Content uploaded by Carlos De las Cuevas
Author content
All content in this area was uploaded by Carlos De las Cuevas
Content may be subject to copyright.
Int J Clin Health Psychol, Vol. 12. Nº 2
LÓPEZ-NÚÑEZ et al. Rutgers Alcohol Problem Index 265
ISSN 1697-2600 print
ISSN 2174-0852 online
2012, Vol. 12, Nº 2, pp. 265-279
© International Journal of Clinical and Health Psychology
Psychometric properties of the parent and teacher
versions of the Strength and Difficulties
Questionnaire (SDQ) in a Spanish sample1
Pedro Javier Rodríguez-Hernández2 (Hospital Universitario N.S. Candelaria, Spain),
Moisés Betancort (Universidad de La Laguna, Spain),
Gustavo Mario Ramírez-Santana (Universidad de La Laguna, Spain),
Ruth García (Institute of Psychiatry, United Kingdom),
Emilio J. Sanz-Álvarez (Universidad de La Laguna, Spain), and
Carlos De las Cuevas-Castresana (Universidad de La Laguna, Spain)
ABSTRACT. The Strengths and Difficulties Questionnaire (SDQ) is a well-known,
widely used and well-established brief screening instrument for detecting psychological
morbidity in childhood. However, no research has been conducted with Spanish samples
to provide data about the validity and reliability of this instrument, or the factorial
structure of the questionnaire. The current instrumental study aims to provide the
psychometric properties of the Spanish version of the SDQ. A sample of 595 children
was evaluated using both teachers and parents as informants. Descriptive statistics for
the SDQ are presented. Cronbach´s alpha coefficient, was calculated for the full global
questionnaire and each of its scales, to examine the internal reliability. Factor analysis
with Varimax Rotation was conducted following the original author´s recommendations.
The factor solution was analyzed using Confirmatory Factor Analysis. Data showed
a satisfactory index of reliability as well as a factorial structure of five dimensions with
some exceptions for some items. We conclude that our study replicates the data found
in other studies in which the goodness of the SDQ as a screening instrument has been
studied.
1The authors would like to thank Professor Robert Goodman (London), for his support and
assistance during the study.
2Correspondece: Child and Adolescent Psychiatric Unit «Diego Matías Guigou y Costa». University
Hospital N.S.Candelaria. Canary Islands (Spain).Email: pedrojavierrodriguezhernandez@yahoo.es
Int J Clin Health Psychol, Vol. 12. Nº 2
266 RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire
KEYWORDS: Strengths and Difficulties Questionnaire. Psychometric properties. Mental
health. Factor analysis. Instrumental study.
RESUMEN. El Strengths and Difficulties Questionnaire (SDQ) es un instrumento de
screening ampliamente utilizado para detectar morbilidad psicológica en la infancia. Sin
embargo, ningún estudio llevado a cabo en muestras españolas ha aportado datos sobre
su validez y fiabilidad, así como sobre la estructura factorial del cuestionario. El
presente estudio instrumental tiene como objetivo informar de los aspectos psicométricos
de la versión española del SDQ. En este trabajo fueron evaluados 595 niños utilizando
a los profesores y los padres como informantes. Se realizaron análisis descriptivos del
SDQ. Se calcularon índices de consistencia interna mediante el Alfa de Cronbach, tanto
para el cuestionario global como para las distintas escalas del mismo. Igualmente se
llevó a cabo un análisis factorial mediante Rotación Varimax siguiendo las recomenda-
ciones de los autores originales del cuestionario. La solución factorial alcanzada fue
sometida a un análisis factorial confirmatorio. Los resultados encontrados muestran una
consistencia interna elevada y una estructura factorial de cinco dimensiones con excep-
ciones puntuales en algunos ítems. Concluimos que el estudio replica los datos encon-
trados con otras muestras en las que se ha examinado la bondad del SDQ como
instrumento de screening.
PALABRAS CLAVE: Cuestionario de capacidades y dificultades. Propiedades
psicométricas. Salud mental. Análisis factorial. Estudio instrumental.
In recent years, a growing number of publications have raised awareness of the
high prevalence of emotional and behavioral disorder in children and adolescents
(Belfer, 2008; Griffin, Guerin, Sharry, and Drumm, 2010; Mansbach-Kleinfeld et al., 2010;
Wille, Bettge, Wittchen, and Ravens-Sieberer, 2008). There is an international consensus
that 1 in 5 children will suffer some degree of mental health problems during their
childhood or adolescence. This estimation has been carried out over different countries
and cultures (Malhotra, Kohli, Kapoor, and Pradhan, 2009; Merikangas, Nakamura, and
Kessler, 2009; Oschilewsky, Gómez, and Belfort, 2010). However, the percentages of
those who make use of mental health services are still relatively low. One of the causes
of this low rate is the lack of instruments that can be easily applied to a wide sample
in order to detect potential emotionally unhealthy children. This is a significant issue,
as preventive policies to detect mental health problems in children could help to avoid
possible psychological and psychiatric problems in adolescence (Belfer, 2008; Goodman,
Ford, Simmons, Gatward, and Meltzer, 2000; Shivram et al., 2009).
Mental health problems in childhood and adolescence can cause academic and
social impairment in the individual and place an emotional burden on families. Therefore,
it is necessary to design convincing instruments that permit fast detection of children
and adolescents at potential risk of developing psychological disorders. This in turn will
allow early interventions to reduce the negative impact (Crone, Vogels, Hoekstra, Treffers,
and Reijneveld, 2008; Estévez and Emler, 2011).
Int J Clin Health Psychol, Vol. 12. Nº 2
RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire 267
The Strengths and Difficulties Questionnaire (SDQ) is a screening instrument broadly
used to detect mental health problems from childhood through adolescence
(Giannakopoulos et al., 2009; Goodman, Renfrew, and Mullick, 2000; Koskelainen, Sourander,
and Kaljonen, 2000; Mullick and Goodman, 2001; Van Widenfelt, Goedhart, Treffers, and
Goodman, 2003). The original English version was created by Robert Goodman at the
Institute of Psychiatry in London (Goodman, 1997). The main use of the SDQ resides
in the possibility of detecting children at potential risk and who could need further
assessment or even treatment.
The SDQ is a brief questionnaire with 25 items divided into 5 scales: Emotional
symptoms, Behavioural problems, Hyperactivity, Peer relationship problems and
Prosocial behaviors. The total difficulty score is generated by adding the scores from
all the scales except the prosocial one. There are three versions of the Questionnaire:
parent, teacher and a self-report version. The SDQ presents several advantages when
compared with similar screening tools; it is brief, easy to fill out, which increases
respondent rates, and it covers the more important psychopathologies in children and
adolescents. It can also evaluate different aspects of behavior using the same instrument.
Additional advantages include the existence of a self report version and the fact that
the same version can be used with different informants, which increases its power of
detecting psychological pathologies. Finally, there is a significant correlation between
the SDQ and other psychopathology scales such us the Rutter and Achenbach
questionnaires (Crone et al., 2008; Goodman, 1997; Goodman and Scott, 1999; Klasen
et al., 2000; Vogels, Crone, Hoekstra, and Reijneveld, 2009).
The SDQ in its different language versions has been widely used in a variety of
studies (Di Riso et al., 2010; Du, Kou, and Coghill, 2008; Giannakopoulos et al., 2009;
Koskelainen et al., 2000; Matsuishi et al., 2008; Mullick and Goodman, 2001; Van
Widenfelt et al., 2003; Woerner, Becker, and Rothenberger, 2004). For instance, it has
been used as a screening tool in prevalence studies (Mullick and Goodman, 2001;
Thabet, Stretch, and Vostanis, 2000; Van Widenfelt et al., 2003; Vogels et al., 2009) and
double face studies (Goodman, Ford et al., 2000; Goodman, Renfrew et al., 2000; Mathai,
Anderson, and Bourne, 2004) aiming to compare or specify the more important psychometric
properties. Recently its use has become widespread as one of the tools used to
dimensionally measure child mental health (Goodman and Goodman, 2009). The sound
data on the psychometric properties of the parent and teacher versions of the SDQ
provides strong support for its use as a screening instrument for research and
epidemiological studies. It has also proved useful for clinical purposes (Becker, Woerner,
Hasselhorn, Banaschewski, and Rothenberger, 2004; Glazebrook, Hollis, Heussler, Goodman,
and Coates, 2003; Mathai, Anderson, and Bourne, 2003; Vostanis, 2006).
The SDQ factor structure and psychometric properties have been widely replicated
in samples from United Kingdom, Europe, Asia, USA and Australia (Du et al., 2008;
Giannakopoulos et al., 2009; Goodman, 2001; Hawes and Dadds, 2004; Matsuishi et al.,
2008; Palmieri and Smith, 2007; Smedje, Broman, Hetta, and Von Knorring, 1999; Woerner
et al., 2004). In the original English version, factor analysis using Varimax rotation with
a British sample yielded 5 factors in both the parent and teacher versions. The psychometric
data indicate a mean internal consistency of .73 as measured with Chronbach’s alpha
Int J Clin Health Psychol, Vol. 12. Nº 2
268 RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire
(Goodman, 2001). Versions in other languages have yielded similar results. In the Swedish
version for parents, the factorial structure using Varimax Rotation yielded the same 5
factors. Some items loaded on another scale. For example, one of the items of the peer
problems scale loads clearly on the behavioural problems scale and one of the items of
the behavioural problems scale loads negatively on the prosocial subscale. In terms of
internal consistency (Cronbach´s alpha), the alpha coefficients show a range between
.51 (peer problems scale) and .76 (total difficulties scale; Smedje et al., 1999). In the
Greek self-report version, Confirmatory Factor Analysis with maximum likelihood procedure
was used to assess the theoretical model for the SDQ. The proposed five-factor structure
was confirmed, although some modifications seem to be necessary in order to increase
the fit. The internal consistency reliability shows Cronbach á values between .50 (peer
problems scale) and .77 (total difficulties score; Giannakopoulos et al., 2009). In the
Chinese versions of the SDQ for parents and teachers, the same five-factor structure
was confirmed using Varimax rotation. In the version for parents, some items loaded
onto other scales. Two of the peer problem items loaded onto the prosocial component
and two of the hyperactivity items loaded onto the behavioural problems scale. For the
teacher version, 8 items loaded onto other scales. In terms of internal consistency,
Cronbach´s alpha coefficients for the parent SDQ scales are between .30 (peer problems)
and .76 (hyperactivity) and for the teacher version, between .48 (peer problems) and .83
(prosocial behaviour).
In conclusion, most studies have shown a strong five-factor structure using teachers,
parents, and self-report informants. Various studies have found that some scales pro-
duce lower ranges of internal reliability mainly due to the few items in each scale
(Giannakopoulos et al., 2009; Hill and Hughes, 2007; Palmieri and Smith, 2007). In Spain,
the SDQ has been broadly used as a clinical instrument. This extensive clinical use
contrasts with the absence of rigorous studies about the internal reliability and factorial
structure of the Spanish version of the SDQ. The aim of this research is to increase
international evidence about the efficacy of the SDQ as an instrument that measures
potential psychopathologies in epidemiological studies. For this propose we followed
the recommendations of Carretero-Dios and Perez (2007). In this study, data was collected
from teachers and parents and from a wide sample of children from the Canary Islands.
Method
Participants
The sample was obtained from all the children aged 7 to 10 who are residents in
the Canary Islands (Spain). The sample was stratified by age using the latest census
on children attending school in the Canary Islands (official data recorded by the Canary
Island Government).
The sample size was determined according to the following statistical conditions;
an expected prevalence of mental health and behavioural problems of 10%, with a
maximum error of 3% and a confidence interval of 95%. To facilitate selection of the
participants, the sample collection was carried out in two stages: First, schools and
classes were randomized. Secondly, we made a randomized selection of the number of
Int J Clin Health Psychol, Vol. 12. Nº 2
RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire 269
students needed for sample size. In addition to this, we included additional participants
in anticipation of possible drop-outs. Gender data was equal during the data collection,
given that all schools are of mixed gender and have similar proportions of boys and
girls. Under all these conditions, the minimum sample required was 392. The actual
selection included 595, incorporating the additional participants. Table 1 shows effect
of gender in each scale.
TABLE 1. Mean and Standard Deviation for subscales and
total scale for male and female.
Teachers Parents
Male Female Sig Male Female Sig
M SD M SD M SD M SD
Total Scale 11.55 7.02 8.62 5.87 ** 12.2 6.3 10.5 6.6 **
Prosocial Scale 7.10 2.31 8.12 1.95 ** 7.9 1.8 8.6 1.6 n.s.
Hyperactivity Scale 4.65 3.08 2.95 2.57 ** 5.0 2.6 3.9 2.7 **
Emotional
Symptoms Scale 2.16 2.18 1.91 1.91 n.s. 2.6 2.3 2.6 2.3 n.s.
Behavioural
Problems Scale 2.35 2.31 1.41 1.79 ** 2.1 1.9 1.9 1.8 **
Peer Problems Scale 1.95 1.94 1.62 1.68 ** 2.5 1.9 2.1 1.7 n.s.
Note. M: Mean, SD: Standard Deviation, Sig: significative, **: p <.01, n.s.: non significant. Authorizations
from the Canary Islands Education Minister was sought.
Instruments
The SDQ is a questionnaire with 25 items divided into 5 scales. The first 4 scales
refer to emotional symptoms, behavioural problems, hyperactivity, and peer relationship
problems. The fifth scale refers to prosocial behaviors. Each of the 25 items is rated as
being Not true (0), Somewhat true (1), or Certainly true (2), and each of the SDQ scales
consists of five items, thus yielding scores between 0 and 10. Four of the SDQ scales
represent problem scores which are added to obtain a total difficulties score. The
questionnaire can be downloaded from internet for free in many different languages. The
official Spanish translations of the parent and teacher versions of the SDQ were used
(http://www.sdqinfo.com). The parent and teacher versions were translated into Spanish
by a bilingual child psychiatrist. After a first translation, the questionnaires were
administered to a group of parents and teachers in Spain in order to combat inherent
weaknesses of cross-cultural adaptation (for example, semantic and scale equivalence)
and with the aim of detecting and clarifying any possible misunderstandings or any
unclear pieces of information. To maximize reliability, the tool was then translated back
from Spanish into English by an independent expert linguist following the recommendation
of the International Test Commission (ITC; 2000).
Procedure
The teachers received the questionnaires from the research team during meetings
when the project aims and outline were presented. Each teacher assessed an average
Int J Clin Health Psychol, Vol. 12. Nº 2
270 RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire
of 25 children. When the children had more than one teacher, the teacher with the
highest number of class contact hours with the child carried out the assessment. The
teachers were responsible for providing the parents with questionnaires and collecting
the completed questionnaires. Where children lived with both parents the questionnaire
was filled in by both parents together. In cases where parents were separated or
divorced the parent who spent most time with the child was responsible for filling in
the questionnaire. In some other cases the questionnaire was filled in by the nearest
relative or carer who spent most time with the child. The teachers returned all questionnaires
by post. The total questionnaire response rate was 89.41% for the parent version and
100% for the teacher version. The sample included 52% males and 48% females.
Data analyses
For this instrumental study (Montero and León, 2007), data managing and analyses
were carried out using R statistical software (http://www.r-project.org/). These analyses
encompass testing the psychometric properties of the Spanish-SDQ, and a Principal
Component Analysis (PCA) over the 25 items scale to verify the factorial structure of
our questionnaire. Measures of internal consistency for the whole scale and subscales
for teachers and parents were carried out using Chronbach’s alpha scores. We ran a
PCA in which 5 factors were isolated attending to the Kaiser criterion (eigenvalues
greater than 1) and visual inspection of the Scree plot. We used Varimax Rotation, as
proposed by the original author (Goodman, 2001) and successive replication in other
populations (Du et al., 2008; Giannakopoulos et al., 2009; Smedje et al., 1999; Woerner
et al., 2004). Finally, a Confirmatory Factor Analysis with maximum likelihood procedure
was done for predicting the theoretical model for the SDQ proposed by Goodman (2001).
Results
Internal reliability for teacher responses
First, the descriptive statistics (Mean and Standard Deviation) and a measure of
discrimination (corrected Item-Total Correlation) for the scale was calculated. For teachers,
the internal reliability analysis achieved a score of .77 for the total scale. Internal
reliability for the other scales ranged from .64 to .85. The lowest internal reliability was
located in the peer problems scale. It seems that there is a certain variability in the items
that make up this scale, especially for the item 11 «Tiene al menos un amigo» («Has
at least one good friend») (Woerner et al., 2004; Smedje et al., 1999; see Tables 2 and
3).
Int J Clin Health Psychol, Vol. 12. Nº 2
RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire 271
TABLE 2. Mean, Standard Deviation and Discrimination indexes for teachers.
Item M SD
r item-
total
Tiene en cuenta los sentimientos de los otros 1.51 0.56 -.30
Es inquieto/a, hiperactivo/a no puede permanecer quieto/a 0.65 0.75 .57
Se queja con frecuencia de dolor de cabeza, estómago, nauseas 0.31 0.58 .25
Comparte frecuentemente con otros niños 1.47 0.57 -.22
Frecuentemente tiene rabietas o mal genio 0.42 0.65 .49
Es más bien solitario/a y tiende a jugar solo/a 0.32 0.57 .22
Por lo general es obediente, suele hacer lo que le piden los adultos 0.60 0.71 .38
Tiene muchas preocupaciones, a menudo parece inquieto/a.. 0.43 0.61 .47
Ofrece ayuda cuando alguien resulta herido, disgustado… 1.53 0.54 -.17
Está continuamente moviéndose y es revoltoso 0.57 0.73 .56
Tiene por lo menos un/a buen/a amigo/a 0.33 0.55 .30
Pelea con frecuencia con otros niños/as o se mete con ellos/as 0.46 0.68 .50
Se siente a menudo infeliz, desanimado o lloroso 0.27 0.53 .45
Por lo general cae bien a los otros niños/as 0.50 0.60 .41
Se distrae con facilidad, su concentración tiende a dispersarse 0.88 0.79 .65
Es nervioso/a o dependiente ante nuevas situaciones…. 0.68 0.71 .61
Trata bien a los niños/as más pequeños/as 1.70 0.52 -.25
A menudo miente o engaña 0.36 0.61 .52
Los otros niños se meten con él/ella o se burlan de él/ella 0.25 0.52 .51
A menudo se ofrece para ayudar (a padres, maestros, otros niños) 1.38 0.64 -.21
Piensa las cosas antes de hacerlas 0.86 0.70 .51
Roba cosas en casa, en la escuela o en otros sitios 0.05 0.24 .22
Se lleva mejor con adultos que con otros niños/as 0.39 0.61 .13
Tiene mucho miedos, se asusta fácilmente 0.37 0.57 .21
Termina lo que empieza, tiene buena concentración 0.86 0.77 .58
Note. M = Mean, SD = Standard Deviation.
TABLE 3. Reliability indexes for each scale and total scale for teachers.
Scales Alpha
Total Scale .77
Emotional S
y
m
p
toms Scale .71
Behavioral Problems Scale .75
H
yp
eractivit
y
Scale .85
Peer Problems Scale .64
Prosocial Scale .83
Internal reliability for parents responses
The descriptive statistics (Mean and Standard Deviation) and a measure of
discrimination (corrected Item-Total Correlation) for the scale was calculated. For parent
responses the total scale achieved a score of .76. The analysis for subscales showed
a range from .58 to .77. In the case of teachers, the scale with the lowest internal
reliability was located in the peer problems scale (see Tables 4 and 5).
Int J Clin Health Psychol, Vol. 12. Nº 2
272 RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire
TABLE 5. Reliability indexes for each scale and total scale for parents.
TABLE 4. Mean, standard deviation and discrimination indexes for parents.
Item M SD
r item-
total
Tiene en cuenta los sentimientos de los otros 1.67 0.50 -.20
Es inquieto/a, hiperactivo/a no puede permanecer quieto/a 0.88 0.79 .43
Se queja con frecuencia de dolor de cabeza, estómago, nauseas 0.48 0.69 .34
Comparte frecuentemente con otros niños 1.59 0.56 -.21
Frecuentemente tiene rabietas o mal genio 0.75 0.73 .50
Es más bien solitario/a y tiende a jugar solo/a 0.34 0.61 .22
Por lo general es obediente, suele hacer lo que le piden los adultos 0.65 0.62 .27
Tiene muchas preocupaciones, a menudo parece inquieto/a.. 0.47 0.62 .43
Ofrece ayuda cuando alguien resulta herido, disgustado… 1.72 0.50 -.08
Está continuamente moviéndose y es revoltoso 0.79 0.79 .49
Tiene por lo menos un/a buen/a amigo/a 0.21 0.47 .14
Pelea con frecuencia con otros niños/as o se mete con ellos/as 0.38 0.61 .38
Se siente a menudo infeliz, desanimado o lloroso 0.37 0.60 .48
Por lo general cae bien a los otros niños/as 0.35 0.53 .18
Se distrae con facilidad, su concentración tiende a dispersarse 1.00 0.78 .56
Es nervioso/a o dependiente ante nuevas situaciones…. 0.68 0.70 .52
Trata bien a los niños/as más pequeños/as 1.81 0.44 -.14
A menudo miente o engaña 0.48 0.62 .48
Los otros niños se meten con él/ella o se burlan de él/ella 0.50 0.66 .49
A menudo se ofrece para ayudar (a padres, maestros, otros niños) 1.47 0.59 -.24
Piensa las cosas antes de hacerlas 0.92 0.64 .31
Roba cosas en casa, en la escuela o en otros sitios 0.05 0.23 .20
Se lleva mejor con adultos que con otros niños/as 0.59 0.71 .41
Tiene mucho miedos, se asusta fácilmente 0.66 0.74 .36
Termina lo que empieza, tiene buena concentración 0.85 0.71 .41
Note. M
=
Mean, SD
=
Standard Deviation.
Scales Alpha
Total Scale .76
Emotional .71
Behavioural .62
H
yp
eractivit
y
Scale .77
Pee
r
Problems Scale .58
Prosocial Scale .69
Principal Component Analysis for SDQ, teacher sample and parents’ sample
The factors were isolated following the Kaiser criterion (eigenvalues greater than
1) and the inspections of the Scree plot. The selected Varimax Rotation criteria increased
the difference between isolated factors without losing explained variance for each item.
For the teacher analysis, the Kaiser-Meyer-Olkin (KMO) measure yielded a value of .89,
with a large value of partial correlation between variables meaning a compact pattern
Int J Clin Health Psychol, Vol. 12. Nº 2
RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire 273
For the parents’ analysis, the solution yielded for parents achieved a .85 value of
KMO index. Bartlett’s Test of Sphericity was significant (p < .05) rejecting the null
hypothesis of an identity matrix. The five-dimensional solution reached 49.2% of explained
variance, although the factorial structure posed more difficulties in its replication. We
found some items which loaded in more than one factor (see Table 7 for items loading
in each factor).
of correlation for a suitable factor analysis (Field, 2009). Bartlett’s Test of Sphericity was
significant (p < .05) rejecting the null hypothesis of an identity matrix. A five-dimensio-
nal solution with 59% of explained variance was accomplished (see Table 6 for items
loading in each factor).
TABLE 6. Factor loading, communalities and uniqueness for teachers as informants.
COMPONENTS
Conduct
Problems
Prosocial
Behavioural Hyperactivity Emotional
Symptoms
Peer
Relationship
Problems
Communalities/
Uniqueness
Behavioural
Problems
5. Temper .73 .60/.40
7. Obedient .49 .60/.40
12. Fights .77 .66/.33
18. Argue with
adult .57 .50/.49
22. Spiteful .35 .16/.83
Prosocial Behaviour
1. Considerate .72 .62/.37
4. Share .72 .55/.44
9. Helpful .81 .67/.32
17.Kind to Kids .71 .55/.43
20. Helps out .73 .59/.40
Hyperactivity
2. Restless .79 .28 .72/.27
10. Fidgeting .78 .28 .72/.27
15. Distractible .78 .79/.20
21. Reflective .65 .65/.34
25. Persistent .75 .74/.25
Emotional
Symptoms
3. Somatic .68 .49/.50
8. Worries .73 .61/.38
13. Unhappy .70 .57/.42
16. Clingy .60 .37 .60/.39
24. Fear .54 .49/.50
Peer Relationship
Problems
6. Solitary .64 .57/.42
11. Good Friend .48 .56/.43
14. Popular .33 .31 .54/.45
19. Bullied .51 .48 .54/.45
23. Best with adult .68 .48/.52
Int J Clin Health Psychol, Vol. 12. Nº 2
274 RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire
TABLE 7. Factor loading, communalities and uniqueness for parents as informants.
Confirmatory Factor Analysis with maximum likelihood procedure was done for
predicting the theoretical model for the SDQ proposed by Goodman (2001). This analysis
was done over the teachers’ responses due to the fact that this factorial structure
showed a more consistent pattern of correlations compared with the parents’ ones. The
independence of error terms was specified, and the factors were allowed to be correlated.
To test the goodness of fit a number of approaches were used. The Comparative Fit
Index (CFI), the Goodness of Fit Index (GFI), the χ2 Goodness of Fit Test and the Root
Mean Square Error of Approximation (RMSEA). For interpreting a good fit of the data,
we assumed that the CFI and GFI must to achieve values close to or greater than .95.
COMPONENTS
Conduct
Problems
Prosocial
Behavioural Hyperactivity Emotional
Symptoms
Peer
Relationship
Problems
Communalities
/Uniqueness
Behavioural
Problems
5. Temper .59 .53/.46
7. Obedient .44 .42/.57
12. Fights .60 .43/.56
18. Argue with
adult .48 .42/.57
22. Spiteful .24 .65 .46/.53
Prosocial
Behaviour
1. Considerate .66 .48/.51
4. Share .61 .40/.59
9. Helpful .63 .42/.57
17. Kind to Kids .65 .45/.54
20. Helps out .59 .50/.49
Hyperactivity
2. Restless .78 <.20 .65/.34
10. Fidgeting .79 <.20 .69/.30
15. Distractible .72 .66/.34
21. Reflective .63 .55/.44
25. Persistent .80 .70/.29
Emotional
Symptoms
3. Somatic .62 .38/.61
8. Worries .67 .49/.50
13. Unhappy .68 .51/.48
16. Clingy .58 .54/.45
24. Fear .63 .42/.57
Peer
Relationship
Problems
6. Solitary .49 <.20 .39/.60
11. Good Friend -.46 <.20 .40/.60
14. Popular -.41 <.20 .35/.64
19. Bullied .43 .44/.55
23. Best with
adult .61 .52/.47
Int J Clin Health Psychol, Vol. 12. Nº 2
RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire 275
For the RMSEA, the value must to be close to or less than .05 indicate a good fit, and
values as high as .08 indicate a reasonable fit (Hu and Bentler, 1999). CFA was carried
out using the EQS 6.0 (Bentler, 1995). Fit indices resulting from the CFA for the five
factors hypothesized model were estimated. The resulting values for CFI, GFI and
RMSEA were .93, .84 and .073 (.068 - .077), respectively. The χ2 goodness of fit yield
significant [χ2(260)= 1,069.83; p < .05)]. The Alfa reliability coefficient was .77. The range
of factors loading was .19 to .76. Being the item «Termina lo que empieza, tiene buena
concentración» and the item «Por lo general cae bien a otros niños» those which the
lowest and highest load respectively.
Discussion
The Spanish version of the SDQ for teachers and parents was administered in a
sample of 595 children with the aim of establishing its psychometric properties. There
was a high response rate which indicates its acceptance in community studies.
With regard to the psychometric properties of this instrument, the values obtained
for the internal reliability for teacher responses are similar to those achieved in other
studies (Du et al., 2008; Giannakopoulos et al., 2009; Goodman, 2001; Hawes and
Dadds, 2004; Matsuishi et al., 2008; Palmieri and Smith, 2007; Smedje et al., 1999;
Woerner et al., 2004). Chronbach’s alpha and other indexes of reliability were adequate
and comparable to other European research studies and those found in the original
English version (Goodman, 2001). For parent data we found minor but acceptable values
compared to the teacher SDQ. Both samples shared high reliability values in the total
difficulties scale and low values for the peer problems scale (Becker et al., 2004; Muris,
Meesters, and Van den Berg, 2003; Stone, Otten, Engels, Vermulst, and Janssens, 2010).
Some authors assume that scales with fewer items have less internal consistency than
more extensive ones (Van Widenfelt et al., 2003). This could explain why the total scale
score has a greater reliability coefficient than the individual scales.
Exploratory factor analysis for teacher responses replicated the original questionnaire.
However some items yielded a load in a different factor than that predicted. The
resulting pattern of main loading was nearly consistent with the factor structure of the
original questionnaire (Goodman, 1997; Goodman, 2001). Finally, taking the result from
the CFA we assume a replication of the values obtained in the literature that indicate
a moderate-adequate fit for the factor structures proposed originally by Goodman (2001).
The pattern of the Exploratory Factor Analysis for teachers’ reveals that items number
2 («Restless, overactive, cannot stay still for long») and number 10 («Constantly
fidgeting or squirming») loaded in the conduct problems scale instead of the hyperactivity
scale. Item number 16 («Nervous or clingy in new situations, easily loses confidence»)
loaded on the hyperactivity scale instead of the emotional symptoms scale. Finally,
items 14 («Generally liked by other children») and 19 («Picked on or bullied by other
children») loaded on the behavioral problems scale instead of the peer problems scale.
Nevertheless, the factor loading differences in these two items is a mere .2 and .3
respectively.
Int J Clin Health Psychol, Vol. 12. Nº 2
276 RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire
A closer inspection of items that loaded in other factor than that predicted showed
that in the case of item 2 the Spanish translation uses the term «hiperactivo» for the
term «overactive». This could create a confound, prompting the item to be saturated in
both the behavioural problems and hyperactivity scales . In the case of item 10 the
translation for «fidgeting» and «squirming» was «moviéndose» and «revoltoso». In
Canary Island culture a child that «moves a lot» and «does not stop» is possibly not
seen as a child with hyperactivity. Instead, this could be seen as a behavior problem.
So this could be the reason for the load of item 10 in the behavioral problems scale.
So it seems that items 2 and 10 must be revised for the Spanish version in the Canary
Islands. There is another theoretical reason for the pattern of load found in items 2 and
10. The terms «restless» and «fidgeting» could be understood as relating to a behavioural
problem and not a cognitive problem. In the hyperactivity scale, the items 15 (distractible),
21 (reflective) and 25 (persistent) are related to cognitive aspects of the hyperactivity
construct. However items 2 and 10 refer to behavioural problems in the hyperactivity
construct. This could be the reason why teachers and parents assign these items to
behavioural problems instead of hyperactivity. In the case of items 14 and 19, clearly
these could be a measure of not just peer problems but problems in general, as in the
behavioural problems scale.
Exploratory factor analysis run on parent data showed that, as in the factorial
structure of the teacher questionnaire, item 2 and item 10 loaded on the behavioural
problem scale instead of the hyperactivity scale. It seems that item 2 and 10 are not
performing properly on the hyperactivity scale. This item seems to be evaluating other
aspects of behaviour not understood as hyperactivity. Item 6 («Rather solitary, tends
to play alone») loaded on the emotional symptoms scale instead of the peer problems
scale. In the case of item 6, being a «solitary» child seems to be understood by the
parents as an emotional problem, not a problem of behavior with others. Items 11 («Has
at least one good friend») and 14 («Generally liked by other children») loaded on the
prosocial behaviour scale with a negative Pearson correlation, instead of the peer
problems scale. It seems that for our sample of parents, to «Have at least one good
friend» and «Generally liked by other children» are more related to prosocial behaviour
than to peer relationship problems. If we check the items of the prosocial behaviour
scale, these items 11 and 14, are related in a negative sense to this scale -»helps others»,
«kind to kids», «helpful», «shares with other children» and «considers others feelings».
Item number 22 («Steals from home, school or elsewhere») loaded on the peer problems
scale, instead of the behavioural problems scale. It seems that parents’ understanding
of «spiteful» was perceived as related to others in the school rather than being seen
as a conduct problem which is exhibited in different contexts. Overall, the data from
parents seems to be less clear in the factorial structure proposed for the SDQ. With
regard to the items saturation, the peer problems scale only has two items with a proper
load. The rest of the items from the scale loaded in other scales. These patterns of
results have been found in most of the samples using parents and teachers as informants
(Becker et al., 2004; Du et al., 2007; Obel et al., 2004; Stone et al., 2010; Woerner et
al., 2004).
Int J Clin Health Psychol, Vol. 12. Nº 2
RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire 277
Although the current study provides the first evaluation in Spanish of the SDQ,
additional research with different samples in Spain are needed in order to accumulate
data about the validity and internal factor structure. As a screening instrument suitable
for professionals in school contexts to evaluate morbidity, the SDQ fits the psychometric
and statistic requisites of this type of instrument. Once children with risk are detected,
a proper evaluation with more fine grained instruments is still needed. The Spanish
version of SDQ revealed psychometric properties comparable to other versions in
different countries. However, testing the predictive role of the scale is necessary in
order to test the factorial structure.
References
Becker, A., Woerner, W., Hasselhorn, M., Banaschewski, T., and Rothenberger, A. (2004).
Validation of the parent and teacher SDQ in a clinical sample. European Child and
Adolescent Psychiatry, 13, 11-16.
Belfer, M.L. (2008). Child and adolescent mental disorders: the magnitude of the problem across
the globe. Journal of Child Psychology and Psychiatry, 49, 226-236.
Bentler, P. M. (1995). EQS Structural Equations Program Manual. Encino, CA: Multivariate
Software, Inc.
Carretero-Dios, H. and Pérez, C. (2007). Standards for the development and the review of
instrumental studies: Considerations about test selection in psychological research.
International Journal of Clinical and Health Psychology, 7, 863-882.
Crone, M.R., Vogels, A.G., Hoekstra, F., Treffers, P.D., and Reijneveld, S.A. (2008). A comparison
of four scoring methods based on the parent-rated Strengths and Difficulties Questionnaire
as used in the Dutch preventive child health care system. BMC Public Health, 8, 106.
Di Riso, D., Salcuni, S., Chessa, D., Raudino, A., Lis, A., and Altoè, G. (2010). The Strenghts
and Difficulties Questionnaire (SDQ). Early evidence of its reliability and validity in a
community sample of Italian children. Personality and Individual Differences, 49, 570-
575.
Du, Y., Kou, J., and Coghill, D. (2008). The vality, reability and normative scores of the parents,
teacher and self report versions of the Strengths and Difficulties Questionnaire in China.
Child and Adolescent Psychiatry and Mental Health, 2, 8.
Estévez, E. and Emler, N.P. (2011). Assessing the links among adolescent and youth offending,
antisocial behaviour, victimization, drug use, and gender. International Journal of Clinical
and Health Psychology, 11, 269-289.
Field, A. (2009). Discovering statistics using SPSS. London: Sage.
Giannakopoulos, G., Tzavara, C., Dimitrakaki, C., Kolaitis, G., Rotsika, V., and Tountas, Y.
(2009). The factor structure of the Strengths and Difficulties Questionnaire (SDQ) in
Greek adolescents. Annals of General Psychiatry, 8, 20.
Glazebrook, C., Hollis, C., Heussler, H., Goodman, R., and Coates, L. (2003). Detecting
emotional and behavioural problems in paediatric clinics. Child: Care, Health and
Development, 29, 141-149.
Goodman, R. (1997). The Strengths and Difficulties Questionnaire: A research note. Journal of
Child Psychology and Psychiatry, 38, 581-586.
Goodman, R. (2001). Psychometric Properties of the Strengths and Difficulties Questionnaire.
Journal of the American Academy of Child and Adolescent Psychiatry, 40, 1337-1345.
Int J Clin Health Psychol, Vol. 12. Nº 2
278 RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire
Goodman, R., Ford, T., Simmons, H., Gatward R., and Meltzer, H. (2000). Using the Strengths
and Difficulties Questionnaire (SDQ) to screen for child psychiatric disorders in a
community sample. British Journal of Psychiatry, 177, 534-539.
Goodman, A. and Goodman, R. (2009). Strengths and difficulties questionnaire as a dimensional
measure of child mental health. Journal of the American Academy of Child and Adolescent
Psychiatry, 48, 400-403.
Goodman, R., Renfrew, D., and Mullick, M. (2000). Predicting type of psychiatric disorder from
Strenghts and Difficulties Questionnaire (SDQ) scores in child mental health clinics in
London and Dhaka. European Child and Adolescent Psychiatry, 9, 129-134.
Goodman, R. and Scott, S. (1999). Comparing the Strengths and Difficulties Questionnaire and
the Child Behavior Checklist: is small beautiful? Journal of Abnormal Child Psychology,
27, 17-24.
Griffin, C., Guerin, S., Sharry, J., and Drumm, M. (2010). A multicentre controlled study of an
early intervention parenting programme for young children with behavioural and
developmental difficulties. International Journal of Clinical and Health Psychology, 10,
279-294.
Hawes, D.J. and Dadds, M.R. (2004). Australian data and psychometric properties of the
Strengths and Difficulties Questionnaire. Australian and New Zealand Journal of Psychiatry,
38, 644-651.
Hill, C.R. and Hughes, J.N. (2007). An Examination of the Convergent and Discriminant Validity
of the Strengths and Difficulties Questionnaire. School Psychology Quarterly, 22, 380-
406.
Hu, L. and Bentler, P.M. (1999) Cutoff criteria for fit indices in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55.
International Test Commission (ITC) (2000). International Guidelines for Test Use. Downloaded
electronically on 2011 December 05 from www.intestcom.org/itc_projects.htm.
Klasen, H., Woerner, W., Wolke, D., Meyer, R., Overmeyer, S., and Kaschnitz, W. (2000).
Comparing the German Versions of the Strenghts and Difficulties Questionnaire (SDQ-
Deu) and the Child Behavior Checklist. European Child and Adolescent Psychiatry, 9,
271-276.
Koskelainen, M., Sourander, A., and Kaljonen, A. (2000). The Strengths and Difficulties
Questionnaire among Finnish school-aged children and adolescents. European Child and
Adolescent Psychiatry, 9, 277-284.
Malhotra, S., Kohli, A., Kapoor, M., and Pradhan, B. (2009). Incidence of childhood psychiatric
disorders in India. Indian Journal of Psychiatry, 51, 101-107.
Mansbach-Kleinfeld, I., Levinson, D., Farbstein, I., Apter, A., Levav, I., and Kanaaneh, R.
(2010). The Israel survey of mental health among adolescents: Aims and methods. Israel
Journal of Psychiatry and Related Sciences, 47, 244-253.
Mathai, J., Anderson, P., and Bourne, A. (2003). Use of the Strengths and Difficulties Questionnaire
as an outcome measure in a child and adolescent mental health service. Australasian
Psychiatry, 11, 334-337.
Mathai, J., Anderson, P., and Bourne, A. (2004). Comparing psychiatric diagnosis generated by
the Strenghts and Difficulties Questionnaire with diagnoses made by clinicians. Australian
and New Zealand Journal of Psychiatry, 38, 639-643.
Matsuishi, T., Nagano, M., Araki, Y., Tanaka, Y., Iwasaki, M., and Yamashita, Y. (2008). Scale
properties of Japanese version of the Strengths and Difficulties Questionnaire (SDQ): A
study of infant and school children in community samples. Brain Development, 30, 410-
415.
Int J Clin Health Psychol, Vol. 12. Nº 2
RODRÍGUEZ-HERNÁNDEZ et al. Strength and Difficulties Questionnaire 279
Merikangas, K.R., Nakamura, E.F., and Kessler, R.C. (2009). Epidemiology of mental disorders
in children and adolescents. Dialogues in Clinical Neurosciences, 11, 7-20.
Montero, I. and León, O.G. (2007). A guide for naming research studies in Psychology. International
Journal of Clinical and Health Psychology, 7, 847-862.
Mullick, M.S. and Goodman, R. (2001). Questionnaire screening for mental health problems in
Bangladeshi children: a preliminary study. Social Psychiatry and Psychiatric Epidemiology,
36, 94-99.
Muris, P., Meesters, C., and Van den Berg, F. (2003). The Strengths and Difficulties Questionnaire
(SDQ): Further evidence for its reliability and validity in a community sample of Dutch
children and adolescents. European Child and Adolescent Psychiatry, 12, 1-8.
Obel, C., Heiervang, E., Rodriguez, A., Heyerdahl, S., Smedje, H., and Sourander, A. (2004). The
Strengths and Difficulties Questionnaire in the Nordic countries. European Child and
Adolescent Psychiatry, 13, 32-39.
Oschilewsky, R.C., Gómez, C.M., and Belfort, E. (2010). Child psychiatry and mental health
in latin America. International Review of Psychiatry, 22, 355-362.
Palmieri, P.A. and Smith, G.C. (2007). Examining the Structural Validity of the Strengths and
Difficulties Questionnaire (SDQ) in a U.S. Sample of Custodial Grandmothers. Psychological
Assessment, 19, 189-198.
Shivram, R., Bankart, J., Meltzer, H., Ford, T., Vostanis, P., and Goodman, R. (2009). Service
utilization by children with conduct disorders: Findings from the 2004 Great Britain child
mental health survey. European Child and Adolescent Psychiatry, 18, 555-563.
Smedje, H., Broman, J.E., Hetta, J., and Von Knorring, A.L. (1999). Psychometric properties of
a Swedish version of the «Strengths and Difficulties Questionnaire». European Child and
Adolescent Psychiatry, 8, 63-70.
Stone, L.L., Otten, R., Engels, R.C., Vermulst, A.A., and Janssens, J.M. (2010). Psychometric
properties of the parent and teacher versions of the Strengths and Difficulties Questionnaire.
Clinical Child and Family Psychology Review, 13, 254-274.
Thabet, A.A., Stretch, D., and Vostanis, P. (2000). Child mental health problems in Arab children:
Application of the Strengths and Difficulties Questionnaire. International Journal of
Social Psychiatry, 46, 266-280.
Van Widenfelt, B.M., Goedhart, A.W., Treffers, P.D., and Goodman, R. (2003). Dutch version
of the Streinghts and Difficulties Questionnaire. European Child and Adolescent Psychiatry,
12, 281-289.
Vogels, A.G., Crone, M.R., Hoekstra, F., and Reijneveld, S.A. (2009). Comparing three short
questionnaires to detect psychosocial dysfunction among primary school children: A
randomized method. BMC Public Health, 9, 489.
Vostanis, P. (2006). Strengths and Difficulties Questionnaire: Research and clinical applications.
Current Opinion in Psychiatry, 19, 367-372.
Wille, N., Bettge, S., Wittchen, H.U., and Ravens-Sieberer, U. (2008). How impaired are children
and adolescents by mental health problems? Results of the BELLA study. European
Child and Adolescent Psychiatry, 17, 42-51.
Woerner, W., Becker, A., and Rothenberger, A. (2004). Normative data and scale properties of
the German parent SDQ. European Child and Adolescent Psychiatry, 13, 3-10.
Received July 27, 2011
Accepted January 10, 2012