SYMPOSIA PAPER Published: 01 January 2005

Using Signal Detection Theory as a Model to Evaluate Release/Retention Criteria in Alpine Skiing


This study uses a methodology similar to Signal Detective Theory (SDT) to empirically evaluate the effectiveness of current release value recommendations as set by ASTM F 939 and ISO 8061. A ski-binding-boot system can be viewed as a signal detection device that must make a binary decision (release or retain) on the basis of a signal (the load passing through the boot-binding interfaces). The study population is what has come to be known as the Sugarbush study. This study of alpine ski injuries began in 1972 and continues to this day. For this analysis, we used the first 30 seasons of data (1972/73 through 2002/03) that consists of 6.08 million skier visits, 17 184 recorded injuries, 3453 controls, and more than 16 000 parking lot surveys. In this analysis, the rate of injuries due to self identified inadvertent releases (also known as ‘False Alarms’ or ‘FAs’) and the rate of mid-shaft tibial fractures (also known as ‘Misses’) were calculated and compared. Several analyses were made. SDT considers both failure modes: misses and false alarms. The results show strong evidence of improvement in SDT first 15 seasons and evidence of deterioration of the standard of care in the last 15 seasons versus the first 15 seasons. When the analysis was limited to the last ten seasons (1993/94 through 2002/03), we found that for all males versus all females the rate of the two failure modes was about equal for females, but males experienced about 50 % more FAs than Misses. This finding is significant at p<0.05. The analysis for different age groups indicates that for the 0–9 age group, the p(Miss) is 14.75 times higher than p(FA), which is significant at p<0.001. For the 10–16 age group, the p(Miss) is 1.26 times higher than p(FA), and this difference is not significant. For the 17–49 age group, the p(Miss) is 0.43 times p(FA), which is significant at p<0.005. For the 50 and older group, p(Miss) is 0.93 times p(FA), which is not significant. The binding release setting points for males and females under 10 appear to be too high in the sense that p(Miss) is much more likely than p(FA). ChiSq = 20.81. For skiers 10–16 years of age, no significant difference was noted between Miss and FA rates for males or females, or combined; thus, there is no basis for recommending any change in the setting recommendations for this group. Males 17–49, as well as 50 and older, are more likely to experience a FA than a Miss (2.1 × to 2.4 × higher). Females 17–49 are also more likely to experience a FA than a Miss (1.57 × higher). For both male and female 17–49 year age groups, the difference in FA and Miss rates is highly significant. Thus, skiers in the 17–49 age range are about twice as likely to experience a FA as a Miss, suggesting that the binding release setting points for these skiers may already be too low because FAs are more likely than Misses. This suggests a need to consider higher, not lower, settings for this age group, especially for males. For skiers 50 and over, there are no significant differences in the rates, so no change is recommended.

Author Information

Shealy, JE
Rochester Institute of Technology, Rochester, NY, USA
Ettlinger, CF
Vermont Safety Research, VT, USA
Johnson, RJ
University of Vermont College of Medicine, Burlington, VT, USA
Price: $25.00
Contact Sales
Reprints and Permissions
Reprints and copyright permissions can be requested through the
Copyright Clearance Center
Developed by Committee: F27
Pages: 1–12
DOI: 10.1520/STP12206S
ISBN-EB: 978-0-8031-5508-4
ISBN-13: 978-0-8031-3490-4