Access to this full-text is provided by Springer Nature.
Content available from Journal of Youth and Adolescence
This content is subject to copyright. Terms and conditions apply.
Journal of Youth and Adolescence (2023) 52:406–418
https://doi.org/10.1007/s10964-022-01704-3
EMPIRICAL RESEARCH
Bullying and Cyberbullying Perpetration and Victimization:
Prospective Within-Person Associations
Antonio Camacho1●Kevin Runions2●Rosario Ortega-Ruiz1●Eva M. Romera 1
Received: 26 August 2022 / Accepted: 1 November 2022 / Published online: 17 November 2022
© The Author(s) 2022
Abstract
Bidirectional associations between bullying and cyberbullying have consistently identified during adolescence. However,
little is known about how this relationship works on the within-person level, after controlling for potential overlap at the
between-person level. This study examined the bidirectional longitudinal associations between bullying and cyberbullying
perpetration and victimization during 18-month period over four time points. A total of 2835 participants, aged 11 to 16
years in time 1 (50% girls; Mage =13.13, SD =1.06) were surveyed. Random intercept cross-lagged analyses revealed the
stability of bullying perpetration and victimization. Cyberbullying victimization predicted inversely bullying and
cyberbullying perpetration. The results indicate spirals of positive long-term associations between bullying (perpetration
and victimization) and cyberbullying perpetration but no long-terms spirals of victimization.
Keywords Within-person ●Longitudinal study ●Adolescents ●Bullying ●Cyberbullying
Introduction
Bullying has been defined as repeated and intentional
aggression by one or more individuals against the victim
who is unable to effectively defend himself/herself (Olweus,
1993; Volk et al., 2014). The new realities of information
and communication technologies have given rise to cyber-
bullying, intentional aggressive acts conducted through these
technologies (Kowalski et al., 2014). Researchers have
recognized the complexity in discriminating between the
two forms of violence in adolescents to the point that an
outstanding question in the literature is the extent of overlap
between bullying and cyberbullying. These questions of
overlap are important to understand the developmental
sequencing of bullying involvement - as perpetrator and as
target of bullying - over time. Questions of overlap also get
at issues of role continuity (e.g., are young people who
engage in cyberbullying more likely to become perpetrators
of offline bullying?) and role inversion (e.g., are targets of
bullying more likely to become perpetrators of bullying?).
Given the negative impact on psychosocial adjustment of
involvement over time in these phenomena there are some
research questions that need to be addressed. Is there stabi-
lity of involvement over time? Does a particular experience
provide an ‘entry point’into further bullying involvement?
Does being a target of bullying lead to perpetration? Does
traditional bullying perpetration tend to migrate to the online
setting, or does cyberbullying prepare students for more
direct acts of bullying? Definitive answers to these questions
have eluded researchers to date, and (largely unrecognized)
limitations of the common methodologies pose severe
challenges to those existing preliminary conclusions held by
researchers. Designed effective intervention for adolescents
involved in bullying requires a clear natural history of the
phenomenon of bullying involvement. This study deploys
recent developments in cross-lagged panel modeling to
provide a close examination of bullying perpetration and
victimization both online and off over time, enabling con-
clusions that do not conflate inter-individual change over
time (relative to one another) and intra-individual change
(relative to oneself). Given that bullying (about the first
years of middle school) and cyberbullying (about the last
years of middle school) peak during adolescence (Kowalski
et al., 2014; Pabian & Vandebosch, 2016), the present study
was a four-wave panel study among middle school adoles-
cents aged 11-16 which tends to peak in around middle
adolescence (Kowalski et al., 2014).
*Eva M. Romera
eva.romera@uco.es
1Universidad de Córdoba, Córdoba, Spain
2The University of Western Australia, Perth, Australia
1234567890();,:
1234567890();,:
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Stability, Role Continuity and Role Inversion in
Online and Offline Bullying Involvement
An overlap between bullying and cyberbullying has been
mainly established by correlational cross-sectional studies
(Baldry et al., 2017; Del Rey et al., 2012; Kowalski et al.,
2014; Modecki et al., 2014). A meta-analysis of long-
itudinal studies provided estimates of the association from
victimization to subsequent perpetration and perpetration to
subsequent victimization (Walters, 2021). But the estimates
from this meta-analysis are based on correlations analyses.
Researchers seeking to understand possible causal influ-
ences on behavior are often interested in cross-lagged path
modeling (CLPM), that estimates concurrent, autoregres-
sion and lagged associations. Based on extant research
using CLPM, autoregressive paths have shown that online
and offline perpetration and victimization are relatively
stable over time (Camacho et al., 2021; Giumetti et al.,
2022; Pabian & Vandebosch, 2016). These studies have
also provided estimates of significant lagged associations
between perpetration and victimization conducted online
and offline. Research on different aspects of bullying (e.g.,
bullying and cyberbullying; perpetration and victimization)
has developed two lines of hypothesizing: role continuity
and role inversion.
In the role continuity hypothesis, involvement in bully-
ing continue and extend it via cyberbullying (Baldry et al.,
2016). Based on research to date, adolescents involved in
bullying perpetration are more likely to engage in cyber-
bullying perpetration later (Chu et al., 2018; Giumetti et al.,
2022; Pabian & Vandebosch, 2016). Similarly, those
involved in cyberbullying perpetration have also been
found to have greater subsequent involvement in offline
perpetration (Pabian & Vandebosch, 2016). Similarly,
higher levels of bullying victimization predict subsequent
increased cyberbullying victimization (Giumetti et al.,
2022; Pabian & Vandebosch, 2016), and vice versa (Chu
et al., 2018; Pabian & Vandebosch, 2016).
Role inversion indicates a process in which those
involved in one role are more likely to become involved in
the other role: victimization may subsequently become
perpetrators, and vice versa (Falla et al., 2022; Lee et al.,
2021; Pabian & Vandebosch, 2016). Perpetrators of bully-
ing may unknowingly make enemies of people more pow-
erful than themselves, thereby becoming targets of bullying
(Malamut et al., 2022). Victims of bullying may come to
learn that perpetration is a path toward perceived popularity
(if not peer acceptance; Strindberg et al., 2020) and begin
bullying other less powerful them themselves. Others may
desire revenge but find retaliation ‘in kind’too difficult due
to imbalances in that modality of power (e.g., physical,
psychological, and social); consequently, the victims may
seek a modality where they are not weak. For example,
young people who experience face-to-face victimization
may use the Internet to bully others and take revenge on
those perceived as bullying them at school (Chu et al.,
2018; Runions et al., 2018). The Internet is potentially
significant for adolescents who perceive themselves as less
empowered, as power imbalance becomes less prominent in
the cyberspace which provides other factors such as anon-
ymity or the technical skills to react to damaging inter-
personal experiences in the offline context. An example
of this is the finding that adolescents with higher levels of
bullying perpetration are at increased subsequent risk of
cyberbullying victimization (Chu et al., 2018).
Addressing Methodological Limitations to
Differentiate Between- and Within-Person Change
To date, most studies examining these hypotheses have
used methodological strategies without a capacity to dif-
ferentiate between inter-individual change and intra-
individual change. Specifically, CLPM has important
methodological concerns as it cannot account for trait-like
(time-invariant) and state-like (time-variant) individual dif-
ferences (Hamaker et al., 2015). Data simulation studies
have shown that this limitation can result in inaccurate
models and thus mistaken conclusions regarding the exis-
tence of a lagged path from one variable to another or may
result in incorrect estimates of the direction of the causal
relationship (Hamaker et al., 2015). Thus, CLPM may
indicate a significant positive path from one variable at one
time to another variable at the subsequent time, but the real
relationship may in fact be negative. Obviously, inaccura-
cies of this scale will lead to the wrong conclusions
regarding causal processes. For an accurate insight into the
causal directionality of effects, a differentiation of the
between- and within-person level over time is needed.
Hamaker et al. (2015) have proposed the random intercept
cross-lagged panel model (RI-CLPM), which account for
the individual differences by distinguishing among
between- and within-person level. By incorporating random
intercepts, the RI-CLPM is able to avoid spurious findings
regarding the presence of any causal relationships, the
temporal priority (and hence likely causal priority) of dif-
ferent variables, and the direction / sign of the estimated
lagged relationship (Hamaker et al., 2015).
To date, few studies of bullying and cyberbullying have
addressed the influence between perpetration and victimi-
zation without confounding inter- and intra-individual. The
stability of traditional (offline) bullying victimization and
perpetration (autoregressive paths) has been reported con-
sistently positive in both CLPM and RI-CLPM (Davis et al.,
2022; Pabian & Vandebosch, 2016; Romera et al., 2021;
Zhu et al., 2022). However, the stability apparent in
cyberbullying perpetration and cyberbullying victimization
Journal of Youth and Adolescence (2023) 52:406–418 407
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
has been based on studies using traditional CLPM (Cama-
cho et al., 2021; Giumetti et al., 2022). Two studies using
RI-CLPM (Boer et al., 2021; Erreygers et al., 2018) have
not replicated this stability, which suggests that the evident
continuity of cyberbullying involvement may be a spurious
finding based on flawed treatment of variance. This finding
- that bullying involvement is relatively stable, whereas
cyberbullying involvement is more sporadic- may be based
on the specific characteristics of cyberbullying. In cyber-
bullying, the aggression not necessarily has to be repeated
over time by the same person (resending videos or images
by others is a way of repeating the aggressive action), there
is a higher probability to disengagement chances, and also
there is not a clear interdependence between the victim and
the aggressor (Huang et al., 2022).
Cross-lagged associations between perpetration and vic-
timization, and the possibility of role inversion, have also
been examined using RI-CLPM. Zhou et al. (2022) exam-
ined prospective reciprocal associations between perpetra-
tion and victimization at the within-person level in a sample
of children. Consistent with previous findings in CLPM,
they found that an increase from T3 in bullying victimiza-
tion compared with their average in the study predicted
higher involvement in bullying perpetration at time 4.
Moreover, bullying perpetration at time 4 also positively
predicted bullying victimization in time 5. These findings
suggest that once the between- and within-person variance
in victimization and perpetration are disaggregated, bidir-
ectional relationships exist. Examining cyberbullying over
three-waves six months apart, Erreygers et al. (2018) also
explored the bidirectional association between perpetration
and victimization for adolescents. In contrast to previous
findings using CLPM (Akgül & Artar, 2020; Camacho
et al., 2021), they did not find significant association
between cyberbullying victimization and subsequent chan-
ges in cyberbullying perpetration. The cross-lagged path
from cyber-perpetration to cyber-victimization was also not
significant. This provides important information on the
possible causal development sequencing of cyberbullying
involvement. To date, however, no studies have examined
the interrelationships of both online and offline bullying
perpetration and victimization, and thus testing the role
continuity and role inversion hypotheses.
Current Study
Although extensive evidence concerning the association
between traditional and cyberbullying perpetration and
victimization has been reported, almost all prior studies
have used statistical approaches that fail to account for
within-person variability (such as CLPM). The dynamic
processes of adolescent bullying and cyberbullying
involvement at the intra-individual level remain unclear. To
address this gap, the aim of the present study was to explore
the bidirectional longitudinal associations between offline
and online bullying perpetration and victimization using
RI-CLPM that accounts for within-person processes. Based
on prior research, it was expected that offline bullying
perpetration and victimization, and cyberbullying perpe-
tration and victimization would show positive between-
individual associations (Hypothesis 1), such that indivi-
duals who score higher than their peers on one at one time
will tend to score higher at subsequent points. At within-
individual stability, it was expected that involvement in
offline bullying perpetration and victimization would be
stable over time, but cyberbullying would not (Hypothesis
2). Based on the cross-lagged paths and the role continuity
hypothesis, it was predicted that online and offline bullying
perpetration would be reciprocally influenced over time
(Hypothesis 3) and bullying and cyberbullying victimiza-
tion would also show significant bidirectionality over time
(Hypothesis 4) (e.g., an increase in bullying perpetration at
one time relative to their average across the four time points
would be associated with higher levels at a later time in
cyberbullying perpetration and vice versa). Finally,
according with role inversion hypothesis, it was expected
that offline victimization and perpetration would be bidir-
ectionally positive associated over time, but not cyberbul-
lying perpetration and victimization (Hypothesis 5). It was
also expected that bullying victimization would predict
cyberbullying perpetration (Hypothesis 6) and bullying
perpetration would be associated with later cyberbullying
victimization (Hypothesis 7).
Methods
Participants and Procedure
The data were drawn from a larger longitudinal study
focusing on characteristics associated with the bullying and
cyberbullying, using a sample of adolescents in secondary
schools in Southern Spain. The study was approved by the
Ethical Committee of the institution of the Spanish authors.
To recruit participants, schools were invited to collaborate
and were informed of the purpose of the project. Once the
management team of each school agreed to participate,
parental consent was obtained (5% of parents did not con-
sent for their children to participate in the study). The
convenience sample comprised 2835 adolescents (50%
girls) between 11- and 16-years old attending Grades 7–9,
recruited in 13 middle schools. Data collection occurred
during school hours via trained psychologists with research
experience. Participants were provided the purpose of the
study and standardized information on the study and their
408 Journal of Youth and Adolescence (2023) 52:406–418
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
participation, emphasizing the voluntary, confidential, and
anonymous nature of the data collection. The procedure was
identical at each time point. Participants completed the
paper-and-pencil questionnaire in approximately 30 min.
Adolescents were assessed four times in 18 months (six
months after each time point). For the purpose of linking
surveys over time, participants were instructed to develop
their own personal code with the initial letters of their name
and date of birth, so that only they would know this iden-
tification and data anonymity would be guaranteed. In the
baseline, the mean age of students in November 2017
(Time 1; T1) was 13.10 years (SD =1.06). The socio-
demographic characteristics for each time point are shown
in Table 1. The average response rate was 88%. Sample
attrition was principally due to students absent from school
on the day of data collection or having moved schools. Due
to the longitudinal design of the study, a test was performed
to examine whether non-participation might be associated
with any of the study variables. Through logistic regression,
no significant differences were reported on the basis of
gender, age, cyberbullying and bullying (both perpetration
and victimization; at each time) predicting higher or lower
participation over time (all ps > 0.05).
Measures
Bullying
The European Bullying Intervention Project Questionnaire
(Ortega-Ruiz et al., 2016) was applied to measure bullying
victimization and perpetration. The questionnaire included
14 items, 7 for each subscale: victimization (e.g., “Someone
has said mean things about me to other people”) and per-
petration (e.g., “I have stolen or broken someone’s things”).
Before completing the questionnaire, participants were
provided the characteristics of bullying (intentionality,
power imbalance, repetition over time) to differentiate it
from other aggressive behaviors. The frequency of the
adolescents’behaviors was addressed with a five response
options as never (0), once or twice (1), once or twice a
month (2), about once a week (3), and more than once a
week (4), and continuous scores were used for analyses. For
each time, internal reliability using McDonald’s omega was
0.86, 0.86, 0.86 and 0.85 for victimization and 0.81, 0.82,
0.81 and 0.77 for perpetration, respectively. The con-
firmatory factor analysis showed good psychometric prop-
erties of the two-factor structure, as proposed in the original
study, with the current sample at T1: χ2=678.430, df =76,
p< 0.001; CFI =0.958, TLI =0.949, RMSEA =0.055,
90% CI [0.051, 0.059], SRMR =0.064.
Cyberbullying
The European Cyberbullying Intervention Project Ques-
tionnaire (Ortega-Ruiz et al., 2016) was applied to mea-
sure cyberbullying victimization and perpetration. The
questionnaire included 22 items, 11 for each subscale:
victimization (e.g., “Someone threatened me through texts
or online messages”) and perpetration (e.g., “Iposted
personal information about someone online”). Adoles-
cents were asked to the frequency of cyberbullying from 0
(never)to4(more than once a week). For each time,
McDonald’s omega was 0.87, 0.85, 0.88 and 0.89 for
victimization and 0.87, 0.89, 0.89 and 0.89 for perpetra-
tion, respectively. The confirmatory factor analysis
showed good psychometric properties of the two-factor
structure, as proposed in the original study, with the
current sample at T1: χ2=1085.311, df =288, p< 0.001;
Table 1 Sample distribution NTime 1 Time 2 Time 3 Time 4
November 2017 May 2018 November 2018 May 2019
n(p.r.) 2835 2657 (94%) 2515 (89%) 2461 (87%) 2357 (83%)
Gender
Girls 50% 50% 50% 51% 51%
Boys 50% 50% 50% 49% 49%
Age (SD) 13.10 (1.06) 13.60 (1.12) 14.01 (1.05) 14.54 (1.06)
School
Rural 52% 52% 52% 52% 53%
Urban 48% 48% 48% 48% 47%
Grade
1 35% 35% 3% 2%
2 34% 34% 36% 37%
3 31% 31% 32% 31%
4 29% 29%
p.r. participation rate
Journal of Youth and Adolescence (2023) 52:406–418 409
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
CFI =0.956, TLI =0.952, RMSEA =0.040, 90% CI
[0.038, 0.043], SRMR =0.067.
Covariates
Age and gender were collected as socio-demographic
information to address any differences in variables based
on these characteristics.
Data Analytic Strategy
Spearman correlations and the Intraclass Correlation
Coefficient were performed in preliminary analyses. To
analyze the prospective relationships between bullying and
cyberbullying by disaggregating the between- and within-
person variance a random intercept cross-lagged panel
model (RI-CLPM; Hamaker et al., 2015) was estimated
using Mplus 8.7 (Muthén & Muthén, 1998). The between-
person level (time-invariant characteristics) captures the
stability in each construct through the random intercept that
is estimated with the score across the four times. The
between-person level includes covariances between the
random intercepts of variables of the study. The variance at
the within-person level (time-variant characteristics) is
captured by the participants’time-to-time deviation from
the individual expected score. These analyses permit con-
clusions regarding whether previous increases or decreases
of face-to-face and online victimization and perpetration
from their own average level across the four time points are
associated with subsequent changes in bullying and
cyberbullying involvement. The within-person level
includes autoregressive paths (e.g., cyberbullying victimi-
zation at T1 on cyberbullying victimization at T2), cross-
lagged paths (e.g., bullying victimization at T1 on cyber-
bullying perpetration at T2), covariances between variables
in T1 (e.g., cyberbullying perpetration at T1 with bullying
perpetration at Time 1), and the residual covariances of the
variables at T2, T3 and T4. Gender and age were included
in the model as time-invariant predictor of observed vari-
ables, as boys have been found to have a higher prevalence
of involvement in perpetration (Smith et al., 2019), while
middle adolescents tend to have higher involvement in
cyberbullying (Camacho et al., 2021).
The data were tested for missingness. Little’s MCAR
(Little, 1988) test was significant (p< 0.001) indicating that
the data were not missing completely at random (MCAR).
Based on the low normed chi-square (χ2/df =1.79) the data
were deemed to be missing at random (MAR) (Bollen,
1989). Therefore, missing data were addressed with full
information maximum likelihood (FIML). Maximum like-
lihood estimation with robust standard errors (MLR) was
used to address the non-normally distributed nature of the
variables. For optimal standard model fit indices,
comparative fit index (CFI) should be above 0.90, and root
mean square error of approximation (RMSEA) should not
exceed 0.08 (Hu & Bentler, 1999). Because the adolescents
were placed in schools, the command “type =complex”
was used to handle the clustering effects on standard errors.
The associations between the study variables were ana-
lyzed through a set of models with the aim to choose the most
parsimonious model whose change in model fitwasnotsig-
nificant (Kline, 2015). First, an unconstrained model (model
1) where the components of the RI-CLPM were freely esti-
mated. Then, a stepwise series of constraints were added to
match the paths over time: autoregressive paths (model 2),
cross-lagged paths (model 3) and correlated changes within-
time (model 4). Significant differences between model fit
comparisons were considered when two of the following
criteria were attained: chi-square difference test at p< 0.05
(Satorra & Bentler, 2001), ΔCFI ≥0.01 and ΔRMSEA ≥0.015
(Chen, 2007). In the lack of differences between the nested
models, the model with the most constraints was retained.
Results
Preliminary Analyses
Descriptive statistics and Spearman’s bivariate correlations
among study variables are reported in Table 2. The cross-
sectional and longitudinal association between bullying and
cyberbullying perpetration and victimization were low-
moderate positive.
In relation to the Intraclass Correlation Coefficient, the
within-person level variance in each measure was higher
than 10% (58% for bullying perpetration; 49% for bullying
victimization; 63% for cyberbullying perpetration; 54% for
cyberbullying victimization). This warrants use of RI-
CLPM to disaggregate the between- and within-person
variance (Hamaker et al., 2015).
Random-Intercept Cross-Lagged Panel Model
Steps were followed to choose the most parsimonious
model before analyzing the prospective relationships
between the study variables (see Table 3). In model 1, paths
were allowed to vary over time. This unconstrained model
showed an excellent model fit. In model 2, the auto-
regressive paths were constrained to be equivalent over
time. This model was not significantly different compared
to the unconstrained model since at least two of the model
fit comparative criteria were not violated (see Table 3). In
model 3, imposing further constraints on cross-lagged paths
did not result in significant differences compared to model
2. For model 4 within-time correlated changes were con-
straining, with no significant differences in model fit
410 Journal of Youth and Adolescence (2023) 52:406–418
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
compared to model 3. Thus model 4 was retained as the
most parsimonious to explore the prospective association
between bullying and cyberbullying.
The results of the RI-CLPM were reported in STDYX
standardized estimates (see Fig. 1). At the between-person
level, the significant positive covariances between random
intercepts of bullying and cyberbullying perpetration and
victimization suggest that adolescents who reported higher
levels of one variable at the four time points also reported
higher levels in other variables compared to other adoles-
cents (Hypothesis 1). At the within-person level, significant
and positive autoregressive paths for both bullying perpe-
tration and victimization indicate that adolescents who
report a higher-than-expected score are likely to report a
higher-than-expected score on subsequent times (Hypoth-
esis 2). Notably, the autoregressive paths of cyberbullying
perpetration and victimization both were not significant (see
Table 4; Hypothesis 2). In relation to cross-lagged paths,
positive cross-lagged reciprocal effects were found across
time between bullying and cyberbullying perpetration
(Hypothesis 3). Offline bullying victimization predicted
subsequent cyberbullying victimization (Hypothesis 4),
providing support for the role continuity hypothesis. With
regard to the role inversion hypotheses (Hypotheses 5-7),
significant negative paths from online victimization to both
bullying and cyberbullying perpetration were discovered
(Hypotheses 5 and 7), while face-to-face victimization
predicted subsequent increases in cyberbullying perpetra-
tion (Hypothesis 6). Regarding the association between
variables within-time, the four study variables were posi-
tively associated (see Table 4).
Based on time-invariant predictors, boys show higher
involvement than girls in both bullying and cyberbullying
perpetration, as in online and face-to-face victimization at T1
Table 2 Spearman’s bivariate correlations and descriptive statistics of study variables
12345678910111213141516
1. Bullying perpetration T1 –
2. Bullying perpetration T2 0.51 –
3. Bullying perpetration T3 0.42 0.47 –
4. Bullying perpetration T4 0.40 0.45 0.50 –
5. Bullying victimization T1 0.55 0.33 0.28 0.29 –
6. Bullying victimization T2 0.31 0.53 0.29 0.29 0.52 –
7. Bullying victimization T3 0.28 0.32 0.55 0.35 0.42 0.49 –
8. Bullying victimization T4 0.25 0.29 0.33 0.56 0.40 0.45 0.54 –
9. Cyberbullying perpetration T1 0.47 0.37 0.31 0.33 0.34 0.22 0.21 0.21 –
10. Cyberbullying perpetration T2 0.38 0.52 0.35 0.34 0.22 0.30 0.21 0.18 0.44 –
11. Cyberbullying perpetration T3 0.31 0.39 0.50 0.38 0.18 0.21 0.31 0.22 0.43 0.47 –
12. Cyberbullying perpetration T4 0.32 0.33 0.39 0.50 0.18 0.16 0.25 0.29 0.36 0.40 0.48 –
13. Cyberbullying victimization T1 0.42 0.30 0.27 0.29 0.53 0.36 0.34 0.32 0.59 0.36 0.32 0.29 –
14. Cyberbullying victimization T2 0.29 0.41 0.28 0.28 0.33 0.46 0.33 0.33 0.34 0.57 0.39 0.34 0.48 –
15. Cyberbullying victimization T3 0.26 0.33 0.38 0.33 0.31 0.36 0.49 0.37 0.34 0.37 0.60 0.37 0.44 0.54 –
16. Cyberbullying victimization T4 0.25 0.28 0.31 0.43 0.31 0.32 0.37 0.50 0.32 0.31 0.37 0.60 0.42 0.45 0.50 –
M0.25 0.28 0.20 0.21 0.55 0.56 0.40 0.42 0.14 0.14 0.12 0.12 0.23 0.21 0.20 0.20
SD 0.42 0.45 0.38 0.36 0.70 0.67 0.58 0.56 0.34 0.36 0.32 0.31 0.43 0.37 0.39 0.38
Range 0–40–40–40–40–40–40–40–40–40–40–40–40–40–40–40–4
All correlations were significant at p< 0.001
Table 3 Model fit statistics of
random intercept cross-lagged
panel model
Model Model fit Model fit comparison
χ2df CFI RMSEA [90% CI] Δχ2(Δdf) ΔCFI ΔRMSEA
Model 1 48.320 38 0.998 0.010 [0.000, 0.018] –––
Model 2 80.170** 46 0.994 0.017 [0.010, 0.023] 24.192 (8)** 0.004 0.007
Model 3 120.834*** 70 0.991 0.017 [0.011, 0.021] 40.605 (24)* 0.003 0.000
Model 4 126.449** 82 0.992 0.014 [0.009, 0.019] 14.496 (12) 0.001 0.003
CFI comparative fit index, CI confidence interval, RMSEA root mean square error of approximation
*p< 0.05; **p< 0.01; ***p< 0.001
Journal of Youth and Adolescence (2023) 52:406–418 411
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
(see Table 5). Older participants in the study showed higher
involvement than younger participants on bullying perpetration
(at T1 and T2), cyberbullying perpetration (from T1 to T3) and
cyberbullying victimization (from T1 to T4) (see Table 5).
Discussion
Research has suggested the role continuity and inversion
hypotheses about the phenomena of bullying and cyberbullying
on the involvement of adolescents in perpetration and victi-
mization. However, the bidirectional influence association has
mainly focused on approaches that may be misleading because
associations within- and between-person were not dis-
aggregated. This study used cutting-edge within-person statis-
tics to correct for limitations with traditional modeling that may
have led to spurious findings. The present research captured
intra-individual longitudinal fluctuations between bullying
perpetration, bullying victimization, cyberbullying perpetration,
and cyberbullying victimization over time after controlling for
time-invariant associations at the between-person level by
using RI-CLPM.
The overlap between bullying perpetration, bullying vic-
timization, cyberbullying perpetration and cyberbullying vic-
timization were reported by the positive associations at a
stable between-person level (Hypothesis 1). These moderate/
strong correlations of random intercepts indicated an overall
time invariance of adolescent bullying involvement of ado-
lescents. Consistent with previous cross-sectional research
(Baldry et al., 2017;DelReyetal.,2012), higher involvement
in one aspect of bullying across the four time points (e.g.,
offline victimization) was associated with high involvement in
the other aspects (e.g., cyberbullying perpetration) overall.
Associations strong overall (ranging from rof 0.68–0.82)
except for the association of bullying victimization and
cyberbullying perpetration, which had a notably more modest
association. However, these findings do not clarify which
features are antecedent or consequence.
Stability of Bullying and Cyberbullying
Once the stable differences between individuals have been
controlled for, a more accurate understanding may be pro-
vided on influence of time-variant variables and likely causal
processes involved. In support of Hypothesis 2 and consistent
with previous research (Boer et al., 2021; Cogo-Moreira et al.,
2021; Davis et al., 2022; Romera et al., 2022), involvement in
perpetration and victimization was stable for offline bullying,
but not in cyberbullying. Such findings for offline bullying are
consistent with those studies even when CLPM was
employed (and thus without separating between- and within-
person variance; Chu et al., 2018; Pabian & Vandebosch,
2016). Using RI-CLPM, the findings of the present study
indicate that there is no direct stability in cyberbullying
involvement over time (Boer et al., 2021; Erreygers et al.,
2018), in contrast with previous research using CLPM
(Camacho et al., 2021;Giumettietal.,2022). The stability
over time of victimization and perpetration may be supported
by the hierarchical nature of offline bullying as a group pro-
cess. However, the involvement of adolescents in cyberbul-
lying tends to have a more sporadic and less sustainable
character over time (Huang et al., 2022;Smaheletal.,2020).
The role of group dynamics in the relative stability of offline
bullying remains to tested more directly but may speak to a
more complex developmental phenomenon than current
models of role inversion and continuity (see below).
.76***
.38***
.62***
Bullying
perpetration
T1
Bullying
perpetration
T2
Bullying
victimization
T1
Bullying
victimization
T2
Cyberbullying
perpetration
T1
Cyberbullying
perpetration
T2
Cyberbullying
victimization
T1
Cyberbullying
victimization
T2
Bullying
perpetration
Bullying
victimization
Cyberbullying
perpetration
Cyberbullying
victimization
Bullying
perpetration
T3
Bullying
victimization
T3
Cyberbullying
perpetration
T3
Cyberbullying
victimization
T3
.19**
.22***
.06*
.13*
.08**
.20***
.24***
.07*
.14**
.10**
.18**
.23***
.07*
-.12***-.11**-.13**
-.10*-.10**-.11**
.12*
.11**
Bullying
perpetration
T4
Bullying
victimization
T4
Cyberbullying
perpetration
T4
Cyberbullying
victimization
T4
Between-person level Within-person level
.68***
.67***
.82***
.08**.10**.09**
Fig. 1 Random intercept cross-lagged panel model. The within-time covariances and the non-significant autoregressive and cross-lagged paths at
the within-person level are not illustrated for simplicity. These associations are reported in Table 4.*p< 0.05; **p< 0.01; ***p< 0.001
412 Journal of Youth and Adolescence (2023) 52:406–418
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Continuity of Perpetration and Victimization across
Bullying and Cyberbullying
Although cyberbullying perpetration was not stable per se in
theseanalyses,therewasastable pattern of bidirectionality
between online and offline bullying at the within-person level,
whereinateachwave,cyberbullyingwaspredictedbysub-
sequent offline bullying and went on to predict the later
increases in offline bullying perpetration, as per Hypothesis 3.
This provides a more conservative replication of previous
studies that used CLPM (Chu et al., 2018;Giumettietal.,
2022; Pabian & Vandebosch, 2016). According to the role
continuity hypothesis (Baldry et al., 2016), adolescents are
expected to adopt the same patterns across phenomena.
The co-construction theory has been applied to account for
how adolescents construct their digital social interactions
in a similar way to their non-digital environment (Sub-
rahmanyam et al., 2006). The online phenomenon may spread
face-to-face perpetration by attacking others on the Internet to
further increase the damage. Regarding the opposite influence,
after online perpetration, adolescents may tend to endorse
surrounding states or normative beliefs that their later invol-
vement in bullying perpetration is an acceptable behavior
(Wright & Li, 2013).
Bullying and cyberbullying victimization were also
expected to be bidirectionally associated (Hypothesis 4). In
Table 4 Standardized
coefficients of within-person
level results in RI-CLPM
Covariances Time 1 Time 2 Time 3 Time 4
Bullying perpetration ↔Bullying victimization 0.41*** 0.41*** 0.52*** 0.55***
Bullying perpetration ↔Cyberbullying perpetration 0.46*** 0.42*** 0.52*** 0.52***
Bullying perpetration ↔Cyberbullying victimization 0.36*** 0.33*** 0.36*** 0.35***
Bullying victimization ↔Cyberbullying perpetration 0.31*** 0.25*** 0.30*** 0.30***
Bullying victimization ↔Cyberbullying victimization 0.52*** 0.35*** 0.39*** 0.37***
Cyberbullying perpetration ↔Cyberbullying
victimization
0.63*** 0.67*** 0.71*** 0.65***
Autoregressive path T1 →T2
ß(SE)
T2 →T3 ß(SE) T3 →T4
ß(SE)
Bullying perpetration →Bullying perpetration 0.18** (0.06) 0.20*** (0.05) 0.19** (0.06)
Bullying victimization →Bullying victimization 0.23*** (0.04) 0.24*** (0.04) 0.22 (0.04)
Cyberbullying perpetration →Cyberbullying
perpetration
0.08 (0.06) 0.08 (0.07) 0.08 (0.07)
Cyberbullying victimization →Cyberbullying
victimization
−0.01 (0.08) −0.01 (0.06) −0.01 (0.06)
Cross-lagged path T1 →T2
ß(SE)
T2 →T3 ß(SE) T3 →T4
ß(SE)
Bullying perpetration →Bullying victimization 0.00 (04) 0.00 (04) 0.00 (04)
Bullying perpetration →Cyberbullying perpetration 0.09** (0.03) 0.10** (0.03) 0.08** (0.03)
Bullying perpetration →Cyberbullying
victimization
0.02 (0.04) 0.02 (0.04) 0.02 (0.03)
Bullying victimization →Bullying perpetration 0.00 (04) 0.00 (04) 0.00 (04)
Bullying victimization →Cyberbullying
perpetration
0.07* (0.04) 0.07* (0.04) 0.06* (0.03)
Bullying victimization →Cyberbullying
victimization
0.11** (0.04) 0.10** (0.03) 0.08** (0.03)
Cyberbullying perpetration →Bullying perpetration 0.12* (0.05) 0.14** (0.05) 0.13* (0.05)
Cyberbullying perpetration →Bullying
victimization
0.04 (0.04) 0.05 (0.05) 0.05 (0.04)
Cyberbullying perpetration →Cyberbullying
victimization
−0.07 (0.07) −0.07 (0.07) −0.06 (0.06)
Cyberbullying victimization →Bullying
perpetration
−0.13** (0.04) −0.11** (0.04) −0.12** (0.04)
Cyberbullying victimization →Bullying
victimization
−0.02 (0.05) −0.02 (0.04) −0.02 (0.05)
Cyberbullying victimization →Cyberbullying
perpetration
−0.11** (0.05) −0.10** (0.04) −0.10* (0.04)
*p< 0.05; **p< 0.01; ***p< 0.001
Journal of Youth and Adolescence (2023) 52:406–418 413
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
line with previous research (Giumetti et al., 2022;Pabian&
Vandebosch, 2016), offline bullying victimization predicted
cyberbullying victimization, suggesting that victimization
may begin in person but migrate to online settings. Among
the possible underlying assumptions for this continuity in
victimization are the spread of offline interpersonal relation-
ships through the Internet and the victim schema model.As
the Internet during these ages is a setting where peers extend
their face-to-face interpersonal relationships, the perpetrator
may spread online to enhance the potential for harming the
victim (Wright & Li, 2013). In addition, victim schema
models (Rosen et al., 2007) may also contribute to role con-
tinuity, as experience of being a target in one setting may
generate negative cognitive biases and maladaptive coping in
peer relationships (Camacho et al., 2022), that lead to greater
perceptions of threat, mistrust and increased likelihood of
victimization both offline and virtual phenomena (Chu et al.,
2018; Rodríguez-de Arriba et al., 2022). However, unlike
previous studies in CLPM (Chu et al., 2018;Pabian&
Vandebosch, 2016), when differences in stable victimization
at between-person level were controlled for, an increase in
online victimization within-person level did not predict a
higher involvement in face-to-face victimization. Previous
studies on internalizing problems arising from victimization
have also found differences between online and face-to-face
victimization. While studies have reported that both bullying
and cyberbullying victimization are subsequently associated
with greater depressive symptoms (Fredrick et al., 2022;He
et al., 2022), such associations are consistent when analyzed
at the within-person level in bullying (Li et al., 2021), but not
in cyberbullying (Boer et al., 2021). Such divergences could
be explained by the role of the online context for adolescents.
In cyberbullying, any person may be exposed to be a victim
even those with greater resilience and adaptative coping
strategies that, in contrast, can be effective to avoid face-to-
face victimization.
Inversion of Victimization and Perpetration across
Bullying and Cyberbullying
According to the role inversion hypothesis (Mishna et al.,
2012), those who become involved in one role (target or
perpetrator of bullying) are more likely to take on the other
role later. Counter to the fifth hypothesis and in contrast to
previous research (Zhou et al., 2022), changes in offline
victimization yielded no direct deviations in offline bullying
perpetration at any wave. However, there was evidence of a
more indirect path whereby being targeted offline was
positively associated with subsequent cyberbullying perpe-
tration (providing support for Hypothesis 6 and consonant
with Chu et al., (2018)), which in turn was positive asso-
ciated with face-to-face perpetration. This suggests that
targeted adolescents may use the more covert cyber setting
as an incubator for bullying perpetration and then move on
–perhaps with growing confidence –to engage in overt
face-to-face bullying. As Ybarra and Mitchell (2004) have
conjectured, for victims of offline bullying, the Internet may
provide a setting for dominating others as compensation for
their own social position. Such a contrast could be based on
the helplessness of the victim in the face-to-face situation.
The power imbalance provides the perpetrator a safe posi-
tion, as the victim may not have enough physical and
psychological strengths to overcome the social gap and
address the perpetrator with revenge. However, the power
imbalance differs in the online phenomenon due to anon-
ymity and disinhibition. The difficulty to identify the per-
petrator and the avoidance of retaliation may develop into
the perception of a lack of responsibility and deindividua-
lization of the behavior. This may provide the victim of
bullying with the courage to engage in cyberbullying per-
petration later as a means of revenge (Runions et al., 2018)
or as a bid to obtain power or social status, which they may
have attributed to their own bullies in their past experiences.
In the present study, it was also expected that bullying
perpetration would be associated with subsequent cyberbully-
ing victimization (Hypothesis 7). In contrast with two previous
studies (Lee et al., 2021; Pabian & Vandebosch, 2016), face-to-
face perpetration did not predict online victimization, while an
increase in cyberbullying victimization predicted decreases in
cyber- and offline bullying perpetration, suggesting that
experiences of cyber-victimization may protect against future
Table 5 Standardized coefficients of the time-invariant predictors in
RI-CLPM
GenderaAge
ßSE ßSE
Bullying perpetration T1 −0.13*** 0.02 0.04* 0.03
Bullying perpetration T2 −0.12*** 0.03 0.04** 0.01
Bullying perpetration T3 −0.12*** 0.03 0.01 0.01
Bullying perpetration T4 −0.10*** 0.02 0.01 0.01
Bullying victimization T1 −0.11** 0.03 0.01 0.02
Bullying victimization T2 −0.03 0.04 −0.04 0.02
Bullying victimization T3 −0.03 0.02 −0.03 0.03
Bullying victimization T4 −0.03 0.02 −0.03 0.02
Cyberbullying perpetration T1 −0.08*** 0.02 0.05*** 0.01
Cyberbullying perpetration T2 −0.06** 0.02 0.11*** 0.02
Cyberbullying perpetration T3 −0.05* 0.02 0.10*** 0.02
Cyberbullying perpetration T4 −0.05** 0.02 0.04 0.03
Cyberbullying victimization T1 −0.05* 0.02 0.05*** 0.01
Cyberbullying victimization T2 −0.01 0.02 0.04*** 0.01
Cyberbullying victimization T3 −0.02 0.02 0.04*** 0.01
Cyberbullying victimization T4 −0.02 0.02 0.02* 0.01
*p< 0.05; **p< 0.01; ***p< 0.001
aGender was coded as: 0 =boy and 1 =girl
414 Journal of Youth and Adolescence (2023) 52:406–418
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
bullying perpetration both offline and online at the individual
level. This finding, in contrast to the significant positive path-
ways from offline victimization to cyberbullying perpetration,
presents a paradox. Once between-person variance is accounted
for, why would an increase in face-to-face victimization lead to
an increase in the involvement in online perpetration, whereas
online victimization leads to a subsequent decrease in perpe-
tration, online and off?
Compared to online victimization, offline victimization is
less strongly associated with increased internalizing problems
(e.g., fear, humiliation, anxiety or depression; Dennehy et al.,
2020). Being cyberbullied, with the constant access to the
victim (24 h a day, 7 days a week) and the permanency of the
evidence of bullying existing on the Internet (see Runions
et al., 2013 for a review) could generate in the victim a feeling
of chronic vulnerability, helplessness, and powerlessness to
address the revenge through perpetration. These processes
may lead individuals to become aware of the harm caused by
perpetration, as well as experiencing the emotions of others,
resulting in a decreased likelihood of subsequent online
aggression towards others (see meta-analysis, Kowalski et al.,
2014). An adolescent who experiences victimization for a
limited period may develop some sensitivity to the aggressive
behaviors of bullying, knowing personally the associated
psychological consequences.
An alternate possibility is that targets of cyberbullying do
not see or project the same accrual of social status (e.g.,
social impact or perceived popularity; Guy et al., 2019) to
the perpetrator as do targets of face-to-face bullying, where
the group dynamics are more evident, and where popularity
may be attributed to one’s tormentor. The perception that
bullying leads to popularity may drive those who are pre-
dominantly bullied offline in a way that diverges from the
experience of cyberbullied adolescents.
Future research might address this issue through models
that examine whether internalizing symptoms or coping
strategies with victimization may play a mediating role in
the effects of cyberbullying victimization and bullying
victimization on perpetration at within-person level.
Limitations and Practical Implications
Though the study presents certain strengths (large sample,
longitudinal data, and within- and between-person level), the
findings bear limitations to be addressed in future research.
First, given that perpetration and victimization were both
treated as discrete variables, the co-occurrence (bully-victim
status) could not be modeled. This is unfortunate, as dual
involvement in both roles is an important phenomenon that this
work is concerned with. Second, only self-report measures
were used, which may have increased shared method variance
and social desirability bias. This bias could be addressed in
future research by using peer or teacher nomination of victims
and perpetrators of traditional bullying. Although the repeated-
measures time intervals (at the beginning and end of the school
year across two school years) provide insight into potential
effects over time, given the dynamic social relationships during
adolescence, a shorter time frame might capture changes to a
better extent, especially in cyberbullying. The sample (between
11 and 16 years at T1; 13-18 at T4) does not allow extra-
polation to the whole period of adolescence, nor does it capture
specific key developmental periods that may influence the
association between variables (e.g., transition from primary to
secondary school, pubertal maturational stage, development of
romantic relationships). Furthermore, future studies could also
consider controlling for factors that could influence the invol-
vement of bullying and cyberbullying (e.g., ethnicity, socio-
economic status, internet use, parental styles, internalizing
symptoms, coping strategies). The sample derives from a
specific context in the south of Spain with a majority Caucasian
population. Future studies could consider more demo-
graphically diverse regions with cross-cultural designs to fur-
ther enhance the relevance of the findings.
Beyond the limitations, the study provides support for
psychoeducational strategies to address bullying and
cyberbullying. Anti-bullying programs aimed at reducing
perpetration and victimization rates in face-to-face phe-
nomenon should target the involvement of schoolchildren in
cyberbullying (Casas et al., 2018). It is also important to
raise awareness that adolescents who are targets of bullying
may come to the erroneous belief that perpetration of bul-
lying is in their best interests. Special emphasis should be
focused on adolescents who experience offline bullying
since such experience could be extended online, and
problem-focused coping strategies to deal with the situation
and with attendant emotional problems (e.g., anxiety and
depression) may be of value. Such interventions should also
address issues of moral sensitivity to adhere to moral
standards and decrease the selective deactivation of the
moral self-regulation process as a means to avoiding the
normalization of violence, which may fuel vicious circles of
perpetration and victimization (Romera et al., 2021).
Conclusions
The association between bullying and cyberbullying
(victimization and perpetration) have been accounted
from the role continuity and inversion hypotheses.Froma
longitudinal approach, the role continuity hypothesis
addresses how previous involvement in bullying may
subsequently extend to cyberbullying phenomena.
Whereas the role inversion hypothesis addresses how the
involvement of adolescents in one phenomenon may lead
to the involvement of other behaviors later (first victimi-
zation and then perpetration or vice versa). The present
Journal of Youth and Adolescence (2023) 52:406–418 415
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
study provides a notable contribution to the literature by
addressing these hypotheses longitudinally using new
statistical approaches that disentangle between- and
within-individual variance over time. At the between-
person level, stable differences across individuals have
highlighted the well-known overlap between bullying and
cyberbullying victimization and perpetration. At the
within-person level, the longitudinal influence between
variables has been established by considering the indivi-
dual and its temporal evolution. Through the state-like
characteristics, the present study considers the influence
between the variables considering that the involvement of
many adolescents in the phenomena may be spontaneous
and connected to a particular situation, which is in line
with previous descriptive studies on bullying and cyber-
bullying. In sum, three conclusions are noteworthy. First,
being involved in bullying, as victim or bully, is posi-
tively long-term associated to cyberbullying perpetration.
Second, there is no long-terms spirals of victimization off
and online. Third, having experiences of cybervictimiza-
tion may predict not being a bully and cyberbully. Taken
together, the findings highlight the importance of con-
sidering both aggressive behaviors (off y online) to
understand and prevent students’involvement. Effective
intervention depends upon a thorough causal under-
standing of how involvement in bullying arises and
evolved over time; this study provides an important step
in explicating these processes. Knowledge of interactive
play between the diverse manifestations is an important
contribution to the shaping of preventive and intervention
strategies adapted to the social reality of adolescents in an
online and offline world that is interconnected in many of
its mechanisms.
Authors’Contributions A.C.: conceived of the manuscript, drafted
the manuscript, performed statistical analyses and participated in the
interpretation of the data; K.R.: participated in the interpretation of the
data, drafted the manuscript and revised the manuscript; R.O.R.:
conceived of the study, drafted the manuscript and revised the
manuscript; E.M.R.: conceived of the study, coordinated the data
collection, conceived of the manuscript, coordinated, and drafted the
manuscript, and participated in the interpretation of the data. All
authors read and approved the final manuscript.
Funding This study was supported by the Spanish National Research
Agency (PDC2021-121741-I00/AEI/10.13039/501100011033), Eur-
opean Union—NextGenerationEU, and by the University of Córdoba
in the Plan Propio de Investigación 2022.
Data Sharing and Declaration The datasets generated and/or analyzed
during the current study are not publicly available but are available
from the corresponding author on reasonable request.
Compliance with Ethical Standards
Conflict of Interest The authors declare no competing interests.
Ethical Approval All procedures performed in studies involving
human participants were in accordance with the ethical standards of
the 1964 Helsinki declaration and its later amendments or comparable
ethical standards. The project was reviewed and approved by Biosafety
and Bioethics Committee of the University of Cordoba.
Informed Consent Written informed consent was obtained from the
parents.
Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as
long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license, and indicate if
changes were made. The images or other third party material in this
article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not
included in the article’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted
use, you will need to obtain permission directly from the copyright
holder. To view a copy of this license, visit http://creativecommons.
org/licenses/by/4.0/.
References
Akgül, G., & Artar, M. (2020). Cyberbullying: Relationship with
developmental variables and cyber victimization. Scandinavian
Journal of Child and Adolescent Psychiatry and Psychology,8,
25–37. https://doi.org/10.21307/sjcapp-2020-004
Baldry, A. C., Farrington, D. P., & Sorrentino, A. (2016). Cyber-
bullying in youth: A pattern of disruptive behaviour. Psico-
logia Educativa,22(1), 19–26. https://doi.org/10.1016/j.pse.
2016.02.001
Baldry, A. C., Farrington, D. P., & Sorrentino, A. (2017). School
bullying and cyberbullying among boys and girls: Roles and
overlap. Journal of Aggression, Maltreatment and Trauma,
26(9), 937–951. https://doi.org/10.1080/10926771.2017.1330793
Boer, M., Stevens, G. W. J. M., Finkenauer, C., de Looze, M. E., &
van den Eijnden, R. J. J. M. (2021). Social media use intensity,
social media use problems, and mental health among adolescents:
Investigating directionality and mediating processes. Computers
in Human Behavior,116, 106645. https://doi.org/10.1016/j.chb.
2020.106645
Bollen, K. A. (1989). Structural equations with latent variables. Wiley
Camacho, A., Ortega-Ruiz, R., & Romera, E. M. (2021). Longitudinal
associations between cybervictimization, anger rumination and
cyberaggression. Aggressive Behavior,47(3), 332–342. https://
doi.org/10.1002/ab.21958
Camacho, A., Ortega-Ruiz, R., & Romera, E. M. (2022). Adoles-
cents’social anxiety dynamics in a latent transition analysis and
its psychosocial effects. International Journal of Clinical and
Health Psychology,22(3), 100311. https://doi.org/10.1016/j.
ijchp.2022.100311
Casas, J. A., Del Rey, R., & Ortega-Ruiz, R. (2018). The ConRed
program: Educating in cybercoexistence and cyberbullying
prevention by improving coexistence projects in schools. In M.
Campbell & S. Bauman (Eds.), Reducing cyberbullying in
schools: International evidence-based best practices (pp.
203–211). Academic Press. https://doi.org/10.1016/B978-0-12-
811423-0.00015-8
416 Journal of Youth and Adolescence (2023) 52:406–418
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Chen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of
measurement invariance. Structural Equation Modeling,14(3),
464–504. https://doi.org/10.1080/10705510701301834
Chu, X. W., Fan, C. Y., Liu, Q. Q., & Zhou, Z. K. (2018). Stability
and change of bullying roles in the traditional and virtual con-
texts: A three-wave longitudinal study in Chinese early adoles-
cents. Journal of Youth and Adolescence,47(11), 2384–2400.
https://doi.org/10.1007/s10964-018-0908-4
Cogo-Moreira, H., Gusmões, J. D., Valente, J. Y., Eid, M., & Sanchez,
Z. M. (2021). Does #Tamojunto alter the dynamic between drug
use and school violence among youth? Secondary analysis from a
large cluster-randomized trial. European Child and Adolescent
Psychiatry.https://doi.org/10.1007/s00787-021-01863-x
Davis, A. C., Farrell, A. H., Brittain, H., Krygsman, A., Arnocky, S.,
& Vaillancourt, T. (2022). The dark triad and bullying in ado-
lescence: A three-wave random intercept cross-lagged panel
analysis. Journal of Research in Personality,96, 104178. https://
doi.org/10.1016/j.jrp.2021.104178
Del Rey, R., Elipe, P., & Ortega-Ruiz, R. (2012). Bullying and
cyberbullying: Overlapping and predictive value of the co-
occurrence. Psicothema,24(4), 608–613
Dennehy, R., Meaney, S., Walsh, K. A., Sinnott, C., Cronin, M., &
Arensman, E. (2020). Young people’s conceptualizations of the
nature of cyberbullying: A systematic review and synthesis of
qualitative research. Aggression and Violent Behavior,51,
101379. https://doi.org/10.1016/j.avb.2020.101379
Erreygers, S., Vandebosch, H., Vranjes, I., Baillien, E., & De Witte, H.
(2018). Positive or negative spirals of online behavior? Exploring
reciprocal associations between being the actor and the recipient of
prosocial and antisocial behavior online. New Media and Society,
20(9), 3437–3456. https://doi.org/10.1177/1461444817749518
Falla, D., Ortega-Ruiz, R., Runions, K., & Romera, E. M. (2022). Why
do victims become perpetrators of peer bullying? Moral disen-
gagement in the cycle of violence. Youth & Society,54(3),
397–418. https://doi.org/10.1177/0044118×20973702
Fredrick, S. S., Nickerson, A. B., & Livingston, J. A. (2022). Ado-
lescent social media use: Pitfalls and promises in relation to
cybervictimization, friend support, and depressive symptoms.
Journal of Youth and Adolescence,51(2), 361–376. https://doi.
org/10.1007/s10964-021-01561-6
Giumetti, G. W., Kowalski, R. M., & Feinn, R. S. (2022). Predictors
and outcomes of cyberbullying among college students: A two
wave study. Aggressive Behavior,48(1), 40–54. https://doi.org/
10.1002/ab.21992
Guy, A., Lee, K., & Wolke, D. (2019). Comparisons between ado-
lescent bullies, victims, and bully-victims on perceived popular-
ity, social impact, and social preference. Frontiers in Psychiatry,
10, 868. https://doi.org/10.3389/fpsyt.2019.00868
Hamaker, E. L., Kuiper, R. M., & Grasman, R. P. P. P. (2015). A
critique of the cross-lagged panel model. Psychological Methods,
20(1), 102–116
He, Y., Chen, S. S., Xie, G. D., Chen, L. R., Zhang, T. T., Yuan, M.
Y., Li, Y. H., Chang, J. J., & Su, P. Y. (2022). Bidirectional
associations among school bullying, depressive symptoms and
sleep problems in adolescents: A cross-lagged longitudinal
approach. Journal of Affective Disorders,298, 590–598. https://
doi.org/10.1016/j.jad.2021.11.038
Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in
covariance structure analysis: Conventional criteria versus new
alternatives. Structural Equation Modeling,6(1), 1–55. https://
doi.org/10.1080/10705519909540118
Huang, J., Huebner, E. S., & Tian, L. (2022). Stability and changes in
traditional and cyberbullying perpetration and victimization in
childhood: The predictive role of depressive symptoms. Journal
of Interpersonal Violence,34, NP17300–NP17324. https://doi.
org/10.1177/08862605211028004
Kline, R. B. (2015). Principles and practice of structural equation
modeling (4th ed.). Guilford Press
Kowalski, R. M., Giumetti, G. W., Schroeder, A. N., & Lattanner, M.
R. (2014). Bullying in the digital age: A critical review and
meta-analysis of cyberbullying research among youth. Psycho-
logical Bulletin,140(4), 1073–1137. https://doi.org/10.1037/a
0035618
Lee, Y., Kim, J., & Song, H. (2021). Do negative emotions matter?
Paths from victimization to cyber and traditional bullying from a
General Strain Theory perspective. Crime and Delinquency.
https://doi.org/10.1177/00111287211052439
Li, X., Huebner, E. S., & Tian, L. (2021). Vicious cycle of emotional
maltreatment and bullying perpetration/victimization among early
adolescents: Depressive symptoms as a mediator. Social Science
and Medicine,291, 114483. https://doi.org/10.1016/j.socscimed.
2021.114483
Little, R. J. A. (1988). A test of missing completely at random for
multivariate data with missing values. Journal of the American
Statistical Association,83(404), 1198–1202. https://doi.org/10.
1080/01621459.1988.10478722
Malamut, S. T., Dawes, M., Lansu, T. A. M., van den Berg, Y., &
Cillessen, A. H. N. (2022). Differences in aggression and alcohol
use among youth with varying levels of victimization and popu-
larity status. Journal of Youth and Adolescence,51(10), 1914–1925.
https://doi.org/10.1007/s10964-022-01649-7
Mishna, F., Khoury-Kassabri, M., Gadalla, T., & Daciuk, J. (2012).
Risk factors for involvement in cyber bullying: Victims, bullies
and bully-victims. Children and Youth Services Review,34(1),
63–70. https://doi.org/10.1016/j.childyouth.2011.08.032
Modecki, K. L., Minchin, J., Harbaugh, A. G., Guerra, N. G., &
Runions, K. C. (2014). Bullying prevalence across contexts: A
meta-analysis measuring cyber and traditional bullying. Journal
of Adolescent Health,55(5), 602–611. https://doi.org/10.1016/j.ja
dohealth.2014.06.007
Muthén, L. K., & Muthén, B. O. (1998). Mplus user’s guide (8th ed.).
Muthén & Muthén
Olweus, D. (1993). Bullying at school. What we know and what we
can do. Blackwell
Ortega-Ruiz, R., Del Rey, R., & Casas, J. A. (2016). Evaluar el bul-
lying y el cyberbullying validación española del EBIP-Q y del
ECIP-Q. Psicologia Educativa,22(1), 71–79. https://doi.org/10.
1016/j.pse.2016.01.004
Pabian, S., & Vandebosch, H. (2016). An investigation of short-term
longitudinal associations between social anxiety and victimiza-
tion and perpetration of traditional bullying and cyberbullying.
Journal of Youth and Adolescence,45(2), 328–339. https://doi.
org/10.1007/s10964-015-0259-3
Rodríguez-de Arriba, M. L., Muñoz-Fernández, N., & Sánchez-
Jiménez, V. (2022). Stability of cyber dating victimization and
psychological adjustment in adolescents: a short-term long-
itudinal study. Psychology Society & Education,14(2), 1–10.
https://doi.org/10.21071/psye.v14i2.142
Romera, E. M., Carmona-Rojas, M., Ortega-Ruiz, R., & Camacho, A.
(2022). Bidirectional association between normative adjustment
and bullying perpetration in adolescence: A prospective long-
itudinal study. Revista de Psicodidáctica,27(2), 132–140. https://
doi.org/10.1016/j.psicoe.2022.03.001
Romera, E. M., Ortega-Ruiz, R., Runions, K., & Camacho, A. (2021).
Bullying perpetration, moral disengagement and need for popu-
larity: Examining reciprocal associations in adolescence. Journal
of Youth and Adolescence, 50(10), 2021–2035. https://doi.org/10.
1007/s10964-021-01482-4
Romera, E. M., Ortega-Ruiz, R., Runions, K., & Falla, D. (2021).
Moral disengagement strategies in online and offline bullying.
Psychosocial Intervention,30(2), 85–93. https://doi.org/10.
12968/pnur.1998.9.16.40
Journal of Youth and Adolescence (2023) 52:406–418 417
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Rosen, P. J., Milich, R., & Harris, M. J. (2007). Victims of their own
cognitions: Implicit social cognitions, emotional distress, and peer
victimization. Journal of Applied Developmental Psychology,
28(3), 211–226. https://doi.org/10.1016/j.appdev.2007.02.001
Runions, K. C., Salmivalli, C., Shaw, T., Burns, S., & Cross, D.
(2018). Beyond the reactive‐proactive dichotomy: Rage, revenge,
reward, and recreational aggression predict early high school
bully and bully/victim status. Aggressive Behavior,44, 501–510.
https://doi.org/10.1002/ab.21770
Runions, K. C., Shapka, J. D., Dooley, J., & Modecki, K. (2013).
Cyber-aggression and victimization and social information pro-
cessing: Integrating the medium and the message. Psychology of
Violence, 3(1). https://doi.org/10.1037/a0030511
Satorra, A., & Bentler, P. M. (2001). A scaled difference chi-square
test statistic for moment structure analysis. Psychometrika,66(4),
507–514
Smahel, D., Machackova, H., Mascheroni, G., Dedkova, L., Staksrud,
E., Ólafsson, K., Livingstone, S., & Hasebrink, U. (2020). EU
Kids Online 2020: Survey results from 19 countries (Issue Feb-
ruary). https://doi.org/10.21953/lse.47fdeqj01ofo
Smith, P. K., López-Castro, L., Robinson, S., & Görzig, A. (2019).
Consistency of gender differences in bullying in cross-cultural
surveys. Aggression and Violent Behavior,45,33–40. https://doi.
org/10.1016/j.avb.2018.04.006
Strindberg, J., Horton, P., & Thornberg, R. (2020). Coolness and
social vulnerability: Swedish pupils’reflections on participant
roles in school bullying. Research Papers in Education,35(5),
603–622. https://doi.org/10.1080/02671522.2019.1615114
Subrahmanyam, K., Smahel, D., & Greenfield, P. (2006). Connecting
developmental constructions to the Internet: Identity presenta-
tion and sexual exploration in online teen chat rooms. Devel-
opmental Psychology,42(3), 395–406. https://doi.org/10.1037/
0012-1649.42.3.395
Volk, A. A., Dane, A. V., & Marini, Z. A. (2014). What is bullying? A
theoretical redefinition. Developmental Review,34(4), 327–343.
https://doi.org/10.1016/j.dr.2014.09.001
Walters, G. D. (2021). School-age bullying victimization and perpe-
tration: A meta-analysis of prospective studies and research.
Trauma, Violence, & Abuse,22(5), 1129–1139. https://doi.org/
10.1177/1524838020906513
Wright, M. F., & Li, Y. (2013). Normative beliefs about aggression
and cyber aggression among young adults: A longitudinal
investigation. Aggressive Behavior,39(3), 161–170. https://doi.
org/10.1002/ab.21470
Ybarra, M. L., & Mitchell, K. J. (2004). Online aggressor/targets,
aggressors, and targets: A comparison of associated youth
characteristics. Journal of Child Psychology and Psychiatry,
45(7), 1308–1316. https://doi.org/10.1111/j.1469-7610.2004.
00328.x
Zhou, J., Huebner, E. S., & Tian, L. (2022). Perceived parental
warmth, peer perpetration, and peer victimization: Unraveling
within-child associations from between-child differences. Pre-
vention Science,23, 295–305. https://doi.org/10.1007/s11121-
021-01325-5
Zhu, X., Griffiths, H., Eisner, M., Hepp, U., Ribeaud, D., & Louise,
M. (2022). Developmental associations between bullying victi-
mization and suicidal ideation and direct self-injurious behavior
in adolescence and early adulthood. The Journal of Child Psy-
chology and Psychiatry,63(7), 820–828. https://doi.org/10.
1111/jcpp.13529
Antonio Camacho is a PhD at the University of Cordoba (Spain). His
major research interests include the risk mechanisms involved in
adolescence aggressive behavior.
Kevin Runions is head of the Social and Emotional Wellbeing team at
the Telethon Kids Institute, University of Western Australia. His
research focuses on child and adolescent social behavior,
disadvantage, and promotion of social and emotional wellbeing
equitably.
Rosario Ortega-Ruiz is an Emeritus Professor at the University of
Cordoba (Spain). Her major research interests include bullying,
cyberbullying and interpersonal violence.
Eva M. Romera is an Associate Professor at the University of
Cordoba (Spain). Her major research interests include social and moral
competence and interpersonal relationships.
418 Journal of Youth and Adolescence (2023) 52:406–418
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
Available via license: CC BY 4.0
Content may be subject to copyright.
Content uploaded by Antonio Camacho
Author content
All content in this area was uploaded by Antonio Camacho on Nov 18, 2022
Content may be subject to copyright.