Journal of Safety Research 53 (2015) 39–43
Contents lists available at ScienceDirect
Journal of Safety Research journal homepage: www.elsevier.com/locate/jsr
Research note
Prevalence and demographic differences in microaccidents and safety behaviors among young workers in Canada☆ Nick Turner, a Sean Tucker, b,⁎ E. Kevin Kelloway c a b c
Haskayne School of Business, University of Calgary, Canada Faculty of Business Administration, University of Regina, Canada Department of Psychology, Saint Mary's University, Canada
a r t i c l e
i n f o
Article history: Received 16 July 2014 Received in revised form 7 January 2015 Accepted 9 March 2015 Available online 27 March 2015 Keywords: Injuries Safety voice Young workers Canada
a b s t r a c t Introduction: The present study examines the self-reported frequency of non-lost work time workplace injuries (“microaccidents”) and the frequency of three types of work-related safety behaviors (i.e., safety voice, safety compliance, and safety neglect) recalled over a four-week period. Method: We analyzed data on microaccidents and safety behaviors from 19,547 young workers (aged 15–25 years, Mdn = 18 years; 55% male) from multiple Canadian provinces. Results: Approximately one-third of all young workers recalled experiencing at least one microaccident at work in the last four weeks. Comparisons across three age groups revealed that younger workers, particularly between the ages of 15–18, reported more frequent microaccidents, less safety voice, less safety compliance, and more safety neglect than workers aged 19–22. This pattern of results also held for comparisons between workers in 19–22 and 23–25 age groups, except for safety voice which did not differ between these two older age groups. In terms of gender, males and females reported the same frequency of microaccidents, but males reported more safety voice, more safety compliance, and more safety neglect than females did. The results and limitations of the present study are discussed. Conclusion: Frequency of microaccidents and safety behavior vary among young worker age sub-groups. © 2015 National Safety Council and Elsevier Ltd. All rights reserved.
1. Introduction Injury statistics show that young workers (aged 15 to 25 years) are more likely to be injured on the job than adults (Salminen, 2004) with young males at highest risk of injury (Breslin & Smith, 2005). In 2011, 31,221 young workers in Canada (Association of Worker Compensation Boards, 2012a) and 116,900 in the United States (Bureau of Labour Statistics, 2012a) missed time away from work due to a workplace injury. That same year, 35 young workers in Canada (Association of Worker Compensation Boards, 2012b) and 380 in the United States (Bureau of Labour Statistics, 2012b) were killed on the job. In contrast to these statistics on severe work injury outcomes, the current paper presents the prevalence of non-lost time work injuries (also known as “microaccidents;” Zohar, 2000) and a range of safety behaviors (Tucker & Turner, 2011) among a large sample of young workers in Canada. Understanding the prevalence of less severe ☆ We presented earlier versions of this paper at the 2012 Congress of the Association of Workers Compensation Boards of Canada, Winnipeg, Canada, and the 30th International Congress of Psychology, Cape Town, South Africa. We thank Phil Germain, Sarah Goodhope, Alex Johnson, Paul Kells, Corey Pocaluyko, and Warren Preece for supporting this project. The first two authors contributed equally to this paper. No conflict of interest exists between the authors and publication of this research. ⁎ Corresponding author at: Faculty of Business Administration, University of Regina, SK, S4S 0A2, Canada. E-mail address: [email protected] (S. Tucker).
http://dx.doi.org/10.1016/j.jsr.2015.03.004 0022-4375/© 2015 National Safety Council and Elsevier Ltd. All rights reserved.
workplace injuries and safety outcomes that are upstream from severe workplace injuries is important for at least three reasons. First, more frequent but less severe forms of injuries are correlated with less frequent but more severe workplace injuries (Heinrich, Petersen, & Ross, 1980). Understanding predictors of more severe workplace injuries is critical in prevention efforts. Second, this study makes an empirical contribution by describing the occurrence of work-related safety outcomes among a large sample of young workers. This is a vulnerable population in the global workforce which, compared to jobs held by older workers, often experiences lower-quality work conditions because of the irregular, transient, low-wage, and part-time characteristics of many of the jobs they hold (International Labour Organization, 2013). Third, while some studies examine variation among young workers safety experiences (e.g., Breslin, Polzer, MacEachen, Morrongiello, & Shannon, 2007; Breslin & Smith, 2005), related research tends to assume the shared experience of work among this broad cohort (Barling & Kelloway, 1999). The current study describes the prevalence of safety behavior and microaccidents by young worker gender and age group (i.e., 15–18 years, 19–22 years, and 23–25 years) in a large sample of Canadian young workers. 2. Theoretical background Research on young worker safety is diverse and interdisciplinary (Runyan, Lewko, & Rauscher, 2012). Studies have identified situational
40
N. Turner et al. / Journal of Safety Research 53 (2015) 39–43
correlates of work-related injuries, such as work pressure (Breslin et al., 2007), gender differences in risk exposure (e.g., Breslin & Smith, 2005), the role of parents in developing children's safety attitudes (Runyan, Schulman, Santo, Bowling, & Agans, 2009), and the prevalence of work-related safety training among employed young people (Zierold, Welsh, & McGeeney, 2012). Absent from this literature is prevalence studies of work-related injuries and safety behaviors based on large samples of young working people. The current study examines two types of safety outcomes: microaccidents (Zohar, 2000) and safety behaviors (Tucker & Turner, 2011). Microaccidents are physical wounds to the body from workrelated activities that do not require time away from work (sometimes called “non-lost-time” injuries), and may require visits to a first aid station. This is in contrast to more severe injuries, variously called “major injuries” or “lost-time injuries” by health and safety regimes, and which necessitate time away from work for off-site medical attention, recuperation, or both. We studied non-lost time injuries for two reasons. First, little is known about non-lost time injury experiences in large samples of young workers; the majority of research on young worker injury prevalence examines lost-time injuries. Second, because microaccidents occur more frequently than lost-time injuries and consistent with Heinrich's iceberg concept (Heinrich et al., 1980), such events are a precursor to less-frequently occurring but more severe injuries in organizations. Studies of lost-time injury claims consistently find that young males between the ages of 19 and 25 experience the highest rate of occupational injuries (e.g., Breslin & Smith, 2005; McCall, Horwitz, & Carr, 2007; WorksafeBC, 2011). At the same time, however, Breslin et al.'s (2007) review of high-quality multivariate studies of predictors of young worker injuries concluded that neither gender nor young worker age was related to self-reported injuries, particularly in the presence of job and work characteristics. Overall, therefore, the findings are unclear with respect to the impact of young worker gender and age on workplace injuries. Workplace injuries in the studies included in Breslin et al.'s (2007) review covered the full spectrum of severity excluding workplace deaths by injury, and given the different frequency distributions between less severe and more severe work injuries described above, the current study focuses on microaccidents among young workers by gender and age. In addition to workplace injuries, two important and widely-studied safety outcomes are safety compliance (defined as “the core safety activities that need to be carried out by individuals to maintain workplace,” Griffin & Neal, 2000, p. 349) and safety participation (defined as voluntary extra-role behaviors that “help to develop an environment that supports safety,” Griffin & Neal, 2000, p. 349). A recent metaanalysis by Clarke (2013) containing 32 adult-aged working samples revealed that higher levels of safety compliance and safety participation were associated with fewer occupational injuries (average rs for both safety compliance and safety participation with occupational injuries = − .21, both with N = 229 reliability-corrected metaanalytic correlations). Recently, Tucker and Turner (2011) broadened this set of safety behaviors to reflect the work experiences of young workers in response to physically dangerous work. In addition to safety compliance, they measured safety neglect (e.g., taking shortcuts or work-arounds) as antithetical to safety compliance, as well as safety voice (e.g., speaking up about hazardous work) as a specific form of safety participation.1 The results of two focus group studies provide insight into the relative frequency of some of these behaviors and gender differences. Breslin et al. (2007) found that young female participants were more likely than young males in their sample to say they speak up about occupational safety concerns, but also that such behavior is commonly 1 Tucker and Turner (2011) also measured safety patience (i.e., waiting for safe conditions to improve), but the two-item measure was not measured in data set due to space restrictions in the questionnaire.
viewed as complaining by supervisors. In contrast, Tucker and Turner's (2013) focus group study (N = 39 participants) found no differences in the frequency of work-related safety voice between young males and young females. However, they noted that “speaking up” behavior (specifically change-oriented voice) is relatively uncommon due to concerns young workers expressed about fears of losing their jobs, supervisor indifference, their relatively short tenures, and felt powerlessness. Safety compliance has been examined with adult working populations, however safety neglect is a new construct and there are no data on its prevalence in either adult or young worker populations. The current study is motivated to extend emergent research on young worker safety behaviors and bring clarity to questions related to gender and age differences in these key safety behaviors. The current study uses a large sample of young Canadian workers to explore the prevalence of microaccidents, safety voice, safety compliance, and safety neglect. Given the mixed evidence of whether young males and young females differ in terms of the frequency of microaccidents and the lack of data on young worker age and gender across this range of safety behaviors, our study explores these potential differences as research questions. 3. Method 3.1. Study context, procedure, and sample Participants were employed Canadians aged 15 to 25 years who responded to a short on-line survey between September 2011 and July 2012. The voluntary survey, which was approved by each of the author's university research ethics board, appeared at the beginning of the Passport to Safety™ (PS) program. PS (http://www. passporttosafety.com/) is an on-line training and assessment tool that is meant to build awareness of workplace hazards, job rights and responsibilities, and work injury prevention approaches among students. It is widely used in Canada, as well as some jurisdictions in the United States and Australia, primarily in high schools as a part of occupational health and safety instruction in the curriculum (Brotherton, 2012). High school teachers who use the PS program in their classes provide each student with a unique personal identification number, and each student logs into the on-line system as either part of in-class instruction or a homework assignment. The short survey they were presented with before taking the PS program included questions about demographic characteristics, work-related injuries, workplace safety behaviors—all taking less than two minutes to complete. Prior to completing the survey, participants had the option to read the PS privacy policy, describing how their responses to the questions may be used in academic research. The data reported in the current study were collected in Canada, primarily from Ontario, the country's most populous province. Over 66,500 people started the short survey prior to beginning the on-line test. After eliminating cases with missing data, participants outside of the study's age range (i.e., over 26 years, inclusive) and unemployed individuals, a sample of 19,547 cases remained (55% male, median age = 18 years). Other data collected from some of these participants was used in a recently published study (Tucker et al., 2014). 3.2. Measures 3.2.1. Demographic information We asked respondents to report their year of birth and gender. 3.2.2. Microaccidents We asked respondents to respond to the following statement — “How many times in the past 4 weeks have you had a ‘minor’ workplace injury (e.g., cut, burn, strain, and sprain) that did not result in you missing time from work?” They responded in six groups ranging from no microaccidents experienced (coded as 0) to five or more microaccidents (coded as 5).
N. Turner et al. / Journal of Safety Research 53 (2015) 39–43
3.2.3. Safety behaviors We used the two highest-loading items from three of Turner and Tucker's (2011) young worker safety behaviors: safety voice (sample item: “Tell my supervisor about hazardous work”), safety compliance (sample item: “Wear protective clothing/equipment”), and safety neglect (sample item: “Take shortcuts that threaten my personal safety”). Respondents responded on a three-point scale: 1 = almost never, 2 = sometimes, 3 = almost always. Since the vast majority of test takers were high school students taking the PS test in a computer classroom, we truncated the response scale from seven to three points to simplify the appearance of the survey and mitigate any confusion.
3.3. Data analysis We used the two highest-loading items from each of Tucker and Turner's (2011) six-item safety voice scale, six-item safety neglect scale, and three-item safety compliance scales. As all items used had truncated three-point ordinal scaling (1 = almost never, 2 = sometimes, 3 = almost always) compared to the seven-point original (1 = almost never to 7 = almost always), we conducted a confirmatory factor analysis on the six items to assess the hypothesized 3-factor model using weighted least squares means and variance adjusted estimation in MPlus 6.11 (Muthén & Muthén, 2011) with listwise deletion. The hypothesized 3-factor model fit the data well, χ2 (6, N = 19,235) = 183.74, p b .001, comparative fit index (CFI) = .996, root mean squared error of approximation (RMSEA) = .039, based on Hu and Bentler's (1998) benchmarks for fit indices. The two items underlying each of the three factors indicated adequate internal consistency (ordinal alphas = .76, .63, and .69, for safety voice, safety compliance, and safety neglect, respectively; Meulman, Van Der Kooij, & Heiser, 2004), recognizing that benchmarks for reliability based on ordinal (vs. continuous) data are more conservative (Zumbo, Gadermann, & Zeisser, 2007). The three-factor model yielded better fit than (a) the two-factor model in which safety compliance formed one factor and safety voice and safety neglect combined to form a single factor (model fit comparisons appear in square brackets), CFI = .965, RMSEA = .103 [Δχ2 (2, N = 19,235) = 1,048.23, p b .001]; (b) the two-factor model in which safety voice formed one factor and safety compliance and safety neglect combined to form a single factor, CFI = .687, RMSEA = .309 [Δχ2 (2, N = 19,235) = 8,988.80, p b .001]; and (c) the two-factor model in which safety neglect formed one factor and safety compliance and safety voice combined to form a single factor, CFI = .682, RMSEA = .311 [Δχ2 (2, N = 19,235) = 9,375.51, p b .001]. As all these measures are self-report, we tested the goodness-of-fit of a one-factor model to test for mono-method bias (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003), which yielded a much weaker fit than the hypothesized three-factor model, CFI = .635, RMSEA = .315 [Δχ2 (3, N = 19,235) = 12,271.06, p b .001], suggesting that the self-reported nature of the scales did not mask construct differences. Taken together, the results of the confirmatory factor analysis, summarized in Table 1, provide support for the existence of three distinct safety behaviors replicating Tucker and Turner's (2011) factor structure for these constructs.
Table 1 Results of confirmatory factor analysis. Model
χ2
df
RMSEA
CFI
Model 1* Model 2 Model 3
183.74 1,231.97 9,172.54
19,229 19,233 19,233
.039 .103 .309
.996 .965 .687
Notes. * Hypothesized model. Model 1 has three factors (voice, neglect, and compliance). Model 2 has two factors (voice (with neglect), and compliance). Model 3 has one factor (voice, neglect, and compliance as a single factor). RMSEA = root-mean-square error of approximation. CFI = comparative fit index.
41
4. Results 4.1. Frequency of workplace injuries between age groups and by gender Slightly over two-thirds of young workers (66.5%) reported no microaccidents in the previous four weeks. Just over a quarter (25.5%) reported between one and two microaccidents in the previous four weeks, with the remaining 8% of young workers sampled reporting three or more microaccidents in the previous four weeks (3 injuries = 3.6%, 4 injuries = 1.5%, 5 or more injuries = 2.9%). Table 2 presents frequency of microaccidents for the overall sample, age groups, and by gender. We conducted a Kruskal–Wallis test to evaluate differences among three age groups (15–18 years, 19–22 years, 23–25 years) on the distribution of microaccidents, χ2 (2, N = 19,537) = 115.45, p b .001. We then conducted Mann–Whitney tests for pairwise differences among combinations of the three age groups, controlling for Type I error across tests with Bonferroni correction. Workers between the ages 15–18 reported more frequent microaccidents than workers between the ages 19–22 (Z = 6.09, p b .001), and workers aged 19–22 reported more frequent microaccidents than workers between ages 23–25 (Z = 6.52, p b .001). Overall, young men and young women reported no difference in frequency of microaccidents (Z = 1.53, p = .13). 4.2. Frequency of safety behaviors by age groups and gender Table 2 presents frequency of the three types of safety behaviors for the overall sample by gender and age group. We conducted a Kruskal– Wallis test to evaluate differences among the three age groups on safety voice, χ2 (2, N = 19,537) = 11.93, p b .01; safety compliance, χ2 (2, N = 19,537) = 67.00, p b .001; and safety neglect scores, χ2 (2, N = 19,537) = 87.29, p b .001. We then conducted Mann–Whitney tests for pairwise differences among the three age groups, controlling for Type I error across tests with Bonferroni correction. In terms of safety voice, workers aged 15–18 reported less safety voice than workers aged 19–22 (Z = − 3.34, p b .01), but safety voice scores in the 19–22 age group did not differ from the scores in the 23–25 age group (Z = − .35, p = .72). In terms of safety compliance, workers aged 15–18 reported less safety compliance than workers aged 19–22 (Z = − 6.62, p = .02), and workers in the 19–22 age group reported less safety compliance than workers in the 23–25 age group (Z = − 2.28, p = .02). In terms of safety neglect, workers aged 15–18 reported more safety neglect than workers aged 19–22 (Z = 4.79, p b .001), and safety neglect scores in the 19–22 age group were higher than safety neglect scores in the 23–25 age group (Z = 6.20, p b .001). Overall, young men reported more safety voice (Z = 2.80, p b .01), more safety compliance (Z = 2.14, p b .05), and substantially more safety neglect (Z = 14.13, p b .001) than young women did. 4.3. Relationships among microaccidents and safety behaviors Table 3 presents the zero-order non-parametric correlations and partial non-parametric correlations (i.e., correlations controlling for age group and gender) among the study variables. Frequency of microaccidents was correlated weakly albeit negatively with safety voice (r = − .03, p b .01) and safety compliance (r = − .05, p b .01), and weakly albeit positively with safety neglect (r = .05, p b .01). Controlling for age group and gender, the pattern and magnitude of associations did not change (rs = −.02, −.03, and .03, ps b .01, respectively). 5. Discussion 5.1. Summary of findings and their implications This is the first study to investigate the prevalence of both microaccidents and a range of safety behaviors in a large sample of young
42
N. Turner et al. / Journal of Safety Research 53 (2015) 39–43
Table 2 Comparison by sex and age.
Overall Sex Male Female Age group 15–18 19–22 23–25
N (listwise deletion)
Mean injuries
95% CI
Mean safety voice
95% CI
Mean safety compliance
95% CI
Mean safety neglect
95% CI
19,547
.65
.63, .67
2.14
2.13, 2.15
2.43
2.42, 2.44
1.30
1.29, 1.31
10,716 8,760
.68 .62
.66, .71 .59, .64
2.16 2.12
2.14, 2.17 2.11, 2.14
2.45 2.41
2.44, 2.46 2.40, 2.43
1.34 1.25
1.33, 1.35 1.24, 1.26
15,083 3,820 664
.68 .58 .30
.67, .70 .55, .62 .24, .36
2.13 2.18 2.15
2.12, 2.14 2.15, 2.20 2.10, 2.21
2.41 2.49 2.53
2.40, 2.42 2.47, 2.51 2.48, 2.58
1.31 1.28 1.15
1.30, 1.32 1.26, 1.29 1.12, 1.18
workers aged 15 to 25 years. The purpose of the study was to assess potential differences in these outcomes by age and gender. The youngest participants (15 to 18 years old) reported more microaccidents than the oldest participants (19 to 25 years old). Further, there was no gender difference in frequency of microaccidents across the age groups. While lost-time injury claim data shows that males aged 20–24 experience the highest rate of injuries (e.g., Breslin & Smith, 2005), Breslin et al.'s (2007) review found that among young worker samples there was no evidence that either age or gender related to self-reported injuries after controlling for job or hazard factors. The current findings on microaccidents are generally consistent with the conclusions of Breslin et al.'s review, except for higher level of microaccidents in the 15- to 18-year old age group. It is important to note that the current study examined over a four-week period frequency of microaccidents – low severity, non-lost-work time injuries in a short, easy-to-remember time period \ whereas the Breslin et al. review combined the full range of injury severity (ranging from “strain/sprains/cuts/lacerations, burns, bruises/contusions, fractured bone, dislocated joint, serious muscle/back pain, blisters” through to “the loss of at least one shift, received hospital treatment, received sutures to a wound,” but excluding fatalities) and across an indeterminate period of time (ranging from “occurrence of injury during summer job” to “ever injured while working”). Furthermore, the findings from the current study are surprising given that, in general, the oldest participants would tend to work more hours than younger participants (Marshall, 2007); thus, in terms of exposure, all else equal, older participants should experience more microaccidents. Controlling for time worked in the last four weeks would have helped to mitigate concern about differences in exposure to possible work hazards. Not only did the youngest survey respondents report experiencing more minor injuries, but also they reported the lowest frequency of both safety voice and safety compliance as well as the highest frequency of safety neglect. These findings are a concern because young employees working their first jobs are vulnerable on a number of fronts, and it is hoped that exposure to unsafe working conditions and direct or vicarious experience of workplace injuries would encourage speaking up about safety concerns in an effort to try and improve workplace safety. This is not always the case, however. Tucker and Turner (2013, 2014) have shown that teenaged workers are more likely to ‘wait and see’ whether unsafe work conditions improve, and that speaking up about Table 3 Spearman correlations among study variables (N = 19,479–19,547).
1. 2. 3. 4 5. 6.
Variable
1.
2.
3.
4.
5.
6.
Age group Gender Safety voice Safety compliance Safety neglect Microaccidents
– −.01 .02 .06 −.06 −.07
– .02 .02 .10 .01
– .48 .03 −.03
.49 – −.08 −.05
.10 .01 – .05
−.02 −.03 .03 –
Note. Zero-order correlations appear below the diagonal. Partial correlations, controlling for age group and gender, appear above the diagonal. Correlations equal or greater than |.02| are significant at the .01 level. Correlations greater than |.01| are significant at the .05 level. Age group: 1 = 15–18 years old; 2 = 19–22 years old; 3 = 23–25 years old. Gender: 1 = male, 0 = female.
safety concerns, while possible, is rare and occurs when unsafe conditions or their consequences are extreme and young workers feel psychologically safe in doing so. Interestingly, in the current study, male participants reported higher safety voice than female participants, which contradicts previous findings using hypothetical experimental scenarios (Tucker & Turner, 2014) and focus groups (Breslin et al., 2007). One possibility is that gender is a proxy for job risk. That is, young males self-select or are assigned work that is objectively more physically risky than the work assigned to female young workers. Past research suggests that adult women compared to adult men face less risk of physical injury at work (Hersch, 1998). Young males may report higher safety voice and higher safety compliance than young females because young males may be assigned work, conduct work, or encounter work situations in which likelihood of physical injury is higher. Thus, by necessity, young males may be more likely to speak up about hazardous work conditions or comply with safety rules to keep safe. Future research is needed to determine whether the differences across studies are due to differences in male and female representation in occupations and work sectors (e.g., service sector vs. construction) and their related risk factors, or methodological factors (e.g., laboratory-based vs. field-based research). Meta-analytic evidence has shown that safety compliance and safety participation are negatively associated with workplace injuries (Clarke, 2013), and this is what we found with safety compliance and safety voice in this study albeit with small effect sizes (absolute value of partial rs ranging from .02 to .03). Furthermore, safety neglect was related to greater frequency of microaccidents, and younger males particularly those in the 15- to 18-year old age group reporting higher safety neglect than young females did. We can only speculate as to why safety neglect may be higher among young males. First, in line with the argument above about males doing different types of work than females, it is possible that young males work in physically riskier conditions, and thus in addition to having more reason to speak up and comply with safety rules, there are also more safety rules to disregard. Second, employees of all ages, but particularly young workers with relatively little experience, may have distorted risk perceptions, resulting in underestimating or overestimating risks (Mbaye & Kouabenan, 2013). More specifically, young people may view themselves as impervious to harm (e.g., Loughlin & Frone, 2004), which helps to explain why young workers often disregard company policies, take shortcuts, and ignore hazards. Relatedly, there is the question of how to reconcile the result showing that males concurrently reported the most frequent safety neglect and safety compliance behaviors — distinct behaviors that one would not expect to be simultaneously higher. However, we note that in absolute terms, males report substantially more frequent safety compliance than safety neglect behavior (Ms = 2.45 vs. 1.35, respectively), and that the two variables are negatively correlated when the sample is limited to male participants (r = −.03, p b .001). 5.2. Study limitations and conclusion Despite the unique characteristics of the current sample, this study has several limitations worth noting. First, the distribution of the age
N. Turner et al. / Journal of Safety Research 53 (2015) 39–43
of the sample is skewed and selection effects are possible. Nearly all participants aged 15 to 18 years complete PS in a high school classroom, while those 19 years and above (i.e., having graduated from high school or were early school-leavers) primarily participating through their employer. It is possible that taking the PS test in the context of a class on workplace health and safety may encourage participants to respond in socially-desirable ways – for example, under-reporting microaccidents or over-reporting safety compliance – and that employers that incorporate PS into job-related training are safer workplaces than employers who do not offer PS training to their employees. This does not explain why the younger age groups reported more microaccidents and less safety compliance; however, it may explain why the older age groups reported fewer microaccidents and exercised safety compliance more frequently. Second, we only collected data on microaccidents (and not lost-time workplace injuries) and, for the safety behaviors, we used shortened scales limited to two items providing participants with three response possibilities. For microaccidents, it is possible that we missed important differences in age or gender based on the sole measurement of low severity injuries. For example, it is possible that young males report more lost-time work injuries than young females report, but we are unable to ascertain this, given that we have only collected non-lost-time injuries (microaccidents). For the measurement of safety behaviors, we necessarily restricted the range of both the focal constructs (only use two items per scale) and their variance (only using three-point response scales instead of the original seven-point response scales), resulting in lower overall scale reliability. However, we used statistical tests that are appropriate for analyzing truncated ordinal scaling. Relatedly, with the exception of the correlation between safety compliance and safety voice (r = .48), the relationships among the safety behaviors were small and likely have little practical significance. Finally, due to the retrospective nature of our injury data (i.e., participants were asked about injuries over the previous four weeks), these results may reflect the impact of injuries on safety attitudes behaviors, rather than the other way around (Stride et al., 2013). Taken together, these findings, which are unique in that they are based on a large sample of employed 15 to 25 year olds, provide support for researchers, policy makers, and managers considering the variety of safety experiences young people have in the workplace. References Association of Worker Compensation Boards (2012a). Retrieved on January 27, 2013 from: http://www.awcbc.org/common/assets/nwisptables/table 3-lti-age-jur.pdf Association of Worker Compensation Boards (2012b). Number of fatalities by age and jurisdiction. Retrieved on January 27, 2013 from: http://www.awcbc.org/common/ assets/nwisptables/table 24-fat-age-jur.pdf Barling, J., & Kelloway, K. E. (1999). Introduction. In J. Barling, & E. K. Kelloway (Eds.), Young workers: Varieties of experience (pp. 3–16). Washington, DC: American Psychological Association. Breslin, C. F., Day, D., Tompa, E., Irvin, E., Bhattacharyya, S., Clarke, J., et al. (2007). Non-agricultural work injuries among youth: A systematic review. American Journal of Preventative Medicine, 32, 151–162. Breslin, C. F., Polzer, J., MacEachen, E., Morrongiello, B., & Shannon, H. (2007). Workplace injury or “part of the job”?: Towards a gendered understanding of injuries and complaints among young workers. Social Science and Medicine, 64, 782–793. Breslin, C. F., & Smith, P. (2005). Age-related differences in work injuries: A multivariate, population-based study. American Journal of Industrial Medicine, 48, 50–56. Brotherton, M. (2012). Passport to safety: From Canada to Australia. Injury Prevention, 18(Suppl. 1), A35–A36. Bureau of Labour Statistics (2012a). Nonfatal occupational injuries and illnesses requiring days away from work 2011. Retrieved January 27, 2013 from: http://www.bls.gov/ news.release/archives/osh2_11082012.pdf Bureau of Labour Statistics (2012b). 2011 Census of Fatal Occupational Injuries. Retrieved January 27, 2013 from: http://www.bls.gov/iif/oshwc/cfoi/cftb0266.pdf
43
Clarke, S. (2013). Safety leadership: A meta-analytic review of transformational and transactional leadership styles as antecedents of safety behaviours. Journal of Occupational and Organizational Psychology, 86, 22–49. Griffin, M. A., & Neal, A. (2000). Perceptions of safety at work: A framework for linking safety climate to safety performance, knowledge, and motivation. Journal of Occupational Health Psychology, 5, 347–358. Heinrich, H. W., Petersen, D., & Ross, N. (1980). Industrial accident prevention (5th ed.). New York: McGraw-Hill. Hersch, J. (1998). Compensating differentials for gender-specific job injury risks. American Economic Review, 88, 598–607. Hu, L., & Bentler, P. M. (1998). Fit indices in covariance structure modeling: Sensitivity to underparameterized model misspecification. Psychological Methods, 3, 424–453. International Labour Organization (2013). Global employment trends for youth 2013: A generation at risk. Geneva: International Labour Organization. Loughlin, C., & Frone, M. R. (2004). Young workers' occupational safety. In J. Barling, & M. R. Frone (Eds.), Psychology of workplace safety (pp. 107–125). Washington, DC: American Psychological Association. Marshall, K. (2007, May). The busy lives of teens. Perspectives on Labour and Income, 8(5), 5–15. Mbaye, S., & Kouabenan, D. R. (2013). Effects of the feeling of invulnerability and the feeling of control on motivation to participate in experience-based analysis, by type of risk. Accident Analysis and Prevention, 51, 310–317. McCall, B. P., Horwitz, I. B., & Carr, B. S. (2007). Adolescent occupational injuries and workplace risks: An analysis of Oregon workers' compensation data 1990–1997. Journal of Adolescent Health, 41, 248–255. Meulman, J. J., Van Der Kooij, A. J., & Heiser, W. J. (2004). Principal components analysis with nonlinear optimal scaling transformations for ordinal and nominal data. In D. Kaplan (Ed.), The sage handbook of quantitative methodology for the social sciences (pp. 49–70). Thousand Oaks, CA: Sage. Muthén, L. K., & Muthén, B. O. (2011). Mplus (Version 6.11) [Computer software]. Los Angeles, CA: Muthén & Muthén. Podsakoff, P. M., MacKenzie, S. M., Lee, J., & Podsakoff, N. P. (2003). Common method variance in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879–903. Runyan, C. W., Lewko, J., & Rauscher, K. (2012). Setting an agenda for advancing young worker safety in the US and Canada. Public Health Reports, 127(3), 246–252. Runyan, C. W., Schulman, M., Santo, J. D., Bowling, J. M., & Agans, R. (2009). Attitudes and beliefs about adolescent work and workplace safety among parents of working adolescents. Journal of Adolescent Health, 44, 349–355. Salminen, S. (2004). Have young workers more injuries than older ones? An international literature review. Journal of Safety Research, 35, 513–521. Stride, C. B., Turner, N., Hershcovis, M. S., Reich, T. C., Clegg, C. W., & Murphy, P. (2013). Negative safety events as correlates of work-safety tension. Safety Science, 53, 45–50. Tucker, S., & Turner, N. (2011). Young worker safety behaviors: Development and validation of new measures. Accident Analysis & Prevention, 43(1), 165–175. Tucker, S., & Turner, N. (2013). Waiting for safety: Responses of young workers in Canada to unsafe work. Journal of Safety Research, 45, 103–110. Tucker, S., & Turner, N. (2014). Safety voice among young workers facing dangerous work: A policy-capturing approach. Safety Science, 62, 530–537. Tucker, S., Diekrager, D., Turner, N., & Kelloway, K. E. (2014). Work-related injury underreporting among young workers: Prevalence, gender differences, and explanations for underreporting. Journal of Safety Research, 50, 67–73. WorksafeBC (2011). Statistics for young workers. Retrieved March 27, 2013 from http:// www2.worksafebc.com/Topics/YoungWorker/Statistics.asp Zierold, K. M., Welsh, E. C., & McGeeney, T. J. (2012). Attitudes of teenagers towards workplace safety training. Journal of Community Health, 37, 1289–1295. Zohar, D. (2000). A group-level model of safety climate: Testing the effect of group climate on microaccidents in manufacturing jobs. Journal of Applied Psychology, 85, 587–596. Zumbo, B. D., Gadermann, A. M., & Zeisser, C. (2007). Ordinal versions of coefficients alpha and theta for Likert-rating scales. Journal of Modern Applied Statistical Methods, 6, 21–29. Dr. Nick Turner is a Professor in the Haskayne School of Business at the University of Calgary. His research interests include occupational health psychology, transformational leadership, and work design. Dr. Sean Tucker is an Associate Professor in the Faculty of Business Administration at the University of Regina. His research interests relate to young worker safety behavior, leadership, and labor relations. Dr. Kevin Kelloway is a Canada Research Chair in Occupational Health and Professor in the Department of Psychology at Saint Mary’s University. He researches leadership, safetyrelated transformational leadership, workplace stress, and the psychology of unionization.
Contents lists available at ScienceDirect
Journal of Safety Research journal homepage: www.elsevier.com/locate/jsr
Research note
Prevalence and demographic differences in microaccidents and safety behaviors among young workers in Canada☆ Nick Turner, a Sean Tucker, b,⁎ E. Kevin Kelloway c a b c
Haskayne School of Business, University of Calgary, Canada Faculty of Business Administration, University of Regina, Canada Department of Psychology, Saint Mary's University, Canada
a r t i c l e
i n f o
Article history: Received 16 July 2014 Received in revised form 7 January 2015 Accepted 9 March 2015 Available online 27 March 2015 Keywords: Injuries Safety voice Young workers Canada
a b s t r a c t Introduction: The present study examines the self-reported frequency of non-lost work time workplace injuries (“microaccidents”) and the frequency of three types of work-related safety behaviors (i.e., safety voice, safety compliance, and safety neglect) recalled over a four-week period. Method: We analyzed data on microaccidents and safety behaviors from 19,547 young workers (aged 15–25 years, Mdn = 18 years; 55% male) from multiple Canadian provinces. Results: Approximately one-third of all young workers recalled experiencing at least one microaccident at work in the last four weeks. Comparisons across three age groups revealed that younger workers, particularly between the ages of 15–18, reported more frequent microaccidents, less safety voice, less safety compliance, and more safety neglect than workers aged 19–22. This pattern of results also held for comparisons between workers in 19–22 and 23–25 age groups, except for safety voice which did not differ between these two older age groups. In terms of gender, males and females reported the same frequency of microaccidents, but males reported more safety voice, more safety compliance, and more safety neglect than females did. The results and limitations of the present study are discussed. Conclusion: Frequency of microaccidents and safety behavior vary among young worker age sub-groups. © 2015 National Safety Council and Elsevier Ltd. All rights reserved.
1. Introduction Injury statistics show that young workers (aged 15 to 25 years) are more likely to be injured on the job than adults (Salminen, 2004) with young males at highest risk of injury (Breslin & Smith, 2005). In 2011, 31,221 young workers in Canada (Association of Worker Compensation Boards, 2012a) and 116,900 in the United States (Bureau of Labour Statistics, 2012a) missed time away from work due to a workplace injury. That same year, 35 young workers in Canada (Association of Worker Compensation Boards, 2012b) and 380 in the United States (Bureau of Labour Statistics, 2012b) were killed on the job. In contrast to these statistics on severe work injury outcomes, the current paper presents the prevalence of non-lost time work injuries (also known as “microaccidents;” Zohar, 2000) and a range of safety behaviors (Tucker & Turner, 2011) among a large sample of young workers in Canada. Understanding the prevalence of less severe ☆ We presented earlier versions of this paper at the 2012 Congress of the Association of Workers Compensation Boards of Canada, Winnipeg, Canada, and the 30th International Congress of Psychology, Cape Town, South Africa. We thank Phil Germain, Sarah Goodhope, Alex Johnson, Paul Kells, Corey Pocaluyko, and Warren Preece for supporting this project. The first two authors contributed equally to this paper. No conflict of interest exists between the authors and publication of this research. ⁎ Corresponding author at: Faculty of Business Administration, University of Regina, SK, S4S 0A2, Canada. E-mail address: [email protected] (S. Tucker).
http://dx.doi.org/10.1016/j.jsr.2015.03.004 0022-4375/© 2015 National Safety Council and Elsevier Ltd. All rights reserved.
workplace injuries and safety outcomes that are upstream from severe workplace injuries is important for at least three reasons. First, more frequent but less severe forms of injuries are correlated with less frequent but more severe workplace injuries (Heinrich, Petersen, & Ross, 1980). Understanding predictors of more severe workplace injuries is critical in prevention efforts. Second, this study makes an empirical contribution by describing the occurrence of work-related safety outcomes among a large sample of young workers. This is a vulnerable population in the global workforce which, compared to jobs held by older workers, often experiences lower-quality work conditions because of the irregular, transient, low-wage, and part-time characteristics of many of the jobs they hold (International Labour Organization, 2013). Third, while some studies examine variation among young workers safety experiences (e.g., Breslin, Polzer, MacEachen, Morrongiello, & Shannon, 2007; Breslin & Smith, 2005), related research tends to assume the shared experience of work among this broad cohort (Barling & Kelloway, 1999). The current study describes the prevalence of safety behavior and microaccidents by young worker gender and age group (i.e., 15–18 years, 19–22 years, and 23–25 years) in a large sample of Canadian young workers. 2. Theoretical background Research on young worker safety is diverse and interdisciplinary (Runyan, Lewko, & Rauscher, 2012). Studies have identified situational
40
N. Turner et al. / Journal of Safety Research 53 (2015) 39–43
correlates of work-related injuries, such as work pressure (Breslin et al., 2007), gender differences in risk exposure (e.g., Breslin & Smith, 2005), the role of parents in developing children's safety attitudes (Runyan, Schulman, Santo, Bowling, & Agans, 2009), and the prevalence of work-related safety training among employed young people (Zierold, Welsh, & McGeeney, 2012). Absent from this literature is prevalence studies of work-related injuries and safety behaviors based on large samples of young working people. The current study examines two types of safety outcomes: microaccidents (Zohar, 2000) and safety behaviors (Tucker & Turner, 2011). Microaccidents are physical wounds to the body from workrelated activities that do not require time away from work (sometimes called “non-lost-time” injuries), and may require visits to a first aid station. This is in contrast to more severe injuries, variously called “major injuries” or “lost-time injuries” by health and safety regimes, and which necessitate time away from work for off-site medical attention, recuperation, or both. We studied non-lost time injuries for two reasons. First, little is known about non-lost time injury experiences in large samples of young workers; the majority of research on young worker injury prevalence examines lost-time injuries. Second, because microaccidents occur more frequently than lost-time injuries and consistent with Heinrich's iceberg concept (Heinrich et al., 1980), such events are a precursor to less-frequently occurring but more severe injuries in organizations. Studies of lost-time injury claims consistently find that young males between the ages of 19 and 25 experience the highest rate of occupational injuries (e.g., Breslin & Smith, 2005; McCall, Horwitz, & Carr, 2007; WorksafeBC, 2011). At the same time, however, Breslin et al.'s (2007) review of high-quality multivariate studies of predictors of young worker injuries concluded that neither gender nor young worker age was related to self-reported injuries, particularly in the presence of job and work characteristics. Overall, therefore, the findings are unclear with respect to the impact of young worker gender and age on workplace injuries. Workplace injuries in the studies included in Breslin et al.'s (2007) review covered the full spectrum of severity excluding workplace deaths by injury, and given the different frequency distributions between less severe and more severe work injuries described above, the current study focuses on microaccidents among young workers by gender and age. In addition to workplace injuries, two important and widely-studied safety outcomes are safety compliance (defined as “the core safety activities that need to be carried out by individuals to maintain workplace,” Griffin & Neal, 2000, p. 349) and safety participation (defined as voluntary extra-role behaviors that “help to develop an environment that supports safety,” Griffin & Neal, 2000, p. 349). A recent metaanalysis by Clarke (2013) containing 32 adult-aged working samples revealed that higher levels of safety compliance and safety participation were associated with fewer occupational injuries (average rs for both safety compliance and safety participation with occupational injuries = − .21, both with N = 229 reliability-corrected metaanalytic correlations). Recently, Tucker and Turner (2011) broadened this set of safety behaviors to reflect the work experiences of young workers in response to physically dangerous work. In addition to safety compliance, they measured safety neglect (e.g., taking shortcuts or work-arounds) as antithetical to safety compliance, as well as safety voice (e.g., speaking up about hazardous work) as a specific form of safety participation.1 The results of two focus group studies provide insight into the relative frequency of some of these behaviors and gender differences. Breslin et al. (2007) found that young female participants were more likely than young males in their sample to say they speak up about occupational safety concerns, but also that such behavior is commonly 1 Tucker and Turner (2011) also measured safety patience (i.e., waiting for safe conditions to improve), but the two-item measure was not measured in data set due to space restrictions in the questionnaire.
viewed as complaining by supervisors. In contrast, Tucker and Turner's (2013) focus group study (N = 39 participants) found no differences in the frequency of work-related safety voice between young males and young females. However, they noted that “speaking up” behavior (specifically change-oriented voice) is relatively uncommon due to concerns young workers expressed about fears of losing their jobs, supervisor indifference, their relatively short tenures, and felt powerlessness. Safety compliance has been examined with adult working populations, however safety neglect is a new construct and there are no data on its prevalence in either adult or young worker populations. The current study is motivated to extend emergent research on young worker safety behaviors and bring clarity to questions related to gender and age differences in these key safety behaviors. The current study uses a large sample of young Canadian workers to explore the prevalence of microaccidents, safety voice, safety compliance, and safety neglect. Given the mixed evidence of whether young males and young females differ in terms of the frequency of microaccidents and the lack of data on young worker age and gender across this range of safety behaviors, our study explores these potential differences as research questions. 3. Method 3.1. Study context, procedure, and sample Participants were employed Canadians aged 15 to 25 years who responded to a short on-line survey between September 2011 and July 2012. The voluntary survey, which was approved by each of the author's university research ethics board, appeared at the beginning of the Passport to Safety™ (PS) program. PS (http://www. passporttosafety.com/) is an on-line training and assessment tool that is meant to build awareness of workplace hazards, job rights and responsibilities, and work injury prevention approaches among students. It is widely used in Canada, as well as some jurisdictions in the United States and Australia, primarily in high schools as a part of occupational health and safety instruction in the curriculum (Brotherton, 2012). High school teachers who use the PS program in their classes provide each student with a unique personal identification number, and each student logs into the on-line system as either part of in-class instruction or a homework assignment. The short survey they were presented with before taking the PS program included questions about demographic characteristics, work-related injuries, workplace safety behaviors—all taking less than two minutes to complete. Prior to completing the survey, participants had the option to read the PS privacy policy, describing how their responses to the questions may be used in academic research. The data reported in the current study were collected in Canada, primarily from Ontario, the country's most populous province. Over 66,500 people started the short survey prior to beginning the on-line test. After eliminating cases with missing data, participants outside of the study's age range (i.e., over 26 years, inclusive) and unemployed individuals, a sample of 19,547 cases remained (55% male, median age = 18 years). Other data collected from some of these participants was used in a recently published study (Tucker et al., 2014). 3.2. Measures 3.2.1. Demographic information We asked respondents to report their year of birth and gender. 3.2.2. Microaccidents We asked respondents to respond to the following statement — “How many times in the past 4 weeks have you had a ‘minor’ workplace injury (e.g., cut, burn, strain, and sprain) that did not result in you missing time from work?” They responded in six groups ranging from no microaccidents experienced (coded as 0) to five or more microaccidents (coded as 5).
N. Turner et al. / Journal of Safety Research 53 (2015) 39–43
3.2.3. Safety behaviors We used the two highest-loading items from three of Turner and Tucker's (2011) young worker safety behaviors: safety voice (sample item: “Tell my supervisor about hazardous work”), safety compliance (sample item: “Wear protective clothing/equipment”), and safety neglect (sample item: “Take shortcuts that threaten my personal safety”). Respondents responded on a three-point scale: 1 = almost never, 2 = sometimes, 3 = almost always. Since the vast majority of test takers were high school students taking the PS test in a computer classroom, we truncated the response scale from seven to three points to simplify the appearance of the survey and mitigate any confusion.
3.3. Data analysis We used the two highest-loading items from each of Tucker and Turner's (2011) six-item safety voice scale, six-item safety neglect scale, and three-item safety compliance scales. As all items used had truncated three-point ordinal scaling (1 = almost never, 2 = sometimes, 3 = almost always) compared to the seven-point original (1 = almost never to 7 = almost always), we conducted a confirmatory factor analysis on the six items to assess the hypothesized 3-factor model using weighted least squares means and variance adjusted estimation in MPlus 6.11 (Muthén & Muthén, 2011) with listwise deletion. The hypothesized 3-factor model fit the data well, χ2 (6, N = 19,235) = 183.74, p b .001, comparative fit index (CFI) = .996, root mean squared error of approximation (RMSEA) = .039, based on Hu and Bentler's (1998) benchmarks for fit indices. The two items underlying each of the three factors indicated adequate internal consistency (ordinal alphas = .76, .63, and .69, for safety voice, safety compliance, and safety neglect, respectively; Meulman, Van Der Kooij, & Heiser, 2004), recognizing that benchmarks for reliability based on ordinal (vs. continuous) data are more conservative (Zumbo, Gadermann, & Zeisser, 2007). The three-factor model yielded better fit than (a) the two-factor model in which safety compliance formed one factor and safety voice and safety neglect combined to form a single factor (model fit comparisons appear in square brackets), CFI = .965, RMSEA = .103 [Δχ2 (2, N = 19,235) = 1,048.23, p b .001]; (b) the two-factor model in which safety voice formed one factor and safety compliance and safety neglect combined to form a single factor, CFI = .687, RMSEA = .309 [Δχ2 (2, N = 19,235) = 8,988.80, p b .001]; and (c) the two-factor model in which safety neglect formed one factor and safety compliance and safety voice combined to form a single factor, CFI = .682, RMSEA = .311 [Δχ2 (2, N = 19,235) = 9,375.51, p b .001]. As all these measures are self-report, we tested the goodness-of-fit of a one-factor model to test for mono-method bias (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003), which yielded a much weaker fit than the hypothesized three-factor model, CFI = .635, RMSEA = .315 [Δχ2 (3, N = 19,235) = 12,271.06, p b .001], suggesting that the self-reported nature of the scales did not mask construct differences. Taken together, the results of the confirmatory factor analysis, summarized in Table 1, provide support for the existence of three distinct safety behaviors replicating Tucker and Turner's (2011) factor structure for these constructs.
Table 1 Results of confirmatory factor analysis. Model
χ2
df
RMSEA
CFI
Model 1* Model 2 Model 3
183.74 1,231.97 9,172.54
19,229 19,233 19,233
.039 .103 .309
.996 .965 .687
Notes. * Hypothesized model. Model 1 has three factors (voice, neglect, and compliance). Model 2 has two factors (voice (with neglect), and compliance). Model 3 has one factor (voice, neglect, and compliance as a single factor). RMSEA = root-mean-square error of approximation. CFI = comparative fit index.
41
4. Results 4.1. Frequency of workplace injuries between age groups and by gender Slightly over two-thirds of young workers (66.5%) reported no microaccidents in the previous four weeks. Just over a quarter (25.5%) reported between one and two microaccidents in the previous four weeks, with the remaining 8% of young workers sampled reporting three or more microaccidents in the previous four weeks (3 injuries = 3.6%, 4 injuries = 1.5%, 5 or more injuries = 2.9%). Table 2 presents frequency of microaccidents for the overall sample, age groups, and by gender. We conducted a Kruskal–Wallis test to evaluate differences among three age groups (15–18 years, 19–22 years, 23–25 years) on the distribution of microaccidents, χ2 (2, N = 19,537) = 115.45, p b .001. We then conducted Mann–Whitney tests for pairwise differences among combinations of the three age groups, controlling for Type I error across tests with Bonferroni correction. Workers between the ages 15–18 reported more frequent microaccidents than workers between the ages 19–22 (Z = 6.09, p b .001), and workers aged 19–22 reported more frequent microaccidents than workers between ages 23–25 (Z = 6.52, p b .001). Overall, young men and young women reported no difference in frequency of microaccidents (Z = 1.53, p = .13). 4.2. Frequency of safety behaviors by age groups and gender Table 2 presents frequency of the three types of safety behaviors for the overall sample by gender and age group. We conducted a Kruskal– Wallis test to evaluate differences among the three age groups on safety voice, χ2 (2, N = 19,537) = 11.93, p b .01; safety compliance, χ2 (2, N = 19,537) = 67.00, p b .001; and safety neglect scores, χ2 (2, N = 19,537) = 87.29, p b .001. We then conducted Mann–Whitney tests for pairwise differences among the three age groups, controlling for Type I error across tests with Bonferroni correction. In terms of safety voice, workers aged 15–18 reported less safety voice than workers aged 19–22 (Z = − 3.34, p b .01), but safety voice scores in the 19–22 age group did not differ from the scores in the 23–25 age group (Z = − .35, p = .72). In terms of safety compliance, workers aged 15–18 reported less safety compliance than workers aged 19–22 (Z = − 6.62, p = .02), and workers in the 19–22 age group reported less safety compliance than workers in the 23–25 age group (Z = − 2.28, p = .02). In terms of safety neglect, workers aged 15–18 reported more safety neglect than workers aged 19–22 (Z = 4.79, p b .001), and safety neglect scores in the 19–22 age group were higher than safety neglect scores in the 23–25 age group (Z = 6.20, p b .001). Overall, young men reported more safety voice (Z = 2.80, p b .01), more safety compliance (Z = 2.14, p b .05), and substantially more safety neglect (Z = 14.13, p b .001) than young women did. 4.3. Relationships among microaccidents and safety behaviors Table 3 presents the zero-order non-parametric correlations and partial non-parametric correlations (i.e., correlations controlling for age group and gender) among the study variables. Frequency of microaccidents was correlated weakly albeit negatively with safety voice (r = − .03, p b .01) and safety compliance (r = − .05, p b .01), and weakly albeit positively with safety neglect (r = .05, p b .01). Controlling for age group and gender, the pattern and magnitude of associations did not change (rs = −.02, −.03, and .03, ps b .01, respectively). 5. Discussion 5.1. Summary of findings and their implications This is the first study to investigate the prevalence of both microaccidents and a range of safety behaviors in a large sample of young
42
N. Turner et al. / Journal of Safety Research 53 (2015) 39–43
Table 2 Comparison by sex and age.
Overall Sex Male Female Age group 15–18 19–22 23–25
N (listwise deletion)
Mean injuries
95% CI
Mean safety voice
95% CI
Mean safety compliance
95% CI
Mean safety neglect
95% CI
19,547
.65
.63, .67
2.14
2.13, 2.15
2.43
2.42, 2.44
1.30
1.29, 1.31
10,716 8,760
.68 .62
.66, .71 .59, .64
2.16 2.12
2.14, 2.17 2.11, 2.14
2.45 2.41
2.44, 2.46 2.40, 2.43
1.34 1.25
1.33, 1.35 1.24, 1.26
15,083 3,820 664
.68 .58 .30
.67, .70 .55, .62 .24, .36
2.13 2.18 2.15
2.12, 2.14 2.15, 2.20 2.10, 2.21
2.41 2.49 2.53
2.40, 2.42 2.47, 2.51 2.48, 2.58
1.31 1.28 1.15
1.30, 1.32 1.26, 1.29 1.12, 1.18
workers aged 15 to 25 years. The purpose of the study was to assess potential differences in these outcomes by age and gender. The youngest participants (15 to 18 years old) reported more microaccidents than the oldest participants (19 to 25 years old). Further, there was no gender difference in frequency of microaccidents across the age groups. While lost-time injury claim data shows that males aged 20–24 experience the highest rate of injuries (e.g., Breslin & Smith, 2005), Breslin et al.'s (2007) review found that among young worker samples there was no evidence that either age or gender related to self-reported injuries after controlling for job or hazard factors. The current findings on microaccidents are generally consistent with the conclusions of Breslin et al.'s review, except for higher level of microaccidents in the 15- to 18-year old age group. It is important to note that the current study examined over a four-week period frequency of microaccidents – low severity, non-lost-work time injuries in a short, easy-to-remember time period \ whereas the Breslin et al. review combined the full range of injury severity (ranging from “strain/sprains/cuts/lacerations, burns, bruises/contusions, fractured bone, dislocated joint, serious muscle/back pain, blisters” through to “the loss of at least one shift, received hospital treatment, received sutures to a wound,” but excluding fatalities) and across an indeterminate period of time (ranging from “occurrence of injury during summer job” to “ever injured while working”). Furthermore, the findings from the current study are surprising given that, in general, the oldest participants would tend to work more hours than younger participants (Marshall, 2007); thus, in terms of exposure, all else equal, older participants should experience more microaccidents. Controlling for time worked in the last four weeks would have helped to mitigate concern about differences in exposure to possible work hazards. Not only did the youngest survey respondents report experiencing more minor injuries, but also they reported the lowest frequency of both safety voice and safety compliance as well as the highest frequency of safety neglect. These findings are a concern because young employees working their first jobs are vulnerable on a number of fronts, and it is hoped that exposure to unsafe working conditions and direct or vicarious experience of workplace injuries would encourage speaking up about safety concerns in an effort to try and improve workplace safety. This is not always the case, however. Tucker and Turner (2013, 2014) have shown that teenaged workers are more likely to ‘wait and see’ whether unsafe work conditions improve, and that speaking up about Table 3 Spearman correlations among study variables (N = 19,479–19,547).
1. 2. 3. 4 5. 6.
Variable
1.
2.
3.
4.
5.
6.
Age group Gender Safety voice Safety compliance Safety neglect Microaccidents
– −.01 .02 .06 −.06 −.07
– .02 .02 .10 .01
– .48 .03 −.03
.49 – −.08 −.05
.10 .01 – .05
−.02 −.03 .03 –
Note. Zero-order correlations appear below the diagonal. Partial correlations, controlling for age group and gender, appear above the diagonal. Correlations equal or greater than |.02| are significant at the .01 level. Correlations greater than |.01| are significant at the .05 level. Age group: 1 = 15–18 years old; 2 = 19–22 years old; 3 = 23–25 years old. Gender: 1 = male, 0 = female.
safety concerns, while possible, is rare and occurs when unsafe conditions or their consequences are extreme and young workers feel psychologically safe in doing so. Interestingly, in the current study, male participants reported higher safety voice than female participants, which contradicts previous findings using hypothetical experimental scenarios (Tucker & Turner, 2014) and focus groups (Breslin et al., 2007). One possibility is that gender is a proxy for job risk. That is, young males self-select or are assigned work that is objectively more physically risky than the work assigned to female young workers. Past research suggests that adult women compared to adult men face less risk of physical injury at work (Hersch, 1998). Young males may report higher safety voice and higher safety compliance than young females because young males may be assigned work, conduct work, or encounter work situations in which likelihood of physical injury is higher. Thus, by necessity, young males may be more likely to speak up about hazardous work conditions or comply with safety rules to keep safe. Future research is needed to determine whether the differences across studies are due to differences in male and female representation in occupations and work sectors (e.g., service sector vs. construction) and their related risk factors, or methodological factors (e.g., laboratory-based vs. field-based research). Meta-analytic evidence has shown that safety compliance and safety participation are negatively associated with workplace injuries (Clarke, 2013), and this is what we found with safety compliance and safety voice in this study albeit with small effect sizes (absolute value of partial rs ranging from .02 to .03). Furthermore, safety neglect was related to greater frequency of microaccidents, and younger males particularly those in the 15- to 18-year old age group reporting higher safety neglect than young females did. We can only speculate as to why safety neglect may be higher among young males. First, in line with the argument above about males doing different types of work than females, it is possible that young males work in physically riskier conditions, and thus in addition to having more reason to speak up and comply with safety rules, there are also more safety rules to disregard. Second, employees of all ages, but particularly young workers with relatively little experience, may have distorted risk perceptions, resulting in underestimating or overestimating risks (Mbaye & Kouabenan, 2013). More specifically, young people may view themselves as impervious to harm (e.g., Loughlin & Frone, 2004), which helps to explain why young workers often disregard company policies, take shortcuts, and ignore hazards. Relatedly, there is the question of how to reconcile the result showing that males concurrently reported the most frequent safety neglect and safety compliance behaviors — distinct behaviors that one would not expect to be simultaneously higher. However, we note that in absolute terms, males report substantially more frequent safety compliance than safety neglect behavior (Ms = 2.45 vs. 1.35, respectively), and that the two variables are negatively correlated when the sample is limited to male participants (r = −.03, p b .001). 5.2. Study limitations and conclusion Despite the unique characteristics of the current sample, this study has several limitations worth noting. First, the distribution of the age
N. Turner et al. / Journal of Safety Research 53 (2015) 39–43
of the sample is skewed and selection effects are possible. Nearly all participants aged 15 to 18 years complete PS in a high school classroom, while those 19 years and above (i.e., having graduated from high school or were early school-leavers) primarily participating through their employer. It is possible that taking the PS test in the context of a class on workplace health and safety may encourage participants to respond in socially-desirable ways – for example, under-reporting microaccidents or over-reporting safety compliance – and that employers that incorporate PS into job-related training are safer workplaces than employers who do not offer PS training to their employees. This does not explain why the younger age groups reported more microaccidents and less safety compliance; however, it may explain why the older age groups reported fewer microaccidents and exercised safety compliance more frequently. Second, we only collected data on microaccidents (and not lost-time workplace injuries) and, for the safety behaviors, we used shortened scales limited to two items providing participants with three response possibilities. For microaccidents, it is possible that we missed important differences in age or gender based on the sole measurement of low severity injuries. For example, it is possible that young males report more lost-time work injuries than young females report, but we are unable to ascertain this, given that we have only collected non-lost-time injuries (microaccidents). For the measurement of safety behaviors, we necessarily restricted the range of both the focal constructs (only use two items per scale) and their variance (only using three-point response scales instead of the original seven-point response scales), resulting in lower overall scale reliability. However, we used statistical tests that are appropriate for analyzing truncated ordinal scaling. Relatedly, with the exception of the correlation between safety compliance and safety voice (r = .48), the relationships among the safety behaviors were small and likely have little practical significance. Finally, due to the retrospective nature of our injury data (i.e., participants were asked about injuries over the previous four weeks), these results may reflect the impact of injuries on safety attitudes behaviors, rather than the other way around (Stride et al., 2013). Taken together, these findings, which are unique in that they are based on a large sample of employed 15 to 25 year olds, provide support for researchers, policy makers, and managers considering the variety of safety experiences young people have in the workplace. References Association of Worker Compensation Boards (2012a). Retrieved on January 27, 2013 from: http://www.awcbc.org/common/assets/nwisptables/table 3-lti-age-jur.pdf Association of Worker Compensation Boards (2012b). Number of fatalities by age and jurisdiction. Retrieved on January 27, 2013 from: http://www.awcbc.org/common/ assets/nwisptables/table 24-fat-age-jur.pdf Barling, J., & Kelloway, K. E. (1999). Introduction. In J. Barling, & E. K. Kelloway (Eds.), Young workers: Varieties of experience (pp. 3–16). Washington, DC: American Psychological Association. Breslin, C. F., Day, D., Tompa, E., Irvin, E., Bhattacharyya, S., Clarke, J., et al. (2007). Non-agricultural work injuries among youth: A systematic review. American Journal of Preventative Medicine, 32, 151–162. Breslin, C. F., Polzer, J., MacEachen, E., Morrongiello, B., & Shannon, H. (2007). Workplace injury or “part of the job”?: Towards a gendered understanding of injuries and complaints among young workers. Social Science and Medicine, 64, 782–793. Breslin, C. F., & Smith, P. (2005). Age-related differences in work injuries: A multivariate, population-based study. American Journal of Industrial Medicine, 48, 50–56. Brotherton, M. (2012). Passport to safety: From Canada to Australia. Injury Prevention, 18(Suppl. 1), A35–A36. Bureau of Labour Statistics (2012a). Nonfatal occupational injuries and illnesses requiring days away from work 2011. Retrieved January 27, 2013 from: http://www.bls.gov/ news.release/archives/osh2_11082012.pdf Bureau of Labour Statistics (2012b). 2011 Census of Fatal Occupational Injuries. Retrieved January 27, 2013 from: http://www.bls.gov/iif/oshwc/cfoi/cftb0266.pdf
43
Clarke, S. (2013). Safety leadership: A meta-analytic review of transformational and transactional leadership styles as antecedents of safety behaviours. Journal of Occupational and Organizational Psychology, 86, 22–49. Griffin, M. A., & Neal, A. (2000). Perceptions of safety at work: A framework for linking safety climate to safety performance, knowledge, and motivation. Journal of Occupational Health Psychology, 5, 347–358. Heinrich, H. W., Petersen, D., & Ross, N. (1980). Industrial accident prevention (5th ed.). New York: McGraw-Hill. Hersch, J. (1998). Compensating differentials for gender-specific job injury risks. American Economic Review, 88, 598–607. Hu, L., & Bentler, P. M. (1998). Fit indices in covariance structure modeling: Sensitivity to underparameterized model misspecification. Psychological Methods, 3, 424–453. International Labour Organization (2013). Global employment trends for youth 2013: A generation at risk. Geneva: International Labour Organization. Loughlin, C., & Frone, M. R. (2004). Young workers' occupational safety. In J. Barling, & M. R. Frone (Eds.), Psychology of workplace safety (pp. 107–125). Washington, DC: American Psychological Association. Marshall, K. (2007, May). The busy lives of teens. Perspectives on Labour and Income, 8(5), 5–15. Mbaye, S., & Kouabenan, D. R. (2013). Effects of the feeling of invulnerability and the feeling of control on motivation to participate in experience-based analysis, by type of risk. Accident Analysis and Prevention, 51, 310–317. McCall, B. P., Horwitz, I. B., & Carr, B. S. (2007). Adolescent occupational injuries and workplace risks: An analysis of Oregon workers' compensation data 1990–1997. Journal of Adolescent Health, 41, 248–255. Meulman, J. J., Van Der Kooij, A. J., & Heiser, W. J. (2004). Principal components analysis with nonlinear optimal scaling transformations for ordinal and nominal data. In D. Kaplan (Ed.), The sage handbook of quantitative methodology for the social sciences (pp. 49–70). Thousand Oaks, CA: Sage. Muthén, L. K., & Muthén, B. O. (2011). Mplus (Version 6.11) [Computer software]. Los Angeles, CA: Muthén & Muthén. Podsakoff, P. M., MacKenzie, S. M., Lee, J., & Podsakoff, N. P. (2003). Common method variance in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879–903. Runyan, C. W., Lewko, J., & Rauscher, K. (2012). Setting an agenda for advancing young worker safety in the US and Canada. Public Health Reports, 127(3), 246–252. Runyan, C. W., Schulman, M., Santo, J. D., Bowling, J. M., & Agans, R. (2009). Attitudes and beliefs about adolescent work and workplace safety among parents of working adolescents. Journal of Adolescent Health, 44, 349–355. Salminen, S. (2004). Have young workers more injuries than older ones? An international literature review. Journal of Safety Research, 35, 513–521. Stride, C. B., Turner, N., Hershcovis, M. S., Reich, T. C., Clegg, C. W., & Murphy, P. (2013). Negative safety events as correlates of work-safety tension. Safety Science, 53, 45–50. Tucker, S., & Turner, N. (2011). Young worker safety behaviors: Development and validation of new measures. Accident Analysis & Prevention, 43(1), 165–175. Tucker, S., & Turner, N. (2013). Waiting for safety: Responses of young workers in Canada to unsafe work. Journal of Safety Research, 45, 103–110. Tucker, S., & Turner, N. (2014). Safety voice among young workers facing dangerous work: A policy-capturing approach. Safety Science, 62, 530–537. Tucker, S., Diekrager, D., Turner, N., & Kelloway, K. E. (2014). Work-related injury underreporting among young workers: Prevalence, gender differences, and explanations for underreporting. Journal of Safety Research, 50, 67–73. WorksafeBC (2011). Statistics for young workers. Retrieved March 27, 2013 from http:// www2.worksafebc.com/Topics/YoungWorker/Statistics.asp Zierold, K. M., Welsh, E. C., & McGeeney, T. J. (2012). Attitudes of teenagers towards workplace safety training. Journal of Community Health, 37, 1289–1295. Zohar, D. (2000). A group-level model of safety climate: Testing the effect of group climate on microaccidents in manufacturing jobs. Journal of Applied Psychology, 85, 587–596. Zumbo, B. D., Gadermann, A. M., & Zeisser, C. (2007). Ordinal versions of coefficients alpha and theta for Likert-rating scales. Journal of Modern Applied Statistical Methods, 6, 21–29. Dr. Nick Turner is a Professor in the Haskayne School of Business at the University of Calgary. His research interests include occupational health psychology, transformational leadership, and work design. Dr. Sean Tucker is an Associate Professor in the Faculty of Business Administration at the University of Regina. His research interests relate to young worker safety behavior, leadership, and labor relations. Dr. Kevin Kelloway is a Canada Research Chair in Occupational Health and Professor in the Department of Psychology at Saint Mary’s University. He researches leadership, safetyrelated transformational leadership, workplace stress, and the psychology of unionization.
