Computer Use and Internet Bullying among US Adolescents: Gender and Grade Differences

Abstract

The purposes of the current study were to examine gender and grade differences in computer use and Internet bullying, and the relationship between computer use and Internet bullying.

Methods

Data were obtained from the Health Behavior in School-Aged Children (HBSC) 2005 Survey, a nationally-representative sample of grades 6 to 10 (N = 7222) in the United States. Multinomial logistic regressions were used for gender and grade differences in computer use. Logistic regression analyses were used for involvement in Internet bullying and victimization, with gender, grade and computer use as predictors.

Results

Results showed that adolescents who spent 2 or more hours per day on computer use were more likely to bully others and to be bullied by others using computers. Females spent more time using computer, but were less likely to use computers to bully others. There was no gender difference in Internet victimization. Computer use increased with grade, but older adolescents were less likely to engage in either Internet bullying or victimization.

Conclusion

Time spent on computer use plays an important role on involvement in Internet bullying and victimization among adolescents.

I. Introduction

According to the 2003 National Center for Education Statistics, over 95% of adolescents in the United States were computer users. Seventy percent of sixth to eighth grade students used the Internet, and this percentage increased to 79% among ninth to twelfth grade students [1]. Accessibility to computer and Internet use may hold tremendous promise for educational purposes and social networking. However, the high prevalence rates of computer and Internet use may also put adolescents at higher risks for Internet bullying, a form of cyber bullying that can be defined as intentional aggression or harassment which is transmitted by computers or the Internet [2].

Based on the Youth Internet Safety Survey, it was estimated that the prevalence of being a target of Internet harassment in the previous year increased by 50% from 6% in 2000 to 9% in 2005 [3]. The increasing prevalence rates of online harassment are important to public health because of their associations with negative psychosocial correlates such as delinquency, elevated depressive symptoms, frequent substance use and increased social anxiety [4, 5, 6]. The current study examines gender and age differences in computer use and Internet bullying, and the association between computer use and internet bullying among U.S. adolescents.

Gender and grade differences in computer use

Mixed findings on gender differences in computer use have been reported. It is known that boys have been more likely to use computers and the Internet than girls [7] but this gender gap has become narrower in recent years [1, 8]. It remains unclear to what extent the gender gap in computer use has closed because gender differences in attitudes towards computer use remain [9]. Continued monitoring of potential gender differences is necessary.

Age or grade differences in computer use are well documented. DeBell and Chapman [1], for instance, found that only 50% of first to fifth graders used the Internet, whereas 70% of sixth to eighth graders and 79% of ninth to twelfth graders used the Internet. Similarly, Colley and Comber [9] reported higher rates of computer use among older adolescents in the United Kingdom. The consistent trend is for increased computer use as children get older.

Gender and grade differences in Internet bullying

As research on cyber bullying or Internet bullying is still in its early stage, there are only a handful of studies on gender and age differences in Internet bullying and the results are mixed. For example, Ybarra and Mitchell [10] found that in the U.S., boys were three times more likely to be frequent perpetrators of online harassment than girls, whereas girls were 50% more likely to be infrequent perpetrators than boys. Similarly, Li [11] found that boys in Canada are more likely to be bullies and cyberbullies than girls. These findings suggest that boys may have a higher chance of engaging in Internet bullying than girls. However, other studies have found no gender differences in both bullying others (bullying) and being bullied by others (victimization) through computers or the Internet [12, 13].

Few studies have examined grade or age differences in cyber or Internet bullying. Evidence thus far indicates that both cyber bullying and victimization may be more prevalent among older adolescents. For example, Ybarra and Mitchell [6] found that a higher prevalence of Internet aggressors were found in 15 to 17 years old than 10 to 14 years old U.S. students. Similarly, Smith and colleagues [14] found that older students in the U.K. were more likely to engage in cyber bullying and to be a victim in bullying through the website. Yet in another study, Slonje and Smith [13] found no significant age differences in cyber bullying or cyber victimization among U.K. students. It remains unclear in the literature to what extent grade is linked to Internet bullying.

Computer use and Internet bullying

There are only a few studies examining the association between computer use and Internet bullying. Based on a U.S. national sample, Ybarra and colleagues [3] reported that Internet use was not associated with online harassment. However, Hinduja and Patchin [12] found that both computer proficiency and time spent online were associated with online offending and victimization. Interestingly, Smith and colleagues [14] found that being a victim but not a bully was associated with Internet use among U.K. students. Research is needed to determine whether frequent computer use is associated with increased chance of Internet bullying among U.S. adolescents.

The frequency of computer use is a potentially modifiable factor that might be associated with Internet bullying and/or victimization. An existing guideline published by the American Academy of Pediatrics suggests a maximum of 2 hours of screen time (television and computer) per day in children and adolescents [15]. It is of interest to examine what level of computer use might put adolescents at a higher risk of Internet bullying and evaluate whether the current recommended screen time limit is adequate to protect children and adolescents from Internet bullying and/or victimization.

Purposes of the current study

The purposes of the current study were to examine gender and grade differences in computer use and Internet bullying and to evaluate the association between computer use and Internet bullying in a diverse and nationally representative sample. As previous studies show that bullying and victimization may represent two distinct behaviors in terms of their associations with gender and computer use, we examine Internet bullying and victimization separately for all analyses in the current study.

II. Methods

A. Sample and Procedure

Data were collected through anonymous self-report questionnaires in the 2005/2006 World Health Organization collaborative Health Behavior in School-aged Children (HBSC) study in the United States. To obtain a nationally-representative sample with controllable estimation errors, the U.S. sampling design was a three-stage stratified design with an oversample of minority students. The study protocol was reviewed and approved by the Institutional Review Board of the Eunice Kennedy Shriver National Institute of Child Health and Human Development. Youth assent and parental consent were obtained as required by the participating school districts.

Information was obtained from 327 schools out of which 97 schools were identified as ineligible schools. Data were collected from students in the remaining 230 schools; 85% (9,016) of the eligible students participated in the HBSC study. The questionnaire containing the Internet bullying items was given to approximately half of the students in grade 6 (randomly selected); therefore, a total of 7,508 students completed the HBSC questionnaire with both computer use and internet bullying items. The software used for the current study, the survey procedures of SAS, version 9.1 [16], fully took into account this sampling design in the analyses.

B. Measures

1) Demographic Variables

Gender was measured as male or female. Grade was measured with five levels: 6 through 10. The family affluence scale, FAS, consisting of four items on family material wealth (i.e, having own bedroom, number of times on a traveling vacation in a year, number of home computers and number of cars owned), was used as the proxy for socioeconomic status (SES). The sum of the four items was calculated and standardized. This scale has shown desirable reliability and validity in previous studies [17].

2) Computer Use

Students were asked “About how many hours a day do you usually use a computer for chatting on-line, internet, emailing, homework etc. in your free time?” Response categories ranged from “None at all” to “About 7 or more hours a day” (Table 1).

TABLE I.

FREQUENCY DISTRIBUTION OF COMPUTER USE

		Weekday		Weekend
Responses	code	N	%	N	%
None at all	0	2045	25.9	2080	27.2
About .5 hr a day	.5	1598	22.1	1320	17.4
About 1 hr a day	1	1388	19.8	1139	16.1
About 2 hrs a day	2	991	14.4	941	13.9
About 3 hrs a day	3	501	7.6	564	8.1
About 4 hrs a day	4	291	4.3	405	6.1
About 5 hrs a day	5	187	2.8	269	4.0
About 6 hrs a day	6	74	1.0	148	2.2
About >=7 hrs a day	7.5	147	2.1	356	5.1

Questions were asked for weekday and weekend separately. Responses were recoded to develop a continuous variable on hours of computer use per day. A weighted average was generated by combining the two items on computer use: weekday *5/7+weekend*2/7. As the distribution was highly skewed, a categorical variable on computer use was generated with four categories: 1) less than 1 hour; 2) 1 to 2 hours; 3) 2 to 3 hours; and 3) more than 3 hours.

3) Internet Bullying

Internet bullying was measured separately for behaviors of bullying others (i.e., bullying) and being bullied by others (i.e., victimization), using the standard format as in the Revised Olweus Bully/Victim Questionnaire [18]. Students were asked “using a computer or e-mail messages or pictures,” how often in the past couple of months 1) did they bully others; and 2) were they bullied by others at school. Response options were “none,” “only once or twice,” “2 or 3 times a month,” “about once a week,” and “several times a week.” For each variable of bullying and victimization, students were categorized into two groups: not involved if they responded “none” and involved otherwise.

C. Analyses

Data analyses consisted of three steps. The first step was to examine gender and grade differences in computer use. A multinomial logistic regression was applied, with the less than 1 hour per day as the reference group, gender and grade as predictors and FAS as a control variable. The second step was to examine gender and grade differences in involvement with Internet bullying and victimization. Logistic regressions were applied separately for bullying and victimization. To test interaction influence of gender and grade in computer use, bullying and victimization, an interaction term of gender with grade was included in each regression analyses. The third step was to repeat these analyses with computer use as an additional predictor in the logistic regressions of internet bullying and victimization. These analyses test both the influence of computer use and the variability in internet bullying by gender and grade, when computer use is controlled. To test interaction between computer use with gender or grade, two interaction terms were included, i.e., computer use with gender and computer use with grade.

III. RESULTS

A. Sample Characteristics

Among the 7,508 adolescents in grade 6 through 10 who completed the survey, 286 were excluded due to missing information on variables included in the current study, resulting in an analytic sample of 7,222. The sample consisted of 47.4% males and 52.6% females. The mean age of the sample was 14.3 years, with a standard deviation of 1.42.

B. Gender and Grade Differences in Computer Use

The frequency distribution of computer use by gender and grade is reported in Table 2.

TABLE II.

FREQUENCY DISTRIBUTION OF COMPUTER USE, BY GENDER AND GRADE

All (N)	Total	Grade 6	Grade 7	Grade 8	Grade 9	Grade 10
	7222	1060	1671	1654	1368	1469
Computer
<1hr	32.7	52.4	44.0	37.1	31.9	32.7
1–2hrs	28.1	24.2	26.7	27.1	28.0	28.1
2–3hrs	17.0	9.8	11.8	12.6	16.1	17.0
>=3hrs	22.2	13.6	17.5	23.2	24.1	22.2
Male (N)	3433	514	792	725	671	731
Computer
<1hr	43.4	60.7	53.7	42.5	36.0	34.7
1–2hrs	25.6	20.2	23.4	25.6	29.0	26.6
2–3hrs	13.7	7.3	10.0	11.4	16.8	18.7
>=3hrs	17.3	11.7	12.9	20.5	18.2	20.0
Female (N)	3789	546	879	929	697	738
Computer
<1hr	32.7	44.2	35.1	32.3	28.1	30.8
1–2hrs	28.6	28.2	29.6	28.5	27.1	29.6
2–3hrs	14.3	12.2	13.5	13.7	15.4	15.3
>=3hrs	24.4	15.4	21.8	25.5	29.4	24.4

It is important to note that 39.2% of adolescents reported using computers 2 or more hours per day. For visual illustration, percentages of heavy computer use, defined as >=2 hrs per day, are plotted for each gender and grade group (Figure 1). Figure 1 suggests a positive linear relation between grade and percentage of heavy or excessive computer use.

Fig 1.

To statistically test gender and grade difference in computer use, the results of multinomial logistic regression are reported in Table 3. As the interaction between gender and grade was not significant, analyses were rerun by excluding the interaction term. The category of less than 1 hour computer use per day was used as the reference group for the multinomial logistic regression. Grade was treated as a continuous variable as shown in Figure 1, indicating a linear relation between grade and computer use. The standardized score of FAS, a proxy for SES, was included as a control variable.

TABLE III.

COMPUTER USE: GENDER, GRADE AND SOCIOECONOMIC DIFFERENCES

	Computer Use
Predictors	OR	95% CI
Gendera
< 1 hour/day (ref)	--	--
1–2 hours/day	1.55	1.30–1.85
2–3 hours/day	1.47	1.22–1.76
>=3 hours/day	1.99	1.65–2.41
Gradeb
< 1 hour/day (ref)	--	--
1–2 hours/day	1.20	1.11–1.29
2–3 hours/day	1.34	1.23–1.46
>=3 hours/day	1.29	1.18–1.42
FASc
< 1 hour/day (ref)	--	--
1–2 hours/day	1.25	1.19–1.32
2–3 hours/day	1.30	1.24–1.36
>=3 hours/day	1.31	1.24–1.38

Notes.

^aMale was the referent for Gender.

^bGrade was treated as a continuous variable as Figure 1 indicating a linear relation between grade and computer use.

^cFAS was a standardized variable, with higher value for higher family affluence.

Results of odds ratio and their corresponding 95% confidence intervals show that females, older adolescents and adolescents from higher family affluence families were related to more frequent computer use. Compared to males, females were almost twice as likely to use computer for three or more hours per day (OR = 1.99, CI = [1.65, 2.41]). Adolescents are 29% more likely to use computer for three or more hours with an increase in one grade level (OR = 1.29, CI = [1.18–1.42]). Adolescents from higher affluent family families are 31% more likely to use computer for three or more hours per day (OR = 1.31, CI = [1.24, 1.38]).

C. Gender and Grade Differences in Internet Bullying and Victimization

Overall, 6.3% of adolescents (7.9% of males and 4.7% of females) reported at least one incidence of bullying others using computers, whereas 8.1% of adolescents (8.0% of males and 8.2% of females) had been bullied using computers. Stratified by gender and grade, the percentages of involvement in internet bullying and victimization are reported in Figure 2 and Figure 3.

Fig. 2.

Fig. 3.

As both graphs suggest nonlinear relations between grade and bullying and victimization, grade was categorized into two age groups, i.e., younger group (grade 6 through 8) and older group (grade 9 and 10) for the following logistic regressions. Since there was no significant interaction between gender and grade on either bullying or victimization, the interaction terms were removed from the final model. Results of the logistic regression of Internet bullying and victimization on gender and grade are reported in Table 4.

TABLE IV.

INTERNET BULLYING: GENDER AND GRADE DIFFERNECES

	Bullying		Victimization
Predictors	OR	95% CI	OR	95% CI
Gendera	0.58	0.46–0.72	1.02	0.83–1.25
Gradeb	0.74	0.53–1.03	0.62	0.47–0.93
FASc	1.00	0.93–1.10	1.03	0.97–1.10

Notes.

^aMale is the referent for Gender.

^bGrade was treated as a categorical variable as shown in Figure 2 and Figure 3, indicating a nonlinear relation between grade and bullying or victimization. Younger adolescents were the referent for Grade.

^cFAS was a standardized variable, with higher value for higher family affluence.

Females were 42% less likely than males to report bullying others on the Internet (OR = 0.58, CI = [0.46, 0.72]). However, there was no gender difference in Internet victimization. There were no grade differences in bullying others, but older adolescents reported less victimization (OR = 0.62, CI = [0.47, 0.93]). There was no FAS difference in involvement in either Internet bullying or victimization.

D. Relation between Computer Use and Internet Bullying

The results of the logistic regressions of Internet bullying on gender, grade, FAS, and computer use are reported in Table 5.

TABLE V.

INTERNET BULLYING AND COMPUTER USE

	Internet Bullying
	Bullying		Victimization
Predictorsa	OR	95% CI	OR	95% CI
Gender	0.54	0.43–0.68	0.95	0.77–1.16
Grade	0.70	0.50–0.96	0.57	0.43–0.76
FAS	0.98	0.92–1.05	1.00	0.94–1.07
Computer Use
< 1 hour/day	1	--	1	--
1–2 hours/day	0.86	0.56–1.31	1.01	0.69–1.48
2–3 hours/day	1.76	1.24–2.52	2.03	1.39–2.95
>=3 hours/day	2.00	1.29–3.08	2.24	1.64–3.05

Notes.

^aMales, younger adolescents and less than 1 hour per day were the referent for gender, grade and computer use respectively. FAS was a standardized continuous variable with higher value for higher family affluence.

Compared to adolescents who used computer less than 1 hour per day, adolescents with 1 to 2 hours computer use did not differ significantly in either Internet bullying or victimization. Adolescents who reported 2 to 3 hours or more than 3 hours of computer use were more likely to get involved in both Internet bullying and victimization. Compared to those who used computer for less than 1 hour per day, adolescents who reported 2 to 3 hours of computer use per day were 76% more likely to report Internet bullying and 103% more likely to report Internet victimization. Adolescents with 3 or more hours of computer use were twice as likely to report Internet bullying and 2.24 times more likely to report Internet victimization than those who used less than 1 hour of computer use per day.

Results on gender, grade, and FAS differences in bullying and victimization were similar to previous section (Table 4), with the only exception of a significant grade difference in bullying when level of computer use was controlled. An increase in one grade level was associated with 30% lower likelihood in bullying others through the Internet.

Because Table 5 suggests a cutoff point of 2 hours per day in the relation between computer use and both bullying and victimization, the logistic regression were rerun with the dichotomous variable of computer use (less than 2 hours vs. 2 or more hours). Results showed that compared to less than 2 hours per day, adolescents who spent 2 or more hours of computer use per day were more likely to bully others (OR = 2.02, 95% CI = [1.49–2.75]) and to be bullied by others via computers (OR = 2.14, 95% CI= [1.68–2.73]).

IV. DISCUSSION

The American Academy of Pediatrics released a guideline of no more than 2 hours of screen time (television and computer) per day in children and adolescents [15]. The current study examined one type of screen time - computer use, and its association with Internet bullying in a nationally representative sample of U.S. youth in grades 6 through 10. The results showed a high proportion of adolescents who use computers 2 or more hours per day, as well as a sizable proportion who were engaged in behaviors of bullying and victimization using computers.

We found that females spent more time on computer use than males. However, females were less likely to bully others using computers than males. This suggests that females may spend more time on computers for pro-social or academic purposes. For instance, emailing, chatting online and doing homework were measured by the computer use item in the present study. On the other hand, a higher likelihood for boys to bully others may reflect a more aggressive nature among boys than girls, which have been consistently found in traditional bullying. Unlike the gender difference in bullying others through the computer, there was no gender difference in experience with Internet victimization, or being bullied by others.

The pattern of a gradual increase in computer use with grade is consistent with previous studies [1]. As the involvement of bullying and victimization by grade was a nonlinear relation, we compared Internet bullying and victimization for younger (grade 6 through 8) versus older adolescents (grade 9 through 10). The results suggest that older adolescents were less likely to bully others or be bullied by others, after controlling for level of computer use. Similar to gender differences in computer use and Internet bullying, this suggests that older adolescents may spend more time on pro-social or academic behaviors via computers. It is important to note that the grade differences in bullying others through the computer were statistically significant only when level of computer use was controlled. This shows the importance of controlling for computer use in studies on variability in Internet bullying or cyber bullying.

Previous studies showed inconsistent results on the association between computer use and Internet bullying [3, 12, 14]. To best grasp the relation between computer use and Internet bullying, we first included computer use as a four-category variable (Table 5). Consistent with the guideline by the American Academy of Pediatrics [15], our results suggest that 2 or more hours of computer use per day would place adolescents at a higher risk of engaging in both Internet bullying and victimization. When level of computer use was dichotomized into less than 2 hours and 2 or more hours, we found that adolescents with 2 or more hours per day were at least as twice as likely to bully others or to be bullied by others using computers. The positive relations between computer use with both Internet bullying and victimization are consistent with the results found by Hinduja and Patchin [12], but are contrary to those by Ybarrra et al. [3] or Smith et al. [14]. Our results suggest that the relation between computer use and internet bullying may not have a linear association, which may explain the mixed findings in previous studies.

It is important to note the limitations of this study. First, the study was based on student self-report. Testing information from multiple sources on computer use and Internet bullying is recommended for future studies. Second, the measure of Internet bullying was limited to incidents that took place at school. Even though this might not influence the association between computer use and Internet bullying due to a high correlation of inside and outside school incidents, the prevalence of Internet bullying may be underestimated. It is recommended to examine Internet bullying both in school and outside school to obtain more precise estimate of prevalence rate of Internet bullying.

Nonetheless, this study contributes to this growing body of literature in at least two important ways. First, we assessed Internet bullying using the standard format as in the Olweus Bully/Victim Questionnaire, which has been used internationally to measure the traditional forms of bullying [18]. Second, we used a large-scale nationally representative sample with sufficient representation from multiple age groups to examine computer use and Internet bullying. This sample allows us to best grasp the linear or nonlinear relationship between grade, computer use and Internet bullying.

V. Conclusion

In conclusion, 39.2% of U.S. adolescents in grade 6 through 10 used computers for 2 or more hours per day and thus exceeded the guideline on limitation of screen time recommended by the American Academy of Pediatrics [15]. By exploring the association between computer use and Internet bullying, we found that adolescents with 2 or more hours of computer use were at least as twice as likely to be involved in Internet bullying or victimization. We also found that boys and younger adolescents were more likely to engage in Internet bullying. An important implication of these findings is the need to encourage a limit of 2 hours of computer use per day, especially among boys and younger adolescents.