Journal of Outdoor Recreation and Tourism 13 (2016) 49–56

Contents lists available at ScienceDirect

Journal of Outdoor Recreation and Tourism

journal homepage: www.elsevier.com/locate/jort

Accident causes and organizational culture among avalanche

professionals
Jerry Johnson a,b,n, Pascal Haegeli c,d, Jordy Hendrikx b, Scott Savage e
a
Political Science, Montana State University, United States
b
Snow and Avalanche Laboratory, Montana State University, United States
c
School for Resource and Environmental Management, Simon Fraser University, Canada
d
Avisualanche Consulting, Canada
e
Sawtooth Avalanche Center, United States

art ic l e i nf o a b s t r a c t

Article history: We report on a study of 392 avalanche professionals (AVPRO). We describe their demographics, organi-
Received 18 March 2015 zational work environment, and the causes and incidence of accidents. We find evidence of strong and
Received in revised form weak cultures of safety among AVPRO organizations and analyze differences between the two with respect
6 November 2015
to avalanche safety work procedures, personal work skills and attitudes, and causes of accidents. Demo-
Accepted 19 November 2015
graphics between the two groups are not different but the perceived causes for accidents were. Those
organizations we classify as having a strong culture of safety are identified by their employees as having
Keywords: better avalanche training and reporting procedures and, more positive working behaviors. With respect to
Avalanche accidents accident causes, we report that “operational pressures” and “management overriding personal judgement”
Organizational culture
during operations were reported by those in organizations with a weak culture of safety as contributing
Culture of safety
factors. Whereas we find value in large scale surveys of the AVPRO industry, we acknowledge that alter-
Avalanche professionals
Worker safety native methods of understanding of organizational culture AVPROs exist and should be utilized.
Risk management
M a n a g e m en t i m p l i c a t i o n s

● AVPRO organizations differ with respect to the culture of safety inherent in the organization. While we
find no organization with no culture of safety, we can identify them as strong vs. weak.
● AVPRO managers should be cognizant that organizational culture may influence how AVPROs do their
job and that causes of accidents may be correlated to a restrictive management culture.
● Communication and better personal decision making are encouraged where an organizational culture
of safety exists.
● High risk workers have a high level of self-efficacy independent of management, indicating that their
professional and personal skills can be enhanced where management recognizes and rewards in-
dependent behavior.
● Because of the dynamic conditions in which the AVPRO mountain community operates, accident
causes are rarely simple and linear. They are more likely to be a combination of personal and orga-
nizational factors. As such, managers and members of the AVPRO community would benefit from
greater understanding of the role of organizational culture in the workplace and could benefit from
research in other risk oriented professions.
& 2015 Elsevier Ltd. All rights reserved.

1. Introduction often work in high-risk environments where on-the-job injury, or

Members of the professional avalanche community (AVPRO)
n
Corresponding author at: Department of Political Science, Wilson Hall, Montana
State University, Bozeman, Montana, 59717, United States.
E-mail address: jdj@montana.edu (J. Johnson).
worse, is a day-to-day possibility. In the United States of America,
since 1950, 59 avalanche workers have been killed by avalanches
while at work and constitute 3.8% of the total avalanche deaths in
North America (Greene, Jamieson, & Logan, 2014). Many more
have been hurt on the job, some with career ending injury.
The AVPRO community has a long history of focusing on the

http://dx.doi.org/10.1016/j.jort.2015.11.003
2213-0780/& 2015 Elsevier Ltd. All rights reserved.
50 J. Johnson et al. / Journal of Outdoor Recreation and Tourism 13 (2016) 49–56
physical aspects of managing avalanche hazard by utilizing well
established technical manuals such as the American Snow Weather,
and Avalanches: Observational Guidelines (SWAG) (Greene et al.,
2010) and the Canadian Observation Guidelines and Recording
Standards for Weather, Snowpack and Avalanches (OGRS) (CAA,
2014). Over the last decade, there has been an increasing aware-
ness of the importance of human factors on avalanche safety and
the number of studies on the human dimension of avalanche
safety is steadily growing (e.g. McCammon, 2004; Haegeli, Haider,
Longland & Beardmore, 2010:186). Whereas the majority of work
in this emerging research field is focused on recreationists (e.g.
Atkins, 2000; McCammon, 2004; Haegeli et al., 2010; Zweifel &
Haegeli, 2014; Hendrikx & Johnson, 2014), there are fewer studies
specifically targeted toward avalanche professionals (e.g., Adams,
2005; Simenhois & Savage, 2010; Stewart-Patterson, 2014; Hen-
drikx, Shelly, & Johnson, 2014). Existing studies (e.g. Adams, 2004)
acknowledge the complex and multifaceted decision environment
of avalanche professionals who must often strike a balance be-
tween personal safety and operational imperatives. This may in-
clude opening ski runs, guiding clients, or teaching courses in
potential avalanche terrain. Studies of avalanche accidents often
focus on the failure of individual decision expertize of while
tending to neglect institutional influences that may play a role in
the incidence of on the job accidents.
In this paper we investigate perceived differences in the orga-
nizational safety culture within the AVPRO community. Specifi-
cally, we examine the role of personal and organizational factors
and the incidence of accidents. Using individual survey data from
AVPROs we identify two types of organizations – those with a
strong safety culture and those with a weak safety culture. We
then examine perceived differences in the causes of accidents
among members of the two types of organizations. We do not
examine accidents rates, severity, or types of accidents.
1.1. Organizational culture
Organizational culture has a rich literature; far beyond the
scope of this paper, but two foci are relevant here. First, how do we
describe organizational culture and second, from which vantage
point do we elucidate our discussion of safety and accidents.
Table 1
Dimensions of organizational culture.
Organizational culture is the behavior of humans within an
organization and the meaning that people attach to those beha-
viors (Smircich, 1983). Schein (2010) refers to these as “norms and
practices”. Organizational culture is described by Deal and Ken-
nedy (2000) as “how work gets done”. Peters and Waterman
(1982) consider innovativeness and productivity. Kotter and Hes-
kett (1992) look to adaptive and unadaptive cultures where
adaptive cultures are those who possess the capacity for organi-
zational learning. Organizational culture can be studied from the
point of view of management, employees, competitors, or custo-
mers. We follow Schein's (1996) examination of organizational
culture from the standpoint of the observer within the organiza-
tion. In our case it is the avalanche professional working in a
setting where avalanche hazard exists.
Schein (2010:26–36) delineates three levels of cultural phenom-
ena visible to the observer within the organization. Each adds to an
understanding of organizational culture. Artifacts are the identifiable
elements of the organization (i.e. uniforms, language). Artifacts pro-
vide personal identity to those in an out of the organization through
the use of professional jargon, behaviors, and appearance. Myths
built around the organization help define it for members and non-
members alike (i.e. FBI as “crime busters”). Espoused Values are rules
of behavior often expressed in official philosophies/policies and
statements of identity such as professionalism and procedures. Va-
lues are often self-reinforcing for the organizational culture. As they
come to define an organizational way of thinking, they may guide the
organization through difficulty (i.e. after a sever accident). Assump-
tions are the deeply embedded, taken-for-granted behaviors, which
are usually unconscious, but constitute the essence of culture. This
level is so well integrated into organizational culture they are difficult
to recognize even by those within. Organizational assumptions are
the way the organization acts, thinks, and perceives their reality. In
practice, Schein (2010) suggests these three levels combine to form a
career anchor—one's self-concept of the organization, one's role in it,
and one's perceptions of one's talents and abilities. In short, by ex-
amining the organizational culture from the insider's point of view
we gain an understanding of how organizational culture affects work.
During our survey design we sought to broadly integrate Schein's
(2010) “Espoused Values” dimension of organizational culture
through queries on ease and openness of communication, training,
 J. Johnson et al. / Journal of Outdoor Recreation and Tourism 13 (2016) 49–56
organizational perception of risk, procedures aimed at assimilating
risk into daily operations, and communication of goals. These ques-
tions are found in Table 1 below.
1.2. Safety culture in organizations
The management leadership literature identifies, along with
several other factors, a “virtuous cycle” of employee perceptions of
the safety system being related to management's commitment to
safety, which, in turn, appear to be related to injury rates (e.g. Goh,
Brown, & Spickett, 2010). Briefly, it is thought that if members of
the organization can maintain a “culture of safety” it prompts all
its members to adopt self-perpetuating safe practices (Carroll,
Rudolph, & Hatakenaka, 2003). The idea is not new and is found in
organization literature over two decades ago (e.g. IAEA, 1991; Cox
& Flin, 1998). However, it is also more complex than it appears.
Zohar (1980) draws on the concept of a “safety climate” to explain
the behavior of individuals’ perception of safety. Climate is described
as the relationship between a number of organizational character-
istics, including management commitment to safety, the openness of
communication links, and the stability of the workforce, and safety
behavior. Thus, there is recognition of a shared understanding re-
garding safety that can be established within organizations between
managers and employees. Roberts (1990) identifies several strategies
that organizations use to build a culture of safety including re-
dundancy, continuous training, and organizational design. Pidgeon
(1991) argues that a “good safety culture” will include not only po-
sitive norms and attitudes regarding safety, but also reflexivity that
allow organizations to discover and learn about potential hazard. This
is reminiscent of Kotter and Heskett's (1992) notion of capacity for
organizational learning.
Guldenmund (2000) suggests that there is no single satisfac-
tory model of a safety culture. Guldenmund (2000) and Schein
(2010) point out that the attributes related to communications,
training, procedures, and accident culture form the basis of any
examination of organizational culture. Each of these elements is a
building block for a culture where safety is paramount in a highly
variable environment. Taken together, these ideas form the basis
of organizational understanding of establishing a safety culture. It
is based largely on procedural practices where cause is identified
and responsibility assigned. Accident avoidance in future cases
relies on changing practices, rules, and behavior modification (e.g.
Zohar and Luria, 2003). However, this limited view of safety cul-
ture fails to take into account the complexity of the social, in-
stitutional, and often political influences on organizational culture.
Recent thinking on safety culture considers “the second story”
to help explain errors or accidents and to take into account its
inherent complexity. Error is not seen as an isolated event (the
first story), rather, error is symptomatic of systemic failure(s)—
errors are symptoms of deeper trouble in the way people do their
work every day (Dekker, 2011). Whereas “first stories” seek to find
proximate causes (and often blame), second stories seek to un-
derstand how the multiple components of the organizational
ecosystem failed to prevent error. An organizational system is
comprised of four levels: parts (mechanicals, equipment), units
(collections of parts – lifts, transport), subsystems (individuals and
the parts they rely on) and the system itself (a collection of the
whole, ecosystem, weather) (Perrow, 1984; Woods & Cook, 2002).
Briefly, an accident can be either component failure that “involve
one or more parts or units” or a system accident that “involve the
unanticipated interaction of multiple failures.”(Perrow, 1984). If,
for example, an accident occurs because “there were too many
people on the slope stressing the snowpack ” (Chabot, 2002:240)
we might ask what transpired that placed that many people on the
slope. We might then ask if placing too many people at risk on the
slope was the result of procedural error, a temporary lapse of the
51
communication system, a personal failure by a line worker, or a
feature of managerial decision making. In the case of a system
failure it could be all four. First story explanations often cause us to
ask who is to blame and stop there, second story explanations
encourage us to look to the context for error and consider the
workplace ecosystem (Maguire, 2014).
2. Data and methods
2.1. Survey design and distribution
Comprehensive surveys of the AVPRO profession are rarely
conducted (Adams, 2005; Johnson & Savage, 2013; Bergeron &
Johnson, 2015). To fill this void, we developed a detailed online
survey to (1) better understand the demographics of the AVPRO
community; (2) examine the institutional settings in which they
work, and (3) study the relationship between organizational cul-
ture and avalanche related accidents.
Survey development followed a three-part process beginning
with literature in the industrial safety literature (e.g., Guldenmund,
2000, 2007; Schein 2010). Questions were rewritten to reflect the
AVPRO profession. We conducted a review of the draft survey via an
interactive poster session for attendees of the 2012 International
Snow Science Workshop (ISSW) in Anchorage, Alaska. The ISSW at-
tracts the global AVPRO community and so provides an efficient
venue for expert input. Approximately 35 respondents of the ava-
lanche education and research community provided comments at
this meeting. The final review of the survey was submitted to a panel
of AVPRO expert reviewers. The survey was comprised of 37 ques-
tions that covered standard demographics, work history and job
description, and organizational culture. Measures of the dimensions
of organizational culture were grounded in the Schein's (2010) three
levels of organizational culture notably, his identification of espoused
values. These included respondents' perceptions of organizational
communication, training procedures, and accident culture.
The survey also asked respondents to think carefully about
their behaviors and those of their organization associated with
close calls and accidents on the job. To examine this aspect of their
organizational culture, we provided participants with a list of 13
accident causes derived from accident reports and our pre test
phase. We asked survey respondents to rank the top three acci-
dent causes in order of importance. The provided list of accident
causes covered all four main dimensions of possible accident
failures: personal, team, organizational, and other. Completion of
the electronic survey required approximately 15–20 min.
There is no single professional organization of the potential re-
spondent pool so we employed a modified convenience sample with
the help of multiple international organizations. The web link to the
survey was sent to professional ski patrols, several professional
avalanche education organizations, ski and mountain guide certifi-
cation programs, and professional membership listserves. Profes-
sional membership organizations included American Avalanche As-
sociation, the Canadian Avalanche Association, and the Association
Nationale pour l’Étude de la Neige et des Avalanches (France). Each of
the participating organizations elicited survey participants by send-
ing emails with the survey hyperlink to their membership lists and/
or posting the survey link on their website or in their electronic
newsletter. We also encouraged respondents to forward the survey
link to their personal contacts (i.e., snowball sampling) to further
increase participation. In this way we expanded the potential re-
spondent pool to New Zealand, Argentina, and Chile.
The survey was posted for six months between September
2012 and February 2013. The relatively long duration for admin-
istration was required in order to capture the winter season in
both the northern and southern hemispheres and to allow ample
52 J. Johnson et al. / Journal of Outdoor Recreation and Tourism 13 (2016) 49–56
time for the snowball sample to be effective. The survey was
confidential and anonymous, and no IP addresses were collected.
All procedures underwent institutional review for human subjects
(IRB) at Montana State University.
2.2. Analysis approach
Of the 485 individuals who started the survey, 401 completed it.
An additional nine participants were eliminated from the analysis
dataset because they did not answer the questions regarding their
employment situation (n¼ 6), answered less than three of the four
questions on organizational culture (n¼2), or provide unrealistic
answers regarding number of witnessed avalanche fatalities (n¼ 1).
This resulted in an analysis data set of 392 survey participants.
First, the characteristics of the survey sample were examined
using basic descriptive statistics (e.g., median, interquartile ranges).
In a second step, primary workplaces of survey participants were
classified as having a strong or weak safety culture based on par-
ticipants’ ratings of the four dimensions of organizational culture of
safety: (a) organizational communication, (b) reflectivity on acci-
dents, (c) procedures, and (d) organization norms. Organizations
with high ratings (Most of the time or Always) in all four dimen-
sions were designated as having a strong safety culture, while or-
ganizations with low ratings (Never, Rarely and Sometimes) on two
or more dimensions are identified as having a weak safety culture.
We considered responses with any other patterns (e.g., only one
low rating, missing values, etc.) to represent organizations with
intermediate safety ratings.
Subsequently, participants’ responses regarding sociodemographic
characteristics of respondents and organizations, avalanche safety
work procedures, and respondent personal work skills and attitude
were examined for differences between workplaces with strong and
weak safety culture. To better highlight the differences between the
extremes, participants representing organizations with intermediate
safety ratings were not included in this comparison. Pearson χ2 tests
were used for comparing nominal data and the Wilcoxon rank-sum
test for ordinal or non-normal data.
Differences in median rankings of individual accident causes
between organizations with strong and weak safety cultures were
also examined using the Wilcoxon rank-sum test. To order the
causes of accidents for all participant from workplaces with strong
and weak safety cultures respectively, weighted overall rankings
were calculated for each cause by multiplying the percentage of
individual ranking by the value of the rank and summing it across
all rank value (i.e., 1*1st% þ2*2nd% þ3*3rd%þ 4*4th%).
For all statistical comparisons, we considered p-values of less
than 0.05 to be statistically significant and p-values between 0.05
and 0.10 to be marginally significant.
3. Results
3.1. Demographics
A total of 392 participants completed the survey and answered
the questions necessary for the analysis. Participants are primarily
from North America (95%), but the sample also includes respondents
from Europe, New Zealand, and South America. Since there is no
accurate knowledge of the total population of potential respondents,
it is impossible to determine a response rate for the survey. Ac-
cordingly, due to the sampling methodology, results cannot be gen-
eralized to the whole of the AVPRO population. The median age of
survey participants is 41 years, 10% of the sample is female. Almost
65% are married and 43% of the survey sample has children. Sixty-
nine percent of participants have an undergraduate or graduate
university degree. Most have a technical or science degree.
3.2. Work setting
Multiple job descriptions populate the AVPRO community. This
includes those who teach avalanche awareness classes to re-
creationalists; avalanche forecasters for commercial ski resorts;
backcountry forecasters; avalanche controllers for departments of
transportation and railroads; commercial mountain and ski guides;
and professional ski patrol. Avalanche educators (46%), guides in-
cluding both motorized and non-motorized access as well as ski and
mountain guiding (45%), and professional ski patrollers (33%) are the
most common occupations among survey participants. All other ca-
tegories are less than 10% of the total job descriptions.
Slightly more than half of respondents (58%) self-identify
working in more than one job description. As a result of the
complexity of the employment matrix, differences among the
various job descriptions is less insightful than it might seem. In
fact, there are likely more similarities between job descriptions
than differences. One aspect of the job all have in common is that
they operate in an environment where physical risk from ava-
lanches is very real and accidents always have the potential of high
consequences. Working conditions are similar as all jobs are car-
ried out in winter conditions where weather, visibility, snowpack
conditions, and terrain hazards define the working environment.
Differences in work related circumstances examined here are the
size of the employer organization, the nature of employment (i.e.
self-employed vs payrolled), as well as their self-stated relation-
ship between employer and employee.
Eighty nine percent of the sample works only in North America.
Sixty eight percent of the sample works exclusively in the US and
eighteen percent only in Canada. One percent of the sample works
in both countries. Five percent work only in a location oversees
(e.g. Europe, New Zealand) and the remaining 6% work in North
America and at least one oversees location. Length of career in the
avalanche industry ranges from 1 to 50 yrs. with a median of 14
years and an interquartile range between 8 and 21 years. The
median tenure at their current primary position is 6.5 years. The
distribution for the number of working days per year is between
0 and 240 days/yr., with the primary peaks around 30 days/yr. (i.e.
one winter season) and 80–100 days/yr. (i.e. two winter seasons)
The majority of survey participants (69%) work for privately
owned companies. Nineteen percent work for either a federal and
state government agency. Most of these organizations (61%) are
well established and have been in existence for more than 30
years. The sizes of the avalanche safety teams employed of these
organizations range from small (1–3; 21%), to medium (4–10;
33%), large (11–20; 20%) and very large ( 420; 26%).
3.3. Organizational culture
An examination of participants' ratings of the dimensions of
organizational culture reveals that most work in an organization
with a strong safety culture (Table 1). Overall, the ratings are
highest for the dimension representing operational procedures (If I
see a potentially dangerous situation, I am confident I can ask op-
erations to stop and reassess.) with a median of 5 and an inter-
quartile range from 5 to 5. The dimension of organization com-
munication (Communication within my company is open and easy.)
received the lowest ratings with a median of 4 and an interquartile
range from 4 to 5.
Our algorithm for classifying organizations with respect to their
safety culture resulted in 241 workplaces of survey participants to
be classified as “strong safety culture” (Most of the time or Always
ratings in all four dimensions) and 47 work places to be classified
as “weak safety culture” (Never, Rarely or Sometimes ratings on
two or more dimensions). One hundred and four workplaces ex-
hibited other patterns (e.g., only one low rating, missing values,
 J. Johnson et al. / Journal of Outdoor Recreation and Tourism 13 (2016) 49–56 53

Table 2
Avalanche safety work procedures.

Question Strong Safety Culture Weak Safety Culture Comparisona

n Median IQR n Median IQR p-Value

My employer has provided me with a formal best practices statement or policy for my job as an 233 2 1–2 43 4 2–4 o 0.001nnn
avalanche professional. Strongly agree (1) – Agree (2) – Neither (3) – Disagree (4) – Strongly dis-
agree (5)
I understand best practice procedures as defined by my employer for my job as an avalanche pro- 231 2 1–2 41 3 2–4 o 0.001nnn
fessional. Strongly agree (1) – Agree (2) – Neither (3) – Disagree (4) – Strongly disagree (5)
Approximately how long was your initial job training for your current position? More than 1 year (1) 236 2 1–4 45 3 1–5 0.101
– 6 to 12 months (2) – 1 to 6 months (3) – A week to a month (4) – Less than one week (5)
My operation conducts avalanche meetings/briefings Daily, before and after each shift (1) – Daily, 233 2 1–3 43 2 2–5 0.013nn
before each shift (2) – Weekly (3) – Monthly (4) – As deemed necessary (5) – Never (6)
The organization I work for has provided me with opportunity to receive the highest level of training 238 2 1–3 45 4 3–4 o 0.001nnn
possible given my position and years on the job. Strongly agree (1) – Agree (2) – Neither (3) –
Disagree (4) – Strongly disagree (5)
When a near miss occurs on the job, my organization takes the opportunity to reflect and/or debrief 199 1 1–2 36 2 1–3 o 0.001nnn
the incident. Always (1) – Most of the time (2) Sometimes (3) – Rarely (4) – Never (5)
When an accident occurs on the job, my organization takes the opportunity to reflect and/or debrief 184 1 1–1 32 1 1–2 o 0.001nnn
the incident. Always (1) – Most of the time (2) Sometimes (3) – Rarely (4) – Never (5)

*po 0.10; **p o 0.05; ***p o 0.01.

a
Wilcoxon rank sum test.

Table 3
Personal skills and attitudes.

Question Strong Safety Culture Weak Safety Culture Comparisona

n Median IQR n Median IQR p-Value

I have good decision making skills when it comes to being safe at my job. Strongly agree (1) – Agree 236 1 1–2 47 1 1–2 0.073
(2) – Neither (3) – Disagree (4) – Strongly disagree (5)
I have a mentor to whom I look for advice and knowledge. Always (1) – Most of the time (2) – 237 2 2–3 47 3 2–3 0.103
Sometimes (3) – Rarely (4) – Never (5)
When I am blasting or ski cutting I assume it will go big and adjust my safe zones accordingly. Always 236 2 1–2 46 2 2–3 o 0.001nnn
(1) – Most of the time (2) – Sometimes (3) – Rarely (4) – Never (5)
I approach each hazard evaluation or mitigation situation with an open mind and no preconceived 239 2 1–2 47 2 2–3 o 0.001nnn
conditions. Always (1) – Most of the time (2) – Sometimes (3) – Rarely (4) – Never (5)
When I make a stability assessment, I try to assimilate as much data as possible. Always (1) – Most of 247 1 1–2 47 2 1–2 0.080n
the time (2) – Sometimes (3) – Rarely (4) – Never (5)
I check my ego at the door when I make group decisions on the job. Always (1) – Most of the time (2) 246 2 1–2 46 2 1–2 0.002nn
– Sometimes (3) – Rarely (4) – Never (5)

*po 0.10; **p o 0.05; ***p o 0.01.

a
Wilcoxon rank sum test.

etc.), which we interpreted as representing workplaces with in-
termediate safety cultures.
3.4. Differences between organizations with a strong or weak safety
culture
The final sample size for the comparison between workplaces
with weak and strong safety culture was 288. Several attributes
were examined with respect to differences between organizations’
safety cultures: sociodemographic characteristics of respondents
and organizations, avalanche safety work procedures, and re-
spondent personal work skills and attitude.
No significant sociodemographic differences between the two
types of organizations emerged among respondents including the
length of the tenure in the current job. Further, there were no dif-
ferences in size, age, or ownership of the two types of organizations.
However, there were observed differences among several op-
erational work practices related to avalanche safety (Table 2).
AVPRO respondents reported organizations identified as having a
strong safety culture were significantly better equipped with formal
best practice statements (p-value o0.001) and that respondents had
a better understanding of organizational best practices (p-value
o0.001). Differences between the lengths of the initial training for
respondents, was just outside of being marginally significant (p-value
0.101). Respondents from organizations with a strong safety culture
felt that they had significantly more opportunity to get the highest
level of training (p-value o0.001). Organizations with a strong safety
culture debriefed job-related near misses and accidents more fre-
quently than organizations with a weak safety culture (p-value
o0.001 for both) and had significantly more frequent avalanche
safety briefings (p-value 0.013).
The differences in avalanche related practices between the two
types of organizations are reflected in personal work skills and
attitudes. Those who work for organizations with a strong safety
culture self-reported higher level skills and more positive attitudes
than those who work for organizations identified as having a weak
safety culture (Table 3).
Our analysis reveals several significant differences between
employees of the two types of organizations. Those who work for
organizations with a strong safety culture self-rated their decision-
making skills slightly higher than their colleagues from the orga-
nizations with a weak safety culture (p-value 0.073). Differences
between the two groups regarding having a mentor were just
slightly outside of being marginally significant (p-value ¼0.103).
Members of organizations with a weak safety culture tended to
rate the question “things will go big when blasting or ski cutting”
54 J. Johnson et al. / Journal of Outdoor Recreation and Tourism 13 (2016) 49–56

Table 4
Rankings of accident causes.

Question Type Strong Safety Culture (n ¼ 441) Weak Safety Culture (n ¼47) Comparisona

Rankn 1st 2nd 3rd 4th Rankb 1st 2nd 3rd 4th p-Value

Operational pressures (open/visit/utilize affected terrain) Organizational 6 6% 8% 5% 81% 1 21% 13% 9% 57% o 0.001nnn
Poor personal decision making Personal 1 24% 16% 15% 45% 2 15% 13% 13% 59% 0.050n
Loss of situational awareness (i.e., what is going on around Personal 2 19% 14% 12% 55% 3 17% 11% 6% 66% ns
you)
Poor communication (between partners or operations) Team 4 13% 17% 9% 61% 4 13% 9% 19% 59% ns
Pre-assumptions (decisions based on past data or Personal 3 14% 14% 16% 56% 5 11% 13% 9% 67% ns
experience)
Hazardous attitudes (macho, anti-authority, impulsive Personal 8 4% 4% 6% 86% 6 6% 17% 9% 68% 0.003nn
behaviour)
Over estimating of one's ability Personal 5 6% 11% 10% 73% 7 6% 6% 9% 79% ns
Management decision that override your judgment Organizational 13 1% 0% 3% 96% 8 2% 6% 9% 83% 0.003nn
Stress/fatigue from long hours/overwork Organizational 7 5% 5% 8% 82% 9 4% 4% 4% 88% ns
Organizational decisions that compromise safety Organizational 11 1% 2% 3% 94% 10 0% 6% 9% 85% 0.061
Sense of responsibility (to the profession) Personal 10 2% 3% 5% 90% 11 2% 2% 4% 92% ns
Competitive with others Team 12 1% 2% 1% 96% 12 2% 0% 0% 98% ns
Bad luck Other 9 3% 3% 6% 88% 13 0% 0% 2% 98% 0.040n

*po 0.05; **p o0.01; ***p o 0.001.

a
Wilcoxon rank sum test
b
Calculated from weighted rankings (1*1st% þ 2*2nd% þ3*3rd%þ 4*4th%)

lower (p-value o0.001) than their counterparts suggesting perhaps
they did not trust the procedures they were expected to perform to
be as effective as their counterparts. Likewise, they reported they
did not approach a hazard situation with an open mind (p-value
o0.001) suggesting they were more likely locked into habit-based
actions rather than goal-directed actions recognized as a condition
for appropriate decision-making. (Soares et al., 2012). Those in weak
safety culture organizations judged their ability to assimilate as
much information in as possible a hazard situation (p-value¼0.08)
lower than their counterparts. Those same respondents reported
they were less likely to check their ego when in-group decision
making settings (p-value¼ 0.002).
With respect to personal orientation on the job, when asked to
rank their job constituencies (myself, company, AVPRO commu-
nity, public), those who work in a strong safety culture organiza-
tion were more likely to place the organization (p-value ¼0.006) as
more important compared to those who work for organizations
with a weak safety culture who tend to rank themselves as more
important (p-value ¼0.03)
3.5. Accidents causes
The rankings of six perceived accident causes were found to be
statistically significant different between organizations with
strong and weak safety culture (Table 4).
Two organization-related causes and two personal-related cau-
ses emerge as noteworthy. Whereas 6% of those in organizations
with a strong safety culture ranked “operational pressures (i.e. to
open runs or utilized closed terrain)” as number six on the list, 21%
of respondents associated with an organization identified as having
a weak safety culture ranked it as number one (p value¼0.001);
53% of this group ranked it number one. Similarly, “management
overriding personal judgement” was not ranked among the top
three causes by 96% of those in organizations with a strong safety
culture while 83% of those who work in a weak safety culture did
not rank it as one of the top three reasons (p value¼ 0.003). Taken
together, we can surmise that in those organizations with a weak
safety culture respondents consider management's role in hazard
supervision to be a contributory factor in accidents and near misses.
Differences in rankings were less distinct but still significant
with respect to personal causes of accidents. Respondents from
both types of organizations ranked “poor personal decision mak-
ing” first and second respectively as the primary cause of accidents
(p value ¼ 0.05). The other statistically significant personal reason
cited was personally hazardous attitudes (macho, anti-authority,
impulsive behavior) (p value ¼ 0.003). These behaviors can possi-
bly be linked to the results reported in Tables 2 and 3.
4. Discussion
We identify two types of organizational cultures—strong and
weak culture of safety—based on respondents’ self reported per-
ceptions of their organization. Safe organizations are reported by
their employees to have better communication, clear goals and
policy statements, better accident prevention, better reporting
practices and, better avalanche safety practices. There is a clear
delineation between the two. We would emphasize here that no
organization was identified as having no safety culture rather; the
perception of strong vs. weak safety culture is a matter of degree.
Congruent with organizational literature that recognizes accidents
are rarely the result of linear cause-effect relationships (e.g., Booth
and Lee, 1995; Leveson, 2004; Dekker, 2014), AVPRO respondents
identify multi dimensions of accidents—some organizational; others
the result of personal actions. We find explanatory power in the
difference in organizational safety culture. Some organizations simply
fail to train, communicate, and as a result, place their employees at
risk. These practices manifested themselves as generally lower skill
levels and higher self-orientation among individuals in organizations
we identify as having a weak safety culture. These traits undermine
organizational cohesion and team orientation.
Safety protocols were found to differ between the two types of
organizations. The issue of operational pressures (to open closed
runs, guide clients in marginal conditions) are evident as a po-
tential cause for accidents in organizations with a weak culture of
safety. This sense of operational pressure can either be magnified,
or diminished through clear operational safety protocols or, vague
and undefined ones. These too were found to vary between types
of organizations. While we did not consider accident rates speci-
fically, the safety culture of an organization will ultimately be re-
flected in the rate of accidents and possibly their severity. This has
been clearly documented in medical and aeronautical settings
 J. Johnson et al. / Journal of Outdoor Recreation and Tourism 13 (2016) 49–56
where accidents are often the result of a combination of in-
dividual, group, and organizational failure. In these industries
safety reforms aimed at all three components has been particularly
effective (see especially: Wilson (2007:102–127) and Gordon,
Mendenhall and O'Connor (2012)).
At issue for organizations identified as having a weak safety
culture is if they can learn from accidents that occur in theirs and
other organizations. Such a culture would display not only an
ability to recover from accidents (i.e. on hill avalanche) but would
learn to avoid similar accidents in the future (Lee, Vargo & Seville,
2013; Everly, 2011). Future analysis would look for evidence of
growth and change toward a culture of safety among those orga-
nizations that are identified as having a weak safety culture.
The findings presented here have real world application. Gen-
erally, organizations where the safety culture is weak tend to seek
to place blame when accidents occur and focus on immediate
causes while discounting those that are complex and difficult to
understand. This further erodes support for the goals of the or-
ganization. The result is that rather than a virtuous cycle of safety,
a downward spiral of less safety, less organizational solidarity, less
organizational learning, and a higher likelihood of accidents may
be the outcome. This refers to Schein’ (2010) concept of the “career
anchor” where one's relationship within the organization is
formed, in part, by the organization itself.
With respect to avalanche safety training, we saw that orga-
nizations with a strong culture of safety confer more training and
expertize to their members. These traits were associated with self-
reported higher-level skills and more positive attitudes toward the
organization. These findings suggest organizations with a strong
culture of safety are more likely to build a virtuous cycle of safety
than weak organizations. Likewise, those who work for organiza-
tions where the culture of safety is weak tended to report a less
positive orientation toward the organization and its constituents;
they tended to not express as much attention to the job.
Most strikingly, we find six of thirteen accident causes to be
different for the two types of organizations. Disturbingly, those who
work for organizations with a weak safety culture rank operational
pressure as the number one cause of accidents most frequently and
ranked “management overriding personal judgement” significantly
higher than those working for organizations that possess a strong
culture of safety. For employees “on the ground”, this form of
management intervention could prove troublesome in the long
term as decision making authority and self-perceived expertize is
usurped. Employees may eventually hand all responsibility over to
distant managers and engage in benign complacency.
Overall, poor personal decision making was ranked very high
by all respondents. Complete understanding of this finding is
problematic. Clearly, all organizations need to engage this facet of
the work environment and work for continued development of
decision-making skills (Berger & Johnson, 2015) and organizations
with a strong culture of safety do. However, it is also possible that
respondents suffer “hindsight bias” and in an effort to deconstruct
the complexity of an accident, they place blame on themselves.
Woods and Cook (2002:138) explain that hindsight bias can be
invoked when we seek to simplify complexity in order to explain a
failure of practice. They observe that this distorted view may lead
to counterproductive “solutions”.
There is a tendency in organizations to look for an ultimate
cause of accidents (i.e. a “smoking gun”). Dekker (2014) refers to
these as the “quick fix” and suggests that in doing so organizations
often focuses on the behavior of individuals rather than potential
problems within the organization. Woods, Dekker, Cook, Johanne-
sen and Sarter (2010:244) observe that “safety is not found in a
single person, device, or department rather; it is a product of a
functional culture of safety”. Conversely, following the logic of the
“second story” (Maguire, 2014), understanding the context for error
55
within the framework of the strong/weak culture of safety typology
may point to organizational understanding for AVPRO accidents.
5. Conclusions
A comprehensive survey aimed at understanding the AVPRO
community of professions is a complex undertaking. Many AVPROs
play multiple roles; they assume a range of responsibilities and
work for multiple organizations. Analysis of the profession re-
vealed significant differences in the safety culture of organizations
that employ AVPROs. Communication, operational pressures, as
well poor personal decision-making were among the causes of
accidents that differed between organizations with a strong cul-
ture of safety and those with a weak culture of safety. Manage-
ment pressure perceived by respondents as intervention serves to
undermine personal autonomy in organizations with a weak cul-
ture of safety and was reported as a major cause of accidents.
Organizations have the capacity to manage these interventions.
Future research could more closely focus on accident causes
between functional groups within the profession (i.e. guides, in-
structors, ski patrol). We did not examine relative accident rates,
severity, or types of accidents due to the nature of convenience
sampling (i.e., we do not know if there is a sample bias) and the
interlocking nature of AVPRO employment (most hold more than
one AVPRO job description). These limitations prevent meaningful
estimation of relative accident rates between AVPRO professions.
Different AVPRO jobs require varying amounts of training, perso-
nal autonomy, and discretion. These investigations could prove
useful for deconstructing causes and eventually reducing the in-
cidence of accidents. At issue might be the degree of variance in
accident rates and types between AVPRO jobs. It may be that a few
job types are responsible for a large number of accidents and that
the causes are relatively narrow in scope.
In addition, various jobs in the industry entail different sets of
trade-offs. The decision to open a transportation corridor must
necessarily consider economic and political factors while opening
a ski run for powder skiing may be largely a social or organiza-
tional concern. These drivers may play an important role in acci-
dent prevention. Where trade-offs are high and tangible, decisions
may be more robust and have greater organizational support.
Where trade offs are perceived to be lower and qualitative, some
organizations may override individual patroller decisions.
Another fruitful area of research could focus on managers
within the AVPRO community. The degree of formal management
training, employment background, and experience must surely
influence organizational culture yet; this area is virtually un-
explored in the AVPRO community.
Exploring the organizational ecosystem via high volume surveys
is problematic. Whereas survey data typically allows researchers to
query respondents on a wide range of issues, in-depth under-
standing often suffers. For example, it is difficult to understand
one's career anchor or role in a complex organization but it can be
accomplished through in-depth interviews albeit with smaller
numbers of participants. An alternative method of investigating
organizational culture and its impact on accidents may be narrative
analysis (Franzosi, 1998) where multiple narratives (stories) from
members of the organization can be collected and used to form
patterns of behavior. Following Schein's (2010) assertion that the
underlying assumptions inherent in an organization are the ulti-
mate source of values and actions, it seems logical to adapt meth-
odologies that allow for those insights. Narrative analysis may be a
promising method for deep exploration of the organizational cul-
ture of safety (Jones, Shanahan, & McBeth, 2014). We suggest that
alternative methodologies in the social sciences may be more in-
sightful than traditional mass surveys. These methods are costly
56 J. Johnson et al. / Journal of Outdoor Recreation and Tourism 13 (2016) 49–56
however in terms of respondent time, coding costs, and expertize.
Finally, we suggest that our reframing the causes of accidents in
the context of both organizational and personal factors is a valu-
able starting point for future investigations into AVPRO workplace
accidents. Other risk-oriented professions have displayed progress
in greater understanding of workplace safety. The AVPRO com-
munity should build and expand on such research by focusing on
the complexity of the workplace rather than continued emphasis
on failure of individual decision making.
References
Adams, L. (2005). A systems approach to human factors and expert decision-making
within the canadian avalanche phenomena (p. 284) Canada: Royal Roads
University.
Adams, L. (2004). Supporting sound decisions: A professional perspective on re-
creational avalanche accident prevention in Canada. In Proceedings of the In-
ternational Snow Science Workshop (pp. 1-10).
Atkins, D. (2000). Human factors in avalanche accidents. In Proceedings of the in-
ternational snow science workshop (pp. 46–51) Big Sky, MT.
Bergeron, A., & Johnson, J. (2015). Ski patrol and avalanche accidents. The Avalanche
Review, 33(3), 25.
Booth, R. T., & Lee, T. R. (1995). The role of human factors and safety culture in safety
management. Proceedings of the Institution of Mechanical Engineers, Part B:
Journal of Engineering Manufacture, 209(5), 393–400.
Carroll, J. S., Rudolph, J. W., & Hatakenaka, S. (2003). Learning from organizational
experience. In Handbook of organizational learning and knowledge management
(p. 600). New York: Wiley Press.
CAA (2014). Observation guidelines and recording standards for weather, snowpack and
avalanches (p. 78) Revelstoke BC, Canada: Canadian Avalanche Association (CAA).
Chabot, D. (2002). Avalanche education for snowmobilers: efforts of the Gallatin
National Forest Avalanche Center. In Proceedings of the international snow sci-
ence workshop (pp. 237–241). Penticton, BC.
Cox, S., & Flin, R. (1998). Safety culture: philosopher's stone or man of straw? Work
Stress, 12(3), 189–201.
Deal, T. E., & Kennedy, A. A. (2000). Corporate cultures: the rites and rituals of cor-
porate life (p. 223) New York: Perseus Publishing.
Dekker, S. (2011). Drift into failure (1st ed.). Burlington: Ashgate.
Dekker, S. (2014). The field guide to understanding human error (3rd ed.). Burlington:
Ashgate Publishing Ltd.
Everly, G. S. (2011). Building a resilient organizational culture. Harvard Business Re-
view. available online athttps://hbrorg/2011/06/building-a-resilient-organizat.
Franzosi, R. (1998). Narrative analysis-or why (and how) sociologists should be
interested in narrative. Annual Review of Sociology (pp. 517–554), 517–554.
Greene, E., Atkins, D., Birkeland, K., Elder, K., Landry, C., Lazar, B., … Williams, K.
(2010). Snow, weather and avalanches: observation guidelines for avalanche
programs in the United States (2nd ed.). Pagosa Springs, CO, USA: American
Avalanche Association online edition.
Greene, E., Jamieson, B., & Logan, S. (2014). Fatal occupational injuries of avalanche
workers in North America. In Proceedings of the international snow science
workshop (pp. 293–299). Banff, AB.
Goh, Y. M., Brown, H., & Spickett, J. (2010). Applying systems thinking concepts in the
analysis of major incidents and safety culture. Safety Science, 48(3), 302–309.
Gordon, S., Mendenhall, P., & O'Connor, B. B. (2012). Beyond the checklist: What else
health care can learn from aviation teamwork and safety (p. 272) Ithaca, New
York: Cornell University Press.
Guldenmund, F. W. (2000). The nature of safety culture: a review of theory and
research. Safety Science, 34(1), 215–257.
Guldenmund, F. W. (2007). The use of questionnaires in safety culture research—an
evaluation. Safety Science, 45(6), 723–743.
Haegeli, P., Haider, W., Longland, M., & Beardmore, B. (2010). Amateur decision-
making in avalanche terrain with and without a decision aid—a stated choice
survey. Natural Hazards, 52(1), 185–209.
Hendrikx, J., & Johnson, J. (2014). Using global crowd-sourced data to understand
travel behavior in avalanche terrain. In Proceedings of the international snow
science workshop (pp. 224–227). Banff, AB.
Hendrikx, J., Shelly, C., & Johnson, J. (2014). Tracking heli-ski guides to understand
decision making in avalanche terrain. In Proceedings of the international snow
science workshop (pp. 1021–1027). Banff, AB.
IAEA (1991). Safety culture: a report by the International Nuclear Safety Advisory
Group (p. 44). INSAG Series 4, Subject Classification: 063-Nucler power plants
STI/PUB/882 (ISBN:92-0-123091-5).
Johnson, J., & Savage, S. (2013). Organizational survey for avalanche professionals:
an interactive poster installation. In Proceedings 2012 international snow science
workshop (pp. 768–769), Anchorage, Alaska.
Jones, M. D., Shanahan, E. A., & McBeth, M. K. (Eds.). (2014). The science of stories:
applications of the narrative policy framework in public policy analysis (p. 304)
New York: Palgrave Macmillan.
Kotter, J. P., & Heskett, J. (1992). Corporate culture and performance (p. 224) New
York: Simon and Schuster.
Lee, A. V., Vargo, J., & Seville, E. (2013). Developing a tool to measure and compare
organizations’ resilience. Natural Hazards Review, 14(1), 29–41.
Leveson, N. (2004). A new accident model for engineering safer systems. Safety
Science, 42(4), 237–270.
McCammon, I. (2004). Heuristic Traps in recreational avalanche accidents: evidence
and implications. Avalanche News, 68, 42–50.
Maguire, L. (2014). The human behind the factor: a brief look at how context in-
forms practice in recreational backcountry users. In Proceedings of the inter-
national snow science workshop (pp. 942-948). Banff, AB.
Perrow, C. (1984). Normal accident theory: living with high risk technologies (1st ed.).
Princeton: Princeton University Press.
Peters, T. J., & Waterman, R. H. (1982). In search of excellence (p. 350) New York:
Harper Business.
Pidgeon, N. F. (1991). Safety culture and risk management in organizations. Journal
of Cross-Cultural Psychology, 22(1), 129.
Roberts, K. H. (1990). Some characteristics of one type of high reliability organi-
zation. Organization Science, 1(2), 160–176.
Schein, E. H. (1996). Career anchors revisited: implications for career development in
the 21st century. The Academy of Management Executive (1993–2005), 10(4), 80–88.
Schein, E. H. (2010). Organizational culture and leadership, Vol. 2. New Jersey: John
Wiley & Sons.
Simenhois, R., & Savage, S. (2010). Compiling and Studying Professional Avalanche
Worker Incident Data to Improve Worker Safety. In Proceedings: 2010 inter-
national snow science workshop.
Smircich, L. (1983). Concepts of culture and organizational analysis. Administrative
Science Quarterly (pp. 339–358), 339–358.
Soares, J. M., Sampaio, A., Ferreira, L. M., Santos, N. C.,Marques, F., Palha, J. A., Cer-
queira, J. J. and Sousa, N. 2012. Stress-induced changes in human decision-
making are reversible. Translational Psychiatry. 2, e131.
Stewart-Patterson, I. (2014). The development of ski guide decision expertise. In Pro-
ceedings of the International Snow Science Workshop (pp. 330–337). Banff, AB.
Wilson, K. A. (2007). Does safety culture predict clinical outcomes? (dissertation).
Orlando: University of Central Florida Available.
Woods, D., & Cook, R. (2002). Nine steps to move forward from error. Cognition
Technology Work, 4, 137–144.
Woods, D., Dekker, S., Cook, R., Johannesen, L., & Sarter, N. (2010). Behind human
error (2nd ed.). Burlington: Ashgate.
Zohar, D. (1980). Safety climate in industrial organizations: theoretical and applied
implications. Journal of Applied Psychology, 65(1), 96–102.
Zohar, D., & Luria, G. (2003). The use of supervisory practices as leverage to improve
safety behavior: a cross-level intervention model. Journal of Safety Research, 34
(5), 567–577.
Zweifel, B., & Haegeli, P. (2014). A qualitative analysis of group formation, leader-
ship and decision making in recreation groups traveling in avalanche terrain.
Journal of Outdoor Recreation and Tourism, 5–6, 17–26.
Jerry Johnson is a professor in the Department of Political Science at Montana State
University in Bozeman, MT. His wide range of research interests include natural
resource management, rural political economy, and decision making in high-risk
environments. He is the co-director of the MSU Tracks project with Jordy Hendrikx
that combines GPS tracking with user surveys to investigate travel behavior of
backcountry recreationists. He is the author of “Knowing Yellowstone: Science in
America's First National Park”. He is also a research fellow with the Foundation for
Research on Economics and the Environment.
Pascal Haegeli is a renowned expert in avalanche safety research and holds the position
of Assistant Professor in the Faculty of Environment at Simon Fraser University in
Vancouver, BC. His primary research and development interests lie at the interface
between the natural and social sciences especially focused on the development of evi-
dence based tools for the recreation and professional backcountry community.
Jordy Hendrikx directs the Snow and Avalanche Laboratory in the Department of
Earth Science at Montana State University in Bozeman, MT. He is a widely re-
cognized expert in mountain environments, snowpack dynamics, extreme event
analysis, and snow hydrology. He directs numerous graduate students in all aspects
of snow science. He serves as the chair of the research committee for the American
Avalanche Association. His research is interdisciplinary, blurring the lines between
snow science and human behavior in high-risk terrain.
Scott Savage is a lifelong snow science professional. He holds degrees in Chemistry
and Molecular Biology from the University of Colorado at Boulder. He was Ava-
lanche Forecaster and Snow Safety Director at Big Sky Resort (Big Sky, MT) for two
decades. He is currently the director of the Sawtooth Avalanche Center in Ketchum,
ID. He is a regular presenter at international avalanche conferences and regional
professional seminars and is a regular contributor to The Avalanche Review. Scott is
a National Avalanche School instructor and the secretary of the American Ava-
lanche Association.