Behavioral Integrity for Safety, Priority of Safety, Psychological Safety, and Patient Safety: A Team-Level Study

Abstract

This article clarifies how leader behavioral integrity for safety helps solve follower's double bind between adhering to safety protocols and speaking up about mistakes against protocols. Path modeling of survey data in 54 nursing teams showed that head nurse behavioral integrity for safety positively relates to both team priority of safety and psychological safety. In turn, team priority of safety and team psychological safety were, respectively, negatively and positively related with the number of treatment errors that were reported to head nurses. We further demonstrated an interaction effect between team priority of safety and psychological safety on reported errors such that the relationship between team priority of safety and the number of errors was stronger for higher levels of team psychological safety. Finally, we showed that both team priority of safety and team psychological safety mediated the relationship between leader behavioral integrity for safety and reported treatment errors. These results suggest that although adhering to safety protocols and admitting mistakes against those protocols show opposite relations to reported treatment errors, both are important to improving patient safety and both are fostered by leaders who walk their safety talk. (PsycINFO Database Record (c) 2012 APA, all rights reserved).

Full text

Running head: LEADER BEHAVIORAL INTEGRITY FOR SAFETY 1
Behavioral Integrity for Safety, Priority of Safety, Psychological Safety, and Patient Safety:
A Team-Level Study
Hannes Leroy
University of Calgary, Haskayne School of Business
Bart Dierynck
Tilburg University
Frederik Anseel
Ghent University
Tony Simons
Cornell University
Jonathon R. B. Halbesleben
The University of Alabama
Deirdre McCaughey
The Pennsylvania State University
Grant T. Savage
University of Alabama at Birmingham
Luc Sels
University of Leuven
In press, Journal of Applied Psychology, June 2012
Do not quote without permission of author; original manuscript (unproofed) for publication.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 1
Abstract
This paper clarifies how leader behavioral integrity for safety helps solve follower’s double
bind between adhering to safety protocols and speaking up about mistakes against protocols.
Path modelling of survey data in 54 nursing teams showed that head nurse behavioral
integrity for safety positively relates to both team priority of safety and psychological safety.
In turn, team priority of safety and team psychological safety were, respectively, negatively
and positively related with the number of treatment errors that were reported to head nurses.
We further demonstrated an interaction effect between team priority of safety and
psychological safety on reported errors such that the relationship between team priority of
safety and the number of errors was stronger for higher levels of team psychological safety.
Finally, we showed that both team priority of safety and team psychological safety mediated
the relationship between leader behavioral integrity for safety and reported treatment errors.
These results suggest that while adhering to safety protocols and admitting mistakes against
those protocols show opposite relations to reported treatment errors, both are important to
improving patient safety and both are fostered by leaders who walk their safety talk.
Keywords: Behavioral integrity, priority of safety, psychological safety, treatment errors,
safety climate, leadership.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 2
Behavioral Integrity for Safety, Priority of Safety, Psychological Safety, and Patient Safety:
A Team-Level Study
Employee reporting of errors is considered to be an indicator of a healthy functioning
organization (Argyris, 1977). Many leaders, however, experience problems in empowering
employees to report errors. In many cases empowerment is an emperor in new clothes
(Argyris, 1998), with leaders advocating strict adherence to company protocols, thus
scrutinously avoiding errors, while at the same time hoping for the reporting of errors against
those same company protocols (Kerr, 1975). As a result, employees may experience a double
bind between these seemingly conflicting behaviors (Argyris, 1977, p. 117): “When
employees adhere to a norm that says "hide errors," they know they are violating another
norm that says "reveal errors." The employees are thus in a double bind”.
This double bind and the role of leaders herein have been shown to be especially
salient in the context of hospitals and patient safety. Katz-Navon, Naveh, and Stern (2009),
for instance, empirically demonstrate that employees may experience a conflict between the
strong enforcement of safety procedures on the one hand and the reporting and learning from
errors on the other hand. Despite this conflict, prior research indicates that a climate of safety
requires both prioritizing existing safety protocols and constructive responses to errors
(Hofmann & Mark, 2006). Prior research also points to the importance of leadership to foster
team priority of safety (Katz-Navon, Naveh, & Stern., 2005; Zohar, 2000, 2002; Zohar &
Luria, 2004; Zohar & Luria, 2010; Zohar & Tenne-Gazit, 2008) as well as team
psychological safety (Edmondson, Edmondson, 1999, 2003, 2004; Nembhard & Edmondson,
2006). However, to date empirical research has remained surprisingly silent about which
leader characteristics may help reconcile the seemingly conflicting demands of closely
adhering to safety protocols and reporting errors against those safety protocols.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 3
Early theorists suggested that leaders can help alleviate the double bind by
communicating more clearly on what they truly value (Bateson, Jackson, Haley, &
Weakland, 1956; Argyris, 1977). More recently, Zohar (2010) highlighted the importance of
leader behavioral integrity for safety when it comes to alleviating ambiguities regarding
safety. Behavioral integrity for safety reflects the extent to which leaders “walk the talk” or
“practice what they preach” concerning safety (Simons, 2002) and is proposed to influence
safety outcomes through two mechanisms (Simons, 2008). First, leader follow-through on
safety values sends unequivocal messages that safety is valued, thus fostering a high team
priority for safety (Zohar, 2010). Second, alignment between words and deeds signals that the
leader’s concern for safety is genuine and trustworthy, thus fostering high psychological
safety within the workteam (Edmondson, 2004). In turn, by fostering both aspects of safety
climate simultaneously, leaders who display behavioral integrity for safety will promote
overall patient safety because their clarity of communication not only enforces safety
protocols but also offers a safe haven where followers feel safe to speak up about errors
against those safety protocols.
Thus, whereas previous research has predominantly looked at how different aspects
of safety climate influence patient safety outcomes (Hofmann & Mark, 2006; Katz-Navon et
al., 2005; Katz-Navon et al., 2009), we aim to extend this line of research by increasing our
understanding of the unique role leaders play in fostering distinct aspects of safety climate
and thereby improving patient safety outcomes. More specifically, we use theory and
research on behavioral integrity (Simons, 2002, 2008) to posit that behavioral integrity for
safety fosters patient safety by alleviating the double bind that followers may experience
between adhering to safety protocols on the one hand and reporting errors against those safety
protocols on the other hand.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 4
Development of Hypotheses
Behavioral Integrity for Safety, Priority of Safety, and Psychological Safety
Simons (1999) argued that leader behavioral integrity is especially important in a work
environment that poses high and diverse demands on employees. When leaders do not
follow-up on the values they espouse, employees receive ambiguous messages about how the
leader values specific behaviors in comparison to other pressing role demands. In support of
this claim, Dineen, Lewicki, and Tomlinson (2006) found that supervisory guidance had a
positive relationship with organizational citizenship behaviors when behavioral integrity was
high and a positive relationship with deviant behavior when behavioral integrity was low.
Supervisory guidance regarding safety has previously been investigated as managerial
safety expectations and safety actions (Zohar, 2000). Recently, Zohar (2010) suggested that
alignment between espoused and enacted safety values is especially important for fostering a
high priority of safety in the work unit. The alignment between words and deeds sends clear
signals to the team about the value of safety in comparison to other pressing role demands.
Hypothesis 1. Leader behavioral integrity for safety positively relates to team priority of
safety.
Simons (2002) argued that keeping one’s word makes an individual more predictable
and is thus a key element to signaling that one is a reliable and trustworthy individual. In
support of this idea, Simons, Friedman, Liu, and McLean Parks (2007) found that perceptions
of leader behavioral integrity are positively related to employees’ trust in leaders. Trust in
the leader is important as it reflects employee’s willingness to be vulnerable based on the
expectation that leaders will not abuse this vulnerability (Rousseau, Sitkin, Burt, & Camerer,
1998). As an example, risky but valuable learning and interpersonal behaviors such as
admitting personal mistakes—which make the employee vulnerable to supervisory
discipline—are less likely to be abused by a trustworthy leader (Edmondson, 1999).

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 5
Edmondson (2004) further advocated that part of creating a psychologically safe
working environment concerns establishing clear boundaries about which behaviors are and
are not valued. In following through on espoused safety values, leaders that display
behavioral integrity for safety establish such clear boundaries (Zohar & Luria, 2010).
Furthermore, when leaders practice the safety values they preach, team members perceive
their leader’s concern for safety as more genuine and therefore are more likely to speak up
about safety issues because they believe their leader values these discussions.
Hypothesis 2. Leader behavioral integrity for safety positively relates to team psychological
safety.
Priority of Safety, Psychological Safety, and Patient Safety
In this paper we conceptualize patient safety as a lack of errors in the treatment of
patients (Katz-Navon et al., 2005; Naveh, Katz-Navon, & Stern, 2005). Previous research has
demonstrated that a positive safety climate consists of both adherence to safety protocols and
constructive responses to errors (Hofmann & Mark, 2006). Katz-Navon et al. (2009) further
disentangle the safety climate dimensions of priority of safety and psychological safety, as
both may have unique and contrasting effects on the number of reported treatment errors. On
the one hand, team priority of safety will relate negatively to the number of reported
treatment errors because team members who prioritize existing safety protocols are more
mindful of safety procedures in the execution of their work (Katz-Navon et al., 2005; Naveh
et al., 2005). Team psychological safety, on the other hand, will relate positively to the
number of reported treatment errors as team members who feel they are in a psychologically
safe environment are more willing to honestly report the number of errors that were
experienced (Edmondson, 2004; Tucker & Edmondson, 2003).
Hypothesis 3. Team priority of safety negatively relates to reported treatment errors.
Hypothesis 4. Team psychological safety positively relates to reported treatment errors.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 6
While both priority of safety and psychological safety have reverse relationships with
reported treatment errors (Katz-Navon et al., 2009), both are necessary ingredients for an
overall safety climate (Hoffman & Mark, 2006). A higher priority for safety protocols
without the willingness to openly admit errors against safety protocols may reflect a fear of
reprimand (e.g. from the leader). In contrast, team priority of safety will reflect a more
genuine concern for existing safety protocols when team members are also willing to report
and learn from their mistakes (Hoffman & Mark, 2006). We posit that team psychological
safety will strengthen the negative relationship between team priority of safety and the
number of reported treatment errors.
Hypothesis 5. Team priority of safety is more negatively related to reported treatment errors
for higher levels of team psychological safety.
Behavioral Integrity, Priority of Safety, Psychological Safety, and Patient Safety
Extending earlier reasoning, we further posit that leader behavioral integrity for safety
will be positively related to reported treatment errors through team priority of safety and team
psychological safety. Simons (2008) proposed that leader behavioral integrity influences
follower outcomes through two complementary mechanisms: it builds trust but it also
clarifies intended direction as the leader sends consistent messages through both words and
actions. Both clear procedures and constructive responses to errors have been shown to
reduce the number of reported treatment errors (Hoffman & Mark, 2006). Similarly, in this
study, we argue that both team psychological safety (a trusting and safe environment) and
team priority of safety (clear directions regarding safety) are important to understand the
effects of leader behavioral integrity for safety on the number of reported treatment errors.
Hypothesis 6. Team priority of safety and team psychological safety mediate the effect of
leader behavioral integrity for safety on reported treatment errors.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 7
We argued that team priority of safety mediates the effects of leader behavioral
integrity for safety on reported treatment errors and that team psychological safety
strengthens the relationship between team priority of safety and reported treatment errors. We
further posit that team psychological safety will also moderate the indirect relationship
between behavioral integrity for safety and reported treatment errors through team priority of
safety. This mediated moderation effect suggests that the indirect effect of leader behavioral
integrity for safety on reported treatment errors is further strengthened by those aspects of
team psychological safety that are not linked to leader consistency of communication (e.g.
leaders actively inviting employee input).
Hypothesis 7. Team psychological safety moderates the effect of leader behavioral integrity
for safety on reported treatment errors through team priority of safety resulting in lower
reported treatment errors when team psychological safety is low, but less so when it is high.
Integrative Model
Figure 1 summarizes the hypothesized relationships of our research model. It suggests
that leader behavioral integrity for safety influences reported treatment errors through two
mechanisms: team priority of safety and team psychological safety. Furthermore, it indicates
that team psychological safety strengthens the relationship between team priority of safety
and reported treatment errors and the indirect relationship between leader behavioral integrity
for safety and reported treatment errors (via team priority of safety). Overall this model
suggests that clarity and consistency of leader communication and behavior regarding safety
(leader behavioral integrity for safety) is important to ensure that employees do not
experience a double bind between adhering to safety procedures and reporting errors against
those safety procedures (Argyris, 1977). Solving that double bind is important, as both team
psychological safety and team priority of safety are important in fostering positive safety
outcomes (Hofmann & Mark, 2006).

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 8
Method
Participants and Procedure
We collected survey data from four Belgian hospitals to test our hypothesized model.
As our research is specified on the team level of analysis, we focused our sampling on
nursing departments within these hospitals. We considered a team to be composed of one
head nurse and a minimum of three nurses who reported directly to this head nurse. We
received permission from hospital nursing directors to conduct our survey with 54 nursing
departments. Paper surveys were distributed to nurses and head nurses within the different
nursing departments, and nurses were asked to deposit them in a sealed box or envelope to
assure anonymity. We collected survey data in the nursing departments at two stages. At
Stage 1, we collected 580 surveys from nurses, which resulted in a response rate of 70%. At
Stage 2, six months later, we used the same procedure to collect survey data from all the head
nurses in the four hospitals on the number of reported treatment errors.
An average of 11 nurses per department responded to our survey, ranging from 3 to 19.
These nurses take care of an average number of 19 patients (SD = 12) at one specific point in
time. The departments have different specializations: surgery, geriatrics, emergency
medicine, pediatrics, maternity, psychiatry, revalidation, dental care, oncology, cardiology,
gastroenterology, orthopedics, radiography, and polyclinics. Head nurses are on average 43
years old (SD = 8), worked in the hospital for 20 years (SD = 9) and have held their position
as a head nurse for 12 years (SD = 9). Fifty-six percent of the head nurses are female and 43
% of the head nurses had worked in the department as a nurse before they served as a head
nurse. Nurses are on average 38 years old (SD = 11), have worked in the hospital for 14 years
(SD = 11) and held their current position for 15 years (SD = 10). On average, a nurse works
in the hospital for 33 hours per week (SD = 8). Seventy-five percent of the nurses are female.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 9
Measures
Behavioral integrity for safety. We constructed a six-item measure based on the
behavioral integrity measure reported in Simons et al. (2007). Leroy, Dierynck, Halbesleben,
Savage, and Simons (2010) validated this instrument showing that behavioral integrity for
safety was correlated with, but factorially distinct from, general behavioral integrity. The
items are ‘Regarding safety, my head nurse delivers the consequences he/she describes.’,
‘When my head nurse lays out safety protocols, he/she makes sure people follow it.’, ‘My
head nurse enforces the safety protocols he/she describes.’, ‘My head nurse always practices
the safety protocols he/she preaches.’, ‘My head nurse does not actually prioritize safety
issues as highly as he/she says he/she does.’ (reversed) and ‘Regarding safety, my head
nurse’s words do not match his/her deeds.’ (reversed).1
Nurses rated these items on a five-point Likert scale ranging from completely disagree
to completely agree. The internal consistency reliability estimate for this scale was 0.93. As
we are interested in team perceptions of the leader, we checked whether we could aggregate
this measure to the team level of analysis. Supporting the aggregation of this measure (Bliese,
2000), we found an average rwg of 0.92 (Mdn = 0.92), an ICC(1) of 0.26 and an ICC(2) of
0.80 and a significant amount of between-group variance F (53, 579) = 4.93, p < 0.01.
Team psychological safety. We measured psychological safety with the seven-item
survey developed by Edmondson (1999). These seven items are measured with a five-point
Likert scale, ranging from completely agree to completely disagree. Example items are: ‘If
you make a mistake in this team, it is often held against you’ (reversed) and ‘Members of this
team are able to bring up problems and tough issues’. The internal consistency reliability
estimate for this scale was 0.80. In support of aggregating this measure to the team level, we
found an average rwg of 0.76 (Mdn = 0.77), ICC(1) = 0.10 and ICC(2) = 0.53 and a significant
amount of between-group variance F (53, 579) = 2.14, p < 0.01.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 10
Team priority of safety. Katz-Navon et al. (2005) validated a seven-item scale
measuring priority of safety that draws on the work of Zohar (2000). This scale measures the
extent to which safety is perceived as a priority within a nursing department in comparison to
other work tasks. Items were measured on a five-point Likert scale ranging from completely
disagree to completely agree. Two examples of items are: ‘It doesn’t matter how the work is
done as long as there are no accidents’ and ‘In order to get the work done, one must ignore
some safety aspects.’ All of the items in this scale are reverse-scored. The internal
consistency reliability estimate for this scale was 0.89. In support of aggregation, we found
an average rwg of 0.74 (Mdn = 0.74), ICC(1) = 0.15 and ICC(2) = 0.65 and a significant
amount of between-group variance F (53, 579) = 2.88, p < 0.01.
Patient safety. We operationalized patient safety as treatment errors in the
performance of an operation, procedure, or test; in the administration of the treatment; in the
dosage or method of using a drug; or as generally inappropriate care that resulted in harm to a
patient (Naveh et al., 2005; Katz-Navon et al., 2005). Common medical errors are adverse
drug events and improper transfusions, surgical injuries and wrong-site surgery, suicides,
restraint-related injuries or death, falls, burns, pressure ulcers, and mistaken patient identities
(Institute of Medicine, 1999). In this study, we collected data on incidents that (1) are broadly
applicable, (2) occur relatively frequently and (3) are subject to underreporting: patient falls,
wound infection, bedsores, and different types of medication errors (medication of wrong
type, at the wrong time, in the wrong amount or applied in the wrong way). A meta-analysis
of Beus, Payne, Bergman, and Arthur (2010) indicates that self-reports and cross-sectional
data can artificially strengthen the relationship between safety climate and safety outcomes.
Accordingly, we asked head nurses to report the treatment errors six months after we
collected the other measures (Mitchell & James, 2001). Over six months head nurses reported
an average of 12 incidents per department (SD = 4).

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 11
Control variables. We controlled for characteristics of the nurses, the nursing teams and
the overall hospital. First, at Stage 1 nurses indicated the average number of patients that
nurses in the team take care of, as reported treatment errors may be a function of nurse
workload (Katz-Navon et al., 2005). Second, at Stage 2 head nurses reported on the average
complexity of patient conditions at the work unit using six items (Hoffman & Mark, 2006) to
control for the fact that reported treatment errors may also be a function of more complex
patient conditions. An example item is ‘How many patients on your unit have complex
problems that are not well understood?’ These items were measured on a five-point Likert
scale ranging from a few (< 20%) to most (> 80%). Third, we controlled for potential
differences between the hospitals (Katz-Navon et al., 2005). Hospital membership of nurses
was derived from the data collection. We found that the addition of these controls did not
meaningfully change our results. Therefore, following the recommendations of Becker
(2005), we omitted these variables from subsequent analyses.
Analyses
We analyzed the data using structural equation modeling in two steps. First, we
conducted a confirmatory factor analysis on our measurement model. The measurement
model showed a good fit to the data (Hu & Bentler, 1998, 1999): χ² (167) = 484.40 (p = 0.00),
SRMR = 0.04, RMSEA = 0.06 and CFI = 0.95. When we alternately constrained each pairwise
factor to unity, we found that, in each case, constraining the factor correlation significantly
worsened the model (p < 0.05). In a next step, we tested the hypothesized, structural
relationships at the team-level of analysis. As multi-level structural equation models are too
parameter intensive for our data (Grizzle, Zablah, Brown, Mowen, and Lee, 2009), we
proceeded using aggregated measures in a path model, correcting for measurement error. We
performed these analyses using the Mplus statistical package (Muthen & Muthen, 2012).

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 12
Results
Main and Interaction Effects
Table 1 depicts the mean, standard deviations and correlation coefficients between our
variables. A path model where behavioral integrity for safety is related to reported treatment
errors through team psychological safety, team priority of safety, and their interaction had a
good fit to the data: χ² (2) = 6.72 (p = 0.03), SRMR = 0.07, RMSEA = 0.02 and CFI = 0.98.
We summarized the relationships between these variables in Figure 1, reporting standardized
beta-coefficients. We found support for Hypothesis 1 that behavioral integrity for safety is
positively related to team priority of safety (β = 0.37, p = 0.01) and for Hypothesis 2 that
behavioral integrity for safety is positively related to team psychological safety (β = 0.34, p =
0.01). In addition, we found support for Hypothesis 3 that team priority of safety is negatively
related to reported treatment errors (β = -0.40, p = 0.01) and Hypothesis 4 that team
psychological safety is positively related to reported treatment errors (β = 0.28, p = 0.02).2
Lastly, we found support for Hypothesis 5 that posited an interaction between team
psychological safety and priority of safety on reported treatment errors (β = -0.35, p = 0.01).
In Figure 2 we specify the nature of the interaction effect. We differentiate between
high and low levels of psychological safety, respectively, one standard deviation above and
one standard deviation below the mean (Aiken & West, 1991). As can be seen in Figure 2,
the relationship between team priority of safety and reported treatment errors is more
negative for higher levels of team psychological safety.
-------------------------------
Insert Table 1 about here
-------------------------------

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 13
----------------------------------------------
Insert Figure 1 and Figure 2 about here
----------------------------------------------
Mediation and Moderated Mediation Effect
Hypothesis 6 asserted that the effect of leader behavioral integrity for safety on
reported treatment errors is mediated through team priority of safety and team psychological
safety. In the path analysis we specified an indirect effect of leader behavioral integrity for
safety on reported treatment errors through team priority of safety (β = -0.13, p < 0.01) and
team psychological safety (β = 0.12, p < 0.01) using a bootstrapping procedure (Preacher &
Hayes, 2008). In addition, we tested an alternative model where leader behavioral integrity
for safety is also directly related to the number of reported treatment errors (James, Mulaik,
& Brett, 2006). This model showed a good fit to the data: χ² (1) = 1.37 (p = 0.24), SRMR =
0.04, RMSEA = 0.03 and CFI = 0.98 but we found no evidence for a direct relationship
between leader behavioral integrity for safety and reported treatment errors (β = -0.08, p =
0.15). This suggests that the model without the direct effect is the more parsimonious and,
thus, the preferred model. Hypothesis 7 asserted that the indirect effect of leader behavioral
integrity for safety on reported treatment errors through team priority of safety is a function
of team psychological safety. This consists of a test of moderated mediation (Muller,
Descartes, Judd, & Yzerbyt, 2005). In support of Hypothesis 7, we found that the conditional
indirect effect was higher at one standard deviation above the mean (ŷ = -4.12, p = 0.03) than
one standard deviation below the mean (ŷ = -3.60, p = 0.03) (Preacher, Rucker, & Hayes,
2007).

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 14
Discussion
The objective of this study was to understand how leaders can reconcile the seemingly
paradoxical elements of enforcing safety protocols while encouraging employee error
reporting. Using theory and research on leader behavioral integrity, we hypothesized that
when leaders live up to espoused safety values, teams will not only attach more importance to
the adherence of safety protocols but also feel psychologically safe to admit mistakes. The
results from a survey study in four hospitals and 54 nursing teams confirm the hypotheses
that head nurse behavioral integrity for safety is related to team perceptions of priority of
safety (Hypothesis 1) as well as psychological safety (Hypothesis 2). These findings imply
that when head nurses are perceived to stay true to the safety values they espouse, they give
clear signals that nurses should place a high priority on safety compared to other demands.
Furthermore, when head nurses’ words and actions align, this signals to nurses that their
concern for safety is genuine, and that it is safe to admit treatment errors. These results
advance previous research in showing that behavioral integrity for safety fosters both team
priority of safety (e.g. Zohar, 2000) and psychological safety (e.g. Edmondson, 1999).
These results are further important as they demonstrate that team priority of safety and
psychological safety have a negative (Hypothesis 3) and positive (Hypothesis 4) relationship
respectively with the number of reported treatment errors. On the one hand, team priority of
safety reflects following safety protocols in daily operations and thus relates to a lower
number of reported treatment errors. On the other hand, team psychological safety reflects an
environment where it is safe to admit mistakes and thus relates to a higher number of reported
treatment errors. These results establish that both aspects are important determinants of an
overall safety climate (Hofmann & Mark, 2006), but they also emphasize the need to
distinguish both aspects from one another as they have a different, seemingly paradoxical,
effect on the number of reported treatment errors (Katz-Navon et al., 2005; 2009).

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 15
To further clarify these opposite effects, we also predicted and found an interaction
effect between team priority of safety and psychological safety (Hypothesis 5). More
specifically, we found that team priority of safety is more negatively related to the number of
reported treatment errors when team psychological safety is high. This suggests that
adherence to safety procedures reflects a genuine concern for safety (and thus has a larger
effect on reported errors) when employees feel safe to speak up about errors. These results
support previous findings that while team psychological safety and priority of safety show
opposing effects on reported treatment errors (Katz-Navon et al., 2009) it is their combination
that has the strongest effect on reported treatment errors (Hofmann & Mark, 2006).
We further demonstrated that the effect of leader behavioral integrity for safety on
reported treatment errors is mediated through the combination of team psychological safety
and priority of safety (Hypothesis 6). In staying true to the safety values they espouse, leaders
foster a safer working environment because they send clear signals that safety should be
prioritized while also fostering an environment supportive of reporting treatment errors.
These results suggest that leader behavioral integrity operates through the complementary
mechanisms of clear communication and fostering a trusting environment (Simons, 2008).
Finally, we found that team psychological safety moderated the indirect relationship between
leader behavioral integrity for safety and reported treatment errors (Hypothesis 7). This
suggest that aspects of psychological safety that are not specific to leader behavioral integrity
for safety may further enhance its positive effects on reported treatment errors.
Overall these results suggest that leaders who follow-up on safety values show a
genuine concern for safety and can expect their team to show a similar genuine commitment
to safety in that they not only adhere to safety protocols but also remain willing to admit
safety mistakes. In this way, behavioral integrity for safety helps resolve the double bind for
team members between adhering to safety protocols and speaking up about mistakes.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 16
Future Research
First, whereas the present study looked at the perceived word-deed alignment of leaders,
additional research is needed to clarify the distinct effects of safety expectations, safety
actions (Zohar, 2000) and their alignment (Simons, 2002). For example, future research
should use surface mapping to indicate how actual alignment of word and deeds influences
perceived alignment. We further expect that safety espousal may influence priority of safety,
but that only their enactment will influence psychological safety.
Second, more research is needed on how specific leader behaviors are related to
employee perceptions that leaders align words and deeds. Future research could extend the
present findings by looking at how authentic leadership fosters behavioral integrity for safety
(Leroy, Simons, & Palanski, in press). Whereas previous research has revealed that
transformational leadership correlates with managerial safety actions (e.g. Barling, Loughlin,
& Kelloway, 2002; Zohar & Luria, 2010), authentic leadership may help us further
understand when leaders enact espoused safety values and invite authentic follower behavior
such as speaking up and admitting mistakes (Gardner, Avolio, & Walumbwa, 2005).
Third, future research should clarify how team psychological safety reduces work-
related errors over the long term. In this study we found that team psychological safety is
positively related to the number of reported treatment errors. Psychological safety may also
lead to a decrease of reported errors over time because an environment supportive of
reporting errors can help employees to learn from mistakes (Edmondson, 1999; Tucker &
Edmondson, 2003; Tucker, Nembhard, & Edmondson, 2007). Future research could use
growth modeling to study these relationships. Building on our findings, we would expect that
whereas psychological safety would be positively related to the average levels of reported
treatment errors in a department at one specific time, psychological safety may lead to a
decrease of reported treatment errors over time.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 17
Limitations
First, we should be careful in drawing causal inferences regarding the direction of the
relationships identified. Some of our data are cross-sectional and thus do not support causal
claims. We have reasonable confidence in our conclusions, however, because a reversed
relationship between leader behavioral integrity for safety and team psychological safety or
priority of safety is unlikely. Leader perceived alignment between words and deeds is
unlikely to be the result of either team priority of safety or team psychological safety.
Nevertheless, additional studies that employ longitudinal designs are needed to determine the
causal direction of the relationships between leader behavioral integrity for safety, team
priority of safety, and team psychological safety.
Next, some of our variables were rated by the same source. This introduces the potential
for common method bias (Podsakoff, Mackenzie, Lee, & Podsakoff, 2003). The risk of such
bias is typically reduced by the adoption of several pro-active strategies (Lindell & Whitney,
2001; Podsakoff et al., 2003). In this study, we included leader-ratings of the number of
reported treatment errors as an outcome and separated these in time from other ratings. We
also used a multi-level research design that confirmed the posited effects at the group level of
analysis, reducing biasing effects that are operative at the individual level of analysis. In
addition, we confirmed the robustness of our findings with a split-sample approach.2
Third, our research is limited to the hospital setting and limited to behavioral integrity
with respect to safety values. This confines the extent to which we can generalize our results
to other industries or other values. Future research could expand on the present findings to
see whether our hypothesized model holds true in other industries or when looking at other
values. For example, future research could further investigate how leader behavioral integrity
addresses the paradox of innovating while also conforming to pre-existing service standards
in a service industry (Palanski & Vogelgesang, 2011).

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 18
Practical Implications
A first implication of our research is that leader behavioral integrity for safety may be
important for creating a climate of safety in the organization. Simons (2008) argued that the
concept of behavioral integrity typically is well received because it has an intuitive appeal
(walking the talk), especially for middle managers who find themselves ‘stuck in the middle’
when it comes to enforcing safety or other procedures proposed by higher management. Our
results suggest that helping these managers maintain their behavioral integrity for safety may
create a safer work environment. In this way leader behavioral integrity not only reflects
leader reliability but promotes a high-reliability organization.
Second, these results suggest that organizations should consider team psychological
safety in addition to safety compliance when evaluating work unit safety. Psychological
safety is important to gain a better understanding as to whether safety is valued out of fear of
punishment or because of a genuine concern for safety. Furthermore, psychological safety is
important to have an accurate assessment of the actual number of errors that occur. In our
study, for example, based on Figure 2 one could conclude that teams low on psychological
safety but high on priority of safety may appear to have the same actual safety performance
as teams high on psychological safety and high of priority of safety. In a team low on
psychological safety however some errors may not be reported, which may create a faulty
perception of actual safety performance (Probst, Bubaker, & Barsotti, 2008).
Conclusion
Our findings suggest that by staying true to the safety values they espouse, leaders can
start to solve the managerial dilemma of providing clear safety directives while encouraging
employees to report errors. This is important as the results of our study indicate that the
combination of both a high priority of safety and a psychologically safe working environment
predicts the number of reported treatment errors in hospitals.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 19
References
Argyris, C. (1977). Double loop learning in organizations: By uncovering their own hidden
theories of action, managers can detect and correct errors. Harvard Business Review, 55,
115-126.
Argyris, C. (1998). Empowerment: the emperor’s new clothes. Harvard Business Review, 76,
98-105.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting
interactions. Newbury Park, CA: Sage.
Barling, J., Loughlin, C. & Kelloway, E. K. (2002). Development and test of a model linking
safety-specific transformational leadership and occupational safety. Journal of Applied
Psychology, 87, 488-496.
Bateson, G., Jackson, D. D., Haley J. & Weakland, J. (1965). Toward a theory of
Schizophrenia, Behavioral Science, 1, 251-264.
Becker, T. E. (2005). Potential problems in the statistical control of variables in
organizational research: A qualitative analysis with recommendations. Organizational
Research Methods, 8, 274-289.
Beus, J. M., Payne, S. C., Bergman, M. E., & Arthur, W. (2010). Safety climate and injuries:
an examination of theoretical and empirical relationships. Journal of Applied
Psychology, 95, 713-27.
Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability:
Implications for Data Aggregation and Analysis. In K. Klein, & S. Kozlowski (Eds.),
Multilevel Theory, Research, and Methods in Organizations. San Francisco, Jossey-Bass.
Dineen, B. R., Lewicki, R. J. & Tomlinson, E. C. (2006). Supervisory guidance and
behavioral integrity: Relationships with employee citizenship and deviant behavior.
Journal of Applied Psychology, 91, 622-635.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 20
Edmondson, A. (1999). Psychological safety and learning behavior in work teams,
Administrative Science Quarterly, 44, 350-383.
Edmondson, A. C. (2003). Speaking up in the operating room: How team leaders promote
learning in interdisciplinary action teams. Journal of Management Studies, 40, 1419-
1452.
Edmondson, A. C. (2004). Learning from failure in health care: Frequent opportunities,
pervasive barriers. Quality and Safety in Health Care, 13, ii3-ii9.
Gardner, W. L., Avolio, B. J., & Walumbwa, F. (2005). Authentic leadership development:
Emergent themes and future directions. In Gardner, W., L., Avolio, B., J., & Walumbwa,
F. (Eds) Authentic leadership theory and practice: Origins, effects and development.
Elsevier; The Netherlands.
Grizzle, J. W., Zablah, A. R., Brown, T. J., Mowen, J. C., & Lee, J. M. (2009). Employee
customer orientation in context: How the environment moderates the influence of
customer orientation on performance outcomes. Journal of Applied Psychology, 94,
1227-1242.
Hofmann, D. A., & Mark, B. (2006). An investigation of the relationship between safety
climate and medication errors as well as other nurse and patient outcomes. Personnel
Psychology, 59, 847-869.
Hu, L. T., & Bentler, P. M. (1998). Fit indices in covariance structure modeling: Sensitivity
to underparameterized model misspecification. Psychological Methods, 3, 424-453.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure
analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6,
1-55.
Institute of Medicine (1999). To err is human: Building a safer health system. Washington
DC: National Academy Press.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 21
James, L. R., Mulaik, S. A., Brett, J. A. (2006). A tale of two methods. Organizational
Research Methods, 9, 233-244.
Katz-Navon, T., Naveh, E., & Stern, Z. (2005). Safety climate in healthcare organizations: A
multidimensional approach. Academy of Management Journal, 48, 1075-1089.
Katz-Navon, T., Naveh E. & Stern, Z. (2009). Active learning: When is more better? The
case of resident physicians' medical errors. Journal of Applied Psychology, 94, 1200-
1209.
Kerr, S. (1975). On the folly of rewarding A, while hoping for B. Academy of Management
Journal, 18, 769-783.
Leroy, H., Dierynck, B., Halbesleben, J., Savage, G., & Simons, T. (2010). Living up to
safety values in healthcare: Effects of behavioral integrity on safety outcomes. Academy
of Management Meeting, Montreal.
Leroy, H., Simons, T., & Palanski, M. (in press). How being true to the self helps leaders
walk the talk: Authentic leadership and behavioral integrity driving follower affective
organizational commitment and work role performance. Journal of Business Ethics,
forthcoming.
Lindell, M., & Whitney, D. J. (2001). Accounting for common method variance in cross-
sectional research designs. Journal of Applied Psychology, 86, 114-121.
Mitchell, T. R., & James, L. R. (2001).Building better theory: Time and specification of when
things happen, Academy of Management Review, 26, 530-547.
Muller, D., Descartes, P., Judd, C. M., & Yzerbyt, V. Y. 2005. When moderation is mediated
and mediation is moderated. Journal of Personality and Social Psychology, 89, 852- 863.
Muthen, L. K., & Muthen, B. O. (2012). Mplus user's guide (6th ed.). Los Angeles, CA:
Muthen & Muthen.
Naveh, E., Katz-Navon, T., & Stern, Z. (2005). Treatment errors in healthcare: A safety

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 22
climate approach. Management Science, 51, 948-960.
Nembhard, I. M. & Edmondson, A. C. (2006). Making it safe: The effects of leader
inclusiveness and professional status on psychological safety and improvement efforts in
health care teams. Journal of Organizational Behavior, 27, 941-966.
Palanski, M. E., & Vogelgesang, G. R. (2011). Virtuous creativity: The effects of leader
behavioral integrity on follower creative thinking and risk taking. Canadian Journal Of
Administrative Sciences, 269, 259-269.
Podsakoff, P. M., Mackenzie, S. B., Lee, J., & Podsakoff, N. P. (2003). Common method
biases in behavioural research: A critical review of the literature and recommended
remedies. Journal of Applied Psychology, 88, 879-903.
Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation
hypotheses: Theory, methods, and prescriptions, Multivariate Behavioral research, 42,
185-227.
Preacher, K. J. & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing
and comparing indirect effects in multiple mediator models. Behavior Research Methods,
40, 879-891.
Probst, T. M., Brubaker, T. L., & Barsotti, A. (2008). Organizational injury rate
underreporting: The moderating effect of organizational safety climate. Journal of
Applied Psychology, 93, 1147-1154.
Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different at all: A
cross-discipline of view of trust. Academy of Management Review, 23, 393-404.
Simons, T. (2002). Behavioral integrity: The perceived alignment between managers' words
and deeds as a research focus. Organization Science, 13, 18-35.
Simons, T. (2008). The integrity dividend: Leading by the power of your word. San
Francisco, Jossey Bass.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 23
Simons, T., Friedman, R., Anne Liu, L., & McLean Parks, J. (2007). Racial differences in
sensitivity to behavioral integrity: Attitudinal consequences, in-group effects, and
“trickle down” among black and non-black employees. Journal of Applied Psychology,
92, 650-666.
Tucker, A. L., & Edmondson A. C. (2003). Why hospitals don't learn from failures:
Organizational and psychological dynamics that inhibit dynamics. California
Management Review, 45, 55-71.
Tucker, A. L., Nembhard, I. M., & Edmondson, A. C. (2007). Implementing new practices:
An empirical study of organizational learning in hospital intensive care units.
Management Science, 53, 894-907.
Zohar, D. (2000). A group-level model of safety climate: Testing the effect of group climate
on microtreatment errors in manufacturing jobs. Journal of Applied Psychology, 85, 587-
596.
Zohar, D. (2002) Modifying supervisory practices to improve subunit safety: A leadership-
based intervention model. Journal of Applied Psychology, 87, 156-163.
Zohar, D. (2010). Thirty years of safety climate research: Reflections and future directions.
Accident Analysis and Prevention, 42, 1517-1522.
Zohar, D., & Luria, G. (2004). Climate as a social-cognitive construction of supervisory
safety practices: Scripts as proxy of behavior patterns. Journal of Applied Psychology,
89, 322-333.
Zohar, D., & Luria, G. (2010). Group leaders as gatekeepers: Testing safety climate
variations across levels. Applied Psychology: An International Review, 59, 647-673.
Zohar, D., & Tenne-Gazit, O. (2008). Transformational Leadership and Group Interaction as
Climate Antecedents: A Social Network Analysis. Journal of Applied Psychology, 93,
744-757.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 24
Footnotes
1 This study also included measures of safety climate (Zohar, 2000) and authentic
leadership (Walumbwa, Avolio, Gardner, Wernsing, & Peterson, 2008). A measurement
model where authentic leadership, safety climate, and leader behavioral integrity for safety
are separate constructs has a reasonable fit to the data (χ2 (87) = 267.51 (p = 0.00), SRMR =
0.04, RMSEA = 0.08, CFI = 0.96) and fit the data better (p < 0.05) than a model where these
measures were set to correlate at 1.0. Furthermore, we found that behavioral integrity for
safety predicted additional explanatory variance in priority of safety when controlling for
authentic leadership (β = 0.26; p < 0.01) and psychological safety when controlling for safety
climate (β = 0.34; p < 0.01). This is consistent with the idea that leader behavioral integrity
for safety adds information to these other measures because it combines the beneficial effects
of a psychologically safe environment and an environment that prioritizes safety regulations.
2 To check for common method bias in these cross-sectional relationships, we used
data from half of each team for leader behavioral integrity for safety on the one hand and
team priority of safety and team psychological safety on the other hand to replicate these
findings. While somewhat smaller, the direction and significance of the relationships confirm
these findings.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 25
Table 1
Means, Standard Deviations and Intercorrelations Among Study Variables.
M
SD
1
2
3
4
1.
Leader behavioral integrity for safety
3.80
.40
.93
2.
Team psychological safety
3.58
.30
.45**
.80
3.
Team priority of safety
3.66
.45
.38**
.24*
.89
4.
Reported treatment errors
11.68
1.37
-.26*
.18
-0.34*
/
Note. Reliability estimates for scales are presented on the diagonal.
* p < 0.10
** p < 0.05

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 26
Leader behavioral integrity for
safety
Team priority of safety
Team psychological safety
Reported treatment errors
0.37**
0.34**
-0.40**
0.28**
- 0.35**
** p < 0.05
Figure 1. Hypothesized model.

LEADER BEHAVIORAL INTEGRITY FOR SAFETY 27
Reported number of
treatment errors
10
11
12
13
14
15
Low priority of safetyHigh priority of safety
High Psychological Safety
Low Psychological Safety
Figure 2. Interaction effect between team priority of safety and psychological safety on the
reported number of treatment errors.