Knowledge exchange and learning from failures in distributed
environments The role of contractor relationship management
and work characteristics
Leif Jarle Gressgård n Kåre Hansen 1
International Research Institute of Stavanger (IRIS) Department of Social Science Thormøhlensgt 55 N5008 Bergen Norway
article info
Article history
Received 10 July 2013
Received in revised form
1 August 2014
Accepted 1 September 2014
Available online 16 September 2014
Keywords
Failurebased learning
Knowledge exchange
Contractor relationship management
Role clarity
Leadership involvement
Empowerment
abstract
Learning from failures is vital for improvement of safety performance reliability and resilience in
organizations In order for such learning to take place in distributed environments knowledge has to be
shared among organizational members at different locations and units This paper reports on a study
conducted in the context of drilling and well operations on the Norwegian Continental Shelf which
represents a highrisk distributed organizational environment The study investigates the relationships
between organizations' abilities to learn from failures knowledge exchange within and between
organizational units quality of contractor relationship management and work characteristics The
results show that knowledge exchange between units is the most important predictor of perceived
ability to learn from failures Contractor relationship management leadership involvement role clarity
and empowerment are also important factors for failurebased learning both directly and through
increased knowledge exchange The results of the study enhance our understanding of how abilities to
learn from failures can be improved in distributed environments where similar work processes take
place at different locations and involve employees from several companies Theoretical contributions
and practical implications are discussed
& 2014 Elsevier Ltd All rights reserved
1 Introduction
Learning from failures is a key organizational process to
improving levels of performance and ensuring safe work conduct
[1–4] Strong abilities to learn from failures are in this respect
found to be a significant characteristic of high reliability organiza
tions (HROs) referring to organizations operating under demand
ing conditions yet manage to avoid major accidents [5] This
underscores that gaining insights from past experiences and using
this knowledge to design more reliable and effective systems are
important facilitators for preparedness for both present and
prospective crises [6–8]
Learning from failures requires sharing of information and knowl
edge about error experiences [1]Efficient knowledge exchange
betweenorganizationalmembersandunitsisthusregardedas
fundamental for this type of organizational learning to occur and
designing a work climate that supports this objective is therefore
important [9] In this paper we argue that knowledge exchange is
particularly important in distributed interorganizational settings
where similar activities take place at different locations and involve
employees from multiple companies Supporting this argument
Wang and Wang [10] claim that knowledge exchange is of particular
importance in emerging distributed organizations as efforts of
improvement like transfer of best practices [1112] in these organiza
tional settings are highly dependent on how well knowledge is
shared between individuals in different units and at different
locations
Despite the growing acknowledgement that learning from
failures is fundamental to organizational life particularly in
highrisk distributed environments [13] there is a scarcity of
studies in this domain [14] Investigations of antecedents of
organizations' abilities to learn from past experiences and mis
takes and what conditions best facilitate such learning in dis
tributed environments are therefore needed In this respect
Carmeli and Gittell [15] argue that research focusing on relational
foundations of failurebased learning is particularly important as
interaction and knowledge exchange between organizational
members are central in learning processes In a work context
characterized by diversity regarding organizational affiliations
emphasizing the interorganizational dimension is important
when considering the relational foundations of learning Factors
Contents lists available at ScienceDirect
journal homepage wwwelseviercomlocateress
Reliability Engineering and System Safety
httpdxdoiorg101016jress201409010
09518320& 2014 Elsevier Ltd All rights reserved
n Corresponding author Tel þ47 55543866
Email addresses ljg@irisno (LJ Gressgård) kaarehansen@irisno (K Hansen)
1 Tel þ47 55543864
Reliability Engineering and System Safety 133 (2015) 167–175
concerning management of suppliers and contractors should
therefore be emphasized in addition to internal work character
istics Addressing this research gap this paper seeks to throw light
on how management of contractor relationships and work char
acteristics influence the degree of knowledge exchange in the
organizational system which again impacts on organizations'
abilities to learn from failures
The context of the study is drilling and well activities on the
Norwegian Continental Shelf (NCS) which is a setting character
ized by distributed interorganizational work This means that
similar work processes take place at different geographical loca
tions and organizational units and that employees from several
companies are involved in the operations Knowledge exchange
within and between various units (eg offshore installations) is
therefore important Further as it is a setting that involves high
risk operations [1617] failurebased learning is highly relevant
The remainder of the paper is organized as follows The theoretical
background and hypotheses are presented first followed by a
description of the research methodology The results are then
presented and discussed including theoretical contributions and
implications for practitioners The paper is ended with study
limitations and directions for future research
2 Theory and hypotheses
21 Knowledge exchange and learning from failures
Organizational learning represents an important mechanism
through which organizations prosper [18] and learning from
failures is recognized as vital for improvement of safety perfor
mance reliability and resilience [1351319–21] A key theoretical
and empirical question is therefore how such learning is enabled
[815] According to Tjosvold et al [22] learning from mistakes
involves recognizing undesired effects and reflecting on conse
quences of actions in order to reduce the probability of their future
occurrence Likewise Hirak et al [7] posit that learning from
failures occurs when unit members reflect on a failed experience
openly discuss why it occurred and identify the work patterns
that need be modified or changed in order to eliminate the root
cause of the problem (p108) Knowledge in organizations is in
this way continuously created altered and discarded as organiza
tional members gain experience and update their understandings
of reality to reflect the lessons that can be drawn [4]
This understanding of learning from failures implies that
collaboration and interaction among individuals and organizations
are fundamental conditions Argote and MironSpektor [23] argue
in this regard that the processes of knowledge acquisition knowl
edge sharing and knowledge combination are central while
Edmondson [24] claims that organizational learning is a process
of change and improvement in organizational actions through
better knowledge and understanding In addition to the profi
ciency of individual employees the exchange of knowledge within
and across units is thus a significant condition for experience
based learning to occur and is related to the abilities of individuals
to benefit from knowledge accumulated by others and also
influences coordination of activities in the organizational system
[18] Exchange of knowledge can in this respect be understood as
the provision or receipt of task information knowhow or feed
back regarding a product or procedure [2526] and may occur
through formal and informal personal interaction or knowledge
management systems Further the construct encompasses both
the processes of knowledge sharing (ie employees providing
knowledge to others) and knowledge seeking (ie employees
searching for knowledge from others) [27]
Knowledge exchange thus appears to be critical for the ability
of organizational members to reflect on their experiences and
can be a significant factor in explaining why organizations vary
dramatically in the rate at which they learn from mistakes [1828]
Supporting this assumption studies and investigations of acci
dents have identified knowledge exchange processes as funda
mental factors According to Pasman et al [29] major accidents
recent years have occurred because of a lack of abilities to absorb
unwanted and unforeseen disturbances and in a study of how
investigations of incidents and accidents in a highhazard setting
were analyzed by the involved companies Doytchev and Hibberd
[30] found that there was limited communication flow between
key stakeholders in various parts of the work processes Research
on HROs also emphasizes that cognitive and organizational sys
tems that promote situational awareness and knowledge sharing
in complex environments can prevent the occurrence of dangerous
situations Knowledge exchange by use of incident reporting
systems may be of particular importance as this may improve
the processes of detection reduction and mitigation of failure in
safetycritical systems [31] Weick and Sutcliffe [5] argue in this
respect that HROs encourage reporting errors they elaborate
experiences of near miss for what can be learned and they are
wary of the potential liabilities of success including complacency
the temptation to reduce margins of safety and the drift into
automatic processing (p 9)
In distributed work environments different units of the same
organization may represent valuable knowledge sources That is
geographically distributed units of the same company are likely to
have similar problems and exchanging solutions is therefore likely
to benefit both the individual units as well as the larger organiza
tion [32] Collection storage and access to experiential knowledge
acquired at one work site can thus be beneficial to other sites [33]
This underscores the importance of knowledge exchange both
within and between units and work locations for organizational
learning to occur We therefore hypothesize that
H1 Firms' abilities to learn from failures are positively related to
knowledge exchange (a) within units and (b) between units
According to Catino and Patriotta [1] failures often stem from
sequential action chains concealed in habitual behavior Likewise
Pasman et al [29] claim that organizational erosive drift is shown to
be responsible for complacent behavior and degradation of safety
attitude This implies that external knowledge (ie knowledge
originating from outside the respective organizational units) may
be necessary in order to enlighten local practices and facilitate
critical reviews of local work conduct Knowledge exchange across
organizational units may thus be of particular importance in order
to avoid drift into failure and facilitate corrections of work We
therefore expect that
H2 Firms' abilities to learn from failures are more strongly positi
vely related to knowledge exchange between units than knowl
edge exchange within units
22 Contractor relationship management and knowledge exchange
Organizational systems where multiple actors are involved in
closelyknit work processes are common in several industries and
imply that tasks conducted by employees of one organization
have to be synchronized with tasks wholly or partly executed by
external actors [3435] In such organizational systems exchange
of knowledge across organizational borders is fundamental in
an organizational safety perspective [36] Scholars argue in this
regard that communication and collaboration among supplychain
members can foster interorganizational learning [37] especially by
exchange of tacit and critical knowledge [38–40]
LJ Gressgård K Hansen Reliability Engineering and System Safety 133 (2015) 167–175168
Offshore drilling and well operations represent one example of a
closelyknit organizational system as almost 70 percent of all
offshore personnel are employed by contractors while the overall
responsibility for operations rests with the operating company [41]
The contractors are often multinational and have experience with
contract work for multiple companies at different locations and thus
possess valuable knowledge on identifying and managing risk
factors and handle challenges that occur during operations Research
has in this respect shown that effective operator–contractor coordi
nation is associated with lower accident rates [16] and increased
safety compliance [42] However attaining wellfunctioning knowl
edge exchange processes in such organizational settings is not
unproblematic Companies do not always provide sufficient infor
mation on their activities and the informal contact between the
interdependent milieus is often negligible [41]
Empirical research results show that firms successful in sup
plier development efforts share information frequently and in
a timely manner with their suppliers [43–45] According to
Rebolledo and Nollet [46] suppliers can be major providers of
knowledge for improvement initiatives and production of goods
and services but valuable interfirm knowledge exchange and
learning is a complex process better achieved under specific
conditions rather than haphazardly This means that there is a
need to identify the characteristics of interorganizational relation
ships that could foster the transfer and application of knowledge
between partners as not all relationships have the same ability to
promote learning across organizational borders [47]
In a study of the effects of relational coordination within
organizational borders Carmeli and Gitell [15] found that goal
congruency shared knowledge and mutual respect resulted in an
organizational climate that enabled organizational members to
engage in learning from failures and concluded that highquality
relationships within organizations increase information proces
sing capacity by connecting employees who play distinct but
interdependent roles In an interorganizational setting Carr and
Kaynak [48] found that interfirm information sharing and sup
plier development are significant factors for firm performance
Relational variables like trust have in this respect been identified
as significant in order for knowledge exchange across organiza
tional borders to take place and prevent potential leakage of
proprietary knowledge [49–52] Collaboration and socialization
at different levels are also expected to support interfirm learning
as close and frequent interaction between employees of different
organizations represents an important mechanism for transfer of
knowledge across the organizational interface [4651] The quality
of relationships management can in other words influence the
level of interaction and knowledge exchange [1553] and be an
enabler of interfirm learning [54]
In distributed environments like offshore drilling and well
operations knowledge exchange occurs both within and between
units (offshore installations) As organizational units may have
limited task interdependencies and interpersonal ties they may
focus their attention at local unit activities [32]Efficient manage
ment of contractor relations may therefore represent a means for
providing external input to local activities Further the work of
contractors and suppliers is in general based on contracts with a
limited duration [42] These employees therefore represent a type
of contingent workforce subject to changes of work locations
based on requests and demands of the client Hence the degree
of unit affiliation and identification of contractor employees is
lower compared to permanent employees of the client In the
context of the present study this means that contractor employees
are more mobile than operator employees (ie a larger number of
contractor employees work at several installations compared to
operator employees) and may therefore represent an important
source for crossunit knowledge exchange [55] The importance of
contractor employees for knowledge exchange across borders is
supported by research showing that employees are less likely to
share and seek knowledge beyond their work unit to the extent
they identify more strongly with the subunit relative to the
organization as a whole [56–58] We therefore hypothesize that
H3 Quality of contractor management is positively related to
knowledge exchange (a) within units and (b) between units
23 Work characteristics and knowledge exchange
Knowledge exchange within and between organizational units
is always ultimately rooted in individual behaviors [59] and is
influenced by both ability and motivational factors [60–62]
Research has in this regard found that characteristics of the work
and work environment are important drivers for knowledge
exchange processes A number of empirical studies have focused
on the influence of various HRM practices on firmlevel knowledge
sharing and creation [6063] and scholars have argued that job
characteristics structure the nature and content of the interrela
tionships between workers by configuring particular patterns of
interaction cooperation and collaboration [64–66] Characteristics
of the work situation may thus influence the extent of knowledge
exchange in the organizational system In this study we focus on
the roles of three factors that previous studies have highlighted as
important for organizational safety in an interorganizational dis
tributed highrisk work environment [426768] Leadership invol
vement role clarity and empowerment
231 Leadership involvement
Barriers may exist that prevent employees from openly and
freely share knowledge about their experiences and mistakes they
have made [6970] According to Edmondson [71] a work climate
that inhibits employees from speaking up with questions con
cerns and challenges and also advice and potential solutions to
problems that the organization faces is detrimental for failure
based learning In contrast a wellfunctioning safety culture of an
organization provides a supportive context for error reporting and
encourages sharing of information and knowledge about experi
ences [1] Research has in this respect shown that leadership is a
central variable that may support expression of views at the
workplace [9] In the context of organizational safety Hirak et al
[7] found that leader inclusiveness is positively associated with
perceptions of psychological safety climate which again facilitate
failurebased learning In a metaanalytic study Nahrgang et al
[72] showed that a supportive work environment is an important
factor for both work engagement and safety Research has also
shown that leadership involvement in work operations has posi
tive effects on safety compliance [42] and assessments of overall
workplace safety [68] In line with this O'Dea and Flin [73] argue
that good safety leadership involves high involvement in safety
initiatives as well as involvement in work operations and frequent
interaction between workers and managers On this basis we
expect that leadership involvement may facilitate knowledge
exchange by means of enhancing work engagement and support
ing a collaborative and open work climate and hypothesize that
H4a Leadership involvement is positively related to knowledge
exchange (a) within units and (b) between units
232 Role clarity
A central element of job design is to identify the relevant tasks
and activities of a job [59] Role clarity encompassing the aspects
of responsibilities authority and competence requirements is
therefore a relevant job design factor It may help employees
seeing their roles in the larger organizational system and thereby
represent a mechanism for increasing shared meaning among
LJ Gressgård K Hansen Reliability Engineering and System Safety 133 (2015) 167–175 169
employees Several scholars have acknowledged that a shared
context or knowledge base represents a significant condition for
the abilities of individuals to recognize understand and share
competencies and resources [74] Clearly defined competence
requirements may also result in employees knowing what to look
for and thus enhance knowledge seeking behavior [7576] Role
clarity may in other words influence employees' abilities to locate
understand interpret and absorb knowledge and thereby lead to
improved absorptive capacity [77] Further by facilitating the
development of shared meaning among employees role clarity is
also likely to influence knowledge contribution processes A clear
understanding of the wider organizational systems that their jobs
are part of may imply that employees more easily understand the
value of their knowledge and are better able to frame their
knowledge in a way that makes sense to potential acquirers
[6178] Thus the extent of role clarity of a job may influence the
abilities of employees to engage in knowledge exchange processes
through increased absorptive capacity and quality of own con
tributions leading to the following hypothesis
H4b Role clarity is positively related to knowledge exchange
(a) within units and (b) between units
233 Empowerment
Factors that influence the experienced meaningfulness and
perceived responsibility of work as well as knowledge of the
results of own work may be important for knowledge sharing
motivation [5979] Employee empowerment is a central factor
in this respect Work systems that are based on employee
involvement and empowerment are often referred to as high
performance work systems [8081] and research has shown that
the job characteristics associated with such systems are positively
related to satisfaction because employees experience meaningful
ness in their work greater responsibilities and control over task
completion and better use of knowledge and skills [808283]
According to Tomer [84] employees are also more cooperative in
highinvolvement work systems and this is supported by Srivas
tava et al [85] who found that empowering leadership fosters
knowledge sharing among team members In the context of
organizational safety research has shown that involvement of
employees including empowerment and delegation of responsi
bility for safety positively influence safety performance [16] and
perceptions of workplace safety [68] Barling et al [80] also argue
that increased work involvement promotes learning and enables
proactive problemsolving and preventive action On basis of
existing research we therefore expect that employee empower
ment represents a significant factor for knowledge exchange
H4c Empowerment is positively related to knowledge exchange
(a) within units and (b) between units
The hypothesized relationships are illustrated in Fig 1
3 Method
31 Description of survey
Data were collected through a webbased survey administered
to personnel involved with drilling and well operations in nine
different companies one operator company and eight of its main
contactors (operating on the Norwegian Continental Shelf) Two
different sets of questionnaires were developed one for the
operator employees and one for contractor employees The ques
tionnaires consisted of a total of 105 items covering various
aspects of the work situation like knowledge exchange job
characteristics perceptions of leadership work compliance etc
For some items the wordings were adapted to the target groups
(ie operator employees and contractor employees) although
measuring the same aspects
Invitation letters describing important details of the study were
distributed to the potential respondents via administrative per
sonnel in each company Distribution lists were provided by the
respective firms but all survey administration and coordination
was handled by the research team The total population was 5856
employees of which 1398 were employees of the operator and
4458 were employees of the contractors The survey was open
during a period of 6 weeks in order to cover all work shifts
offshore During this period two reminders were sent The total
number of responses obtained from these two groups was 2653
resulting in a response rate of 45 (880 responses from operator
employees and 1773 responses from contractor employees giving
response rates of 63 and 40 respectively)
32 Measures
Measures of contractor relationship management and work char
acteristics were based on previous research [4267] and adapted to
the context and requirements of this particular study The following
items were included Leadership involvement My leader participates
actively in planning and preparing the work My leader system
atically follows up the execution of the work My leader contributes
to a good cooperation between unitsinvolved groups Role clarity
The responsibilities of my position are unambiguously documen
ted The authority of my position is unambiguously documented
The skill requirements for my present position are clearly documen
ted Empowerment I am able to utilize my expertise and abilities in
my present position Iamsufficiently involved inhave a say on
decision related to my work situation I receive the necessary
training to handle new work tasks and responsibilitiesAsthe
sample involved employees of both the operator and contractors
two different sets of items were applied to measure the construct of
contractor relationship management Operator respondents In my
unit we closely follow up contractorssuppliers we work with In
my unit we systematically follow up the feedback we receive from
Fig 1 Research model with hypotheses
LJ Gressgård K Hansen Reliability Engineering and System Safety 133 (2015) 167–175170
supplierscontractors The supplierscontractors we work with have
received the training they need to carry out their tasks in a safe
manner Contractor respondents [Name of the operator]'s follow
up of the company I work for is good Our feedback to [Name of the
operator] is systematically followed up [Name of operator] makes
sure we get the training required to accomplish our tasks in a safe
manner All items were measured by use of a 6point scale ranging
from totally disagree (1) to fully agree (6)
Factor analysis of the items resulted in a 4factor solution that
accounted for 71 of the total variance All items had sufficient
loadings (above 040) on a single factor (no variables had factor
loading above 040 on more than one factor) Cronbach's alphas
for the four constructs were 088 (leadership involvement) 081
(role clarity) 068 (empowerment) and 079 (contractor relation
ship management) The alpha score for the empowerment scale
was 068 which is slightly lower than the suggested threshold of
070 [86] but values down to 060 can be deemed acceptable [87]
Prior to the knowledge exchange questions the respondents
were given the following information regarding this particular
topic We are interested in the exchange of advice and informa
tion participation in transfer of experiences and other forms of
knowledge transfer This concerns your daily accomplishment of
tasks questions about methods and choice of technology etc
They were thereafter asked to rate the extent to which such
knowledge exchange occurs in your entityinstallation and
between entitiesinstallations on a 6point scale ranging from
none to very much The means and standard deviations were
4789 and 4298 respectively In order to measure organizations'
abilities to learn from failures the respondents were asked to rate
the extent to which they perceive their company to learn from
mistakes being made (In my company we learn from mistakes)
The mean and standard deviation of this measure were 44710)
The constructs of knowledge sharing and failurebased learning
may involve different aspects depending on work situation and
context As described in section 21 exchange of knowledge may
involve provision or receipt of task information knowhow or feed
back regarding a product or procedure [2526] and may occur
through formal and informal personal interaction or standardized
knowledge exchange systems Failurebased learning refers to the use
of relevant knowledge for improvement of work by identifying and
discussing work patterns that need to be modified or changed [7]
These constructs are therefore measured by global singleitem ques
tions as such measures allow a respondent to consider all aspects
and individual preferences of the certain aspects of the construct
being measured [88][p79]Askingsingleitemquestionsthus
assumes that respondents automatically consider different aspects
of the construct and thereby also ignore aspects that are not relevant
to their situations and differentially weight the relevant aspects to
provide a single rating [8990]
4 Results
Table 1 lists the correlation statistics of the variables included
in the study From the table we find that all correlations are
moderate which indicates that multicollinearity between the
independent variables is not a problem
In order to test H1 and H2 multiple regression analysis with
knowledge exchange within units and knowledge exchange
between units as predictor variables and ability to learn from
failures as dependent variable was first conducted The results
show that both predictors have significant effects on the depen
dent variable (t¼48 β¼012 p¼000 for knowledge exchange
within units and t¼115 β¼028 p¼000 for knowledge exchange
between units R2¼013) These results provide preliminary sup
port for the hypotheses However as the effects of these variables
may depend on the other predictor variables in the model
(illustrated by the dotted lines in Fig 1) hierarchical multiple
regression analysis including work characteristics and contractor
relationship management (as predictors) was conducted in order
to test whether knowledge exchange variables had significant
effects above and beyond the other predictors Leadership involve
ment role clarity and empowerment were included in step (1)
contractor relationship management was included in step (2) and
knowledge exchange within and between units were included in
step (3) in the regression The results are presented in Table 2
Table 2 shows that knowledge exchange between units is the
most important predictor of perceived ability to learn from fail
ures (β¼018) and has a significant effect above and beyond the
other predictor variables Empowerment (β¼017) and role clarity
(β¼016) are also important variables for firms' abilities to learn
from failures The total model explains 26 of the variance in the
dependent variable and we see that introduction of the knowl
edge exchange variables (in step (3) of the regression) leads to a
slight but significant increase in explained variance (up from 23
in step (2)) We also see that knowledge exchange within units
turns out insignificant when the other predictors are included in
the model These results provide support for H1b and H2 while
Table 1
Pearson's correlation matrix
Variable Mean SD 1 2345
1 Leadership involvement 458 095
2 Role clarity 453 092 044n
3 Empowerment 488 067 045n 053n
4 Contractor relationship management 443 082 035n 029n 045n
5 Knowledge exchange within units 465 089 028n 024n 032n 032n
6 Knowledge exchange between units 417 098 026n 026n 031n 030n 061n
n All correlations are significant at the po001 level
Table 2
Effects of knowledge exchange contractor relationship management and work
characteristics on firms' abilities to learn from failures
Predictor variables Dependent variable Ability to learn
from failures
Step 1 Step 2 Step 3
t β t β t β
Leadership involvement 60 013n 47 011n 39 009n
Role clarity 78 019n 77 18n 69 016n
Empowerment 103 25n 80 20n 69 17n
Contractor rel management 59 13n 42 09n
Knowledge exchange within units 12 03
Knowledge exchange between units 75 18n
ΔR2 02 03
ΔF 343n 457n
R2 021 023 026
F 1883n 1521n 1210n
n p o0001 β represents standardized β coefficients
LJ Gressgård K Hansen Reliability Engineering and System Safety 133 (2015) 167–175 171
H1a is not supported We should also notice that contractor
relationship management has a direct effect on perceived ability
to learn from failures (β¼009) in addition to the indirect effect
through knowledge exchange between units (β¼017 see Table 3)
H3 and H4 were tested by use of two separate hierarchical
multiple regression analyses with knowledge exchange within
units and knowledge exchange between units as dependent
variables In step (1) of the analyses the work characteristics (ie
leadership involvement role clarity and empowerment) were
included while contractor relationship management was included
in step (2) The results of the analyses are shown in table 3
Table 3 shows that contractor relationship management has
significant positive effects on knowledge exchange within units
(β¼018) and knowledge exchange between units (β¼017) H3a
and H3b are thus supported By introducing this variable in the
model explained variance increases from 13 to 16 for knowl
edge exchange within units and from 13 to 15 for knowledge
exchange between units We further see that all work character
istics have significant effects on the dependent variables except
for role clarity on knowledge exchange within units This means
that H4a–a H4a–b H4b–b H4c–a and H4c–b are supported while
H4b–a should be rejected Finally the results show that contractor
relationship management is the most important predictor of
knowledge exchange both within and between unitsinstallations
As the arguments leading to the expected relationship between
contractor relationship management and knowledge exchange
between units (cf discussion in section 22) to some extent were
based on the expectation of differences in mobility of contractor
and operator employees this represents an important premise for
H3b Our expectation is confirmed as descriptive statistics show
that 41 of operator employeesrespondents work at one specific
installation while the corresponding number for contractor
employees is 31 The remaining employees work at several
installations (or not at any specific installation) Contractor
employees thus move around (between locations) to a larger
extent than operator employees and may therefore represent
important sources for distribution of knowledge across borders
In sum the results of the tests of hypotheses show that only
knowledge exchange between units has a positive significant
effect on firms' abilities to learn from failures which leads to
rejection of H1a while H1b and H2 are supported H3a and H3b are
also supported as contractor relationship management has posi
tive effects on knowledge exchange both within and between
units Finally the results provide support for the hypothesized
effects of work characteristics (leadership support role clarity and
empowerment) except for H4b–a (lack of significant relationship
between role clarity and knowledge exchange within units)
5 Discussion
The results of this study increase our understanding of varia
tions in firms' abilities to learn from failures and thus make a
contribution to answering the question how such learning is
enabled [15] The results show in this respect that knowledge
exchange between units is central in complex distributed inter
organizational systems An explanation of this observation can be
related to the importance of diversity of opinions and perspectives
in order to build and maintain organizational resilience in complex
systems According to Reason [91] organizations that experience
incidents and accidents often focus on active (human) failures
rather than trying to dig deeper and uncover problematic latent
conditions However there is isomorphism between error com
plexity and technicalorganizational complexity [92] and diversity
in causes of failures (ie failures involving multiple factors and
complex interactions) will therefore force organizations to avoid
such simple explanations With reference to this perspective
knowledge exchange between units may increase the heteroge
neity of perspectives and thereby force organizations to look
beyond simple causes and responses Diversity of knowledge bases
and opinions thus reduce the tendency of organizations to focus
on the surface when attempting to learn from failures and may
therefore promote the organizational function logic instead of
individual blame [93] when explaining the origins and dynamics
of failures This is further consistent with the view that diverse
information stimulates constructive conflict around issues which
leads people to deliberate about appropriate action [92] This again
should lead to a better understanding of the problem and to
solutions that reduce future failures Knowledge diversity may
thus reduce the likelihood of failures by increasing employees'
abilities to adopt to changes and handle unplanned situations and
thereby promote resilience that allow people to produce success
when failure threatens [94] [p2]
Interpretation of the results along this line of thought is
consistent with Weick's [95] discussion of requisite variety and
emphasis on the benefits of increasing system variety for reducing
errors Such variety can be increased through several mechanisms
like promoting individual diversity focus on facetoface interaction
modes reduction of bureaucratic rigidity and promoting individual
discretion over decisions [92] Applying this theoretical lens our
results may indicate that contractor employees represent a source
for individual diversity and that the heterogeneity of perspectives
that result from efficient contractor relationship management is
positive for system variety This argument should be related to the
claim of Goodman and Darr [32] that geographically distributed
units of the same company are likely to have similar problems and
Table 3
Effects of contractor relationship management and work characteristics on knowledge exchange
Predictor variables Dependent variable Knowledge exchange…
…within unitsinstallations …between unitsinstallations
Step 1 Step 2 Step 1 Step 2
t β t β t β t β
Leadership involvement 69 016n 55 013n 52 012n 38 009n
Role clarity 18 004 16 004 45 011n 44 011n
Empowerment 91 23n 62 016n 80 020n 53 014n
Contractor rel management 81 018n 75 017n
ΔR2 003 002
ΔF 653n 557n
R2 013 016 013 015
F 1068n 989n 1028n 931n
n p o0001 β represents standardized β coefficients
LJ Gressgård K Hansen Reliability Engineering and System Safety 133 (2015) 167–175172
solutions but may focus on local unit activities due to their limited
task interdependencies or interpersonal ties In such work contexts
contractor employees may help increase crossunit knowledge ex
change because of higher levels of mobility The results thus provide
support for the view that the quality of interfirm relationships is an
enabler of learning [465154]
The results also show that work characteristics are important for
knowledge exchange and firms' abilities to learn from failures Role
clarity and empowerment were the most significant factors in this
respect which indicates that knowledge exchange processes and
organizational learning to some extent depend on employees'
understanding of their own position and work role in the wider
organizational system It further provides support for the view that
highperformance work systems represent supportive environ
ments for knowledge sharing and that participative management
[73] promotes failurebased learning With reference to Weick's
[95] emphasis on system variety it can be argued that these work
characteristics positively influence individual discretion through
better understanding of own capabilities and responsibilities as
well as motivation and work involvement Our study thus responds
to the call for research by Grant [64] and Foss et al [59] that there is
a need for studies that provide a deeper understanding of a wider
set of job characteristics than the limited set defined largely by
Hackman and Oldham's [96] model
On a practical level the results of the study increase our
understanding of forces for and against knowledge exchange in
distributed environments and thereby also improve our knowl
edge of how organizational learning processes (particularly learn
ing from failures) in distributed environments can be supported
This is of particular importance in complex and highrisk systems
like offshore drilling as the isomorphism between error complex
ity and technical complexity [92] makes it fundamental to imple
ment mechanisms to uncover problematic latent and interacting
system components and thereby prevent organizations from
making simple conclusions regarding failure causes The results
show in this respect that organizations should aim at developing
efficient systems for followup of contractors as well as designing
work environments that promote employees' empowerment and
understanding of work roles These factors thus represent impor
tant capabilities that enable organizations to extract distribute
and apply useful information from failures made in various parts
of the organizational system Contractor relationship management
and work characteristics should therefore be emphasized in
incident learning systems [13] and riskhazard analyses [9798]
in order to increase our understanding of human and organiza
tional factors as well as initiatives toward resilient collaboration
[99] to promote crossborder knowledge exchange leading to
improvement of organizational performance
6 Conclusions
Overall the results show that firms' abilities to learn from
failures in distributed environments are influenced by the degree
of knowledge exchange between units and the quality of con
tractor relationship management and work characteristics (role
clarity empowerment and leadership involvement) The study
underscores that researchers and practitioners within the field of
system reliability and safety need to focus on these aspects
However the conclusion should be considered together with the
research limitations First the study was based on crosssectional
data which means that causality between variables cannot be
tested statistically and is therefore solely based on theoretical
reasoning Second use of global indicators of knowledge exch
ange and firms' abilities to learn from failures also represents a
significant limitation resulting in crude measures of the variables
That is both knowledge exchange and failurebased learning
can involve different aspects (ie types of knowledge that are
exchanged and failures that are made) and the measures do not
capture this variation Also the measures applied assess the
respondents' perceptions rather than objective indicators Future
research should therefore apply more rigorous (subjective and
objective) measures of both knowledge exchange and failure
based learning and thereby focusing on various aspects of the
concepts Consequences of such learning should also be studied
like changes in behavior routines etc Building on the call for
research by Grant [64] and Foss et al [59] there is also need for
studies that investigate how knowledge exchange and failure
based learning in distributed environments are influenced by a
wider set of work characteristics (eg work autonomy specializa
tion task identity etc) and various aspects of contractor relation
ships (eg trust power opportunism etc) Finally we should note
that knowledge exchange is measured by employees' assessments
(perceptions) of the extent of knowledge exchange within and
across units Future research should therefore more directly
measure knowledge sharing behavior and investigate antecedents
and effects of this regarding learning from failures and safety
behavior (eg risk assessments)
Acknowledgments
The paper is based on data from a research project funded by
an operator company in Norway Representatives from the com
pany were involved in the questionnaire design process (indicator
development)
References
[1] Catino M Patriotta G Learning from errors cognition emotions and safety
culture in the Italian Air Force Organ Stud 201334437–67
[2] Carmeli A Sheaffer Z How learning leadership and organizational learning
from failures enhance perceived organizational capacity to adapt to the task
environment J Appl Behav Sci 200844468–89
[3] Ron N Lipshitz R Popper M How organizations learn postflight reviews in
an F16 fighter squadron Organ Stud 2006271069–89
[4] Madsen PM Desai V Failing to learn The effects of failure and success on
organizational learning in the global orbital launch vehicle industry Acad
Manage J 201053451–76
[5] Weick KE Sutcliffe K Managing the unexpected resilient performance in an
age of uncertainty San Francisco CA Jossey Bass 2007
[6] Carmeli A Schaubroeck J Organisational crisispreparedness the importance
of learning from failures Long Range Plan 200841177–96
[7] Hirak R Peng AC Carmeli A Schaubroeck JM Linking leader inclusiveness to
work unit performance The importance of psychological safety and learning
from failures Leadersh Q 201223107–17
[8] Tucker AL Edmondson AC Why hospitals don't learn from failures organiza
tional and psychological dynamics that inhibit system change Calif Manage
Rev 20034555–72
[9] Edmondson AC Learning from failure in health care frequent opportunities
pervasive barriers Qual Saf Health Care 2004133–9
[10] Wang Z Wang N Knowledge sharing innovation and firm performance
Expert Syst Appl 2012398899–908
[11] Szulanski G Exploring internal stickiness impediments to the transfer of best
practice within the firm Strategic Manag J 19961727–43
[12] Szulanski G The process of knowledge transfer a diachronic analysis of
stickiness Organ Behav Hum Decis Processes 2000829–27
[13] Cooke DL Rohleder TR Learning from incidents from normal accidents to
high reliability Syst Dynam Rev 200622213–39
[14] Baumard P Starbuck WH Learning from failures why it may not happen
Long Range Plan 200538281–98
[15] Carmeli A Gittell JH Highquality relationships psychological safety and
learning from failures in work organizations J Organ Behav 200930709–29
[16] Mearns K Whitaker SM Flin R Safety climate safety management practice
and safety performance in offshore environments Saf Sci 200341641–80
[17] Rosness R Blakstad HC Forseth U Dahle IB Wiig S Environmental conditions
for safety work – theoretical foundations Saf Sci 2012501967–76
[18] Reagans R Argote L Brooks D Individual experience and experience working
together predicting learning rates from knowing who knows what and
knowing how to work together Manage Sci 200551869–81
LJ Gressgård K Hansen Reliability Engineering and System Safety 133 (2015) 167–175 173
[19] Turner BA Toft B Organizational learning from disasters In Smith D Elliot
D editors Key readings in crisis management systems and structures for
prevention and recovery London Ruthledge 2006 p 191–204
[20] Størseth F Tinmannsvik RK The critical reaction learning from accidents Saf
Sci 2012501977–82
[21] Mahler J Casamayou MH Organizational learning at NASA The Challenger
and Columbia accidents Washington DC Georgetown University Press 2009
[22] Tjosvold D Ziyou Y Chun H Team learning from mistakes the contribution
of cooperative goals and problemsolving J Manage Stud 2004411223–45
[23] Argote L MironSpektor E Organizational learning from experience to
knowledge Organ Sci 2011221123–37
[24] Edmondson AC The local and variegated nature of learning in organizations
a grouplevel perspective Organ Sci 200213128–46
[25] Hansen MT The searchtransfer problem the role of weak ties in sharing
knowledge across organization subunits Adm Sci Q 19994482–111
[26] Cummings JN Work groups structural diversity and knowledge sharing in a
global organization Manage Sci 200450352–64
[27] Wang S Noe RA Knowledge sharing a review and directions for future
research Hum Resour Manage Rev 201020115–31
[28] Edmondson AC Bohmer RM Pisano GP Disrupted routines team learning
and new technology implementation in hospitals Adm Sci Q
200146685–716
[29] Pasman HJ Knegtering B Rogers WJ A holistic approach to control process
safety risks possible ways forward Reliab Eng Syst Saf 201311721–9
[30] Doytchev D Hibberd RE Organizational learning and safety in design
experiences from German industry J Risk Res 200912295–312
[31] Johnson CW Failure in SafetyCritical Systems A Handbook of Accident and
Incident Reporting Glasgow Scotland University of Glasgow Press 2003
[32] Goodman PS Darr ED Computeraided systems and communities mechan
isms for organizational learning in distributed environments MIS Q
199822417–40
[33] van Wijk R Jansen JJP Lyles MA Inter and intraorganizational knowledge
transfer a metaanalytic review and assessment of its antecedents and
consequences J Manage Stud 200845830–53
[34] DavisBlake A Broschak JP Outsourcing and the changing nature of work
Annu Rev Sociol 200935321–40
[35] Anand N Daft RL What is the right organization design Organ Dyn
200736329–44
[36] Read C BP and the Macondo spill the complete story New York NY Palgrave
Mcmillan 2011
[37] Powell WW Koput KW SmithDoerr L Interorganizational collaboration and
the locus of innovation networks of learning in biotechnology Adm Sci Q
199641116–45
[38] Grant RM Prospering in dynamicallycompetitive environments organiza
tional capability as knowledge integration Organ Sci 19967375–87
[39] Paulraj A Lado AA Chen IJ Interorganizational communication as a relational
competency antecedents and performance outcomes in collaborative buyer—
supplier relationships J Oper Manage 20082645–64
[40] Kogut B Zander U Knowledge of the firm combinative capabilities and the
replication of technology Organ Sci 19923383–97
[41] Petroleum Safety Authority Norway (PSA) The contractors important role big
HSE responsibility Available at 〈httpwwwpsanonewsthecontractorsim
portantrolebighseresponsibilityarticle2926878html〉
[42] Dahl Ø Olsen E Safety compliance on offshore platforms a multisample
survey on the role of perceived leadership involvement and work climate Saf
Sci 20135417–26
[43] Humphreys PK Li WL Chan LY The impact of supplier development on
buyer–supplier performance Omega 200432131–43
[44] Li W Humphreys PK Yeung ACL Edwin Cheng TC The impact of specific
supplier development efforts on buyer competitive advantage an empirical
model Int J Prod Econ 2007106230–47
[45] Li W Humphreys PK Yeung ACL Cheng TCE The impact of supplier
development on buyer competitive advantage a path analytic model Int J
Prod Econ 2012135353–66
[46] Rebolledo C Nollet J Learning from suppliers in the aerospace industry Int J
Prod Econ 2011129328–37
[47] Mayer KJ Teece DJ Unpacking strategic alliances the structure and purpose of
alliance versus supplier relationships J Econ Behav Organ 200866106–27
[48] Carr AS Kaynak H Communication methods information sharing supplier
development and performance Int J Oper Prod Manage 200727346–70
[49] Bessant J Kaplinsky R Lamming R Putting supply chain learning into practice
Int J Oper Prod Manage 200323167–84
[50] Inkpen AC Learning through joint ventures a framework of knowledge
aquisition J Manage Stud 2000371019–43
[51] Kale P Singh H Perlmutter H Learning and protection of proprietary assets in
strategic alliances Building relational capital Strategic Manage J 200021
217–37
[52] Spekman RE Spear J Kamauff J Supply chain competency learning as a key
component Supply Chain Manage An Int J 2002741–55
[53] Kellogg KC Orlikowski WJ Yates J Life in the trading zone Structuring
coordination across boundaries in postbureaucratic organizations Organ Sci
20061722–44
[54] Amesse F Dragoste L Nollet J Ponce S Issues on partnering evidences from
subcontracting in aeronautics Technovation 200121559–69
[55] Song J Almeida P Wu G Learningbyhiring when is mobility more likely to
facilitate interfirm knowledge transfer Manage Sci 200349351–65
[56] Burgess D What motivates employees to transfer knowledge outside their
work unit J Bus Commun 200542324–48
[57] Fisher RJ Maltz E Jaworski BJ Enhancing communication between marketing
and engineering the moderating role of relative functional identification
J Market 19976154–70
[58] Tsai W Social structure of coopetition within a multiunit organization
coordination competition and intraorganizational knowledge sharing Organ
Sci 200213179–90
[59] Foss NJ Minbaeva DB Pedersen T Reinholt M Encouraging knowledge
sharing among employees how job design matters Hum Resour Manage
200948871–93
[60] Minbaeva D Pedersen T Björkman I Fey CF Park HJ MNC knowledge transfer
subsidiary absorptive capacity and HRM J Int Bus Stud 200334586–99
[61] Reinholt MIA Pedersen T Foss NJ Why a central network position isn't
enough The role of motivation and ability for knowledge sharing in employee
networks Acad Manage J 2011541277–97
[62] Argote L McEvily B Reagans R Managing knowledge in organizations an
integrative framework and review of emerging themes Manage Sci
200349571–82
[63] Cabrera Á Collins WC Salgado JF Determinants of individual engagement in
knowledge sharing Int J Hum Resour Manage 200617245–64
[64] Grant AM Relational job design and the motivation to make a prosocial
difference Acad Manage Rev 200732393–417
[65] Stewart GL Barrick MR Team structure and performance assessing the
mediating role of intrateam process and the moderating role of task type
Acad Manage J 200043135–48
[66] Wageman R Interdependence and group effectiveness Adm Sci Q
199540145–80
[67] Tharaldsen JE Knudsen K Næss S Monitoring integration and measuring
progress In Colman HL Stensaker I Tharaldsen JE editors A Merger of
Equals The Integration of Statoil and Hydro's Oil & Gas Activities Bergen
Fagbokforlaget 2011
[68] Tharaldsen JE Stensaker I Gressgård LJ The impact of organizational integra
tion on safety trust and identity In Colman HL Stensaker I Tharaldsen JE
editors A Merger of Equals The Integration of Statoil and Hydro's Oil & Gas
Activities Bergen Fagbokforlaget 2011
[69] Husted K Michailova S Diagnosing and fighting knowledgesharing hostility
Organ Dyn 20023160–73
[70] Zhao B Olivera F Error reporting in organizations Acad Manage Rev
2006311012–30
[71] Edmondson AC Psychological safety trust and learning in organizations A group
level lens In Kramer RM Cook KS editors Trust and Distrust in Organizations
Dilemmas and Approaches New York Russell Sage Foundation 2004
[72] Nahrgang JD Morgeson FP Hofmann DA Safety at work a Metaanalytic
investigation of the link between job demands job resources burnout
engagement and safety outcomes J Appl Psychol 20119671–94
[73] O'Dea A Flin R Site managers and safety leadership in the offshore oil and gas
industry Saf Sci 20013739–57
[74] Kang SC Morris SS Snell SA Relational archetypes organizational learning
and value creation extending the human resource architecture Acad Manage
Rev 200732236–56
[75] He W Wei KK What drives continued knowledge sharing An investigation
of knowledgecontribution and seeking beliefs Decis Support Syst
200946826–38
[76] Kankanhalli A Tan BCY Wei KK Understanding seeking from electronic
knowledge repositories an empirical study J Am Soc Inf Sci Technol
2005561156–66
[77] Cohen WM Levinthal DA Absorptive capacity a new perspective on learning
and innovation Adm Sci Q 199035128–52
[78] Reagans R McEvily B Network structure and knowledge transfer the effects
of cohesion and range Adm Sci Q 200348240–67
[79] Parker S Wall T Job and work design organizing work to promote wellbeing
and effectiveness London Sage 1998
[80] Barling J Kelloway EK Iverson RD Highquality work job satisfaction and
occupational injuries J Appl Psychol 200388276–83
[81] Lawler E The ultimate advantage creating the highinvolvement organization
San Francisco JosseyBass 1992
[82] Berg P The effects of high performance work practices on job satisfaction in
the United States steel industry Relat IndInd Relat 199954111–34
[83] Godard J High performance and the transformation of work The implications
of alternative work practices for the experience and outcomes of work Ind
Labor Relat Rev 200154776–805
[84] Tomer JF Understanding highperformance work systems The joint contribu
tion of economics and human resource management Journal of Socio
Economics 20013063–73
[85] Srivastava A Bartol KM Locke EA Empowering leadership in management
teams effects on knowledge sharing efficacy and performance Acad Manage
J 2006491239–51
[86] Nunnally JC Psychometric theory New York McGrawHill 1978
[87] Hair JF Anderson RE Tatham RL Black WC Multivariate data analysis 5th ed New
Jersey PrenticeHall 1998
[88] Nagy MS Using a singleitem approach to measure facet job satisfaction
J Occup Organ Psychol 20027577–86
[89] Fuchs C Diamantopoulos A Using singleitem measures for construct mea
surement in management research Conceptual issues and application guide
lines Die Betriebswirtschaft 200969195–210
LJ Gressgård K Hansen Reliability Engineering and System Safety 133 (2015) 167–175174
[90] de Boer A van Lanschot J Stalmeier P van Sandick J Hulscher J de Haes J
et al Is a singleitem visual analogue scale as valid reliable and responsive as
multiitem scales in measuring quality of life Qual Life Res 200413311–20
[91] Reason J Managing the risks of organizational accidents Sydney Ashgate
1997
[92] Haunschild PR Sullivan BN Learning from complexity effects of prior
accidents and incidents on airlines' learning Adm Sci Q 200247609–43
[93] Catino M A review of literature individual blame vs organizational function
logics in accident analysis J Conting Crises Manage 20081653–62
[94] Hollnagel E Woods DD Leveson N Resilience engineering concepts and
precepts Aldershot UK Ashgate 2006
[95] Weick KE Organizational culture as a source of high reliability Calif Manage
Rev 198729112–27
[96] Hackman JR Oldham GR Work redesign reading MA AddisonWesley 1980
[97] Skogdalen JE Vinnem JE Quantitative risk analysis of oil and gas drilling
using Deepwater Horizon as case study Reliab Eng Syst Saf 201210058–66
[98] Skogdalen JE Vinnem JE Quantitative risk analysis offshore—human and
organizational factors Reliab Eng Syst Saf 201196468–79
[99] Skjerve AB Kaarstad M Størseth F Wærø I Grøtan TO Planning for resilient
collaboration at a new petroleum installation – a case study of a coaching
approach Saf Sci 2012501952–9
LJ Gressgård K Hansen Reliability Engineering and System Safety 133 (2015) 167–175 175