What Determines Work Discipline and Performance? An Empirical Study in Indonesia

Abstract

The purpose of this research is to look into the effects of organizational culture and training and development on work discipline and performance. The data for this study was directly obtained from employees of a municipal water corporation in Medan, Indonesia, with a total of 204 participants. Partial Least Square Structural Equation Modeling (PLS-SEM) was applied for data analysis. The results showed that organizational culture and training and development positively and significantly affect performance. However, organizational culture and training & development positively affect employees' work discipline, albeit insignificantly. The findings of this study suggest that organizational culture and training and development play a critical role in shaping work discipline and performance in organizations in Indonesian settings. Therefore, the finding of this research engage all leaders in the organization to conduct training and development more intensively. Although it seems to have costly, this will have a good impact on the organization in the long run. Furthermore, the authors also suggest the creation of a solid organizational culture for every organization to foster excellent performance. However, each organization should choose its own acceptable organizational culture because it is possible that the organizational culture that works in one context does not work in another.

1. Introduction

Understanding the determinants of work discipline and performance is essential so that an organization can run smoothly and efficiently. Many studies have been devoted to examining what exactly determines both of them. One of the earliest studies was conducted by Waterman and Peters (1982), who found a significant link between particular culture and performance. In a recent study, Martinez et al. (2015) also had similar findings. Organizational culture is known to be not the only antecedent to performance but also training and development. This is because training & development have a fundamental role in achieving organizational goals by combining their interests and those of workers (Stone & Stone, 2013). Nowadays, training is one of the most important factors in the business world because it can increase the efficiency and effectiveness of employees and an organization. This can be important because training can enhance employees’ capabilities. Attention needs to be given to employee performance; as previous literature has found; organizational performance depends on employee performance. So, to improve it, training is given to the employees.

Regarding work discipline, there has been broad discussion about the influence of organizational culture on work discipline. Organizational culture can be considered to be a series of shared feelings, values, norms, expectations, and assumptions that manifest in behavior and connect people to the organization (Hofstede, 1993). Several studies have proven that organizational learning is an antecedent of work discipline (Franklin & Pagan, 2006). Aside from organizational culture, the variables that were also assumed to be antecedents were training and development. Unfortunately, however, how these variables influence discipline has not been studied, particularly using a sample comprising organizations in Southeast Asia.

Therefore, this study aims to capture a broader set related to work discipline and performance, especially among employees of a municipal water company in Indonesia. The majority of these research topics still concentrate on Western populations, and we are hoping we can gain a better understanding of Asian populations.

2. Literature Review

2.1. Training & Development

To improve performance, it is critical to provide training and development. As a result, Michael (2000) asserted that any organization’s training should be as well-designed as feasible. Although the expense of enhancing performance through training and development appears to be high at first, it will pay off in the long run. However, providing training and development cannot be done arbitrarily since it must take into account numerous characteristics of the people who will be trained and developed, such as status, title, and job description. When done correctly and prudently, the influence on training and development is not insignificant, because when people in an organization get training and development programs, the changes they experience tend to be permanent, resulting in improved job performance.

2.2. Organizational Culture

Culture can be interpreted as a system that is publicly and collectively accepted operating for a given group at a given time (White, 1949). However, if culture focuses on organizational context, then culture can be defined as a set of attributes that describe an organization and differentiate a firm from others. Organizational culture plays a vital role in an organization because culture guides how employees in particular organizations do certain things (Franklin & Pagan, 2006). In addition, there has long been an assumption that in organizations that have a “strong culture” with a well integrated and effective set of specific values, beliefs, and behavior patterns, employees who are in the organization will produce higher productivity than organizations that do not have a “strong culture”(Denison, 1984).

2.3. Work Discipline

When discussing disciplines in an organizational context, scholars provide different definitions. There is an emphasis on the pursuit of productivity, but there is also a focus on compliance with policies or regulations (Franklin & Pagan, 2006). An example is Belohlav (1985) who stated that creating and keeping employees productive is a function of the disciplinary process. On the other hand, Werther Jr and Davis (1981) define discipline as management actions to promote compliance with organizational standards. In creating discipline within an organization, policymakers generally take two approaches, the first is the progressive sanctions approach and the positive approach (Franklin & Pagan, 2006). Regardless of the approach used in an organization, supervisors generally work closely with managers and human resource departments. However, it is better if organizations use informal strategies to enforce work discipline, as suggested by Franklin and Pagan (2006), to examine cultural cues to know how to handle discipline problems.

2.4. Performance

Scholars that specialize in organizational research have multiple definitions of performance, just as they do in other fields. In general, academics define performance as a set of efficiency metrics that includes both transactional and input/ output efficiency (Stannack, 1996). According to Shayya (2018), job performance is defined as an individual’s ability to complete a task as well as their level of motivation. This is similar to Sarmiento et al. (2007), who stated that job success is determined by the abilities that individuals possess, whether they were gained naturally or via effort. Performance is about the motivation to apply these skills to work effectively, not just the hard skill. Because knowledge employees now account for more than two-thirds of the workforce, it is hard to compare their performance to that of labor workers (Shayya, 2018). That is why nowadays, there are no universally agreed-upon methods or tools to measure performance (Ramírez & Nembhard, 2004).

2.5. Hypotheses

Based on the literature review that has been collected, thus, the following hypotheses are proposed:

H1: Training & development has a positive and significant effect related to work discipline.

H2: Training & development has a positive and significant effect related to performance.

H3: Organizational culture has a positive and significant effect related to work discipline.

H4: Organizational culture has a positive and significant effect related to performance.

3. Research Methods and Materials

3.1. Measurements

Fifty items were generated to reflect the four constructs. The response format was a 5-point Likert-type scale ranging from strongly agree to strongly disagree. To improve the reliability and validity of the questionnaire, the authors used multiple-item measures for all of the study variables (Hair et al., 2016). In addition, to establish face validity to improve the format and content of the questionnaire items for clarity and readability, we initially distributed a pilot questionnaire to one of the branches that were not part of the study sample (Sekaran & Bougie, 2016). In the beginning, each construct has ten items to be measured. However, in the end, fourteen were used to measure all the constructs because the rest had inadequate factor loading and AVE (Figure 1).

Figure 1: Research Model

3.2. Population and Sample Size

The population in this study was all employees who worked at the municipal water company in the city of Medan, Indonesia. Several can be used as a benchmark in taking the number of samples for SEM-PLS statistical analysis. The sample size must be at least ten times larger than the number of indicators used to estimate a construct or ten times the structural model that leads to a construct. However, this basis was still considered not precise enough. Thus, the authors refer to Hair et al. (2016) who recommended that the sample size be adjusted according to power analysis. This is why, to determine the size of the sample that would be suitable for power analysis, the authors used the help of G * power analysis software. We used error measurements of type one and two at α = 0.05 and β = 0.95, while the effect size = 0.15, and the number of predictors as the model offered by the researcher is 4. The settings and results produced by the G * power application can be seen in Figure 2.

Figure 2: Power Results for Required Sample Size

Figure 2 explains that, at an error probability of 0.05 and a confidence level of 95%, the smallest sample required is 107 samples. This shows that the size of the sample in this study is more than sufficient because it has a population of 204 samples.

3.3. Data Collection

Data was collected using a questionnaire survey distributed directly to the office of the municipal water company in Medan, in both electronic and paper-based form. A total of 180 respondents’ answers (all samples) were collected.

3.4. Data Analysis

Partial Least Square Structural Equation Modeling (PLS-SEM) was employed for data analysis in this study. Although covariance-based structural equation modeling (CB-SEM) has dominated previous research as a method for analyzing complicated interrelationships between observed and latent variables, in recent years, the number of studies using PLS-SEM has increased rapidly by comparison (Hair et al., 2016). PLS-SEM has now been widely applied in many social science disciplines, including in the field of management (Hair et al., 2012, 2019). Besides, the PLS-SEM analysis method is also desirable to many researchers because it allows them to evaluate complicated models with many constructs, indicators, and structural paths without forcing distributional assumptions on the data (Hair et al., 2019).

Two main stages were carried out in analyzing the output results on Smart PLS v. 3.2.9, namely evaluation of measurement models and evaluation of the structural model (Hair et al., 2016). The rationales for both evaluations will be explained in the next session.

4. Results

4.1. Evaluation of Measurement Models

The initial stage was testing the measurement model. Measurement model evaluation measures the reliability and validity of the constructs along with their corresponding items. Three aspects determine whether a measurement model is accepted or not, namely convergent validity, internal consistency reliability, and discriminant validity. Convergent validity is the degree to which a measure correlates positively with alternative measures of the same construct (Hair et al., 2019). The convergent validity required loading factors to exceed 0.5 and Average Variance Extracted (AVE) to exceed 0.5. Meanwhile, internal consistency reliability is a form of reliability used to judge the consistency of results across items in the same test and determine whether the items measuring a construct are similar in their scores (Hair et al., 2016). This requires composite reliability > 0.6, as well as the Cronbach’s Alpha (Hair et al., 2019). The last aspect is discriminant validity which is the degree to which a construct is genuinely distinct from other constructs according to empirical standards (Hair et al., 2019). The cross-loadings and Fornell-Larcker criterion are typically used to assess discriminant validity. Nevertheless, current research that has critically studied the performance of cross-loadings and the Fornell-Larcker criterion for discriminant validity has found that neither method reliably recognizes discriminant validity issues (Henseler et al., 2015). As a remedy, Henseler et al. (2015) have suggested using the Heterotrait-monotrait ratio (HTMT). For the threshold level, the Heterotrait-Monotrait ratio (HTMT) confidence interval must not include 1, while a lower and consequently more conservative threshold value of 0.85 seems justified (Henseler et al., 2015).

In the Smart PLS analysis, the authors used a bootstrapping of 5, 000 sub-samples as recommended by Hair et al. (2019). In the first analysis, the measurement model did not meet the requirements because it had a low AVE value, so there were several indicators with low loading factors that were removed. After the new model was formed, we ran the PLS algorithm for the second time. As we can see in Table 1, the results demonstrated that all constructs present adequate convergent validity, with loadings and AVE exceeding 0.5. Internal consistency reliability also exceeded the threshold, with composite reliability and Cronbach’s alpha exceeding 0.6. With regard to discriminant validity (Table 2), HTMT was applied, and the measurement results showed that there is no single construct that includes 0.85 in HTMT.

Table 1: Results Summary For Convergent Validity and Internal Consistency Reliability

4.2. Evaluation of the Structural Model

After the construct measures are proved to be reliable and valid, the next step is to assess the structural model results. According to Hair et al. (2016), when examining the structural model, it is important to realize that PLS-SEM is different from CB-SEM, which estimates parameters so that the differences between the sample covariances and those predicted by the theoretical/conceptual model are minimized. The goodness-of-fit measures, such as the chi square statistic or the several fit indices compared with CB-SEM, are not fully transferrable to PLS-SEM. Instead, the fundamental criteria for evaluating the structural model in PLS-SEM are the path coefficients, R² values, f² effect size, and SRMR.

Structural model evaluation is conducted to examine the path between constructs based on the offered hypothesis. As recommended by Hair et al. (2016), the authors used bootstrapping with 5, 000 subsamples, two-tailed, and a 0.05 significant level to produce the standard error and t-statistics for the sample. As shown in Table 3 and Figure 3, the structural model evaluation results showed that the two main paths are significant. Table 3 also shows that the path relationship between organizational culture and performance is β = 0.318, p = 0.001. This indicates that organizational culture has a positive significant effect on performance. On the other hand, organizational culture shows that there is a positive but no significant effect on work performance, β = 0.169, p = 0.307. Regarding training and development, this variable has positive and significant effect, with β = -0.183, p = 0.037. Meanwhile, training & development have a positive but not significant effect on work discipline, β = 0.168, p = 0.280.

Table 3: Results Summary for Structural Model Evaluation

Note: *p-value < 0.1; **p-value < 0.05; ***p-value < 0.001. Significant at the 0.05 level.

Figure 3: Structural Model with Loading Factor, Path Coefficients, and r Square

Next, the most commonly used standard in assessing the structural model is the coefficient of determination (R² value). The coefficient depicts the amount of variance in the endogenous constructs described by all of the exogenous constructs associated with it (Hair et al., 2016). The value varies from 0 to 1. While it is challenging to present rules of thumb for adequate R², 0.20 is nevertheless considered adequate (Hair et al., 2016). As we can see from Table 3, the R² coefficient is 0.170 for work performance and 0.074 for work discipline, so it means the R² is adequate.

Furthermore, the magnitude of the effect size of the predictor constructs was evaluated using ƒ² effect size. Guidelines for evaluating ƒ² are that values of 0.02, 0.15, and 0.35, sequentially express small, medium, and large effects (Cohen, 2013). As a consequence, from Table 4 we can conclude that all variables have a small effect. SRMR was also assessed to know the root mean square difference between the observed correlations and the model-implied correlations (Hair et al., 2016).

Table 4: Results of Latent Variable Correlations

5. Discussion

This research investigated the influence of training & development, including organizational culture, on work discipline and performance. We used SEM-PLS to analyze the data. The results obtained confirm the reliability and validity of the measurement model (see Table 1 and Table 2).

From the structural model evaluation, it was first observed that the R² coefficient is 0.170 for the work performance construct, and 0.074 for the work discipline construct, which is adequate. Regarding the hypothesis testing, the empirical results for the samples showed that organizational culture has a positive and significant effect on performance. The findings are consistent with previous research that found Table 4: Results of Latent Variable Correlations links between those two variables in senior management team culture in British acute hospitals (Jacobs et al., 2013). Furthermore, according to Shayya (2018), organizational culture does have a role in determining performance levels in different units in public education. The findings of Sopiah et al. (2021) suggest that organizational culture can explain employee performance in Islamic banks in Indonesia when compared to research conducted in similar sample locations. Previous research findings in Indonesia show that organizational culture has a major impact on performance (Munawaroh et al., 2021). The importance of organizational culture in organizational growth in Indonesia is shown by the validation of identical results in the same scenario. Not only that, Diana et al. (2021), who reported identical results, also corroborate this conclusion. These findings support the idea that organizational culture has an impact on employee performance in Indonesian demographics.

Furthermore, corporate culture was discovered to have a good impact on employee work discipline. In contrast to the impact of organizational culture on performance, this variable has a minor but beneficial impact on work discipline. Crow and Hartman (2002) found that healthcare companies that overlook the detrimental features of their culture risk not just having bad employee relations but also being unable to properly implement discipline. The findings are consistent with Franklin and Pagan (2006), who found empirical evidence that informal tactics are popular when supervisors confront discipline challenges, and that cultural factors can explain why disciplinary strategies are varied.

In addition, the PLS findings revealed that training and development have a significant positive impact on performance. The findings of this study are in line with earlier research that has found that strong training and development improves organizational performance (Niazi, 2011). This is in line with prior research, which found that training and development were positively connected with employee performance and effectiveness and maintained a statistically significant association (Asfaw et al., 2015). Furthermore, the findings revealed that training and development had a positive (although minor) impact on work discipline. Based on the findings of Bharata (2016), this study finds that compensation and work discipline training have a significant impact on employee performance.

6. Conclusion

The capability of an organization’s resources has been shown to have a significant effect on its success. The cause of this is a combination of factors, including work discipline and performance. Training and development, as well as company culture, have a good impact on work discipline, according to this study. These two variables, on the other hand, have a positive and significant impact on employee performance. That is why organizations and businesses must pay close attention to training and development, as well as the organizational culture in place.

The authors urge that senior management or organizational leaders continue to perform employee development through training based on the findings of this study. Although this method appears to be costly, it will return to their respective organizations in the long run by improving work discipline and performance provided by employees to their organizations. Furthermore, previous literature suggests that the change in the skills they gain is likely to be permanent. In addition, the author advises that every firm establish a strong organizational culture. However, it is also important to consider what type of organizational culture is acceptable for each organization, as the organizational culture in one organization may not necessarily match the organizational culture in another.

The data collected is limited to a sample of employees who work at the municipal water company in Medan, Indonesia, hence this study has limitations as well. To corroborate the findings of this study, more research with larger sample size is required. Future researchers should think about doing longitudinal research to back up their conclusions.

*Acknowledgements:

[1] The authors thank the respondents for their willingness to fill out the questionnaire.

[2] Ethical approval was obtained from the Institute for Research and Community Service of Universitas Pembangunan Panca Budi. Completion of the questionnaire was also taken to indicate the respondents consented to take part in the study.