1. What did the results reveal? were the statistics significant or not and how do you know? provide actual numerical information (e.g., results of the statistical technique described).

Results Numbers and severity A 'yes' response to the question, 'Have you ever been bullied during your school years?' was provided by 81.13 percent of males and 72.16 percent of females (76.85 percent across sexes; hereafter called 'yes- responders'). A chi-square test revealed no male vs female differences in responding to this item (x2 = 2.29, 2d.f., p = .13). No age-related 8 Hoover et at; Victims' Perceptions of B ullying difference in the reporting of bullying episodes accrued, nor did age interact with sex. This may be due to the lack of age variability in the sample. Even though the denition of bullying specied prolonged occur- rence, some participants may have placed unwarranted emphasis on isolated incidents. Consequently, students were also asked to evaluate the impact (severity) of personally experienced bullying on a four-point Likert-type scale ranging from 1 (not at all a problem) to 4 (a very severe problem) in the somatic, social, emotional, academic and familial domains. When the combined scale's hypothetical midpoint of 2.5 was used to dene a severe response to mobbing, 15 males met this criterion (14.15 percent of all male subjects); 14 of 97 females also met this criterion (14.43 percent); clearly, no gender difference was observed. One-way between-subject ANOVAS were calculated for gender (SAS, 1984; General Linear Models Procedure). The .05 alpha level was adjusted to .0125 via the Bonferroni method in order to control for family-wide error rate (Pedhazur, 1982; this method was employed as appropriate in subsequent analyses). Using this criterion, the only sex difference was for 'emotional problem' (F [1,146] = 9.14, p = .003). Females scored higher (more problematic) in the 'emotionally troubled' domain (M males = 2.34, SD = 1.02; M females = 2.83, SD = 0.94). Males and females felt that bullying produced more problems in the social and emotional (M across sexes = 2.56, SD = 0.98) domains, with physical (M = 1.78, SD = 0.93), academic (M = 1.78, SD = 0.99) and familial (M = 1.64, SD = 0.88) problems rated less severe. The average score for all female yes-responders was past the midpoint (toward the 'problem' end of the scale) in the emotional domain. Given the emphasis on academic achievement in the American school reform movement, the number of students responding past the midpoint of the severity scale on the academic item was tabulated. On the basis of this criterion, 20 males (18.87 percent of the total sample) felt that school bullying pejoratively affected academic performance; 14 females, or 14.43 percent of the sample, fell into the same category. Worst ages for bullying Yes-responders were asked to rate the school grade during which they were most troubled with bullying (because American students are accustomed to speaking in terms of grade rather than age). Results are shown by gender in Figure 1 (American school grades are shown as average entry age). Because relatively fewer students had the opportu- nity to nominate grades 11 or 12 (ages 16, 38 percent, or 17, 14 percent) as most troublesome for bullying the prevalence of bullying could have been systematically underestimated for these ages. However, among the 29 16-year-old yes-responders, none listed ages 15 or 16 as problem 9 School Psychology International (1992), Vol. 13 M 0 Female Victims Male Victims .5 0| N of Nominations as Worst Year 01 ES 0, 56 7891011121314151617 Age of Worst Bullying Figure l Nominations as worst age for bullying by gender years; among 21 17-year-old yes-responders, only three nominated ages 15, 16 or 17. Thus, combined data probably provide a clear and accurate temporal picture of victimization by bullies. There was no signicant relationship between either respondent age and prevalence of victimization or between respondent age and year nominated as most difcult. Students felt they were most at risk of peer victimization from approximately ages 1014, corresponding to the late elementary and middle school years in the US. Though the pattern of problematic years for bullying is similar across gender, there is some indication that problems begin and persist a year later in males. Bullying behaviors Yes-responders rated ve forms of bullying. Ridicule/teasing, other verbal attacks, practical jokes, vandalism of property and physical attack were evaluated on a scale running from 1 (never occurred) to 4 (always occurred). One-way ANOVAs were calculated to test for 10 Hoover et al.. Victims' Perceptions of B ullying differences between male and female yes-responders on the ve behav- ioral dimensions. A signicant between-gender difference for ratings of 'ridicule and teasing' was observed (F[1,148] = 8.11, p = .005), with females (M = 3.15, SD = 0.80) scoring higher than males (M = 2.76, SD = 0.85). In the eyes of females, bullying takes a more verbal form than is true for males. Support for this is provided by a trend for males to experience more physical aggression, though this was not statistically interpretable given the adjustment for family-wide error rate (F[1,139] = 4.32, p = .040). Across gender, ridicule/teasing was the behavior most associated with bullying, followed by other verbal harassment (M = 2.60, SD = 1.04), practical jokes (M = 2.23, SD = 1.07), physical attack (M = 1.67, SD = 0.93) and property vandalism (M = 1.60, SD = 0.86). Characteristics perceived to motivate bullying Twenty-six possible factors motivating bullying were presented to subjects who were asked to rate them on a scale from 1 (a primary reason that I was bullied) to 4 (I was not bullied for this reason). The ve items rated highest by males were 'didn't t in', 'physically weak', 'short tempered', 'who my friends were', and 'the clothes I wore'. Female victims frequently attributed their status to 'didn't t in', 'facial appearance', 'cried/was emotional', 'overweight' and 'good grades'. Few qualitative differences between genders emerged; one item appeared only among the ten highest rated by males ('too short') and one item appeared only on the female list ('I cried/was emotional'). 'Good grades' was frequently cited as a motivator for peer abuse, while bad grades and placement in special education were listed infrequently. It is possible that fewer special education students or individuals with a history of enrollment in special or remedial educa- tion participated in the study. Responses of school personnel Subjects viewing themselves as victims were asked to rate the response of school officials to bullying episodes. Eight yes-responders (5.6 per- cent) felt that school personnel responded very well, while 40 (28 percent) felt that school personnel responded adequately. By far the majority of self-reported victims (95 or 66.4 percent) responded that ofcials responded to episodes of bullying poorly. A chi-square test revealed that these respouses were not randomly distributed (12, 2 d.f. = 81.2, p