General Versus Specific Methods for Classifying U.S. Students' Bullying Involvement: Investigating Classification Agreement, Prevalence Rates, and Concurrent Validity

The purpose of this study is to investigate the differential functionality of 2 general self-report items compared with 18 specific self-report items for classifying U.S. students' bullying involvement. First, by use of four samples from the Health Behavior in School-aged Children survey (HBSC; 1997–1998, N = 15,686; 2001–2002, N = 14,817; 2005–2006, N = 9,227; and 2009–2010, N = 12,642), prevalence rates of bullying involvement were calculated with classifications derived from general and specific items. Next, by use of only the most recent HBSC sample (2009–2010), concurrent validity of the classifications derived from general and specific items was investigated by exploring classification agreement as well as the association of each classification with self-reported student well-being indicators (i.e., life satisfaction, overall health, attitude toward school, academic performance, and general attention). Findings indicate that both bullying involvement classifications yielded drastically different prevalence rates but that trends for both methods suggested bullying involvement is decreasing over time. Results also show that general and specific classification methods had poor agreement but that there were no substantive differences in their associations with student well-being indicators. Implications for research and practice are discussed.