Peer-reviewed article
Lifetime Prevalence of Nonsuicidal Self-Injury in Adolescents: Reliability of a Dichotomous Survey Item.
Aspeqvist E, Andersson H, Dahlström Ö, Korhonen L, Zetterqvist M ·
Summary
Written by AI, not yet AICS's view · claude-haiku-4-5-20251001This study examined how reliably adolescents report whether they have ever engaged in nonsuicidal self-injury (NSSI) when asked the same question multiple times over 16 months. The researchers followed 253 Swedish teenagers aged around 14 years and found that one in five changed their answer at some point. Notably, more than half of those who initially said they had self-injured later denied it. The findings suggest that single survey questions about lifetime NSSI history produce only moderately reliable answers and should be interpreted cautiously.
Key points
- 18.97% of adolescents changed their answer about lifetime NSSI history across five measurements over 16 months
- 8.95% of those who initially denied NSSI later acknowledged it
- 57.69% of those who initially acknowledged NSSI later answered negatively
- Fleiss' kappa across all five measurements indicated moderate agreement
- Single-item dichotomous survey questions may not reliably capture NSSI prevalence in adolescents
Limitations
The study was conducted in a single Swedish community sample and examines only the reliability of measurement, not the actual prevalence or causes of NSSI. It does not explain why adolescents changed their answers or whether this reflects memory issues, changes in how they interpret the question, or actual changes in behaviour.
Abstract
Introduction Nonsuicidal self-injury (NSSI) is a widespread phenomenon among adolescents. Accurate estimates of prevalence rates when screening for NSSI are essential both in research and clinical settings. However, high variability raises the question of the reliability of self-reported NSSI prevalence. Methods The reliability of a dichotomous survey item on history of NSSI was investigated longitudinally using data from five time points across 16 months in a community sample (N = 253) of Swedish adolescents (age M = 14.21 years, SD = 0.56, 58.5% female). Results About one in five participants (18.97%) changed their answer compared to their initial response. Of those who answered negatively on lifetime prevalence of NSSI at baseline, 8.95% later acknowledged NSSI, while 57.69% of those initially acknowledging NSSI later answered negatively. Fleiss' kappa across all five measurements indicated moderate agreement. Cohen's kappa between baseline and 12 months indicated moderate agreement. Conclusions In this adolescent sample, the reliability of a single-item question on NSSI history was moderate. More than half of those who initially acknowledged a history of NSSI answered negatively at a later time point. Interpretation of self-reported lifetime prevalence rates of NSSI in adolescents using dichotomous survey items needs to be made cautiously.