An Exploratory Analysis of Different Problematic Internet Use Profiles in Cybervictims,Cyberbullies,and Cyberbully Victims

Abstract

The Internet has brought about a paradigm shift in the lives of many people, especially adolescents. While it has opened great possibilities, it has also led to various risks such as cyberbullying and problematic Internet use (PIU). These two constructs have been extensively researched individually and jointly, but the existence of different profiles of problematic use according to the role a person assumes in the context of cyberbullying has not yet been explored. Therefore, the main aim of this study is to analyze the different PIU profiles of those who have been cybervictims, cyberbullies, and cyberbully victims. An analytical and cross-sectional study was conducted on 25,341 adolescents between 10 and 18 years of age (14.60 ± 1.68). The Cyberbullying Triangulation Questionnaire (CTQ) and the Spanish version of the Generalized and Problematic Internet Use Scale-2 (GPIUS2) were used. The results indicate that cybervictims (6.4 percent), cyberbullies (4.3 percent), and cyberbully victims (2.7 percent) have different profiles of PIU (p < 0.001). Two common profiles emerge from the three roles: one of nonproblematic use and the other of severe problematic use. Participants who presented severe problematic use are the ones who obtained higher scores in cybervictimization and cyberaggression, particularly in the case of cyberbully victims. Furthermore, this profile is 7.6 (IC99 percent:6.11–9.44) times more likely to present PIU than noninvolved adolescents. These results are relevant when planning cyberbullying-focused interventions and programs because of the association between cyberbullying and general PIU.

Introduction

The digital society is a source of development and opportunities for people. However, it also entails risks, especially for adolescents, such as cyberbullying and Problematic Internet Use (PIU). Cyberbullying is a violent and intentional act that is carried out repeatedly over a long period of time through the use of technologies by one or more individuals toward another person who has difficulty defending himself or herself.¹ Cyberbullying behaviors include online name-calling, denigration, impersonation, exclusion, as well as revealing secrets and taking pictures and videos and spreading them.² The effects of cyberaggression are magnified and multiplied through the use of social networks, which increases the feeling of helplessness and the inability to escape. Furthermore, this behavior usually occurs outside the school context on any day of the week and at any time.¹ The mean prevalence of cybervictimization is 15 percent while for cyberaggression, it is 16 percent, according to some meta-analyses, although the figures vary considerably.³ In fact, other review studies suggest that cybervictimization ranges from 1 percent to 61.1 percent.⁴ It should also be noted that the role of the cyberbully victim is gaining attention and becoming increasingly important,⁵ partly because the psychological consequences associated with this role are usually more severe than for those who are cybervictims or cyberbullies only.⁶

PIU is a complex construct that has been approached from different perspectives⁷ and that is conceptualized in this research as a dysfunctional use of the Internet by an individual.^8,9 This dysfunctional use is characterized by a preference for online social interaction and the regulation of mood through the Internet, which in turn increases the likelihood of deficient self-regulation (characterized by cognitive preoccupation and compulsive Internet use) that can lead to various negative consequences in the person's life.¹⁰ While it is difficult to establish the overall prevalence of PIU due to the many different definitions and assessment tools used,⁷ in the European context, the prevalence ranges from 14.3 percent to 54.9 percent,¹¹ while studies in the Spanish context place the prevalence of severe PIU between 2.4 percent and 4.9 percent.^12,13 The prevalence of at-risk users varies from 16.3 percent ¹⁴ to 18 percent.^12,13 Profiles of problematic Internet users have also been studied, with at least two approaches. The first of these, based on the GPIU and Caplan's model,^9,10 has established the existence of four profiles: nonproblematic users, mood regulation users, problematic users, and severe problematic users.¹² Other authors have established four additional profiles based on other theoretical approaches: first steppers, trainees, sensible users, and heavy users.¹⁵

The association between cyberbullying and PIU has been explored and established in different transversal^16–18 and longitudinal studies.^19,20 For instance, a cross-cultural study¹⁷ found an association between scores in PIU and cybervictimization and cyberaggression, particularly for those adolescents who showed a compulsive use and used Internet to regulate their mood. In a longitudinal study with Spanish adolescents, Gámez-Guadix et al.²¹ found that PIU predicted cyberperpetration 6 months later, although cyberbullying perpetration at T1 was not related to PIU at T2. Regarding victimization it was found that cybervictimization at T1 predicted PIU 6 months later, but PIU at T1 was not associated with cybervictimization at T2.¹⁹

This complex relation could be explained by several theoretical frameworks, one of such would be the Problem-behavior theory.²² According to this theory, those adolescents that participate in one risky or problematic activity are likely to engage in other risky behaviors. Moreover, PIU is linked to an increased use of the Internet and therefore higher exposure to any Internet-related risks such as cyberbullying victimization or perpetration. Furthermore, the relationship between PIU and cyberbullying could be explained according to compensatory Internet use theory,²³ which provides an integrated framework to understand PIU and postulates that stressors will drive certain adolescents to compulsively use the Internet as an alternative to cope with their negative emotions.²⁰

However, to date, the different profiles of problematic users based on their cyberbullying role has not been explored. Therefore, the main objective of this study is to analyze the different PIU profiles of cybervictims, cyberbullies, and cyberbully victims.

Based on the reviewed literature this study poses the next hypothesis: there will be a significant positive relationship between PIU cybervictimization and cyberaggression.^16–20 In addition, considering the gap in the literature described above, the following research question is raised: “Are there differential profiles of problematic use of the Internet for victims, aggressors, or aggressive victims of cyberbullying?.”

Materials and Methods

Design and participants

An analytical and cross-sectional study was carried out in a northern Spanish Autonomous Community. The sample comprised 25,341 participants, of whom 49.9 percent were girls (n = 12,569). The sampling was random and representative of the reference population with a margin of error lower than 0.1 percent (CI 99 percent). The mean age is 14.60 ± 1.68.

Assessment tools

For the analysis of the variables under study, the following instruments were used to collect data about the participants' experiences in the previous 5 months (the start of the course).

(1)

Victimization and aggression scales of the Cyberbullying Triangulation Questionnaire (CTQ).²⁴ The cybervictimization scale comprises 11 items (e.g., “Someone has posted humiliating images of me on the Internet”). The cyberaggression scale includes the same 11 items (e.g., “I have posted humiliating images of a classmate on the Internet”) and four additional items (15 items in total) that describe actions that could not be evaluated from the perspective of the cybervictim (e.g., “I have sent links with humiliating images of other people for them to see”). The CTQ uses a Likert scale with three alternative answers (0 = never, 1 = occasionally, 2 = often). The reliability of each subscale is presented in Table 1.

(2)

Spanish version of the Generalized and Problematic Internet Use Scale-2 (GPIUS2)^9,10 comprised 15 items on a six-point Likert scale ranging from 1 (completely disagree) to 6 (completely agree). Is composed of five factors: preference for online social interaction (e.g., “I prefer to interact with other people through the Internet rather than communicating face to face”), mood regulation (e.g., “I've used the Internet to talk to others when I've felt lonely”), negative consequences (e.g., “My use of the Internet has hindered the control of my life”), cognitive preoccupation (e.g., “I would feel lost if I couldn't connect to the Internet”), and compulsive use (e.g., “When I'm not on the Internet, it's hard to resist the urge to connect”). The reliability of each subscale is presented in Table 1.

Table 1.

Correlations, Descriptive Statistics, and Reliability of Cyberbullying and Problematic Internet Use Dimensions

	1.	2.	3.	4.	5.	6.	7.	ω
1. Cybervictimization	—							0.79
2. Cyberaggression	0.449	—						0.86
3. Online Preference	0.190	0.149	—					0.85
4. Mood Regulation	0.236	0.167	0.483	—				0.83
5. Cognitive Preoccupation	0.224	0.255	0.445	0.468	—			0.80
6. Compulsive Use	0.196	0.211	0.391	0.428	0.727	—		0.83
7. Negative Consequences	0.276	0.256	0.427	0.382	0.591	0.606	—	0.76
Mean	0.74	0.96	5.60	8.15	6.58	5.86	4.58	—
SD	1.60	2.11	3.35	4.56	3.90	3.34	2.69	—
Range	0–22	0–30	3–18	3–18	3–18	3–18	3–18	—

ω = McDonald's Omega. All correlations are significant at p < 0.001.

Procedure

A total of 156 state-funded schools (i.e., entirely funded or concerted) were approached through the [hidden for review]. The collaboration of the students and the schools was voluntary and anonymous. A total of 115 (73 percent) schools accepted to take part in the study (82 public schools and 33 concerted schools).

The study was performed with the authorization of the participants, principals, and the political-educational institutions. Through official channels, the schools submitted a passive consent form informing the legal guardians of the students about the research. Those who did not wish to allow their child to participate returned the signed form. This occurred in less than 1 percent of the sample of those schools that took part in the study.

The collection of data was performed on the online platform SurveyMonkey^® using the computer rooms of each center. The average time to complete the questionnaires was 15 minutes.

The project was approved by the ethics committee of Principado de Asturias (Ref.59/17).

Statistical approach

First, each participant involved in cyberbullying was classified in one of these exclusive roles: cybervictim, cyberbully, and cyberbully victim. For this purpose, the criteria used by González-Cabrera et al.²⁴ were followed. Specifically, those with total scores equal to or greater than three in the cybervictimization scale were classified as cybervictims, while total scores equal to or greater than four in the cyberaggression scale were considered cyberbullies. Those who met both conditions were assigned to the cyberbully victim group. These three categories are mutually exclusive (i.e., no participant was assigned to two categories simultaneously).

With regard to PIU, to establish the problematic category, the cutoff point established by Machimbarrena et al.¹²—a score equal to or higher than 52—was used.

To test the hypotheses of the study (i.e., relationship between suffering problematic use and cybervictimization and cyberaggression), Pearson correlations performed between the direct scores and analysis of odds ratio were calculated between the aforementioned cutoff score of GPIU and being a cybervictim, a cyberbully, and a cyberbullying victim. To answer the research question three Latent Profile Analyses for each subgroup were carried out separately based on the five PIU dimensions, and ANOVA/MANOVA were used to compare their scores.

Regarding the Latent Profile Analysis (LPA), the model that presented the best fit was selected based on several criteria such as Bayesian Information Criterion, Akaike Information Criterion, Approximate Weight of Evidence, Entropy, p-value of the Lo-Mendell-Rubin Adjusted LRT, and the interpretability of the results following the criteria outlined by Akogul and Erisoglu.²⁵

In addition, the following analyses were performed: descriptive analysis (frequencies, means, and standard deviations) and analysis of reliability (omega coefficient). Due to the great number of comparisons, and to limit type I error, only values equal to or less than 0.001 were considered statistically significant. The analyses were performed using SPSS v.25 and the tidyLPA package.²⁶

Results

The prevalence of participants who are cybervictims was 6.4 percent (n = 1647), 4.3 percent (n = 1111) were cyberbullies, and 2.7 percent (n = 699) were cyberbully victims. Table 1 shows the correlations obtained between the cybervictimization and cyberaggression subscales and the dimensions of the PIU, in addition to their descriptive statistics. In addition, PIU correlated positively with cybervictimization (r = 0.288, p < 0.001) and cyberaggression (r = 0.263, p < 0.001).

Cybervictims were 4.1 times more likely to present PIU than noninvolved adolescents (OR = 4.14 [3.50–4.90]), while cyberbullies were 3.2 times more likely (OR = 3.21 [2.59–3.97]), and cyberbully victims were 7.6 times more likely to present PIU than noninvolved adolescents (OR = 7.59 [6.1–9.44]).

Subsequently, LPAs were performed for each group (cybervictims, cyberbullies, and cyberbully victims) based on the direct scores obtained in the five PIU dimensions. Table 2 presents the values for the different models, with the four-profile model being the most appropriate for cybervictims and the three-profile model for cyberbullies and cyberbully victims.

Table 2.

Fit of the Profile Models Based on the Problematic Internet Use Dimensions for Cybervictims, Cyberbullies, and Cyberbully Victims

	# pro files	AIC	BIC	S-BIC	LL	AWE	Lo-Mendell (p)	Entr.	Prob Min-max	N per group
	# pro files	AIC	BIC	S-BIC	LL	AWE	Lo-Mendell (p)	Entr.	Prob Min-max	1	2	3	4	5
Cybervictims										1647
	2	24,940.78	25,081.36	24,998.76	−12,444.39	24,898.21	441.99 (0.00)	0.85	0.90–0.97	1244	403
	3	24,759.15	24,932.17	24,830.51	−12,347.58	24,709.30	189.37 (0.00)	0.82	0.80–0.96	1143	215	289
	4	24,469.21	24,674.66	24,553.94	−12,196.60	24,412.07	295.30 (0.00)	0.87	0.81–0.97	1112	253	158	154
	5	24,381.04	24,618.93	24,479.15	−12,146.52	24,316.62	97.97 (0.02)	0.77	0.81–0.87	606	492	265	154	130
Cyberbullies										1111
	2	16,467.62	16,597.96	16,515.38	−8359.96	16,729.01	297.23 (0.01)	0.87	0.88–0.98	924	187
	3	16,313.72	16,474.14	16,372.50	−8207.81	16,429.26	162.05 (0.00)	0.83	0.79–0.96	784	165	162
	4	16,203.28	16,393.77	16,273.07	−8124.86	16,267.90	98.52 (0.35)	0.86	0.85–0.96	719	179	167	46
Cyberbully victims										699
	2	11,037.94	11,156.75	11,074.19	−5792.97	11,594.63	114.69 (0.00)	0.77	0.85–0.96	520	179
	3	10,971.58	11,117.80	11,016.20	−5453.79	10,920.61	76.42 (0.00)	0.82	0.85–0.88	295	239	165
	4	10,914.54	11,088.18	10,967.52	−5419.27	10,855.92	67.65 (0.08)	0.87	0.79–0.97	510	108	50	31

The selected model is shown in boldface.

AIC, Akaike Information Criterion; BIC, Bayesian Information Criterion; S-BIC, sample-size adjusted BIC; AWE, approximate weight of evidence; LL, logarithm likelihood; Lo-mendell, Lo-Mendell-Rubin Adjusted LRT Test; Entr., Entropy; Prob min-max: Classification Probabilities for the Most Likely Latent Class Membership.

The results of the multivariate analysis of variance (MANOVA) comparing the scores of the profiles in the different variables confirm statistically significant differences in the five dimensions of the PIU between the four profiles among cybervictims (Pillai's trace = 1.39, F(15, 4923) = 286.00, p < 0.001, η_p² = 0.466), the three profiles among cyberbullies (Pillai's trace = 1.20, F(10, 2210) = 330.00, p < 0.001, η_p² = 0.599) and cyberbully victims (Pillai's trace = 1.30, F(10, 1386) = 74.86, p < 0.001, η_p² = 0.651), and the 10 profiles altogether (Pillai's trace = 1.45, F(45, 17,235) = 156.305, p < 0.001, η_p² = 0.290) in the scores for the five dimensions of the PIU. Scores for each profile and the results of the MANOVA dimensions can be found in Table 3. The profiles are graphically represented through their standardized scores in Figure 1.

FIG. 1.

Standardized scores in the five dimensions of PIU by profile for cybervictims, cyberbullies, and cyberbully victims. CV-NoPIU, cybervictims- No PIU; CV-POSI, cybervictims preference for online social interaction; CV-AVOI, cybervictims online social interaction avoidant; CV-SevPIU, cybervictims severe PIU; CB-NoPIU, cyberbullies -NoPIU; CB-DSR, cyberbullies deficient self regulation; CB-SevPIU, cyberbullies severe PIU; CBV-NoPIU, cyberbully-victims No PIU; CBV-SevPIU, cyberbully-victims severe PIU; CBV-MRU, cyberbully-victims mood regulation user. The displayed scores are standardized.

Table 3.

Mean Scores and Standard Deviations in Problematic Internet Use Dimension by Profiles and ANOVA Over Each Role

Role	PIU Profile	Online preference	Mood regulation	Cognitive preoccupation	Compulsive use	Negative consequences
Role	PIU Profile	M (SD)	M (SD)	M (SD)	M (SD)	M (SD)
CV	No PIU ^CV1	5.05	9.71	5.90	6.85	4.73
	No PIU ^CV1	(2.17)	(4.44)	(3.28)	(3.80)	(1.99)
	POSI ^CV2	13.32	13.82	7.43	8.45	5.01
	POSI ^CV2	(2.43)	(3.74)	(3.57)	(3.91)	(1.85)
	Avoidance ^CV3	4.86	12.74	9.37	11.32	12.08
	Avoidance ^CV3	(1.85)	(3.74)	(4.17)	(3.80)	(2.20)
	Severe PIU ^CV4	12.97	14.11	11.76	12.56	12.15
	Severe PIU ^CV4	(2.78)	(3.71)	(3.71)	(3.73)	(2.47)
	F(df)η²	F _{(3, 1643)} = 1342.68^***; η² = 0.710	F _{(3, 1643)} = 107.66^***; η² = 0.164	F _{(3, 1643)} = 159.74^***; η² = 0.226	F _{(3, 1643)} = 141.59^***;η² = 0.205	F _{(3, 1643)} = 1015.80^***; η² = 0.650
	Post hocGames-Howell	CV1 < CV2, CV4CV2> CV3; CV3<CV4	CV1<CV2, CV3, CV4	CV1<CV2, CV3, CV4CV2<CV3, CV4; CV3<CV4	CV1<CV2, CV3, CV4; CV2<CV3, CV4	CV1<CV3, CV4; CV2<CV3, CV4
CB	No PIU ^CB1	5.58	8.66	5.40	6.64	4.62
	No PIU ^CB1	(2.94)	(4.30)	(2.31)	(3.39)	(1.86)
	DSR ^CB2	7.06	10.74	13.34	12.99	5.59
	DSR ^CB2	(3.57)	(4.58)	(2.43)	(3.36)	(2.15)
	Severe PIU ^CB3	9.45	11.82	12.00	12.70	12.08
	Severe PIU ^CB3	(3.97)	(4.02)	(3.34)	(3.93)	(2.52)
	F(df)η²	F _{(2, 1108)} = 103.98^***;η² = 0.158	F _{(2, 1108)} = 45.59^***;η² = 0.0.76	F _{(2, 1108)} = 981.723^***;η² = 0.639	F _{(2, 1108)} = 369.72^***;η² = 0.400;	F _{(2, 1108)} = 936.92^***;η² = 0.628
	Post hocGames-Howell	CB1<CB2< CB3	CB1< CB2, CB3	CB1 < CB2, CB3CB2 > CB3	CB1 < CB2, CB3	CB1< CB2< CB3
CBV	No PIU ^CBV1	6.56	7.48	7.28	7.71	5.99
	No PIU ^CBV1	(3.15)	(2.72)	(3.33)	(3.58)	(2.52)
	MR User ^CBV2	7.83	14.57	8.85	10.18	5.59
	MR User ^CBV2	(4.24)	(2.43)	(3.95)	(4.16)	(2.30)
	Severe PIU ^CBV3	10.02	13.22	12.56	13.25	13.73
	Severe PIU ^CBV3	(4.40)	(3.29)	(3.55)	(3.57)	(2.49)
	F(df)η²	F _{(2, 696)} = 42.59^***;η² = 0.109	F _{(2, 696)} = 485.22^***;η² = 0.582	F _{(2, 696)} = 114.23^***;η² = 0.247	F _{(2, 696)} = 114.227^***;η² = 0.248	F _{(2, 696)} = 665.36^***;η² = 0.657
	Post hocGames-Howell	CBV1< CB2< CB3	CBV1 < CVB2 < CVB3	CBV1< CBV2< CBV3	CBV1< CBV2< CBV3	CBV1, CBV2 < CBV3

***

p < 0.001.

CV, cybervictims; CB, cyberbullies; CVB, cyberbully victims; POSI, preference for online social interaction; DSR, deficient self regulation; MR, mood regulation; M, mean; SD, standard deviation; F, Fishers F; df , degrees of freedom; η², eta squared.

A profile of nonproblematic use (NoPIU) can clearly be observed in the three roles of cyberbullying. Similarly, a profile with severe PIU (SevPIU), that presents high scores in all five dimensions—especially relevant are its high scores in negative consequences—can be found in the three roles. In the case of cybervictims, two additional profiles were found: (1) a profile that combines a low preference for online social interaction (z-score = −0.22) and high scores on negative Internet consequences (z-score = 2.77), cognitive preoccupation (z-score = 1.12), and compulsive use (z-score = 1.21) (Online Interaction Avoidance Profile; CV-AVOID) and (2) a profile that presents almost opposite characteristics, with a high preference for online social interaction (z-score = 2.29) and minimal consequences (z-score = 0.16) (Preference for online social interaction; CV-POSI). In the group of cyberbullies, a third profile emerges where deficient self-regulation dominates (high cognitive preoccupation and compulsive use), but there are few negative consequences (z-score = 0.36) (Deficient Self-Regulation; CB-DSR). Finally, in cyberbully victims, a profile emerges with relatively high scores on all dimensions, especially in mood regulation (z-score = 1.41), but with low scores on negative consequences (z-score = 0.37) (Mood Regulation User; CBV-MRU).

Finally, the ANOVA analysis between the cybervictimization and cyberaggression scores and the PIU profiles for each role (Table 4) revealed that those who were assigned to the Severe PIU profile reported the highest scores in cybervictimization and/or cyberaggression in the three roles.

Table 4.

ANOVA of Scores in Cybervictimization and Cyberaggression by Latent Profile Analysis Problematic Internet Use Profile

Role	PIU Profile	Cyber victimization	Cyber aggression
Role	PIU Profile	M (SD)	M (SD)
CV	No PIU ^CV1	4.25 (2.06)	—
	POSI ^CV2	4.28 (1.78)	—
	Avoidance ^CV3	4.70 (1.80)	—
	Severe PIU ^CV4	5.17 (3.07)	—
	F(df)	F _{(3, 1643)} ₌ 9.81^***	—
	η²	η² = 0.81	—
	Post hoc Games-Howell	CV1 < CV4	—
CB	No PIU ^CB1	—	5.83 (3.41)
	DSR ^CB2	—	5.37 (2.43)
	Severe PIU ^CB3	—	7.33 (6.03)
	F(df)	—	F _{(2, 1108)} = 13.08^***
	η²	—	η² = 0.23
	Post hoc Games-Howell	—	CB2 < CB3
CBV	No PIU ^CBV1	5.15 (2.96)	7.74 (4.84)
	MR User ^CBV2	4.56 (1.91)	6.20 (3.27)
	Severe PIU ^CBV3	7.22 (4.73)	9.64 (5.93)
	F(df)	F _{(2, 696)} ₌ 36.16^***	F _{(2, 696)} ₌ 26.50^***
	η²	η² = 0.94	η² = 0.71
	Post hoc Games-Howell	CBV1 > CBV2CBV2 < CBV3	CBV1 > CBV2,CBV1 < CVB3CBV2 < CBV3

***

p < 0.001.

M, mean; SD, standard deviation; F, fishers F; df, degrees of freedom; η² = eta squared.

Discussion

This article contributes to the understanding of the relationship between cyberbullying and problematic use of the Internet from a previously unaddressed perspective. While both constructs have been analyzed individually and jointly in numerous articles,^{13–18,27,28} no other study has analyzed the different PIU profiles in cybervictims, cyberbullies, and cyberbully victims.

Other works that have explored problematic use profiles in broad and general samples^12,15 have identified two clearly defined groups: users without problems and users with severe problems.¹² The focus of this study, however, is on three subsamples that present an Internet- related risk, namely, being a cybervictim, a cyberbully, or a cyberbully victim. As in other studies, the existence of the aforementioned two profiles (No PIU and Severe PIU) is observed in all three roles.^12,15 However, scores for most of the dimensions of PIU use are slightly over the mean in the No PIU profiles in all three roles (particularly in cybervictims and cyberbully victims).

In the group of cybervictims, in addition to the two aforementioned profiles, a third profile emerges that we label “online social interaction avoidant” (CV-AVOI) because it is characterized by a below-average online preference but with moderate scores in mood regulation, poor self-regulation, and high negative consequences. This profile may fit a person who is being victimized online and, thus, avoiding Internet use and interacting with others online, which could be a coping strategy.^29,30 In contrast, a fourth profile among cybervictims has a strong online preference and moderate mood regulation that presents neither deficient self-regulation nor negative consequences. It could be hypothesized that the impact of the negative behavior in these victims is probably lower or that they use different strategies to regulate their emotions.^31,32 Among these four profiles, the ones with severe PIU (SevPIU) obtain the highest cybervictimization scores.

In the group of cyberbullies, one profile in which high scores on poor self-regulation as characterized by Caplan stands out, this group shows high scores on compulsive Internet use and excessive concern about it, but this does not seem to lead to negative consequences in the person's life.^9,10 Along these lines, other studies have shown the scarce psychological consequences of the role of pure cyberbully at an emotional and cognitive level.^6,33 Nevertheless, this profile is the one that obtains the lowest score in cyberaggression, together with those that do not present PIU problems. Finally, the severe PIU profile is again the profile that obtains the highest scores in cyberaggression.

In the group of cyberbully victims, a profile emerges that could be similar to what Machimbarrena et al.¹² termed mood regulation use, which could be similar to the profile of entertainment users identified in other studies.¹⁵ This profile is notable for their use of the Internet to regulate their mood but without any suggestion of poor self-regulation or negative consequences. The severe PIU group within the group of cyberbully victims has the highest scores in both cybervictimization and cyberaggression of the entire sample (higher than those of cybervictims and cyberbullies). This situation is common to other psychosocial problems where the group that combines problems of cybervictimization and cyberaggression usually presents greater psychological effects and more associated problems,^34–38 in this case, PIU. This falls in line with the problem-behavior theory²² according to which adolescents with multiple risks (such as being a cybervictim and/or a cyberbully) are more likely to present additional risks and issues.^39,40 However causality cannot be inferred from our data; therefore, the compensatory Internet use theory could not be examined.

In addition, it should be noted that a clear relationship exists between the roles associated with cyberbullying and those associated with PIU. As such, the possibility of presenting PIU varies between OR = 3.21 in the case of cyberbullies to OR = 7.59 in the case of cyberbully victims. The trend of these data is consistent with other studies,¹⁶ although the values found in this research are significantly higher and confirm the posed hypothesis.

This research has several theoretical and practical implications for professionals. On the one hand, the problem of cyberbullying is associated with PIU, so knowing the profile of problematic use can help in determining the approach of an intervention in school and clinical settings. Likewise, it is important that cyberbullying prevention programs also contemplate the adequate management of Internet use beyond the recognition of inappropriate behavior. The relationship between a relational risk such as cyberbullying and another risk associated with a dysfunctional use of technology suggests the need to explore the common foundations of these problems.

It should be noted that this study has some limitations that should be taken into consideration. First, the sole use of self-report questionnaires may influence the results due to social desirability. Second, some samples of problematic use profiles are small, even if the initial sample was particularly high. Finally, this is a novel exploratory approach to this problem and should be taken as such. Therefore, we propose that future studies use a longitudinal design and aim to include more informants along with other Internet risks. Furthermore, it would also be of great interest to study the coping styles associated with these profiles.

In conclusion, there are different profiles of Internet use according to the role played in cyberbullying, with cyberbully victims reporting the most negative consequences regarding their use of the Internet. Likewise, a relationship exists between a person's problematic use of the Internet and their role as a cybervictim, cyberbully, or cyberbully victim.

Footnotes

Acknowledgments

The authors are grateful for the collaboration of the “Consejería de Educación y Cultura del Principado de Asturias” and the directive teams, teachers, family members, and students who have participated.

Author Disclosure Statement

There are no financial, industrial, or other relationships that may constitute a conflict of interest concerning this work.

Funding Information

This research was funded by the Spanish Ministry of Economy, Industry and Competitiveness, RTI2018-094212-B-I00: (CIBER-AACC), and by the International University of La Rioja, Project “Cyberpsychology (Triennium 2017-2020)”.

References

Tokunaga

. Following you home from school: A critical review and synthesis of research on cyberbullying victimization. Computers in Human Behavior, 2010; 26:277–287.

Kowalski

, Giumetti

, Schroeder

, et al. Bullying in the digital age: a critical review and meta-analysis of cyberbullying research among youth. Psychological Bulletin, 2014; 140:1073–1137.

Modecki

, Minchin

, Harbaugh

, et al. Bullying prevalence across contexts: a meta-analysis measuring cyber and traditional bullying. Journal of Adolescent Health, 2014; 55:602–611.

Brochado

, Soares

, Fraga

. A Scoping Review on Studies of Cyberbullying Prevalence Among Adolescents. Trauma, Violence, & Abuse, 2017; 18:523–531.

Lozano-Blasco

, Cortés-Pascual

, Latorre-Martínez

. Being a cybervictim and a cyberbully – The duality of cyberbullying: a meta-analysis. Computers in Human Behavior, 2020; 111:106444.

Zych

, Ortega-Ruiz

, Del Rey

. Systematic review of theoretical studies on bullying and cyberbullying: facts, knowledge, prevention, and intervention. Aggression and Violent Behavior, 2015; 23:1–21.

Tokunaga

, Rains

. A review and meta-analysis examining conceptual and operational definitions of problematic internet use: Problematic internet use. Human Communication Research, 2016; 42:165–199.

Caplan

. Problematic Internet use and psychosocial well-being: development of a theory-based cognitive–behavioral measurement instrument. Computers in Human Behavior, 2002; 18:553–575.

Caplan

. Theory and measurement of generalized problematic Internet use: a two-step approach. Computers in Human Behavior, 2010; 26:1089–1097.

10.

Gámez-Guadix

, Orue

, Calvete

. Evaluation of the cognitive-behavioral model of generalized and problematic Internet use in Spanish adolescents. Psicothema, 2013; 25:299–306.

11.

Laconi

, Kaliszewska-Czeremska

, Gnisci

, et al. Cross-cultural study of Problematic Internet Use in nine European countries. Computers in Human Behavior, 2018; 84:430–440.

12.

Machimbarrena

, González-Cabrera

, Ortega-Barón

, et al. Profiles of problematic internet use and its impact on adolescents' health-related quality of life. International Journal of Environmental Research and Public Health, 2019; 16:3877.

13.

Machimbarrena

, Calvete

, Fernández-González

, et al. Internet risks: An overview of victimization in cyberbullying, cyber dating abuse, sexting, online grooming and problematic internet use. International Journal of Environmental Research and Public Health, 2018; 15:2471.

14.

Gómez

, Rial

, Braña

, et al. Screening of problematic internet use among spanish adolescents: Prevalence and related variables. Cyberpsychology, Behavior, and Social Networking, 2017; 20:259–267.

15.

Rial

, Gómez

, Picón

, et al. Identification and Characterization of Adolescent Internet User's Profiles. The Spanish Journal of Psychology, 2015; 18:E50.

16.

Jung

Y-E

, Leventhal

, Kim

, et al. Cyberbullying, problematic internet use, and psychopathologic symptoms among Korean Youth. Yonsei Medical Journal, 2014; 55:826.

17.

Yudes-Gómez

, Baridon-Chauvie

, González-Cabrera

J-M

. Cyberbullying and problematic Internet use in Colombia, Uruguay and Spain: cross-cultural study. Comunicar, 2018; 26:49–58.

18.

Zsila Á, Orosz

, Király

, et al. Psychoactive substance use and problematic internet use as predictors of bullying and cyberbullying victimization. International Journal of Mental Health and Addiction, 2018; 16:466–479.

19.

Gámez-Guadix

, Orue

, Smith

, et al. Longitudinal and reciprocal relations of cyberbullying with depression, substance use, and problematic internet use among adolescents. Journal of Adolescent Health, 2013; 53:446–452.

20.

Liu

, Liu

, Yuan

. Cyberbullying victimization and problematic Internet use among Chinese adolescents: longitudinal mediation through mindfulness and depression. Journal of Health

Psychology

. 2020. [Epub ahead of print]; DOI: 10.1177/1359105320934158.

21.

Gámez-Guadix

, Borrajo

, Almendros

. Risky online behaviors among adolescents: longitudinal relations among problematic Internet use, cyberbullying perpetration, and meeting strangers online. Journal of Behavioral Addictions, 2016; 5:100–107.

22.

Jessor

. Risk behavior in adolescence: a psychosocial framework for understanding and action. Journal of Adolescent Health, 1991; 12:597–605.

23.

Kardefelt-Winther

. A conceptual and methodological critique of internet addiction research: towards a model of compensatory internet use. Computers in Human Behavior, 2014; 31:351–354.

24.

González-Cabrera

, León-Mejía

, Machimbarrena

, et al. Psychometric properties of the cyberbullying triangulation questionnaire: a prevalence analysis through seven roles. Scandinavian Journal of Psychology, 2019; 60:160–168.

25.

Akogul

, Erisoglu

. An approach for determining the number of clusters in a model-based cluster analysis. Entropy, 2017; 19:452.

26.

Rosenberg

, Beymer

, Anderson

, et al. tidyLPA: an r package to easily carry out latent profile analysis (LPA) using open-source or commercial software. Journal of Open Source Software, 2019; 4:978.

27.

Brighi

, Menin

, Skrzypiec

, et al. Young, Bullying, and Connected. Common Pathways to Cyberbullying and Problematic Internet Use in Adolescence. Frontiers in Psychology, 2019; 10:1467.

28.

Rial

, Golpe

, Isorna

, et al. Minors and problematic Internet use: evidence for better prevention. Computers in Human Behavior, 2018; 87:140–145.

29.

Machackova

, Cerna

, Sevcikova

, et al. Effectiveness of coping strategies for victims of cyberbullying. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 2013; 7:3.

30.

Šléglová. V, Cerna A. Cyberbullying in adolescent victims: perception and coping. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 2011; 5:2.

31.

Arató. N, Zsidó AN, Lénárd K, et al. Cybervictimization and cyberbullying: The role of socio-emotional skills. Frontiers in Psychiatry, 2020; 11:248.

32.

Marín-López

, Zych

, Ortega-Ruiz

, et al. Relations among online emotional content use, social and emotional competencies and cyberbullying. Children and Youth Services Review, 2020; 108:104647.

33.

Zych

, Ortega-Ruiz

, Marín-López

. Cyberbullying: a systematic review of research, its prevalence and assessment issues in Spanish studies. Psicología Educativa, 2016; 22:5–18.

34.

Cook

, Williams

, Guerra

, et al. Predictors of bullying and victimization in childhood and adolescence: a meta-analytic investigation. School Psychology Quarterly, 2010; 25:65–83.

35.

Tippett

, Wolke

. Socioeconomic Status and Bullying: a Meta-Analysis. American Journal of Public Health, 2014; 104:e48–e59.

36.

Lereya

, Samara

, Wolke

. Parenting behavior and the risk of becoming a victim and a bully/victim: a meta-analysis study. Child Abuse & Neglect, 2013; 37:1091–1108.

37.

González-Cabrera

, Machimbarrena

, Ortega-Barón

, et al. Joint association of bullying and cyberbullying in health-related quality of life in a sample of adolescents. Quality of Life Research, 2019; 29:941–952.

38.

Holt

, Vivolo-Kantor

, Polanin

, et al. Bullying and suicidal ideation and behaviors: a meta-analysis. Pediatrics, 2015; 135:e496–e509.

39.

González-Cabrera

, Machimbarrena

, Fernández-González

, et al. Health-related quality of life and cumulative psychosocial risks in adolescents. Youth & Society 2019. [Epub ahead of print]; DOI: 10.1177/0044118X19879461.

40.

Prochaska

, Spring

, Nigg

. Multiple health behavior change research: an introduction and overview. Preventive Medicine, 2008; 46:181–188.