Bullying occurs when a student is exposed, repeatedly and over time, to negative actions on the part of one or more other students.1,2 It usually begins in elementary school and peaks in middle school, with a decline in later school years.3 Exposure to bullying can occur either as a bully, victim, witness, or combination thereof. The detrimental impact of bullying is substantial and is of concern as it may negatively affect the victim’s scholastic achievement, desire to attend school, and self-esteem.4–7 Bullying behavior has been linked to the physical, psychosocial, and social well-being of children.3,4,8,9 Whereas bullying impacts the mental health of bullies, it also affects victims and bully-victims.4,10–14
Bullying/victimization has a higher prevalence among children with special health care needs, such as those with cancer,15 speech/language impairment,16 visual impairment (VI),17 and psychological impairment.18 Children with VI have been reported to be twice as likely to be bullied as their peers.19 This could be attributed to the fact that accessing visual information is a significant advantage for developing social understanding and that children with VI face difficulties with this skill.20 With increasing emphasis on inclusive education (i.e., children with special educational needs are educated alongside their typically developing peers in the same school), children with VI are perhaps more vulnerable to bullying and can have more difficulties forming social relationships with peers than typically developing peers. Given that bullying occurs as early as in elementary school years, implementing antibullying intervention programs (such as the Olweus Bullying Prevention Programs,21–23 Steps to Respect,24 and Let’s Get Real) very early, before bullying becomes a part of school culture, might improve the school’s ability to carry out its educational mission by improving the student’s ability to focus on learning and establishing an atmosphere of respect early on in life. Given that eye care providers, including optometrists, are highly likely to come across children with VI in their setups, they may play a pivotal role by working with schools to assist with the implementation of antibullying policies. However, before implementing antibullying programs, teachers, parents, and eye care professionals will require to identify the bullied student or the at-risk children (such as those with identifiable causes of VI, e.g., microphthalmos, and children who have been suggested to use low-vision devices). A few instruments for the assessment of bullying have been developed.25–27 One such instrument is the Olweus Bully/Victim Questionnaire (OBVQ).28 It has been widely used for understanding the prevalence of bullying29 and for evaluating the effects of bullying precautions and interventions.30 The OBVQ was developed, albeit using classical test theory, to measure victimization (i.e., being bullied) and bullying (bullying others) in normally sighted schoolchildren but has been revalidated using Rasch analysis (revised OBVQ) in Greek Cypriot normally sighted primary school students.31 Among other advantages of Rasch analysis, the most important include the greater insights it provides into the psychometric properties of an instrument, especially, if the items of the instruments target the spectrum of the overall construct being measured, that is, bullying/victimization.
Instruments should be ideally used only in those populations whom they have been developed for and validity in other populations cannot be assumed. However, instruments have been successfully used in populations other than those it was developed for.32,33 Encouraged by the results of such studies, and coupled with the benefits of Rasch analysis, we were interested in determining whether the OBVQ that was originally developed for the normally sighted children would demonstrate adequate measurement properties, using Rasch analysis, in school-aged children with VI in India. Therefore, this constituted the aim of our study. Having a reliable and valid instrument for the assessment of bullying could help in many ways: (1) measure the incidence of bullying in children with VI; (2) identify at-risk children; (3) make decisions about the most suitable kind of intervention for a particular student or for a group of students; (4) and develop antibullying intervention programs that could be used by school administrators, teachers, students, and/or parents.
The revised OBVQ consists of 40 items that assess various aspects of bully/victim problems:31 Specifically, these items relate to three forms of bullying: physical, verbal, and indirect, racial, sexual forms of bullying harassment; initiation of various forms of bullying other students; where the bullying occurs; probullying and provictim attitudes; and the extent to which teachers, peers, and parents are informed about and react to the bullying.34 Of the 40 items, we used only 16 items for purposes of Rasch analysis (Table 1). The reason for this choice was that these 16 items comprise the two main dimensions (eight items each) of the OBVQ: being victimized (i.e., the act of bullying against the child who is responding to the instrument) and bullying others (i.e., expression of bullying behavior against others by this child).35 Consequently, these two dimensions form two parts, I and II, of the OBVQ, wherein part I consists of eight items (5 to 12) that can be used to assess the extent to which the child is a victim of bullying, and part II consists of eight items (25 to 32) that can be used to assess the extent to which the child initiates an act of bullying against other children. Furthermore, these 16 items are similar to those used in the earlier Rasch analysis in a sample of Cyprus school-aged children.31 All the 16 items are rated by the participants on a five-category Likert scale, and the response categories are (1) It hasn’t happened to me in the past couple of months, (2) Only once or twice, (3) Two or three times a month, (4) About once a week, and (5) Several times a week. A higher score on both the parts indicates that the child is being victimized more often or the child bullies the other students more often. The remaining 24 items comprise the demographic details and other aspects such as by whom the child was bullied, who did the child bully, where the bullying has occurred, whether it was reported, any necessary action taken by the parent/guardian or teachers, and so on. The entire instrument was translated for use in this study into local languages (Hindi and Telugu) using standard accepted procedures. Rather than a literal translation of the instrument, we aimed for conceptual equivalence (cultural adaptation) to suit the local needs of the population. For example, given that the word “sexual” is not culturally appropriate for younger children in India, we replaced it with “vulgar.”
Participants were children with VI referred to the Vision Rehabilitation Centres, L. V. Prasad Eye Institute, India. Ethical approval was obtained from the L. V. Prasad Eye Institute Ethics Committee for Human Research. The purpose of the study was explained to each child and his or her parent/guardian, and informed consent was obtained from both the child and the parent/guardian. After informed consent, the revised OBVQ was either self- or interviewer administered (face to face) to the participant in a quiet room (without the presence of the parent) before any clinical examination.
Included participants were aged 8 to 16 years, had VI (best-corrected visual acuity in better eye, <20/60 or ≥20/60 with restricted visual fields) from any cause, had presence of single disability (visual), and could respond to the items on the instrument. Participants with additional disability (physical, hearing, etc.) were excluded, as those with total blindness in both eyes (bilateral absence of light perception). The study was conducted in accordance with the tenets of the Declaration of Helsinki. The sociodemographic data were extracted from the clinical records.
After administration of the OBVQ, all the participants underwent a comprehensive low-vision examination, the details of which have been previously published in this journal.36 Habitual monocular and binocular visual acuity was recorded for all participants using a high-contrast letter chart based on logMAR principles.
Rasch analysis37 was conducted using the Andrich Rating Scale model38 with Winsteps software (version 3.68.0).39 Rasch analysis transforms ordinal scores onto a logit (i.e., log odds) scale and allows for interval-level measurement. The procedures of Rasch analysis have been described in detail by the authors elsewhere in this journal.33,40,41 Four fundamental indicators were used to evaluate instrument quality.42 These included (1) behavior of response categories; (2) measurement precision or the extent to which the items reliably distinguish distinct levels of bullying in the participants (using person separation reliability [PSR] minimum recommended value, 0.80); (3) targeting or the extent to which the set of items is of appropriate endorsability (i.e., level of agreement or disagreement) for the level of participant’s bullying, and this was investigated using the person-item map; (4) fit or the extent that items in the OBVQ measured a single construct, that is, unidimensionality (using fit statistics with acceptable infit mean square [MnSq] range of 0.70 to 1.30 and principal component analysis [PCA] of residuals with the criterion that the first contrast should have an eigenvalue of >2.0 to be considered as second dimension, which was greater than the magnitude seen with random data).
Adequate PSR (0.80) constituted the minimum acceptable measurement properties of the Rasch model for both parts of the OBVQ to be termed as a measure.
Descriptive statistics were analyzed using SPSS software (version 16.0; Chicago, IL).
One hundred fifty children with VI provided responses to the OBVQ, and the sociodemographic characteristics of these participants are provided in Table 2. The instrument was interviewer administered in most of the participants (83%). The mean age (SD) of the participants was 11.6 (2.2) years (range, 8 to 16 years), and a little more than two-thirds (n = 103 [69%]) were male. There was a male preponderance in accordance with the slightly skewed distribution among sexes presenting to our center.41
Psychometric Properties of the OBVQ
The first step in the analysis of both parts I and II of the OBVQ was to check category response thresholds. The thresholds were disordered for both parts, which required reordering by category reorganization before further analysis. The percentage of category utilization for part I were 43% (1, “it hasn’t happened to me in the past couple of months”), 14% (2, “only once or twice”), 13% (3, two or three times a month”), 10% (4, “about once a week”), and 21% (5, “several times a week”). As is evident, the intermediate categories were used infrequently. Therefore, we began with examining each of them. To begin with, category 2 was underused; at no point was it more likely to be chosen over any other category. Therefore, we initially combined category 2 with 3 (“it has only happened once or twice” with “two or three times a month”), resulting in four response categories (1, 2, 2, 3, 4). However, disordering persisted, so further category collapsing was performed; revised category 2 was combined with 3 (“once or twice/two or three times a month” with “about once a week”). Finally, this resulted in three response categories (1, 2, 2, 2, 3), and ordered thresholds were obtained. In addition, the category utilization for the intermediate categories improved. The percentage category utilization rates were 43% (1, “it hasn’t happened to me in the past couple of months”), 37% (2, “only once or twice/two or three times a month/about once a week”), and 21% (3, “several times a week”).
Similarly, category reorganization was performed for part II. The percentage category utilization rates before category reorganization were 69% (1, “it hasn’t happened to me in the past couple of months”), 9% (2, “only once or twice”), 12% (3, two or three times a month”), 4% (4, “about once a week”), and 6% (5, “several times a week”). As is evident here, not only the intermediate categories but the end (upper) categories were also used infrequently. Therefore, we began with examining each of them. To begin with, we combined category 3 with 4 (“two or three times a month” and “about once a week”) that resulted in a four-response category format, that is, 1, 2, 3, 3, 4 (containing a revised category 3, i.e., two or three times a month/once a week). However, disordering of thresholds continued, so category 2 (“it has only happened once or twice”) was combined with the revised category 3. Finally, this category reorganization resulted in three response categories (1, 2, 2, 2, 3), and ordered thresholds were obtained. In addition, the category utilization for the intermediate categories improved, although the upper end category remained underutilized. However, we decided to retain the final three-category format given the ordered behavior of the thresholds. The percentage category utilization rates were 69% (1, “it hasn’t happened to me in the past couple of months”), 26% (2, “only once or twice/two or three times a month/about once a week”), and 6% (3, “several times a week”).
Unidimensionality and Targeting
All the items in part I fit the Rasch model well, indicating that, together, they were measuring a single underlying trait (i.e., victimization). This was reinforced by the PCA of the residuals that showed that the variance explained by the measures was comparable for the empirical calculation (41.1%) and by the model (39.9%). The unexplained variance explained by the first contrast was 1.7 eigenvalue units, which was greater than the magnitude seen with random data. Taken together, these findings confirmed the unidimensionality of part I. Targeting was −0.58 logits for part I, indicating that the item endorsability was matched well to the participant’s victimization. This is evident from the person-item map, where it can be seen that most of the items are concentrated where the participants are located (Fig. 1).
In part II, one item demonstrated misfit (infit MnSq 1.42; “I kept him/her out of things on purpose, excluded him or her from my group of friends, or completely ignored him or her”). Principal component analysis of residuals showed that the variance explained by the measures was comparable for the empirical calculation (34.4%) and by the model (35.3%). The unexplained variance explained by the first contrast was 1.9 eigenvalue units, which was greater than the magnitude seen with random data. The misfitting item was deleted, and following this, all the remaining items fit the Rasch model. Principal component analysis of residuals continued to demonstrate unidimensionality. Taken together, these findings suggest that part II was unidimensional.
However, targeting was −1.97 logits for part II, indicating that the items were poorly matched with the bullying behavior of the participants.
The PSR was 0.64 and 0.12 (0.19 for eight items for parts I and II, respectively. The person separation index was 2.10 and 0.98 (0.83 for eight items) for parts I and II, respectively. Taken together, these results indicate that instrument could not distinguish among at least three strata of the participant’s victimization and bullying others. Because of the larger number of participants, as compared with the items, as expected, the item reliabilities were comparatively higher (0.96 for part I and 0.80 for part II) than PSR.
The results of our study indicated that the revised OBVQ failed to meet the expectations of the Rasch model in this population of children with VI. The fundamental shortcoming of both parts of the OBVQ was that they lacked adequate measurement precision, that is, they did not possess sufficient discriminatory ability to reliably differentiate among children with VI who were being victimized and who bullied others, as was evidenced by inadequate PSR. Both the parts were only able to differentiate the participants into two groups, that is, lower versus higher endorsability of being bullied and bullying others. The associated lower reliability suggested that the user cannot have enough confidence in the item or person estimates. These findings are, however, at variance with the previous Rasch analysis of the OBVQ, albeit in normally sighted Cyprus children, that reported a PSR of 0.84 for both parts I and II.31
Intuitively, children with VI are more likely to be bullied rather than bully others. In our sample, most (70%) of these children were not involved in bullying other children. Therefore, the population was more homogeneous in terms of not being involved in bullying others. Given this finding, it is perhaps not surprising that part II of the OBVQ (bullying others) had a low discriminatory ability. As expected, the results of this aspect of the revised OBVQ (part II) are at odds with the Cyprus study that included healthy children who were bullied as well as were involved in bullying others.31
Although the PSR was inadequate among children with VI, it was relatively closer to the minimum recommended value of 0.8 for part I (being victimized) than for II (bullying others). It is plausible that PSR for part I may reach a value of 0.8 in another population, thereby, rendering it a reliable scale. Nonetheless, a simple strategy to increase the reliability would be to add items so as to increase the range of victimization behaviors that are associated with chronic VI in schoolchildren. However, addition of items lies within the purview of the developers of the OBVQ, so it was not pursued in the present study.
More importantly, all the items of part I fit the Rasch model, indicating that they were all contributing toward the measurement of a single construct, that is, this part was unidimensional. This was further reiterated by PCA of residuals. Unidimensionality is a fundamental requirement of the Rasch model. Indeed, recent observations that even slight multidimensionality can affect person estimates have further emphasized the importance of unidimensionality as a measurement property of an instrument.43 Although conventionally fit statistics have been used to assess dimensionality of an instrument, as has been done in the earlier Rasch analysis by the authors in the Cyprus study, fit statistics alone is insufficient and PCA of residuals is recommended for confirmation of unidimensionality.44 Therefore, this is the first study to have used PCA of residuals to investigate the dimensionality of the OBVQ. In addition, the behavior of the rating scale was not investigated in the earlier study. Among the various benefits offered by Rasch analysis is the detailed assessment of the use of response categories,42 and these were not used as was intended to by the developers for either part of the OBVQ. The five-category response format required shortening to a three-category format in the present study. Given that similar modifications have been recommended for other instruments used in children, our results are not surprising.41,45 Taken together, all these studies perhaps indicate that, as compared with adults, children can remember a relatively smaller number of categories.
Targeting of part I of the OBVQ was good (<1.0 logit46; Fig. 1), indicating that the items in this part matched well with the participants’ victimization behaviors. Therefore, it seems that given such good targeting, small modifications in part I such as addition of only a few appropriate items to improve its discriminatory ability (i.e., measurement precision) in children with VI would render it a reliable measure even in this population. Given that the OBVQ was developed in 1978, a time when usage of Internet and social media was not popular among children, items related to verbal, physical, and relational forms of bullying were only included. However, items related to cyberbullying47 (bullying over social media) should now be added given the ubiquity of Internet usage and social media among children in the present era. In addition, given that there were almost no items to target those at the lower end of the person-item map in Fig. 1 (i.e., those with less victimization, n = 22), a possible recommendation would be to add items in this range of less victimization. Targeting was considerably suboptimal in the Cyprus study of normally sighted children, thereby, forcing the authors to recommend the addition of difficult items to improve targeting.31
By comparison, part II of the OBVQ demonstrated item misfit (albeit a single item), indicating that it had an item that was not in tandem with the remaining items in the measurement of a single underlying construct. However, its deletion restored unidimensionality. Nonetheless, its targeting was suboptimal, indicating that the items in part II were not matched with the participants’ bullying behaviors (i.e., they were easily endorsed) and more difficult items were required. The concepts toward bullying have been reported to vary across countries and cultures,48 and we speculate that this could perhaps be one of the reasons for poor targeting in our population.
Other investigators that have used questionnaires in children have also found suboptimal functioning of questionnaires when they were used in populations other than those they were developed, and such reports exist for both adult as well as for pediatric populations. For example, Gothwal et al.33 used the Visual Function and Quality of Life (VF and QoL) Questionnaire developed specifically for an Indian adult cataract population to assess visual disability and QoL in an Australian cataract population. However, the authors found that the QoL aspect of the questionnaire was not a valid scale in the Australian sample. Lamoureux et al.45 used the pediatric quality of life inventory (PedsQL 4.0) to assess the impact of refractive errors on health-related quality of life (HRQoL) in preschool children in Singapore. The authors found that the PedsQL was not a valid psychometric scale to effectively evaluate the impact of refractive errors on HRQoL in that population.
The strengths of this study are the relatively large sample size from a single low-vision rehabilitation center of a tertiary facility. The use of Rasch analysis to validate the revised OBVQ available in local languages is another strong point of this study. There are some limitations of our study that deserve attention. First, the cross-sectional design prevents any causal relationships from being determined. Second, the mode of administration of the OBVQ (self- vs. interviewer administered) could have had some influence on the results. This is critical because children are reluctant to inform adults about being bullied,49 and this may have led to underreporting of bullying in our study. Poor literary skills of some participants and time constraint in the clinic were major considerations for interviewer administration. Third, given that our study was confined to a single urban tertiary eye care facility in South India and that bullying behaviors are known to vary across regions and cultures, there exists the possibility that the study population, patients presenting to a rehabilitation center, is different from students in a classroom setting and in the community. Consequently, our results may not apply to other cultures. However, there are no similarly designed studies for comparative purposes. These data may simply reflect the pervasiveness and complexity of the bullying culture that exists in schools. Fourth, despite the standard procedures including cognitive debriefing with a representative sample of children with VI that we adopted in the cultural adaptation and translations of the OBVQ, it is likely that the meaning of certain phrases and words was lost in translation. For example, we had difficulty in finding a suitable word for “sexual” that was consistently understood by children of all age groups. Fifth, most of the participants spoke and understood the local language (Telugu); we also had children from other states who had to be administered the OBVQ in another language (Hindi). It should be borne in mind that the way bullying behaviors and problems are expressed can vary from one region of the country to another. Sixth, the lack of items related to cyberbullying (perhaps more prevalent and harmful than bullying in person) in the OBVQ deserves to be highlighted. In our study, it is unlikely that children may have considered bullying via social media while responding to the items in the revised OBVQ. Finally, the range of VI among the participants was unequal across the mild, moderate, and severe VI groups; a little more than two-thirds of the participants had moderate VI, indicating a relatively large homogeneous sample, and this could have impacted our results. However, similar distribution of VI among study participants has been reported previously by us and by others from rehabilitation centers elsewhere, indicating the referral pattern of patients to these centers. Nonetheless, our study findings are useful in providing directions for future studies to assess bullying in children with VI.
It is acknowledged that there may be other factors associated with bullying involvement that have not been accounted for in our study—for example, self-esteem, HRQoL, social well-being, anxiety, and depression. Future work is needed to investigate their potential influence further.
In conclusion, this study demonstrated that the revised OBVQ is not a valid psychometric scale to assess bullying/victimization among children with VI in India. The instrument failed the fundamental prerequisite for it to be termed a reliable and valid measure. Future studies could consider if the revised OBVQ (at least part I that assesses being victimized) can be psychometrically reengineered for use in this population by including additional items. For example, items that address cyberbullying should be added given the recent boom in the information and communication technology sector all over the world. Another strategy to identify potential items could be to conduct focus group discussions involving children with VI of different age groups. However, there are other instruments such as the Life in School 50 ([LIS] secondary school version) and the My Life in School checklist16 (primary age version of LIS), Verbal Bullying Index (developed from LIS), Bullying and Friendship Interview schedule,27 and School Bullying Scales (SBS)51 that could equally be used. However, none of them, except the SBS, have been validated using Rasch analysis, albeit in a healthy population. However, similar to the OBVQ, the SBS will also need to be revalidated in children with VI given that it was developed for a different population. Although other instruments such as the LIS and so on can also be subjected to Rasch analysis, it can be a vicious cycle with little benefit, if any. An alternative and perhaps a superior strategy would be to develop and validate a bullying instrument in children with VI given that there is no robust measure to assess bullying/victimization in this population. The prevalence and burden of bullying among children with special educational needs such as those with VI makes this a pressing issue for rehabilitation practice, policy, and research.
Vijaya K. Gothwal
L. V. Prasad Eye Institute
Kallam Anji Reddy Campus
L. V. Prasad Marg Banjara Hills
Hyderabad 500034 Andhra Pradesh
This study was supported by the Hyderabad Eye Research Foundation, Hyderabad, India.
The authors have no personal financial interest in the development, production, or sale of any device discussed herein.
Received December 17, 2012; accepted March 19, 2013.
1. Fekkes M, Pijpers FI, Verloove-Vanhorick SP. Bullying behavior and associations with psychosomatic complaints and depression in victims. J Pediatr 2004; 144: 17–22.
2. Olweus D. Bullying at school: basic facts and effects of a school based intervention program. J Child Psychol Psychiatry 1994; 35: 1171–90.
3. Pellegrini AD, Long JD. A longitudinal study of bullying, dominance, and victimization during the transition from primary school through secondary school. Br J Dev Psychol 2002; 20: 259–80.
4. Nansel TR, Overpeck M, Pilla RS, Ruan WJ, Simons-Morton B, Scheidt P. Bullying behaviors among US youth: prevalence and association with psychosocial adjustment. JAMA 2001; 285: 2094–100.
5. Iuvonen J, Nishina A, Graham S. Peer harassment, psychological adjustment and school functioning in early adolescence. J Educ Psychol 2000; 92: 349–59.
6. Mynard H, Joseph S. Bully/victim problems and their association with Eysenck’s personality dimensions in 8 to 13 year-olds. Br J Educ Psychol 1997; 67 (Pt. 1): 51–4.
7. Dawkins JL. Bullying, physical disability and the paediatric patient. Dev Med Child Neurol 1996; 38: 603–12.
8. Salmon G, James A, Smith DM. Bullying in schools: self-reported anxiety, depression, and self esteem in secondary school children
. BMJ 1998; 317: 924–5.
9. Natvig GK, Albrektsen G, Qvarnstrom U. School-related stress experience as a risk factor for bullying behavior. J Youth Adolescence 2001; 30: 561–75.
10. Olweus D. Bullying at school: long-term outcomes for the victims and an effective school-based intervention program. In: Huesmann LR, ed. Aggressive Behavior: Current Perspectives. New York, NY: Plenum Press; 1994: 97–130.
11. Hawker DS, Boulton MJ. Twenty years’ research on peer victimization and psychosocial maladjustment: a meta-analytic review of cross-sectional studies. J Child Psychol Psychiatry 2000; 41: 441–55.
12. Schwartz D, Gorman AH, Nakamoto J, Toblin RL. Victimization in the peer group and children
’s academic functioning. J Educ Psychol 2005; 97: 425–35.
13. Luster T, Small SA, Lower R. The correlates of abuse and witnessing abuse among adolescents. J Interpers Violence 2002; 17: 1323–40.
14. Olweus D. Bullying among school children
: intervention and prevention. In: Peters R, McMahon RJ, Quinsey VL, ed. Aggression and Violence Through the Life Span. Thousand Oaks, CA: Sage Publications; 1992: 100–25.
15. Lahteenmaki PM, Huostila J, Hinkka S, Salmi TT. Childhood cancer patients at school. Eur J Cancer 2002; 38: 1227–40.
16. Conti-Ramsden G, Botting N. Social difficulties and victimization in children
with SLI at 11 years of age. J Speech Lang Hear Res 2004; 47: 145–61.
17. Horwood J, Waylen A, Herrick D, Williams C, Wolke D. Common visual defects and peer victimization in children
. Invest Ophthalmol Vis Sci 2005; 46: 1177–81.
18. Kumpulainen K, Rasanen E. Children
involved in bullying at elementary school age: their psychiatric symptoms and deviance in adolescence. An epidemiological sample. Child Abuse Negl 2000; 24: 1567–77.
19. Dunn J. Children
’s Friendships: The Beginnings of Intimacy. Malden, MA: Blackwell; 2004.
21. Olweus D. Bully-victim problems among school children
: Basic facts and effects of a school-based intervention program. In: Pepler DJ, Rubin KH, ed. The Development and Treatment of Childhood Aggression. Hillsdale, NJ: L. Erlbaum Associates; 1991: 411–48.
23. Olweus D. New Positive Results with the Olweus Bullying Prevention Program in 37 Oslo schools. Report. Bergen, Norway: Research Center for Health Promotion, University of Bergen; 2005.
24. Brown EC, Low S, Smith BH, Haggerty KP. Outcomes from a school-randomized controlled trial of Steps to Respect: a bullying prevention program. School Psychol Rev 2011; 40: 423–43.
25. Olweus D. Aggression in the Schools: Bullies and Whipping Boys. Washington, DC: Hemisphere Publishing Corporation; 1978.
26. Rigby K, Slee PT. Dimensions of interpersonal relation among Australian children
and implications for psychological well-being. J Soc Psychol 1993; 133: 33–42.
27. Wolke D, Woods S, Bloomfield L, Karstadt L. The association between direct and relational bullying and behaviour problems among primary school children
. J Child Psychol Psychiatry 2000; 41: 989–1002.
28. Olweus D. The Revised Olweus Bully/Victim Questionnaire
. Bergen, Norway: Research Center for Health Promotion, University of Bergen; 1996.
29. Currie C, Gabhainn SN, Godeau E, Roberts C, Smith R, Currie D, eds. Inequalities in young people’s health. Copenhagen, Denmark: World Health Organization; 2008.
30. Crothers LM, Levinson EM. Assessment of bullying: a review of methods and instruments. J Counsel Devel 2004; 82: 496–503.
31. Kyriakides L, Kaloyirou C, Lindsay G. An analysis of the Revised Olweus Bully/Victim Questionnaire
using the Rasch measurement model. Br J Educ Psychol 2006; 76: 781–801.
32. Pesudovs K, Caudle LE, Rees G, Lamoureux EL. Validity of a visual impairment
questionnaire in measuring cataract surgery outcomes. J Cataract Refract Surg 2008; 34: 925–33.
33. Gothwal VK, Wright TA, Lamoureux EL, Pesudovs K. Rasch analysis
of visual function and quality of life questionnaires. Optom Vis Sci 2009; 86: 1160–8.
34. Olweus D. Bully/victim problems in school: facts and intervention. Eur J Psych Educ 1997; 12: 495–510.
35. Bendixen M, Olweus D. Measuement of antisocial behavior in early adolescence and adolescence: psychometric properties and substantive findings. Criminal Behav Ment Health 1999; 9: 323–54.
36. Marella M, Gothwal VK, Pesudovs K, Lamoureux E. Validation of the visual disability questionnaire (VDQ) in India
. Optom Vis Sci 2009; 86: 826–35.
37. Rasch G. Probabilistic Models for Some Intelligence and Attainment Tests. Chicago, IL: University of Chicago Press; 1960.
38. Andrich DA. A rating scale formulation for ordered response categories. Psychometrika 1978; 43: 561–73.
39. Linacre JM. WINSTEPS Rasch Measurement [computer program]. Chicago, IL: Winsteps.com; 2009.
40. Gothwal VK, Wright TA, Lamoureux EL, Pesudovs K. Rasch analysis
of the quality of life and vision function questionnaire. Optom Vis Sci 2009; 86: 836–44.
41. Gothwal VK, Sumalini R, Bharani S, Reddy SP, Bagga DK. The second version of the L. V. Prasad-functional vision questionnaire. Optom Vis Sci 2012; 89: 1601–10.
42. Pesudovs K, Burr JM, Harley C, Elliott DB. The development, assessment, and selection of questionnaires. Optom Vis Sci 2007; 84: 663–74.
43. Tennant A, Pallant JF. Unidimensionality matters (a tale of two Smiths?). Rasch Meas Trans 2006;20:1048–51. Available at: http://www.rasch.org/rmt/rmt201c.htm
. Accessed April 3, 2013.
44. Gothwal VK, Wright TA, Lamoureux EL, Pesudovs K. Activities of Daily Vision Scale: what do the subscales measure? Invest Ophthalmol Vis Sci 2010; 51: 694–700.
45. Lamoureux EL, Marella M, Chang B, Dirani M, Kah-Guan AE, Chia A, Young TL, Wong TY, Saw SM. Is the pediatric quality of life inventory valid for use in preschool children
with refractive errors? Optom Vis Sci 2010; 87: 813–22.
46. Gothwal VK, Wright TA, Lamoureux EL, Pesudovs K. Cataract symptom score questionnaire: Rasch revalidation. Ophthalmic Epidemiol 2009; 16: 296–303.
47. Smith PK, Mahdavi J, Carvalho M, Fisher S, Russell S, Tippett N. Cyberbullying: its nature and impact in secondary school pupils. J Child Psychol Psychiatry 2008; 49: 376–85.
48. Smith PK, Cowie H, Olafsson RF, Liefooghe AP, Almeida A, Araki H, del Barrio C, Costabile A, Dekleva B, Houndoumadi A, Kim K, Olafsson RP, Ortega R, Pain J, Pateraki L, Schafer M, Singer M, Smorti A, Toda Y, Tomasson H, Wenxin Z. Definitions of bullying: a comparison of terms used, and age and gender differences, in a fourteen-country international comparison. Child Dev 2002; 73: 1119–33.
49. Craig WM. The relationship among bullying, victimization, depression, anxiety, and aggression in elementary school children
. Personality Individ Diff 1998; 24: 121–30.
50. Arora CM, Thompson DA. Defining bullying for a secondary school. Educ Child Psychol 1987; 4: 110–20.
51. Cheng YY, Chen LM, Liu KS, Chen YL. Development and psychometric evaluation of the School Bullying Scales: a Rasch measurement approach. Educ Psych Meas 2011; 71: 200–16.