The Hazardous Event Scale (HES;
Hunter, 1995) is a 10-item scale that assesses participants’ involvement
in hazardous aviation events. This scale is based on the generally held
premise that in most instances errors and other hazardous events do not
lead to accidents because, for any of a number of reasons, the causal
chain is broken. However, had the circumstances been slightly different
(a little less fuel on board, a little heavier load, visibility slightly
worse), then an accident might have occurred. It has been suggested that,
disregarding any arguments about accident-prone personalities, those individuals
who place themselves in hazardous situations more often are more likely
to eventually experience an accident. The HES is proposed as either a
predictor of eventual accident involvement, or as a surrogate measures
in studies in which the preferred criterion (actual accident involvement)
cannot be used.
Participants are asked to indicate
how often during the previous 24 months they had been involved in one
of the 10 events.
The response scale ranges from
0 to 6 or more. Responses of 6 or more are recorded as 6. Higher scores
indicate the person had experienced more hazardous events.
NOTE: Some implementations
of this scale only use “0, 1, 2, 3, 4 or more” as the response
In the earliest study that included the items comprising the HES, Hunter
(1995) conducted a nationwide statistical probability survey of 19,657
pilots. The response alternatives were 0, 1, 2, 3, 4, 5, and 6 or more.
Responses for the nine hazardous events items (excluding accident involvement
and items that did not appear in subsequent studies) were summed to produce
a HES score (M = 6.89, SD = 5.91). For this sample, the HES correlated
r = .23 (p < .01) with self-reported accident involvement. Coefficient
alpha for this sample and nine-item scale was 0.65.
As part of a web-based study of risk perception and risk tolerance among
pilots, Hunter (2002) obtained item responses from 394 pilots on the ten-item
version of the HES. To create a HES score, we excluded the item that asked
about accident involvement and computed the sum of the nine remaining
items. The resulting HES score had a mean = 3.16 (SD = 3.42). Pilots’
responses to the first (excluded) item from the10-item scale served as
the criterion. As before, that variable was coded as zero or one (no accident
reported, one or more accidents reported). Of the 394 respondents, 18
(4.6%) reported having been in an aviation accident during the previous
24 months. The HES score correlated r = .32 (p < .01) with this measure
of self-reported accident involvement. Coefficient alpha for this sample
and nine-item scale was 0.65.
Lubner, Adams, Hunter, Sindoni, and Hellman (2003) conducted a retrospective,
anonymous, case-control survey of 4,000 active U.S. pilots with a 42%
response rate. As part of this study, they administered seven items from
Hunter’s (1995) HES. They computed a HES score by taking the sum
of the responses for those seven items. The mean for this HES score was
5.14 (SD = 4.96). For this study, two measures of accident involvement
were available: self-reported accident involvement and case-control status
based on NTSB records. Of the 1,333 pilots, 597 (44.8%) reported having
been in an aviation accident at some point during their flying career.
The HES score correlated r = .24 (p < .01) for this self-report measure.
Correlating the HES score with case-control status yielded a correlation
r = .26 (p < .01) for a sample 1,401 pilots. Coefficient alpha for
this sample and seven-item scale was 0.64.
The ten-item HES was administered as part of a study to develop and evaluate
a situational judgment test (SJT) for general aviation pilots (Hunter,
2003). The SJT and the HES were administered to 467 participants over
the Internet. Item response data on the HES were available for 446 participants
and these data were analyzed, as described above, to create a HES score
(M = 3.06; SD = 3.39) based on nine of the items. The item concerning
accident involvement was recoded (zero or one) as before. Of the 446 pilots,
24 (5.4%) reported having been in an accident during the previous 24 months.
The correlation between HES and the self-reported accident status was
r = .30 (p < .01). Coefficient alpha for this sample and nine-item
scale was 0.65.
Hunter, D. R. (1995). Airman
research questionnaire: Methodology and overall results. (DOT/FAA/AM-95/27).
Washington, DC: Federal Aviation Administration.
Hunter, D. R. (2002) Risk perception and risk tolerance in aircraft pilots.
(DOT/FAA/AM-02/17). Washington, DC: Federal Aviation Administration.
Hunter, D.R. (2003). Measuring general aviation pilot judgment using a
situational judgment technique. International Journal of Aviation Psychology,
Hunter, D.R. (2006). Risk perception among general aviation pilots. International
Journal of Aviation Psychology, 16, 135-144.
Lubner, M., Adams. R., Hunter, D. R., Sindoni, R.& Hellman, F. (2003)
Risks for aviation occurrences among U.S. pilots by pilot training. 12th
International Symposium on Aviation Psychology, Dayton, OH: April 14-17,