357

Bullying is a subtype of aggression that can adopt many forms

(physical, verbal, and relational) and that includes among its main
features an imbalance of power, intentionality and persistence
over time (Olweus, 1993). Its prevalence varies between different
countries and studies, ranging from 4% (Daz-Aguado, Martnez-
Arias, & Martn-Babarro, 2010) to 28% (Rivers & Smith, 1994),
with some authors placing it at approximately 15% (Smith &
Shu, 2000). Most of these studies have used self-reports as their
data collection methodology. However, the use of questionnaires
based on peer reports reduces measurement errors and increases
reliability because they provide scores that are based on multiple
informants (Jimerson, Swearer, & Espelage, 2009). Correlations
between self- and peer reports are generally very low, between
.14 and .42 (Achenbach, McConaughy, & Howell, 1987; Juvonen,
Nishina, & Graham, 2001). Furthermore, due to peer reports,
respondents tend to overcome the secrecy and silence code about
bullying that often exists in the classroom.
The evidence consistently points to the fact that boys are more
prone to adopt both the role of bully and of victim than are girls
(Cerezo, 2000; Lucas-Molina, Pulido-Valero, & Solbes-Canales,
2011). During bullying episodes, there are different roles and
degrees of participation, from active students supporting the
main aggressors to others who oppose and are annoyed with such
behavior, as well as students with neutral and/or indifferent attitudes
(Salmivalli, Lagerspetz, Bjrkqvist, sterman, & Kaukiainen,
1996). Isolation and the lack of friends in the group increase the
risk of victimization (Huttunen, Salmivalli, & Lagerspetz, 1996);
rejected students are also more frequently bullied (Lucas-Molina
et al., 2011; Salmivalli et al., 1996). This relationship between
victimization and rejection by the peer group has led to the
incorporation of sociometry as a tool to assess bullying.
ISSN 0214 - 9915 CODEN PSOTEG
Copyright 2014 Psicothema
www.psicothema.com
Assessment and detection of peer-bullying through analysis
of the group context
Javier Martn Babarro
Universidad Complutense de Madrid
Abstract Resumen
Background: The assessment of bullying requires an analysis both of the
main proles involved in this phenomenon and of the social context in
which it occurs. By considering both aspects, this study develops a scale
that, in addition to individual information, incorporates a representation of
the group structure of the classroom. Method: A large sample composed
of 11,561 students (mean age = 11.12 years, girls = 49.2%) from 108
schools completed the Sociescuela Scale by peer reports. An analysis of the
internal structure and reliability of the scale was performed, as well as of
the students social networks. Results: Factor analysis yielded ve factors:
Victimization, Acceptance, Prosociality, Withdrawal, and Aggressiveness.
Boys showed more victimization and aggressiveness than girls. The results
obtained enable us to: (a) evaluate a series of individual proles associated
with involvement in bullying and their sociometric status, and (b) position
them on a social map of each classroom. Conclusions: The data suggested
that the scale is reliable and valid for use in the detection of bullying and its
applied nature facilitates the design of school interventions.
Keywords: Sociometry, bullying, social network analysis, scale.
Evaluacin y deteccin del acoso escolar a travs del anlisis del contexto
grupal. Antecedentes: la evaluacin del acoso escolar requiere tanto de un
anlisis de los principales perles involucrados como del contexto social
en el que se produce. Considerando ambos aspectos, este estudio desarrolla
una escala en la que adems de la informacin individual, se incorpora
una representacin de la estructura grupal del aula. Mtodo: una amplia
muestra compuesta por 11.561 estudiantes (edad media = 11,12, chicas
= 49,2%) de 108 centros educativos completaron la escala Sociescuela a
travs de heteroinforme. Se analizaron la estructura interna y la abilidad
de la escala, as como las redes sociales del grupo. Resultados: el
anlisis factorial distingui cinco factores: Prosocialidad, Retraimiento,
Agresividad, Victimizacin y Aceptacin. Los chicos mostraron un mayor
nivel de victimizacin y de agresividad. Los resultados obtenidos permiten:
(1) evaluar una serie de perles individuales asociados a la participacin
en el acoso escolar, as como su estatus sociomtrico, y (2) situarlos en
un mapa social de cada clase. Conclusiones: los anlisis sugieren que la
escala es able y vlida para ser utilizada en la deteccin del acoso escolar.
Su carcter aplicado facilita el diseo de intervenciones en los centros
educativos.
Palabras clave: sociometra, acoso escolar, anlisis de redes sociales,
escala.
Psicothema 2014, Vol. 26, No. 3, 357-363
doi: 10.7334/psicothema2014.85

Received: April 8, 2014 Accepted: April 10, 2014
Corresponding author: Javier Martn Babarro
Facultad de Psicologa
Universidad Complutense de Madrid
28223 Madrid (Spain)
e-mail: jbabarro@psi.ucm.es
Javier Martn Babarro
358
With this in mind, several authors have proposed increasingly
elaborate instruments based on peer reports, analyzing participant
roles in bullying (Salmivalli et al., 1996; Lucas-Molina et al.,
2011) or collecting information on victimization behaviors and
perceptual attributes associated with major bullying proles
(Bjrkqvist & sterman, 1995). The most complete scales have
also incorporated information on social preferences (Daz-
Aguado, 1986; Cerezo, 2000). However, despite the variety of
scales published to date, all of them are based on the analysis of
individual proles, but none of them analyzes the structure of the
group in which episodes of bullying occur. Social preferences,
friendships among students, and the main bullying-related
proles are elements included within a larger framework, such as
the social networks of the classroom (Gifford-Smith & Brownell,
2003). Additionally, over 80% of cases originate in the class group
(Salmivalli & Peets, 2008). Hence, evaluating these networks is
essential to understand this phenomenon. Another relevant aspect
is that the existing scales (Cerezo, 2000; Daz-Aguado, 1986) base
their sociometric information on social preference, which does not
represent a real tie among students as do friendship groups.
The need to analyze more specically the group context
surrounding bullying motivated the elaboration of this instrument.
Thus, this study develops the Sociescuela Scale in order to identify
the main individual proles associated with bullying and to
observe these proles embedded in a social map of the classroom.
The main advantage of this instrument is that, for the rst time, a
technique borrowed from social network analysis for developing
social maps (NEGOPY, Social Cognitive Mapping) is applied to a
scale of bullying in order to analyze class groups. Items about real
ties among students thus provide information about the cliques
formed. A second advantage compared with previously published
scales is that it offers more comprehensive information on the
sociometric prole by using widely adopted correction procedures
such as nominations (Coie, Dodge, & Coppotelli, 1982; Newcomb
& Bukowski, 1983) and ratings (Maassen, Akkermans, & van der
Linden, 1996).
Method
Participants
The sample was obtained through a non-random sampling (N
= 11,561 students) and using a cross-sectional design within a
wider assessment effort conducted in 108 schools in four regions
of Spain (Madrid, Castile and Leon, Andalusia and Castile-La
Mancha). The mean age was 11.32 (SD = 1.44) years and 49.2%
(n = 5,688) were girls. In the sample, 29.3% (n = 3,387) were in
the fth grade and 24.5% (n = 2,832) were in the sixth grade of
primary education. In addition, 21.3% (n = 2,462) were in the rst
year and 24.9% (n = 2,878) were in the second year of secondary
education.
Instrument
The instrument is divided into three subscales the items of
which were collected through peer nominations. The rst group
of questions regarding bullying victimization was created from
items adapted from questionnaires obtained by self-reports of
victimization (Defensor del Pueblo [Public Ombudsman], 2000).
The second group contained sociometric questions using items
adapted from Daz-Aguado (1986). Finally, a set of questions was
based on the method of perceived attributes, using items from
Sutton and Smith (1999).
In the rst phase, an initial validation of the 34-item
questionnaire was performed on a sample of 3,160 adolescents
from 25 schools (Martn-Babarro, 2011). The initial scale based
on peer reports was then compared with other scales obtained by
self-reports of victimization (Defensor del Pueblo, 2000). The
ndings showed correlations of .22 for physical victimization, .31
for verbal victimization, and .39 for relational victimization. These
results are similar to those found by other authors (Achenbach et
al., 1987; Juvonen et al., 2001). The scale was also compared with a
scale of behavioral disorders, ESPERI (Parellada, San-Sebastin,
& Martnez-Arias, 2009), nding correlations between the bully
role and the factors of impulsivity-inattention (.25), dissocial (.35)
and hyperactivity (.31) as pointed out in prior studies (Valdivia-
Peralta, Fonseca-Pedrero, Gonzlez-Bravo, & Lemos-Girldez,
2014). Subsequently, in a second phase, which comprises the
present study, the questionnaire was reduced to 20 items and a
software application was developed in the PHP language version
5.3.14, with an Apache server and a MySQL database.
Regarding the number of nominations, some authors have found
questionnaires to be more reliable and valid when an unlimited
number of nominations is applied (Terry, 2000). In our research,
after a rst phase in which an unlimited number of nominations
was applied to all items, we decided to maintain a high number of
nominations in the sociometric questions (up to 12 nominations).
This is because the vast majority of students tended to show a
sharp decline in their responses as of the eighth nomination.
The scores of each question were weighted by the order of each
nomination and compared with unweighted scores, but the results
showed no signicant differences.
Victimization. A group of three items was proposed to evaluate
victimization: one aimed at measuring physical victimization,
another for verbal victimization and a third for relational
victimization. A proportional score was then calculated for each
item by dividing the number of nominations received for each
student by the number of classmates who answered the item.
Finally, an average score using all three questions was calculated,
obtaining a range between 0 and 1 (.00 E .33, M = .01, SD
= .03).
Acceptance. This subscale consists of ve items based on
sociometry. First, through a nominations method, two items were
used, one for obtaining positive nominations and another for
negative nominations, with a maximum of 12 nominations. Second,
a rating method was applied in which all student were required to
rate all of their classmates (Which of your classmates do you like
or dislike?). For this assessment, a 7-point Likert scale was used.
The correction method proposed by Asher and Dodge (1986) was
used, through which positive ratings were computed as positive
nominations and the lowest ratings as negative nominations.
Social preferences (SP) in the ratings procedure were calculated
from the scores of elections (E) and rejections (R) (SP = E - R).
Finally, two items for measuring friendship groups (Which
classmates are your friends?) and frequent interaction groups
(Which classmates do you hang out with?), with a maximum of
12 nominations per item, were used. Proportional scores were then
calculated for each item as in the previous subscale and a nal
average score using all ve elements was obtained (-.29 E .67,
M = .13, SD = .11).
Assessment and detection of peer-bullying through analysis of the group context
359
Additionally, the software calculates the categories of
sociometric status (popular, rejected, isolated, controversial,
average) with three methods. For the nominations procedure, a
method based on standard scores (Coie et al., 1982) and a second
method based on probabilistic scores (Newcomb & Bukowski,
1983) were applied. The latter is the more conservative in
allocations, forming smaller, more homogeneous and extreme
groups. For the ratings procedure, a method based on probability
criteria with a two-dimensional classication (Maassen et al.,
1996) was applied.
Social map. Using items to measure friendship or frequent
interaction groups and techniques based on social network
analysis (e.g., NEGOPY, Social Cognitive Mapping), a graphical
representation of the group structure was created. By means
of binary data (0 = no choice, 1 = choice), a logarithm was
applied, based on the following guidelines: (a) a dyad is formed
by reciprocal nominations among its members, (b) a group is
identied when at least three children have reciprocal nominations
with other members of the same group and are all linked by ties
of friendship throughout the group, and (c) a student who has not
answered the test, but has received two or more nominations from
members of a group, is incorporated into that group. This creates
various situations and enables us to detect participants who are
members of a social group and participants who are not members
of any group. In turn, the latter are divided into isolated students
and students who are liaisons between groups.
Perceived attributes. This subscale consists of 12 items
regarding various perceived attributes with a maximum of three
nominations per item. The items are related to the key roles that
typically appear in episodes of bullying: (a) a prole of withdrawal
is often associated with a victimization role, (b) an aggressive
prole is often associated with a bully role, and (c) a prosocial
prole is usually present in most cooperative students with high
social status in the group (Daz-Aguado, 1986; Lucas-Molina et
al., 2011). Each factor consisted of four items. Proportional scores
were calculated for each item, and a mean score ranging from 0
to 1 was obtained for Withdrawal (.00 E .93, M = .07, SD
= .11), Aggressiveness (.00 E .86, M = .05, SD = 0.10), and
Prosociality (.00 E .91, M = .11, SD = .12).
Procedure
Students participated voluntarily, and parental consent was
requested. The test lasted 20 to 25 minutes. It can be implemented
individually; however, for this research it was conducted
collectively, by classroom groups and in the computer rooms in
each school. Students responded to questions that appeared on an
array of photographs and names of classmates.
Data analysis
Data analysis was conducted with SPSS 20.0 (SPSS Inc.,
2011) and Factor (Lorenzo-Seva & Ferrando, 2006). Descriptive
statistics of items were calculated, followed by exploratory
factor analysis using Promax rotation to extract the scale factors.
Subsequently, the internal consistency of each factor obtained
was calculated with Cronbachs alpha coefcient. Students t-test
analysis of the differences in factors by gender was performed.
Finally, Pearson correlations between the factors obtained were
calculated.
Results
Item analysis
Table 1 shows the descriptive statistics for the items. Mean
scores tended to be lower for Items 1, 2, and 3 about victimization,
with respective values of .02, .01 and .01. Items 4 and 6, involving
positive peer nominations, showed the highest mean values (.32
and .75, respectively). The items were highly and positively
skewed, except for Item 6, which was negatively skewed. Kurtosis
showed high values and a leptokurtic distribution for most items,
except for Item 4, which was platykurtic.
Exploratory factor analysis

Because the distributional properties of items are more likely
to be skewed than continuous and normally distributed, the
questionnaire was analyzed based on polychoric correlations
through exploratory factor analysis, using the weighted least
squares estimation method. The Promax oblique rotation was also
applied because of the relationship among some of the factors.
The ndings revealed a ve-factor solution (Victimization,
Acceptance, Prosociality, Withdrawal, and Aggressiveness), with
high loadings of these items on all of the factors (Table 2). A
measure of sampling adequacy, Kaiser-Meyer-Olkin (KMO = .80),
and Bartletts sphericity test (
2
[91] = 5441.4, p<.001) indicated
the appropriateness of factor analysis. Subsequently, the reliability
Table 1
Composition of Sociescuela and the descriptive statistics of the items
Items Mean SD Skewness Kurtosis
1. Which classmate is hit or physically mistreated
because of his/her weakness in the group? (1)
.02 .06 5.7 43.7
2. Which classmate is insulted or humiliated by
others? (1)
.01 .04 4.7 30.5
3. Which classmate is isolated or ignored by
others? (1)
.01 .04 4.2 27.1
4. Who do you like to sit next to? (2) .32 .18 .47 -.38
5. Who do you dislike to sit next to? (2)* .21 .18 1.12 1.01
6. Who do you like or dislike? (2) .75 .18 -1.29 2.07
7. Who do you hang out with? (2) .23 .13 .66 1.08
8. Who are your friends? (2) .21 .13 .78 1.17
9. Who has a good relationship with teachers? (3) .09 .11 1.68 4.34
10. Who treats others well? (3) .11 .12 1.79 4.5
11. Who helps others? (3) .11 .13 1.9 4.9
12. Who is polite and respectful? (3) .10 .12 1.73 4.2
13. Who is bossy? (4) .05 .10 2.8 10.71
14. Who has a bad relationship with teachers? (4) .06 .14 3.2 11.04
15. Who disturbs others? (4) .05 .13 3.30 12.39
16. Who is more aggressive? (4) .04 .11 3.77 16.85
17. Who is often sad? (5) .06 .12 3.05 11.23
18. Who is fearful? (5) .08 .10 3.21 12.39
19. Who is shy? (5) .05 .13 3.60 14.24
20. Who has problems communicating? (5) .10 .17 2.28 5.33
(1) Victimization (up to 12 nominations), (2) Acceptance (up to 12 nominations), (3)
Prosociality (up to three nominations), (4) Aggressiveness (up to three nominations), (5)
Withdrawal (up to three nominations)
* Note: item is reverse scored
Javier Martn Babarro
360
coefcients of these factors were calculated with Cronbachs alpha
coefcient (Table 3). As the purpose of the scale was to provide
a series of proles and a level of victimization in order to locate
them on a social map of the class, we did not calculate a second-
order factor analysis.
Sex differences in the factors were analyzed with Students
t-test (Table 4). In all cases, the variables presented homogeneous
variances and therefore, contrast with equal variances was applied
(df = 11,561) to the responses of all participants. All the analyses
yielded signicant differences. Boys obtained higher levels of
Victimization and Aggressiveness than girls (Victimization:
.07 and .03; Aggressiveness: .32 and .12, for boys and girls,
respectively), both with a medium effect size (Cohens d = .53
and .50, respectively). In turn, girls obtained higher scores on
Acceptance and Prosociality than boys (Acceptance: .85 and .79;
Prosociality: .57 and .35, for girls and boys, respectively), with a
low effect size (Cohens d = .42 and .43, respectively). Girls also
showed a higher level of Withdrawal than boys (.33 vs. .24) with
a lack of an effect size. Subsequently, ANOVA was conducted to
determine the differences in each factor according to grade level,
nding no signicant differences.
Correlations

Pearsons correlation analysis between the factors obtained was
also carried out (Table 5). Statistically signicant relationships
between all factors were found, highlighting the positive
relationships between Victimization and Withdrawal and between
Prosociality and Acceptance. A moderately negative relationship
between Aggressiveness and Acceptance was observed, and
negative correlations between Acceptance and Victimization and
between Acceptance and Withdrawal were also found.
Social map

The software application displays a visual representation of three
proles associated with bullying: Withdrawal, Aggressiveness,
and Prosociality. This highlights students with scores above the
80
th
percentile in each factor, marking them with different frames
(Figure 1).
A social network analysis allows us to build a social map
through questions about friendship or frequent interaction groups.
The above-mentioned individual proles related to bullying and
sociometric information were incorporated into this social map
(Figure 2).
Discussion

This study analyzes a tool for the assessment of bullying based
on previous tests (Defensor del Pueblo, 2000; Daz-Aguado,
1986; Sutton & Smith, 1999) and on the consideration that this
phenomenon is conditioned by social context and not exclusively
by the bully-victim interaction. Scales for measuring bullying
Table 2
Rotated factor matrix
F1 F2 F3 F4 F5
Item 1 .701
Item 2 .750
Item 3 .696
Item 4 .865
Item 5 .754
Item 6 .740
Item 7 .864
Item 8 .891
Item 9 .801
Item 10 .392 .640
Item 11 .334 .727
Item 12 .703
Item 13 .753
Item 14 .809
Item 15 .853
Item 16 .872
Item 17 .624
Item 18 .739
Item 19 .792
Item 20 .651
Note: Items with loadings between -.30 and .30 were not considered
Table 3
Reliability coefcients of the subscales
Cronbachs alpha
Victimization .73
Acceptance .86
Prosociality .84
Withdrawal .78
Aggressiveness .89
Table 4
Descriptive statistics for each subscale
Sex
Total
M (SD) Boys
M (SD)
Girls
M (SD)
T Cohens d
Victimization .02(.04) .01(.03) 12.38*** .53 .01(.03)
Acceptance .79(.71) .85(.68) -5.32***
.42

.82(.69)
Prosociality .09(.10) .14(.13) -24.29*** .43 .11(.12)
Withdrawal .06(.10) .08(.12) -10.57*** .18 .07(.11)
Aggressiveness .08(.13) .03(.06) 25.51*** .50 .05(.10)
* p<.05; ** p<.01; *** p<.001
Table 5
Correlations between the factors obtained
Victimiza-
tion
Prosocia-
lity
Aggressive-
ness
With-
drawal
Accep-
tance
Victimization
Prosociality -.117
**

Aggressiveness .062
**
-.246
**

Withdrawal .225
**
.076
**
-.170
**

Acceptance -.352
**
.454
**
-.269
**
-.291
**

* p<.05; ** p<.01
Assessment and detection of peer-bullying through analysis of the group context
361
Figure 1. The gure represents the students sorted by their class group. Students with any of the specied attributes have frames around their pictures
Figure 2. Representation of the structure of the class group
Javier Martn Babarro
362
have gradually evolved, going from lists based on victimization
and aggressive behaviors obtained from self-reports to
increasingly complex questionnaires obtained from peer-reports
that assess both victimization and the main bullying-related
proles and social context information. The tool presented herein
aims to improve the existing scales in the area of social context
information.
The analysis of individual variables showed adequate
psychometric properties of the instrument, with a solid internal
consistency in the different subscales. The test showed three
proles associated with the main roles involved in bullying. First,
we found a Prosocial prole with characteristics such as having
good relationships with teachers, treating others well, or being
willing to help others, which had a high level of Acceptance in the
group, concurring with previous results (Salmivalli et al., 1996;
Lucas-Molina et al., 2011; Sutton & Smith, 1999). Second, the
analysis revealed an Aggressive prole related to features such as
being dominant and disruptive as well as having poor relationships
with teachers, which showed a low level of Acceptance, in line
with the ndings of previous research (Daz-Aguado, 1986;
Lucas-Molina et al., 2011). Third, a prole of Withdrawal was
related to characteristics such as shyness and introversion in
peer relationships as well as low self-esteem. This factor also
seemed to be associated with higher levels of Victimization and
lower Acceptance, which could be classied as the passive victim
type corresponding with the results found in previous research
(Cerezo, 2000; Daz-Aguado, 1986). The results also revealed a
signicant inuence of sex, with boys showing higher levels of
Aggressiveness and Victimization, and girls presenting more
Prosociality and Acceptance, which parallels prior national (Daz-
Aguado & Martnez-Arias, 2013; Lucas-Molina et al., 2011) and
international research (Olweus, 1993).
This study aimed to develop a tool for practical application
in schools in order to detect cases of bullying and to facilitate
intervention through the analysis of social context variables.
Therefore, in addition to the analysis of the previous variables, the
most relevant aspect is that it creates a social map of each classroom
on which are placed the individual proles associated with
bullying and sociometry. This allows us to carry out a qualitative
analysis of each case, which cannot be conducted only with scales
based on individual proles. Details of a case of victimization,
such as knowing the groups of friends in the classroom, the peer
groups support of the bullies, or whether the rejection of the
victim can be generalized to the entire classroom, are central
issues to understand the social dynamics. This information helps
us to intervene more effectively, for instance, by identifying the
appropriate classmates to participate in a peer-helper program or
by organizing the class groups from one academic year to the next
in order to reduce the level of bullying. The scale also permits the
interpretation of social status by showing rejection and popularity
as categorical instead of quantitative variables, as in other scales
(Daz-Aguado, 1986). Moreover, it uses more extensively adopted
correction procedures (Coie et al., 1982; Newcomb & Bukowski,
1983; Maassen et al., 1996) for the sociometric information, which
allows us to obtain a more complex categorization of social status
compared with previous scales (Cerezo, 2000), for example, the
controversial status. In addition, Sociescuela incorporates a rating
method that provides more comprehensive information about the
social afnity among all group members in comparison with items
based on nominations.
Nevertheless, some limitations should be noted for future
lines of research. First, despite the large number of participants,
the sample is not representative, which means that the ndings
cannot be generalized to the entire Spanish population. Another
weakness is the lack of items about cyberbullying, which
represents an important bullying subtype. The inclusion of this
information would allow us to compare bullying occurring in the
classroom with that occurring outside the classroom through new
technologies. It is necessary to study the stability of the measure
through a longitudinal study, and, owing to the relevance of the
social context, it would also be important to observe its inuence
on the stability through multilevel analysis.
As a general conclusion, the instrument presented herein can
be considered to be a useful tool for the detection of bullying,
and it could be a rst step in the development of instruments
that incorporate an analysis of the group structure in which this
phenomenon occurs. However, a study with a representative
sample should be carried out. The applicability and social utility
of this scale and software, which has been used in 348 schools, are
also noteworthy.
References
Achenbach, T.M., McConaughy S.H., & Howell, C.T. (1987). Child/
adolescent behavioral and emotional problems: Implications of
cross-informant correlations for situational specicity. Psychological
Bulletin, 101(2), 213-232.
Asher, S.R., & Dodge, K.A. (1986). Identifying children who are rejected
by their peers. Developmental Psychology, 22, 444-449.
Bjrkqvist, K., & sterman, K. (1995). PECOBE Peer Estimated Conict
Behavior: An inventory for the measurement of conict behaviour in
school children. Vasa, Finland: bo Akademi University.
Cerezo, F. (2000). Bull-S. Test de evaluacin de la agresividad entre
escolares [Bull-S. Test to assess aggressiveness among students].
Madrid: Albor-Cohs.
Coie, J.D., Dodge, K.A., & Coppotelli, H. (1982). Dimensions and types
of social status: A cross-age perspective. Developmental Psychology,
18, 557-570.
Defensor del Pueblo (2000). Informe sobre violencia escolar: el maltrato
entre iguales en la Educacin Secundaria Obligatoria [Report on school
violence: Bullying in compulsory secondary education]. Prepared by C.
del Barrio, E. Martn, I. Montero, L. Hierro, I. Fernndez, H. Gutirrez
and E. Ochata on behalf of the Spanish Committee of UNICEF.
Madrid: Publicaciones de la Ocina del Defensor del Pueblo.
Daz-Aguado, M.J. (1986). El papel de la interaccin entre iguales en la
adaptacin escolar y el desarrollo social [The role of peer interaction
in school adjustment and social development]. Madrid: C.I.D.E.
Daz-Aguado, M.J., & Martnez-Arias, R. (2013). Peer bullying and
disruption-coercion escalations in student-teacher relationship.
Psicothema, 25(2), 206-213.
Daz-Aguado, M.J., Martnez-Arias, R., & Martn-Babarro, J. (2010).
Estudio estatal de la convivencia escolar en la Educacin Secundaria
Obligatoria [State study of coexistence in compulsory secondary
education]. Madrid: MEPSYD.
Gifford-Smith, M.E., & Brownell, C.A. (2003). Childhood peer
relationships: Social acceptance, friendships, and social network.
Journal of School Psychology, 41, 235-284.
Assessment and detection of peer-bullying through analysis of the group context
363
Huttunen, A., Salmivalli, C., & Lagerspetz, K.M. (1996). Friendship
networks and bullying in schools. Annals of the New York Academy of
Sciences, 794(1), 355-359.
Jimerson, S.R., Swearer, S.M., & Espelage, D.L. (Eds.) (2009). Handbook
of bullying in schools: An international perspective. New York, NY:
Routledge.
Juvonen, J., Nishina, A., & Graham, S. (2001). Self-views versus peer
perceptions of victim status among early adolescents. In J. Juvonen & S.
Graham (Eds.), Peer harassment in school: A plight of the vulnerable
and victimized (pp. 105-124). New York, NY: Guilford Press.
Lorenzo-Seva, U., & Ferrando, P.J. (2006). FACTOR: A computer program
to t the Exploratory Factor Analysis model. Behavior Research
Methods, Instruments, & Computers, 38, 88-91.
Lucas-Molina, B., Pulido-Valero, R., & Solbes-Canales, I. (2011). Violencia
entre iguales en Educacin Primaria: el papel de los compaeros y
su relacin con el estatus sociomtrico [Peer violence in primary
education: The role of peers and its relationship with sociometric
status]. Psicothema, 23(2), 245-251.
Maassen, G., Akkermans, W., & van der Linden, J. (1996). Two-
dimensional sociometric status determination with rating scales.
Small Group Research, 27, 56-78.
Martn-Babarro, J. (2011). Sociometra y exclusin en la Enseanza
Secundaria Obligatoria [Sociometry and exclusion in secondary
compulsory education]. Unpublished doctoral thesis. Departamento
de Metodologa de las Ciencias del Comportamiento. Universidad
Complutense de Madrid.
Newcomb, A.F., & Bukowski, W. (1983). Social impact and social preference
as determinants of childrens peer group status. Developmental
Psychology, 19, 856-867.
Olweus, D. (1993). Bullying at school: What we know and what we can do.
Oxford, UK: Blackwell.
Parellada, M.J., San-Sebastin, J., & Martnez-Arias, R. (2009). Manual
Esperi de trastornos del comportamiento en nios y adolescentes
[Esperi Handbook of Behavior Disorders in Children and Adolescents].
Madrid: EOS.
Rivers, I., & Smith, P.K. (1994). Types of bullying behaviour and their
correlates. Aggressive Behavior, 20(5), 359-368.
Salmivalli, C., Lagerspetz, K., Bjrkqvist, K., sterman, K., & Kaukiainen,
A. (1996). Bullying as a group process: Participant roles and their
relations to social status within the group. Aggressive Behavior, 22,
1-15.
Salmivalli, C., & Peets, K. (2008). Bullies, victims, and bully-victim
relationships. In K. Rubin, W. Bukowski, & B. Laursen (Eds.),
Handbook of peer interactions, relationships, and groups (pp. 322-
340). New York, NY: Guilford Press.
Smith, P.K., & Shu, S. (2000). What good schools can do about bullying
ndings from a survey in English schools after a decade of research
and action. Childhood, 7(2), 193-212.
SPSS, Inc. (2011). IBM SPSS Statistics for Windows, Version 20.0. IBM
Corporation, Armonk, NY.
Sutton, J., & Smith, P.K. (1999). Social cognition and bullying: Social
inadequacy or skilled manipulation? British Journal of Developmental
Psychology, 17, 435-450.
Terry, R. (2000). Recent advances in measurement theory and the use
of sociometric techniques. In A.H.N. Cillessen & W.M. Bukowski
(Eds.), Recent advances in the measurement of acceptance and
rejection in the peer system (pp. 27-53). San Francisco, CA: Jossey-
Bass.
Valdivia-Peralta, M., Fonseca-Pedrero, E., Gonzlez-Bravo, L., & Lemos-
Girldez, S. (2014). Psychometric properties of the AQ Aggression
Scale in Chilean students. Psicothema, 26(1), 39-46.