Occupational Safety and Health Practices and Performance of Small and Medium-Size Enterprises in Kenya

Occupational Safety and Health Practices and Performance of Small and Medium-Size Enterprises in Kenya

Florence K. Guantai^*., Hezron M. Osano., Tabitha G. Murerwa., Phares O. Ochola

Department of Business and Management Studies, Technical University of Kenya

DOI: https://dx.doi.org/10.47772/IJRISS.2025.909000193

Received: 20 September 2025; Accepted: 29 September 2025; Published: 04 October 2025

ABSTRACT

This study investigated the influence of occupational safety and health (OSH) practices on the performance of SMEs in Kenya. The objectives were to:  examine the effect of employee wellness on SME performance, and analyze the relationship between the working environment and SME performance .To test the hypothesis, a descriptive and inferential research design was adopted, with a sample of 294 organizations selected from a population of 1,118 SMEs through stratified random sampling. In each SME, the manager or human resource officer was identified as respondent. Data were analyzed using an ordinal logistic regression model. Findings revealed that both employee wellness and working environment, had a positive and statistically significant effect on SME performance, with the working environment exerting the greatest influence. The study concludes that SMEs should adopt comprehensive OSH practices by promoting employee wellness and providing conducive working environment to reduce accidents thereby enhancing performance. Further research is recommended to compare OSH adoption between SMEs and large enterprises.

Keywords: Occupational Safety and Health Practices, Performance, Small and Medium-Size Enterprises, Risks management

INTRODUCTION

Empirical studies and global agencies such as International Labor Organization (ILO) and World Health Organization (WHO) contend that safety, health and welfare of employees, who comprise of approximately half of the global population, is fundamental to workers and their relations, as well as to the sustainability, productivity and competitiveness of an enterprise hence, to the growth of the national and global economy (DeLoatch, 2020). It is estimated that every year, the world loses $ 1.25 trillion, approximately 4 percent of GDP as a result of occupational accidents and sicknesses (Ayub et al., 2012). According to Dugolli (2021), poor safety and health consequently adversely affect organizations as they are forced to cater for compensation covers in case of accidents, losses due to poor production, and legal fees as well as deal with diminished morale. Nzuve et al., (2012) argue that utmost OSH standard is vital in eliminating safety and health hazards and risks.

The World Health Organization (WHO, 2021) defines Occupational Safety and Health (OSH) as the discipline concerned with anticipating, identifying, assessing, and preventing workplace hazards or exposures that may cause harm to employees, while also considering the potential effects on the environment and surrounding communities. Traditionally, OSH emphasized the physical work environment, focusing mainly on physical, chemical, biological, and ergonomic risks. However, the modern perspective integrates additional dimensions such as psychosocial factors, organizational culture, and workplace practices, as well as their broader linkages to public health, all of which can significantly affect employee well-being (ILO, 2021).

Murithi (2019) argue that whether businesses are formalized or not, it is important that governments    guarantee safety of all employees mainly through sensitization and trainings (Murithi 2021) To promote wellness, organizations implement workplace interventions such as awareness programs, health education, counseling, and the creation of supportive environments. Examples include fitness facilities, health coaching, non-smoking policies, employee assistance programs, and nutritious cafeteria options (Julander, 2014). Effective communication and information systems also play a pivotal role in improving safety outcomes and strengthening employees’ perception of management’s OSH commitment (Gyekye et al., 2012). A comprehensive safety management system encompassing policies, procedures, and strategies further enhances organizational reliability and coordination (Fernández-Muniz et al., 2009).

The overreaching goal of OSH management is to minimize risks and protect employees from exposure to harmful substances, unsafe practices, and workplace accidents. Gbadago et al. (2017) argue that protecting employees’ safety requires robust hazard prevention and management systems that address four broad categories of hazards: physical (e.g., noise, poor lighting, high temperatures), biological (e.g., microorganisms, infections, waste), chemical (e.g., fumes, gases, dust), and psychosocial (e.g., stress, harassment, workplace violence). Increasingly, psychosocial risks are recognized as a major challenge to modern OSH and must be managed effectively (Gehad et al., 2021). Such hazards directly impact employees’ mental health and their capacity to cope with stressful workplace dynamics (Chirico et al., 2019; Ruiz et al., 2021).

Statement of the Problem

• In most developing countries, SMEs are the backbone to the economic growth and development of those countries (Varianou-Mikellidou et al., 2019). In the European Union, SMEs represent 99 percent of all enterprises and provide about 75 million jobs regionally (Colvin, 2013). In Kenya, they provide about 80% of employment opportunities (Bowen, Morara, & Mureithi, 2009).
• Despite the critical role SMEs play in the economy, they unfortunately report a 90 to 95% failure rate due to a number of social impediments and poor implementation of policy (Kersten et al., 2017). According to an economic survey (2018), Kenya witnessed a drop of its GDP in the manufacturing sector from 9.1 percent in 2016 to 8.4 percent in 2017. The decline in performance was partly due to increased number of occupational accidents (KNBS, 2017). Studies indicate that OSH issues considerably impact negatively on SME performance due to limited financial and technical capacity to implement effective safety measures (Arocena & Nüñez, 2019; Okpala, 2020). Inadequate OSH implementation leads to poor employee performance, higher turnover rates, and reduced business continuity (Bello & Oyekunle, 2021; Patel et al., 2023).

The ILO, (2017) survey estimates that approximately 317 million people experience occupational accidents, out of which 6,300 are fatal. Additionally, approximately 160 million cases of occupational diseases reported lead to loss of employee productivity, absence from duty and a rise in employee work injury compensation costs and medical expenses employers (Singh et al., 2019; Murithi, 2021). These challenges highlight that OSH has been ignored and consequently has turned out to be an impediment to achieving sustainable development for SMEs (Khan et al., 2019; Yuen et al., 2021). In addition, some empirical studies indicate that workplace injuries and illnesses reduce overall firm efficiency, leading to decreased employee morale and increased operational costs (Shikdar & Al-Hadhrami, 2020; Oluwaseun et al., 2022). To perform optimally, employers must ensure employees are fit physically, mentally, and emotionally. Ensuring OSH of workers   significantly impact on organization’s sustainability, competitiveness, and productivity (Osei Boakye et al., 2021).

Several empirical studies have not thoroughly examined the connection between SMEs’ performance and OSH practices, which has resulted in a failure to address important gaps in the relationship. Although Harnois and Gabriel (2018) found employee satisfaction and well-being significant in improving performance, their study did not include the working environment and ergonomics as independent variables, which this study aimed to explore. Another related study by Gunyomi and Bruning (2016) conducted in Nigeria found human capital development and occupational health and safety closely correlated to performance, but it omitted other independent variables this study intends to explore. Moreover, Ombasyi (2019) study in Kenya points substantial association between mental wellness and performance but focused solely on the psychological well-being of employees, omitting other independent variables explored in the current study. Similarly, recent studies have emphasized the importance of OSH compliance but have not explored the moderating effect of government policy on the relationship between OSH and Performance of SMEs or its influence on SMEs long-term sustainability (Aminu & Ezekiel, 2020; Mugambi et al., 2022).

The purpose of this study was therefore to analyze the influence of occupational safety and health practices on the performance of small and medium-size enterprises in Kenya. The study placed emphasis on underpinning aspects of government policy as a moderating variable. By addressing these gaps, this study would enhance OSH awareness among organizations and recommend the best OSH practices to boost performance in SMEs.

Objectives of the Study

1. To investigate the influence of Employee Wellness on performance of small and medium size enterprises in Kenya.
2. To analyze the influence of the working environment on performance of small and medium- size enterprises in Kenya.

Research Questions

1. Does Employee Wellness influence performance of small and medium size enterprises in Kenya.
2. Does working environment influence performance of small and medium- size enterprises in Kenya.

Research Hypotheses

1. H Employee Wellness has no statistically significant relationship with performance of small and medium-size enterprises in Kenya.
2. H_02. Working environment has no statistically significant relationship with performance of small and medium- size enterprises in Kenya.

LITERATURE REVIEW

Theoretical Review

The research was anchored on   Behavior Reasoning Theory and Heinrich’s Domino Theory. The theories were identified specifically to give a deeper understanding of occupational safety and health practices and organizational performance.

Behavioral Reasoning Theory (BRT)

In contemporary times various social science and management disciplines have increasingly focused on understanding behavioral dimensions (Sahu et al., 2020). Historically, behavioral theories were widely applied in consumer behavior studies to analyze the factors influencing user intentions and conduct. According to Westaby (2005), many modern behavioral theories are rooted in two foundational models: the Theory of Reasoned Action (TRA) and the Theory of Planned Behavior (TPB). TRA posits two categories of beliefs—behavioral beliefs and normative beliefs—that shape attitudes, subjective norms, intentions, and ultimately behavior. TPB, an extension of TRA, adds a third category, control beliefs, introducing the construct of perceived behavioral control alongside attitudes and subjective norms. Together, these factors shape intention, which subsequently predicts behavior (Ajzen, 1991).

The TPB has been applied in occupational safety and health (OSH) interventions, such as studies involving healthcare workers in the United Kingdom, where it was used to examine the frequency of safe manual patient-handling practices (Rickett et al., 2006; Guerin & Sleet, 2020). TPB posits that attitudes, subjective norms, and perceived behavioral control directly influence an individual’s intention to perform a behavior. Within the workplace context, TPB highlights that occupational illnesses and injuries have long-term physical, emotional, and economic consequences for employees, families, and communities. Thus, interventions that reduce disease and injury risks must incorporate behavior change strategies at individual, organizational, community, and societal levels (Sahu, Padhy, & Dhir, 2020).

Despite their contributions, TPB and TRA have been criticized for limitations in predicting and generalizing behavioral decision-making, which led to the development of Behavioral Reasoning Theory (BRT) (Gilal et al., 2019; Hagger et al., 2002). BRT extends traditional behavioral intention models such as TRA and TPB (Ajzen, 1991; Haktanir, 2019) and emerged as a promising theoretical advancement in marketing research, later finding relevance in other fields, including occupational safety and health (Sundtoft, 2018; Guerin & Sleet, 2020).

A key distinction of BRT is its introduction of the construct of reasons (for and against), which enriches understanding of decision-making mechanisms (Sahu et al., 2020). BRT has been applied in diverse contexts, including green purchasing (Sreen et al., 2023), electronic waste management (Dhir et al., 2021), and organic food procurement (Tandon et al., 2020). However, this research remains dispersed across disciplines such as psychology, sociology, marketing, and organizational behavior, and scholars have recommended further exploration of BRT in varied contexts (Sahu et al., 2020). In this regard, the present study applies BRT to explain the relationship between OSH practices and the performance of SMEs in Kenya.

Similar to TPB, BRT assumes that intentions directly influence actions, and that attitudes, perceived control, and subjective norms strongly shape intentions, as illustrated in Figure 2.1. Behavioral theories offer valuable tools for developing and evaluating interventions designed to enhance employee safety and health. In OSH, BRT has proven useful in designing prevention programs and identifying multilevel factors that either promote or hinder workplace safety (Sundtoft, 2018; Guerin & Sleet, 2020).

However, like TPB and TRA, BRT assumes that individuals can protect themselves even in hazardous environments, a premise that has been critiqued for contributing to a “blaming the victim” perspective. Behavioral science research has emphasized that individual behaviors, while critical in preventing illness and injury, must complement mechanical and environmental safety measures (Sahu, Padhy, & Dhir, 2020). Claudy et al. (2013) refined BRT by replacing the concept of global motives with a more nuanced understanding of attitudes, which enhanced its utility for empirical research.

Despite calls for broader application, relatively few studies have anchored OSH research in behavioral theories (Rebecca & Sleet, 2020). Regulatory frameworks, such as those in the United States, increasingly support the application of BRT, recognizing that organizational and group-level factors strongly shape individual attitudes and behaviors toward safety, which in turn affect accident rates. OSH measures, however, have traditionally emphasized the physical environment, thereby creating tension between “active” behavioral strategies and “passive” mechanical

Heinrich’s Domino Theory

The second theory underpinning this study is the Domino Theory, advanced by Heinrich (1931). The theory conceptualizes accidents through the analogy of dominos toppling one after another, thereby creating a sequence of events. However, removing a prime cause prevents the commencement of this chain reaction (Herskovitz & DeCamp, 2019). Heinrich argues that all accidents are directly linked to deficiencies in safety programs such as poor human judgment, unsafe working conditions, and insufficient safety training, among other factors. The theory identifies two main factors that contribute to unsafe workplaces: work-related factors like job overload, wear and tear, poor-quality tools, and defective work design or maintenance; and individual faults, such as negligence, anxiety, uneasiness, egocentrism, and ignorance of safety practices. Heinrich asserts that a person’s social environment that is, the circumstances in which they are brought up and socialized can either inherit or acquire character flaws (Herskovitz & DeCamp, 2019).

Heinrich elaborates on his theory through four stages. The first stage, social environment and ancestry, addresses personality traits that may lead to undesirable actions, such as obstinacy, greed, or irresponsibility. These traits may be transmitted through heredity or shaped by the social environment. Both aspects commonly referred to as the nature versus nurture debate contribute to personal faults and form the foundation of accident causation. This stage aligns with the assumptions of social learning theories (Wang & Yan, 2019).

The second stage, faults of a person, refers to personality characteristics or behavioral tendencies that predispose individuals to accidents. For example, a short temper may result in impulsive actions or disregard for safety protocols. Similarly, recklessness or ignorance of safety standards can increase vulnerability to injury. Wang and Yan (2019) found a significant correlation between such character flaws including temper, inconsiderateness, ignorance, and carelessness and accident causation.

The third stage, unsafe acts or conditions, represents the most immediate and visible precursor to an accident. Unlike the first two stages, this stage is directly connected to the accident event. It may involve risky actions such as operating machinery without proper caution, ignoring guardrails, or misusing personal protective equipment. These omissions or commissions are often the proximate causes of accidents. Consequently, unsafe acts and conditions are regarded as the most practical points for intervention, akin to removing one domino from the sequence (Alli, 2019).

The fourth stage is the accident itself, which occurs when an unintended and undesirable event takes place. This stage is followed by the final stage, injury, which reflects the adverse consequences of the accident. Accidents typically arise from the convergence of multiple triggers, producing a cascading chain of events (Dugolli, 2021). However, whether an accident results in injury often depends on probability; not all accidents lead to harm. Similarly, the negative traits highlighted in stage two may not manifest in supportive environments, indicating that context can mitigate risks. From a policy perspective, the most effective approach is to remove at least one domino from the chain, thereby fostering a safety-oriented subculture and promoting accident prevention. Such an environment not only reduces hazards but also enhances organizational performance (Sabet et al., 2021).

Despite its continuing relevance, the Domino Theory has been critiqued. Scholars argue that it places disproportionate emphasis on the causes of accidents and injuries while overlooking the broader benefits of maintaining safe workplaces (Reese, 2018; Bayram, 2022). Furthermore, critics contend that the model’s complexity calls for greater clarity on how safety measures directly influence employee and organizational outcomes (Sabet et al., 2021). Nonetheless, the theory remains valuable for contemporary research. Implementing protective measures can reduce hazard exposure by interrupting the accident sequence at earlier stages. The greater the number of preventive measures in place, the more effectively organizations can eliminate risks, enabling employees to perform productively without persistent concerns about safety (Mutegi et al., 2023). Reduced accident incidence also fosters positive employee attitudes, greater job satisfaction, and ultimately improved productivity (Aswathappa, 2015). Thus, the Domino Theory remains highly relevant in explaining workplace accidents and guiding safety practices in modern organizational contexts. The theory underscores the importance of implementing wellness programs, such as safety training, to equip employees with the knowledge, skills, and attitudes necessary to adhere to safety procedures and prevent workplace accidents. Counseling programs can also play a critical role in correcting unsafe behaviors, thereby reducing the likelihood of accidents. Furthermore, the provision of personal protective equipment (PPE) serves as a preventive measure against unsafe acts or hazardous conditions. In addition, the theory emphasizes the need to maintain a conducive working environment through effective occupational safety and health (OSH) hazard management systems, which minimize accidents and, in turn, enhance organizational performance. A key component of this process is the identification of potential safety and health hazards including physical, biological, chemical, and psychosocial risks which constitutes the first step toward accident prevention.

Empirical Review

Employee Wellness and Performance

A number of scholars on employee wellness agree that wellness entails a sense of contentment and positive attitudes toward one’s job (Kyriazoglou, 2015; Dolyle, L’Estrange, & Bauman, 2016; Ryan & Deci, 2001; Ryff & Keyes, 1995). Keeman et al. (2017) assert that wellness refers to the quality or state of being healthy not only in the body but also in the mind, particularly when achieved through deliberate efforts to promote employee well-being. Workplace health program activities are often undertaken at the organizational level to enhance employee productivity by improving health. These programs may emphasize individual behavioral change, consistent with the Behavior Reasoning Theory (BRT), or target organizational policies and environments that support healthy behaviors (Philips & Gully, 2014).

Empirical research has provided diverse perspectives on the influence of employee wellness on productivity. As a crucial component of workers’ social well-being, Isidro (2024) investigated the role that organisations play in wellness initiatives. The study examined the mediating functions of commitment and job satisfaction as well as the effects of wellness initiatives and perceived organisational support on organisational performance. Snowball sampling was used to collect data from 137 respondents, and Partial Least Squares (PLS) structural equation modelling was used for analysis. The results showed that wellness initiatives, perceived organisational support, and worker performance were significantly positively correlated, with dedication and job satisfaction acting as important mediating factors.

A comparison of Isidro’s (2024) study with the current research highlights several gaps. While the former employed snowball sampling, the current study adopted stratified random sampling for broader representativeness. Moreover, whereas job satisfaction and commitment were used as mediators in Isidro’s work, the present study introduces government policy as a moderating variable. Methodologically, the current study applied ordinal logistic regression, which is more suitable for analyzing ordinal outcomes influenced by multiple explanatory variables, particularly when assumptions of continuity and normality cannot be met (Chau-Kuang & Hughes, 2004).

Abdi et al. (2020) studied the effect of OSH on employee performance at Afghanistan International Bank. The findings revealed that OSH significantly influences employee well-being and security, with particular emphasis on managing critical health and safety risks. Key concerns identified included fire safety, building safety, occupational health, and machine safety. The study argued that in cases of non-compliance, organizations must act promptly and collaborate with stakeholders to mitigate risks and improve performance. Data were collected from 205 employees selected through stratified sampling using validated questionnaires. Results demonstrated that several perceived factors—such as Employee Assistance Programs (EAP), Employee Wellness Programs (EWP), Health and Safety Policies (HSP), Health and Safety Inspections (HSI), Health and Safety Audits (HSA), and Health and Safety Training (HST)—significantly influence employee performance. While the study provided valuable insights, its focus on the service sector limited its generalizability to other industries.

Unlike earlier studies, Senso (2017) investigated the effect of leadership, employee involvement, and training on OSH in Temeke Municipality, Tanzania. Using a cross-sectional case study approach, data were collected from 168 respondents purposively sampled from a population of 312 employees. Analysis using bivariate correlation and multiple regression revealed that leadership, involvement, and training positively influenced OSH implementation, explaining 76.4% of the variance in health and safety practices. Furthermore, cooperation between management and employees was identified as a key factor in preventing workplace accidents and diseases. Supporting these findings, Harnois and Gabriel (2018) and Dugolli (2021, as cited in Alli, 2019) emphasized that satisfied and well-supported employees contribute significantly to organizational performance. They argued that effective organizational structures and information systems are necessary to reduce stress and improve wellness. Despite these contributions, none of these studies jointly examined wellness, and working conditions,  as predictors of performance, revealing contextual and conceptual gaps that the current study addresses in Kenya.

Mungania et al. (2016) investigated the impact of wellness initiatives on banking industry organisational performance in Kenya. Regression analysis was used to examine data gathered from all 44 banks registered with the Central Bank of Kenya using a cross-sectional survey approach. Findings revealed that wellness programs enhanced performance, though work-related stress caused by long hours and unplanned schedules posed significant challenges. While the study highlighted important stressors, it was limited to the banking sector, leaving other industries unexplored. Moreover, no moderating variable was included, and data were analyzed using simple regression. In contrast, the current study applied ordinal logistic regression for greater analytical rigor and considered government policy as a moderator.

Mungania et al. (2016) anchored their study on Work–Family Border Theory (Clark, 2000), which explains how individuals negotiate boundaries between work and family. Although useful, this theory has been critiqued for being gender-insensitive (Jacobs & Gerson, 2004) and narrowly focused on family and work domains, excluding other aspects of wellness such as leave benefits, leisure, and counseling. Given these theoretical limitations, the present study adopted more relevant frameworks—BRT and Domino Theory, which provide broader explanations of the wellness–performance relationship.

Ombasyi (2019) examined the influence of mental wellness on employee performance at Brand Design Development Limited in Kenya. Using a descriptive survey design and census sampling of 50 employees, correlation and regression analyses revealed that psychological well-being significantly enhanced performance. Stress arising from work overload was identified as a major determinant of both individual and organizational performance. The study recommended that organizations integrate counseling into human resource management to address stressors and improve wellness. However, it was limited by its small scope and reliance on Human Capital Theory. The current study extends this by sampling SMEs across 14 sectors and applying more comprehensive theoretical models, including Domino Theory, and Behaviour Reasoning Theory.

Similarly, Senthil and Harshitha (2021) investigated employee well-being and organizational performance in the manufacturing sector in Bangalore, India. Using a survey of 100 respondents, they found that well-being positively influenced performance. Nevertheless, hesitancy among some respondents to share information limited the reliability of findings, and the study’s focus on Bangalore reduced its generalizability to other contexts. These limitations further underscore the need for research in Kenya, where wellness is examined as one of the predictors of SME performance.

Working Environment and Performance

An increasing body of empirical studies conducted internationally, regionally, and nationally affirms that providing a safe and healthy working environment enhances organizational performance. While such evidence exists, scholars including Hafeez et al. (2019), Claudine Umugwaneza, Nkechi, and Mugabe (2019), Maryjoan and Tom (2019), Dunmade et al. (2019), Agboola (2022), Oluoch (2017), Mwangangi (2018), and Mwangi and Waiganjo (2017), among others, highlight persistent research gaps on the relationship between the working environment and organizational performance. These gaps necessitated the present study, which examines the influence of the working environment as one of the key variables affecting SME performance.

Using employee health as a mediating variable, Hafeez et al. (2019) investigated the effects of behavioral environmental factors (BEF) and physical environmental factors (PEF) on worker productivity. 250 employees of Pakistani software companies were given questionnaires to complete in order to gather data. The findings showed that whereas employee health improved by 33% with a one-unit rise in BEF, it improved by 35% with a one-unit increase in PEF. Additionally, there was an 80% increase in staff productivity for every unit increase in employee health. These findings are consistent with Cottini and Ghinetti (2012), who argued that poor workplace environments contribute to stress-related illnesses, absenteeism, and reduced employee commitment, ultimately undermining organizational productivity and growth. Conversely, safe and conducive work environments were found to enhance productivity, commitment, and financial stability, thereby promoting organizational growth. Similar observations were made by Bhatti (2018) and Mattson, Melder, and Horowitz (2016), who emphasized that supportive work environments significantly motivate employees.

Despite these insights, Hafeez et al. (2019) identified notable knowledge gaps. Although physical and behavioral environmental factors were shown to significantly influence employee health, and employee health positively mediated the relationship between workplace environment factors and performance, the study recommended further exploration of alternative mediating variables beyond employee health. Additionally, the adoption of a cross-sectional research design limited causal inferences, prompting calls for future studies to employ alternative research designs. These knowledge and methodological gaps underscore the need for the current study.

Using safety culture as an intervening variable, Shan et al. (2022) also investigated the connection between job satisfaction and occupational health risk perception in China. The study, which was based on the Job Demands-Resources model and Resource Conservation theory, used a structured questionnaire that was given to 237 managers and production line personnel. The results showed that job satisfaction was significantly impacted negatively by perceived occupational health risks. Additionally, it was discovered that the association between job satisfaction and occupational health risk perception was mediated by organisational commitment and work stress. While the study contributed to OSH scholarship by highlighting the importance of risk awareness and encouraging employee participation in workplace safety, it nonetheless exhibited theoretical and methodological gaps. Furthermore, given that the study was conducted in China, where workplace characteristics differ significantly from those in Kenya, there is a need for similar research within the Kenyan context to enable meaningful comparisons.

Ayalew and Demissie (2020) examined the impact of occupational health and safety hazard control programs including chemical, biological, psychological, and physical controls on organizational productivity in a tannery factory in Bahir Dar, Ethiopia. The results showed that all of the other programs significantly increased production, with chemical hazard controls having the biggest impact, except for biological hazard controls. A stratified sampling technique was used to choose 112 workers from various departments. Questionnaires were used to gather data, and multiple regression analysis, correlation, and descriptive statistics were used for analysis. The study recommended that factory managers design and implement appropriate strategies for each hazard control program to enhance productivity. In particular, managers were advised to prioritize chemical hazard controls due to their significant contribution to productivity while minimizing the costs associated with biological hazard controls, which had no measurable effect on performance.

Agboola (2022) looked at how workers at the Warri Refining and Petrochemical Company (WRPC), located in Delta State, Nigeria, felt about occupational safety measures’ effects on both individual and organisational performance. Using a cross-sectional approach, the study gathered information from 236 randomly chosen individuals in various occupational cadres. More than half of the respondents affirmed that occupational safety measures had been effectively implemented at WRPC, and nearly all participants agreed that safety training had enhanced individual performance. According to the study’s findings, workplace safety practices improve employee and organisational performance. However, its recommendations emphasized lower-cadre staff, thereby limiting its scope. Furthermore, assessing performance from the employees’ perspective, while useful, may not constitute an entirely objective measure. The present study addressed this limitation by focusing on the entire workforce rather than a single cadre.

Yamoah and Nsowah (2024) investigated how employee dedication, safety, and OSH policies affected the performance of Ghanaian manufacturing companies. 162 workers from three businesses in the Awutu Senya District provided the data. The study demonstrated through correlation, regression, and mediation analysis that OSH practices had a beneficial impact on employee commitment and performance, with commitment acting as a partial mediating factor in this relationship.

Maryjoan and Tom (2019) looked into how worker job performance was affected by industrial safety and health in a few cement companies in Cross River State, Nigeria. Employing a survey design, the study collected data from 100 employees selected through simple random sampling. Results confirmed a significant positive correlation between safety and health strategies and job performance. The study recommended that organizations provide adequate safety and health strategies to reduce turnover and enhance performance. However, the study demonstrated contextual gaps, as it omitted wellness and  working conditions,  which are addressed in the present research. Methodologically, Maryjoan and Tom (2019) employed regression analysis, whereas the current study applied ordinal logistic regression to examine predictor–outcome relationships. Additionally, the cement industry context in Nigeria differs substantially from the SME context in Kenya, underscoring a geographical and sectoral gap.

In a same vein, Dunmade et al. (2019) evaluated how OSH practices affected worker performance in the Nigerian state of Kwara’s steel and kam wire industries. Out of 318 employees, 177 responses were selected at random from a simple random sample. Structured questionnaires were used to gather data, and multiple regression models and descriptive statistics were used for analysis. The findings demonstrated that safety and health measures significantly enhance employee performance, reduce accidents, and mitigate workplace stressors. The authors recommended employee involvement in decision-making and regular inspection and maintenance of OSH facilities. Nonetheless, the study exhibited methodological limitations, as it relied solely on simple random sampling and did not incorporate advanced analytical methods. In contrast, the current study employed stratified random sampling and ordinal logistic regression for more robust analysis.

Mutegi et al. (2023) investigated the connection between worker productivity and workplace safety in Kenyan manufacturing companies. Ergonomics, emergency response, safety training, and risk transfer were the study’s independent variables, and productive time, job completion, and value addition were used to gauge productivity. 124 businesses from 14 manufacturing subsectors were sampled for the study using a cross-sectional survey design. Findings revealed significant positive associations for three of the independent variables. The study contributed practical insights for designing safety programs but recommended further research employing alternative statistical techniques.

A theoretical investigation into the impact of occupational safety and health (OSH) procedures on the working environment in Kenya’s water service sector was carried out by Oluoch (2017). The study looked at the connection between organisational performance and OSH practices. The results showed that employee productivity in manufacturing companies was positively correlated with occupational safety policies. A sample of 80 workers was selected from the Lake Victoria South Water Service Board and Kisumu County Water Service Provider (Kisumu Water and Sewerage Company Limited), with an emphasis on workers involved in the building and maintenance of water and sanitation facilities. The study established that enhancing staff awareness of OSH practices improved the work environment, r(76) = 0.363, p = 0.001, CI = 95%. Furthermore, a reduction in exposure to hazards and risks was associated with improvements in the work environment, r(76) = -0.095, p = 0.413, CI = 95%. The study also highlighted collaboration and teamwork as integral components of a supportive workplace environment (Boustras & Guldenmund, 2018). To identify respondents, both census and simple random sampling techniques were applied, while data were collected through self-administered semi-structured questionnaires. The study concluded that organizations enhance their effectiveness by reducing the frequency and severity of occupational hazards, illnesses, and workplace injuries, mitigating stress-related conditions, and improving overall work standards for employees.

A review of Oluoch’s (2017) findings alongside those of Ayalew and Demissie (2020) reveals divergent perspectives regarding the influence of biological and chemical hazards on organizational performance. While Ayalew and Demissie reported that biological hazards were insignificant in influencing performance but chemical hazards were critical, Oluoch found that both biological and chemical hazards significantly affected the management of the water service industry. He added that both hazards should receive immediate attention, while other risks may be handled later. He suggested that future research concentrate on the effects of the biological and chemical risks that are already present on workers. These contradictory results highlight the need for additional study to fill in the information gaps regarding the connection between organisational success and the workplace environment. Moreover, Oluoch’s study was confined to the water sector, excluding other industries within the manufacturing sector, thereby highlighting a contextual gap. The present study sought to address this gap by investigating occupational safety and health practices across 14 SME sectors.

With a focus on Penta Flowers Limited in Thika Sub-County, Mwangi and Waiganjo (2017) investigated the impact of occupational health and safety (OHS) measures on worker performance in the flower business. One of their main goals was to evaluate how OSHA training and workers’ attitudes towards safety affected their performance, even though their larger study focused on OSH procedures. The study used an explanatory sequential design and a mixed-methods methodology. A sample of 200 people was chosen, including 130 general employees, 50 supervisors, and 20 managers. The results showed that performance was significantly impacted by both safety instruction and favourable staff attitudes. The study recommended the adoption of an integrated training framework to provide all employees with consistent and accessible safety education, alongside initiatives to foster positive safety attitudes as a means of reducing workplace injuries. Mwangi and Waiganjo further argued that employees represent the most valuable assets of organizations, capable of driving greater performance when effectively supported. However, they cautioned that lapses in OHS measures expose workers to health and safety hazards that diminish performance. Inadequate protective measures not only increase absenteeism, accidents, and occupational illnesses but may also result in permanent disability.

These findings align with prior empirical studies, which have also established a positive relationship between OHS practices, job satisfaction, and employee performance (Oduor, 2021). Nonetheless, some scholars caution that such results cannot be generalized across small and medium size enterprises in Kenya, as the relationship between OHS practices and performance vary across regions, industries, and national contexts (Baah, 2015). A safe and productive work environment therefore requires the establishment of clear policies and procedures to safeguard employee health. Fundamental to these policies are the identification of hazards, implementation of appropriate controls in line with regulatory standards, and the provision of continuous safety education and training (Kirwan et al., 2013)

In another Kenyan context, Mwangangi (2018) investigated the link between OSH practices and employee performance at Kenya Power and Lighting Company (KPLC). The study considered training and employee attitudes as independent variables and confirmed that safe working environments require well-structured policies and procedures. Policy implications highlighted the importance of hazard identification, compliance with government standards, and safety training (Kirwan et al., 2013). However, the study was anchored on the Distraction Theory and Goal-Freedom Alertness Theory, creating a theoretical gap that the current research addressed by adopting the Behavior Reasoning Theory (BRT),. Moreover, the KPLC-focused study had limited generalizability due to its small sample size, while the current study addressed this limitation by analyzing 295 firms distributed nationally.

Burton (2010) advanced the Healthy Workplace Framework and Model, emphasizing the critical role of employee mental and physical health in organizational success. From a business perspective, ensuring employee well-being through health protection and promotion is strongly associated with long-term competitiveness. Burton argued that organizations with robust OSH records and healthier, more satisfied employees achieve superior performance outcomes. This is consistent with Cottini and Ghinetti (2012), who noted that poor workplace environments foster absenteeism and stress-related illnesses, reducing employee responsibility and organizational progress. Conversely, safe and supportive environments enhance productivity, commitment, and financial well-being. This model reinforces the empirical evidence reviewed, which generally supports the positive relationship between OSH practices and organizational performance. However, despite these contributions, several theoretical, methodological, and contextual gaps remain. Without adequate health interventions, OSH risks continue to pose significant challenges for SMEs (Burton, 2010).

Conceptual Framework

The study was anchored on the conceptual framework Shown in Fig. 2 which described the relationship between predictor variables and dependent variables including the moderating variable. The framework identifies occupational safety and health practices as the independent variables and performance of SMEs as the dependent variable. Employee wellness and working environment were identified as the key sub-variables of occupational safety and health practices, Employee Wellness focused on welfare services provided by SMEs such as the provision of Personal protective equipment, counselling and stress management program including safety training offered in organization. The relationship between wellness and performance of SMEs is shown by H_1. Where an effective program is in place, performance of SMEs is likely to improve. Healthier employees are more productive, take fewer sick and disability days, and are at a lower risk of many serious health problems (Phillips & Gully, 2014). Many employers pursue wellness programs for productivity gains in terms of reduced errors, improved efficiency and better decision-making (Sullivan, 2013). A workplace wellness program benefits both the employee and the employer in the long run. In a nutshell, comprehensive workplace wellness program can over time, produce three key benefits: less absences, high productivity and worker satisfaction and retention (Bray, 2016).

The working environment analyzed hazard identification, risk analysis and risk control measures. The link between the working environment and performance of SMEs is depicted by H₂ as evidenced in the literature review, organizations that generally conduct a risk analysis and puts mitigating measures against OSH hazards report fewer cases of accidents and employee’s productivity is enhanced. However, from the behavioral theory model, accidents occurrence is influenced by human behaviors and error (Guerin & Sleet, 2020).

RESEARCH METHODOLOGY

Current study adopted positivist philosophy. According to positivism, only information derived from measurement and observation is trustworthy. According to Saunders, Lewis, and Thornhill (2023), the concept is founded on actual facts, objectivity, neutrality, measurement, and the validity of results. According to positivism, the researcher’s duty is restricted to gathering and objectively interpreting data, with the goal of producing conclusions that are easily observable, quantifiable, and amenable to statistical analysis (Dudovskiy, 2022).

To gather information from small and medium-sized businesses in Kenya, the study used an inferential research design and a cross-sectional descriptive survey. Descriptive research, according to Stangor (2015), asks people about their implementations, attitudes, behaviours, or values in order to get data that depicts an actual reality. Inferential design was used to analyses the relationship between the variables and allow the testing of hypothesis of the study. Cross-sectional data was gathered using online questionnaires. Sullivan, Parvanta and Begin (2007) and Higgins (2014) recommend cross-sectional research design in situations where primary data is necessary, time is a limitation and longitudinal data unavailable. This design is particularly beneficial when gathering data from several respondents concurrently, permitting a wide and comparative analysis (Blumberg, Cooper & Schindler, 2014).

The study applied the Yamane (1967) formulae to obtain an appropriate sample size from the target population of 1118 SMEs, at a 5% estimation error giving sample of 294. The 294-sample size was ideal representation of the target population as it surpassed the 20% recommended by mills and Gay (2014). Using each sector’s proportionate sample, simple random sampling using a random number table was used to select the firms from which data was collected. Self-administered structured questionnaires were used to collect primary data from the study’s selected SMEs respondents. Prior to conducting actual research, a pilot study was conducted to ascertain how occupational safety and health standards affected the performance of SMEs in Kenya. According to Yin (2011), pilot studies help researchers test and improve one or more aspects of a final study, such as its design, fieldwork procedures, data collection tools, or analysis plans.

Internal consistency was calculated and test score reliability for structured questionnaires was estimated using Cronbach’s alpha coefficient estimate. The validity of instruments was measured using Content Validity Index (CVI). Two assessors/experts in the field of study were used to rate the content in the questionnaire. These experts assisted in assessing the phrasing of the questions to avoid ambiguity. The extent to which the questions were related to the topic of the study was assessed and the CVI computed using the formula shown below. The estimation for validity was 0.75 and above confirming validity.

The study applied Ordinal Logistic Regression (OLR) to model the relationship between the ordinal response variable and one or explanatory variables. OLR was suitable since it separates data into two sets made up of first- and second-order constructs, with binary dependent variables and nominal/ordinal independent variables (Ochola, 2013). OLR’s presumption that the variables are ordinal must be true for it to be applied to data analysis in any study. The Wald Chi-square was used for logistic regression analysis (Osano & Kesusu, 2016). The model was based on cumulative distribution function (Erkan & Zeki,2014). Proportional odds models were estimated using cumulative probabilities (Kleinbaum and Ananth, 1997). McFadden’s, the ratio of the likelihoods test, and the Cox and Snell test were used to evaluate the coefficient of determination

FINDINGS AND DISCUSSIONS

Employee Wellness Rating

To adequately measure the employee wellness rating, the variable was broken down into nine sub constructs. The respondents were thus asked to rate the latter on the Likert scale SA-strongly Agree, A-Agree, N-Neutral, Disagree, and SD-Strongly Disagree with the results being as in table 1

Table 1:  Employee Wellness Rating Results

It is evident from the results that majority of the respondents, 76% confirm that that small and medium enterprises employees are provided with adequate PPE while a mere 15.2% disagree that employees are provided with adequate PPE.  Similarly, a majority agreement  77.9% of the respondents confirm that most organizations have medical insurance cover for their employees  This trend can be observed  with respect to adequacy of First aid room where a commanding 85.5% were in agreement, ,  provision of special programs for people  abled differently (57.2%) , Medical checkups and OSH trainings (58.1%),  availability of Washing and changing areas for staffs (58.1%.medical checkups and biometric screening (69%) and  that OSH trainings (79.1%). However, contrary to the above, a significant majority of the respondents were in disagreement on the adequacy of gym facilities (62.8%, counseling, EAP services (62.6%).

These findings present some consistence’s with the empirical studies on employee wellness and performance.   Wellness programs and perceived organisational and employee performance were shown to be significantly correlated by Isidro (2024), with commitment and job satisfaction acting as a strong mediating factor between the two variables. In a similar vein, Abdi et al. (2020) discovered that employee performance is highly influenced by the categories of recognised factors that are perceived, including Health and Safety Policy (HSP), Health and Safety Inspections (HSI), Health and Safety Audits (HSA), Employee Wellness Programs (EWP), Employee Assistance Program (EAP), and Health and Safety Training (HST).

Work Environment Rating

To adequately measure the working environment, 15 sub constructs inform of responses were presented. The respondents were thus asked to rate the latter on the Likert scale SA-strongly Agree, A-Agree, N-Neutral, Disagree, and SD-Strongly Disagree with the results presented as shown in table 2.

Table 2. Descriptive Statistics for Working Environment

From the results, it is evident that majority of the respondents, 74.7% confirm that Hazards are frequently identified while a slight 15.2% disagree that hazards are frequently identified. employees are provided with adequate PPE.  From the results it is evident that the work environment in most of the organizations is conducive as evidenced by most respondents agreeing, risks are classified for ease in management; at 83.8% agreement and the work environment is generally safe; as confirmed by agreement level among respondents of 78.6%. The respondents confirmed lack of evidence of risk analysis and safety at 44.5% agreement and 48.5% disagreement while inspections are conducted regularly as confirmed by majority, 76.8% of the respondents. While majority of the respondents 80.4% were confident that OSH Audit is done annually, majority, 77.7 % of the respondent confirmed that accidents are reported to management with 78.2% of the respondent exceeding confidence that investigations are done when the accidents occur. Majority, 84.3% of the respondents confirmed that dangerous materials are eliminated from work environment; as the management ensures employees strictly follow the procedures at work as evidenced by acceptance level of 80.3%. As risk control measures, there are in place adequate PPEs as confirmed by 79% of the respondents and Cautions and notices to employees was average as confirmed by just 58% of the respondents. There was very high level of insurance of OSH risks as evident in the very high agreement rate of 80.4% among respondents. There was evidence that employees are compensated in case of a risk. When asked whether the Management has put in place a safe and health   work environment, 78.6% felt there were adequate measures put in place. Lastly, employees are committed towards accident prevention in their work stations as revealed high acceptance rate of 71.2%.

Performance of SMEs

To adequately measure organizational performance sub variables. The respondents were thus asked to rate the latter on the Likert scale SA-strongly Agree, A-Agree, N-Neutral, Disagree, and SD-Strongly Disagree with the results presented as shown in table 6

The analysis was based on the performance of SMEs sub variables: reduction of operational costs, Reduced absenteeism, Increased job satisfaction, Improved productivity, reduced medical bills, Ongoing health & safety improvement and availability of safety information as shown in the conceptual framework. This ensured all aspects all aspects of SMEs performance were adequately covered to determine the relationship between OSH   and performance of SMEs.

Table 6: SMEs Performance Rating Results

Findings shown in table 4.8 reveal that the organizations experience improved employee morale, job satisfaction and productivity of staff at 81.4% agreement. The organizations experience reduced accidents in work place; at 86.5% and reduced absenteeism and lateness of staff at 66.8%. The respondents confirmed that the organizations adhere to government OSH regulations and records improved OSH as evidenced by high acceptance level of 67.2%. Similarly, the respondents contend that the organizations experience efficiency of workers; at 79% acceptance level. Similarly, the respondents were confident that the organizations realized positive psychological outcome among staff at 77.5% and reduced medical bills at 80.8%. In the study, SMEs registered acceptable performance as evidenced by higher percentage responses of strongly agree and agree (above 70% in all categories confirming the proposition of behavioral theories that acceptable performance is achieved through comfortable workforce.

Correlation Analysis.

The results of the Correlation analysis examined the degree of relationship between the dependent variable and independent variables using Pearson correlation coefficient (r), which yield a statistic that range between -1 to 1. Gogtay and Thatte (2017), states that the bigger the r (absolute value), the stronger the correlation between the two variables. Where the correlation coefficient is positive (+ve), the impression is that there is a positive association between two variables and vice versa.  If r has zero value, it shows that there is no correlation between the two variables. The Pearson correlation results between the Performance of SMEs and the predictor variables indicate working environment (.606) had the highest significant correlation followed by employee wellness (.371). From the results there is evidence that relationship between the predictors was significant as all coefficients are positive. The findings are consistent with Oluoch (2017) study that established significant correlation between occupational safety practices and the performance of staff in SMEs

Table 7. Correlation

The table 7 shows that there was correlation between employee wellness and working environment with a correlation coefficient of (0.778) .All coefficients were significant with p-values of 0.000 which is less than 0.05, the level of significance.

Ordinal regression Analysis

This study used Ordinal regression model for purposes of predicting the performance of SMEs, given the two independent variables; Employee Wellness and Working Environment. Ordinal regression is used to envisage the dependent variable with ‘ordered’ multiple categories and independent variables (Zikmund 2014).    In other words, it is used to accelerate the interaction of dependent variables (with numerous ordered ranks) with one or more predictor variables. The analysis was intended to   estimate two ordinal regression models in line with the research objectives, the main effect model to test the hypothesis   listed under each  below.

H₀₁: Employee Wellness has no significant effect on performance of small and medium size enterprises in Kenya.

H₀₂: Working environment has no significant effect on performance of small and medium size enterprises in Kenya.

Proportional Odds Assumption Test

One barrier towards using ordinal regression methods lies in the understanding and validation of the assumption of proportional odds. In order to assess the suitability of the ordinal regression analysis in this study, the proportional odds assumption was tested using test of parallel lines. In this study, Table 8 shows the results of the proportional odds test using test of parallel lines.

Table 8 Test of Parallel Lines

From the model, null hypothesis which stated that the slope coefficients in the model are the same across the response categories, the rejection of the hypotheses at p value less than 0.05, would suggest that the model assumption of parallel lines was violated. This would indicate that the proposed Ordinal Regression methodology is inappropriate as a modeling technique. In the case where the value is greater than 0.05, the assumption holds and the model is said to be non-significant hence appropriate.

In this study, with a χ²_((33,229))=”30.267″ , p=0.604 (<0.05) this therefore suggests that there is no evidence from the sample to reject the parallel lines assumption leading to the conclusion that the Ordinal regression model is appropriate for the study as stipulated by William (2016). For the model, the proportional odds assumption therefore appears to have held because of the significance of the Chi-Square statistic (.604 < .05)

Main Effect of Ordinal Regression Model Estimation Results  

The subsection presents Model Fitting Information, Goodness of Fit, Pseudo R^2, Classification Table and the parameter estimates The main effect analysis was organized in terms of overall fit of Estimated Model using Model Fitting Information, Evaluation of the Goodness of Fit of the Model using Goodness-of-Fit test, Evaluation of the power of the variance explanatory power of the model using the Pseudo R-square test, evaluating of the predictive power of the model using evaluation of the overall effect of the Predictor variables using the Likelihood Ratio Test statistics and the Parameter Estimates Results and the classification table. The findings are as summarized in this section.

Overall Fit of Estimated Ordinal Regression Model

The first step in ordinal regression model estimation was to assess whether the estimated model has a better predictive power than the theoretical intercept only model.  It is sufficient to note that if the p-value is less than the selected level of significance, in this case 0.05, then the model fits the data significantly better than the theoretical model otherwise. The model with the predictors (full model) does not fit the data better than a model without the parameters (the theoretical model). To achieve this, the model fitting information was derived with results as in table 9

Table 9 Model Fitting Information

The outcome shows the model’s parameters for which the model fit is computed. “Intercept Only” refers to a model that predicts the outcome variable by fitting an intercept without controlling for any predictor factors. A model that incorporates the designated predictor variables and whose coefficient has been calculated through an interactive procedure that maximises the outcome’s log likelihood is referred to be “final.” It is reasonable to assume that the “Final” model will outperform the “Intercept Only” model by incorporating the predictor variables and optimizing the log likelihood of the result. This is typically observed in the variations in the models’ -2 (Log Likelihood) values. The amount of variability in the data that can be explained is measured by log likelihood.

The improvement or otherwise is tested using the chi-square value corresponding to the final model which should be significant for a conclusion that the final model is better than the intercept only model to be drawn.  The results summarized in table 4.11, the -2Log Likelihood for intercept only value of 313.852 compared to that of the model of 236.390 , which is a difference of 77.462 , a change which statistically significant (χ^2_((3,229))=”77.462″,p=0.001<0.05) (.This indicates that the final model gives a significant improvement over the baseline intercept-only model and hence a better predictive of performance of SMEs than the intercept only case. The null hypothesis was thus rejected which can be interpreted as the full model explains a significant amount of the original variability.

Evaluation of the Goodness of Fit of the Model

The Goodness-of-Fit table (Table 10) is the following table in the output. The model’s Pearson’s chi-square statistic and a different chi-square statistic based on the deviation are both included in this table. The purpose of these statistics is to determine if the fitted model and the observed data are consistent. The null hypothesis, which states that the fit is as good as the null hypothesis, comes first. If the null hypothesis assumption of a good fit, is rejected (conventionally if p<.05), then the model does not fit the data well.

Table 10: Goodness-of-Fit

From the results in Table 4.12, the result gives a Pearson chi-square of χ^²(97))=”269.846,p”=”0.000″<0.05  implying the model does not fit the data well which same conclusion   would be drawn from the Deviance test statistic test result χ²((97))=”196.148,p”=”0.001″<0.05

Fagerland M, W.M & Hosmer, D (2016), suggests that the above result should not be used as a panacea to assess the validity of the model. The authors opine that the chi-square is highly likely to be significant when sample size is large, as it is in this study and that in such circumstances the conventional p value of 5% may need to be set lower for rejecting the assumption of a good fit. Most significantly however is that while the chi-square can be very useful for models with a small number of categorical explanatory variables, the statistic is highly sensitive to empty cells as is evidenced in the this study results. The authors therefore propose that whatever the goodness of fit test results, other methods such as measures of association, like the pseudo R², should in addition be used.

Assessment of Proportion of Variability in Outcome Variable Explained by The Predictors

The coefficient of determination (also known as pseudo-R²) in multiple regression analysis indicates how much of the variation in the outcome variable can be explained by the predictor variables. In other words, Adjusted R² in linear regression summarises the percentage of outcome variance that can be explained by the explanatory variables; higher R² values mean that a better model can account for more variation in the outcome, up to a maximum of 1. However, the same R² statistic that can be computed for linear regression cannot be computed for logistic and ordinal regression models. Thus, Cox and Snell, Nagelkerke, and McFadden are the three approximations that are calculated (Table 11).

In this study, the R-square presented in Table 12 similarly presents information about how much variance is explained by the independent variable using three statistical measures, the Cox and Snell, Nagelkerke‟s, and McFadden‟s pseudo-R² statistics.

Table 12: Pseudo R-Squ

The Pseudo r2 values (Nagelkerke. = 0.795=80%) presented indicate that the estimated main effect model with its three independent variables explains a relatively large proportion of the variation well above the threshold of 70% pseudo _r2 value.  This therefore indicates that the model with employee wellness and working environment is a good predictor of performance of small and medium size enterprises in Kenya.

Evaluation of the Predictive Performance of the Estimated Model

The next stage of the model estimation   is to examine the predictions generated by the estimated model. In this context, what is of interest is how often the model can produce correct predicted categories based on the values of the predictor variables i.e. employee wellness and working environment. In ordinal logistics regression, this is achieved through the construction of a classification table also known as confusion table by cross-tabulating the predicted categories with the actual dependent variable categories.

An ideal model would display only values on the diagonal-correctly classifying all cases. Tallying across the rows denotes the number of cases in each category in the actual data and inputting down the columns embodies the number of cases in each category as classified by the complete model. The crucial part of information is the overall percentage in the lower right corner of the classification table as shown in table 12

Table 12: Classification Table

Source: Author (2025)

It is critical to note that in interpreting the table, while values in the leading diagonal indicate correct predictions’ those off the diagonal are the incorrect cases. In this study therefore:

1. 1 (100%) of the 1 individual whose response was strongly disagree in the SME performance rating was correctly classified.
2. 27 (87.1%) of the 31 individuals whose responses was “neutral” on the SME performance rating was correctly classified
3. 100 (89.3%) of the 112 individuals whose responses was “Agree” on the SME performance rating was correctly classified
4. 46 (59.0%) of the 78 individuals whose responses was “Strongly Agree” on the SME performance rating was correctly classified
5. Overall, however, 76. % of the cases are classified correctly.

Evaluation of the Effect of the Predictor Variables on the Mses. Performance

The Parameter estimates is the core of the output, showing specifically the relationship between the explanatory and the outcome variables and if this relationship is statistically significant or not specifically. The Table 13 presents estimates in terms of the values of the regression coefficients and intercepts, together with the corresponding standard errors, Wald test, p-values and Odd Ratio (EXP (β)

Table 13 Parameter Estimates

The Parameter Estimates table indicates the coefficients, their standard errors, the Wald test and associated p-values (Sig.), The first section of the output the threshold shown at the top of the parameter estimates output, are the intercepts, sometimes referred to as the cut-points that represent the threshold of the ordinal logistic regression. They indicate where the latent variable is cut to make the three groups that are observed in the data.  The threshold coefficients represent the intercepts, specifically the point (in terms of a logit) where health and safety status might be predicted into the three categories.

For the ordinal logistic regression model, there are n-1 intercepts, where n is the number of categories in the dependent variable (performance of SMEs). In this case, with n-5 categories, there are therefore 4 intercepts corresponding to 1- Strongly Disagree to 4 = Agree.

The second section of the output the location, are the independent variables employee wellness and working environment sometimes referred to as the predictor variables of the ordinal logistic regression. It is instructive to note that all the explanatory variables of factor type of ordinal nature are measured in terms of 1- strongly disagree, 2=disagree, 3= neural, 4=agree and 5=strongly agree. The relative value of the coefficient for these factor levels can give important insights into the effect of the predictors in the model.  Generally, in ordinal regression, the explanatory variable with the largest coefficient with a corresponding p-value less than the chosen significant level of 0.05 is considered the most significant influential factor. Since p values are less than alpha level, they are statistically significant; otherwise not.

Column “EXP (B)” in the table 14 above represents the variations in odds for a unit increase of the corresponding independent variable. Odd ratios lower than 1 correspond to decreases and odds ratios more than 1.0 corresponds to increases in odds. It can further be interpreted that when odds ratios are nearer to 1.0, it suggests that a unit change in the independent variable has no effect the outcome variable. They are the odds ratio of the predictors hence an exponential of the coefficients. It shows the probability of an event happening rather than the association between the outcome variable and the predictor variable, that is, the effects of the predictor variable on the odds ratio. Further, the degree of free (df) was established as 1 since this model had only one degree of freedom for each predictor.

To establish the estimated regression models using the study results, the following variable codes as defined in chapter 3 is used:

y_j = the j^th rating category where j=1,2,3,4 wish j=5 being the reference category

x_ij = the i^th independent variable for i-1,2,3 and j=1,2,3,4 with j=5 being the reference category

Using the above definition and noting that 1=1 represents employee wellness, 1=2 working environment and 1=3 ergonometric, the following estimated ordinal regression models on the basis of parameter estimates coefficients for the independent variables in the location section and the outcome variable coefficients in threshold section (Eqs.4.1 – Eqs. 4.4):

y_j=1= -27.063 + 56.361X₁₁ -1.665X₁₂ + .126₁₃ – 0.002X₁₄-40.614X₂₁ -3.587X₂₂ -5.575X₂₃ -1.631X₂₄ ———————-

_{Equation 4.1}

y_j=21 = -9.927 + 56.361X₁₁ -1.665X₁₂ + .126₁₃ – 0.002X₁₄-40.614X₂₁ -3.587X₂₂ -5.575X₂₃ -1.631X₂₄ ————————

_{Equation 4.2}

y_j=3= -5.489 + 56.361X₁₁ -1.665X₁₂ + .126₁₃ – 0.002X₁₄-40.614X₂₁ -3.587X₂₂ -5.575X₂₃ -1.631X₂₄ –_{Equation 4.3}

y_j=4 = + 56.361X₁₁ -1.665X₁₂ + .126₁₃ – 0.002X₁₄-40.614X₂₁ -3.587X₂₂ -5.575X₂₃ -1.631X_{24 —————————-Equation 4.4}

The results above indicate that working environment has a statistically significant effect on SMEs performance   at three of the four rating categories, 2 – Disagree (b=-3.587, SE=1.336, Wald=7.209 p=.007<0.05). 3-Neutral (b=–5.575, SE=1.171, Wald=22.664, p=.000<0.05).  and 4=Agree (b=–1.631, SE=.365, Wald=22.664, p=.000<0.05).   Assuming no change in employee wellness practices, the results further indicate that, firms where respondents either disagreed or neutral or agreed   on the level of investment in working environment were respectively 3.587, 5.575 and 1.631 less likely to perform better compared to cases where there was strong agreement with the level of focus on working environment

The likelihood of   working environment affecting  performance  is  exp (b) 0.00, 0.03 0.00 and 0.20  with the coefficient of -5.575 and  equivalent to odds ratio of  at P=000(<0.05) is the most influential factor in influencing performance of SMEs.).Each of the coefficients are generally interpreted as the predicted change in log odds being in a higher rating category on the dependent variable SMEs Performance (controlling for the other independent variable as per unit increase in the independent variable. The assessment of the effect of each of the Predictor Variables on the SMEs performance is explained below.

The above estimates represent the relationship between the independent variable (employee wellness(X1) and working Environment(X2), and the dependent variable (Performance of SMEs).

They display the degree of surge or decrease (if the coefficient is negative) in the estimated log odds which would be estimated by a 1-unit increase or reduction in the predictor variable, taking all other factors constant.

From table 15 above, it was established that at 0.05 significance level, employee wellness factors were found not significant in determining performance of SMES as their p-values at the four categories Sign. were more than 0.05.

Employee Wellness and Performance of SMES

Regarding Employee Wellness and performance of SMEs, the study was designed to test the following null hypothesis:

H₀₁: Employee Wellness has no significant effect on performance of small and medium size enterprises in Kenya.

Respondents rating data was collected on the extent to which the organization provides Employee Wellness services/facilities and the extent to which these contribute to SMEs performance. Compared to those with very high investment and controlling for working environment, firms who put little effort in employee wellness services/facilities in intervention (1=Strongly disagree rating category of employee wellness), are more likely to have improved performance.  This result is however not statistically significant (b= 56.361 〖χ^2〗_((1))=0.000,p=0.986>0.05). For firms with higher levels of investment on employee wellness equivalent to 2=disagree, and 3 Neutral, = were similarly found to have, albeit none statistically significant effect on performance as respectively demonstrated by (b= -1.665 〖χ^2〗_((1))=2.851,p=0.091>0.05) and (b= 0.126, 〖χ^2〗_((1))=0.037,p=0.848>0.05). Lastly, again compared to those with very high investment, those with moderately high levels of investments in employee wellness equivalent to 4=Agree rating surprisingly were less likely to have improved performance as demonstrated by (b= -0.002〖χ^2〗_((1))=0.000,p=0.997>0.05) with  the effect not being statistically significant.  This result leads to the conclusion that there is no evidence from the sample to lead to the rejection of the null hypothesis that employee Wellness has no significant effect on performance of small and medium size enterprises in Kenya. However, while controlling for the effect of working environment, employee wellness   was found to have the least coefficient although with a statistically significant effect on SMEs performance. This therefore leads to the null hypothesis being rejected and a conclusion made that employee wellness has positive relationship with performance of small and medium-size enterprises in Kenya. The findings are consistent with Senso (2017 study that established leadership, employee involvement, and employee training have an effect on the implementation of health and safety practices. Training was identified as one of the sub constructs of wellness in the current study from this study it can be deducted wellness comprises of other sub constructs like leadership that was not included in the current study. Similarly Ombasyi (2019)  supports the current study   that   positive and highly significant correlation exists  between employee performance and stress with a mean of 3.575 indicating that stress in the workplace occurs when employees are presented with high job demands and   employee wellbeing has a significant role in enhancing organization performance  as demonstrated by a mean of 3.925 and finally  psychological wellbeing influence the future performance of individuals as shown by a mean of 4.225. the study recommends that organizations should embrace a healthy workforce through the human resource and counselling department to augment general performance of the organization. Counselling and support services were examined as sub constructs of employee wellness in the current study. Another study supporting the findings of the current study is Senthil & Harshitha, (2021). They found employee well-being within the organization was very effective in enhancing employee performance within the organization. towards health and safety procedures at work place. These results provide a link between theory and practice and enhance appreciation of the need for organizations to ensure employee wellness. The finding is consistent with the proposition of Herskovitz & DeCamp (2019) that organizations need sound occupational safety and health hazards management practices and that employee wellness is associated with improved organizational performance. and health hazards such as the physical, biological, chemical and psychosocial hazards would be the first step towards accident prevention (Seyyed & Zahra, 2012).

Employee wellness promotes overall worker safety and health (Sundtoft, 2018). person’s proactive role in protecting themselves in the presence of hazards and other barriers in their environments is an outcome of physical and emotional wellness (Sahu, Padhy, & Dhir, 2020). Similarly, theory of planned behavior presents that workplace ailments and injuries affect not only the employees, but also their families, and communities. Organizations must therefore ensure effective interventions and programs to aid employees improve on their health status and manage behavior change at individual and organizational level in order to realize positive performance (Sahu, Padhy, & Dhir, 2020).

Although employee wellness was found to have   least significant effect on SMEs performance as compared to working environment, there was significant correlation between wellness and performance of SMEs Hence the Null hypothesis H01: Employee Wellness has no significant effect on performance of small and medium size enterprises in Kenya. was therefore rejected. This implies that the alternative hypothesis 01 was accepted. Leading to the conclusion that employee wellness has significant influence on performance of SMEs.

Working Environment and Performance of Small and Medium-Size Enterprises

The second hypothesis in the current study was;

 H0₁: Working environment has no relationship with performance of small and medium-size enterprises in Kenya.

From the results, the working environment factors signs indicate that the sign values are less than the p-value at category the three categories except for category 1 compared to employee wellness. This Implies that working environment significantly affects organizational performance controlling for the wellness variable.

Conducting factor analysis on working environment, respondents rating data was collected on the extent to which the organization provides an ideal working environment and the extent to which this contribute to SMEs performance. Compared to the other investment and controlling for employee wellness SMEs investment in working environment in intervention 1= strongly disagree rating category and intervention are not statistically significant as their p-value is greater than level of significant 0.05. The results equivalent rating category 3 -Neutral, and 4-Agree = and 5-strongly agree were found to have, albeit none statistically significant effect on performance as respectively demonstrated by:

(b= -3.587〖χ^2〗_((1))=7209,p=0.007<0.05) and (b24= -1.631〖χ^2〗_((1))=19.919,p=0.001<0.05).

While controlling for the effect of employee wellness, working environment (β=”1.225,Wald 23.007,p”=”0.000 “(<“0.05”) was found to have the largest positive coefficient in the overall parameter estimates with a statistically significant effect on SMEs performance. Working environment therefore was ranked the most important factor affecting performance of SMEs.  The null hypothesis was rejected and a contrary conclusion was made that working environment has significant relationship with performance of small and medium-size enterprises in Kenya. There is a strong association between factors that affect health status, even when p-values are less than alpha level. Environmental factors such as creating a health hazard free environment, with ideal temperatures, well cleaned offices, provision of clean lavatories, early detection of health risks among other environmental factors are seen to be significant.

The Findings of the current study are consistent with a number of empirical studies explored in chapter two. Gbadago et al. (2017) study found healthy and safe work environment to have the capability to increase organizational productivity and growth. Similarly, Hafeez et al. (2019) established that physical and behavioral environmental factors positively affect employee health which in turn significantly affects employee Performance. Other scholars Bhatti (2018); Mattson, Melder, and Horowitz (2016), illustrated that the environment of the workplace had enhanced consequences by motivating employees. The goal of Safety and health management practices is to minimize the hazard thereby ensuring safe and healthy work environment (Gbadago et al., 2017) and enhancing employee delivery towards desired outcome and performance (Gehad et al., 2021). Chirico et al. (2019) stablished that work environment  influences employees’ psychological being and ability to contribute and that risky work environment may adversely affect organizational performance, an argument further cemented by Oluoch (2017) who reported positive relationship between occupational safety practices and the performance of staff in SMEs, arguing that organizations become considerably more effective by lowering the occurrence and intensity of occupational hazards, illnesses, workplace injuries, stress-related sickness, and enhancing the standard of work for their employees. Consistent with Shan et al. (2022) the results of the study found that    perceived occupational health risks significantly negatively affected job satisfaction, and both work stress and organizational commitment mediate the relationships between perceived occupational health risks and job satisfaction.  Finally, a study by Ayalew and Demissie (2020) found except biological hazard control program, the three occupational hazard control programs (chemical psychological and physical) have positive and substantial effect on organizational productivity, with the chemical hazard control programme having the strongest impact. Agboola (2022) study concludes that occupational safety measures were perceived to have positively affected workers’ and organizational performance.

Linking the study findings with Heinrich’s theory, work environment may shape worker personality and personal characteristics consistent with safety regulations or standard operating procedures (Wang & Yan, 2019) According to the theory, there is a significant correlation between employees’ character and job outcomes which in turn predict performance (Wang & Yan, (2019). Regrettably, environmental errors in an employee’s family or life may cause the secondary personal defects, which contribute to unsafe acts which can affect employee performance adversely.

Therefore, with the ordinal regression analysis result demonstrating that working environment is positively significant to performance of SMEs, and evidence of the various studies in support of this finding, there is clear evidence that working environment is very critical in enhancing performance of SMEs. hence the null hypothesis H₀₁: Working environment has no relationship with performance of small and medium-size enterprises in Kenya is rejected. This Implies that working environment significantly affects performance of SMEs controlling for the wellness Variable.

Table:14.  Summary of objectives, hypothesis, conclusion and findings

CONCLUSION

The study investigated the relationship between occupational health and safety practices and performance of SMEs in Kenya with a focus on employee wellness and working environment as predictors. From the research findings, conclusions were made as summarized in this section.

Relationship between Employee Wellness and Performance

The estimate coefficient and Wald test statistic confirmed that employee wellness has positive relationship with performance of SMs.  The study concludes that employee wellness is a significant determinant of performance in SMEs in Kenya. Employee Wellness is associated with enhanced productivity, reduced accidents, low cases of absenteeism and enhanced   motivation which in turn lead to improved work outcome and consequently enhanced organizational performance. However, compared to working environment employee wellness was found to be least significant as the determinant of organizational performance.

Welfare services such as provision of personal protective equipment, first aid facilities, medical insurance were rated as key in increasing the satisfaction and comfort of employees. Most organizations trained their staff as they associated training with improved performance. Lastly counselling and support services including were seen to play an insignificant role in most small and medium enterprises under study. This could have been attributed to the fact that many SMEs are not very keen on stress management in their organizations. Similarly, few SMEs provide gym facilities. It is presumed that employees can go for walks or exercise on their own to keep fit.   The conclusion is that while employee wellness leads to better performance generally, counseling and gym services have not taken root in many organizations. Therefore, the Null hypothesis that employee wellness has no significant relationship with performance of SMEs is rejected and the alternative hypothesis that employee wellness has significant relationship with performance of is accepted. However, this relationship was found not to be as robust as that of working environment.

Relationship between Working Environment and Performance

Study findings led to conclusion that working environment has the most significant effect on performance of SMEs. These findings imply that organizations with conducive environment where hazards are identified quite regularly, risk assessment conducted and control measures taken, reduces operational costs through reduced accidents and medical bills, lessens absenteeism and experience better performance. Majority of SMEs reported having safety procedures in place, conducting risks assessment regularly, reporting accidents and adhering to safety and health regulations at all times. Therefore, SMEs that strive for better working environment create comfortable working condition, which in turn enhances employee productivity and consequently enhances organizational performance. This conclusion is in concurrence with the model for the need for occupational safety and health in SMEs (adopted from Burton (2010). Thus the study rejects the null hypothesis that the working environment has no significant relationship with performance of SMEs and accepts the alternative hypothesis that working environment has significant relationship with performance of SMEs in Kenya. Investing in working environment improves the OSH in organizations leading to improved performance.

RECOMMENDATIONS OF THE STUDY

SMEs should stress on the importance of adherence to safety and health regulations and ensure total commitment to ensuring employees welfare. There is need to provide   employee’ wellness by putting into place measures that make employees feel that their physical and emotional wellness are taken care of. Provision of adequate OSH facilities and programs to ensure employees health needs are catered for is critical the organizations should ensure emotional and physical fitness of employees through special facilities and dedicate programs. Lastly, the organizations should ensure employees with special needs feel comfortable working and relating with colleagues within the organization. However, SMEs need to channel more effort in counselling and support services as most organizations invested east on them as compared to other wellness services,

Secondly, the SMEs should ensure the work environment is safe and conducive for the employees. Hazardous situations within the work environment must be identified and eliminated through a proper risk management strategy and procedures. Where risks occur, SMEs must have arrangements to mitigate and compensate those injured. In addition, organizations should ensure adequate safety gears are made available in the work environment to ensure optimum safety and comfort in the work environment.

Recommendations for Further Research

The study was narrowed to SMEs registered by the Kenya Association of Manufacturers (2022) and conclusions   drawn from the selected sample size of the population.  The study therefore   recommends a comparative analysis to be conducted between SMEs and large organizations There is need to explore a further research on employee wellness as the independent variable to determine the influence of counselling and support services and gym facilities among other wellness subcontracts to find out their level of investment and their impact on occupational health and safety of employees.

REFERENCES

1. Alli, B. O. (2019). Fundamental principles of occupational health and safety (4th ed.). International Labour Office.
2. Aminu, M. I., & Ezekiel, A. A. (2020). Occupational safety and health compliance in small and medium enterprises: Implications for sustainable business growth. Journal of Business and Safety Studies, 12(3), 45-60.
3. Arocena, P., & Núñez, I. (2019). Occupational health and safety management and firm performance. Safety Science, 113, 257-267. https://doi.org/10.1016/j.ssci.2018.12.002
4. Ayalew, A., & Demissie, Y. (2020). The effect of occupational health and safety program on organizational productivity: In case of Bahir Dar Tannery Factory. International Journal of Scientific and Research Publications (IJSRP), 10(2), 779-798. http://dx.doi.org/10.29322/IJSRP.10.02.2020.p98100 IJSRP
5. Bello, A. S., & Oyekunle, O. A. (2021). The impact of occupational safety measures on employee performance in SMEs. African Journal of Business and Management, 15(4), 78-95.
6. Chirico, F., Heponiemi, T., Pavlova, M., Zaffina, S., & Magnavita, N. (2019). Psychosocial risk prevention in a global occupational health perspective: A descriptive analysis. International Journal of Environmental Research and Public Health, 16(14), 2470. https://doi.org/10.3390/ijerph16142470 MDPI
7. Cooper, M. D. (2000). Towards a model of safety culture. Saf. Sci., 36, 111–136.
8. Gbadago, P., Amedome, S. N., & Honyenuga, B. Q. (2017). The impact of occupational health and safety measures on employee performance at the South Tongu District Hospital. Global Journal of Medical Research, 17(5), 13–19. https://globaljournals.org/GJMR_Volume17/3-The-Impact-of-Occupational-Health.pdf
9. Guerin, R., & Sleet, D. (2020). Using behavioral theory to enhance occupational safety and health: Applications to health care workers. American Journal of Lifestyle Medicine, 14(1), 67–75. https://doi.org/10.1177/1559827619861011.
10. Gunyomi, P. A., & Bruning, N. S. (2016). Human capital development and occupational health and safety: Impact on SME performance. Journal of Workplace Learning, 28(7), 430–450. https://doi.org/10.1108/JWL-11-2015-0082
11. Haktanir, G. (2019). Prediction of safety-related behaviour among Turkish nurses: An application of theory of planned behaviour and effects of safety climate perceptions. Social Sciences Journal, 8(1), 13–21. https://doi.org/10.11648/j.ss.20190801.13
12. Harshitha, R., & Senthil, K. (2021). Workplace hazards in SMEs: A comparative analysis with large enterprises. International Journal of Occupational Health, 27(2), 112–130. https://doi.org/10.1177/10773587211001234
13. International Labour Organization. (2010). ILO Declaration on fundamental principles and rights at work and its follow-up. International Labour Conference, 7–22.
14. International Labour Organization. (2017). Global trends in occupational safety and health: Survey report. Geneva, Switzerland: ILO Publications.
15. International Labour Organization. (2021). How can occupational safety and health be managed? Geneva, Switzerland: International Labour Organization.
16. Kersten, R., Harms, J., Liket, K., & Maas, K. (2017). Small firms, large impact? A systematic review of the SME finance literature. World Development, 97, 330-348. https://doi.org/10.1016/j.worlddev.2017.04.012
17. Kenya National Bureau of Statistics (KNBS). (2017). Economic survey 2018. Nairobi, Kenya: KNBS.
18. Kenya National Bureau of Statistics (KNBS). (2019). Micro, small, and medium enterprises (MSME) survey 2019. Nairobi, Kenya: KNBS.
19. Kirwan, M., Matthews, A., & Scott, P. (2013). The impact of the work environment of nurses on patient safety outcomes: A multi-level modelling approach. International Journal of Nursing
20. Lim, S., Cortina, L. M., & Magley, V. J. (2016). Personal and workgroup incivility: Impact on work and health outcomes. Journal of Applied Psychology, 93(1), 95–107. https://doi.org/10.1037/0021-9010.93.1.95
21. Maryjoan, U. I., & Ebitu, E. T. (2016). Effects of industrial safety and health on employees’ job performance in selected cement companies in Cross River State, Nigeria. International Journal of Business and Management Review, 4(3), 49–56. Retrieved from https://www.eajournals.org
22. Ministry of Labour, Kenya. (2007). Occupational Safety and Health Act, 2007. Kenya Gazette Supplement No. 111 (Act No. 15). Nairobi, Kenya: Government Press.
23. Mugambi, J. K., Muthama, N. J., & Mwiti, S. M. (2022). Occupational health and safety compliance among SMEs in Kenya: A policy perspective. Kenyan Journal of Business and Economics, 9(1), 22–35. https://doi.org/10.37284/kjbe.9.1.699
24. Murithi, G. (2021). An overview of workplace safety and health in Kenya. Proceedings of the 5th Kenya Institute for Public Policy Research and Analysis (KIPPRA) Annual Conference. Nairobi, Kenya: KIPPRA.
25. Murithi, J. (2021). The effect of workplace safety policies on SME productivity: Evidence from Kenya. African Journal of Economic and Social Studies, 14(2), 98–115.
26. Mutegi, T., Mugambi, P. J., & Kinyua, J. M. (2023). Workplace safety and employee productivity of manufacturing firms in Kenya. African Journal of Business Management, 17(4), 112–121. https://doi.org/10.5897/AJBM2023.9601
27. Mwangi, J. W., & Waiganjo, E. (2017). Influence of occupational health and safety on employee performance in the flower industry in Kenya: A case study of Penta Flowers Limited, Thika Sub-County. Strategic Journal of Business & Change Management, 4(3), 191–208. https://www.strategicjournals.com/index.php/journal/article/view/490
28. Oduor, P. (2021). Occupational health and safety practices on employee performance in beverage manufacturing firms in Nairobi County, Kenya. Social Sciences Journal, 5(1), 2–13.
29. Olabode, O., Adesanya, R., & Bakare, A. (2017). Ergonomics awareness and employee performance: An exploratory study. Economic and Environmental Studies, 17(2), 813–829. https://doi.org/10.25167/ees.2017.43.19
30. Oluoch, I. (2017). Assessment of the occupational safety and health management practices in water service industry in Kisumu County, Kenya (Doctoral dissertation). University of Nairobi.
31. Ombasyi, L. (2019). Effect of mental wellness on performance of employees within Brand Design Development Limited in Kenya (Doctoral dissertation). United States International University–Africa.
32. Ombasyi, P. (2019). The relationship between mental wellness and employee performance: A case study of SMEs in Kenya. East African Journal of Business and Innovation, 5(2), 58–74.
33. Osano, H., & Kesusu, J. (2016). Factors influencing health information seeking behaviour on Internet: Case of MBA students of Africa Nazarene University, Kenya. Global Journal of Health Sciences, 1(1), 55–69.
34. Sahu, A. K., Padhy, R. K., & Dhir, A. (2020). Envisioning the future of behavioral decision-making: A systematic literature review of behavioral reasoning theory. Australasian Marketing Journal, 28(4), 145–159. https://doi.org/10.1016/j.ausmj.2020.05.001
35. Sahu, A. K., Padhy, R. K., & Dhir, A. (2022). Determinants and barriers of implementing lean manufacturing practices in MSMEs: A behavioural reasoning theory perspective. Production Planning and Control, 33(12), 1197–1213. https://doi.org/10.1080/09537287.2020.1857449
36. Saunders, M., Lewis, P., & Thornhill, A. (2012). Research methods for business students (6th ed.). Pearson Education Limited.
37. Senthil, A., & Harshitha, L. (2021). Impact of employee well-being on organizational performance in workplace. Wesleyan Journal of Research, 14(30), 27–32.
38. Seyyed, S., & Zahra, J. (2012). Major theories of construction accident causation models. International Journal of Advances in Engineering & Technology, 4(2), 53–66.
39. Shan, B., Liu, X., Gu, A., & Zhao, R. (2022). The effect of occupational health risk perception on job satisfaction. International Journal of Environmental Research and Public Health, 19(4), 2111. https://doi.org/10.3390/ijerph19042111
40. Singh, P., Ghosh, D., & Rahman, M. (2019). The role of occupational health and safety in business performance: A meta-analysis. Journal of Business and Safety Science, 24(2), 156–172.
41. Siegel, A., Antony, J., Garza-Reyes, J., & Cherra, J. (2019). Integrated green lean approach and sustainability for SMEs: From literature review to a conceptual framework. Journal of Cleaner Production, 240, 118205. https://doi.org/10.1016/j.jclepro.2019.118205
42. Silaparasetti, R. (2021). The impact of occupational health and safety factors on workers’ behavior in construction projects in Oman. Safety Science, 136, 105156. https://doi.org/10.1016/j.ssci.2020.104947
43. Singh, R., Luthra, S., Mangla, S., & Uniyal, S. (2019). Applications of information and communication technology for sustainable growth of SMEs in India food industry. Resources, Conservation and Recycling, 147, 10–18. https://doi.org/10.1016/j.resconrec.2019.04.014
44. Sullivan, S. (2017). Wellness programs. Journal of Applied Medical Sciences, 6(1), 1–10.
45. Sundtoft, H. (2018). Social influence and safe behavior in manufacturing. Copenhagen Business School Journal, 109(1), 111–120.
46. Tandon, A., Dhir, A., Kaur, P., Kushwah, S., & Salo, J. (2020). Behavioral reasoning perspectives on organic food purchase. Appetite, 154, 104786. https://doi.org/10.1016/j.appet.2020.104786
47. Umugwaneza, C., Nkechi, I. E., & Mugabe, J. B. (2019). Effect of workplace safety and health practices on employee commitment and performance in steel manufacturing companies in Rwanda. European Journal of Business and Management Research, 4(5). https://doi.org/10.24018/ejbmr.2019.4.5.84
48. Waddell, G., & Burton, A. K. (2006). Is work good for your health and well-being? Department of Work and Pensions. https://www.gov.uk/government/publications/is-work-good-for-your-health-and-well-being
49. https://www.iiste.org/Journals/index.php/EJBM/article/view/42127
50. Wang, H. (2008). Safety factors and leading indicators in shipping organizations: Tanker and container operations (Doctoral dissertation). Rensselaer Polytechnic Institute.
51. Westaby, J. D. (2005). Behavioral reasoning theory: Identifying new linkages underlying intentions and behavior. Organizational Behavior and Human Decision Processes, 98(2), 97–120. https://doi.org/10.1016/j.obhdp.2005.07.003
52. Williams, R. A. (2020). Ordinal independent variables. In P.
53. Yamoah, E. E., & Nsowah, I. (2022). Performance in Warri refining and petrochemical company, Delta State, Nigeria. Nigerian Medical Journal, 61(3), 144–150. https://doi.org/10.4103/nmj.NMJ_146_19