You are on page 1of 13

Available online at www.sciencedirect.

com

Safety Science 46 (2008) 7991


www.elsevier.com/locate/ssci

The development of a considerate and responsible


safety attitude in work teams
Christopher D.B. Burt *, Bridgit Sepie, Gretchen McFadden
Department of Psychology, University of Canterbury, Private Bag 4800, Christchurch, New Zealand

Received 22 March 2006; received in revised form 15 October 2006; accepted 23 October 2006

Abstract

Safety research has identied the potential importance of having considerate and responsible employees who care about
the safety of their co-workers [e.g. Burt, C.D.B., Gladstone, K.L, Grieve, K.R., 1998. Development of the considerate and
responsible employee (CARE) scale. Work and Stress 12 (4), 362369]. This study had the general aim of identifying vari-
ables that are related to employees having a caring attitude towards their co-workers safety. Variables examined as pos-
sibly related to caring were co-worker knowledge, opener ability, team tenure, co-worker and supervisor support, group
orientation and group cohesion. The relationship between caring and the personal support dimension of contextual perfor-
mance was also examined. Study 1 sampled workers in the forestry and construction industries, and found that caring was
positively correlated to group cohesion and to the amount of knowledge an employee had about their co-workers. Study 2
sampled workers in a road construction and a power generation company, and found that caring was positively correlated
with team tenure and the personal support dimension of contextual performance. Implications of these ndings are dis-
cussed in terms of the recruitment of team members, and the management of the caring aspect of team safety climate.
2006 Elsevier Ltd. All rights reserved.

Keywords: Caring; Safety climate

1. Introduction

Safety in the workplace can be managed from a number of perspectives, including ergonomic research,
which oers good advice about the design of safe systems, by adopting management systems to improve
safety (e.g., Cox and Cox, 1996), and considering individuals disposition towards safety and risk taking at
recruitment. In contrast, relatively little is known about how individual employees inuence the safety of their
colleagues. That is, by what mechanisms does employee A help to ensure the safety of employee B? Another
way of framing this question is to ask what leads employee A to care about employee Bs safety? Answering
this question could potentially lead to interventions which could result in improved workplace safety,

*
Corresponding author. Tel.: +64 3 3642231; fax +64 3 3642181.
E-mail address: Christopher.burt@canterbury.ac.nz (C.D.B. Burt).

0925-7535/$ - see front matter 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ssci.2006.10.005
80 C.D.B. Burt et al. / Safety Science 46 (2008) 7991

especially where employees work in teams. This paper reports two studies which attempted to identify vari-
ables which are related to the development of caring attitudes.
Variation in employees caring about their co-workers safety is potentially one dimension of a work teams
safety climate. Since the 1980s numerous studies have examined safety climate (e.g., Brown and Holmes, 1986;
Dedobbeleer and Beland, 1991; Hofmann and Stetzer, 1996; Mearns et al., 1998, 2003; Cheyne et al., 1998,
2002; Zohar, 1980, 2000). Consequently, there is now a growing body of evidence which suggests that safety
climate inuences safety practices (Zohar, 1980), unsafe behaviour (Hofmann and Stetzer, 1996), accidents
(Mearns et al., 1998, 2003; Zohar, 2000), and is a useful predictive indicator of safety performance (Flin
et al., 2000). While it seems appropriate to conclude that safe work behaviour is facilitated by a positive safety
climate (Clarke, 2000; Cooper and Phillips, 2004), little is known about the existence or operation of caring
within a safety climate.
Geller and colleagues (e.g., Roberts and Geller, 1995; Geller et al., 1996; Geller, 2001) were perhaps the
rst to formally recognise the importance, and potential use, of promoting safety through co-worker rela-
tions. They coined the term actively caring which refers to employees caring enough about the safety
of others to act accordingly (Geller et al., 1996). That is, actively caring requires employees to go beyond
the call of duty to identify environmental hazards and unsafe work practices and then implement appropri-
ate corrective actions when unsafe conditions or behaviours are observed (Roberts and Geller, 1995, p. 54).
Actively caring also relates to a phenomenon that Hofmann and Stetzer (1996) called approach intentions,
dened as the tendency for team members to approach one another regarding safety related activities.
Geller notes that actively caring might overcome (or perhaps supplement) the need for management to
constantly monitor safety related behaviour, by employees taking this role/responsibility to help ensure their
co-workers safety.
Burt et al. (1998) developed the considerate and responsible employee (CARE) scale which is a reliable
measure of employees attitudes toward their co-workers safety a measure of caring. The care scale measures
workers attitudes (strongly disagree to strongly agree) to 21 statements relating to behaviours, which if per-
formed would increase co-workers safety. The scale items (which all load on a single factor) relate to identi-
fying, removing and discussing hazards, to reporting and discussing accidents, to understanding and
communicating safety information, to helping and warning co-workers, and to recognising co-workers limits.
While other attitudinal safety measures include items regarding co-workers attitudes to safety (e.g., Donald
and Canter, 1994), these tap into an employees perception of their co-workers safety attitudes rather than
how workers feel about actually engaging in specic behaviours aimed at enhancing their co-workers safety.
Although other attitudinal measures include a question or two regarding co-workers safety, these items tend
to focus on the single component of safety communication (e.g., (Cox and Coxs, 1991) scale includes the item
Individuals should encourage colleagues to work safety).
We argue that a strong caring work environment would result in an improvement in safety. While equip-
ment design considerations, management driven training and policies, and careful recruitment of competent
workers are appropriate areas to target in relation to safety, it is still the actual workers that are faced with
the workplace risks and hazards. If the dominant attitude held by each co-worker is to care about each other,
then there should be a reduction in work-place hazards with each worker actively identifying, removing or
warning others of these hazards (not because it is company policy to do so but because they know that
the next person to encounter the hazard is likely to be one of their co-workers). There should be much more
communication about past accidents (and near miss situations), and how these can be avoided in the future.
There should also be an explicit recognition that each co-worker has limits, and a belief that providing
assistance to a co-worker, when needed, is an appropriate action within the team. The key issue for this
research was to determine the factors which are related to an employee having a caring attitude towards their
co-workers that determine care scale scores.
Obviously caring is most relevant in work settings which involve teams where each individuals behaviour
can potentially inuence the safety of other team members. Given this, we suggest there are a number of fac-
tors that drive the development of a caring attitude. Fig. 1 shows these relationships. Our rst suggestion is
that as co-workers gain more knowledge of team members, the social fabric of the team will develop (e.g.,
friendships, cohesion, support, and social relationships), and caring will develop. That is, simply getting to know
the team members that an employee is placed with may help develop that employees caring attitudes. Our
C.D.B. Burt et al. / Safety Science 46 (2008) 7991 81

Personality:
Opener Ability

Work values:
Group
Orientation Acquisition of Development of: Development of:
Knowledge about Friendships Considerate and
Co-workers Cohesion Responsible Safety
Support Attitudes - CARING
Social Relationships
Socialisation into
Team

Team Safety Climate Outcomes:


Team Tenure Increased Safety Communication
Contextual
Performance: Hazard Identification & Removal
Personal Support Employees Assisting Each Other

Less Accidents & Near Misses

Fig. 1. A model of the development of a caring attitude.

second suggestion is that as the social relationship aspects of a team develop (e.g., friendships, cohesion, and
support develops within the team) they operate in a reciprocal fashion, where stronger social relationships
lead to increased caring and further knowledge acquisition, and both of these increase the strength of social
relationships, which in turn leads to more caring and so on. Furthermore, we suggest that caring may be par-
tially determined by an individuals personality (opener ability) and work values (group orientation) through
their inuence on the employees acquisition of knowledge about their co-workers. Finally, we suggest that
socialisation into a team (perhaps via formal induction training), and time (team tenure) should provide
opportunities for co-workers to get to know each other. We explore each of these suggestions in some detail
below.
As an employee gets to know his/her co-workers the greater will be the likelihood that he/she will develop a
caring attitude towards them. In a new employment situation, or in the formation of a new project team, the
individuals involved will often have little or no knowledge of others in the team; in essence they are strangers.
Social psychology research has demonstrated that individuals tend not to help strangers (e.g., the bystander
apathy eect see Garcis et al. (2002)), but are likely to help when the other is a friend (e.g., Bell et al.,
1995). The formation of a work team probably places individuals somewhere between complete strangers
and friends, in that the simple labelling of a group of individuals as belonging to a team begins the process
of bonding the group together. As the closeness of employee A and Bs relationship increases, which we argue
can be measured by examining how much they know about each other and their level of social interaction (in
this research we measured social interaction by frequency of conversations), the concern that employee A and
B have for each others safety may increase. Put another way, we suggest that to begin caring about a co-
worker, an employee has to get to know them.
The literature on organisational socialisation is pertinent here. Comer (1991) found that new employees
who form peer relationships acquire information faster and this facilitates socialisation into the organisation
(see also Kramer, 1996; Morrison, 1993). Arguably, information acquisition during socialisation (induction
training) is not just about the organisation but also about the personal life of co-workers: their families, atti-
tudes, and interests etc. More importantly, if the development of co-worker knowledge and the level of social
relationships within a team are related to the development of positive safety related attitudes (caring), it might
be possible to design interventions (particularly associated with the socialisation process into the team/orga-
nisation) which might speed the process up. This might be particularly important in occupations, such as for-
est harvesting, where individual team members are restricted by operational techniques and safety equipment
(hearing protection) as to how much they can communicate with each other on the job. These interventions
would be aimed at developing the social fabric of a work team faster than might be the case if acquisition of
knowledge about ones co-workers was simply allowed to develop in a haphazard manner. In Study 1 we
82 C.D.B. Burt et al. / Safety Science 46 (2008) 7991

explored the relationships between caring and how much employees know about each other (measured by
7 knowledge questions), how frequently employees talk about work and non-work issues, how they rate their
relationships (i.e. simply a work colleague through to best friend) and how long they have known their co-
workers (team tenure). The main hypothesis that was tested is that the more an employee knows about their
co-workers, and the closer their social relationship, the more they may care for their co-workers safety.
So far we have suggested that the development of caring for ones team members begins with the acquisition
of knowledge about who the employee is working with. This, we would argue, is paralleled by the development
of an increasingly complex social relationship structure within the work team. While not directly considering
caring attitudes, research has highlighted the importance of the relationships in a work setting for safety out-
comes. Simard and Marchand (1997) found that social relationships at the shop oor level were the best pre-
dictors of compliance with safety regulations. In particular, social support appears to be a key issue, with
Goldberg et al. (1991) nding that, when co-worker support was present, workers were more likely to partic-
ipate in plant safety programs. Similarly, Iverson and Erwin (1997) found that supervisory and co-worker sup-
port predicted occupational injury, while Oliver et al. (2002) found that both supervisor and co-worker
support played a key role in the prevention of accidents. Furthermore, Burt et al. (1998) found that caring
was positively related to group cohesion. In light of these ndings it is possible that in a team setting, social
support may impact on occupational safety via the development of caring attitudes. That is, more co-worker,
and perhaps to a lesser extent supervisor, support may be related to more caring for co-workers safety. Study
2 attempted to provide some insight into this area by examining the relationship between levels of co-worker
and supervisor support and caring.
We suggested that an employees personality and values might predispose them to develop a caring atti-
tude towards their co-workers though their inuence on the employees development of knowledge about
team members. Given the suggestion that knowledge of co-workers may be important in the development
of caring, Opener ability (which is a personality variable that might predispose an employee to engage with
his/her team) was examined in Study 2. Opener ability refers to the tendency to open up or illicit disclo-
sure from others (Miller et al., 1983). Miller et al. (1983) found that in comparison to low openers, high
openers consistently gain more personal information about people regardless of the targets regular self-
disclosure level. Because of this, it is possible that in the work setting, those who rate highly on opener
ability will gain more personal information about their co-workers which in turn could facilitate caring.
Thus, it was hypothesised that opener ability will be positively correlated to co-worker knowledge and to
care scale scores.
We also included a measure ofGroup orientation in Study 2. Group orientation refers to a category of work
values that centre around social relations with peers, supervisors, and others and is concerned with how much
one values socialising at work (De Vos et al., 2003; De Vos et al., 2005). De Vos et al. (2005) suggest that peo-
ple high in group orientation are more likely to invest in creating a social network at work, look for social
values within their work situation, and help or have concern for others at work. Thus, we predicted that group
orientation would be positively correlated with knowledge acquisition and caring.
Finally, Clarke (2003) noted that caring may be a safety-specic type of contextual performance dimension.
Contextual performance is dened as extra role discretionary behaviors that in general are not explicitly
related to a companys formal reward system, but that contribute to organizational outcomes through their
eect on the psychological, social, and organizational context (Motowidlo, 2003; Van Scotter et al., 2000).
Borman et al. (2001) reported that contextual performance has three dimensions personal support, organiza-
tional support, and conscientiousness initiative with personal support characterized by behaviors such as help-
ing others, cooperating, informing others, courtesy etc. Essentially, employees that score highly on contextual
performance go beyond what is formally required by the organization. They engage in behaviors which you
might not expect to nd in their job description. Morgeson et al. (2005) have noted the importance of contex-
tual performance for team performance, and looked at variables which predict contextual performance in a
recruitment context. If Clarkes (2003) suggestion is correct, and caring is a type of contextual performance,
then variables that predict contextual performance may also predict caring. To investigate Clarkes suggestion
we included a measure of contextual performances personal support dimension in Study 2 (and show a link
between caring and contextual performance in Fig. 1), and predicted a positive correlation between this mea-
sure and care scale scores.
C.D.B. Burt et al. / Safety Science 46 (2008) 7991 83

2. Study 1

2.1. Method

2.1.1. Participants
Ninety-six employees working in teams from eight organisations in the forestry and construction industries
were invited to participate in the study, 80 employees returned completed questionnaires (less than 1% missing
data) giving a response rate of 83 percent. Seventy-seven participants were male (mean age 36.6 years) and 2
were female (mean age 29 years), one participant did not report their age and gender. The overall mean tenure
with their organisation was 74.7 months (SD = 92.1), and with their current team was 32.8 months
(SD = 51.57). The forestry and construction industries were chosen because of their high accident rates and
operational use of teams (Feyer et al., 2001).

2.1.2. Materials
The study questionnaire contained a demographic section and 3 other measurement scales which are
described below. The demographic section contained questions on the participants age, gender, their tenure
with the organisation, and their tenure with the particular team they were currently working within. While the
content of each questionnaire was identical, the scales were combined in ve dierent random orders to min-
imise the potential eect of common method variance (Crampton and Wagner, 1994). Instructions for com-
pleting the questionnaire and an informed consent statement were printed on the cover page.

2.1.3. Care scale


The care scale is a 21 item measure where responses to statements such as Safety comes from worker co-
operation and Co-workers should discuss changes that could improve safety are made on a 5-point rating
scale where 1 = strongly disagree and 5 = strongly agree (the care scale, including scoring instructions is pub-
lished in Burt et al., 1998). After reverse scoring items, ratings are summed to give a nal score where a larger
score indicates more co-worker consideration and responsibility towards co-worker safety. Possible scores
range from 21 to 105. Burt et al. (1998) reported a coecient alpha of .91, and a test re-test reliability of
.62 for the care scale. A coecient alpha of .81 was obtained in the present study.

2.1.4. Co-worker knowledge/socialising


A self-report measure of the level of knowledge an employee had of other members of their team, and the
frequency of engaging in conversation about various topics, and the nature of their relationship, was devel-
oped. Participants were asked to write down the initials of ve of their team members (the collection of data
on ve team members was aimed at getting a representative view of participants perception of their knowl-
edge of, and social relationships with, their team while not being overly taxing in terms of time to complete
the questionnaire), and answer 7 knowledge questions for each of these co-workers. These 7 questions took
the form do you know . . .. . .. about X, in relation to the nominated co-workers birthday, marital status,
place of last holiday, last job, interests and hobbies, type of car they drive, and number of children. These ques-
tions required a yes or no answer (with a score of 1 for a yes and zero for a no answer). The responses to the 7
knowledge questions were summed across the 5 team members thus giving a possible knowledge score range
from 0 to 35. The next 5 questions required ratings on a 7-point scale (1 = never, 2 = once every few months,
3 = once a month, 4 = once a week, 5 = several times a week, 6 = around once a day, and 7 = more than
once a day) as to how frequently the participant discussed politics, sport, family issues, any other non-work
related issues, and work related matters with each of the nominated 5 team members. Responses to the four
non-work related conversation questions were summed across the 5 team members giving a conversation score
range of 20140. Ratings for the work related discussion question were also summed across the 5 team mem-
bers to give a possible work discussion score range of 535. Finally, participants were also asked to rate their
relationship with each of the 5 team members using a 10 point scale where 1 = just a team member and
10 = team member and close friend. These ratings were summed to form a relationship score with a possible
range of 550.
84 C.D.B. Burt et al. / Safety Science 46 (2008) 7991

Table 1
Means, standard deviations and Pearson product moment correlations
Variable Mean SD Care Knowledge Conversation Work discussion Relationship Group
score score score score cohesion
Care score 90.5 8.3
Knowledge 24.3 5.7 .34**
Conversation 71.2 22.3 .19 .44**
Work discussion 26.8 8.4 .08 .21* .36**
Relationship 23.2 11.4 .06 .46** .29* .20
Group cohesion 90.5 8.3 .24* .31** .23* .17 .34**
Team tenure 32.4 51.5 .18 .22* .01 .12 .17 .19
*
p < .05, two-tailed.
**
p < .01, two-tailed.

2.1.5. Group cohesion


Group cohesion was measured using the group attitude scale (Evans and Jarvis, 1986), which measures
group members attraction to a group, and their desire to identify with and be an accepted member of the
group. The group attitude scale consists of 20 items which participants are required to answer on a 9-point
scale where 1 = agree and 9 = disagree. After reverse coding, ratings are summed across the 20 items to give
a total score (potential range from 20 to 180) where a greater score indicates higher group cohesion. The group
attitude scale was reported to have internal consistency ranging from a = .90 to .97 across three studies (Evans
and Jarvis, 1986), and a value of a = .86 was found in the present study.

2.1.6. Procedure
The study questionnaire was distributed to the forestry participants by their team supervisor in their work-
place (including a sealable envelope) and the questionnaires were returned to the researchers by the organisa-
tion. Questionnaires (including a sealable envelope) were distributed to the construction workers by the second
author during paid work time, and collected immediately after completion.

2.2. Results

Means, standard deviations and Pearson product moment correlations were calculated for the study vari-
ables and these are shown in Table 1. Inspection of Table 1 indicates that as predicted care scale scores were
positively correlated with co-worker knowledge, as well as with group cohesion. A number of other variables
correlated with co-worker knowledge, and these relationships appear readily interpretable and consistent with
our model (Fig. 1). In particular, co-worker knowledge has signicant positive correlates with conversation
scores, work discussion scores, team tenure, relationship scores, and group cohesion.
We regressed conversation scores, work discussion scores, relationship scores, team tenure and group cohe-
sion onto knowledge scores. The overall model was signicant F(5,74) = 7.859, p < .01 and accounted for
30.2% of the variance in knowledge scores (adjusted R2 = .272), with the conversation score (b = .31,
p < .01) and relationship score (b = .29, p < .01) contributing signicantly to the model. The remaining betas
were team tenure .14, work discussion .10, and group cohesion .00.

3. Study 2

Study 2 had four aims. First to replicate the nding from Study 1 that knowledge of co-workers is related
to caring. Secondly, to determine if opener ability and group orientation are related to co-worker knowl-
edge. Thirdly, to examine the prediction that co-worker and supervisor support would be related to caring.
Finally, to examine the relationship between the personal support dimension of contextual performance and
caring.
C.D.B. Burt et al. / Safety Science 46 (2008) 7991 85

3.1. Method

3.1.1. Participants
A total of 192 questionnaires were distributed to employees from a road construction company and a
power company, with a total of 104 useable (less than 1% missing data) questionnaires returned, giving a
response rate of 54.1%. The response rate is lower than that obtained for Study 1, and seems to be due to
slightly more questionnaires being returned with a lot of missing data (making them unusable), and to the
use of postal returns with the power company which seems to have reduced compliance. There were 95 male
participants (mean age 37.5 years), and 9 female participants (mean age 32.8 years). On average the partici-
pants had worked for their organisation for 63.8 month (SD 73.6), and with their current team for 49.7
months (SD 63.1).

3.1.2. Materials
The study questionnaire contained a cover page with instructions and an informed consent statement, six
measurement scales, and a demographic information section. The demographic section included questions
relating to participants age, gender, tenure with their current team and tenure with the organisation. To help
control for common method variance, several dierent orders of the measurement scales were used.

3.1.3. Care scale


The care scale, as described in Study 1, was used to measure workers attitudes toward their co-workers
safety. The coecient alpha calculated for the present study was .85.

3.1.4. Co-worker knowledge


Co-worker knowledge was measured for 5 co-workers using the same 7 question format used in Study 1.
Thus, co-worker knowledge varied on a possible score range from zero to 35, with a higher score indicating a
higher level of co-worker knowledge.

3.1.5. Opener ability


Opener ability was measured using Miller et al.s (1983) Opener Scale. This scale consists of ten items (e.g.,
I have been told Im a good listener, and I encourage people to tell me about themselves), each accom-
panied by a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree). The scales possible score range
is 1050, with a high score indicating the respondent is a high opener. Miller et al. (1983) reported the scale
had a reliability of a = .79, and sound test-retest reliability. A coecient alpha of a = .75 was calculated for
this study.

3.1.6. Supervisor and co-worker support


Supervisor and Co-worker support were measured using the appropriate four items from Caplan et al.s
(1975) social support scale. The measure requires two ratings on a 5-point scale (1 = Not at all to 5 = Very
much) for each of the four items, one rating for your immediate supervisor and one for co-workers (e.g.,
How much do each of these people go out of their way to do things to make your work life easier for you?).
Each support scale has a potential score range of 420 with a higher score indicating higher perceived levels of
support from the person in question. Fields (2002) reported a coecient alpha for the co-worker subscale of
.79, while the supervisor subscale reliability is reported to range from .86 to .91 (Fields, 2002). The present
study calculated comparable internal reliability statistics of a = .73 and a = .87 for the co-worker and super-
visor subscales, respectively.

3.1.7. Group orientation


The six-item group orientation scale as developed by De Vos et al. (2003); De Vos et al. 2005 was used in
this study. Respondents were asked to indicate on a 5-point scale the importance (1 = not at all important to
5 = extremely important) that they attached to each item (e.g., To have social contact with other people at
86 C.D.B. Burt et al. / Safety Science 46 (2008) 7991

work and To have people around you who have time for a chat). The potential score range of this scale is
630, with a high score indicating a high value given to group orientation at work. The group orientation
scales coecient alpha calculated for the present study was a = .76.

3.1.8. Contextual performance


Contextual performance was measured using Van Scotter et al.s (2000) interpersonal facilitation measure
which is a seven-item measure of contextual performances personal support dimension. It measures the help-
ful, considerate, and co-operative aspects of contextual performance with questions such as When at work
how likely are you to help someone without being asked?. Coecient alpha for the personal support measure
has been reported to be .89 (Van Scotter et al., 2000). The reliability statistic calculated for the present study
was a = .81.

3.1.9. Procedure
Participants from the Roading company completed the study questionnaire during work time at either the
beginning of a monthly departmental meeting, or at a worksite after a team meeting. On completion, one of
the researchers collected the questionnaires. Participants from the Power company were from several dierent
locations throughout New Zealand. A representative from the human resources department in each location
distributed the study questionnaire to participants, along with a postage paid envelope addressed to the
researchers.

3.2. Results

A total of 13 missing data points were evident across the 104 cases for the care scale, contextual perfor-
mance measure, opener ability scale and support scales, and item mean imputation was used to correct this.
Means, standard deviations and Pearson product moment correlations were calculated for the study variables
and these are shown in Table 2. The rst issue we examined was whether caring might be a safety specic con-
textual performance dimension. Inspection of Table 2 indicates that a signicant positive correlation was
obtained between care scale scores and the personal support dimension of contextual performance. This result
suggests that caring might well be a safety-specic contextual performance dimension. Further support for this
was found in an exploratory principal component factor analysis using oblimin Kaiser normalization rotation
(KaiserMeyerOlkin measure of sampling adequacy = .811, Bartletts test of sphericity = 1961.1), which pro-
duced a clear two factor solution with the care scale items and contextual performance personal support items
loading on separate factors.

Table 2
Means, standard deviations and Pearson product moment correlations
Variable Mean SD Care Knowledge Group Opener Worker Supervisor Interpersonal
score score orientation ability support support facilitation
Care score 92.9 7.6
Knowledge 24.3 6.1 .13
Group orientation 23.8 3.7 .01 .08
Opener ability 36.0 5.0 .19 .20* .32**
Co-worker support 16.5 2.5 .15 .03 .18 .26*
Supervisor support 15.2 4.2 .13 .03 .09 .11 .30**
Contextual performance: personal 29.3 3.6 .49** .24** .24** .47** .31** .18
support
Team tenure 49.7 63.1 .25** .08 .05 .01 .16 .01 .22*
*
p < .05, two-tailed.
**
p < .01, two-tailed.
C.D.B. Burt et al. / Safety Science 46 (2008) 7991 87

3.2.1. Co-worker knowledge and caring


Inspection of Table 2 indicates that, in contrast to Study 1, care scores were not signicantly correlated with
the overall knowledge scores (although the correlation was positive). To explore this relationship further we
correlated the scores for each of the 7 knowledge questions with care scores: birthday r = .04; martial status
r = .08; last holiday r = .22, p < .05; last job r = .18; hobbies r = .03; car r = .02; children r = .16. While
these results are mixed, there is some indication that caring and knowledge of co-workers are related. Table
2 also shows a signicant correlation between team tenure and caring, which is consistent with the idea that
having time to get to know co-workers is important for the development of caring. Furthermore, co-worker
and supervisor support, and opener ability were positively correlated (although not signicantly) with caring.
The correlations shown in Table 2 indicate that as predicted knowledge scores were signicantly correlated
with opener ability. There was also a signicant correlation between knowledge scores and personal support
which is consistent with the idea that the personal support dimension of contextual performance is character-
ized by cooperating with and informing others. However, contrary to our suggestion, group orientation
showed no relationship with knowledge scores, although it was signicantly correlated with opener ability.

4. General discussion

The relationships we studied are likely to be primarily reciprocal, rather than uni-directional causal rela-
tionships. Knowledge of co-workers is suggested to be a prerequisite for the development of caring, and if car-
ing does develop (and we discuss below why it may not), this is likely to lead to further knowledge acquisition
and so on. In a similar way the relationship and support aspects of a team are, we suggest, likely to feed into
each other: as the social fabric of a team strengthens, caring is likely to develop, and as co-workers perceive
(see) each other caring about their safety, the social fabric is likely to further strengthen. Opener ability and
group orientation are suggested to be facilitators of the above processes, leading an individual to seek more
knowledge about co-workers (or at least seek this knowledge faster), and to engage with their team, thus
enhancing perceived support and the teams social fabric.
The hypothesis that gaining knowledge of ones co-workers may be a prerequisite for caring about them
arises from the very fundamental ndings in social psychology that we tend to avoid helping strangers, but
do help friends. Study 1 generally supported this hypothesis, but the results from Study 2 were not as convinc-
ing. The mean knowledge scores from the two studies were in fact identical, and the mean care scores across
the two studies were also similar (see Tables 1 and 2). So why was the knowledge/caring link weaker in Study
2? There are dierent types of knowledge a person may acquire about their co-workers. Such knowledge could
be perceived as positive, negative, or neutral. For example, a person may nd out positive information about a
co-workers family life such as the number of children they have, and/or they could nd out negative infor-
mation such as, the co-worker abuses their spouse. Our research focused on positive co-worker knowledge.
It seems intuitive that having positive knowledge of co-workers would increase concern for their safety. How-
ever, for this to happen there might also have to be an absence of negative knowledge. Put another way, an
employee might gain knowledge of their co-workers and in some cases this might be associated with caring,
but in other cases where the acquired knowledge leads the individual to decide they do not really like the indi-
vidual (or agree with some of their attitudes and values) no caring relationship might develop. We would thus
still argue that one of the fundamental steps to the development of caring within a work team is the initial
process of getting to know ones co-workers, but it might be that on occasions the nature of the acquired
knowledge actually precludes the development of caring. In support of this suggestion Burt et al. (1998) found
that caring was positively related to co-worker satisfaction. Hence, when examining the relationship between
co-worker knowledge and caring, a measure of co-worker liking/satisfaction may need to be included. At a
practical level, members of a work team that get to know each other may develop caring attitudes, but perhaps
only if they actually like each other.
Another variable which may inuence the relationship between co-worker knowledge and caring, both the
initial formation of a caring attitude, and its ongoing maintenance/development, is trust. An understanding of
the importance of trust to work-place safety seems to be only just immerging in the literature (see Flin and
Burns, 2004; Burns et al., in press). A degree of trust in managements commitment to safety might be required
for caring to begin, and any management activities which disrupt this trust relationships may well interrupt the
88 C.D.B. Burt et al. / Safety Science 46 (2008) 7991

maintenance and further development of a caring attitude. For example, one form of caring which we noted in
the introduction is to give a co-worker assistance with a task if it appears the task is dicult for them (and thus
potentially risky). If employees do not trust managements commitment to safety, and feel there is a possibility
they could be reprimanded for leaving their work to assist a co-worker, then the caring attitude-behaviour link
might be broken. A similar example can be developed for hazard reporting worker A sees a hazard which
was created (perhaps unknowingly) by worker B, and which worker C is likely to encounter during the day.
Worker As reaction (if they have a caring attitude) is to alert management and co-workers to the hazard and
ensure its removal. However, if they feel that management will in some way reprimand worker B for initially
creating the hazard, then this lack of trust in managements commitment to safety may disrupt the caring atti-
tude-behaviour linkage. Future research might well explore the relationship between trust (both within team
and towards management) and caring.
Finding that team tenure was positively related to caring (a signicant relationship in Study 2 and a positive
correlation in Study 1) suggests that employee turnover is not only an inconvenience with regards to nding
and training a new employee but also has a safety cost. When people leave a team (the organisation) they take
with them the caring attitudes that may have developed while being with the team, they also probably disrupt
the social fabric of the team. The loss of a single team member and his/her replacement could expose the
remaining team members to risk as initially the remaining team members will not know if they can rely
on (trust) the new team member to care about their safety (nor will they know whether they should care about
the safety of the new employee). As we argue that the induction process can be an opportunity for acquisition
of co-worker knowledge (and the development of caring), the risks associated with bringing new members into
a team may be able to be partially mediated by ensuring that the induction process provides for at least some
relationship building between the new employee/s and the team.
Opener ability refers to a personality variable associated with getting others to disclose personal informa-
tion about themselves (see Miller et al., 1983). The results from Study 2 indicated that opener ability was sig-
nicantly correlated with co-worker knowledge. Miller et al.s (1983) explanation of the characteristics
associated with higher openers suggests this nding is not surprising. They note that high openers are able
to get others to feel relaxed around them, have an interest in listening to others, and have strong interpersonal
skills. Hence, high openers do not simply ask questions but rather use a combination of skills to elicit infor-
mation from others. While it is still perhaps the case that liking a co-worker (or not) may play a part in
whether caring develops, it is also probably the case that individuals who are high openers would reach either
one of these conclusions faster simply because they explore who they are working with faster than low
openers.
We had expected that group orientation would operate in a similar way to opener ability. As noted group
orientation refers to a category of work values that centre around social relations with peers, supervisors, and
others (De Vos et al., 2003). De Vos et al. (2005) suggest that people high in group orientation are more likely
to socialise with others at work and help or have concern for them. Study 2 did nd that group orientation was
signicantly correlated with opener ability, and with the personal support dimension of contextual perfor-
mance, which are results consistent with De Vos et al. (2005) description of the construct, but it was not
related to caring in any appreciable way. This result was unexpected, but as in the case of co-worker knowl-
edge, an orientation towards the group might not be sucient if it turns out that the group members are not
particularly likable.
Prior research has linked co-worker and supervisor support to a decrease in workplace accidents (Oliver
et al., 2002), and both have been suggested to be predictive of injury rates (Iverson and Erwin, 1997). While
not previously examined, it was suggested that in a team setting co-worker support in particular may be
related to caring. The results from Study 2 show positive correlations between both co-worker and supervisor
support and caring, but neither correlation was signicant. Opener ability was correlated signicantly with co-
worker support, which might be expected as higher openers could be thought of as supportive employees who
perhaps engender a reciprocal supportive relationship from their co-workers.
While Study 2 did not produce the pattern of relationships we expected in relation to care scores, the vari-
ables that we expected to be related to care scores all correlated signicantly with the personal support dimen-
sion of contextual performance. That is, co-worker knowledge, group orientation, and co-worker support
(and indeed opener ability) were all signicantly correlated with personal support. Furthermore, care scores
C.D.B. Burt et al. / Safety Science 46 (2008) 7991 89

and personal support were strongly related, yet according to the factor analysis are separate constructs, sup-
porting Clarkes (2003) suggestion that actively caring may be a safety-specic type of contextual performance
dimension. In general personal support refers to helpful, considerate, and co-operative types of behaviours
that are directed towards co-workers and it appears that the variables that we expected would be related to
caring (where helpful, considerate and cooperative behaviours are safety related) do in fact relate to personal
support. Our discussion of trust in managements commitment to safety provides one explanation of why these
variables did not relate to caring. While management might encourage personal support because it has many
positive benets, safety is a more complex issue with considerations of legal compliance (accident investiga-
tion) issues and production costs which may see less overt management support for the types of caring behav-
iours we have described. Thus it may be one thing to be helpful and considerate towards ones co-workers, but
is sometimes a more dicult thing to go beyond this and also care about their safety.
The observed relationship between caring and personal support is theoretically and practically important as
it may provide further insights into the caring concept in general, and point towards ways to manage it. The
recent paper by Morgeson et al. (2005) which examined contextual performance in relation to selecting indi-
viduals into a team setting, argued that contextual performance may be an essential competency when indi-
viduals work in a collaborative and highly interdependent team setting. They reviewed meta-analytic
studies on personality and job performance, and hypothesised that the personality characteristics of conscien-
tiousness, extraversion, agreeableness and emotional stability would be positively related to the personal sup-
port dimension of contextual performance in a team setting (note Morgeson et al.s, 2005 used the term
interpersonal facilitation, and used a measure which was similar but not identical to ours). Morgeson et al.
(2005) results supported these hypotheses at the bivariate level, and overall they found that social skills, con-
scientiousness, extraversion and teamwork knowledge incrementally predicted personal support. From this
work, and our ndings of a strong relationship between caring and the personal support dimension of contex-
tual performance, it is likely that these predictors might also be useful for predicting caring.
The results from this study have several implications for researchers and safety practitioners. The ndings
suggest that both individual and social relations variables are important for understanding caring. Further-
more, future research might show that trust in managements commitment to safety is related to caring. These
results and suggestions are consistent with Cheyne et al.s (1998, 2002) safety climate model which proposes
that a broad range of variables (i.e. organisational, work environment, group, and individual) can inuence
safety climate. We believe that identifying a strong link between caring and contextual performance provides
further evidence of the importance of contextual performance in team work settings. Furthermore, if caring is
a strong correlate of the personal support dimension of contextual performance it does provide a very good
argument for utilising predictors of personal support, such as those identied in the work of Morgeson et al.
(2005) when recruiting for teams working in high risk situations.
In terms of future research directions we have already noted possible linkages between trust and caring.
Another important consideration is to demonstrate that caring actually does have a positive eects on team
safety. A number of variables might be measured including accidents, near misses, active safety communica-
tion, perceived risk, hazard identication, hazard removal, and oering assistance to others. Our prediction is
that teams with a strong caring climate should have lower accident rates and fewer near misses, and should
actively communicate safety issues more frequently, should perceive that risk is lower (but not to a point where
true risk is being masked), should identify hazards and removed them from the work environment faster, and
should demonstrate greater frequency of oers of assistance to co-workers when compared to teams with a
weak caring climate.

References

Bell, J., Grekul, J., Lamba, N., Minas, C., 1995. The impact of cost on student helping behaviour. Journal of Social Psychology 135 (1),
4956.
Borman, W.C., Buck, D.E., Hanson, M.A., Motowidlo, S.J., Stark, S., Drasgow, F., 2001. An examination of the comparative reliability,
validity, and accuracy of performance ratings made using computerized adaptive rating scales. Journal of Applied Psychology 86 (5),
965973.
Brown, R.L., Holmes, H., 1986. The use of factor-analytic procedure for assessing the validity of an employee safety climate model.
Accident Analysis and Prevention 18 (6), 455470.
90 C.D.B. Burt et al. / Safety Science 46 (2008) 7991

Burns, C., Mearns, K., McGeorge, P. Explicit and implicit trust within safety culture. Risk Analysis, in press.
Burt, C.D.B., Gladstone, K.L., Grieve, K.R., 1998. Development of the considerate and responsible employee (CARE) scale. Work and
Stress 12 (4), 362369.
Caplan, R.D., Cobb, S., French, J.R., 1975. Relationships of cessation of smoking with job stress, personality and social support. Journal
of Applied Psychology 60 (2), 211219.
Cheyne, A., Cox, S., Oliver, A., Tomas, J.M., 1998. Modelling safety climate in the predictions of levels of safety activity. Work and Stress
12, 255271.
Cheyne, A., Oliver, A., Tomas, J.M., Cox, S., 2002. The architecture of employee attitudes to safety in the manufacturing sector. Personnel
Review 31 (6), 649670.
Clarke, S., 2003. The contemporary workforce: implications for organisational safety culture. Personnel Review 32 (1), 4057.
Clarke, S.G., 2000. Safety culture: underspecied and overrated. International Journal of Management Reviews 2 (1), 6590.
Comer, D.R., 1991. Organizational newcomers acquisition of information from peers. Management Communication Quarterly 5 (1), 6489.
Cooper, M.D., Phillips, R.A., 2004. Exploratory analysis of the safety climate and safety behavior relationship. Journal of Safety Research
35, 497512.
Cox, S., Cox, T., 1996. Safety Systems and People. Butterworth-Heinemann, Oxford.
Cox, S., Cox, T., 1991. The structure of employee attitudes to safety: a European example. Work and Stress 5, 93106.
Crampton, S.M., Wagner, J.A., 1994. Perceptpercept ination in microorganizational research: an investigation of prevalence and eect.
Journal of Applied Psychology 79 (1), 6776.
De Vos, A., Buyens, D., Schalk, R., 2005. Making sense of a new employment relationship: psychological related information seeking and
the role of work values and locus of control. International Journal of Selection and Assessment 13 (1), 4152.
De Vos, A., Buyens, D., Schalk, R., 2003. Psychological contract development during organizational socialization: adaptation to reality
and the role of reciprocity. Journal of Organizational Behavior 24, 537559.
De Vos, A., Buyens, D., Schalk, R., 2005. Making sense of a new employment relationship: psychological contract-related information
seeking and the role of work values and locus of control. International Journal of Selection and Assessment 13 (1), 4152.
Dedobbeleer, N., Beland, F., 1991. A safety climate measure for construction sites. Journal of Safety Research 22, 97103.
Donald, I., Canter, D., 1994. Employee attitudes and safety in the chemical industry. Journal of Loss and Prevention 7 (3), 203208.
Evans, N.J., Jarvis, P.A., 1986. The group attitude scale: a measure of attraction to group. Small Group Behaviour 17, 203216.
Feyer, A.M., Williamson, A.M., Stout, N., Driscoll, T., Usher, H., Langley, J.D., 2001. Comparison of work related fatal industries in the
United States, Australia, and New Zealand: method and overall ndings. Injury Prevention 7, 2228.
Fields, D.L., 2002. Taking the Measure of Work: a Guide to Validated Scales for Organizational Research and Diagnosis. Sage
Publications, London.
Flin, R., Burns, C., 2004. The role of trust in safety management. Human Factors and Aerospace Safety 4 (4), 277287.
Flin, R., Mearns, K., OConner, P., Bryden, R., 2000. Measuring safety climate: identifying the common features. Safety Science 34, 177
192.
Garcis, S.M., Weaver, K., Moskowitz, G.B., Darley, J.D., 2002. Crowded minds: the implicit bystander eect. Journal of Personality and
Social Psychology 83 (4), 843853.
Geller, E.S., 2001. A total safety culture: from a corporate achievement to a global vision. Behavior and Social Issues 11 (1), 18.
Geller, E.S., Roberts, D.S., Gilmore, M.R., 1996. Predicting propensity to actively care for occupational safety. Journal of Safety
Research 27 (1), 18.
Goldberg, A., Dar-El, E., Rubin, A., 1991. Threat perception and the readiness to participate in safety programs. Journal of
Organizational Behavior 12 (2), 109122.
Hofmann, D.A., Stetzer, A., 1996. A cross-level investigation of factors inuencing unsafe behaviours and accidents. Personnel
Psychology 49 (2), 307339.
Iverson, R.D., Erwin, P.J., 1997. Predicting occupational injury: the role of aectivity. Journal of Occupational and Organizational
Psychology 70 (2), 113128.
Kramer, M.W., 1996. A longitudinal study of peer communication during job transfers: the impact of frequency, quality and network
multiplexity on adjustment. Human Communication Research 23 (1), 5986.
Mearns, K., Flin, R., Gordon, R., Fleming, M., 1998. Measuring safety climate on oshore installations. Work and Stress 12 (3), 238254.
Mearns, K., Whitaker, S.M., Flin, R., 2003. Safety climate, safety management practice and safety performance in oshore environments.
Safety Science 41 (8), 641680.
Miller, L.C., Berg, J.H., Archer, R.L., 1983. Openers: individuals who elicit intimate self-disclosure. Journal of Personality and Social
Psychology 44 (6), 12341244.
Morgeson, F.P., Reider, M.H., Campion, M.A., 2005. Selecting individuals in team settings: the importance of social skills, personality
characteristics and teamwork knowledge. Personnel Psychology 58 (3), 583611.
Morrison, E.W., 1993. Longitundinal study of the eects of information seeking on newcomers socialization. Journal of Applied
Psychology 78 (2), 173183.
Motowidlo, S.J., 2003. Job performance. In: Borman, W.C., Ilgen, D.R., Klimoski, R.J. (Eds.), Comprehensive handbook of psychology,
Industrial and Organizational Psychology, vol. 12. Wiley, New York, NY, pp. 3953.
Oliver, A., Cheyne, A., Tomas, J.M., Cox, S., 2002. The eects of organizational and individual factors on occupational accidents. Journal
of Occupational and Organizational Psychology 75 (4), 473488.
Roberts, D.S., Geller, E.S., 1995. An actively caring model for occupational safety: a eld test. Applied and Preventative Psychology 4,
5359.
C.D.B. Burt et al. / Safety Science 46 (2008) 7991 91

Simard, M., Marchand, A., 1997. Workgroups propensity to comply with safety rules: the inuence of macro organisational factors.
Ergonomics 40 (2), 172188.
Van Scotter, J.R., Motowidlo, S.J., Cross, T.C., 2000. Eects of task performance and contextual performance on systemic rewards.
Journal of Applied Psychology 85 (4), 526535.
Zohar, D., 1980. Safety climate in industrial organizations: theoretical and applied implications. Journal of Applied Psychology 65 (1),
96102.
Zohar, D., 2000. A group-level model of safety climate: testing the eect of group climate on microaccidents in manufacturing jobs.
Journal of Applied Psychology 85 (4), 587596.

You might also like