chapter  2
17 Pages

Finding out about bullying and why statistics can be misleading

WithPeter K. Smith

Meta-analyses make sense once a large number of individual researches on a topic have been published. The great majority present quantitative data, with a small minority presenting qualitative data (or occasionally, both). These two kinds of data have their own advantages and disadvantages; actually a mixed methods approach can sometimes be an excellent way forward. Interview material can provide detailed information on how someone has experienced bullying, how they felt, what they did about it and what happened. It can also give insight into how things developed over time – how the bullying started, how it changed. The majority of research studies, however, have used quantitative methods – essentially, getting numerical data and often using statistics to see how generalizable the findings are. The two main methods have been questionnaire surveys and peer nominations. A number of surveys have attempted cross-national comparisons, giving the same questionnaire to young people in a range of countries across the globe.