Research design refers to the way researchers plot their study in terms of the findings they intend to draw from it (Miles, & Huberman 1994, p. 56). The qualitative research aims at exploring the nature and effects of indirect aggression among teenage girls in details. It also intends to study the existent intervention measures and further come up with more effective intervention schemes that will ensure the containment of the vice.
To achieve its objectives, it uses a co-relational design, which relates aggression with social and psychological factors such as rejection, loneliness, depression and isolation. As a result, the researchers are able to come up with a model containing the explanations of why girls indulge in indirect aggression, what they do while indulging in it and the effects of the aggression on victims and bystanders.
They also provide the characteristics in victims that make them susceptible to victimization, and intervention measures that are existent, as well as those that need further investigation and application. Evidently, this is a wise design to use, as it incorporates every aspect of the study within an observable model while also linking the various categories to portray the way they relate to each other.
Moreover, such a co-relational design leaves room for inclusion of new categories that a researcher may think of later in the research. For instance, in this case, the characteristics of the victims of indirect aggression were an afterthought, yet it fit neatly into the final model, and provided additional insight into the nature of the indirect aggression.
The quantitative study’s main aim was to establish the relationship between recurrent peer victimization and the onset of self-reported symptoms of anxiety and / or depression in early teen years (Salmon, & Smith 1998, p. 984).
To achieve this purpose, the researchers applied a prospective study, which was a cohort study carried out over two years. This design was not the best that could have been applied for such a research especially because it does not give enough time for the participants to exhibit generalizable results.
Two years is not sufficient to come up with a report on the effect of bullying on the emotional well-being of teenagers because for most of these adolescents, bullying had occurred way before grade eight and would continue to occur long after grade nine, with possibly radical effects on their emotional wellbeing. Setting such tight parameters compromise a study, as the information collected seems too much regulated, and one can miss important peripheral occurrences.
However, the current study was relevant because of this design, for the scope of the study covered as well as the results were accurate, and the conclusions drawn were correct. It would be better to incorporate more time or duration in future so that the participants report on general bullying experiences throughout their lives so that one can draw conclusive results on the effects of the said bullying on the emotional wellbeing of the participants.
A good example of this deficit is evident in the self-reported incidences of symptoms of anxiety and/or depression during the first and second waves of the study in year 8. The researchers used to make the conclusion that a weak emotional build-up does not automatically make the person susceptible to bullying, or is not the ingredient necessary for the formation of a vicious cycle.
Whereas this conclusion may seem correct, it is not entirely conclusive because the researchers did not look into the previous lives of the students with these self-reported symptoms (Salmon, & Smith 1998, p. 485). No reference was evident as to whether they had been victims of previous bullying although the conclusion of the study asserts just that.
The method (s) used in the respective researches in another ground for establishing their strengths and weaknesses. The qualitative research used focus groups to hold discussions among the girls. It also used paired interviews as extensions of focus groups, to follow up on individuals who had more information to give, and the girls preferred for interview in pairs instead of an individual interview.
In addition, it applied individual interviews for the key teachers, who were guaranteed to be in close connection with the girls and hence aware of the nature of indirect aggression.
The selected teachers were school counselors, grade level coordinators and pastoral (homeroom) teachers. The researchers selected a pilot group, or control group of six sixteen-year old students from a coeducational government middle-class school in metropolitan Adelaide, Southern Australia.
This group assisted in developing and revising of the procedures for use in the research. It also used the findings derived from them alongside those of the main group. The main group consisted of fifty-four 15-year old girls randomly selected from two catholic schools in Southern Australia. They constituted the focus group members, who were 6-8 per group. The pair interviewees were grade 10 girls who voluntarily offered to participate and they were six in total.
Finally, key teachers were chosen based on “those from whom the girls may seek support during friendship or personal difficulty” (Owens, Shute, & Slee 2000, p. 77) and individual interviews were administered on these to establish their understanding on the nature, effects and possible intervention strategies for indirect aggression among teenage girls. The focus groups lasted 45 minutes during which the moderators administered a vignette / case study to each participant and then read it aloud before initiating discussion.
The case study was on a fifteen-year-old girl called Jo and the participants easily identified with the case at hand, making it easier to gather information from them. The teachers had been briefed by the first author on the nature and objectives of the study making it unnecessary to use the Jo vignette during the interviews.
The focus groups were conducted in a tutorial building close to the classrooms in both schools. This method was highly effective, as will be proved by the findings obtained. However, the sample was too small (less than 100 participants) which makes the data collected less generalizable. Consequently, although very conclusive, one cannot project the findings of this research on all teenage girls.
This method was also very accommodating, as it did not restrict the responses of the participants. In addition to the semi-structured focus groups, it administered unstructured pair interviews where the participants could talk about other issues that they could not have raised within the structure of the focus group discussions. It also incorporated the opinions if teachers which acts as a further validity of the findings of this research.
The quantitative research also had a control group, “a cluster randomized control design used for allocation of education districts to intervention or control status” (Bond, Carlin, Thomas, Rubin, & Patton 2001, p. 480). The report provides a detailed description of the sampling process. The researcher used self-administered questionnaires, which they presented to the participating students on researchers’ laptops. That is the first weaknesses.
There is a possibility that some students may not have known how to navigate on a laptop, which consequently could have compromised the data input. The students had 40 minutes to fill out these questionnaires, another weakness, the time necessary may have surpassed this limit, and finally, the students who were absent at the time of the survey were contacted via telephone or their study conducted on a later date.
This may also have compromised the accuracy of the data obtained due to a non-uniform collection procedure. However, this study involved a larger sample of students from a wider geographical area. It had 3623 students participating from different schools in 12 districts. This makes the findings more generalizable and hence more applicable in terms of policy or regulatory reform considerations.
The strength of this study was in the incorporation of an “ethics approval from the Royal Children’s Hospital Ethics in human research committee, as well as from the Victorian department of Education, Employment, and Training, and the Catholic Education Office” (Owens, Shute, & Slee 2000, p. 80). It is not clear whether the qualitative study did not obtain such approval or did not report on it but such approval is crucial for the validation of any study.
The approach to analysis varied for both types of research. For instance, in the qualitative research, the focus group discussions and the interviews were audio taped, and the transcripts were analyzed after collection of data was complete. The analysis was conducted by a NUD.
IST software program which had been preset with the predetermined categories of the study, and so the information was simply input into these categories. This is advantageous in that qualitative studies often gather a lot of information (bulky), which may be difficult to analyze, but the use of a software program assists in saving time and other resources such as human labor.
However, it can be disadvantageous because many misrepresentations could occur during data input and due to intrinsic shortcomings of such a program that in turn compromise the accuracy of the data collected. However, the researchers subdivided the data collected into major categories that represented the objectives that the researchers had set out to achieve.
The researchers judged the rigor of the study by applying four criteria: credibility – judging whether the participants can identify with the subject of the study as a relevant human experience. Requesting 12 participating students to write 10-page summaries on the study, and seven teachers to compile a 20-page summary report on the study did this.
Another criterion is ‘fittingness’ which refers to the applicability of the findings to other people outside the study. The researchers did this by collecting feedback and conducting follow-up interviews with six teenage girls from three schools that were not part of the study, and six teachers from six unrelated secondary schools in the vicinity. Third is ‘auditability’ or the ability of another researcher to follow the procedures of the research and get similar results.
This was possible based on the detailed systematic instructions on the procedure for this research, the availability of materials justifying the study, and an explanation of the analytical procedure. Finally, ‘confirmability’, which is satisfied when all the above three criteria are present. In the quantitative analysis, there were also subcategories, which fell under certain definitions of terms.
The researchers conducted the analysis on all the information collected. However, they based the results “on data gathered from the participants who gave information on victimization and mental health status for all the waves (2365) and for those whose missing data was for either wave 1 or 2 only (194)” (Owens, Shute, & Slee 2000, p. 89).
The researchers based the analysis of the second group’s data on the assumption that there had been no bullying or symptoms of anxiety or depression in the wave. Multiple logistic regressions modeled, and hence controlled the potential confounding effects, which ensured the accuracy of the findings.
This is a necessary tool to use in such a case of a ‘cause and effect’ study to be sure that a certain phenomenon is the only possible cause of a respective effect, otherwise, if there are multiple possibilities of other explanations in terms of causes, then the alleged cause will lose viability. This is a strength depicted by this quantitative research. The study also used various variables to depict various relationships and the occurrence of phenomena.
For instance, ‘victimization’, as used referred to the group that responded in the affirmative to having faced any of the “four types of recent bullying, namely teasing, having rumors spread about them, deliberate exclusion, and experiencing threats and violence” (Bond et al 2001, p. 481). The researchers deduced ‘Recurrent victimization’ from respondents who self reported symptoms of anxiety or depression during both study times (wave 1 and wave 2) in year 8.
The researchers evaluated Mental health status using “a structured psychiatric interview for non-clinical populations” (Bond et al 2001, p. 482). They used a simpler version of it, which seemed easy for adolescents to understand. Those participants who scored >_12 at both waves in year 8 were deemed to have ‘recurrent self-reported symptoms of anxiety or depression” (Bond et al 2001, p. 484).
Another subcategory was ‘social relations’, described as “having someone to talk to or depend on when angry or upset,” or “when having a tough time,” and “having someone who knows one well and worth trusting with private feelings and thoughts” (Bond et al 2001, p. 482). Finally, the last variable was ‘family measures’, which defined societal systems into integral households, alienated or detached parents, additional situations, and customs arrived at from the tongue spoken in the various abodes.
Clearly, this research type had a more conclusive analysis than the qualitative analysis with various interrelated variables present to contribute to or simply explain the findings. However, since from the beginning, it restricted itself to certain parameters of study in terms of scope, these additional variables do not fulfill their optimal potential of examining other possible relationships affecting the results.
Results from the qualitative research were a reflection of the set out objectives of the study. They tabled detailed explanations of why teenage girls engage in indirect aggression, how they do it, and the effects that this vice has on victims and bystanders, as well as characteristics of susceptible victims, and the existent intervention measures. These were presented in a simple, easily understandable format, which included testimonials from the participants as further backing of the validity of the results.
However, the general presentation of these results does not portray any correlation between various phenomena. They were simply explanations of certain behavioral types, and they fail to put the reader in perspective of the big picture. They are not specific, for instance they do not indicate how many of the 54 participating students felt that intervention was “a joke” (Owens, Shute, & Slee 2000, p. 76). This generalization of data makes it lose some of its authority.
For the quantitative study, the researchers presented the results in the form of statistics. The sample consisted of 3623 students, of these, 2860 participated in at least one of the waves of data, and 2559 provided data for analysis (Bond et al 2001, p. 483). There were significant differences in socio-demographic factors more common among boys, students who belonged to non-intact families, and students whose first language was not English.
The prevalence of victimization was 49% in wave 1, 51% in wave 2, and 42% in wave 3 (Bond et al 2001, p. 483). Even after the adjustment for social relations factors, a history of victimization remained the main cause of self-reported anxiety or depression. It was also evident that the effect of bullying on mental health status was most apparent in girls, probably because boys react differently to aggression (Lagerspetz, Bjorkqvist, & Peltonen 1988, p. 409).
These findings seem more detailed based on their portrayal as better compared to the generalized and oversimplified findings of the qualitative research (Miles, & Huberman 1994, p. 45). This is because they report on a specified range or scope of information. In addition to this, one can generalize them because the researchers used a larger sample.
The conclusion and recommendations made in the qualitative research included a suggestion for a study of peer meditation as an intervention measure. The researchers arrived at this point after a realization that policies, regulations and adult interference, as ways of redressing indirect aggression, are counterproductive. The quantitative research concluded that victimization adversely affects the mental health status of adolescents, and especially teenage girls.
It recommended a study into the effect of reducing victimization on the onset of symptoms of anxiety and depression among adolescents (Olweus 1991, p.43). In my opinion, this is a rather obvious suggestion worth seeing, as the study at hand already proved victimization as responsible for these mental conditions. The restrictive parameters limited the quantitative researchers from making any other relevant recommendation, a situation that was not similar to the qualitative researchers who were free to recommend any field.
Reference List
Bond, L. et al., 2001. Does bullying cause emotional problems? A prospective study of young teenagers. BMJ , 323 (12), pp. 480-484.
Lagerspetz, K., Bjorkqvist, K., & Peltonen, T., 1988. Is indirect aggression typical of females? gender differences in aggressiveness in 11 to 12 yearold children. Aggressive Behavior , 14 (3), pp. 403-414.
Miles, B., & Huberman, A., 1994. Qualitative data analysis: an expanded sourcebook. 2nd ed. Thousand Oaks, CA: Sage Publishers.
Olweus, D., 1991. Bully/victim problems among schoolchildren: basic facts and effects of a school based intervention program. Hillsdale, NJ: Lawrence Erlbaum Associates.
Owens, L., Shute, R., & Slee, P., 2000. Guess what I just heard: Indirect Aggression among Teenage Girls in Australia. Aggressive Behaviour , 26 (2), pp. 67-83.
Salmon, G., & Smith, G., 1998. Bullying in Schools: Self-Reported anxiety, depression, and self esteem in secondary school children. BMJ , 317 (5), pp. 984-985.