .970 .954 .623 442.03 649.89
[.044 − .054] [.173 − .299]
Nota: χ2: Ji-Cuadrado; gl: grados de libertad; χ2 / gl: χ2 relativo; RMR: Raíz Cuadrática Media Residual; GFI: Índice de Bondad de Ajuste; AGFI:
Índice de Bondad de Ajuste Ajustado; RMSEA: Error Cuadrático Medio de Aproximación; AIC: Criterio de Información de Akaike; CAIC: Crite-
rio de Información Consistente de Akaike; ECVI: Índice de Validación Cruzada Esperada.
Fuente: Elaboración propia, a partir de IBM SPSS AMOS 23.
Validación del constructo de autodeterminación a través de la escala ARC-INICO para adolescentes
153
revista española de pedagogía
Jesús-Miguel MUÑOZ-CANTERO y Luisa LOSADA-PUENTE
Sin restricciones 470.906* 120 .966 .949 .663 .718 .036 594.9
solución de cuatro factores resulta adecuada tamente las variables del modelo, así como
para estos subgrupos de la muestra. con respecto al modelo (Tabla 5).
revista española de pedagogía
Autodeterminación 1
Autonomía .846 1
156
Validación del constructo de autodeterminación a través de la escala ARC-INICO para adolescentes
157
Jesús-Miguel MUÑOZ-CANTERO y Luisa LOSADA-PUENTE
Así mismo, en relación al AFC mul- Este estudio presenta algunas impli-
ti-grupo, se ha obtenido un ajuste ade- caciones para las prácticas educativas. La
cuado de los datos al modelo de factor educación tiene un papel especial en la ca-
de orden superior en ambas muestras: pacitación y el apoyo a los estudiantes, es-
adolescentes con y sin Trastornos del pecialmente a los más vulnerables, para que
Desarrollo Neurológico. Ello sugiere la puedan adquirir el control y la responsabili-
viabilidad de este modelo multidimensio- dad de sus acciones y decidir cómo vivir sus
nal para medir la autodeterminación de propias vidas (Arellano y Peralta, 2013; Lee
todos los estudiantes. Además, la compa- et al., 2012; Wehmeyer et al., 2012; Wehme-
ración entre los grupos refleja diferencias yer y Shogren, 2018; etc.). Dichos apoyos ha-
significativas de mayor nivel de autode- cen referencia a aspectos como «la provisión
terminación en aquellos que no presen- de intervenciones profesionales, la creación
tan Trastornos del Desarrollo Neurológi- de entornos y la prestación de estrategias de
co frente a los que sí presentan alguno. apoyos individualizados» (Schalock, 2018,
En este sentido, varios estudios revelan p. 12) que requieren una evaluación previa
peores resultados de autodeterminación que permita determinar la necesidad de apo-
en alumnos con diversas necesidades de yo y una evaluación final de la eficacia de las
apoyo (Cho, Wehmeyer y Kinston, 2013; intervenciones sobre los resultados persona-
Chou, Wehmeyer, Palmer y Lee, 2016; les. De ahí la importancia de desarrollar y
Vega et al., 2013). validar herramientas sólidas de diagnósti-
co para, a partir de ellas, diseñar acciones
Estas evidencias sugieren que la Es- orientadas a la mejora educativa, personal
cala ARC-INICO es una escala válida y y social.
año 77, nº 272, enero-abril 2019, 143-162
presentación adecuada de este construc- para corroborar o refutar los datos obte-
to. Al respecto, es posible referirse a otros nidos en el presente estudio, teniendo en
estudios con resultados similares. Por cuenta también las limitaciones constata-
ejemplo, Verdugo et al. (2014) y Vicente das en este, tales como:
et al. (2015) han llevado a cabo estudios
de las propiedades psicométricas de esta a) La amenaza a la posibilidad de generalizar
escala, mediante análisis factorial explo- estos resultados dado que las personas par-
ratorio y confirmatorio, que apoyan una ticipantes son solo estudiantes gallegos, lo
estructura compuesta de cuatro factores que podría subsanarse ampliando la mues-
independientes. Del mismo modo, estos tra a otras comunidades autónomas.
estudios apoyan los resultados hallados
en relación a los valores de los coeficien- b) El uso de una medida de autoinforme pro-
tes de correlación entre factores de pri- cedente del propio instrumento de evalua-
mer orden y con respecto al factor de se- ción, y que puede contener los sesgos que
gundo orden. se derivan de la deseabilidad social.
158
Validación del constructo de autodeterminación a través de la escala ARC-INICO para adolescentes
Chou, Y.-C., Wehmeyer, M. L., Palmer, S. B. y Lee, Peralta, F. y Arellano, A. (2014). La autodetermina-
J. (2016). Comparisons of self-determination ción de las personas con discapacidad intelectual:
among students with autism, intellectual dis- Situación actual en España. Revista CES Psico-
ability, and learning disabilities: A multivari- logía, 7 (2), 59-77. Recuperado de http://revistas.
able analysis. Focus on Autism and Other De- ces.edu.co/index.php/psicologia/article/view/2891
velopmental Disabilities, 14, 1-9. doi: https:// Prieto, G. y Delgado, A. R. (2010). Fiabilidad y
doi.org/10.1177/1088357615625059 validez. Papeles del Psicólogo, 31 (1), 67-74.
Elosua, P. (2005). Evaluación progresiva de la in- Recuperado de www.papelesdelpsicologo.es/
varianza factorial entre las versiones original resumen?pii=1797 (Consultado el 16/02/2017).
y adaptada de una escala de autoconcepto. Psi- Schalock, R. L. (2018). Seis ideas que están cam-
cothema, 17 (2), 356-362. Recuperado de http:// biando el campo de las discapacidades inte-
www.psicothema.com/psicothema.asp?id=3112 lectuales y del desarrollo en todo el mundo.
(Consultado el 12/02/2017). Siglo Cero: Revista Española sobre Discapaci-
dad Intelectual, 49 (1), 7-19. doi: http://dx.doi.
Griffin, L. K., Adams, N. y Little, T. D. (2017).
org/10.14201/scero2018491719
Self-determination theory, identity develop-
Shogren, K. A. y Wehmeyer, M. L. (2016). Self-de-
ment, and adolescence. En M. L. Wehmeyer,
termination and goal attainment. En M. L. We-
K. A. Shogren, T. D. Little y S. J. López (Eds.),
hmeyer y K. A. Shogren (Eds.), Handbook of
Development of self-determination through the
research-based practices for educating students
life-course (pp. 189-196). Nueva York: Springer with intellectual disability (pp. 255-273). Nue-
Link. va York: Routledge.
Kelloway, E. K. (2014). Using Mplus for structural Shogren, K. A., Wehmeyer, M. L. y Burke, K. M.
equation modeling: A researcher’s guide. Nue- (2017). Self-determination. En K. A. Shogren,
va York: Sage. M. L. Wehmeyer y N. N. Singh (Eds.), Hand-
Kline, R.B. (2015). Principles and practice of struc- book of Positive Psychology in Intellectual and
tural equation modeling. Nueva York: The Developmental Disabilities (pp. 49-64). Nueva
Guildford Press. York: Springer.
año 77, nº 272, enero-abril 2019, 143-162
Lee, Y., Wehmeyer, M., Palmer, S., William-Diehm, Shogren, K. A., Wehmeyer, M. L., Palmer, S. B.,
K., Davies, D. y Stock, S. (2012). Examining Forber-Pratt, A. J., Little, T. J. y Lopez, S.
individual and instruction-related predictors (2015). Causal agency theory: Reconceptual-
of the self-determination of students with dis- izing a functional model of self-determination.
Education and Training in Autism and Devel-
revista española de pedagogía
Vega, C., Gómez-Vela, M., Fernández-Pulido, R. Wehmeyer, M. L. (1995). The arc’s self-determina-
y Badía, M. (2013). El papel del contexto edu- tion scale: procedural guidelines. Texas: The
cativo en la autodeterminación. Análisis de su ARC of United States.
influencia en el proceso de transición a la vida Wehmeyer, M. L. (1999). A functional model
adulta de alumnos con discapacidad intelectual. of self-determination: describing develop-
Revista Iberoamericana de Educación, 63, 19- ment and implementing instruction. Fo-
33. Recuperado de https://rieoei.org/RIE/article/ cus on Autism and Other Developmental
view/418 (Consultado el 22/06/2017). Disabilities, 14 (1), 53-61. doi: https://doi.
org/10.1177/108835769901400107
Verdugo, M. A., Vicente, E., Fernández-Pulido,
R., Gómez-Vela, M., Wehmeyer, M. L. y Gui- Wehmeyer, M. L. (2003). A functional theory of
self-determination: model overview. En M. L.
llén, V. M. (2015). Evaluación psicométrica de
Wehmeyer, B. Abery, D. E. Mithaug y R. Stan-
la escala ARC-INICO de autodeterminación
cliffe (Eds.), Theory in Self-Determination:
para adolescentes con discapacidad intelectual. foundations for Educational Practice (pp. 182-
International Journal of Clinical and Health 201). Springfield, IL: Charles C. Thomas.
Psychology, 15 (2), 149-159. doi: https://doi. Wehmeyer, M. L. (2015). Framing the future:
org/10.1016/j.ijchp.2015.03.001 Self-determination. Remedial and Special
Verdugo, M. A., Vicente, E. M., Gómez, M., Fernán- Education, 36 (1), 20-23. doi: https://doi.
dez, R., Wehmeyer, M. L., Badía, M., … y Calvo, org/10.1177/0741932514551281
M. I. (2014). Escala ARC-INICO de evaluación Wehmeyer, M. L. y Abery, B. H. (2013). Self-deter-
de la autodeterminación: Manual de aplicación mination and choice. Intellectual and Develop-
y corrección. Salamanca: INICO. mental Disabilities, 51 (5), 399-411. doi: https://
Vicente, E., Guillén, V. M., Gómez, L., Ibáñez, A. doi.org/10.1352/1934-9556-51.5.399
y Sánchez, S. (2018). Elaborando una escala Wehmeyer, M. L., Field, S. y Thoma, C. A. (2012).
de autodeterminación a partir del consenso Self-determination and adolescent transition
entre expertos. Siglo Cero: Revista Española education. En M. L. Wehmeyer y K.W. Webb
sobre Discapacidad Intelectual, 1, 138-139. doi: (Coords.), Handbook of adolescent transition
https://orcid.org/0000-0001-5502-1771 https://orcid.org/0000-0003-2300-9537
año 77, nº 272, enero-abril 2019, 143-162
revista española de pedagogía
162
revista española de pedagogía
año 77, nº 272, enero-abril 2019
Spanish Journal of Pedagogy
year 77, n. 272, January-April 2019
Sumario*
Table of Contents**
Estudios
Studies
Francisco López Rupérez, Isabel García García, David Luque
Eva Expósito Casas Desarrollos interpretativos de la filosofía
Rendimiento en ciencias, concepciones epistémicas de la educación en la tradición anglófona:
y vocaciones STEM en las comunidades autónomas un intento de sistematización
españolas. Evidencias desde PISA 2015, políticas y Interpretive developments of the philosophy
prácticas de mejora of education in the anglophone tradition:
Performance in Sciences, epistemic conceptions and STEM an attempt to systematise them 67
vocations in the Spanish Autonomous Communities. Evidence
from PISA 2015, improvement policies and practices 5 Notas
Notes
Zaida Espinosa Zárate
El cultivo de la creatividad para el diálogo María Moralo, Manuel Montanero
intercultural Aprendizaje con y sin error en estudiantes con TEA
Cultivating creativity for intercultural dialogue 29 Learning with and without errors in students with ASD 85
* Todos los artículos están también publicados en inglés en la página web de la revista: https://revistadepedagogia.org.
** All the articles are also published in English on the web page of the journal: https://revistadepedagogia.org.
Oihane Fernández-Lasarte, Eider Goñi, Igor Camino, Informaciones
Estibaliz Ramos-Díaz El Profesor Giuseppe Mari. In memoriam (Emanuele
Apoyo social percibido e implicación escolar Balduzzi); III Conferencia Internacional de EuroSoTL:
del alumnado de educación secundaria “Explorando nuevos campos a través de un enfoque
Perceived social support and school engagement académico de la enseñanza y el aprendizaje”;
in secondary students 123 Una visita a la hemeroteca (Ana González-Benito);
Una visita a la red (David Reyero). 191
Jesús-Miguel Muñoz-Cantero, Luisa Losada-Puente
Validación del constructo de autodeterminación
a través de la escala ARC-INICO para adolescentes Instrucciones para los autores
Validation of the construct of self-determination through Instructions for authors 201
the ARC-INICO scale for teenagers 143
143 EV
Jesús-Miguel MUÑOZ-CANTERO and Luisa LOSADA-PUENTE
In addition, the factorial invariance meas- del desarrollo. Este estudio pretende evaluar
ure shows the utility of model 2 to compare sus propiedades psicométricas con población
scores according to the presence or absence adolescente gallega, comprobando su equi-
of neurological development disorders. In valencia tanto para su uso con jóvenes con
conclusion, the ARC-INICO scale provides Trastornos del Desarrollo Neurológico como
an important basis for decisions making re- sin ellos. Se emplea una muestra de 2 220
garding the design of care programs, through estudiantes. La estructura de la escala fue
the development of resources, guidelines and estudiada mediante Análisis Factorial Con-
strategies, and provides information for the firmatorio, usando la propuesta original con
differential provision of said resources and una estructura factorial de orden superior
supports. correlacionada con cuatro factores, y un
modelo unifactorial que asume la unidimen-
Keywords: self-determination, students, sionalidad de la autodeterminación. Res-
questionnaire, test reliability, test validity. pecto a la fiabilidad, presenta una elevada
consistencia interna global y en sus seccio-
Resumen: nes. Aunque el modelo unifactorial ofrece
La autodeterminación se sitúa como un un ajuste aceptable (Modelo 1: GFI = .958,
buen predictor de la calidad de vida, enten- AGFI = .941; RMSEA = .057), es superior
dida como una estrategia que pretende in- en el modelo de orden superior (Modelo 2:
crementar y mejorar las prácticas educativas GFI = .970, AGFI = .954; RMSEA = .049).
centradas en las necesidades de la persona Además, la medida de invarianza factorial
a nivel global, y a lo largo de su desarrollo muestra la utilidad del modelo 2 para com-
year 77, n. 272, January-April 2019, 143-162
vital. De ahí que la evaluación de este cons- parar puntuaciones según la presencia o no
tructo haya experimentado notables avances de Trastornos del Desarrollo Neurológico.
en nuestro país, fruto de la acogida de mo- En conclusión, la ARC-INICO ofrece una
delos teóricos internacionales y del diseño base importante para la toma de decisiones
revista española de pedagogía
144 EV
Validation of the construct of self-determination through the ARC-INICO scale for teenagers
needs present in the classroom and indi- Likert-type scale for the autonomy sec-
vidual needs to establish individual and tion, and a 4-point scale in the other
group action plans. sections. It is a questionnaire for that
the participants complete themselves,
revista española de pedagogía
research personnel. After collecting the other words, to be able to study the rela-
data, the questionnaires were reviewed tionships between latent variables or in-
and ones with five or more unanswered dicators (that are not directly observable),
items were eliminated (176 question- it is necessary to use observable indicators
naire eliminated). (the answers to questionnaire items). To
do this, a previously-bounded theoretical
2.4. Information analysis process structure must be taken as a basis. In the
To study the factorial structure of case of this study, it refers to the following
the ARC-INICO scale, item parcelling models:
was used with the aim of reducing the
breadth of the questionnaire and the idio- • Model 1. Unidimensional model: this
syncratic influence of the items measured assumes the unidimensionality of
individually (Bandalos, 2002; Bandalos & the self-determination construct; in
Finley, 2012). The parcelling of the items other words, for each section, there
was based on the recommendations in the is a single factor in which all of the
literature (Little, 2013; Little, Cunning- variables measured are saturated
ham, Shahar, & Widaman, 2002): (a) se- (parcels).
lection of conceptually similar items and
(b) ones that show a strong unidimen- • Model 2. Higher-order factorial mod-
sionality in exploratory factor analysis el: a structure with a higher level of
(EFA). abstraction is imposed, based on the
influence of a higher-order factor
same time in the higher-order factor (hi- The eigenvalues and percentage of var-
erarchical model) (Kelloway, 2014; Kline, iance explained were calculated, corrob-
2015). orating the hypothesis that each parcel
represents a unidimensional structure
Finally, the invariance of the scale with eigenvalues above one and with
was analysed to compare the results of over 50 % of the variance explained by
the two different groups (adolescents each one. In addition, all of the parcels
with neuro-development disorders and displayed adequate or high reliability,
ones without) to establish whether evaluated using Cronbach’s Alpha index
the measurements obtained from both (α ≤. 80).
groups are similar, and so whether its
use is valid for both groups (Elosua, Regarding univariate normality, the
2005). A multi-group analysis was per- measurements of distribution of skew-
formed using a progressive process ness and kurtosis are used, which make
(Byrne, 2008; Elosua, 2005): config- it possible to identify how the data sets
ural invariance (the pattern of factor group or separate around a central point.
loadings is the same), metric invari- The skew values z(G1) for the parcels
ance (the pattern of loadings and the are greater and lesser at ± 1.96, except
factorial weights are equal), and scalar in parcel P1_4, where z(G1) = .29 indi-
invariance (this also entails equal var- cates a symmetrical distribution. There-
iances for errors). To estimate the fit fore, the null hypothesis of a symmetri-
of the data, various indexes were used cal distribution is rejected in almost all
year 77, n. 272, January-April 2019, 143-162
that make it possible to select the model of these cases. Furthermore, calculating
from the two models considered that has the kurtosis indicates that the null hy-
the smaller discrepancy from the true pothesis that the distribution is mesokur-
model. These indexes are the Akaike in- tic (z[G2]> ± 1.96) is not fulfilled in
revista española de pedagogía
formation criterion (AIC) and Bentler’s most cases, except for parcels P2_1 (z
comparative fit index (CFI). [G2] = 1.587), P2_3 (z [G2] = −1067),
P3_2 (z [G2] = −1,442), P3_3 (z
These analyses were supported by the [G2] = 1,712) and P4_1 (z [G2] = 1,376).
IBM Statistical Package for the Social In the skew and kurtosis comparison set,
Sciences (IBM SPSS) version 23, and the the assumption of univariate normality
confirmatory factor analyses were done is not met as in all cases k2 > 5.98. Fi-
using the IBM SPSS AMOS 23.0 program. nally, the Kolmogorov-Smirnov goodness-
of-fit test, with Lilliefors’ modification,
indicates rejection of the null hypothesis
3. Results of normality of the parcels as, for a 95 %
3.1. Preliminary analysis confidence level, all of the values obtained
Table 1 shows the final matrix com- are p < .005. Therefore, the hypothesis
prising 13 parcels that represent the that the data are from a normal univari-
four sections of the ARC-INICO scale. ate distribution was rejected.
148 EV
Table 1. Analysis of the unidimensionality and univariate normality comparisons of the ARC-INICO scale.
% Skew Kurtosis K-S test
vari-
Eigen- α
ance
Sections Parcels values stand-
ex-
a ardised z ET z (G1) z ET z (G2) K2 K-Sª p
plained
b
P1_1 4.95 82.442 .957 .28 .052 5.46 −.43 .104 −4.16 47.16 .337 .00
P1_2 4.45 55.637 .884 −.79 .052 −15.23 −.40 .104 −3.87 246.92 .179 .00
Autonomy
P1_3 4.27 71.179 .917 −.18 .052 −3.40 −.87 .104 −8.35 81.24 .188 .00
P1_4 3.72 74.476 .913 .02 .052 0.29 .30 .104 2.86 8.24 .244 .00
P2_1 4.81 96.146 .990 −.23 .052 −4.48 .17 .104 1.59 22.60 .319 .00
Self-
P2_2 3.84 96.006 .986 −.39 .052 −7.52 .22 .104 2.09 60.89 .288 .00
regulation
P2_3 2.45 59.861 .857 −.32 .052 −6.15 −.11 .104 −1.07 39.01 .294 .00
P3_1 3.87 77.439 .911 −.48 .052 −9.31 .26 .104 2.53 93.04 .147 .00
Empowerment P3_2 3.64 90.924 .967 −.27 .052 −5.21 −.15 .104 −1.44 29.24 .253 .00
P3_3 2.88 98.648 .978 −.44 .052 −8.46 .18 .104 1.71 74.54 .311 .00
P4_1 3.79 94.781 .972 −.25 .052 −4.87 .14 .104 1.38 25.04 .307 .00
Self-
P4_2 3.45 86.288 .947 −.98 .052 −18.92 .71 .104 6.83 404.69 .204 .00
knowledge
P4_3 2.95 73.803 .881 −.71 .052 −13.67 .52 .104 5.01 212.05 .174 .00
Note:
a. Eigenvalues: these reflect, in decreasing order, the degree of variation in the set of variables, which explains the factor, or in other words,
the amount of information each variable provides to the factor. The sum of all of the eigenvalues being equal to the number of variables
introduced in the analysis.
b. % Variance explained: the percentage of variance is represented by the set of items included.
Initials and Statistics: α − Cronbach’s Alpha index; K-S Comparison: Kolmogorov-Smirnov test comparison to calculate the normality of the
distribution; z − Statistical value referring to the standard deviation obtained for one variable, compared with the value of the mean; SE:
Validation of the construct of self-determination through the ARC-INICO scale for teenagers
149 EV
revista española de pedagogía
year 77, n. 272, January-April 2019, 143-162
150 EV
Table 2. Correlation between parcels on the ARC-INICO scale.
P1_1 P1_2 P1_3 P1_4 P2_1 P2_2 P2_3 P3_1 P3_2 P3_3 P4_1 P4_2 P4_3
P1_1 1
P1_2 .399** 1
P3_3 .303** .466** .385** .205** .300** .376** .358** .349** .359** 1
P4_1 .202** .393** .318** .194** .352** .387** .311** .387** .315** .320** 1
Jesús-Miguel MUÑOZ-CANTERO and Luisa LOSADA-PUENTE
P4_2 .137** .157** .110** .129** .228** .233** .227** .274** .186** .098** .224** 1
P4_3 −.012 .074** .054* .081** .153** .144** .067** .224** .141** .096** .241** .188** 1
Note: * p < .05 ** p < .001. This value refers to the probability associated with the statistic, which gives information about acceptance of the
hypothesis that there are statistically significant correlations, with an associated confidence level of 95 %.
Source: Own elaboration.
Validation of the construct of self-determination through the ARC-INICO scale for teenagers
To test multivariate normality, the evaluate and test goodness of fit (Kello-
normality and outliers test in IBM SPSS way, 2014; Kline, 2015): firstly, the chi-
AMOS 23 was performed. The value of squared statistic (χ2) and its relative
the multivariate kurtosis (g2 = 11.91; version (χ2 / df) to evaluate the general
cr = 14.20) indicates that the variables fit, where a non-significant χ2 and val-
display a kurtosis which is significantly ues of χ2 / df < 2 indicate a good fit; and
different from a normal multivariate anal- secondly, given these indexes’ sensitivi-
ysis. The analysis of multivariate outliers ty to variations in sample size, addition
using the Mahalanobis distance figure (D2) indexes were used to evaluate the abso-
indicates the presence of 15 extreme val- lute partial fit: root mean square resid-
ues (p < .001). It was decided not to ex- ual (RMR ≤. 08 shows an adequate fit);
clude or convert them, as they are errors goodness of fit index (GFI) and adjust-
that reflect the idiosyncrasy of the stu- ed goodness of fit index (AGFI), which
dents sampled. should have a value of .90 or more. Fur-
thermore, the parsimony of fit was eval-
Mulicollinearity was verified by calcu- uated using the root mean square error
lating the correlation matrix, as shown in of approximation (RMSEA ≤.08 shows
Table 2. The remaining values show the an adequate fit) and non-nested models,
absence of multicollinearity in the data, which indicate that the model has a lower
with values lower than .90; the highest cor- discrepancy with the true model (Akaike
relation coefficient found was rxy = .593 information criterion, AIC and consistent
between parcels P1_2 and P1_3. Akaike information criterion, CAIC) and
Graph 1 shows the initial standardised observed variables that can be explained
solution for model 1. All of the coefficients by the latent factor (self-determination)
are significant (t values significantly dif- approximates to the appropriate value, so
ferent from 0), with prediction errors that long as the variables that are far from the
vary from .01 to .60, so that the squared optimal values are eliminated. The factor
correlation coefficients (r2) vary from .99 loadings range between .07 and .78, six of
to .40. Almost all of the figures for r2 are them being lower than .6. This could in-
greater than .5, with the exception of dicate that the single-factor model is not
P1_2 and P1_3. These results indicate sufficient to reproduce the original covar-
that the proportion of the variance in the iance matrix.
When analysing the model’s fit with the 3.2.2. Model 2: Higher-Order Factorial
empirical data, the figures for which are Model
shown in Table 3, deficient indexes of fit This model derives from a proposal
can be seen (χ2 528.929; χ2 / df = 8.145; p for self-determination as a factorial and
<.000), and a partial fit that can be improved hierarchical construct in which, within
in the evaluation of the model using the a general self-determination factor (sec-
RMR and RMSEA indexes (RMR = .648; ond-order factor), there are four factors
RMSEA = .057). There is also a specification (first-order factors) grouped by the par-
error (p = .007). In contrast with this, some cels corresponding to autonomy, self-reg-
absolute partial fit indexes have satisfactory ulation, empowerment, and self-knowl-
results (GFI = .958; AGFI = .941). edge.
152 EV
Table 3. Goodness of fit statistics.
Partial fit
Absolute fit
.057 .262
Model 1 528.929 65 .000 8.137 .648 .958 .941 .007 580.93 755.27
[.052 − .061] [.230 − .297]
.049 .199
Model 2 420.1 60 .000 6.334 .553 .970 .954 .623 442.03 649.89
[.044 − .054] [.173 − .299]
Note: χ2: chi-square; df: degrees of freedom; χ2 / df: relative χ2; RMR: root mean square residual; GFI: goodness of fit index; AGFI: adjusted
goodness of fit index; RMSEA: root mean square error of approximation; AIC: Akaike information criterion; CAIC: consistent Akaike informa-
tion criterion; ECVI: expected cross validation index.
Source: Own elaboration based on IBM SPSS AMOS 23.
Validation of the construct of self-determination through the ARC-INICO scale for teenagers
153 EV
Jesús-Miguel MUÑOZ-CANTERO and Luisa LOSADA-PUENTE
From the empirical perspective, this variables in the observed variables dis-
model establishes independence relat- play fairly high values (range: .80−.42),
ing to the four sections pf ARC-INICO, with loadings greater than 0.4 (except for
which are grouped into a single general P4_2 = .29 and P4_3 = .16). The same cir-
second-order factor. Graph 2 shows the cumstance is seen in the factor loadings of
standardised solution, which shows pre- the endogenous variables compared with
diction errors varying from .03−.64 with the exogenous variables (range .84−.94),
coefficients of determination of .36 to .97. although their prediction errors are also
The factor loadings for the endogenous high (.70 ≤ e ≤ .89).
Graph 2. Standardised parameters of Model 2 (higher-order factorial model).
year 77, n. 272, January-April 2019, 143-162
The empirical results, shown in Ta- the existence of first-order factors that
revista española de pedagogía
ble 3, indicate that this model has a bet- correspond with the four sections defined
ter fit than the previous one, with higher rationally in the test.
values (GFI = .970; AGFI = .954; RM-
SEA = .049, and RMR = .553) with a 3.3. Comparison of models
PCLOSE value = .623 indicating a good Finally, regarding the parsimonious in-
fit of the data for a 90 % confidence lev- dexes of fit that compare the non-nested
el. In addition, if we take into account models, Akaike’s AIC and its consistent
the magnitude of χ2 / df (420.1 / 60), version (CAIC) are interpreted in such a
it should be noted that this model has a way that their lowest value fits the spec-
better fit than model 1, given its smaller ified model better (West, Taylor, & Wu,
size. As was expected, in accordance with 2015). Likewise, the interpretation of the
the theoretical backing of the multidi- expected cross validation index (ECVI) is
mensional models of self-determination, based on the comparison between models,
it can be said that this model has a better assuming that the model with the lower
fit with the data, as it takes into account value is the one with the greatest poten-
154 EV
Validation of the construct of self-determination through the ARC-INICO scale for teenagers
tial for replication (Browne & Cudeck, disorders. The indexes of fit obtained are
1993). Consequently, the comparison be- shown in Table 4. These make it possible
tween the models based on these indexes to accept the equivalence of the basic meas-
indicated a better fit for model 2, as shown urement models between the two samples.
in Table 3. Although the chi-squared value exceeds that
required to accept the invariance hypothesis,
3.4. Factorial Invariance Analysis the other indexes contradict this conclusion
Based on Abalo, Lévy, Rial, and Varela (GFI = .966, AGFI = .949, RMSEA = .036;
(2006), the same model was estimated (Mod- AIC = 594.9; CFI = .718) which enables us
el 2) for two samples of students depending to accept the base invariance model (unre-
on whether they have neuro-development stricted model).
Without restrictions 470.906* 120 .966 .949 .663 .718 .036 594.9
Metric invariance 587.051* 129 .958 .940 .580 .632 .040 693.1
Scalar invariance 587.669 132 .958 .942 .580 .633 .039 687.7
Note: *p < .05.
Source: Own elaboration based on IBM SPSS AMOS 23.
In both cases, confirmatory factor tracted from the analysis. This figure
analysis (CFA) results display a good fit is excellent for the model in general
for the higher-order factorial model. The (CR = .924), and adequate in the sec-
subgroup with the better fit is the one ond-order factor (CR = .799), and in the
comprising students who do not have neu- first-order factors (.696 ≤ CR ≥ .808).
ro-developmental disorders. In any case, Secondly, the general saturation of the
the generally uniform fit between the scale is studied with the aim of test-
groups suggested that the four-factor solu- ing convergent validity using McDon-
tion is appropriate for these subgroups ald’s ω, which gives excellent results
from the sample. for the total scale (ω = .922), and ad-
equate results for the second-order fac-
3.5. Reliability and correlation tor (ω = 552), and for each first-order
between factors factor (ω = .821−.700). Finally, the cor-
Finally, the reliability and validity relation coefficients between the latent
of the final model and the correlations variables in Model 2 were calculated,
between the latent variables were test- which indicate the extent to which the
ed. Firstly, composite reliability (CR) model’s variables vary jointly, and the
was calculated, which indicates the extent to which they vary with regards
consistency of all of the constructs ex- to the model (Table 5).
and self-determination.
Self-
Self- Empower- Self-
determina- Autonomy
regulation ment knowledge
tion
Self-determination 1
revista española de pedagogía
Autonomy .846 1
The results show reasonably satis- mination construct has a very good rela-
factory values given that the correlation tionship with the variables that define it,
coefficients between first-order factors reaching 84.6 % (r = .846) with regards
(r = .309 − .593), indicate a relation- to autonomy, 82 % (r = .820) regarding
ship of 30.9 % to 59.3 % between autono- self-regulation, 79.8 % (r = .798) regard-
my, self-regulation, empowerment, and ing empowerment, and 63.9 % (r = .639)
self-knowledge. Similarly, the self-deter- regarding self-knowledge.
156 EV
Validation of the construct of self-determination through the ARC-INICO scale for teenagers
This evidence suggests that the for designing actions intended to lead
ARC-INICO scale is a valid and reliable to educational, personal, and social im-
scale for studying self-determination in provement.
adolescence, with Wehmeyer’s function-
al model (1999, 2003) being an adequate We conclude by emphasising the im-
representation of this construct. In this portance of continuing with this line of
respect, it is possible to refer to other research in order to corroborate or refute
studies with similar results. For example, the data obtained in this study, also tak-
Verdugo et al. (2014) and Vicente et al. ing into account the limitations stated in
(2015) carried out studies of the psycho- it, such as:
metric properties of this scale, through
exploratory and confirmatory factor anal- a) The threat to the possibility of gen-
ysis, which support a structure compris- eralising these results given that all
ing four independent factors. Similarly, of the participants are Galician stu-
these studies support the results found dents. This could be overcome by
in relation to the values of the correla- expanding the sample to include stu-
tion coefficients between first-order fac- dents from other autonomous regions
tors and with regards to the second-order of Spain.
factor. b) The use of a self-report measure, which
derives from the instrument being
This study has several implications evaluated itself, and which could con-
for educational practices. Education has tain biases deriving from social desira-
year 77, n. 272, January-April 2019, 143-162
and decide how to live their own lives idence including the foundations for
(Arellano & Peralta, 2013; Lee et al., constructing a new theoretical model
2012; Wehmeyer et al., 2012; Wehmeyer and for developing evaluation and pro-
& Shogren, 2018; etc.). These supports motion tools.
refer to aspects such as «the provision
of professional interventions, the cre- One example of this is causal agency
ation of settings and the presentation theory (Shogren, Wehmeyer, Palmer, &
of individualised support strategies» Forber-Pratt, 2015; Shogren et al., 2016;
(Schalock, 2018, p. 12), which require Shogren, Wehmeyer, & Burke, 2017),
prior evaluation that makes it possible which proposes an extension of the func-
to determine the need for support and tional model, focussing on the theoretical
a final evaluation of the impact of the and practical reformulation of the original
interventions on personal results. Con- model and giving particular importance
sequently, it is important to develop and to the individual’s capacity for action, de-
validate sold diagnostic tools as a basis fined as being able to «act as the primary
158 EV
Validation of the construct of self-determination through the ARC-INICO scale for teenagers
Kline, R. B. (2015). Principles and practice of struc- M. L. Wehmeyer, & N. N. Singh (Eds.), Hand-
tural equation modeling. New York: The Guild- book of Positive Psychology in Intellectual and
ford Press. Developmental Disabilities (pp. 49-64). New
Lee, Y., Wehmeyer, M., Palmer, S., William-Diehm, York: Springer.
K., Davies, D., & Stock, S. (2012). Examining Shogren, K. A., Wehmeyer, M. L., Palmer, S. B.,
individual and instruction-related predictors of Forber-Pratt, A. J., Little, T. J., & Lopez, S.
the self-determination of students with disabil- (2015). Causal agency theory: Reconceptu-
ities: multiple regression analysis. Remedial & alizing a functional model of self-determi-
Special Education, 33 (3), 150-161. doi: https:// nation. Education and Training in Autism
doi.org/10.1177/0741932510392053 and Developmental Disabilities, 50 (3), 251-
Little, T. D. (2013). Longitudinal structural equa- 263. Retrieved from http://www.jstor.org/
tion modeling. New York: Guilford Press. stable/24827508 (Consulted on 16/11/2017).
Little, T. D., Cunningham, W. A., Shahar, G., & Shogren, K. A., Wehmeyer, M. L., Palmer, S. B.,
Widaman, K. F. (2002). To parcel or not to & Paek, Y. (2013). Exploring personal and
parcel: Exploring the question, weighing school environment characteristics that pre-
the merits. Structural Equation Modeling, dict self-determination. Exceptionality: a Spe-
9, 151-173. doi: http://dx.doi.org/10.1207/ cial Education Journal, 21 (3), 147-157. doi:
S15328007SEM0902_1 https://doi.org/10.1080/09362835.2013.802231
Murumbardó, C., Guàrdia, J., & Giné, C. (2018). Shogren, K. A., Wehmeyer, M. L., Schalock, R. L., &
Autodeterminación: midiendo el constructo en
Thompson, J. R. (2017). Reframing educational
jóvenes con y sin discapacidad. Siglo Cero: Re-
supports for students with intellectual disabil-
vista Española sobre Discapacidad Intelectual,
ity through strengths-based approaches. In M.
1, 137-138. doi: http://dx.doi.org/10.14201/
L. Wehmeyer, & K. A. Shogren (Eds.), Hand-
scero20180
book of research-based practices for educating
Peralta, F., & Arellano, A. (2014). La autodetermi-
students with intellectual disability (pp. 17-31).
year 77, n. 272, January-April 2019, 143-162
validez. Papeles del Psicólogo, 31 (1), 67-74. adulta de alumnos con discapacidad intelec-
Retrieved from www.papelesdelpsicologo.es/ tual. Revista Iberoamericana de Educación, 63,
resumen?pii=1797 (Consulted on 16/02/2017). 19-33. Retrieved from https://rieoei.org/RIE/
Schalock, R. L. (2018). Seis ideas que están cam- article/view/418 (Consulted on 22/06/2017).
biando el campo de las discapacidades inte- Verdugo, M. A., Vicente, E., Fernández-Pulido,
lectuales y del desarrollo en todo el mundo. R., Gómez-Vela, M., Wehmeyer, M. L., & Gui-
Siglo Cero: Revista Española sobre Discapaci- llén, V. M. (2015). Evaluación psicométrica de
dad Intelectual, 49 (1), 7-19. doi: http://dx.doi. la escala ARC-INICO de autodeterminación
org/10.14201/scero2018491719 para adolescentes con discapacidad intelectual.
Shogren, K. A., & Wehmeyer, M. L. (2016). International Journal of Clinical and Health
Self-determination and goal attainment. In M. Psychology, 15 (2), 149-159. doi: https://doi.
L. Wehmeyer, & K. A. Shogren (Eds.), Hand- org/10.1016/j.ijchp.2015.03.001
book of research-based practices for educating Verdugo, M. A., Vicente, E. M., Gómez, M.,
students with intellectual disability (pp. 255- Fernández, R., Wehmeyer, M. L., Badía, M., …
273). New York: Routledge. & Calvo, M. I. (2014). Escala ARC-INICO de
Shogren, K. A., Wehmeyer, M. L., & Burke, K. M. evaluación de la autodeterminación: Manual
(2017). Self-determination. In K. A. Shogren, de aplicación y corrección. Salamanca: INICO.
160 EV
Validation of the construct of self-determination through the ARC-INICO scale for teenagers
Vicente, E., Guillén, V. M., Gómez, L., Ibáñez, A., velopmental Disabilities, 51 (5), 399-411. doi:
& Sánchez, S. (2018). Elaborando una escala https://doi.org/10.1352/1934-9556-51.5.399
de autodeterminación a partir del consenso Wehmeyer, M. L., Field, S., & Thoma, C. A. (2012).
entre expertos. Siglo Cero: Revista Española Self-determination and adolescent transition
sobre Discapacidad Intelectual, 1, 138-139. doi: education. In M. L. Wehmeyer, & K. W. Webb
http://dx.doi.org/10.14201/scero20180 (Coords.), Handbook of adolescent transition
Vicente, E., Verdugo, M. A., Gómez-Vela, M., education for youth with disabilities (pp. 171-
Fernández-Pulido, R., & Guillén, V. (2015). Pro- 190). London: Routledge.
piedades psicométricas de la escala ARC-INI- Wehmeyer, M. L., & Kelchner, K. (1995). The Arc’s
CO para evaluar la autodeterminación. Revis- self-determination scale. Arlington: The Arc
ta Española de Orientación y Psicopedagogía, National of United States.
26 (1), 8-24. doi: https://doi.org/10.5944/reop. Wehmeyer, M. L., & Shogren, K. A. (2018). Self-de-
vol.26.num.1.2015.14339 termination and positive psychological aspects
Vicente, E., Verdugo, M. A., Gómez-Vela, M., of social psychology. In D. S. Dunn (Ed.), Posi-
Fernández-Pulido, R., & Guillén, V. (2017). tive psychology: Established and emerging is-
Personal characteristics and school contextual sues. New York: Rutledge.
variables associated with student self-determi- West, S. G., Taylor, A. B., & Wu, W. (2015). Model
nation in Spanish context. Journal of Intellec- fit and model selection in structural equation
tual & Developmental Disability, 42 (3), 1-12 modeling. In R. H. Hoyle (Ed.), Handbook of
doi: https://doi.org/10.3109/13668250.2017.13 structural equation modeling (pp. 209-231).
10828 New York: The Guilford Press.
Walker, H. M., Calkins, C., Wehmeyer, M. L.,
Walker, L., Bacon, A., Palmer, S. B., … & Jon-
hson, D. R. (2011). A social-ecological approach Authors’ biographies
to promote self-determination. Exceptionality: Jesús Miguel Muñoz Cantero is a
A Special Education Journal, 19 (1), 6-18. doi: PhD from the Universidad de Santiago de
https://orcid.org/0000-0003-2300-9537
year 77, n. 272, January-April 2019, 143-162
revista española de pedagogía
162 EV
revista española de pedagogía
año 77, nº 272, enero-abril 2019
Spanish Journal of Pedagogy
year 77, n. 272, January-April 2019
Table of Contents
Sumario
Studies Notes
Estudios Notas
Francisco López Rupérez, Isabel García García, María Moralo, Manuel Montanero
Eva Expósito Casas Learning with and without errors in students with ASD
Performance in science, epistemic conceptions, and STEM Aprendizaje con y sin error en estudiantes con TEA 85
vocations in Spain’s autonomous communities: evidence
Antonio Portela Pruaño, José Miguel Nieto Cano, Ana
from PISA 2015, improvement policies, and practices
Torres Soto
Rendimiento en Ciencias, concepciones epistémicas y
Re-engagement in education and training of young
vocaciones STEM en las comunidades autónomas españolas.
people who leave education early: the importance
Evidencias desde PISA 2015, políticas y prácticas de mejora 5
of earlier prior trajectories
Zaida Espinosa Zárate La reincorporación formativa de jóvenes
Cultivating creativity for intercultural dialogue que abandonan tempranamente la educación:
El cultivo de la creatividad para el diálogo intercultural 29 relevancia de su trayectoria previa 103
This is the English version of the research articles and book reviews published orig-
inally in the Spanish printed version of issue 272 of the revista española de pe-
dagogía. The full Spanish version of this issue can also be found on the journal's
website http://revistadepedagogia.org.