Linking safety climate perception to types of behavior

Abstract

BACKGROUND: Certain behavior types can drive the emergence of an unsafe act at the workplace which may be reflected by workers’ perceptions toward safety climate.

OBJECTIVE: The purpose of this study was to explore the relationship between workers’ safety climate perceptions (NOSACQ-50 questionnaire) and their behavior types (based on DISC personality traits: Dominance, Influence, Steadiness, and Conscientiousness).

METHODS: NOSACQ-50 and DISC assessment questionnaires were distributed on employees of four different companies (N = 755) working at hazardous workplaces in different occupational fields: fertilizer, coal mining, and oil & gas.

RESULTS: Four of the 48 correlation coefficients were statistically significant in which the strongest correlation was found between safety commitment and Influence (I) trait in public self (r = –0.119, p < 0.01). There were two groups of DISC’s profile that have a lower perception toward safety climate. The first group was indicated by high scores on Influence (I) trait in public self (17.2% of respondents), and the latter was characterized by low scores on Dominance (D) trait and high scores on Conscientiousness (C) trait in private self (2.8% of respondents).

CONCLUSIONS: Lower perception toward safety climate was found to be correlated with two groups of DISC’s profile. Workers characterized with these two profiles are considered to have priority needs for training or to be assigned to less a hazardous workplace.

Keywords

Workplace accident accident prevention safety climate behavior type

1 Introduction

Accidents are still prevalent at the workplace. In Indonesia, the government social security agency (PT. Jamsostek) reported that about 7 people died each day in workplace accidents in 2013 [1]. As seen in Table 1, the number of work-related accidents has been increasing during the last four years, though the rates slightly decline. Financially, costs of occupational accidents have also been rising with an increase of about 150% in the payment for an accident claim during 2010–2013 [1 –4]. The actual rates and costs are expected to be higher because the data are taken from only about 10% of Indonesian workers who are currently registered as a member of the agency insurance program [1]. Most of unregistered members are assumed to be from the informal sectors, with more vulnerable working conditions [5].

Many injuries are believed to be preventable, although many people consider injury as inevitable [6, 7]. Therefore, a number of efforts have been established to reduce the risk of injury at the workplace. The efforts can be attributed to the application of safety science and public health approaches with a collaboration among disciplines and sectors [8]. Continuous efforts are required to produce more effective strategies, programs, controls and technologies for injury prevention.

Research has focused on identifying personal individual characteristics, such as personality traits, that are initially associated with accident proneness. Organizational factors such as work climate, organization and leadership are also considered to be important in affecting safety performance [9, 10]. An integrative model of workplace safety has been proposed, describing how safety climate and personality characteristics (e.g. conscientiousness) affect safety knowledge and safety motivation, and how safety knowledge and safety motivation influence safety performance and safety outcomes [11]. In this work, a correlation model is constructed to better understanding the link between individual’s behavior type and safety climate.

Safety climate is generally defined as workers’ shared perceptions of management and workgroup safety policies, procedures, and practices at a given time [12]. Safety climate can be used as an antecedent to safety performance in organization [11, 13]. Safety climate is the observable manifestation of safety culture [14], and one of the predictors for safety behavior [15 –17]. An improvement in the climate is believed to improve the perceived risk, and would leading to a safer way of conducting a job. Among safety climate assessments, the Nordic Occupational Safety Climate Questionnaire (NOSACQ-50) is a tool that has been widely used in several countries, and is available in several languages [18].

In this study, we believe that certain types of behavior can drive the emergence of an unsafe act. A classical theory of Marston [19] mentioned that behavior expressions of emotion can be categorized into tetra-logy based on their perceptions toward themselves in relations with their environment (surface personality traits). These four behavior dimensions are called Dominance (D), Inducement (later called Influence, I), Submission (later called Steadiness S), and Compliance (later called Conscientiousness, C) [20]. Variation in each dimension will direct the behaviors shown by an individual in public self, private self, and perceived self [21]. Public self reflects the social image that has been showed by an individual in front of others. Private self represents behaviors exhibited in stressful situation, while perceived self reflects a mental picture of self.

People who score high in the intensity of the Dominance (D) scores (high D individuals) in perceived self and public self (low in private self) tend to use power as their first consideration, and they love challenges. High I individuals in perceived self and public self (low in private self) tend to focus on people, and are associated with enthusiasm, persuasiveness, and optimism. High S individuals in perceived self and public self (low in private self) tend to fear change and try to restore harmony. Finally, high C individuals in perceived self and public self (low in private self) tend to strive for perfection, are cautious, careful, exacting, and systematic [20, 21].

Previous studies have investigated relationships between safety climate and safety behavior [14, 17], and their associations to work related accidents [15, 16]. However, it is still not known how the results can be used for injury prevention at the workplace. The purpose of this study was to investigate the relationships between behavior types and safety climate dimensions. Marston’s tetra-logy differentiation was used as the basis of behavior type, referring to the DISC (Dominance, Influence, Steadiness, and Conscientiousness) traits. NOSACQ-50 was utilized as the instrument for measuring safety climate perception. Results of this study were expected to provide a better understanding of the relationships between individual’s behavior type and their perception toward their behavior regarding safetyclimate.

2 Methods

2.1 Measures

2.1.1 NOSACQ-50

As previously mentioned, NOSACQ-50 was used to assess safety climate [18]. The english version of NOSACQ-50 was translated into Bahasa Indonesia, and validated using back-translation by an expert translator. The questionnaire consists of 50 items, and is divided into seven dimensions. In this study, only three dimensions of the NOSACQ-50 safety climate related to perception toward workers’ self and behavior were analyzed, including: 1) workers’ safety commitment (Dimension-1), 2) workers’ safety priority and risk non-acceptance (Dimension-2), and 3) peer safety communication, learning, and trust in safety ability (Dimension-3). The other four dimensions reflect perceptions on management policies, procedures and practices, and are less relevant to characterize individual behaviors.

2.1.2 Behavior type

DISC (Marston Model Indonesia) was chosen as the basis for behavior type assessment. The instrument has been standardized into Bahasa Indonesia. Furthermore, it was chosen due to its simplicity(a single-page questionnaire) and only requiring about 4–7 minutes to be completed. The instrument contains 24 blocks with four adjectives in each block. Respondent’s answers represent four dimensions: Dominance (D), Influence (I), Steadiness (S), and Conscientiousness (C).

2.2 Respondents

Respondents in this study were 755 field workers of four companies with hazardous workplaces. The respondents consisted of 200 workers of a fertilizer company, 255 workers of a coal mining company, and 320 workers of two oil & gas companies. These three types of occupational field have been reported to be among the top five sectors with the most frequent workplace accidents in Indonesia [5]. The participation was voluntary, but only employees of the companies were included. Subcontractors or part-time workers were excluded. Those with a working experience of less than 6 months were also excluded. Among 755 survey data collected, only 680 survey data (87.7%) were considered complete. Respondents’ demographic characteristics can be seen in Table 2.

2.3 Data collecting procedures

The two questionnaires were distributed by gathering groups of the respondents (maximum of 60 individuals) in a room. After an explanation about the study was given, the respondents were asked to read the first page of the instrument, describing the purpose of the study, anonymity and confidentiality of their data, and their right to refrain from answering any one particular question, a group of questions, or the entire questionnaire, prior to giving their consent. The respondents were asked to provide their demographic data before completing the questionnaires (NOSACQ-50 and DISC) guided by the authors. Since the survey required anonymity, demographic data collected were only age, sex, educational background, working experience and job position (staff or manager).

All data were kept confidentially to the companies. If needed, assistance was given to explain wording or terminology. The respondents took about 30 minutes to complete both questionnaires. NOSACQ-50 was firstly given, in order to provide a frame of reference for DISC.

The average score of each NOSACQ-50’s dimension was computed if at least 50% of the total items in each dimension were completed. Perception toward safety was obtained from the dimensions’ average score. Its construct validity was calculated using Pearson product moment correlations coefficient, and internal consistency reliability was tested using Alpha Cronbach. Pearson’s validity testing resulted in correlation coefficients between 0.395–0.745 for each NOSACQ-50’s items with their dimensions using α= 1%. Alpha Cronbach was 0.919 and between 0.574–0.794 for each NOSACQ-50’s dimension.

The DISC questionnaire was interpreted by a certified psychologist to obtain each respondent’s DISC graph in perceived self, public self and private self. Based on the DISC graphs, the relative level of each DISC dimension to the midline was determined (categorized to high, fairly high, fairly low, orlow).

Correlations between NOSACQ-50 and DISC dimensions were computed using Jaspen of multiserial correlations (Jaspen’s M) since the type of data obtained from NOSACQ-50 was interval and DISC was ordinal. Based on the computed correlation values, the DISC graphs that had an association with a lower perception toward safety climate were identified.

3 Results

In general, there were no DISC trait differences according to the default category among the respondents’ profile criteria in perceived self, public self and private self, although the scores have different ranges (Table 3). More than 80% of the respondents were identified to be 1) low D in perceived self and public self (but high D in private self); 2) high S in perceived self and public self (but low S in private self); 3) high C in perceived self, high in public self (but low C in private self); and 4) low I in perceived self and public self (but fairly low I in private self).

If the respondents were grouped based on their company type, those who work in coal mining and fertilizer companies had lower scores of S in public self (categorized as fairly high S) compared to those work in the oil & gas company (categorized ashigh S).

Results of the safety climate perception for the three selected NOSACQ-50 dimensions can generally be described as follows (Table 4). Across the companies, the scores of NOSACQ-50’s Dimension-1 and Dimension-3 were in the range of 3.00–3.30. No significant differences were found across the companies in the scores of these two dimensions. Based on the NOSACQ-50 website [22], a score in this range means fairly good level of safety climate with slight need of improvement. The average score of NOSACQ-50’s Dimension-2 was 2.75, 2.96 and 2.99, respectively for the fertilizer company, the two oil & gas companies, and the coal mining company, showing a fairly low level of safety climate with need of improvement [22].

Results of the correlation between the dimensions of DISC and NOSACQ-50’s can be seen in Table 5. Four of the 48 correlation coefficients were statistically significant. Interestingly, the correlation was unique for each “self”, in which no significant correlation was found for perceived self (Fig. 1, 2). The strongest correlation was found between I in public self and Dimension-1 of NOSACQ-50 (r = –0.119, p < 0.01). S had no correlation with any dimension of NOSACQ-50.

Based on the DISC graphs, there were seven profiles with high I in public self, as shown in Fig. 3. The profiles belong to 17.2% of the total respondents (9% respondents of fertilizer company, 25% respondents of coal mining company and 19% respondents of oil & gas companies), indicating low scores on Dimension-1 of NOSACQ-50.

In addition, there were three profiles of those with both low D and high C in private self as shown in Fig. 4. The profiles belong to 2.8% of the total respondents (1.7% respondents of fertilizer company, 2.6% respondents of coal mining company and 3.9% respondents of oil & gas companies), indicating low scores on Dimension-2 and Dimension-3 of NOSACQ-50.

4 Discussion

This study was aimed to investigate relationships between surface personality traits and safety climate perception. Our main assumption is that safety climate can be used as an antecedent to safety performance in organizations [11, 13]. Our results indicated that the correlations were significant on four pair relationships, indicating the composition of affective disposition that can be associated with engagement in the safety climate perceptions. Results of this study support work of Christian et al. [11]. They concluded that personality characteristics (e.g. conscientiousness, locus of control, risk taking) as parts of distal person-related factors are correlated with safety performance and safety outcomes. Among the four personality traits, Influence (I) trait seemed to be the highest predictor for safety climate perceptions.

Influence (I) trait in public self had significant correlations with workers’ safety commitment (Dimension-1 of NOSACQ-50) and their perception toward safety climate (both coefficient correlations were negative). This means that individuals with high I in public self seem to have lower safety commitment and lower perception toward safety climate. This result can be explained as follows. Individuals with high I (people with high Influence factors) in public self, with their social images characterized by a desire of an acknowledgement of their ability and freedom from controls, have a tendency towards a lower safety commitment. On the other hand, Individuals with their social images reflected as habitually silent, holding back emotions, and drawing back from publicity (low I in public self), have a tendency towards a higher safety commitment.

In general, people with high I in public self have social images as enthusiastic, impulsive, emotional, and influential; while those with low I in public self have social images as pessimistic, aloof, and discerning. It appears that the low I trait is more apt to describe a person who commits themselves with safety. Thus, someone who is willing to take responsibility to achieve higher safety performance would be more likely to possess a less active personality as their social images.

Similarly, Conscientiousness (C) trait in private self was found to have a negative correlation with safety climate of peer safety communication, learning, and trust in safety ability (Dimension-3 of NOSACQ-50). People with low C in private self are described to be perfectionists and adhere to rules, regulation, and structure in stressful situation. This correlation can be interpreted as the follows. Individuals that have a high standard, following customary usage, and tendency to be careful and tactful in stressful situation, tend to communicate, learn, and trust their peers.

The last significant correlation was found between Dominance (D) trait in private self and workers’ safety priority and risk non-acceptance (Dimension-2 of NOSACQ-50). This dimension of safety climate measures workers’ perceptions of how they themselves relate to safety at work (e.g. prioritizing safety before production goals, not resigning to hazardous conditions, not accepting risk-taking, not showing fearlessness) [18]. Those with low D in private self are said to be demanding and love to control stressful situation, while high D individuals in private self are said to be indecisive and prefer to be a team player in stressful situation. Individuals with the first set (ready to take serious risk and forcing their desires in stressful situations) consider safety and risk as a challenge and take risks unwisely and dangerously. In additions, low D individuals have a tendency to perceive risk to be acceptable and workers’ safety not a priority in stressful situation. On the other hand, high D trait in stressful situation is more apt to describe a person who prioritizes safety.

Although it is not statistically significant, it is worth noting that there are several tendencies of correlations, found in this study, between personality traits and safety climate perception. Further analysis can be done to interpret the correlation tendency between Influence (I) trait in perceived self and workers’ safety priority and risk non-acceptance dimension; between Conscientiousness (C) trait in perceived self and public self with perception toward safety climate; and between Dominance (D) trait in public self and perception toward safety climate.

Influence (I) trait in perceived self had a tendency to negatively correlate with workers’ safety priority and risk non-acceptance dimension. This suggests that individuals who perceive their behaviors as sociable, are fond of the company of others, are easily aroused and have a tendency to accept risk and put aside safety. On the other hand, individuals who prioritize safety and disallow risk tend to perceive their behavior as self-conscious, practice looking on the dark side of things, and are habitually silent or uncommunicative.

In contrast to the negative correlations between Conscientiousness (C) trait in private scale and peer safety communication, learning, and trust in safety ability, C trait in perceived self had a tendency to positively correlate with perception toward safety climate. It appears that low C scores in perceived self tend to have lower scores on perception toward safety climate. Thus, people who perceived them self with a tendency to be tactless and rebellious tend to act an unsafe behavior.

Results of this study related to perception toward safety climate suggest that high C individuals in public self, who reflect the social image as conventional, analytical, dissecting a whole into its part in order to discover their nature, have a tendency to reflect a more positive perception toward safety climate. On the other hand, low C individuals in public self, characterized by bold resistance to authority, lack of delicacy handling situations that require diplomacy or sensitivity and not easily affected by opinions of others as their social images, have a tendency to reflect a lower perception toward safety climate. This result supports a previous study regarding behavior type and accident experiences [23].

The last correlation tendency was found between Dominance (D) trait in public self and perception toward safety climate. High D scores in public self have a social image that are said to be egocentric, demanding, risk taker, and self-reliant, while low D scorers are generally described to be realistic, weigh up the pros and cons, conservative and dependent. It appears that high D trait of social image is more apt to describe a person who perceive safety climate unfavorably and can lead to an unsafe behavior.

What do all of the above results mean to us? As previously mentioned, we hoped that the better understanding of workers’ personality traits (both advantages and limitations) can be used by safety practitioners in: 1) selecting individuals who will work on high risk jobs, 2) determining workers who have priority needs for training, and 3) selecting individuals who will be assigned to less hazardous workplace. Workers who have a greater tendency towards accidents can be assigned to jobs with lower accident risks. Those who have lower perception toward safety climate would need more training than others. This study suggests that several individual characteristics may need to get more focus during selection, training and safety monitoring. They are those with high I in public self, high C in private self, and low D in private self. Though no one would totally match with those risk traits, priority can be given to those match with at least one trait.

In the first case, we found that there are seven profiles of those with high I scores in public self, as shown in Fig. 3. This study suggests that they may have a lower safety commitment and/or lower perception toward safety climate. In this study, they accounted for 17.2% of the respondents whose social image can be described as enthusiastic, impulsive, emotional, and influential. They may have a tendency of being unwilling to take responsibility to maintain and strive to achieve higher safety levels. Among those seven profiles, those with low D scores, high S scores and high C scores seem to be dominant.

In the second case, we found that there are three profiles of those with low D scores in private self and high C scores in private self, as shown in Fig. 4. The present study suggests that they may have a lower perception of safety priority & risk non-acceptance and lower peer safety communication, learning, and trust in safety ability. They accounted for 2.8% of the respondents who are straightforward, ready to take serious risk, forcing their desire in stressful situations. In additions, they may have a tendency to perceive risk to be acceptable and workers’ safety is not a priority in stressful situations. They are also tactless and rebellious in stressful situation thus tend to be uncommunicative, reluctant to learn, and distrust their peers’ safety ability.

Those individuals grouped in both cases may benefit from additional training or to be assigned to less hazardous workplaces. This is an alternative to more effective practices for injury prevention at the workplace. Since the traits have been found to have significant correlations with the dimensions of safety climate, an improvement in safety culture is also expected.

Several limitations are worth noting. The main concern is related to the respondents’ capability. As shown in Table 2, most of the respondents, who can be categorized as “blue collar workers’, have a formal education of only high school or less. This may affect their understanding of the content in the questionnaires, although necessary explanations and guidance were given during the questionnaire session.

Another issue is related to how the respondents interpreted the questionnaires. Respondents’ behavior type can be biased due to the influence of general opinions or ideal norms (social desirability). In this study, the sample size is larger when compared to studies by Lin and Lin (2012) with n = 20 [24], Puccio and Grivas (2009) with n = 137 [25], Keren et al. (2009) with n = 117 [26], and Neal et al. (2000) with n = 525 [17] to reduce this issue. Further study, however, is suggested in a more controlled experimental situation to collect data related to respondents’ behavior types.

Despite the limitations, this study has made several contributions. We found that the safety climate assessment tool, specifically the Bahasa Indonesia version, can be used as an alternative, standardized tool for measuring safety climate. This study also contributes to improving our understanding about the relationship between surface personality traits, specifically traits that are assessed using DISC and safety climate perceptions.

5 Conclusion

Lower perception toward safety climate was found to be correlated with the following behavior types: high scores in public self on Influence (I) trait, high score in private self on Conscientiousness (C) trait, and low score in private self on Dominance (D) trait. These results, hopefully, can be considered during selection, training, and placement of workers to minimize risk of accidents related to human behavior.

Conflict of interest

The authors have no conflict of interest to report.

References

PT Jamsostek (The Indonesian Social Security Agency). Annual Reports 2013 Available from http://www.jamsostek.co.id

PT Jamsostek (The Indonesian Social Security Agency). Annual Reports 2010 Available from http://www.jamsostek.co.id

PT Jamsostek (The Indonesian Social Security Agency). Annual Reports 2011 Available from http://www.jamsostek.co.id

PT Jamsostek (The Indonesian Social Security Agency). Annual Reports 2012 Available from http://www.jamsostek.co.id

Asia Monitor Resource Center. OSH Legal Resource Handbook 2012. Asia Monitor Resource Center, Hongkong, 2012.

Nordlof

, Wiitavaara

, Windblad

, Wijk

, Westerling

. Safety culture and reasons for risk-taking at a large steel-manufacturing company: Investigating the workers perspective. Safety Science 2015;73:126–35.

Nordlof

. What is safety culture and risk-taking like at a large steel manufacturing company. Work 2012;41:5831–32.

Stout

, Linn

. Occupational injury prevention research: Progress and priorities. Injury Prevention 2002;8(Suppl IV):iv9–iv14.

Reiman

, Pietikainen

, Oedewald

, Gotcheva

. System modeling with the DISC framework: Evidence from safety-critical domains. Work 2012;41:3018–25.

10.

Neal

, Griffin

. A study of the lagged relationships among safety climate, safety motivation, safety behavior, and accidents at the individual and group levels. Journal of Applied Psychology 2006;91(4):946–53.

11.

Christian

, Bradley

, Wallace

, Burke

. Workplace safety: A Meta-analysis of the roles of person and situation factors. Journal of Applied Psychology 2009;94(5):1103–27.

12.

Zohar

. Safety climate and beyond: A multi-level multi-climate framework. Safety Science 2008;46:287–376.

13.

Griffin

, Neal

. Perceptions of safety at work: A framework for linking safety climate to safety performance, knowledge, and motivation. Journal of Occupational Health Psychology 2000;5(3):347–58.

14.

Mearns

, Whitaker

, Flin

. Safety climate, safety management practice and safety performance in offshore environments. Safety Science 2003;41:641–80.

15.

Brown

, Willis

, Prussia

. Predicting safe employee behavior in the steel industry: Development and test of a sociotechnical model. Journal of Operations Management 2000;18:445–65.

16.

Guastello

, Gershon

RRM

, Murphy

. Catastrophe model for the exposure to blood borne pathogens and other accidents in health care settings. Accident Analysis and Prevention 1999;31:739–49.

17.

Neal

, Griffin

, Hart

. The Impact of organizational climate on safety climate and individual behavior. Safety Science 2000;34:99–109.

18.

Kines

, Lappalainen

, Mikkelsen

, Olsen

, Pousette

, Tharaldsen

, Tómasson

, Törner

. Nordic safety climate questionnaire (NOSACQ-50): A new tool for diagnosing occupational safety climate. International Journal of Industrial Ergonomics 2011;41:634–46.

19.

Marston

. Emotions of normal people. Great Britain: Devonshire Press;1928.

20.

Inscape Publishing. DISC^® classic and models of personality research report; 1996.

21.

Authentic School of Behavioral Studies. DISC Course; 2008.

22.

National Research Centre for the Working Environment (NRCWE). Interpreting the Nordic Occupational Safety Climate Questionnaire NOSACQ-50 results. Available from http://www.arbejdsmiljoforskning.dk/en/publikationer/spoergeskemaer/nosacq-50/how-to-use-nosacq-50/interpreting-nosacq-50-results

23.

Anatasia

, Yassierli . Linking accident case and behavior type. Proceeding of Ergonomic Conference, Bandung, Indonesia; 2012.

24.

Lin

, Lin

. Investigating the relationship among DISC personality characteristics, job stress and job performance in F&B Industry. Presented at BIA Conference, Sapporo, July 3-5 2012.

25.

Puccio

, Grivas

. Examining the relationship between personality traits and creativity style. Creativity and Innovation Management 2009;18:247–55.

26.

Keren

, Mills

, Freeman

, Shelley

. Can level of safety climate predict level of orientation toward safety in decision making task? Safety Science 2009;47:1312–23.